diff --git a/Documentation/devicetree/bindings/display/multi-inno,mi0283qt.txt b/Documentation/devicetree/bindings/display/multi-inno,mi0283qt.txt new file mode 100644 index 000000000000..eed48c3d4875 --- /dev/null +++ b/Documentation/devicetree/bindings/display/multi-inno,mi0283qt.txt @@ -0,0 +1,27 @@ +Multi-Inno MI0283QT display panel + +Required properties: +- compatible: "multi-inno,mi0283qt". + +The node for this driver must be a child node of a SPI controller, hence +all mandatory properties described in ../spi/spi-bus.txt must be specified. + +Optional properties: +- dc-gpios: D/C pin. The presence/absence of this GPIO determines + the panel interface mode (IM[3:0] pins): + - present: IM=x110 4-wire 8-bit data serial interface + - absent: IM=x101 3-wire 9-bit data serial interface +- reset-gpios: Reset pin +- power-supply: A regulator node for the supply voltage. +- backlight: phandle of the backlight device attached to the panel +- rotation: panel rotation in degrees counter clockwise (0,90,180,270) + +Example: + mi0283qt@0{ + compatible = "multi-inno,mi0283qt"; + reg = <0>; + spi-max-frequency = <32000000>; + rotation = <90>; + dc-gpios = <&gpio 25 0>; + backlight = <&backlight>; + }; diff --git a/Documentation/devicetree/bindings/display/panel/boe,nv101wxmn51.txt b/Documentation/devicetree/bindings/display/panel/boe,nv101wxmn51.txt new file mode 100644 index 000000000000..b258d6a91ec6 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/boe,nv101wxmn51.txt @@ -0,0 +1,7 @@ +BOE OPTOELECTRONICS TECHNOLOGY 10.1" WXGA TFT LCD panel + +Required properties: +- compatible: should be "boe,nv101wxmn51" + +This binding is compatible with the simple-panel binding, which is specified +in simple-panel.txt in this directory. diff --git a/Documentation/devicetree/bindings/display/panel/netron-dy,e231732.txt b/Documentation/devicetree/bindings/display/panel/netron-dy,e231732.txt new file mode 100644 index 000000000000..c6d06b5eab51 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/netron-dy,e231732.txt @@ -0,0 +1,7 @@ +Netron-DY E231732 7.0" WSVGA TFT LCD panel + +Required properties: +- compatible: should be "netron-dy,e231732" + +This binding is compatible with the simple-panel binding, which is specified +in simple-panel.txt in this directory. diff --git a/Documentation/devicetree/bindings/display/panel/panel.txt b/Documentation/devicetree/bindings/display/panel/panel.txt new file mode 100644 index 000000000000..e2e6867852b8 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/panel.txt @@ -0,0 +1,4 @@ +Common display properties +------------------------- + +- rotation: Display rotation in degrees counter clockwise (0,90,180,270) diff --git a/Documentation/devicetree/bindings/display/panel/tianma,tm070jdhg30.txt b/Documentation/devicetree/bindings/display/panel/tianma,tm070jdhg30.txt new file mode 100644 index 000000000000..eb9501a82e25 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/tianma,tm070jdhg30.txt @@ -0,0 +1,7 @@ +Tianma Micro-electronics TM070JDHG30 7.0" WXGA TFT LCD panel + +Required properties: +- compatible: should be "tianma,tm070jdhg30" + +This binding is compatible with the simple-panel binding, which is specified +in simple-panel.txt in this directory. diff --git a/Documentation/devicetree/bindings/interrupt-controller/snps,archs-idu-intc.txt b/Documentation/devicetree/bindings/interrupt-controller/snps,archs-idu-intc.txt index 0dcb7c7d3e40..944657684d73 100644 --- a/Documentation/devicetree/bindings/interrupt-controller/snps,archs-idu-intc.txt +++ b/Documentation/devicetree/bindings/interrupt-controller/snps,archs-idu-intc.txt @@ -15,6 +15,9 @@ Properties: Second cell specifies the irq distribution mode to cores 0=Round Robin; 1=cpu0, 2=cpu1, 4=cpu2, 8=cpu3 + The second cell in interrupts property is deprecated and may be ignored by + the kernel. + intc accessed via the special ARC AUX register interface, hence "reg" property is not specified. diff --git a/Documentation/devicetree/bindings/net/mediatek-net.txt b/Documentation/devicetree/bindings/net/mediatek-net.txt index c010fafc66a8..c7194e87d5f4 100644 --- a/Documentation/devicetree/bindings/net/mediatek-net.txt +++ b/Documentation/devicetree/bindings/net/mediatek-net.txt @@ -7,7 +7,7 @@ have dual GMAC each represented by a child node.. * Ethernet controller node Required properties: -- compatible: Should be "mediatek,mt7623-eth" +- compatible: Should be "mediatek,mt2701-eth" - reg: Address and length of the register set for the device - interrupts: Should contain the three frame engines interrupts in numeric order. These are fe_int0, fe_int1 and fe_int2. diff --git a/Documentation/devicetree/bindings/net/phy.txt b/Documentation/devicetree/bindings/net/phy.txt index ff1bc4b1bb3b..fb5056b22685 100644 --- a/Documentation/devicetree/bindings/net/phy.txt +++ b/Documentation/devicetree/bindings/net/phy.txt @@ -19,8 +19,9 @@ Optional Properties: specifications. If neither of these are specified, the default is to assume clause 22. - If the phy's identifier is known then the list may contain an entry - of the form: "ethernet-phy-idAAAA.BBBB" where + If the PHY reports an incorrect ID (or none at all) then the + "compatible" list may contain an entry with the correct PHY ID in the + form: "ethernet-phy-idAAAA.BBBB" where AAAA - The value of the 16 bit Phy Identifier 1 register as 4 hex digits. This is the chip vendor OUI bits 3:18 BBBB - The value of the 16 bit Phy Identifier 2 register as diff --git a/Documentation/devicetree/bindings/vendor-prefixes.txt b/Documentation/devicetree/bindings/vendor-prefixes.txt index 16d3b5e7f5d1..f98498df2365 100644 --- a/Documentation/devicetree/bindings/vendor-prefixes.txt +++ b/Documentation/devicetree/bindings/vendor-prefixes.txt @@ -187,6 +187,7 @@ mpl MPL AG mqmaker mqmaker Inc. msi Micro-Star International Co. Ltd. mti Imagination Technologies Ltd. (formerly MIPS Technologies Inc.) +multi-inno Multi-Inno Technology Co.,Ltd mundoreader Mundo Reader S.L. murata Murata Manufacturing Co., Ltd. mxicy Macronix International Co., Ltd. @@ -196,6 +197,7 @@ nec NEC LCD Technologies, Ltd. neonode Neonode Inc. netgear NETGEAR netlogic Broadcom Corporation (formerly NetLogic Microsystems) +netron-dy Netron DY netxeon Shenzhen Netxeon Technology CO., LTD nexbox Nexbox newhaven Newhaven Display International @@ -296,6 +298,7 @@ technologic Technologic Systems terasic Terasic Inc. thine THine Electronics, Inc. ti Texas Instruments +tianma Tianma Micro-electronics Co., Ltd. tlm Trusted Logic Mobility topeet Topeet toradex Toradex AG diff --git a/Documentation/gpu/index.rst b/Documentation/gpu/index.rst index 367d7c36b8e9..f81278a7c2cc 100644 --- a/Documentation/gpu/index.rst +++ b/Documentation/gpu/index.rst @@ -11,6 +11,7 @@ Linux GPU Driver Developer's Guide drm-kms-helpers drm-uapi i915 + tinydrm vga-switcheroo vgaarbiter diff --git a/Documentation/gpu/tinydrm.rst b/Documentation/gpu/tinydrm.rst new file mode 100644 index 000000000000..a913644bfc19 --- /dev/null +++ b/Documentation/gpu/tinydrm.rst @@ -0,0 +1,42 @@ +========================== +drm/tinydrm Driver library +========================== + +.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-core.c + :doc: overview + +Core functionality +================== + +.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-core.c + :doc: core + +.. kernel-doc:: include/drm/tinydrm/tinydrm.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-core.c + :export: + +.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-pipe.c + :export: + +Additional helpers +================== + +.. kernel-doc:: include/drm/tinydrm/tinydrm-helpers.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-helpers.c + :export: + +MIPI DBI Compatible Controllers +=============================== + +.. kernel-doc:: drivers/gpu/drm/tinydrm/mipi-dbi.c + :doc: overview + +.. kernel-doc:: include/drm/tinydrm/mipi-dbi.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/tinydrm/mipi-dbi.c + :export: diff --git a/Documentation/media/uapi/cec/cec-func-close.rst b/Documentation/media/uapi/cec/cec-func-close.rst index 8267c31b317d..895d9c2d1c04 100644 --- a/Documentation/media/uapi/cec/cec-func-close.rst +++ b/Documentation/media/uapi/cec/cec-func-close.rst @@ -33,11 +33,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - Closes the cec device. Resources associated with the file descriptor are freed. The device configuration remain unchanged. diff --git a/Documentation/media/uapi/cec/cec-func-ioctl.rst b/Documentation/media/uapi/cec/cec-func-ioctl.rst index 9e8dbb118d6a..7dcfd178fb24 100644 --- a/Documentation/media/uapi/cec/cec-func-ioctl.rst +++ b/Documentation/media/uapi/cec/cec-func-ioctl.rst @@ -39,11 +39,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - The :c:func:`ioctl()` function manipulates cec device parameters. The argument ``fd`` must be an open file descriptor. diff --git a/Documentation/media/uapi/cec/cec-func-open.rst b/Documentation/media/uapi/cec/cec-func-open.rst index af3f5b5c24c6..0304388cd159 100644 --- a/Documentation/media/uapi/cec/cec-func-open.rst +++ b/Documentation/media/uapi/cec/cec-func-open.rst @@ -46,11 +46,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - To open a cec device applications call :c:func:`open()` with the desired device name. The function has no side effects; the device configuration remain unchanged. diff --git a/Documentation/media/uapi/cec/cec-func-poll.rst b/Documentation/media/uapi/cec/cec-func-poll.rst index cfb73e6027a5..6a863cfda6e0 100644 --- a/Documentation/media/uapi/cec/cec-func-poll.rst +++ b/Documentation/media/uapi/cec/cec-func-poll.rst @@ -39,11 +39,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - With the :c:func:`poll()` function applications can wait for CEC events. diff --git a/Documentation/media/uapi/cec/cec-intro.rst b/Documentation/media/uapi/cec/cec-intro.rst index 4a19ea5323a9..07ee2b8f89d6 100644 --- a/Documentation/media/uapi/cec/cec-intro.rst +++ b/Documentation/media/uapi/cec/cec-intro.rst @@ -3,11 +3,6 @@ Introduction ============ -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - HDMI connectors provide a single pin for use by the Consumer Electronics Control protocol. This protocol allows different devices connected by an HDMI cable to communicate. The protocol for CEC version 1.4 is defined @@ -31,3 +26,15 @@ control just the CEC pin. Drivers that support CEC will create a CEC device node (/dev/cecX) to give userspace access to the CEC adapter. The :ref:`CEC_ADAP_G_CAPS` ioctl will tell userspace what it is allowed to do. + +In order to check the support and test it, it is suggested to download +the `v4l-utils `_ package. It +provides three tools to handle CEC: + +- cec-ctl: the Swiss army knife of CEC. Allows you to configure, transmit + and monitor CEC messages. + +- cec-compliance: does a CEC compliance test of a remote CEC device to + determine how compliant the CEC implementation is. + +- cec-follower: emulates a CEC follower. diff --git a/Documentation/media/uapi/cec/cec-ioc-adap-g-caps.rst b/Documentation/media/uapi/cec/cec-ioc-adap-g-caps.rst index 2b0ddb14b280..a0e961f11017 100644 --- a/Documentation/media/uapi/cec/cec-ioc-adap-g-caps.rst +++ b/Documentation/media/uapi/cec/cec-ioc-adap-g-caps.rst @@ -29,11 +29,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - All cec devices must support :ref:`ioctl CEC_ADAP_G_CAPS `. To query device information, applications call the ioctl with a pointer to a struct :c:type:`cec_caps`. The driver fills the structure and diff --git a/Documentation/media/uapi/cec/cec-ioc-adap-g-log-addrs.rst b/Documentation/media/uapi/cec/cec-ioc-adap-g-log-addrs.rst index b878637e91b3..09f09bbe28d4 100644 --- a/Documentation/media/uapi/cec/cec-ioc-adap-g-log-addrs.rst +++ b/Documentation/media/uapi/cec/cec-ioc-adap-g-log-addrs.rst @@ -35,11 +35,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - To query the current CEC logical addresses, applications call :ref:`ioctl CEC_ADAP_G_LOG_ADDRS ` with a pointer to a struct :c:type:`cec_log_addrs` where the driver stores the logical addresses. diff --git a/Documentation/media/uapi/cec/cec-ioc-adap-g-phys-addr.rst b/Documentation/media/uapi/cec/cec-ioc-adap-g-phys-addr.rst index 3357deb43c85..a3cdc75cec3e 100644 --- a/Documentation/media/uapi/cec/cec-ioc-adap-g-phys-addr.rst +++ b/Documentation/media/uapi/cec/cec-ioc-adap-g-phys-addr.rst @@ -35,11 +35,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - To query the current physical address applications call :ref:`ioctl CEC_ADAP_G_PHYS_ADDR ` with a pointer to a __u16 where the driver stores the physical address. diff --git a/Documentation/media/uapi/cec/cec-ioc-dqevent.rst b/Documentation/media/uapi/cec/cec-ioc-dqevent.rst index e256c6605de7..6e589a1fae17 100644 --- a/Documentation/media/uapi/cec/cec-ioc-dqevent.rst +++ b/Documentation/media/uapi/cec/cec-ioc-dqevent.rst @@ -30,11 +30,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - CEC devices can send asynchronous events. These can be retrieved by calling :c:func:`CEC_DQEVENT`. If the file descriptor is in non-blocking mode and no event is pending, then it will return -1 and diff --git a/Documentation/media/uapi/cec/cec-ioc-g-mode.rst b/Documentation/media/uapi/cec/cec-ioc-g-mode.rst index 4f5818b9d277..e4ded9df0a84 100644 --- a/Documentation/media/uapi/cec/cec-ioc-g-mode.rst +++ b/Documentation/media/uapi/cec/cec-ioc-g-mode.rst @@ -31,11 +31,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - By default any filehandle can use :ref:`CEC_TRANSMIT`, but in order to prevent applications from stepping on each others toes it must be possible to obtain exclusive access to the CEC adapter. This ioctl sets the diff --git a/Documentation/media/uapi/cec/cec-ioc-receive.rst b/Documentation/media/uapi/cec/cec-ioc-receive.rst index bdf015b1d1dc..dc2adb391c0a 100644 --- a/Documentation/media/uapi/cec/cec-ioc-receive.rst +++ b/Documentation/media/uapi/cec/cec-ioc-receive.rst @@ -34,11 +34,6 @@ Arguments Description =========== -.. note:: - - This documents the proposed CEC API. This API is not yet finalized - and is currently only available as a staging kernel module. - To receive a CEC message the application has to fill in the ``timeout`` field of struct :c:type:`cec_msg` and pass it to :ref:`ioctl CEC_RECEIVE `. diff --git a/Documentation/power/states.txt b/Documentation/power/states.txt index 8a39ce45d8a0..008ecb588317 100644 --- a/Documentation/power/states.txt +++ b/Documentation/power/states.txt @@ -35,9 +35,7 @@ only one way to cause the system to go into the Suspend-To-RAM state (write The default suspend mode (ie. the one to be used without writing anything into /sys/power/mem_sleep) is either "deep" (if Suspend-To-RAM is supported) or "s2idle", but it can be overridden by the value of the "mem_sleep_default" -parameter in the kernel command line. On some ACPI-based systems, depending on -the information in the FADT, the default may be "s2idle" even if Suspend-To-RAM -is supported. +parameter in the kernel command line. The properties of all of the sleep states are described below. diff --git a/MAINTAINERS b/MAINTAINERS index adfc60d9c456..706ba94dcae0 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1091,7 +1091,7 @@ F: arch/arm/boot/dts/aspeed-* F: drivers/*/*aspeed* ARM/ATMEL AT91RM9200, AT91SAM9 AND SAMA5 SOC SUPPORT -M: Nicolas Ferre +M: Nicolas Ferre M: Alexandre Belloni M: Jean-Christophe Plagniol-Villard L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) @@ -1773,7 +1773,7 @@ F: drivers/soc/renesas/ F: include/linux/soc/renesas/ ARM/SOCFPGA ARCHITECTURE -M: Dinh Nguyen +M: Dinh Nguyen S: Maintained F: arch/arm/mach-socfpga/ F: arch/arm/boot/dts/socfpga* @@ -1783,7 +1783,7 @@ W: http://www.rocketboards.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux.git ARM/SOCFPGA CLOCK FRAMEWORK SUPPORT -M: Dinh Nguyen +M: Dinh Nguyen S: Maintained F: drivers/clk/socfpga/ @@ -2175,56 +2175,56 @@ F: include/linux/atm* F: include/uapi/linux/atm* ATMEL AT91 / AT32 MCI DRIVER -M: Ludovic Desroches +M: Ludovic Desroches S: Maintained F: drivers/mmc/host/atmel-mci.c ATMEL AT91 SAMA5D2-Compatible Shutdown Controller -M: Nicolas Ferre +M: Nicolas Ferre S: Supported F: drivers/power/reset/at91-sama5d2_shdwc.c ATMEL SAMA5D2 ADC DRIVER -M: Ludovic Desroches +M: Ludovic Desroches L: linux-iio@vger.kernel.org S: Supported F: drivers/iio/adc/at91-sama5d2_adc.c ATMEL Audio ALSA driver -M: Nicolas Ferre +M: Nicolas Ferre L: alsa-devel@alsa-project.org (moderated for non-subscribers) S: Supported F: sound/soc/atmel ATMEL XDMA DRIVER -M: Ludovic Desroches +M: Ludovic Desroches L: linux-arm-kernel@lists.infradead.org L: dmaengine@vger.kernel.org S: Supported F: drivers/dma/at_xdmac.c ATMEL I2C DRIVER -M: Ludovic Desroches +M: Ludovic Desroches L: linux-i2c@vger.kernel.org S: Supported F: drivers/i2c/busses/i2c-at91.c ATMEL ISI DRIVER -M: Ludovic Desroches +M: Ludovic Desroches L: linux-media@vger.kernel.org S: Supported F: drivers/media/platform/soc_camera/atmel-isi.c F: include/media/atmel-isi.h ATMEL LCDFB DRIVER -M: Nicolas Ferre +M: Nicolas Ferre L: linux-fbdev@vger.kernel.org S: Maintained F: drivers/video/fbdev/atmel_lcdfb.c F: include/video/atmel_lcdc.h ATMEL MACB ETHERNET DRIVER -M: Nicolas Ferre +M: Nicolas Ferre S: Supported F: drivers/net/ethernet/cadence/ @@ -2236,32 +2236,32 @@ S: Supported F: drivers/mtd/nand/atmel_nand* ATMEL SDMMC DRIVER -M: Ludovic Desroches +M: Ludovic Desroches L: linux-mmc@vger.kernel.org S: Supported F: drivers/mmc/host/sdhci-of-at91.c ATMEL SPI DRIVER -M: Nicolas Ferre +M: Nicolas Ferre S: Supported F: drivers/spi/spi-atmel.* ATMEL SSC DRIVER -M: Nicolas Ferre +M: Nicolas Ferre L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Supported F: drivers/misc/atmel-ssc.c F: include/linux/atmel-ssc.h ATMEL Timer Counter (TC) AND CLOCKSOURCE DRIVERS -M: Nicolas Ferre +M: Nicolas Ferre L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Supported F: drivers/misc/atmel_tclib.c F: drivers/clocksource/tcb_clksrc.c ATMEL USBA UDC DRIVER -M: Nicolas Ferre +M: Nicolas Ferre L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Supported F: drivers/usb/gadget/udc/atmel_usba_udc.* @@ -3567,7 +3567,7 @@ F: drivers/infiniband/hw/cxgb3/ F: include/uapi/rdma/cxgb3-abi.h CXGB4 ETHERNET DRIVER (CXGB4) -M: Hariprasad S +M: Ganesh Goudar L: netdev@vger.kernel.org W: http://www.chelsio.com S: Supported @@ -4154,7 +4154,7 @@ F: Documentation/gpu/i915.rst INTEL GVT-g DRIVERS (Intel GPU Virtualization) M: Zhenyu Wang M: Zhi Wang -L: igvt-g-dev@lists.01.org +L: intel-gvt-dev@lists.freedesktop.org L: intel-gfx@lists.freedesktop.org W: https://01.org/igvt-g T: git https://github.com/01org/gvt-linux.git @@ -4245,6 +4245,12 @@ S: Supported F: drivers/gpu/drm/mediatek/ F: Documentation/devicetree/bindings/display/mediatek/ +DRM DRIVER FOR MI0283QT +M: Noralf Trønnes +S: Maintained +F: drivers/gpu/drm/tinydrm/mi0283qt.c +F: Documentation/devicetree/bindings/display/multi-inno,mi0283qt.txt + DRM DRIVER FOR MSM ADRENO GPU M: Rob Clark L: linux-arm-msm@vger.kernel.org @@ -9737,7 +9743,7 @@ S: Maintained F: drivers/pinctrl/pinctrl-at91.* PIN CONTROLLER - ATMEL AT91 PIO4 -M: Ludovic Desroches +M: Ludovic Desroches L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) L: linux-gpio@vger.kernel.org S: Supported @@ -10196,7 +10202,6 @@ F: drivers/media/tuners/qt1010* QUALCOMM ATHEROS ATH9K WIRELESS DRIVER M: QCA ath9k Development L: linux-wireless@vger.kernel.org -L: ath9k-devel@lists.ath9k.org W: http://wireless.kernel.org/en/users/Drivers/ath9k S: Supported F: drivers/net/wireless/ath/ath9k/ @@ -13067,7 +13072,7 @@ F: drivers/input/serio/userio.c F: include/uapi/linux/userio.h VIRTIO CONSOLE DRIVER -M: Amit Shah +M: Amit Shah L: virtualization@lists.linux-foundation.org S: Maintained F: drivers/char/virtio_console.c diff --git a/Makefile b/Makefile index 098840012b9b..503dae1de8ef 100644 --- a/Makefile +++ b/Makefile @@ -1,8 +1,8 @@ VERSION = 4 PATCHLEVEL = 10 SUBLEVEL = 0 -EXTRAVERSION = -rc5 -NAME = Anniversary Edition +EXTRAVERSION = -rc8 +NAME = Fearless Coyote # *DOCUMENTATION* # To see a list of typical targets execute "make help" @@ -797,7 +797,7 @@ KBUILD_CFLAGS += $(call cc-option,-Werror=incompatible-pointer-types) KBUILD_ARFLAGS := $(call ar-option,D) # check for 'asm goto' -ifeq ($(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-goto.sh $(CC)), y) +ifeq ($(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-goto.sh $(CC) $(KBUILD_CFLAGS)), y) KBUILD_CFLAGS += -DCC_HAVE_ASM_GOTO KBUILD_AFLAGS += -DCC_HAVE_ASM_GOTO endif diff --git a/arch/arc/include/asm/delay.h b/arch/arc/include/asm/delay.h index a36e8601114d..d5da2115d78a 100644 --- a/arch/arc/include/asm/delay.h +++ b/arch/arc/include/asm/delay.h @@ -26,7 +26,9 @@ static inline void __delay(unsigned long loops) " lp 1f \n" " nop \n" "1: \n" - : : "r"(loops)); + : + : "r"(loops) + : "lp_count"); } extern void __bad_udelay(void); diff --git a/arch/arc/kernel/head.S b/arch/arc/kernel/head.S index 689dd867fdff..8b90d25a15cc 100644 --- a/arch/arc/kernel/head.S +++ b/arch/arc/kernel/head.S @@ -71,14 +71,14 @@ ENTRY(stext) GET_CPU_ID r5 cmp r5, 0 mov.nz r0, r5 -#ifdef CONFIG_ARC_SMP_HALT_ON_RESET - ; Non-Master can proceed as system would be booted sufficiently - jnz first_lines_of_secondary -#else + bz .Lmaster_proceed + ; Non-Masters wait for Master to boot enough and bring them up - jnz arc_platform_smp_wait_to_boot -#endif - ; Master falls thru + ; when they resume, tail-call to entry point + mov blink, @first_lines_of_secondary + j arc_platform_smp_wait_to_boot + +.Lmaster_proceed: #endif ; Clear BSS before updating any globals diff --git a/arch/arc/kernel/mcip.c b/arch/arc/kernel/mcip.c index 9274f8ade8c7..9f6b68fd4f3b 100644 --- a/arch/arc/kernel/mcip.c +++ b/arch/arc/kernel/mcip.c @@ -93,11 +93,10 @@ static void mcip_probe_n_setup(void) READ_BCR(ARC_REG_MCIP_BCR, mp); sprintf(smp_cpuinfo_buf, - "Extn [SMP]\t: ARConnect (v%d): %d cores with %s%s%s%s%s\n", + "Extn [SMP]\t: ARConnect (v%d): %d cores with %s%s%s%s\n", mp.ver, mp.num_cores, IS_AVAIL1(mp.ipi, "IPI "), IS_AVAIL1(mp.idu, "IDU "), - IS_AVAIL1(mp.llm, "LLM "), IS_AVAIL1(mp.dbg, "DEBUG "), IS_AVAIL1(mp.gfrc, "GFRC")); @@ -175,7 +174,6 @@ static void idu_irq_unmask(struct irq_data *data) raw_spin_unlock_irqrestore(&mcip_lock, flags); } -#ifdef CONFIG_SMP static int idu_irq_set_affinity(struct irq_data *data, const struct cpumask *cpumask, bool force) @@ -205,12 +203,27 @@ idu_irq_set_affinity(struct irq_data *data, const struct cpumask *cpumask, return IRQ_SET_MASK_OK; } -#endif + +static void idu_irq_enable(struct irq_data *data) +{ + /* + * By default send all common interrupts to all available online CPUs. + * The affinity of common interrupts in IDU must be set manually since + * in some cases the kernel will not call irq_set_affinity() by itself: + * 1. When the kernel is not configured with support of SMP. + * 2. When the kernel is configured with support of SMP but upper + * interrupt controllers does not support setting of the affinity + * and cannot propagate it to IDU. + */ + idu_irq_set_affinity(data, cpu_online_mask, false); + idu_irq_unmask(data); +} static struct irq_chip idu_irq_chip = { .name = "MCIP IDU Intc", .irq_mask = idu_irq_mask, .irq_unmask = idu_irq_unmask, + .irq_enable = idu_irq_enable, #ifdef CONFIG_SMP .irq_set_affinity = idu_irq_set_affinity, #endif @@ -243,36 +256,14 @@ static int idu_irq_xlate(struct irq_domain *d, struct device_node *n, const u32 *intspec, unsigned int intsize, irq_hw_number_t *out_hwirq, unsigned int *out_type) { - irq_hw_number_t hwirq = *out_hwirq = intspec[0]; - int distri = intspec[1]; - unsigned long flags; - + /* + * Ignore value of interrupt distribution mode for common interrupts in + * IDU which resides in intspec[1] since setting an affinity using value + * from Device Tree is deprecated in ARC. + */ + *out_hwirq = intspec[0]; *out_type = IRQ_TYPE_NONE; - /* XXX: validate distribution scheme again online cpu mask */ - if (distri == 0) { - /* 0 - Round Robin to all cpus, otherwise 1 bit per core */ - raw_spin_lock_irqsave(&mcip_lock, flags); - idu_set_dest(hwirq, BIT(num_online_cpus()) - 1); - idu_set_mode(hwirq, IDU_M_TRIG_LEVEL, IDU_M_DISTRI_RR); - raw_spin_unlock_irqrestore(&mcip_lock, flags); - } else { - /* - * DEST based distribution for Level Triggered intr can only - * have 1 CPU, so generalize it to always contain 1 cpu - */ - int cpu = ffs(distri); - - if (cpu != fls(distri)) - pr_warn("IDU irq %lx distri mode set to cpu %x\n", - hwirq, cpu); - - raw_spin_lock_irqsave(&mcip_lock, flags); - idu_set_dest(hwirq, cpu); - idu_set_mode(hwirq, IDU_M_TRIG_LEVEL, IDU_M_DISTRI_DEST); - raw_spin_unlock_irqrestore(&mcip_lock, flags); - } - return 0; } diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c index 88674d972c9d..2afbafadb6ab 100644 --- a/arch/arc/kernel/smp.c +++ b/arch/arc/kernel/smp.c @@ -90,22 +90,37 @@ void __init smp_cpus_done(unsigned int max_cpus) */ static volatile int wake_flag; +#ifdef CONFIG_ISA_ARCOMPACT + +#define __boot_read(f) f +#define __boot_write(f, v) f = v + +#else + +#define __boot_read(f) arc_read_uncached_32(&f) +#define __boot_write(f, v) arc_write_uncached_32(&f, v) + +#endif + static void arc_default_smp_cpu_kick(int cpu, unsigned long pc) { BUG_ON(cpu == 0); - wake_flag = cpu; + + __boot_write(wake_flag, cpu); } void arc_platform_smp_wait_to_boot(int cpu) { - while (wake_flag != cpu) + /* for halt-on-reset, we've waited already */ + if (IS_ENABLED(CONFIG_ARC_SMP_HALT_ON_RESET)) + return; + + while (__boot_read(wake_flag) != cpu) ; - wake_flag = 0; - __asm__ __volatile__("j @first_lines_of_secondary \n"); + __boot_write(wake_flag, 0); } - const char *arc_platform_smp_cpuinfo(void) { return plat_smp_ops.info ? : ""; diff --git a/arch/arc/kernel/unaligned.c b/arch/arc/kernel/unaligned.c index abd961f3e763..5f69c3bd59bb 100644 --- a/arch/arc/kernel/unaligned.c +++ b/arch/arc/kernel/unaligned.c @@ -241,8 +241,9 @@ int misaligned_fixup(unsigned long address, struct pt_regs *regs, if (state.fault) goto fault; + /* clear any remanants of delay slot */ if (delay_mode(regs)) { - regs->ret = regs->bta; + regs->ret = regs->bta & ~1U; regs->status32 &= ~STATUS_DE_MASK; } else { regs->ret += state.instr_len; diff --git a/arch/arm/boot/dts/Makefile b/arch/arm/boot/dts/Makefile index f10fe8526239..01d178a2009f 100644 --- a/arch/arm/boot/dts/Makefile +++ b/arch/arm/boot/dts/Makefile @@ -617,7 +617,7 @@ dtb-$(CONFIG_ARCH_ORION5X) += \ orion5x-lacie-ethernet-disk-mini-v2.dtb \ orion5x-linkstation-lsgl.dtb \ orion5x-linkstation-lswtgl.dtb \ - orion5x-lschl.dtb \ + orion5x-linkstation-lschl.dtb \ orion5x-lswsgl.dtb \ orion5x-maxtor-shared-storage-2.dtb \ orion5x-netgear-wnr854t.dtb \ diff --git a/arch/arm/boot/dts/imx1.dtsi b/arch/arm/boot/dts/imx1.dtsi index b792eee3899b..2ee40bc9ec21 100644 --- a/arch/arm/boot/dts/imx1.dtsi +++ b/arch/arm/boot/dts/imx1.dtsi @@ -18,6 +18,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { gpio0 = &gpio1; diff --git a/arch/arm/boot/dts/imx23.dtsi b/arch/arm/boot/dts/imx23.dtsi index ac2a9da62b6c..43ccbbf754a3 100644 --- a/arch/arm/boot/dts/imx23.dtsi +++ b/arch/arm/boot/dts/imx23.dtsi @@ -16,6 +16,14 @@ / { #size-cells = <1>; interrupt-parent = <&icoll>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { gpio0 = &gpio0; diff --git a/arch/arm/boot/dts/imx25.dtsi b/arch/arm/boot/dts/imx25.dtsi index 831d09a28155..acd475659156 100644 --- a/arch/arm/boot/dts/imx25.dtsi +++ b/arch/arm/boot/dts/imx25.dtsi @@ -14,6 +14,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx27.dtsi b/arch/arm/boot/dts/imx27.dtsi index 9d8b5969ee3b..b397384248f4 100644 --- a/arch/arm/boot/dts/imx27.dtsi +++ b/arch/arm/boot/dts/imx27.dtsi @@ -19,6 +19,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx28.dtsi b/arch/arm/boot/dts/imx28.dtsi index 3aabf65a6a52..d6a2190b60ef 100644 --- a/arch/arm/boot/dts/imx28.dtsi +++ b/arch/arm/boot/dts/imx28.dtsi @@ -17,6 +17,14 @@ / { #size-cells = <1>; interrupt-parent = <&icoll>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &mac0; diff --git a/arch/arm/boot/dts/imx31.dtsi b/arch/arm/boot/dts/imx31.dtsi index 85cd8be22f71..23b0d2cf9acd 100644 --- a/arch/arm/boot/dts/imx31.dtsi +++ b/arch/arm/boot/dts/imx31.dtsi @@ -12,6 +12,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { serial0 = &uart1; diff --git a/arch/arm/boot/dts/imx35.dtsi b/arch/arm/boot/dts/imx35.dtsi index 9f40e6229189..d0496c65cea2 100644 --- a/arch/arm/boot/dts/imx35.dtsi +++ b/arch/arm/boot/dts/imx35.dtsi @@ -13,6 +13,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx50.dtsi b/arch/arm/boot/dts/imx50.dtsi index fe0221e4cbf7..ceae909e2201 100644 --- a/arch/arm/boot/dts/imx50.dtsi +++ b/arch/arm/boot/dts/imx50.dtsi @@ -17,6 +17,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx51.dtsi b/arch/arm/boot/dts/imx51.dtsi index 33526cade735..1ee1d542d9ad 100644 --- a/arch/arm/boot/dts/imx51.dtsi +++ b/arch/arm/boot/dts/imx51.dtsi @@ -19,6 +19,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx53.dtsi b/arch/arm/boot/dts/imx53.dtsi index ca51dc03e327..2e516f4985e4 100644 --- a/arch/arm/boot/dts/imx53.dtsi +++ b/arch/arm/boot/dts/imx53.dtsi @@ -19,6 +19,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx6dl.dtsi b/arch/arm/boot/dts/imx6dl.dtsi index 1ade1951e620..7aa120fbdc71 100644 --- a/arch/arm/boot/dts/imx6dl.dtsi +++ b/arch/arm/boot/dts/imx6dl.dtsi @@ -137,7 +137,7 @@ &gpio3 { &gpio4 { gpio-ranges = <&iomuxc 5 136 1>, <&iomuxc 6 145 1>, <&iomuxc 7 150 1>, <&iomuxc 8 146 1>, <&iomuxc 9 151 1>, <&iomuxc 10 147 1>, - <&iomuxc 11 151 1>, <&iomuxc 12 148 1>, <&iomuxc 13 153 1>, + <&iomuxc 11 152 1>, <&iomuxc 12 148 1>, <&iomuxc 13 153 1>, <&iomuxc 14 149 1>, <&iomuxc 15 154 1>, <&iomuxc 16 39 7>, <&iomuxc 23 56 1>, <&iomuxc 24 61 7>, <&iomuxc 31 46 1>; }; diff --git a/arch/arm/boot/dts/imx6qdl.dtsi b/arch/arm/boot/dts/imx6qdl.dtsi index 89b834f3fa17..e7d30f45b161 100644 --- a/arch/arm/boot/dts/imx6qdl.dtsi +++ b/arch/arm/boot/dts/imx6qdl.dtsi @@ -16,6 +16,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx6sl.dtsi b/arch/arm/boot/dts/imx6sl.dtsi index 19cbd879c448..cc9572ea2860 100644 --- a/arch/arm/boot/dts/imx6sl.dtsi +++ b/arch/arm/boot/dts/imx6sl.dtsi @@ -14,6 +14,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec; diff --git a/arch/arm/boot/dts/imx6sx.dtsi b/arch/arm/boot/dts/imx6sx.dtsi index 10f333016197..dd4ec85ecbaa 100644 --- a/arch/arm/boot/dts/imx6sx.dtsi +++ b/arch/arm/boot/dts/imx6sx.dtsi @@ -15,6 +15,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { can0 = &flexcan1; diff --git a/arch/arm/boot/dts/imx6ul.dtsi b/arch/arm/boot/dts/imx6ul.dtsi index 39845a7e0463..53d3f8e41e9b 100644 --- a/arch/arm/boot/dts/imx6ul.dtsi +++ b/arch/arm/boot/dts/imx6ul.dtsi @@ -15,6 +15,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { ethernet0 = &fec1; diff --git a/arch/arm/boot/dts/imx7s.dtsi b/arch/arm/boot/dts/imx7s.dtsi index 8ff2cbdd8f0d..be33dfc86838 100644 --- a/arch/arm/boot/dts/imx7s.dtsi +++ b/arch/arm/boot/dts/imx7s.dtsi @@ -50,6 +50,14 @@ / { #address-cells = <1>; #size-cells = <1>; + /* + * The decompressor and also some bootloaders rely on a + * pre-existing /chosen node to be available to insert the + * command line and merge other ATAGS info. + * Also for U-Boot there must be a pre-existing /memory node. + */ + chosen {}; + memory { device_type = "memory"; reg = <0 0>; }; aliases { gpio0 = &gpio1; diff --git a/arch/arm/boot/dts/orion5x-lschl.dts b/arch/arm/boot/dts/orion5x-linkstation-lschl.dts similarity index 98% rename from arch/arm/boot/dts/orion5x-lschl.dts rename to arch/arm/boot/dts/orion5x-linkstation-lschl.dts index 947409252845..ea6c881634b9 100644 --- a/arch/arm/boot/dts/orion5x-lschl.dts +++ b/arch/arm/boot/dts/orion5x-linkstation-lschl.dts @@ -2,7 +2,7 @@ * Device Tree file for Buffalo Linkstation LS-CHLv3 * * Copyright (C) 2016 Ash Hughes - * Copyright (C) 2015, 2016 + * Copyright (C) 2015-2017 * Roger Shimizu * * This file is dual-licensed: you can use it either under the terms @@ -52,7 +52,7 @@ #include / { - model = "Buffalo Linkstation Live v3 (LS-CHL)"; + model = "Buffalo Linkstation LiveV3 (LS-CHL)"; compatible = "buffalo,lschl", "marvell,orion5x-88f5182", "marvell,orion5x"; memory { /* 128 MB */ diff --git a/arch/arm/boot/dts/stih407-family.dtsi b/arch/arm/boot/dts/stih407-family.dtsi index c8b2944e304a..ace97e8576db 100644 --- a/arch/arm/boot/dts/stih407-family.dtsi +++ b/arch/arm/boot/dts/stih407-family.dtsi @@ -680,6 +680,7 @@ dwc3: dwc3@9900000 { phy-names = "usb2-phy", "usb3-phy"; phys = <&usb2_picophy0>, <&phy_port2 PHY_TYPE_USB3>; + snps,dis_u3_susphy_quirk; }; }; diff --git a/arch/arm/configs/ezx_defconfig b/arch/arm/configs/ezx_defconfig index ea316c4b890e..d3f1768840e2 100644 --- a/arch/arm/configs/ezx_defconfig +++ b/arch/arm/configs/ezx_defconfig @@ -64,8 +64,8 @@ CONFIG_NETFILTER=y CONFIG_NETFILTER_NETLINK_QUEUE=m CONFIG_NF_CONNTRACK=m CONFIG_NF_CONNTRACK_EVENTS=y -CONFIG_NF_CT_PROTO_SCTP=m -CONFIG_NF_CT_PROTO_UDPLITE=m +CONFIG_NF_CT_PROTO_SCTP=y +CONFIG_NF_CT_PROTO_UDPLITE=y CONFIG_NF_CONNTRACK_AMANDA=m CONFIG_NF_CONNTRACK_FTP=m CONFIG_NF_CONNTRACK_H323=m diff --git a/arch/arm/configs/imote2_defconfig b/arch/arm/configs/imote2_defconfig index 18e59feaa307..7f479cdb3479 100644 --- a/arch/arm/configs/imote2_defconfig +++ b/arch/arm/configs/imote2_defconfig @@ -56,8 +56,8 @@ CONFIG_NETFILTER=y CONFIG_NETFILTER_NETLINK_QUEUE=m CONFIG_NF_CONNTRACK=m CONFIG_NF_CONNTRACK_EVENTS=y -CONFIG_NF_CT_PROTO_SCTP=m -CONFIG_NF_CT_PROTO_UDPLITE=m +CONFIG_NF_CT_PROTO_SCTP=y +CONFIG_NF_CT_PROTO_UDPLITE=y CONFIG_NF_CONNTRACK_AMANDA=m CONFIG_NF_CONNTRACK_FTP=m CONFIG_NF_CONNTRACK_H323=m diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index ce131ed5939d..ae738a6319f6 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -600,7 +600,7 @@ static int gpr_set(struct task_struct *target, const void *kbuf, const void __user *ubuf) { int ret; - struct pt_regs newregs; + struct pt_regs newregs = *task_pt_regs(target); ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, &newregs, diff --git a/arch/arm/mach-imx/mmdc.c b/arch/arm/mach-imx/mmdc.c index 699157759120..c03bf28d8bbc 100644 --- a/arch/arm/mach-imx/mmdc.c +++ b/arch/arm/mach-imx/mmdc.c @@ -60,7 +60,6 @@ #define to_mmdc_pmu(p) container_of(p, struct mmdc_pmu, pmu) -static enum cpuhp_state cpuhp_mmdc_state; static int ddr_type; struct fsl_mmdc_devtype_data { @@ -82,6 +81,7 @@ static const struct of_device_id imx_mmdc_dt_ids[] = { #ifdef CONFIG_PERF_EVENTS +static enum cpuhp_state cpuhp_mmdc_state; static DEFINE_IDA(mmdc_ida); PMU_EVENT_ATTR_STRING(total-cycles, mmdc_pmu_total_cycles, "event=0x00") diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index 3a2e678b8d30..0122ad1a6027 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -610,9 +610,9 @@ static int __init early_abort_handler(unsigned long addr, unsigned int fsr, void __init early_abt_enable(void) { - fsr_info[22].fn = early_abort_handler; + fsr_info[FSR_FS_AEA].fn = early_abort_handler; local_abt_enable(); - fsr_info[22].fn = do_bad; + fsr_info[FSR_FS_AEA].fn = do_bad; } #ifndef CONFIG_ARM_LPAE diff --git a/arch/arm/mm/fault.h b/arch/arm/mm/fault.h index 67532f242271..afc1f84e763b 100644 --- a/arch/arm/mm/fault.h +++ b/arch/arm/mm/fault.h @@ -11,11 +11,15 @@ #define FSR_FS5_0 (0x3f) #ifdef CONFIG_ARM_LPAE +#define FSR_FS_AEA 17 + static inline int fsr_fs(unsigned int fsr) { return fsr & FSR_FS5_0; } #else +#define FSR_FS_AEA 22 + static inline int fsr_fs(unsigned int fsr) { return (fsr & FSR_FS3_0) | (fsr & FSR_FS4) >> 6; diff --git a/arch/arm64/boot/dts/amlogic/meson-gx.dtsi b/arch/arm64/boot/dts/amlogic/meson-gx.dtsi index eada0b58ba1c..0cbe24b49710 100644 --- a/arch/arm64/boot/dts/amlogic/meson-gx.dtsi +++ b/arch/arm64/boot/dts/amlogic/meson-gx.dtsi @@ -55,6 +55,24 @@ / { #address-cells = <2>; #size-cells = <2>; + reserved-memory { + #address-cells = <2>; + #size-cells = <2>; + ranges; + + /* 16 MiB reserved for Hardware ROM Firmware */ + hwrom_reserved: hwrom@0 { + reg = <0x0 0x0 0x0 0x1000000>; + no-map; + }; + + /* 2 MiB reserved for ARM Trusted Firmware (BL31) */ + secmon_reserved: secmon@10000000 { + reg = <0x0 0x10000000 0x0 0x200000>; + no-map; + }; + }; + cpus { #address-cells = <0x2>; #size-cells = <0x0>; diff --git a/arch/arm64/boot/dts/amlogic/meson-gxbb-odroidc2.dts b/arch/arm64/boot/dts/amlogic/meson-gxbb-odroidc2.dts index 5d28e1cdc998..c59403adb387 100644 --- a/arch/arm64/boot/dts/amlogic/meson-gxbb-odroidc2.dts +++ b/arch/arm64/boot/dts/amlogic/meson-gxbb-odroidc2.dts @@ -151,6 +151,18 @@ ðmac { status = "okay"; pinctrl-0 = <ð_rgmii_pins>; pinctrl-names = "default"; + phy-handle = <ð_phy0>; + + mdio { + compatible = "snps,dwmac-mdio"; + #address-cells = <1>; + #size-cells = <0>; + + eth_phy0: ethernet-phy@0 { + reg = <0>; + eee-broken-1000t; + }; + }; }; &ir { diff --git a/arch/arm64/crypto/aes-modes.S b/arch/arm64/crypto/aes-modes.S index c53dbeae79f2..838dad5c209f 100644 --- a/arch/arm64/crypto/aes-modes.S +++ b/arch/arm64/crypto/aes-modes.S @@ -193,15 +193,16 @@ AES_ENTRY(aes_cbc_encrypt) cbz w6, .Lcbcencloop ld1 {v0.16b}, [x5] /* get iv */ - enc_prepare w3, x2, x5 + enc_prepare w3, x2, x6 .Lcbcencloop: ld1 {v1.16b}, [x1], #16 /* get next pt block */ eor v0.16b, v0.16b, v1.16b /* ..and xor with iv */ - encrypt_block v0, w3, x2, x5, w6 + encrypt_block v0, w3, x2, x6, w7 st1 {v0.16b}, [x0], #16 subs w4, w4, #1 bne .Lcbcencloop + st1 {v0.16b}, [x5] /* return iv */ ret AES_ENDPROC(aes_cbc_encrypt) @@ -211,7 +212,7 @@ AES_ENTRY(aes_cbc_decrypt) cbz w6, .LcbcdecloopNx ld1 {v7.16b}, [x5] /* get iv */ - dec_prepare w3, x2, x5 + dec_prepare w3, x2, x6 .LcbcdecloopNx: #if INTERLEAVE >= 2 @@ -248,7 +249,7 @@ AES_ENTRY(aes_cbc_decrypt) .Lcbcdecloop: ld1 {v1.16b}, [x1], #16 /* get next ct block */ mov v0.16b, v1.16b /* ...and copy to v0 */ - decrypt_block v0, w3, x2, x5, w6 + decrypt_block v0, w3, x2, x6, w7 eor v0.16b, v0.16b, v7.16b /* xor with iv => pt */ mov v7.16b, v1.16b /* ct is next iv */ st1 {v0.16b}, [x0], #16 @@ -256,6 +257,7 @@ AES_ENTRY(aes_cbc_decrypt) bne .Lcbcdecloop .Lcbcdecout: FRAME_POP + st1 {v7.16b}, [x5] /* return iv */ ret AES_ENDPROC(aes_cbc_decrypt) @@ -267,24 +269,15 @@ AES_ENDPROC(aes_cbc_decrypt) AES_ENTRY(aes_ctr_encrypt) FRAME_PUSH - cbnz w6, .Lctrfirst /* 1st time around? */ - umov x5, v4.d[1] /* keep swabbed ctr in reg */ - rev x5, x5 -#if INTERLEAVE >= 2 - cmn w5, w4 /* 32 bit overflow? */ - bcs .Lctrinc - add x5, x5, #1 /* increment BE ctr */ - b .LctrincNx -#else - b .Lctrinc -#endif -.Lctrfirst: + cbz w6, .Lctrnotfirst /* 1st time around? */ enc_prepare w3, x2, x6 ld1 {v4.16b}, [x5] - umov x5, v4.d[1] /* keep swabbed ctr in reg */ - rev x5, x5 + +.Lctrnotfirst: + umov x8, v4.d[1] /* keep swabbed ctr in reg */ + rev x8, x8 #if INTERLEAVE >= 2 - cmn w5, w4 /* 32 bit overflow? */ + cmn w8, w4 /* 32 bit overflow? */ bcs .Lctrloop .LctrloopNx: subs w4, w4, #INTERLEAVE @@ -292,11 +285,11 @@ AES_ENTRY(aes_ctr_encrypt) #if INTERLEAVE == 2 mov v0.8b, v4.8b mov v1.8b, v4.8b - rev x7, x5 - add x5, x5, #1 + rev x7, x8 + add x8, x8, #1 ins v0.d[1], x7 - rev x7, x5 - add x5, x5, #1 + rev x7, x8 + add x8, x8, #1 ins v1.d[1], x7 ld1 {v2.16b-v3.16b}, [x1], #32 /* get 2 input blocks */ do_encrypt_block2x @@ -305,7 +298,7 @@ AES_ENTRY(aes_ctr_encrypt) st1 {v0.16b-v1.16b}, [x0], #32 #else ldr q8, =0x30000000200000001 /* addends 1,2,3[,0] */ - dup v7.4s, w5 + dup v7.4s, w8 mov v0.16b, v4.16b add v7.4s, v7.4s, v8.4s mov v1.16b, v4.16b @@ -323,18 +316,12 @@ AES_ENTRY(aes_ctr_encrypt) eor v2.16b, v7.16b, v2.16b eor v3.16b, v5.16b, v3.16b st1 {v0.16b-v3.16b}, [x0], #64 - add x5, x5, #INTERLEAVE + add x8, x8, #INTERLEAVE #endif - cbz w4, .LctroutNx -.LctrincNx: - rev x7, x5 + rev x7, x8 ins v4.d[1], x7 + cbz w4, .Lctrout b .LctrloopNx -.LctroutNx: - sub x5, x5, #1 - rev x7, x5 - ins v4.d[1], x7 - b .Lctrout .Lctr1x: adds w4, w4, #INTERLEAVE beq .Lctrout @@ -342,30 +329,39 @@ AES_ENTRY(aes_ctr_encrypt) .Lctrloop: mov v0.16b, v4.16b encrypt_block v0, w3, x2, x6, w7 + + adds x8, x8, #1 /* increment BE ctr */ + rev x7, x8 + ins v4.d[1], x7 + bcs .Lctrcarry /* overflow? */ + +.Lctrcarrydone: subs w4, w4, #1 bmi .Lctrhalfblock /* blocks < 0 means 1/2 block */ ld1 {v3.16b}, [x1], #16 eor v3.16b, v0.16b, v3.16b st1 {v3.16b}, [x0], #16 - beq .Lctrout -.Lctrinc: - adds x5, x5, #1 /* increment BE ctr */ - rev x7, x5 - ins v4.d[1], x7 - bcc .Lctrloop /* no overflow? */ + bne .Lctrloop + +.Lctrout: + st1 {v4.16b}, [x5] /* return next CTR value */ + FRAME_POP + ret + +.Lctrhalfblock: + ld1 {v3.8b}, [x1] + eor v3.8b, v0.8b, v3.8b + st1 {v3.8b}, [x0] + FRAME_POP + ret + +.Lctrcarry: umov x7, v4.d[0] /* load upper word of ctr */ rev x7, x7 /* ... to handle the carry */ add x7, x7, #1 rev x7, x7 ins v4.d[0], x7 - b .Lctrloop -.Lctrhalfblock: - ld1 {v3.8b}, [x1] - eor v3.8b, v0.8b, v3.8b - st1 {v3.8b}, [x0] -.Lctrout: - FRAME_POP - ret + b .Lctrcarrydone AES_ENDPROC(aes_ctr_encrypt) .ltorg diff --git a/arch/arm64/kernel/topology.c b/arch/arm64/kernel/topology.c index 23e9e13bd2aa..655e65f38f31 100644 --- a/arch/arm64/kernel/topology.c +++ b/arch/arm64/kernel/topology.c @@ -11,6 +11,7 @@ * for more details. */ +#include #include #include #include @@ -209,7 +210,12 @@ static struct notifier_block init_cpu_capacity_notifier = { static int __init register_cpufreq_notifier(void) { - if (cap_parsing_failed) + /* + * on ACPI-based systems we need to use the default cpu capacity + * until we have the necessary code to parse the cpu capacity, so + * skip registering cpufreq notifier. + */ + if (!acpi_disabled || cap_parsing_failed) return -EINVAL; if (!alloc_cpumask_var(&cpus_to_visit, GFP_KERNEL)) { diff --git a/arch/parisc/include/asm/bitops.h b/arch/parisc/include/asm/bitops.h index 3f9406d9b9d6..da87943328a5 100644 --- a/arch/parisc/include/asm/bitops.h +++ b/arch/parisc/include/asm/bitops.h @@ -6,7 +6,7 @@ #endif #include -#include /* for BITS_PER_LONG/SHIFT_PER_LONG */ +#include #include #include #include @@ -17,6 +17,12 @@ * to include/asm-i386/bitops.h or kerneldoc */ +#if __BITS_PER_LONG == 64 +#define SHIFT_PER_LONG 6 +#else +#define SHIFT_PER_LONG 5 +#endif + #define CHOP_SHIFTCOUNT(x) (((unsigned long) (x)) & (BITS_PER_LONG - 1)) diff --git a/arch/parisc/include/uapi/asm/bitsperlong.h b/arch/parisc/include/uapi/asm/bitsperlong.h index e0a23c7bdd43..07fa7e50bdc0 100644 --- a/arch/parisc/include/uapi/asm/bitsperlong.h +++ b/arch/parisc/include/uapi/asm/bitsperlong.h @@ -3,10 +3,8 @@ #if defined(__LP64__) #define __BITS_PER_LONG 64 -#define SHIFT_PER_LONG 6 #else #define __BITS_PER_LONG 32 -#define SHIFT_PER_LONG 5 #endif #include diff --git a/arch/parisc/include/uapi/asm/swab.h b/arch/parisc/include/uapi/asm/swab.h index e78403b129ef..928e1bbac98f 100644 --- a/arch/parisc/include/uapi/asm/swab.h +++ b/arch/parisc/include/uapi/asm/swab.h @@ -1,6 +1,7 @@ #ifndef _PARISC_SWAB_H #define _PARISC_SWAB_H +#include #include #include @@ -38,7 +39,7 @@ static inline __attribute_const__ __u32 __arch_swab32(__u32 x) } #define __arch_swab32 __arch_swab32 -#if BITS_PER_LONG > 32 +#if __BITS_PER_LONG > 32 /* ** From "PA-RISC 2.0 Architecture", HP Professional Books. ** See Appendix I page 8 , "Endian Byte Swapping". @@ -61,6 +62,6 @@ static inline __attribute_const__ __u64 __arch_swab64(__u64 x) return x; } #define __arch_swab64 __arch_swab64 -#endif /* BITS_PER_LONG > 32 */ +#endif /* __BITS_PER_LONG > 32 */ #endif /* _PARISC_SWAB_H */ diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index a8ee573fe610..281f4f1fcd1f 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -164,7 +164,6 @@ config PPC select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE select HAVE_ARCH_HARDENED_USERCOPY select HAVE_KERNEL_GZIP - select HAVE_CC_STACKPROTECTOR config GENERIC_CSUM def_bool CPU_LITTLE_ENDIAN @@ -484,6 +483,7 @@ config RELOCATABLE bool "Build a relocatable kernel" depends on (PPC64 && !COMPILE_TEST) || (FLATMEM && (44x || FSL_BOOKE)) select NONSTATIC_KERNEL + select MODULE_REL_CRCS if MODVERSIONS help This builds a kernel image that is capable of running at the location the kernel is loaded at. For ppc32, there is no any diff --git a/arch/powerpc/include/asm/cpu_has_feature.h b/arch/powerpc/include/asm/cpu_has_feature.h index b312b152461b..6e834caa3720 100644 --- a/arch/powerpc/include/asm/cpu_has_feature.h +++ b/arch/powerpc/include/asm/cpu_has_feature.h @@ -23,7 +23,9 @@ static __always_inline bool cpu_has_feature(unsigned long feature) { int i; +#ifndef __clang__ /* clang can't cope with this */ BUILD_BUG_ON(!__builtin_constant_p(feature)); +#endif #ifdef CONFIG_JUMP_LABEL_FEATURE_CHECK_DEBUG if (!static_key_initialized) { diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h index a34c764ca8dd..233a7e8cc8e3 100644 --- a/arch/powerpc/include/asm/mmu.h +++ b/arch/powerpc/include/asm/mmu.h @@ -160,7 +160,9 @@ static __always_inline bool mmu_has_feature(unsigned long feature) { int i; +#ifndef __clang__ /* clang can't cope with this */ BUILD_BUG_ON(!__builtin_constant_p(feature)); +#endif #ifdef CONFIG_JUMP_LABEL_FEATURE_CHECK_DEBUG if (!static_key_initialized) { diff --git a/arch/powerpc/include/asm/module.h b/arch/powerpc/include/asm/module.h index cc12c61ef315..53885512b8d3 100644 --- a/arch/powerpc/include/asm/module.h +++ b/arch/powerpc/include/asm/module.h @@ -90,9 +90,5 @@ static inline int module_finalize_ftrace(struct module *mod, const Elf_Shdr *sec } #endif -#if defined(CONFIG_MODVERSIONS) && defined(CONFIG_PPC64) -#define ARCH_RELOCATES_KCRCTAB -#define reloc_start PHYSICAL_START -#endif #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_MODULE_H */ diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h index 0d4531aa2052..dff79798903d 100644 --- a/arch/powerpc/include/asm/reg.h +++ b/arch/powerpc/include/asm/reg.h @@ -649,9 +649,10 @@ #define SRR1_ISI_N_OR_G 0x10000000 /* ISI: Access is no-exec or G */ #define SRR1_ISI_PROT 0x08000000 /* ISI: Other protection fault */ #define SRR1_WAKEMASK 0x00380000 /* reason for wakeup */ -#define SRR1_WAKEMASK_P8 0x003c0000 /* reason for wakeup on POWER8 */ +#define SRR1_WAKEMASK_P8 0x003c0000 /* reason for wakeup on POWER8 and 9 */ #define SRR1_WAKESYSERR 0x00300000 /* System error */ #define SRR1_WAKEEE 0x00200000 /* External interrupt */ +#define SRR1_WAKEHVI 0x00240000 /* Hypervisor Virtualization Interrupt (P9) */ #define SRR1_WAKEMT 0x00280000 /* mtctrl */ #define SRR1_WAKEHMI 0x00280000 /* Hypervisor maintenance */ #define SRR1_WAKEDEC 0x00180000 /* Decrementer interrupt */ diff --git a/arch/powerpc/include/asm/stackprotector.h b/arch/powerpc/include/asm/stackprotector.h deleted file mode 100644 index 6720190eabec..000000000000 --- a/arch/powerpc/include/asm/stackprotector.h +++ /dev/null @@ -1,40 +0,0 @@ -/* - * GCC stack protector support. - * - * Stack protector works by putting predefined pattern at the start of - * the stack frame and verifying that it hasn't been overwritten when - * returning from the function. The pattern is called stack canary - * and gcc expects it to be defined by a global variable called - * "__stack_chk_guard" on PPC. This unfortunately means that on SMP - * we cannot have a different canary value per task. - */ - -#ifndef _ASM_STACKPROTECTOR_H -#define _ASM_STACKPROTECTOR_H - -#include -#include -#include - -extern unsigned long __stack_chk_guard; - -/* - * Initialize the stackprotector canary value. - * - * NOTE: this must only be called from functions that never return, - * and it must always be inlined. - */ -static __always_inline void boot_init_stack_canary(void) -{ - unsigned long canary; - - /* Try to get a semi random initial value. */ - get_random_bytes(&canary, sizeof(canary)); - canary ^= mftb(); - canary ^= LINUX_VERSION_CODE; - - current->stack_canary = canary; - __stack_chk_guard = current->stack_canary; -} - -#endif /* _ASM_STACKPROTECTOR_H */ diff --git a/arch/powerpc/include/asm/xics.h b/arch/powerpc/include/asm/xics.h index f0b238516e9b..e0b9e576905a 100644 --- a/arch/powerpc/include/asm/xics.h +++ b/arch/powerpc/include/asm/xics.h @@ -44,6 +44,7 @@ static inline int icp_hv_init(void) { return -ENODEV; } #ifdef CONFIG_PPC_POWERNV extern int icp_opal_init(void); +extern void icp_opal_flush_interrupt(void); #else static inline int icp_opal_init(void) { return -ENODEV; } #endif diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index 23f8082d7bfa..f4c2b52e58b3 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -19,10 +19,6 @@ CFLAGS_init.o += $(DISABLE_LATENT_ENTROPY_PLUGIN) CFLAGS_btext.o += $(DISABLE_LATENT_ENTROPY_PLUGIN) CFLAGS_prom.o += $(DISABLE_LATENT_ENTROPY_PLUGIN) -# -fstack-protector triggers protection checks in this code, -# but it is being used too early to link to meaningful stack_chk logic. -CFLAGS_prom_init.o += $(call cc-option, -fno-stack-protector) - ifdef CONFIG_FUNCTION_TRACER # Do not trace early boot code CFLAGS_REMOVE_cputable.o = -mno-sched-epilog $(CC_FLAGS_FTRACE) diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c index 0601e6a7297c..195a9fc8f81c 100644 --- a/arch/powerpc/kernel/asm-offsets.c +++ b/arch/powerpc/kernel/asm-offsets.c @@ -91,9 +91,6 @@ int main(void) DEFINE(TI_livepatch_sp, offsetof(struct thread_info, livepatch_sp)); #endif -#ifdef CONFIG_CC_STACKPROTECTOR - DEFINE(TSK_STACK_CANARY, offsetof(struct task_struct, stack_canary)); -#endif DEFINE(KSP, offsetof(struct thread_struct, ksp)); DEFINE(PT_REGS, offsetof(struct thread_struct, regs)); #ifdef CONFIG_BOOKE diff --git a/arch/powerpc/kernel/eeh_driver.c b/arch/powerpc/kernel/eeh_driver.c index d88573bdd090..b94887165a10 100644 --- a/arch/powerpc/kernel/eeh_driver.c +++ b/arch/powerpc/kernel/eeh_driver.c @@ -545,7 +545,7 @@ static void *eeh_pe_detach_dev(void *data, void *userdata) static void *__eeh_clear_pe_frozen_state(void *data, void *flag) { struct eeh_pe *pe = (struct eeh_pe *)data; - bool *clear_sw_state = flag; + bool clear_sw_state = *(bool *)flag; int i, rc = 1; for (i = 0; rc && i < 3; i++) diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S index 5742dbdbee46..3841d749a430 100644 --- a/arch/powerpc/kernel/entry_32.S +++ b/arch/powerpc/kernel/entry_32.S @@ -674,11 +674,7 @@ BEGIN_FTR_SECTION mtspr SPRN_SPEFSCR,r0 /* restore SPEFSCR reg */ END_FTR_SECTION_IFSET(CPU_FTR_SPE) #endif /* CONFIG_SPE */ -#if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP) - lwz r0,TSK_STACK_CANARY(r2) - lis r4,__stack_chk_guard@ha - stw r0,__stack_chk_guard@l(r4) -#endif + lwz r0,_CCR(r1) mtcrf 0xFF,r0 /* r3-r12 are destroyed -- Cort */ diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_64.c index bb1807184bad..0b0f89685b67 100644 --- a/arch/powerpc/kernel/module_64.c +++ b/arch/powerpc/kernel/module_64.c @@ -286,14 +286,6 @@ static void dedotify_versions(struct modversion_info *vers, for (end = (void *)vers + size; vers < end; vers++) if (vers->name[0] == '.') { memmove(vers->name, vers->name+1, strlen(vers->name)); -#ifdef ARCH_RELOCATES_KCRCTAB - /* The TOC symbol has no CRC computed. To avoid CRC - * check failing, we must force it to the expected - * value (see CRC check in module.c). - */ - if (!strcmp(vers->name, "TOC.")) - vers->crc = -(unsigned long)reloc_start; -#endif } } diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index 04885cec24df..5dd056df0baa 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -64,12 +64,6 @@ #include #include -#ifdef CONFIG_CC_STACKPROTECTOR -#include -unsigned long __stack_chk_guard __read_mostly; -EXPORT_SYMBOL(__stack_chk_guard); -#endif - /* Transactional Memory debug */ #ifdef TM_DEBUG_SW #define TM_DEBUG(x...) printk(KERN_INFO x) diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index ec47a939cbdd..ac83eb04a8b8 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -2834,6 +2834,9 @@ static void __init prom_find_boot_cpu(void) cpu_pkg = call_prom("instance-to-package", 1, 1, prom_cpu); + if (!PHANDLE_VALID(cpu_pkg)) + return; + prom_getprop(cpu_pkg, "reg", &rval, sizeof(rval)); prom.cpu = be32_to_cpu(rval); diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c index 6fd30ac7d14a..62a50d6d1053 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -253,8 +253,11 @@ int do_page_fault(struct pt_regs *regs, unsigned long address, if (unlikely(debugger_fault_handler(regs))) goto bail; - /* On a kernel SLB miss we can only check for a valid exception entry */ - if (!user_mode(regs) && (address >= TASK_SIZE)) { + /* + * The kernel should never take an execute fault nor should it + * take a page fault to a kernel address. + */ + if (!user_mode(regs) && (is_exec || (address >= TASK_SIZE))) { rc = SIGSEGV; goto bail; } @@ -390,20 +393,6 @@ int do_page_fault(struct pt_regs *regs, unsigned long address, #endif /* CONFIG_8xx */ if (is_exec) { - /* - * An execution fault + no execute ? - * - * On CPUs that don't have CPU_FTR_COHERENT_ICACHE we - * deliberately create NX mappings, and use the fault to do the - * cache flush. This is usually handled in hash_page_do_lazy_icache() - * but we could end up here if that races with a concurrent PTE - * update. In that case we need to fall through here to the VMA - * check below. - */ - if (cpu_has_feature(CPU_FTR_COHERENT_ICACHE) && - (regs->msr & SRR1_ISI_N_OR_G)) - goto bad_area; - /* * Allow execution from readable areas if the MMU does not * provide separate controls over reading and executing. diff --git a/arch/powerpc/mm/pgtable-radix.c b/arch/powerpc/mm/pgtable-radix.c index cfa53ccc8baf..34f1a0dbc898 100644 --- a/arch/powerpc/mm/pgtable-radix.c +++ b/arch/powerpc/mm/pgtable-radix.c @@ -65,7 +65,7 @@ int radix__map_kernel_page(unsigned long ea, unsigned long pa, if (!pmdp) return -ENOMEM; if (map_page_size == PMD_SIZE) { - ptep = (pte_t *)pudp; + ptep = pmdp_ptep(pmdp); goto set_the_pte; } ptep = pte_alloc_kernel(pmdp, ea); @@ -90,7 +90,7 @@ int radix__map_kernel_page(unsigned long ea, unsigned long pa, } pmdp = pmd_offset(pudp, ea); if (map_page_size == PMD_SIZE) { - ptep = (pte_t *)pudp; + ptep = pmdp_ptep(pmdp); goto set_the_pte; } if (!pmd_present(*pmdp)) { diff --git a/arch/powerpc/mm/tlb-radix.c b/arch/powerpc/mm/tlb-radix.c index 61b79119065f..952713d6cf04 100644 --- a/arch/powerpc/mm/tlb-radix.c +++ b/arch/powerpc/mm/tlb-radix.c @@ -50,9 +50,7 @@ static inline void _tlbiel_pid(unsigned long pid, unsigned long ric) for (set = 0; set < POWER9_TLB_SETS_RADIX ; set++) { __tlbiel_pid(pid, set, ric); } - if (cpu_has_feature(CPU_FTR_POWER9_DD1)) - asm volatile(PPC_INVALIDATE_ERAT : : :"memory"); - return; + asm volatile(PPC_INVALIDATE_ERAT "; isync" : : :"memory"); } static inline void _tlbie_pid(unsigned long pid, unsigned long ric) @@ -85,8 +83,6 @@ static inline void _tlbiel_va(unsigned long va, unsigned long pid, asm volatile(PPC_TLBIEL(%0, %4, %3, %2, %1) : : "r"(rb), "i"(r), "i"(prs), "i"(ric), "r"(rs) : "memory"); asm volatile("ptesync": : :"memory"); - if (cpu_has_feature(CPU_FTR_POWER9_DD1)) - asm volatile(PPC_INVALIDATE_ERAT : : :"memory"); } static inline void _tlbie_va(unsigned long va, unsigned long pid, diff --git a/arch/powerpc/platforms/powernv/smp.c b/arch/powerpc/platforms/powernv/smp.c index c789258ae1e1..eec0e8d0454d 100644 --- a/arch/powerpc/platforms/powernv/smp.c +++ b/arch/powerpc/platforms/powernv/smp.c @@ -155,8 +155,10 @@ static void pnv_smp_cpu_kill_self(void) wmask = SRR1_WAKEMASK_P8; idle_states = pnv_get_supported_cpuidle_states(); + /* We don't want to take decrementer interrupts while we are offline, - * so clear LPCR:PECE1. We keep PECE2 enabled. + * so clear LPCR:PECE1. We keep PECE2 (and LPCR_PECE_HVEE on P9) + * enabled as to let IPIs in. */ mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1); @@ -206,8 +208,12 @@ static void pnv_smp_cpu_kill_self(void) * contains 0. */ if (((srr1 & wmask) == SRR1_WAKEEE) || + ((srr1 & wmask) == SRR1_WAKEHVI) || (local_paca->irq_happened & PACA_IRQ_EE)) { - icp_native_flush_interrupt(); + if (cpu_has_feature(CPU_FTR_ARCH_300)) + icp_opal_flush_interrupt(); + else + icp_native_flush_interrupt(); } else if ((srr1 & wmask) == SRR1_WAKEHDBELL) { unsigned long msg = PPC_DBELL_TYPE(PPC_DBELL_SERVER); asm volatile(PPC_MSGCLR(%0) : : "r" (msg)); @@ -221,6 +227,8 @@ static void pnv_smp_cpu_kill_self(void) if (srr1 && !generic_check_cpu_restart(cpu)) DBG("CPU%d Unexpected exit while offline !\n", cpu); } + + /* Re-enable decrementer interrupts */ mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) | LPCR_PECE1); DBG("CPU%d coming online...\n", cpu); } diff --git a/arch/powerpc/sysdev/xics/icp-opal.c b/arch/powerpc/sysdev/xics/icp-opal.c index 60c57657c772..f9670eabfcfa 100644 --- a/arch/powerpc/sysdev/xics/icp-opal.c +++ b/arch/powerpc/sysdev/xics/icp-opal.c @@ -120,18 +120,49 @@ static void icp_opal_cause_ipi(int cpu, unsigned long data) { int hw_cpu = get_hard_smp_processor_id(cpu); + kvmppc_set_host_ipi(cpu, 1); opal_int_set_mfrr(hw_cpu, IPI_PRIORITY); } static irqreturn_t icp_opal_ipi_action(int irq, void *dev_id) { - int hw_cpu = hard_smp_processor_id(); + int cpu = smp_processor_id(); - opal_int_set_mfrr(hw_cpu, 0xff); + kvmppc_set_host_ipi(cpu, 0); + opal_int_set_mfrr(get_hard_smp_processor_id(cpu), 0xff); return smp_ipi_demux(); } +/* + * Called when an interrupt is received on an off-line CPU to + * clear the interrupt, so that the CPU can go back to nap mode. + */ +void icp_opal_flush_interrupt(void) +{ + unsigned int xirr; + unsigned int vec; + + do { + xirr = icp_opal_get_xirr(); + vec = xirr & 0x00ffffff; + if (vec == XICS_IRQ_SPURIOUS) + break; + if (vec == XICS_IPI) { + /* Clear pending IPI */ + int cpu = smp_processor_id(); + kvmppc_set_host_ipi(cpu, 0); + opal_int_set_mfrr(get_hard_smp_processor_id(cpu), 0xff); + } else { + pr_err("XICS: hw interrupt 0x%x to offline cpu, " + "disabling\n", vec); + xics_mask_unknown_vec(vec); + } + + /* EOI the interrupt */ + } while (opal_int_eoi(xirr) > 0); +} + #endif /* CONFIG_SMP */ static const struct icp_ops icp_opal_ops = { diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 7447ba509c30..12020b55887b 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -963,6 +963,11 @@ static int s390_fpregs_set(struct task_struct *target, if (target == current) save_fpu_regs(); + if (MACHINE_HAS_VX) + convert_vx_to_fp(fprs, target->thread.fpu.vxrs); + else + memcpy(&fprs, target->thread.fpu.fprs, sizeof(fprs)); + /* If setting FPC, must validate it first. */ if (count > 0 && pos < offsetof(s390_fp_regs, fprs)) { u32 ufpc[2] = { target->thread.fpu.fpc, 0 }; @@ -1067,6 +1072,9 @@ static int s390_vxrs_low_set(struct task_struct *target, if (target == current) save_fpu_regs(); + for (i = 0; i < __NUM_VXRS_LOW; i++) + vxrs[i] = *((__u64 *)(target->thread.fpu.vxrs + i) + 1); + rc = user_regset_copyin(&pos, &count, &kbuf, &ubuf, vxrs, 0, -1); if (rc == 0) for (i = 0; i < __NUM_VXRS_LOW; i++) diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 7a1897c51c54..d56ef26d4681 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -202,7 +202,7 @@ static inline pgste_t ptep_xchg_start(struct mm_struct *mm, return pgste; } -static inline void ptep_xchg_commit(struct mm_struct *mm, +static inline pte_t ptep_xchg_commit(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pgste_t pgste, pte_t old, pte_t new) { @@ -220,6 +220,7 @@ static inline void ptep_xchg_commit(struct mm_struct *mm, } else { *ptep = new; } + return old; } pte_t ptep_xchg_direct(struct mm_struct *mm, unsigned long addr, @@ -231,7 +232,7 @@ pte_t ptep_xchg_direct(struct mm_struct *mm, unsigned long addr, preempt_disable(); pgste = ptep_xchg_start(mm, addr, ptep); old = ptep_flush_direct(mm, addr, ptep); - ptep_xchg_commit(mm, addr, ptep, pgste, old, new); + old = ptep_xchg_commit(mm, addr, ptep, pgste, old, new); preempt_enable(); return old; } @@ -246,7 +247,7 @@ pte_t ptep_xchg_lazy(struct mm_struct *mm, unsigned long addr, preempt_disable(); pgste = ptep_xchg_start(mm, addr, ptep); old = ptep_flush_lazy(mm, addr, ptep); - ptep_xchg_commit(mm, addr, ptep, pgste, old, new); + old = ptep_xchg_commit(mm, addr, ptep, pgste, old, new); preempt_enable(); return old; } diff --git a/arch/sparc/include/asm/mmu_context_64.h b/arch/sparc/include/asm/mmu_context_64.h index b84be675e507..d0317993e947 100644 --- a/arch/sparc/include/asm/mmu_context_64.h +++ b/arch/sparc/include/asm/mmu_context_64.h @@ -35,15 +35,15 @@ void __tsb_context_switch(unsigned long pgd_pa, static inline void tsb_context_switch(struct mm_struct *mm) { __tsb_context_switch(__pa(mm->pgd), - &mm->context.tsb_block[0], + &mm->context.tsb_block[MM_TSB_BASE], #if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) - (mm->context.tsb_block[1].tsb ? - &mm->context.tsb_block[1] : + (mm->context.tsb_block[MM_TSB_HUGE].tsb ? + &mm->context.tsb_block[MM_TSB_HUGE] : NULL) #else NULL #endif - , __pa(&mm->context.tsb_descr[0])); + , __pa(&mm->context.tsb_descr[MM_TSB_BASE])); } void tsb_grow(struct mm_struct *mm, diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c index 3bebf395252c..4d0248aa0928 100644 --- a/arch/sparc/kernel/irq_64.c +++ b/arch/sparc/kernel/irq_64.c @@ -1021,7 +1021,7 @@ static void __init alloc_one_queue(unsigned long *pa_ptr, unsigned long qmask) unsigned long order = get_order(size); unsigned long p; - p = __get_free_pages(GFP_KERNEL, order); + p = __get_free_pages(GFP_KERNEL | __GFP_ZERO, order); if (!p) { prom_printf("SUN4V: Error, cannot allocate queue.\n"); prom_halt(); diff --git a/arch/sparc/kernel/sstate.c b/arch/sparc/kernel/sstate.c index c59af546f522..3caed4023589 100644 --- a/arch/sparc/kernel/sstate.c +++ b/arch/sparc/kernel/sstate.c @@ -43,8 +43,8 @@ static const char poweroff_msg[32] __attribute__((aligned(32))) = "Linux powering off"; static const char rebooting_msg[32] __attribute__((aligned(32))) = "Linux rebooting"; -static const char panicing_msg[32] __attribute__((aligned(32))) = - "Linux panicing"; +static const char panicking_msg[32] __attribute__((aligned(32))) = + "Linux panicking"; static int sstate_reboot_call(struct notifier_block *np, unsigned long type, void *_unused) { @@ -76,7 +76,7 @@ static struct notifier_block sstate_reboot_notifier = { static int sstate_panic_event(struct notifier_block *n, unsigned long event, void *ptr) { - do_set_sstate(HV_SOFT_STATE_TRANSITION, panicing_msg); + do_set_sstate(HV_SOFT_STATE_TRANSITION, panicking_msg); return NOTIFY_DONE; } diff --git a/arch/sparc/kernel/traps_64.c b/arch/sparc/kernel/traps_64.c index 4bc10e44d1ca..dfc97a47c9a0 100644 --- a/arch/sparc/kernel/traps_64.c +++ b/arch/sparc/kernel/traps_64.c @@ -2051,6 +2051,73 @@ void sun4v_resum_overflow(struct pt_regs *regs) atomic_inc(&sun4v_resum_oflow_cnt); } +/* Given a set of registers, get the virtual addressi that was being accessed + * by the faulting instructions at tpc. + */ +static unsigned long sun4v_get_vaddr(struct pt_regs *regs) +{ + unsigned int insn; + + if (!copy_from_user(&insn, (void __user *)regs->tpc, 4)) { + return compute_effective_address(regs, insn, + (insn >> 25) & 0x1f); + } + return 0; +} + +/* Attempt to handle non-resumable errors generated from userspace. + * Returns true if the signal was handled, false otherwise. + */ +bool sun4v_nonresum_error_user_handled(struct pt_regs *regs, + struct sun4v_error_entry *ent) { + + unsigned int attrs = ent->err_attrs; + + if (attrs & SUN4V_ERR_ATTRS_MEMORY) { + unsigned long addr = ent->err_raddr; + siginfo_t info; + + if (addr == ~(u64)0) { + /* This seems highly unlikely to ever occur */ + pr_emerg("SUN4V NON-RECOVERABLE ERROR: Memory error detected in unknown location!\n"); + } else { + unsigned long page_cnt = DIV_ROUND_UP(ent->err_size, + PAGE_SIZE); + + /* Break the unfortunate news. */ + pr_emerg("SUN4V NON-RECOVERABLE ERROR: Memory failed at %016lX\n", + addr); + pr_emerg("SUN4V NON-RECOVERABLE ERROR: Claiming %lu ages.\n", + page_cnt); + + while (page_cnt-- > 0) { + if (pfn_valid(addr >> PAGE_SHIFT)) + get_page(pfn_to_page(addr >> PAGE_SHIFT)); + addr += PAGE_SIZE; + } + } + info.si_signo = SIGKILL; + info.si_errno = 0; + info.si_trapno = 0; + force_sig_info(info.si_signo, &info, current); + + return true; + } + if (attrs & SUN4V_ERR_ATTRS_PIO) { + siginfo_t info; + + info.si_signo = SIGBUS; + info.si_code = BUS_ADRERR; + info.si_addr = (void __user *)sun4v_get_vaddr(regs); + force_sig_info(info.si_signo, &info, current); + + return true; + } + + /* Default to doing nothing */ + return false; +} + /* We run with %pil set to PIL_NORMAL_MAX and PSTATE_IE enabled in %pstate. * Log the event, clear the first word of the entry, and die. */ @@ -2075,6 +2142,12 @@ void sun4v_nonresum_error(struct pt_regs *regs, unsigned long offset) put_cpu(); + if (!(regs->tstate & TSTATE_PRIV) && + sun4v_nonresum_error_user_handled(regs, &local_copy)) { + /* DON'T PANIC: This userspace error was handled. */ + return; + } + #ifdef CONFIG_PCI /* Check for the special PCI poke sequence. */ if (pci_poke_in_progress && pci_poke_cpu == cpu) { diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c index 6ef688a1ef3e..7ff1b0c86a8e 100644 --- a/arch/x86/crypto/aesni-intel_glue.c +++ b/arch/x86/crypto/aesni-intel_glue.c @@ -1085,9 +1085,9 @@ static void aesni_free_simds(void) aesni_simd_skciphers[i]; i++) simd_skcipher_free(aesni_simd_skciphers[i]); - for (i = 0; i < ARRAY_SIZE(aesni_simd_skciphers2) && - aesni_simd_skciphers2[i].simd; i++) - simd_skcipher_free(aesni_simd_skciphers2[i].simd); + for (i = 0; i < ARRAY_SIZE(aesni_simd_skciphers2); i++) + if (aesni_simd_skciphers2[i].simd) + simd_skcipher_free(aesni_simd_skciphers2[i].simd); } static int __init aesni_init(void) @@ -1168,7 +1168,7 @@ static int __init aesni_init(void) simd = simd_skcipher_create_compat(algname, drvname, basename); err = PTR_ERR(simd); if (IS_ERR(simd)) - goto unregister_simds; + continue; aesni_simd_skciphers2[i].simd = simd; } diff --git a/arch/x86/events/intel/rapl.c b/arch/x86/events/intel/rapl.c index 17c3564d087a..22ef4f72cf32 100644 --- a/arch/x86/events/intel/rapl.c +++ b/arch/x86/events/intel/rapl.c @@ -161,7 +161,13 @@ static u64 rapl_timer_ms; static inline struct rapl_pmu *cpu_to_rapl_pmu(unsigned int cpu) { - return rapl_pmus->pmus[topology_logical_package_id(cpu)]; + unsigned int pkgid = topology_logical_package_id(cpu); + + /* + * The unsigned check also catches the '-1' return value for non + * existent mappings in the topology map. + */ + return pkgid < rapl_pmus->maxpkg ? rapl_pmus->pmus[pkgid] : NULL; } static inline u64 rapl_read_counter(struct perf_event *event) @@ -402,6 +408,8 @@ static int rapl_pmu_event_init(struct perf_event *event) /* must be done before validate_group */ pmu = cpu_to_rapl_pmu(event->cpu); + if (!pmu) + return -EINVAL; event->cpu = pmu->cpu; event->pmu_private = pmu; event->hw.event_base = msr; @@ -585,6 +593,20 @@ static int rapl_cpu_online(unsigned int cpu) struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu); int target; + if (!pmu) { + pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); + if (!pmu) + return -ENOMEM; + + raw_spin_lock_init(&pmu->lock); + INIT_LIST_HEAD(&pmu->active_list); + pmu->pmu = &rapl_pmus->pmu; + pmu->timer_interval = ms_to_ktime(rapl_timer_ms); + rapl_hrtimer_init(pmu); + + rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu; + } + /* * Check if there is an online cpu in the package which collects rapl * events already. @@ -598,27 +620,6 @@ static int rapl_cpu_online(unsigned int cpu) return 0; } -static int rapl_cpu_prepare(unsigned int cpu) -{ - struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu); - - if (pmu) - return 0; - - pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); - if (!pmu) - return -ENOMEM; - - raw_spin_lock_init(&pmu->lock); - INIT_LIST_HEAD(&pmu->active_list); - pmu->pmu = &rapl_pmus->pmu; - pmu->timer_interval = ms_to_ktime(rapl_timer_ms); - pmu->cpu = -1; - rapl_hrtimer_init(pmu); - rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu; - return 0; -} - static int rapl_check_hw_unit(bool apply_quirk) { u64 msr_rapl_power_unit_bits; @@ -803,29 +804,21 @@ static int __init rapl_pmu_init(void) /* * Install callbacks. Core will call them for each online cpu. */ - - ret = cpuhp_setup_state(CPUHP_PERF_X86_RAPL_PREP, "perf/x86/rapl:prepare", - rapl_cpu_prepare, NULL); - if (ret) - goto out; - ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_RAPL_ONLINE, "perf/x86/rapl:online", rapl_cpu_online, rapl_cpu_offline); if (ret) - goto out1; + goto out; ret = perf_pmu_register(&rapl_pmus->pmu, "power", -1); if (ret) - goto out2; + goto out1; rapl_advertise(); return 0; -out2: - cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE); out1: - cpuhp_remove_state(CPUHP_PERF_X86_RAPL_PREP); + cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE); out: pr_warn("Initialization failed (%d), disabled\n", ret); cleanup_rapl_pmus(); @@ -836,7 +829,6 @@ module_init(rapl_pmu_init); static void __exit intel_rapl_exit(void) { cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_RAPL_ONLINE); - cpuhp_remove_state_nocalls(CPUHP_PERF_X86_RAPL_PREP); perf_pmu_unregister(&rapl_pmus->pmu); cleanup_rapl_pmus(); } diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c index 8c4ccdc3a3f3..1ab45976474d 100644 --- a/arch/x86/events/intel/uncore.c +++ b/arch/x86/events/intel/uncore.c @@ -100,7 +100,13 @@ ssize_t uncore_event_show(struct kobject *kobj, struct intel_uncore_box *uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu) { - return pmu->boxes[topology_logical_package_id(cpu)]; + unsigned int pkgid = topology_logical_package_id(cpu); + + /* + * The unsigned check also catches the '-1' return value for non + * existent mappings in the topology map. + */ + return pkgid < max_packages ? pmu->boxes[pkgid] : NULL; } u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event) @@ -764,30 +770,6 @@ static void uncore_pmu_unregister(struct intel_uncore_pmu *pmu) pmu->registered = false; } -static void __uncore_exit_boxes(struct intel_uncore_type *type, int cpu) -{ - struct intel_uncore_pmu *pmu = type->pmus; - struct intel_uncore_box *box; - int i, pkg; - - if (pmu) { - pkg = topology_physical_package_id(cpu); - for (i = 0; i < type->num_boxes; i++, pmu++) { - box = pmu->boxes[pkg]; - if (box) - uncore_box_exit(box); - } - } -} - -static void uncore_exit_boxes(void *dummy) -{ - struct intel_uncore_type **types; - - for (types = uncore_msr_uncores; *types; types++) - __uncore_exit_boxes(*types++, smp_processor_id()); -} - static void uncore_free_boxes(struct intel_uncore_pmu *pmu) { int pkg; @@ -1058,86 +1040,6 @@ static void uncore_pci_exit(void) } } -static int uncore_cpu_dying(unsigned int cpu) -{ - struct intel_uncore_type *type, **types = uncore_msr_uncores; - struct intel_uncore_pmu *pmu; - struct intel_uncore_box *box; - int i, pkg; - - pkg = topology_logical_package_id(cpu); - for (; *types; types++) { - type = *types; - pmu = type->pmus; - for (i = 0; i < type->num_boxes; i++, pmu++) { - box = pmu->boxes[pkg]; - if (box && atomic_dec_return(&box->refcnt) == 0) - uncore_box_exit(box); - } - } - return 0; -} - -static int first_init; - -static int uncore_cpu_starting(unsigned int cpu) -{ - struct intel_uncore_type *type, **types = uncore_msr_uncores; - struct intel_uncore_pmu *pmu; - struct intel_uncore_box *box; - int i, pkg, ncpus = 1; - - if (first_init) { - /* - * On init we get the number of online cpus in the package - * and set refcount for all of them. - */ - ncpus = cpumask_weight(topology_core_cpumask(cpu)); - } - - pkg = topology_logical_package_id(cpu); - for (; *types; types++) { - type = *types; - pmu = type->pmus; - for (i = 0; i < type->num_boxes; i++, pmu++) { - box = pmu->boxes[pkg]; - if (!box) - continue; - /* The first cpu on a package activates the box */ - if (atomic_add_return(ncpus, &box->refcnt) == ncpus) - uncore_box_init(box); - } - } - - return 0; -} - -static int uncore_cpu_prepare(unsigned int cpu) -{ - struct intel_uncore_type *type, **types = uncore_msr_uncores; - struct intel_uncore_pmu *pmu; - struct intel_uncore_box *box; - int i, pkg; - - pkg = topology_logical_package_id(cpu); - for (; *types; types++) { - type = *types; - pmu = type->pmus; - for (i = 0; i < type->num_boxes; i++, pmu++) { - if (pmu->boxes[pkg]) - continue; - /* First cpu of a package allocates the box */ - box = uncore_alloc_box(type, cpu_to_node(cpu)); - if (!box) - return -ENOMEM; - box->pmu = pmu; - box->pkgid = pkg; - pmu->boxes[pkg] = box; - } - } - return 0; -} - static void uncore_change_type_ctx(struct intel_uncore_type *type, int old_cpu, int new_cpu) { @@ -1177,12 +1079,14 @@ static void uncore_change_context(struct intel_uncore_type **uncores, static int uncore_event_cpu_offline(unsigned int cpu) { - int target; + struct intel_uncore_type *type, **types = uncore_msr_uncores; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, pkg, target; /* Check if exiting cpu is used for collecting uncore events */ if (!cpumask_test_and_clear_cpu(cpu, &uncore_cpu_mask)) - return 0; - + goto unref; /* Find a new cpu to collect uncore events */ target = cpumask_any_but(topology_core_cpumask(cpu), cpu); @@ -1194,12 +1098,82 @@ static int uncore_event_cpu_offline(unsigned int cpu) uncore_change_context(uncore_msr_uncores, cpu, target); uncore_change_context(uncore_pci_uncores, cpu, target); + +unref: + /* Clear the references */ + pkg = topology_logical_package_id(cpu); + for (; *types; types++) { + type = *types; + pmu = type->pmus; + for (i = 0; i < type->num_boxes; i++, pmu++) { + box = pmu->boxes[pkg]; + if (box && atomic_dec_return(&box->refcnt) == 0) + uncore_box_exit(box); + } + } return 0; } +static int allocate_boxes(struct intel_uncore_type **types, + unsigned int pkg, unsigned int cpu) +{ + struct intel_uncore_box *box, *tmp; + struct intel_uncore_type *type; + struct intel_uncore_pmu *pmu; + LIST_HEAD(allocated); + int i; + + /* Try to allocate all required boxes */ + for (; *types; types++) { + type = *types; + pmu = type->pmus; + for (i = 0; i < type->num_boxes; i++, pmu++) { + if (pmu->boxes[pkg]) + continue; + box = uncore_alloc_box(type, cpu_to_node(cpu)); + if (!box) + goto cleanup; + box->pmu = pmu; + box->pkgid = pkg; + list_add(&box->active_list, &allocated); + } + } + /* Install them in the pmus */ + list_for_each_entry_safe(box, tmp, &allocated, active_list) { + list_del_init(&box->active_list); + box->pmu->boxes[pkg] = box; + } + return 0; + +cleanup: + list_for_each_entry_safe(box, tmp, &allocated, active_list) { + list_del_init(&box->active_list); + kfree(box); + } + return -ENOMEM; +} + static int uncore_event_cpu_online(unsigned int cpu) { - int target; + struct intel_uncore_type *type, **types = uncore_msr_uncores; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, ret, pkg, target; + + pkg = topology_logical_package_id(cpu); + ret = allocate_boxes(types, pkg, cpu); + if (ret) + return ret; + + for (; *types; types++) { + type = *types; + pmu = type->pmus; + for (i = 0; i < type->num_boxes; i++, pmu++) { + box = pmu->boxes[pkg]; + if (!box && atomic_inc_return(&box->refcnt) == 1) + uncore_box_init(box); + } + } /* * Check if there is an online cpu in the package @@ -1389,38 +1363,16 @@ static int __init intel_uncore_init(void) if (cret && pret) return -ENODEV; - /* - * Install callbacks. Core will call them for each online cpu. - * - * The first online cpu of each package allocates and takes - * the refcounts for all other online cpus in that package. - * If msrs are not enabled no allocation is required and - * uncore_cpu_prepare() is not called for each online cpu. - */ - if (!cret) { - ret = cpuhp_setup_state(CPUHP_PERF_X86_UNCORE_PREP, - "perf/x86/intel/uncore:prepare", - uncore_cpu_prepare, NULL); - if (ret) - goto err; - } else { - cpuhp_setup_state_nocalls(CPUHP_PERF_X86_UNCORE_PREP, - "perf/x86/intel/uncore:prepare", - uncore_cpu_prepare, NULL); - } - first_init = 1; - cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_STARTING, - "perf/x86/uncore:starting", - uncore_cpu_starting, uncore_cpu_dying); - first_init = 0; - cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE, - "perf/x86/uncore:online", - uncore_event_cpu_online, uncore_event_cpu_offline); + /* Install hotplug callbacks to setup the targets for each package */ + ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE, + "perf/x86/intel/uncore:online", + uncore_event_cpu_online, + uncore_event_cpu_offline); + if (ret) + goto err; return 0; err: - /* Undo box->init_box() */ - on_each_cpu_mask(&uncore_cpu_mask, uncore_exit_boxes, NULL, 1); uncore_types_exit(uncore_msr_uncores); uncore_pci_exit(); return ret; @@ -1429,9 +1381,7 @@ module_init(intel_uncore_init); static void __exit intel_uncore_exit(void) { - cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_UNCORE_ONLINE); - cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_UNCORE_STARTING); - cpuhp_remove_state_nocalls(CPUHP_PERF_X86_UNCORE_PREP); + cpuhp_remove_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE); uncore_types_exit(uncore_msr_uncores); uncore_pci_exit(); } diff --git a/arch/x86/include/asm/microcode.h b/arch/x86/include/asm/microcode.h index 38711df3bcb5..2266f864b747 100644 --- a/arch/x86/include/asm/microcode.h +++ b/arch/x86/include/asm/microcode.h @@ -140,6 +140,7 @@ extern void __init load_ucode_bsp(void); extern void load_ucode_ap(void); void reload_early_microcode(void); extern bool get_builtin_firmware(struct cpio_data *cd, const char *name); +extern bool initrd_gone; #else static inline int __init microcode_init(void) { return 0; }; static inline void __init load_ucode_bsp(void) { } diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index 1be64da0384e..e6cfe7ba2d65 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -104,6 +104,7 @@ struct cpuinfo_x86 { __u8 x86_phys_bits; /* CPUID returned core id bits: */ __u8 x86_coreid_bits; + __u8 cu_id; /* Max extended CPUID function supported: */ __u32 extended_cpuid_level; /* Maximum supported CPUID level, -1=no CPUID: */ diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 1e35dd06b090..bd6b8c270c24 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -1875,7 +1875,6 @@ static struct irq_chip ioapic_chip __read_mostly = { .irq_ack = irq_chip_ack_parent, .irq_eoi = ioapic_ack_level, .irq_set_affinity = ioapic_set_affinity, - .irq_retrigger = irq_chip_retrigger_hierarchy, .flags = IRQCHIP_SKIP_SET_WAKE, }; @@ -1887,7 +1886,6 @@ static struct irq_chip ioapic_ir_chip __read_mostly = { .irq_ack = irq_chip_ack_parent, .irq_eoi = ioapic_ir_ack_level, .irq_set_affinity = ioapic_set_affinity, - .irq_retrigger = irq_chip_retrigger_hierarchy, .flags = IRQCHIP_SKIP_SET_WAKE, }; @@ -2117,6 +2115,7 @@ static inline void __init check_timer(void) if (idx != -1 && irq_trigger(idx)) unmask_ioapic_irq(irq_get_chip_data(0)); } + irq_domain_deactivate_irq(irq_data); irq_domain_activate_irq(irq_data); if (timer_irq_works()) { if (disable_timer_pin_1 > 0) @@ -2138,6 +2137,7 @@ static inline void __init check_timer(void) * legacy devices should be connected to IO APIC #0 */ replace_pin_at_irq_node(data, node, apic1, pin1, apic2, pin2); + irq_domain_deactivate_irq(irq_data); irq_domain_activate_irq(irq_data); legacy_pic->unmask(0); if (timer_irq_works()) { diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 1d3167269a67..2b4cf04239b6 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -309,8 +309,22 @@ static void amd_get_topology(struct cpuinfo_x86 *c) /* get information required for multi-node processors */ if (boot_cpu_has(X86_FEATURE_TOPOEXT)) { + u32 eax, ebx, ecx, edx; - node_id = cpuid_ecx(0x8000001e) & 7; + cpuid(0x8000001e, &eax, &ebx, &ecx, &edx); + + node_id = ecx & 0xff; + smp_num_siblings = ((ebx >> 8) & 0xff) + 1; + + if (c->x86 == 0x15) + c->cu_id = ebx & 0xff; + + if (c->x86 >= 0x17) { + c->cpu_core_id = ebx & 0xff; + + if (smp_num_siblings > 1) + c->x86_max_cores /= smp_num_siblings; + } /* * We may have multiple LLCs if L3 caches exist, so check if we diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 9bab7a8a4293..ede03e849a8b 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1015,6 +1015,7 @@ static void identify_cpu(struct cpuinfo_x86 *c) c->x86_model_id[0] = '\0'; /* Unset */ c->x86_max_cores = 1; c->x86_coreid_bits = 0; + c->cu_id = 0xff; #ifdef CONFIG_X86_64 c->x86_clflush_size = 64; c->x86_phys_bits = 36; diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index 00ef43233e03..537c6647d84c 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -1373,20 +1373,15 @@ static unsigned long mce_adjust_timer_default(unsigned long interval) static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; -static void __restart_timer(struct timer_list *t, unsigned long interval) +static void __start_timer(struct timer_list *t, unsigned long interval) { unsigned long when = jiffies + interval; unsigned long flags; local_irq_save(flags); - if (timer_pending(t)) { - if (time_before(when, t->expires)) - mod_timer(t, when); - } else { - t->expires = round_jiffies(when); - add_timer_on(t, smp_processor_id()); - } + if (!timer_pending(t) || time_before(when, t->expires)) + mod_timer(t, round_jiffies(when)); local_irq_restore(flags); } @@ -1421,7 +1416,7 @@ static void mce_timer_fn(unsigned long data) done: __this_cpu_write(mce_next_interval, iv); - __restart_timer(t, iv); + __start_timer(t, iv); } /* @@ -1432,7 +1427,7 @@ void mce_timer_kick(unsigned long interval) struct timer_list *t = this_cpu_ptr(&mce_timer); unsigned long iv = __this_cpu_read(mce_next_interval); - __restart_timer(t, interval); + __start_timer(t, interval); if (interval < iv) __this_cpu_write(mce_next_interval, interval); @@ -1779,17 +1774,15 @@ static void __mcheck_cpu_clear_vendor(struct cpuinfo_x86 *c) } } -static void mce_start_timer(unsigned int cpu, struct timer_list *t) +static void mce_start_timer(struct timer_list *t) { unsigned long iv = check_interval * HZ; if (mca_cfg.ignore_ce || !iv) return; - per_cpu(mce_next_interval, cpu) = iv; - - t->expires = round_jiffies(jiffies + iv); - add_timer_on(t, cpu); + this_cpu_write(mce_next_interval, iv); + __start_timer(t, iv); } static void __mcheck_cpu_setup_timer(void) @@ -1806,7 +1799,7 @@ static void __mcheck_cpu_init_timer(void) unsigned int cpu = smp_processor_id(); setup_pinned_timer(t, mce_timer_fn, cpu); - mce_start_timer(cpu, t); + mce_start_timer(t); } /* Handle unconfigured int18 (should never happen) */ @@ -2566,7 +2559,7 @@ static int mce_cpu_dead(unsigned int cpu) static int mce_cpu_online(unsigned int cpu) { - struct timer_list *t = &per_cpu(mce_timer, cpu); + struct timer_list *t = this_cpu_ptr(&mce_timer); int ret; mce_device_create(cpu); @@ -2577,13 +2570,13 @@ static int mce_cpu_online(unsigned int cpu) return ret; } mce_reenable_cpu(); - mce_start_timer(cpu, t); + mce_start_timer(t); return 0; } static int mce_cpu_pre_down(unsigned int cpu) { - struct timer_list *t = &per_cpu(mce_timer, cpu); + struct timer_list *t = this_cpu_ptr(&mce_timer); mce_disable_cpu(); del_timer_sync(t); diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c index 6a31e2691f3a..079e81733a58 100644 --- a/arch/x86/kernel/cpu/microcode/amd.c +++ b/arch/x86/kernel/cpu/microcode/amd.c @@ -384,8 +384,9 @@ void load_ucode_amd_ap(unsigned int family) reget: if (!get_builtin_microcode(&cp, family)) { #ifdef CONFIG_BLK_DEV_INITRD - cp = find_cpio_data(ucode_path, (void *)initrd_start, - initrd_end - initrd_start, NULL); + if (!initrd_gone) + cp = find_cpio_data(ucode_path, (void *)initrd_start, + initrd_end - initrd_start, NULL); #endif if (!(cp.data && cp.size)) { /* diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c index 2af69d27da62..73102d932760 100644 --- a/arch/x86/kernel/cpu/microcode/core.c +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -46,6 +46,8 @@ static struct microcode_ops *microcode_ops; static bool dis_ucode_ldr = true; +bool initrd_gone; + LIST_HEAD(microcode_cache); /* @@ -190,21 +192,24 @@ void load_ucode_ap(void) static int __init save_microcode_in_initrd(void) { struct cpuinfo_x86 *c = &boot_cpu_data; + int ret = -EINVAL; switch (c->x86_vendor) { case X86_VENDOR_INTEL: if (c->x86 >= 6) - return save_microcode_in_initrd_intel(); + ret = save_microcode_in_initrd_intel(); break; case X86_VENDOR_AMD: if (c->x86 >= 0x10) - return save_microcode_in_initrd_amd(c->x86); + ret = save_microcode_in_initrd_amd(c->x86); break; default: break; } - return -EINVAL; + initrd_gone = true; + + return ret; } struct cpio_data find_microcode_in_initrd(const char *path, bool use_pa) @@ -247,9 +252,16 @@ struct cpio_data find_microcode_in_initrd(const char *path, bool use_pa) * has the virtual address of the beginning of the initrd. It also * possibly relocates the ramdisk. In either case, initrd_start contains * the updated address so use that instead. + * + * initrd_gone is for the hotplug case where we've thrown out initrd + * already. */ - if (!use_pa && initrd_start) - start = initrd_start; + if (!use_pa) { + if (initrd_gone) + return (struct cpio_data){ NULL, 0, "" }; + if (initrd_start) + start = initrd_start; + } return find_cpio_data(path, (void *)start, size, NULL); #else /* !CONFIG_BLK_DEV_INITRD */ diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c index 3f329b74e040..8325d8a09ab0 100644 --- a/arch/x86/kernel/cpu/microcode/intel.c +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -41,7 +41,7 @@ static const char ucode_path[] = "kernel/x86/microcode/GenuineIntel.bin"; -/* Current microcode patch used in early patching */ +/* Current microcode patch used in early patching on the APs. */ struct microcode_intel *intel_ucode_patch; static inline bool cpu_signatures_match(unsigned int s1, unsigned int p1, @@ -607,12 +607,6 @@ int __init save_microcode_in_initrd_intel(void) struct ucode_cpu_info uci; struct cpio_data cp; - /* - * AP loading didn't find any microcode patch, no need to save anything. - */ - if (!intel_ucode_patch || IS_ERR(intel_ucode_patch)) - return 0; - if (!load_builtin_intel_microcode(&cp)) cp = find_microcode_in_initrd(ucode_path, false); @@ -628,7 +622,6 @@ int __init save_microcode_in_initrd_intel(void) return 0; } - /* * @res_patch, output: a pointer to the patch we found. */ diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c index e4e97a5355ce..de7234401275 100644 --- a/arch/x86/kernel/fpu/core.c +++ b/arch/x86/kernel/fpu/core.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include @@ -183,7 +184,8 @@ void fpstate_init(union fpregs_state *state) * it will #GP. Make sure it is replaced after the memset(). */ if (static_cpu_has(X86_FEATURE_XSAVES)) - state->xsave.header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT; + state->xsave.header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | + xfeatures_mask; if (static_cpu_has(X86_FEATURE_FXSR)) fpstate_init_fxstate(&state->fxsave); diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 85e87b46c318..dc6ba5bda9fc 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -352,6 +352,7 @@ static int hpet_resume(struct clock_event_device *evt, int timer) } else { struct hpet_dev *hdev = EVT_TO_HPET_DEV(evt); + irq_domain_deactivate_irq(irq_get_irq_data(hdev->irq)); irq_domain_activate_irq(irq_get_irq_data(hdev->irq)); disable_irq(hdev->irq); irq_set_affinity(hdev->irq, cpumask_of(hdev->cpu)); diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 46732dc3b73c..99b920d0e516 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -433,9 +433,15 @@ static bool match_smt(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o) int cpu1 = c->cpu_index, cpu2 = o->cpu_index; if (c->phys_proc_id == o->phys_proc_id && - per_cpu(cpu_llc_id, cpu1) == per_cpu(cpu_llc_id, cpu2) && - c->cpu_core_id == o->cpu_core_id) - return topology_sane(c, o, "smt"); + per_cpu(cpu_llc_id, cpu1) == per_cpu(cpu_llc_id, cpu2)) { + if (c->cpu_core_id == o->cpu_core_id) + return topology_sane(c, o, "smt"); + + if ((c->cu_id != 0xff) && + (o->cu_id != 0xff) && + (c->cu_id == o->cu_id)) + return topology_sane(c, o, "smt"); + } } else if (c->phys_proc_id == o->phys_proc_id && c->cpu_core_id == o->cpu_core_id) { diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index e41af597aed8..37e7cf544e51 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -1356,6 +1356,9 @@ void __init tsc_init(void) (unsigned long)cpu_khz / 1000, (unsigned long)cpu_khz % 1000); + /* Sanitize TSC ADJUST before cyc2ns gets initialized */ + tsc_store_and_check_tsc_adjust(true); + /* * Secondary CPUs do not run through tsc_init(), so set up * all the scale factors for all CPUs, assuming the same @@ -1386,8 +1389,6 @@ void __init tsc_init(void) if (unsynchronized_tsc()) mark_tsc_unstable("TSCs unsynchronized"); - else - tsc_store_and_check_tsc_adjust(true); check_system_tsc_reliable(); diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index d0db011051a5..728f75378475 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -286,13 +286,6 @@ void check_tsc_sync_source(int cpu) if (unsynchronized_tsc()) return; - if (tsc_clocksource_reliable) { - if (cpu == (nr_cpu_ids-1) || system_state != SYSTEM_BOOTING) - pr_info( - "Skipped synchronization checks as TSC is reliable.\n"); - return; - } - /* * Set the maximum number of test runs to * 1 if the CPU does not provide the TSC_ADJUST MSR @@ -380,14 +373,19 @@ void check_tsc_sync_target(void) int cpus = 2; /* Also aborts if there is no TSC. */ - if (unsynchronized_tsc() || tsc_clocksource_reliable) + if (unsynchronized_tsc()) return; /* * Store, verify and sanitize the TSC adjust register. If * successful skip the test. + * + * The test is also skipped when the TSC is marked reliable. This + * is true for SoCs which have no fallback clocksource. On these + * SoCs the TSC is frequency synchronized, but still the TSC ADJUST + * register might have been wreckaged by the BIOS.. */ - if (tsc_store_and_check_tsc_adjust(false)) { + if (tsc_store_and_check_tsc_adjust(false) || tsc_clocksource_reliable) { atomic_inc(&skip_test); return; } diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index d153be8929a6..e52c9088660f 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3182,6 +3182,7 @@ static void fill_xsave(u8 *dest, struct kvm_vcpu *vcpu) memcpy(dest, xsave, XSAVE_HDR_OFFSET); /* Set XSTATE_BV */ + xstate_bv &= vcpu->arch.guest_supported_xcr0 | XFEATURE_MASK_FPSSE; *(u64 *)(dest + XSAVE_HDR_OFFSET) = xstate_bv; /* diff --git a/arch/x86/mm/dump_pagetables.c b/arch/x86/mm/dump_pagetables.c index ea9c49adaa1f..8aa6bea1cd6c 100644 --- a/arch/x86/mm/dump_pagetables.c +++ b/arch/x86/mm/dump_pagetables.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -406,6 +407,7 @@ static void ptdump_walk_pgd_level_core(struct seq_file *m, pgd_t *pgd, } else note_page(m, &st, __pgprot(0), 1); + cond_resched(); start++; } diff --git a/arch/x86/platform/efi/efi_64.c b/arch/x86/platform/efi/efi_64.c index 319148bd4b05..2f25a363068c 100644 --- a/arch/x86/platform/efi/efi_64.c +++ b/arch/x86/platform/efi/efi_64.c @@ -268,6 +268,22 @@ int __init efi_setup_page_tables(unsigned long pa_memmap, unsigned num_pages) efi_scratch.use_pgd = true; + /* + * Certain firmware versions are way too sentimential and still believe + * they are exclusive and unquestionable owners of the first physical page, + * even though they explicitly mark it as EFI_CONVENTIONAL_MEMORY + * (but then write-access it later during SetVirtualAddressMap()). + * + * Create a 1:1 mapping for this page, to avoid triple faults during early + * boot with such firmware. We are free to hand this page to the BIOS, + * as trim_bios_range() will reserve the first page and isolate it away + * from memory allocators anyway. + */ + if (kernel_map_pages_in_pgd(pgd, 0x0, 0x0, 1, _PAGE_RW)) { + pr_err("Failed to create 1:1 mapping for the first page!\n"); + return 1; + } + /* * When making calls to the firmware everything needs to be 1:1 * mapped and addressable with 32-bit pointers. Map the kernel diff --git a/arch/xtensa/kernel/setup.c b/arch/xtensa/kernel/setup.c index 848e8568fb3c..8fd4be610607 100644 --- a/arch/xtensa/kernel/setup.c +++ b/arch/xtensa/kernel/setup.c @@ -419,7 +419,7 @@ subsys_initcall(topology_init); void cpu_reset(void) { -#if XCHAL_HAVE_PTP_MMU +#if XCHAL_HAVE_PTP_MMU && IS_ENABLED(CONFIG_MMU) local_irq_disable(); /* * We have full MMU: all autoload ways, ways 7, 8 and 9 of DTLB must diff --git a/block/blk-lib.c b/block/blk-lib.c index f8c82a9b4012..ed1e78e24db0 100644 --- a/block/blk-lib.c +++ b/block/blk-lib.c @@ -306,11 +306,6 @@ int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, if (ret == 0 || (ret && ret != -EOPNOTSUPP)) goto out; - ret = __blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, - ZERO_PAGE(0), biop); - if (ret == 0 || (ret && ret != -EOPNOTSUPP)) - goto out; - ret = 0; while (nr_sects != 0) { bio = next_bio(bio, min(nr_sects, (sector_t)BIO_MAX_PAGES), @@ -369,6 +364,10 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, return 0; } + if (!blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, + ZERO_PAGE(0))) + return 0; + blk_start_plug(&plug); ret = __blkdev_issue_zeroout(bdev, sector, nr_sects, gfp_mask, &bio, discard); diff --git a/crypto/algapi.c b/crypto/algapi.c index df939b54b09f..1fad2a6b3bbb 100644 --- a/crypto/algapi.c +++ b/crypto/algapi.c @@ -356,6 +356,7 @@ int crypto_register_alg(struct crypto_alg *alg) struct crypto_larval *larval; int err; + alg->cra_flags &= ~CRYPTO_ALG_DEAD; err = crypto_check_alg(alg); if (err) return err; diff --git a/crypto/algif_aead.c b/crypto/algif_aead.c index f849311e9fd4..533265f110e0 100644 --- a/crypto/algif_aead.c +++ b/crypto/algif_aead.c @@ -661,9 +661,9 @@ static int aead_recvmsg_sync(struct socket *sock, struct msghdr *msg, int flags) unlock: list_for_each_entry_safe(rsgl, tmp, &ctx->list, list) { af_alg_free_sg(&rsgl->sgl); + list_del(&rsgl->list); if (rsgl != &ctx->first_rsgl) sock_kfree_s(sk, rsgl, sizeof(*rsgl)); - list_del(&rsgl->list); } INIT_LIST_HEAD(&ctx->list); aead_wmem_wakeup(sk); diff --git a/drivers/acpi/acpica/tbdata.c b/drivers/acpi/acpica/tbdata.c index 82b0b5710979..b0399e8f6d27 100644 --- a/drivers/acpi/acpica/tbdata.c +++ b/drivers/acpi/acpica/tbdata.c @@ -852,23 +852,18 @@ acpi_tb_install_and_load_table(acpi_physical_address address, ACPI_FUNCTION_TRACE(tb_install_and_load_table); - (void)acpi_ut_acquire_mutex(ACPI_MTX_TABLES); - /* Install the table and load it into the namespace */ status = acpi_tb_install_standard_table(address, flags, TRUE, override, &i); if (ACPI_FAILURE(status)) { - goto unlock_and_exit; + goto exit; } - (void)acpi_ut_release_mutex(ACPI_MTX_TABLES); status = acpi_tb_load_table(i, acpi_gbl_root_node); - (void)acpi_ut_acquire_mutex(ACPI_MTX_TABLES); -unlock_and_exit: +exit: *table_index = i; - (void)acpi_ut_release_mutex(ACPI_MTX_TABLES); return_ACPI_STATUS(status); } diff --git a/drivers/acpi/acpica/tbinstal.c b/drivers/acpi/acpica/tbinstal.c index 5fdf251a9f97..01e1b3d63fc0 100644 --- a/drivers/acpi/acpica/tbinstal.c +++ b/drivers/acpi/acpica/tbinstal.c @@ -217,6 +217,10 @@ acpi_tb_install_standard_table(acpi_physical_address address, goto release_and_exit; } + /* Acquire the table lock */ + + (void)acpi_ut_acquire_mutex(ACPI_MTX_TABLES); + if (reload) { /* * Validate the incoming table signature. @@ -244,7 +248,7 @@ acpi_tb_install_standard_table(acpi_physical_address address, new_table_desc.signature.integer)); status = AE_BAD_SIGNATURE; - goto release_and_exit; + goto unlock_and_exit; } /* Check if table is already registered */ @@ -279,7 +283,7 @@ acpi_tb_install_standard_table(acpi_physical_address address, /* Table is still loaded, this is an error */ status = AE_ALREADY_EXISTS; - goto release_and_exit; + goto unlock_and_exit; } else { /* * Table was unloaded, allow it to be reloaded. @@ -290,6 +294,7 @@ acpi_tb_install_standard_table(acpi_physical_address address, * indicate the re-installation. */ acpi_tb_uninstall_table(&new_table_desc); + (void)acpi_ut_release_mutex(ACPI_MTX_TABLES); *table_index = i; return_ACPI_STATUS(AE_OK); } @@ -303,11 +308,19 @@ acpi_tb_install_standard_table(acpi_physical_address address, /* Invoke table handler if present */ + (void)acpi_ut_release_mutex(ACPI_MTX_TABLES); if (acpi_gbl_table_handler) { (void)acpi_gbl_table_handler(ACPI_TABLE_EVENT_INSTALL, new_table_desc.pointer, acpi_gbl_table_handler_context); } + (void)acpi_ut_acquire_mutex(ACPI_MTX_TABLES); + +unlock_and_exit: + + /* Release the table lock */ + + (void)acpi_ut_release_mutex(ACPI_MTX_TABLES); release_and_exit: diff --git a/drivers/acpi/nfit/core.c b/drivers/acpi/nfit/core.c index 2f82b8eba360..7361d00818e2 100644 --- a/drivers/acpi/nfit/core.c +++ b/drivers/acpi/nfit/core.c @@ -2704,6 +2704,7 @@ static int acpi_nfit_flush_probe(struct nvdimm_bus_descriptor *nd_desc) struct acpi_nfit_desc *acpi_desc = to_acpi_nfit_desc(nd_desc); struct device *dev = acpi_desc->dev; struct acpi_nfit_flush_work flush; + int rc; /* bounce the device lock to flush acpi_nfit_add / acpi_nfit_notify */ device_lock(dev); @@ -2716,7 +2717,10 @@ static int acpi_nfit_flush_probe(struct nvdimm_bus_descriptor *nd_desc) INIT_WORK_ONSTACK(&flush.work, flush_probe); COMPLETION_INITIALIZER_ONSTACK(flush.cmp); queue_work(nfit_wq, &flush.work); - return wait_for_completion_interruptible(&flush.cmp); + + rc = wait_for_completion_interruptible(&flush.cmp); + cancel_work_sync(&flush.work); + return rc; } static int acpi_nfit_clear_to_send(struct nvdimm_bus_descriptor *nd_desc, diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index 9b6cebe227a0..54abb26b7366 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -674,14 +674,6 @@ static void acpi_sleep_suspend_setup(void) if (acpi_sleep_state_supported(i)) sleep_states[i] = 1; - /* - * Use suspend-to-idle by default if ACPI_FADT_LOW_POWER_S0 is set and - * the default suspend mode was not selected from the command line. - */ - if (acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0 && - mem_sleep_default > PM_SUSPEND_MEM) - mem_sleep_default = PM_SUSPEND_FREEZE; - suspend_set_ops(old_suspend_ordering ? &acpi_suspend_ops_old : &acpi_suspend_ops); freeze_set_ops(&acpi_freeze_ops); diff --git a/drivers/acpi/video_detect.c b/drivers/acpi/video_detect.c index 02ded25c82e4..7f48156cbc0c 100644 --- a/drivers/acpi/video_detect.c +++ b/drivers/acpi/video_detect.c @@ -305,17 +305,6 @@ static const struct dmi_system_id video_detect_dmi_table[] = { DMI_MATCH(DMI_PRODUCT_NAME, "Dell System XPS L702X"), }, }, - { - /* https://bugzilla.redhat.com/show_bug.cgi?id=1204476 */ - /* https://bugs.launchpad.net/ubuntu/+source/linux-lts-trusty/+bug/1416940 */ - .callback = video_detect_force_native, - .ident = "HP Pavilion dv6", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"), - DMI_MATCH(DMI_PRODUCT_NAME, "HP Pavilion dv6 Notebook PC"), - }, - }, - { }, }; diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 9cd0a2d41816..c2d3785ec227 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -1702,6 +1702,8 @@ unsigned ata_exec_internal_sg(struct ata_device *dev, if (qc->err_mask & ~AC_ERR_OTHER) qc->err_mask &= ~AC_ERR_OTHER; + } else if (qc->tf.command == ATA_CMD_REQ_SENSE_DATA) { + qc->result_tf.command |= ATA_SENSE; } /* finish up */ @@ -4356,10 +4358,10 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = { { "ST380013AS", "3.20", ATA_HORKAGE_MAX_SEC_1024 }, /* - * Device times out with higher max sects. + * These devices time out with higher max sects. * https://bugzilla.kernel.org/show_bug.cgi?id=121671 */ - { "LITEON CX1-JB256-HP", NULL, ATA_HORKAGE_MAX_SEC_1024 }, + { "LITEON CX1-JB*-HP", NULL, ATA_HORKAGE_MAX_SEC_1024 }, /* Devices we expect to fail diagnostics */ diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c index 823e938c9a78..2f32782cea6d 100644 --- a/drivers/ata/sata_mv.c +++ b/drivers/ata/sata_mv.c @@ -4132,6 +4132,9 @@ static int mv_platform_probe(struct platform_device *pdev) host->iomap = NULL; hpriv->base = devm_ioremap(&pdev->dev, res->start, resource_size(res)); + if (!hpriv->base) + return -ENOMEM; + hpriv->base -= SATAHC0_REG_BASE; hpriv->clk = clk_get(&pdev->dev, NULL); diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c index 4497d263209f..ac350c518e0c 100644 --- a/drivers/base/firmware_class.c +++ b/drivers/base/firmware_class.c @@ -558,9 +558,6 @@ static void fw_load_abort(struct firmware_priv *fw_priv) struct firmware_buf *buf = fw_priv->buf; __fw_load_abort(buf); - - /* avoid user action after loading abort */ - fw_priv->buf = NULL; } static LIST_HEAD(pending_fw_head); @@ -713,7 +710,7 @@ static ssize_t firmware_loading_store(struct device *dev, mutex_lock(&fw_lock); fw_buf = fw_priv->buf; - if (!fw_buf) + if (fw_state_is_aborted(&fw_buf->fw_st)) goto out; switch (loading) { diff --git a/drivers/base/memory.c b/drivers/base/memory.c index dacb6a8418aa..fa26ffd25fa6 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -389,33 +389,33 @@ static ssize_t show_valid_zones(struct device *dev, { struct memory_block *mem = to_memory_block(dev); unsigned long start_pfn, end_pfn; + unsigned long valid_start, valid_end, valid_pages; unsigned long nr_pages = PAGES_PER_SECTION * sections_per_block; - struct page *first_page; struct zone *zone; int zone_shift = 0; start_pfn = section_nr_to_pfn(mem->start_section_nr); end_pfn = start_pfn + nr_pages; - first_page = pfn_to_page(start_pfn); /* The block contains more than one zone can not be offlined. */ - if (!test_pages_in_a_zone(start_pfn, end_pfn)) + if (!test_pages_in_a_zone(start_pfn, end_pfn, &valid_start, &valid_end)) return sprintf(buf, "none\n"); - zone = page_zone(first_page); + zone = page_zone(pfn_to_page(valid_start)); + valid_pages = valid_end - valid_start; /* MMOP_ONLINE_KEEP */ sprintf(buf, "%s", zone->name); /* MMOP_ONLINE_KERNEL */ - zone_can_shift(start_pfn, nr_pages, ZONE_NORMAL, &zone_shift); + zone_can_shift(valid_start, valid_pages, ZONE_NORMAL, &zone_shift); if (zone_shift) { strcat(buf, " "); strcat(buf, (zone + zone_shift)->name); } /* MMOP_ONLINE_MOVABLE */ - zone_can_shift(start_pfn, nr_pages, ZONE_MOVABLE, &zone_shift); + zone_can_shift(valid_start, valid_pages, ZONE_MOVABLE, &zone_shift); if (zone_shift) { strcat(buf, " "); strcat(buf, (zone + zone_shift)->name); diff --git a/drivers/base/power/runtime.c b/drivers/base/power/runtime.c index 872eac4cb1df..a14fac6a01d3 100644 --- a/drivers/base/power/runtime.c +++ b/drivers/base/power/runtime.c @@ -966,13 +966,13 @@ int __pm_runtime_idle(struct device *dev, int rpmflags) unsigned long flags; int retval; - might_sleep_if(!(rpmflags & RPM_ASYNC) && !dev->power.irq_safe); - if (rpmflags & RPM_GET_PUT) { if (!atomic_dec_and_test(&dev->power.usage_count)) return 0; } + might_sleep_if(!(rpmflags & RPM_ASYNC) && !dev->power.irq_safe); + spin_lock_irqsave(&dev->power.lock, flags); retval = rpm_idle(dev, rpmflags); spin_unlock_irqrestore(&dev->power.lock, flags); @@ -998,13 +998,13 @@ int __pm_runtime_suspend(struct device *dev, int rpmflags) unsigned long flags; int retval; - might_sleep_if(!(rpmflags & RPM_ASYNC) && !dev->power.irq_safe); - if (rpmflags & RPM_GET_PUT) { if (!atomic_dec_and_test(&dev->power.usage_count)) return 0; } + might_sleep_if(!(rpmflags & RPM_ASYNC) && !dev->power.irq_safe); + spin_lock_irqsave(&dev->power.lock, flags); retval = rpm_suspend(dev, rpmflags); spin_unlock_irqrestore(&dev->power.lock, flags); @@ -1029,7 +1029,8 @@ int __pm_runtime_resume(struct device *dev, int rpmflags) unsigned long flags; int retval; - might_sleep_if(!(rpmflags & RPM_ASYNC) && !dev->power.irq_safe); + might_sleep_if(!(rpmflags & RPM_ASYNC) && !dev->power.irq_safe && + dev->power.runtime_status != RPM_ACTIVE); if (rpmflags & RPM_GET_PUT) atomic_inc(&dev->power.usage_count); diff --git a/drivers/bcma/bcma_private.h b/drivers/bcma/bcma_private.h index f642c4264c27..168fa175d65a 100644 --- a/drivers/bcma/bcma_private.h +++ b/drivers/bcma/bcma_private.h @@ -45,6 +45,9 @@ int bcma_sprom_get(struct bcma_bus *bus); void bcma_core_chipcommon_early_init(struct bcma_drv_cc *cc); void bcma_core_chipcommon_init(struct bcma_drv_cc *cc); void bcma_chipco_bcm4331_ext_pa_lines_ctl(struct bcma_drv_cc *cc, bool enable); +#ifdef CONFIG_BCMA_DRIVER_MIPS +void bcma_chipco_serial_init(struct bcma_drv_cc *cc); +#endif /* CONFIG_BCMA_DRIVER_MIPS */ /* driver_chipcommon_b.c */ int bcma_core_chipcommon_b_init(struct bcma_drv_cc_b *ccb); diff --git a/drivers/bcma/driver_chipcommon.c b/drivers/bcma/driver_chipcommon.c index b4f6520e74f0..62f5bfa5065d 100644 --- a/drivers/bcma/driver_chipcommon.c +++ b/drivers/bcma/driver_chipcommon.c @@ -15,8 +15,6 @@ #include #include -static void bcma_chipco_serial_init(struct bcma_drv_cc *cc); - static inline u32 bcma_cc_write32_masked(struct bcma_drv_cc *cc, u16 offset, u32 mask, u32 value) { @@ -186,9 +184,6 @@ void bcma_core_chipcommon_early_init(struct bcma_drv_cc *cc) if (cc->capabilities & BCMA_CC_CAP_PMU) bcma_pmu_early_init(cc); - if (IS_BUILTIN(CONFIG_BCM47XX) && bus->hosttype == BCMA_HOSTTYPE_SOC) - bcma_chipco_serial_init(cc); - if (bus->hosttype == BCMA_HOSTTYPE_SOC) bcma_core_chipcommon_flash_detect(cc); @@ -378,9 +373,9 @@ u32 bcma_chipco_gpio_pulldown(struct bcma_drv_cc *cc, u32 mask, u32 value) return res; } -static void bcma_chipco_serial_init(struct bcma_drv_cc *cc) +#ifdef CONFIG_BCMA_DRIVER_MIPS +void bcma_chipco_serial_init(struct bcma_drv_cc *cc) { -#if IS_BUILTIN(CONFIG_BCM47XX) unsigned int irq; u32 baud_base; u32 i; @@ -422,5 +417,5 @@ static void bcma_chipco_serial_init(struct bcma_drv_cc *cc) ports[i].baud_base = baud_base; ports[i].reg_shift = 0; } -#endif /* CONFIG_BCM47XX */ } +#endif /* CONFIG_BCMA_DRIVER_MIPS */ diff --git a/drivers/bcma/driver_mips.c b/drivers/bcma/driver_mips.c index 96f171328200..89af807cf29c 100644 --- a/drivers/bcma/driver_mips.c +++ b/drivers/bcma/driver_mips.c @@ -278,9 +278,12 @@ static void bcma_core_mips_nvram_init(struct bcma_drv_mips *mcore) void bcma_core_mips_early_init(struct bcma_drv_mips *mcore) { + struct bcma_bus *bus = mcore->core->bus; + if (mcore->early_setup_done) return; + bcma_chipco_serial_init(&bus->drv_cc); bcma_core_mips_nvram_init(mcore); mcore->early_setup_done = true; diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index b2bdfa81f929..265f1a7072e9 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -197,13 +197,13 @@ struct blkfront_info /* Number of pages per ring buffer. */ unsigned int nr_ring_pages; struct request_queue *rq; - unsigned int feature_flush; - unsigned int feature_fua; + unsigned int feature_flush:1; + unsigned int feature_fua:1; unsigned int feature_discard:1; unsigned int feature_secdiscard:1; + unsigned int feature_persistent:1; unsigned int discard_granularity; unsigned int discard_alignment; - unsigned int feature_persistent:1; /* Number of 4KB segments handled */ unsigned int max_indirect_segments; int is_ready; @@ -2223,7 +2223,7 @@ static int blkfront_setup_indirect(struct blkfront_ring_info *rinfo) } else grants = info->max_indirect_segments; - psegs = grants / GRANTS_PER_PSEG; + psegs = DIV_ROUND_UP(grants, GRANTS_PER_PSEG); err = fill_grant_buffer(rinfo, (grants + INDIRECT_GREFS(grants)) * BLK_RING_SIZE(info)); @@ -2323,13 +2323,16 @@ static void blkfront_gather_backend_features(struct blkfront_info *info) blkfront_setup_discard(info); info->feature_persistent = - xenbus_read_unsigned(info->xbdev->otherend, - "feature-persistent", 0); + !!xenbus_read_unsigned(info->xbdev->otherend, + "feature-persistent", 0); indirect_segments = xenbus_read_unsigned(info->xbdev->otherend, "feature-max-indirect-segments", 0); - info->max_indirect_segments = min(indirect_segments, - xen_blkif_max_segments); + if (indirect_segments > xen_blkif_max_segments) + indirect_segments = xen_blkif_max_segments; + if (indirect_segments <= BLKIF_MAX_SEGMENTS_PER_REQUEST) + indirect_segments = 0; + info->max_indirect_segments = indirect_segments; } /* @@ -2652,6 +2655,9 @@ static int __init xlblk_init(void) if (!xen_domain()) return -ENODEV; + if (xen_blkif_max_segments < BLKIF_MAX_SEGMENTS_PER_REQUEST) + xen_blkif_max_segments = BLKIF_MAX_SEGMENTS_PER_REQUEST; + if (xen_blkif_max_ring_order > XENBUS_MAX_RING_GRANT_ORDER) { pr_info("Invalid max_ring_order (%d), will use default max: %d.\n", xen_blkif_max_ring_order, XENBUS_MAX_RING_GRANT_ORDER); diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c index 6ce5ce8be2f2..87fba424817e 100644 --- a/drivers/char/hw_random/core.c +++ b/drivers/char/hw_random/core.c @@ -92,7 +92,6 @@ static void add_early_randomness(struct hwrng *rng) mutex_unlock(&reading_mutex); if (bytes_read > 0) add_device_randomness(rng_buffer, bytes_read); - memset(rng_buffer, 0, size); } static inline void cleanup_rng(struct kref *kref) @@ -288,7 +287,6 @@ static ssize_t rng_dev_read(struct file *filp, char __user *buf, } } out: - memset(rng_buffer, 0, rng_buffer_size()); return ret ? : err; out_unlock_reading: @@ -427,7 +425,6 @@ static int hwrng_fillfn(void *unused) /* Outside lock, sure, but y'know: randomness. */ add_hwgenerator_randomness((void *)rng_fillbuf, rc, rc * current_quality * 8 >> 10); - memset(rng_fillbuf, 0, rng_buffer_size()); } hwrng_fill = NULL; return 0; diff --git a/drivers/cpufreq/brcmstb-avs-cpufreq.c b/drivers/cpufreq/brcmstb-avs-cpufreq.c index 4fda623e55bb..c94360671f41 100644 --- a/drivers/cpufreq/brcmstb-avs-cpufreq.c +++ b/drivers/cpufreq/brcmstb-avs-cpufreq.c @@ -784,8 +784,19 @@ static int brcm_avs_target_index(struct cpufreq_policy *policy, static int brcm_avs_suspend(struct cpufreq_policy *policy) { struct private_data *priv = policy->driver_data; + int ret; - return brcm_avs_get_pmap(priv, &priv->pmap); + ret = brcm_avs_get_pmap(priv, &priv->pmap); + if (ret) + return ret; + + /* + * We can't use the P-state returned by brcm_avs_get_pmap(), since + * that's the initial P-state from when the P-map was downloaded to the + * AVS co-processor, not necessarily the P-state we are running at now. + * So, we get the current P-state explicitly. + */ + return brcm_avs_get_pstate(priv, &priv->pmap.state); } static int brcm_avs_resume(struct cpufreq_policy *policy) @@ -954,9 +965,9 @@ static ssize_t show_brcm_avs_pmap(struct cpufreq_policy *policy, char *buf) brcm_avs_parse_p1(pmap.p1, &mdiv_p0, &pdiv, &ndiv); brcm_avs_parse_p2(pmap.p2, &mdiv_p1, &mdiv_p2, &mdiv_p3, &mdiv_p4); - return sprintf(buf, "0x%08x 0x%08x %u %u %u %u %u %u %u\n", + return sprintf(buf, "0x%08x 0x%08x %u %u %u %u %u %u %u %u %u\n", pmap.p1, pmap.p2, ndiv, pdiv, mdiv_p0, mdiv_p1, mdiv_p2, - mdiv_p3, mdiv_p4); + mdiv_p3, mdiv_p4, pmap.mode, pmap.state); } static ssize_t show_brcm_avs_voltage(struct cpufreq_policy *policy, char *buf) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index f91c25718d16..50bd6d987fc3 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1235,6 +1235,25 @@ static void intel_pstate_hwp_enable(struct cpudata *cpudata) cpudata->epp_default = intel_pstate_get_epp(cpudata, 0); } +#define MSR_IA32_POWER_CTL_BIT_EE 19 + +/* Disable energy efficiency optimization */ +static void intel_pstate_disable_ee(int cpu) +{ + u64 power_ctl; + int ret; + + ret = rdmsrl_on_cpu(cpu, MSR_IA32_POWER_CTL, &power_ctl); + if (ret) + return; + + if (!(power_ctl & BIT(MSR_IA32_POWER_CTL_BIT_EE))) { + pr_info("Disabling energy efficiency optimization\n"); + power_ctl |= BIT(MSR_IA32_POWER_CTL_BIT_EE); + wrmsrl_on_cpu(cpu, MSR_IA32_POWER_CTL, power_ctl); + } +} + static int atom_get_min_pstate(void) { u64 value; @@ -1845,6 +1864,11 @@ static const struct x86_cpu_id intel_pstate_cpu_oob_ids[] __initconst = { {} }; +static const struct x86_cpu_id intel_pstate_cpu_ee_disable_ids[] = { + ICPU(INTEL_FAM6_KABYLAKE_DESKTOP, core_params), + {} +}; + static int intel_pstate_init_cpu(unsigned int cpunum) { struct cpudata *cpu; @@ -1875,6 +1899,12 @@ static int intel_pstate_init_cpu(unsigned int cpunum) cpu->cpu = cpunum; if (hwp_active) { + const struct x86_cpu_id *id; + + id = x86_match_cpu(intel_pstate_cpu_ee_disable_ids); + if (id) + intel_pstate_disable_ee(cpunum); + intel_pstate_hwp_enable(cpu); pid_params.sample_rate_ms = 50; pid_params.sample_rate_ns = 50 * NSEC_PER_MSEC; @@ -2005,7 +2035,8 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy) limits = &performance_limits; perf_limits = limits; } - if (policy->max >= policy->cpuinfo.max_freq) { + if (policy->max >= policy->cpuinfo.max_freq && + !limits->no_turbo) { pr_debug("set performance\n"); intel_pstate_set_performance_limits(perf_limits); goto out; @@ -2047,6 +2078,17 @@ static int intel_pstate_verify_policy(struct cpufreq_policy *policy) policy->policy != CPUFREQ_POLICY_PERFORMANCE) return -EINVAL; + /* When per-CPU limits are used, sysfs limits are not used */ + if (!per_cpu_limits) { + unsigned int max_freq, min_freq; + + max_freq = policy->cpuinfo.max_freq * + limits->max_sysfs_pct / 100; + min_freq = policy->cpuinfo.max_freq * + limits->min_sysfs_pct / 100; + cpufreq_verify_within_limits(policy, min_freq, max_freq); + } + return 0; } diff --git a/drivers/crypto/ccp/ccp-dev-v5.c b/drivers/crypto/ccp/ccp-dev-v5.c index e2ce8190ecc9..612898b4aaad 100644 --- a/drivers/crypto/ccp/ccp-dev-v5.c +++ b/drivers/crypto/ccp/ccp-dev-v5.c @@ -959,7 +959,7 @@ static irqreturn_t ccp5_irq_handler(int irq, void *data) static void ccp5_config(struct ccp_device *ccp) { /* Public side */ - iowrite32(0x00001249, ccp->io_regs + CMD5_REQID_CONFIG_OFFSET); + iowrite32(0x0, ccp->io_regs + CMD5_REQID_CONFIG_OFFSET); } static void ccp5other_config(struct ccp_device *ccp) diff --git a/drivers/crypto/ccp/ccp-dev.h b/drivers/crypto/ccp/ccp-dev.h index 830f35e6005f..649e5610a5ce 100644 --- a/drivers/crypto/ccp/ccp-dev.h +++ b/drivers/crypto/ccp/ccp-dev.h @@ -238,6 +238,7 @@ struct ccp_dma_chan { struct ccp_device *ccp; spinlock_t lock; + struct list_head created; struct list_head pending; struct list_head active; struct list_head complete; diff --git a/drivers/crypto/ccp/ccp-dmaengine.c b/drivers/crypto/ccp/ccp-dmaengine.c index 6553912804f7..e5d9278f4019 100644 --- a/drivers/crypto/ccp/ccp-dmaengine.c +++ b/drivers/crypto/ccp/ccp-dmaengine.c @@ -63,6 +63,7 @@ static void ccp_free_chan_resources(struct dma_chan *dma_chan) ccp_free_desc_resources(chan->ccp, &chan->complete); ccp_free_desc_resources(chan->ccp, &chan->active); ccp_free_desc_resources(chan->ccp, &chan->pending); + ccp_free_desc_resources(chan->ccp, &chan->created); spin_unlock_irqrestore(&chan->lock, flags); } @@ -273,6 +274,7 @@ static dma_cookie_t ccp_tx_submit(struct dma_async_tx_descriptor *tx_desc) spin_lock_irqsave(&chan->lock, flags); cookie = dma_cookie_assign(tx_desc); + list_del(&desc->entry); list_add_tail(&desc->entry, &chan->pending); spin_unlock_irqrestore(&chan->lock, flags); @@ -426,7 +428,7 @@ static struct ccp_dma_desc *ccp_create_desc(struct dma_chan *dma_chan, spin_lock_irqsave(&chan->lock, sflags); - list_add_tail(&desc->entry, &chan->pending); + list_add_tail(&desc->entry, &chan->created); spin_unlock_irqrestore(&chan->lock, sflags); @@ -610,6 +612,7 @@ static int ccp_terminate_all(struct dma_chan *dma_chan) /*TODO: Purge the complete list? */ ccp_free_desc_resources(chan->ccp, &chan->active); ccp_free_desc_resources(chan->ccp, &chan->pending); + ccp_free_desc_resources(chan->ccp, &chan->created); spin_unlock_irqrestore(&chan->lock, flags); @@ -679,6 +682,7 @@ int ccp_dmaengine_register(struct ccp_device *ccp) chan->ccp = ccp; spin_lock_init(&chan->lock); + INIT_LIST_HEAD(&chan->created); INIT_LIST_HEAD(&chan->pending); INIT_LIST_HEAD(&chan->active); INIT_LIST_HEAD(&chan->complete); diff --git a/drivers/crypto/chelsio/chcr_algo.c b/drivers/crypto/chelsio/chcr_algo.c index 2ed1e24b44a8..b4b78b37f8a6 100644 --- a/drivers/crypto/chelsio/chcr_algo.c +++ b/drivers/crypto/chelsio/chcr_algo.c @@ -158,7 +158,7 @@ int chcr_handle_resp(struct crypto_async_request *req, unsigned char *input, case CRYPTO_ALG_TYPE_AEAD: ctx_req.req.aead_req = (struct aead_request *)req; ctx_req.ctx.reqctx = aead_request_ctx(ctx_req.req.aead_req); - dma_unmap_sg(&u_ctx->lldi.pdev->dev, ctx_req.req.aead_req->dst, + dma_unmap_sg(&u_ctx->lldi.pdev->dev, ctx_req.ctx.reqctx->dst, ctx_req.ctx.reqctx->dst_nents, DMA_FROM_DEVICE); if (ctx_req.ctx.reqctx->skb) { kfree_skb(ctx_req.ctx.reqctx->skb); @@ -1362,8 +1362,7 @@ static struct sk_buff *create_authenc_wr(struct aead_request *req, struct chcr_wr *chcr_req; struct cpl_rx_phys_dsgl *phys_cpl; struct phys_sge_parm sg_param; - struct scatterlist *src, *dst; - struct scatterlist src_sg[2], dst_sg[2]; + struct scatterlist *src; unsigned int frags = 0, transhdr_len; unsigned int ivsize = crypto_aead_ivsize(tfm), dst_size = 0; unsigned int kctx_len = 0; @@ -1383,19 +1382,21 @@ static struct sk_buff *create_authenc_wr(struct aead_request *req, if (sg_nents_for_len(req->src, req->assoclen + req->cryptlen) < 0) goto err; - src = scatterwalk_ffwd(src_sg, req->src, req->assoclen); - dst = src; + src = scatterwalk_ffwd(reqctx->srcffwd, req->src, req->assoclen); + reqctx->dst = src; + if (req->src != req->dst) { err = chcr_copy_assoc(req, aeadctx); if (err) return ERR_PTR(err); - dst = scatterwalk_ffwd(dst_sg, req->dst, req->assoclen); + reqctx->dst = scatterwalk_ffwd(reqctx->dstffwd, req->dst, + req->assoclen); } if (get_aead_subtype(tfm) == CRYPTO_ALG_SUB_TYPE_AEAD_NULL) { null = 1; assoclen = 0; } - reqctx->dst_nents = sg_nents_for_len(dst, req->cryptlen + + reqctx->dst_nents = sg_nents_for_len(reqctx->dst, req->cryptlen + (op_type ? -authsize : authsize)); if (reqctx->dst_nents <= 0) { pr_err("AUTHENC:Invalid Destination sg entries\n"); @@ -1460,7 +1461,7 @@ static struct sk_buff *create_authenc_wr(struct aead_request *req, sg_param.obsize = req->cryptlen + (op_type ? -authsize : authsize); sg_param.qid = qid; sg_param.align = 0; - if (map_writesg_phys_cpl(&u_ctx->lldi.pdev->dev, phys_cpl, dst, + if (map_writesg_phys_cpl(&u_ctx->lldi.pdev->dev, phys_cpl, reqctx->dst, &sg_param)) goto dstmap_fail; @@ -1711,8 +1712,7 @@ static struct sk_buff *create_aead_ccm_wr(struct aead_request *req, struct chcr_wr *chcr_req; struct cpl_rx_phys_dsgl *phys_cpl; struct phys_sge_parm sg_param; - struct scatterlist *src, *dst; - struct scatterlist src_sg[2], dst_sg[2]; + struct scatterlist *src; unsigned int frags = 0, transhdr_len, ivsize = AES_BLOCK_SIZE; unsigned int dst_size = 0, kctx_len; unsigned int sub_type; @@ -1728,17 +1728,19 @@ static struct sk_buff *create_aead_ccm_wr(struct aead_request *req, if (sg_nents_for_len(req->src, req->assoclen + req->cryptlen) < 0) goto err; sub_type = get_aead_subtype(tfm); - src = scatterwalk_ffwd(src_sg, req->src, req->assoclen); - dst = src; + src = scatterwalk_ffwd(reqctx->srcffwd, req->src, req->assoclen); + reqctx->dst = src; + if (req->src != req->dst) { err = chcr_copy_assoc(req, aeadctx); if (err) { pr_err("AAD copy to destination buffer fails\n"); return ERR_PTR(err); } - dst = scatterwalk_ffwd(dst_sg, req->dst, req->assoclen); + reqctx->dst = scatterwalk_ffwd(reqctx->dstffwd, req->dst, + req->assoclen); } - reqctx->dst_nents = sg_nents_for_len(dst, req->cryptlen + + reqctx->dst_nents = sg_nents_for_len(reqctx->dst, req->cryptlen + (op_type ? -authsize : authsize)); if (reqctx->dst_nents <= 0) { pr_err("CCM:Invalid Destination sg entries\n"); @@ -1777,7 +1779,7 @@ static struct sk_buff *create_aead_ccm_wr(struct aead_request *req, sg_param.obsize = req->cryptlen + (op_type ? -authsize : authsize); sg_param.qid = qid; sg_param.align = 0; - if (map_writesg_phys_cpl(&u_ctx->lldi.pdev->dev, phys_cpl, dst, + if (map_writesg_phys_cpl(&u_ctx->lldi.pdev->dev, phys_cpl, reqctx->dst, &sg_param)) goto dstmap_fail; @@ -1809,8 +1811,7 @@ static struct sk_buff *create_gcm_wr(struct aead_request *req, struct chcr_wr *chcr_req; struct cpl_rx_phys_dsgl *phys_cpl; struct phys_sge_parm sg_param; - struct scatterlist *src, *dst; - struct scatterlist src_sg[2], dst_sg[2]; + struct scatterlist *src; unsigned int frags = 0, transhdr_len; unsigned int ivsize = AES_BLOCK_SIZE; unsigned int dst_size = 0, kctx_len; @@ -1832,13 +1833,14 @@ static struct sk_buff *create_gcm_wr(struct aead_request *req, if (sg_nents_for_len(req->src, req->assoclen + req->cryptlen) < 0) goto err; - src = scatterwalk_ffwd(src_sg, req->src, req->assoclen); - dst = src; + src = scatterwalk_ffwd(reqctx->srcffwd, req->src, req->assoclen); + reqctx->dst = src; if (req->src != req->dst) { err = chcr_copy_assoc(req, aeadctx); if (err) return ERR_PTR(err); - dst = scatterwalk_ffwd(dst_sg, req->dst, req->assoclen); + reqctx->dst = scatterwalk_ffwd(reqctx->dstffwd, req->dst, + req->assoclen); } if (!req->cryptlen) @@ -1848,7 +1850,7 @@ static struct sk_buff *create_gcm_wr(struct aead_request *req, crypt_len = AES_BLOCK_SIZE; else crypt_len = req->cryptlen; - reqctx->dst_nents = sg_nents_for_len(dst, req->cryptlen + + reqctx->dst_nents = sg_nents_for_len(reqctx->dst, req->cryptlen + (op_type ? -authsize : authsize)); if (reqctx->dst_nents <= 0) { pr_err("GCM:Invalid Destination sg entries\n"); @@ -1923,7 +1925,7 @@ static struct sk_buff *create_gcm_wr(struct aead_request *req, sg_param.obsize = req->cryptlen + (op_type ? -authsize : authsize); sg_param.qid = qid; sg_param.align = 0; - if (map_writesg_phys_cpl(&u_ctx->lldi.pdev->dev, phys_cpl, dst, + if (map_writesg_phys_cpl(&u_ctx->lldi.pdev->dev, phys_cpl, reqctx->dst, &sg_param)) goto dstmap_fail; @@ -1937,7 +1939,8 @@ static struct sk_buff *create_gcm_wr(struct aead_request *req, write_sg_to_skb(skb, &frags, src, req->cryptlen); } else { aes_gcm_empty_pld_pad(req->dst, authsize - 1); - write_sg_to_skb(skb, &frags, dst, crypt_len); + write_sg_to_skb(skb, &frags, reqctx->dst, crypt_len); + } create_wreq(ctx, chcr_req, req, skb, kctx_len, size, 1, @@ -2189,8 +2192,8 @@ static int chcr_gcm_setkey(struct crypto_aead *aead, const u8 *key, unsigned int ck_size; int ret = 0, key_ctx_size = 0; - if (get_aead_subtype(aead) == - CRYPTO_ALG_SUB_TYPE_AEAD_RFC4106) { + if (get_aead_subtype(aead) == CRYPTO_ALG_SUB_TYPE_AEAD_RFC4106 && + keylen > 3) { keylen -= 4; /* nonce/salt is present in the last 4 bytes */ memcpy(aeadctx->salt, key + keylen, 4); } diff --git a/drivers/crypto/chelsio/chcr_core.c b/drivers/crypto/chelsio/chcr_core.c index 918da8e6e2d8..1c65f07e1cc9 100644 --- a/drivers/crypto/chelsio/chcr_core.c +++ b/drivers/crypto/chelsio/chcr_core.c @@ -52,6 +52,7 @@ static struct cxgb4_uld_info chcr_uld_info = { int assign_chcr_device(struct chcr_dev **dev) { struct uld_ctx *u_ctx; + int ret = -ENXIO; /* * Which device to use if multiple devices are available TODO @@ -59,15 +60,14 @@ int assign_chcr_device(struct chcr_dev **dev) * must go to the same device to maintain the ordering. */ mutex_lock(&dev_mutex); /* TODO ? */ - u_ctx = list_first_entry(&uld_ctx_list, struct uld_ctx, entry); - if (!u_ctx) { - mutex_unlock(&dev_mutex); - return -ENXIO; + list_for_each_entry(u_ctx, &uld_ctx_list, entry) + if (u_ctx && u_ctx->dev) { + *dev = u_ctx->dev; + ret = 0; + break; } - - *dev = u_ctx->dev; mutex_unlock(&dev_mutex); - return 0; + return ret; } static int chcr_dev_add(struct uld_ctx *u_ctx) @@ -202,10 +202,8 @@ static int chcr_uld_state_change(void *handle, enum cxgb4_state state) static int __init chcr_crypto_init(void) { - if (cxgb4_register_uld(CXGB4_ULD_CRYPTO, &chcr_uld_info)) { + if (cxgb4_register_uld(CXGB4_ULD_CRYPTO, &chcr_uld_info)) pr_err("ULD register fail: No chcr crypto support in cxgb4"); - return -1; - } return 0; } diff --git a/drivers/crypto/chelsio/chcr_crypto.h b/drivers/crypto/chelsio/chcr_crypto.h index d5af7d64a763..7ec0a8f12475 100644 --- a/drivers/crypto/chelsio/chcr_crypto.h +++ b/drivers/crypto/chelsio/chcr_crypto.h @@ -158,6 +158,9 @@ struct ablk_ctx { }; struct chcr_aead_reqctx { struct sk_buff *skb; + struct scatterlist *dst; + struct scatterlist srcffwd[2]; + struct scatterlist dstffwd[2]; short int dst_nents; u16 verify; u8 iv[CHCR_MAX_CRYPTO_IV_LEN]; diff --git a/drivers/crypto/qat/qat_c62x/adf_drv.c b/drivers/crypto/qat/qat_c62x/adf_drv.c index bc5cbc193aae..5b2d78a5b5aa 100644 --- a/drivers/crypto/qat/qat_c62x/adf_drv.c +++ b/drivers/crypto/qat/qat_c62x/adf_drv.c @@ -233,7 +233,7 @@ static int adf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) &hw_data->accel_capabilities_mask); /* Find and map all the device's BARS */ - i = 0; + i = (hw_data->fuses & ADF_DEVICE_FUSECTL_MASK) ? 1 : 0; bar_mask = pci_select_bars(pdev, IORESOURCE_MEM); for_each_set_bit(bar_nr, (const unsigned long *)&bar_mask, ADF_PCI_MAX_BARS * 2) { diff --git a/drivers/crypto/qat/qat_common/adf_accel_devices.h b/drivers/crypto/qat/qat_common/adf_accel_devices.h index e8822536530b..33f0a6251e38 100644 --- a/drivers/crypto/qat/qat_common/adf_accel_devices.h +++ b/drivers/crypto/qat/qat_common/adf_accel_devices.h @@ -69,6 +69,7 @@ #define ADF_ERRSOU5 (0x3A000 + 0xD8) #define ADF_DEVICE_FUSECTL_OFFSET 0x40 #define ADF_DEVICE_LEGFUSE_OFFSET 0x4C +#define ADF_DEVICE_FUSECTL_MASK 0x80000000 #define ADF_PCI_MAX_BARS 3 #define ADF_DEVICE_NAME_LENGTH 32 #define ADF_ETR_MAX_RINGS_PER_BANK 16 diff --git a/drivers/crypto/qat/qat_common/qat_hal.c b/drivers/crypto/qat/qat_common/qat_hal.c index 1e480f140663..8c4fd255a601 100644 --- a/drivers/crypto/qat/qat_common/qat_hal.c +++ b/drivers/crypto/qat/qat_common/qat_hal.c @@ -456,7 +456,7 @@ static int qat_hal_init_esram(struct icp_qat_fw_loader_handle *handle) unsigned int csr_val; int times = 30; - if (handle->pci_dev->device == ADF_C3XXX_PCI_DEVICE_ID) + if (handle->pci_dev->device != ADF_DH895XCC_PCI_DEVICE_ID) return 0; csr_val = ADF_CSR_RD(csr_addr, 0); @@ -716,7 +716,7 @@ int qat_hal_init(struct adf_accel_dev *accel_dev) (void __iomem *)((uintptr_t)handle->hal_cap_ae_xfer_csr_addr_v + LOCAL_TO_XFER_REG_OFFSET); handle->pci_dev = pci_info->pci_dev; - if (handle->pci_dev->device != ADF_C3XXX_PCI_DEVICE_ID) { + if (handle->pci_dev->device == ADF_DH895XCC_PCI_DEVICE_ID) { sram_bar = &pci_info->pci_bars[hw_data->get_sram_bar_id(hw_data)]; handle->hal_sram_addr_v = sram_bar->virt_addr; diff --git a/drivers/dma/cppi41.c b/drivers/dma/cppi41.c index d5ba43a87a68..200828c60db9 100644 --- a/drivers/dma/cppi41.c +++ b/drivers/dma/cppi41.c @@ -153,6 +153,8 @@ struct cppi41_dd { /* context for suspend/resume */ unsigned int dma_tdfdq; + + bool is_suspended; }; #define FIST_COMPLETION_QUEUE 93 @@ -257,6 +259,10 @@ static struct cppi41_channel *desc_to_chan(struct cppi41_dd *cdd, u32 desc) BUG_ON(desc_num >= ALLOC_DECS_NUM); c = cdd->chan_busy[desc_num]; cdd->chan_busy[desc_num] = NULL; + + /* Usecount for chan_busy[], paired with push_desc_queue() */ + pm_runtime_put(cdd->ddev.dev); + return c; } @@ -317,12 +323,12 @@ static irqreturn_t cppi41_irq(int irq, void *data) while (val) { u32 desc, len; - int error; - error = pm_runtime_get(cdd->ddev.dev); - if (error < 0) - dev_err(cdd->ddev.dev, "%s pm runtime get: %i\n", - __func__, error); + /* + * This should never trigger, see the comments in + * push_desc_queue() + */ + WARN_ON(cdd->is_suspended); q_num = __fls(val); val &= ~(1 << q_num); @@ -343,9 +349,6 @@ static irqreturn_t cppi41_irq(int irq, void *data) c->residue = pd_trans_len(c->desc->pd6) - len; dma_cookie_complete(&c->txd); dmaengine_desc_get_callback_invoke(&c->txd, NULL); - - pm_runtime_mark_last_busy(cdd->ddev.dev); - pm_runtime_put_autosuspend(cdd->ddev.dev); } } return IRQ_HANDLED; @@ -447,6 +450,15 @@ static void push_desc_queue(struct cppi41_channel *c) */ __iowmb(); + /* + * DMA transfers can take at least 200ms to complete with USB mass + * storage connected. To prevent autosuspend timeouts, we must use + * pm_runtime_get/put() when chan_busy[] is modified. This will get + * cleared in desc_to_chan() or cppi41_stop_chan() depending on the + * outcome of the transfer. + */ + pm_runtime_get(cdd->ddev.dev); + desc_phys = lower_32_bits(c->desc_phys); desc_num = (desc_phys - cdd->descs_phys) / sizeof(struct cppi41_desc); WARN_ON(cdd->chan_busy[desc_num]); @@ -457,20 +469,26 @@ static void push_desc_queue(struct cppi41_channel *c) cppi_writel(reg, cdd->qmgr_mem + QMGR_QUEUE_D(c->q_num)); } -static void pending_desc(struct cppi41_channel *c) +/* + * Caller must hold cdd->lock to prevent push_desc_queue() + * getting called out of order. We have both cppi41_dma_issue_pending() + * and cppi41_runtime_resume() call this function. + */ +static void cppi41_run_queue(struct cppi41_dd *cdd) { - struct cppi41_dd *cdd = c->cdd; - unsigned long flags; + struct cppi41_channel *c, *_c; - spin_lock_irqsave(&cdd->lock, flags); - list_add_tail(&c->node, &cdd->pending); - spin_unlock_irqrestore(&cdd->lock, flags); + list_for_each_entry_safe(c, _c, &cdd->pending, node) { + push_desc_queue(c); + list_del(&c->node); + } } static void cppi41_dma_issue_pending(struct dma_chan *chan) { struct cppi41_channel *c = to_cpp41_chan(chan); struct cppi41_dd *cdd = c->cdd; + unsigned long flags; int error; error = pm_runtime_get(cdd->ddev.dev); @@ -482,10 +500,11 @@ static void cppi41_dma_issue_pending(struct dma_chan *chan) return; } - if (likely(pm_runtime_active(cdd->ddev.dev))) - push_desc_queue(c); - else - pending_desc(c); + spin_lock_irqsave(&cdd->lock, flags); + list_add_tail(&c->node, &cdd->pending); + if (!cdd->is_suspended) + cppi41_run_queue(cdd); + spin_unlock_irqrestore(&cdd->lock, flags); pm_runtime_mark_last_busy(cdd->ddev.dev); pm_runtime_put_autosuspend(cdd->ddev.dev); @@ -705,6 +724,9 @@ static int cppi41_stop_chan(struct dma_chan *chan) WARN_ON(!cdd->chan_busy[desc_num]); cdd->chan_busy[desc_num] = NULL; + /* Usecount for chan_busy[], paired with push_desc_queue() */ + pm_runtime_put(cdd->ddev.dev); + return 0; } @@ -1150,8 +1172,12 @@ static int __maybe_unused cppi41_resume(struct device *dev) static int __maybe_unused cppi41_runtime_suspend(struct device *dev) { struct cppi41_dd *cdd = dev_get_drvdata(dev); + unsigned long flags; + spin_lock_irqsave(&cdd->lock, flags); + cdd->is_suspended = true; WARN_ON(!list_empty(&cdd->pending)); + spin_unlock_irqrestore(&cdd->lock, flags); return 0; } @@ -1159,14 +1185,11 @@ static int __maybe_unused cppi41_runtime_suspend(struct device *dev) static int __maybe_unused cppi41_runtime_resume(struct device *dev) { struct cppi41_dd *cdd = dev_get_drvdata(dev); - struct cppi41_channel *c, *_c; unsigned long flags; spin_lock_irqsave(&cdd->lock, flags); - list_for_each_entry_safe(c, _c, &cdd->pending, node) { - push_desc_queue(c); - list_del(&c->node); - } + cdd->is_suspended = false; + cppi41_run_queue(cdd); spin_unlock_irqrestore(&cdd->lock, flags); return 0; diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c index 740bbb942594..7539f73df9e0 100644 --- a/drivers/dma/pl330.c +++ b/drivers/dma/pl330.c @@ -1699,7 +1699,6 @@ static bool _chan_ns(const struct pl330_dmac *pl330, int i) static struct pl330_thread *pl330_request_channel(struct pl330_dmac *pl330) { struct pl330_thread *thrd = NULL; - unsigned long flags; int chans, i; if (pl330->state == DYING) @@ -1707,8 +1706,6 @@ static struct pl330_thread *pl330_request_channel(struct pl330_dmac *pl330) chans = pl330->pcfg.num_chan; - spin_lock_irqsave(&pl330->lock, flags); - for (i = 0; i < chans; i++) { thrd = &pl330->channels[i]; if ((thrd->free) && (!_manager_ns(thrd) || @@ -1726,8 +1723,6 @@ static struct pl330_thread *pl330_request_channel(struct pl330_dmac *pl330) thrd = NULL; } - spin_unlock_irqrestore(&pl330->lock, flags); - return thrd; } @@ -1745,7 +1740,6 @@ static inline void _free_event(struct pl330_thread *thrd, int ev) static void pl330_release_channel(struct pl330_thread *thrd) { struct pl330_dmac *pl330; - unsigned long flags; if (!thrd || thrd->free) return; @@ -1757,10 +1751,8 @@ static void pl330_release_channel(struct pl330_thread *thrd) pl330 = thrd->dmac; - spin_lock_irqsave(&pl330->lock, flags); _free_event(thrd, thrd->ev); thrd->free = true; - spin_unlock_irqrestore(&pl330->lock, flags); } /* Initialize the structure for PL330 configuration, that can be used @@ -2122,20 +2114,20 @@ static int pl330_alloc_chan_resources(struct dma_chan *chan) struct pl330_dmac *pl330 = pch->dmac; unsigned long flags; - spin_lock_irqsave(&pch->lock, flags); + spin_lock_irqsave(&pl330->lock, flags); dma_cookie_init(chan); pch->cyclic = false; pch->thread = pl330_request_channel(pl330); if (!pch->thread) { - spin_unlock_irqrestore(&pch->lock, flags); + spin_unlock_irqrestore(&pl330->lock, flags); return -ENOMEM; } tasklet_init(&pch->task, pl330_tasklet, (unsigned long) pch); - spin_unlock_irqrestore(&pch->lock, flags); + spin_unlock_irqrestore(&pl330->lock, flags); return 1; } @@ -2238,12 +2230,13 @@ static int pl330_pause(struct dma_chan *chan) static void pl330_free_chan_resources(struct dma_chan *chan) { struct dma_pl330_chan *pch = to_pchan(chan); + struct pl330_dmac *pl330 = pch->dmac; unsigned long flags; tasklet_kill(&pch->task); pm_runtime_get_sync(pch->dmac->ddma.dev); - spin_lock_irqsave(&pch->lock, flags); + spin_lock_irqsave(&pl330->lock, flags); pl330_release_channel(pch->thread); pch->thread = NULL; @@ -2251,7 +2244,7 @@ static void pl330_free_chan_resources(struct dma_chan *chan) if (pch->cyclic) list_splice_tail_init(&pch->work_list, &pch->dmac->desc_pool); - spin_unlock_irqrestore(&pch->lock, flags); + spin_unlock_irqrestore(&pl330->lock, flags); pm_runtime_mark_last_busy(pch->dmac->ddma.dev); pm_runtime_put_autosuspend(pch->dmac->ddma.dev); } diff --git a/drivers/firmware/efi/libstub/fdt.c b/drivers/firmware/efi/libstub/fdt.c index 921dfa047202..260c4b4b492e 100644 --- a/drivers/firmware/efi/libstub/fdt.c +++ b/drivers/firmware/efi/libstub/fdt.c @@ -187,6 +187,7 @@ static efi_status_t update_fdt_memmap(void *fdt, struct efi_boot_memmap *map) struct exit_boot_struct { efi_memory_desc_t *runtime_map; int *runtime_entry_count; + void *new_fdt_addr; }; static efi_status_t exit_boot_func(efi_system_table_t *sys_table_arg, @@ -202,7 +203,7 @@ static efi_status_t exit_boot_func(efi_system_table_t *sys_table_arg, efi_get_virtmap(*map->map, *map->map_size, *map->desc_size, p->runtime_map, p->runtime_entry_count); - return EFI_SUCCESS; + return update_fdt_memmap(p->new_fdt_addr, map); } /* @@ -300,22 +301,13 @@ efi_status_t allocate_new_fdt_and_exit_boot(efi_system_table_t *sys_table, priv.runtime_map = runtime_map; priv.runtime_entry_count = &runtime_entry_count; + priv.new_fdt_addr = (void *)*new_fdt_addr; status = efi_exit_boot_services(sys_table, handle, &map, &priv, exit_boot_func); if (status == EFI_SUCCESS) { efi_set_virtual_address_map_t *svam; - status = update_fdt_memmap((void *)*new_fdt_addr, &map); - if (status != EFI_SUCCESS) { - /* - * The kernel won't get far without the memory map, but - * may still be able to print something meaningful so - * return success here. - */ - return EFI_SUCCESS; - } - /* Install the new virtual address map */ svam = sys_table->runtime->set_virtual_address_map; status = svam(runtime_entry_count * desc_size, desc_size, diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig index 90bc65d07a35..88e01e08e279 100644 --- a/drivers/gpu/drm/Kconfig +++ b/drivers/gpu/drm/Kconfig @@ -263,6 +263,8 @@ source "drivers/gpu/drm/mxsfb/Kconfig" source "drivers/gpu/drm/meson/Kconfig" +source "drivers/gpu/drm/tinydrm/Kconfig" + # Keep legacy drivers last menuconfig DRM_LEGACY diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index 92de3991fa56..3ee95793d122 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -94,3 +94,4 @@ obj-$(CONFIG_DRM_ARCPGU)+= arc/ obj-y += hisilicon/ obj-$(CONFIG_DRM_ZTE) += zte/ obj-$(CONFIG_DRM_MXSFB) += mxsfb/ +obj-$(CONFIG_DRM_TINYDRM) += tinydrm/ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 96ad79627dbb..c1b913541739 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1037,7 +1037,6 @@ struct amdgpu_uvd { bool use_ctx_buf; struct amd_sched_entity entity; uint32_t srbm_soft_reset; - bool is_powergated; }; /* @@ -1066,7 +1065,6 @@ struct amdgpu_vce { struct amd_sched_entity entity; uint32_t srbm_soft_reset; unsigned num_rings; - bool is_powergated; }; /* @@ -1484,6 +1482,9 @@ struct amdgpu_device { spinlock_t gtt_list_lock; struct list_head gtt_list; + /* record hw reset is performed */ + bool has_hw_reset; + }; static inline struct amdgpu_device *amdgpu_ttm_adev(struct ttm_bo_device *bdev) @@ -1702,7 +1703,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) int amdgpu_gpu_reset(struct amdgpu_device *adev); bool amdgpu_need_backup(struct amdgpu_device *adev); void amdgpu_pci_config_reset(struct amdgpu_device *adev); -bool amdgpu_card_posted(struct amdgpu_device *adev); +bool amdgpu_need_post(struct amdgpu_device *adev); void amdgpu_update_display_priority(struct amdgpu_device *adev); int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c index d9def01f276e..821f7cc2051f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c @@ -100,7 +100,7 @@ static bool igp_read_bios_from_vram(struct amdgpu_device *adev) resource_size_t size = 256 * 1024; /* ??? */ if (!(adev->flags & AMD_IS_APU)) - if (!amdgpu_card_posted(adev)) + if (amdgpu_need_post(adev)) return false; adev->bios = NULL; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index fa34dcae392f..d9e5aa4a79ef 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -834,16 +834,18 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, case CHIP_TOPAZ: if (((adev->pdev->device == 0x6900) && (adev->pdev->revision == 0x81)) || ((adev->pdev->device == 0x6900) && (adev->pdev->revision == 0x83)) || - ((adev->pdev->device == 0x6907) && (adev->pdev->revision == 0x87))) + ((adev->pdev->device == 0x6907) && (adev->pdev->revision == 0x87))) { + info->is_kicker = true; strcpy(fw_name, "amdgpu/topaz_k_smc.bin"); - else + } else strcpy(fw_name, "amdgpu/topaz_smc.bin"); break; case CHIP_TONGA: if (((adev->pdev->device == 0x6939) && (adev->pdev->revision == 0xf1)) || - ((adev->pdev->device == 0x6938) && (adev->pdev->revision == 0xf1))) + ((adev->pdev->device == 0x6938) && (adev->pdev->revision == 0xf1))) { + info->is_kicker = true; strcpy(fw_name, "amdgpu/tonga_k_smc.bin"); - else + } else strcpy(fw_name, "amdgpu/tonga_smc.bin"); break; case CHIP_FIJI: @@ -858,9 +860,10 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, ((adev->pdev->device == 0x67ff) && ((adev->pdev->revision == 0xcf) || (adev->pdev->revision == 0xef) || - (adev->pdev->revision == 0xff)))) + (adev->pdev->revision == 0xff)))) { + info->is_kicker = true; strcpy(fw_name, "amdgpu/polaris11_k_smc.bin"); - else + } else strcpy(fw_name, "amdgpu/polaris11_smc.bin"); } else if (type == CGS_UCODE_ID_SMU_SK) { strcpy(fw_name, "amdgpu/polaris11_smc_sk.bin"); @@ -874,9 +877,10 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, (adev->pdev->revision == 0xe4) || (adev->pdev->revision == 0xe5) || (adev->pdev->revision == 0xe7) || - (adev->pdev->revision == 0xef))) + (adev->pdev->revision == 0xef))) { + info->is_kicker = true; strcpy(fw_name, "amdgpu/polaris10_k_smc.bin"); - else + } else strcpy(fw_name, "amdgpu/polaris10_smc.bin"); } else if (type == CGS_UCODE_ID_SMU_SK) { strcpy(fw_name, "amdgpu/polaris10_smc_sk.bin"); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 57301f5936fa..d2d0f60ff36d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -83,6 +83,13 @@ int amdgpu_cs_get_ring(struct amdgpu_device *adev, u32 ip_type, } break; } + + if (!(*out_ring && (*out_ring)->adev)) { + DRM_ERROR("Ring %d is not initialized on IP %d\n", + ring, ip_type); + return -EINVAL; + } + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 944ba0d3874a..6abb238b25c9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -619,25 +619,29 @@ void amdgpu_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc) * GPU helpers function. */ /** - * amdgpu_card_posted - check if the hw has already been initialized + * amdgpu_need_post - check if the hw need post or not * * @adev: amdgpu_device pointer * - * Check if the asic has been initialized (all asics). - * Used at driver startup. - * Returns true if initialized or false if not. + * Check if the asic has been initialized (all asics) at driver startup + * or post is needed if hw reset is performed. + * Returns true if need or false if not. */ -bool amdgpu_card_posted(struct amdgpu_device *adev) +bool amdgpu_need_post(struct amdgpu_device *adev) { uint32_t reg; + if (adev->has_hw_reset) { + adev->has_hw_reset = false; + return true; + } /* then check MEM_SIZE, in case the crtcs are off */ reg = RREG32(mmCONFIG_MEMSIZE); if (reg) - return true; + return false; - return false; + return true; } @@ -665,7 +669,7 @@ static bool amdgpu_vpost_needed(struct amdgpu_device *adev) return true; } } - return !amdgpu_card_posted(adev); + return amdgpu_need_post(adev); } /** @@ -2071,7 +2075,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) amdgpu_atombios_scratch_regs_restore(adev); /* post card */ - if (!amdgpu_card_posted(adev) || !resume) { + if (amdgpu_need_post(adev)) { r = amdgpu_atom_asic_init(adev->mode_info.atom_context); if (r) DRM_ERROR("amdgpu asic init failed\n"); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index ec7037a48b6e..51d759463384 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -504,13 +504,7 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, list_for_each_entry(entry, list, head) { struct amdgpu_bo *bo = container_of(entry->bo, struct amdgpu_bo, tbo); - - /* if anything is swapped out don't swap it in here, - just abort and wait for the next CS */ - if (!amdgpu_bo_gpu_accessible(bo)) - goto error; - - if (bo->shadow && !amdgpu_bo_gpu_accessible(bo->shadow)) + if (amdgpu_gem_va_check(NULL, bo)) goto error; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index 95e026a4a2de..346e80a7119b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -1296,7 +1296,8 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) if (!adev->pm.dpm_enabled) return; - amdgpu_display_bandwidth_update(adev); + if (adev->mode_info.num_crtc) + amdgpu_display_bandwidth_update(adev); for (i = 0; i < AMDGPU_MAX_RINGS; i++) { struct amdgpu_ring *ring = adev->rings[i]; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 1154b0a8881d..4c6094eefc51 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -529,6 +529,9 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_ case TTM_PL_TT: break; case TTM_PL_VRAM: + if (mem->start == AMDGPU_BO_INVALID_OFFSET) + return -EINVAL; + mem->bus.offset = mem->start << PAGE_SHIFT; /* check if it's visible */ if ((mem->bus.offset + mem->bus.size) > adev->mc.visible_vram_size) diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index 7c39b538dc0e..c4d4b35e54ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -1176,6 +1176,7 @@ static int cik_gpu_pci_config_reset(struct amdgpu_device *adev) if (RREG32(mmCONFIG_MEMSIZE) != 0xffffffff) { /* enable BM */ pci_set_master(adev->pdev); + adev->has_hw_reset = true; r = 0; break; } diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 1cf1d9d1aec1..5b24e89552ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -3737,9 +3737,15 @@ static void dce_v11_0_encoder_add(struct amdgpu_device *adev, default: encoder->possible_crtcs = 0x3; break; + case 3: + encoder->possible_crtcs = 0x7; + break; case 4: encoder->possible_crtcs = 0xf; break; + case 5: + encoder->possible_crtcs = 0x1f; + break; case 6: encoder->possible_crtcs = 0x3f; break; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c index 762f8e82ceb7..e9a176891e13 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c @@ -627,11 +627,8 @@ static const struct drm_encoder_helper_funcs dce_virtual_encoder_helper_funcs = static void dce_virtual_encoder_destroy(struct drm_encoder *encoder) { - struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); - - kfree(amdgpu_encoder->enc_priv); drm_encoder_cleanup(encoder); - kfree(amdgpu_encoder); + kfree(encoder); } static const struct drm_encoder_funcs dce_virtual_encoder_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index e3589b55a1e1..1f9354541f29 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -1983,6 +1983,14 @@ static void gfx_v7_0_gpu_init(struct amdgpu_device *adev) WREG32(mmPA_CL_ENHANCE, PA_CL_ENHANCE__CLIP_VTX_REORDER_ENA_MASK | (3 << PA_CL_ENHANCE__NUM_CLIP_SEQ__SHIFT)); WREG32(mmPA_SC_ENHANCE, PA_SC_ENHANCE__ENABLE_PA_SC_OUT_OF_ORDER_MASK); + + tmp = RREG32(mmSPI_ARB_PRIORITY); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS0, 2); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS1, 2); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS2, 2); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS3, 2); + WREG32(mmSPI_ARB_PRIORITY, tmp); + mutex_unlock(&adev->grbm_idx_mutex); udelay(50); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 35f9cd83b821..67afc901905c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -3898,6 +3898,14 @@ static void gfx_v8_0_gpu_init(struct amdgpu_device *adev) PA_SC_FIFO_SIZE__SC_HIZ_TILE_FIFO_SIZE__SHIFT) | (adev->gfx.config.sc_earlyz_tile_fifo_size << PA_SC_FIFO_SIZE__SC_EARLYZ_TILE_FIFO_SIZE__SHIFT)); + + tmp = RREG32(mmSPI_ARB_PRIORITY); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS0, 2); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS1, 2); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS2, 2); + tmp = REG_SET_FIELD(tmp, SPI_ARB_PRIORITY, PIPE_ORDER_TS3, 2); + WREG32(mmSPI_ARB_PRIORITY, tmp); + mutex_unlock(&adev->grbm_idx_mutex); } @@ -7260,7 +7268,7 @@ static void gfx_v8_0_ring_emit_ce_meta_init(struct amdgpu_ring *ring, uint64_t c static union { struct amdgpu_ce_ib_state regular; struct amdgpu_ce_ib_state_chained_ib chained; - } ce_payload = {0}; + } ce_payload = {}; if (ring->adev->virt.chained_ib_support) { ce_payload_addr = csa_addr + offsetof(struct amdgpu_gfx_meta_data_chained_ib, ce_payload); @@ -7287,7 +7295,7 @@ static void gfx_v8_0_ring_emit_de_meta_init(struct amdgpu_ring *ring, uint64_t c static union { struct amdgpu_de_ib_state regular; struct amdgpu_de_ib_state_chained_ib chained; - } de_payload = {0}; + } de_payload = {}; gds_addr = csa_addr + 4096; if (ring->adev->virt.chained_ib_support) { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index e2b0b1646f99..0635829b18cf 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -254,6 +254,9 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) } WREG32(mmHDP_REG_COHERENCY_FLUSH_CNTL, 0); + if (adev->mode_info.num_crtc) + amdgpu_display_set_vga_render_state(adev, false); + gmc_v6_0_mc_stop(adev, &save); if (gmc_v6_0_wait_for_idle((void *)adev)) { @@ -283,7 +286,6 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } gmc_v6_0_mc_resume(adev, &save); - amdgpu_display_set_vga_render_state(adev, false); } static int gmc_v6_0_mc_init(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index f15df99f0a06..b34cefc7ebd5 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -401,7 +401,8 @@ static void uvd_v4_2_stop(struct amdgpu_device *adev) break; mdelay(1); } - break; + if (status & 2) + break; } for (i = 0; i < 10; ++i) { @@ -411,7 +412,8 @@ static void uvd_v4_2_stop(struct amdgpu_device *adev) break; mdelay(1); } - break; + if (status & 0xf) + break; } /* Stall UMC and register bus before resetting VCPU */ @@ -424,7 +426,8 @@ static void uvd_v4_2_stop(struct amdgpu_device *adev) break; mdelay(1); } - break; + if (status & 0x240) + break; } WREG32_P(0x3D49, 0, ~(1 << 2)); @@ -723,7 +726,8 @@ static int uvd_v4_2_set_powergating_state(void *handle, if (state == AMD_PG_STATE_GATE) { uvd_v4_2_stop(adev); if (adev->pg_flags & AMD_PG_SUPPORT_UVD && amdgpu_dpm == 0) { - if (!(RREG32_SMC(ixCURRENT_PG_STATUS) & 0x4)) { + if (!(RREG32_SMC(ixCURRENT_PG_STATUS) & + CURRENT_PG_STATUS__UVD_PG_STATUS_MASK)) { WREG32(mmUVD_PGFSM_CONFIG, (UVD_PGFSM_CONFIG__UVD_PGFSM_FSM_ADDR_MASK | UVD_PGFSM_CONFIG__UVD_PGFSM_POWER_DOWN_MASK | UVD_PGFSM_CONFIG__UVD_PGFSM_P1_SELECT_MASK)); @@ -733,7 +737,8 @@ static int uvd_v4_2_set_powergating_state(void *handle, return 0; } else { if (adev->pg_flags & AMD_PG_SUPPORT_UVD && amdgpu_dpm == 0) { - if (RREG32_SMC(ixCURRENT_PG_STATUS) & 0x4) { + if (RREG32_SMC(ixCURRENT_PG_STATUS) & + CURRENT_PG_STATUS__UVD_PG_STATUS_MASK) { WREG32(mmUVD_PGFSM_CONFIG, (UVD_PGFSM_CONFIG__UVD_PGFSM_FSM_ADDR_MASK | UVD_PGFSM_CONFIG__UVD_PGFSM_POWER_UP_MASK | UVD_PGFSM_CONFIG__UVD_PGFSM_P1_SELECT_MASK)); diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index 46e715193924..ad8c02e423d4 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -825,12 +825,10 @@ static int uvd_v5_0_set_powergating_state(void *handle, if (state == AMD_PG_STATE_GATE) { uvd_v5_0_stop(adev); - adev->uvd.is_powergated = true; } else { ret = uvd_v5_0_start(adev); if (ret) goto out; - adev->uvd.is_powergated = false; } out: @@ -844,7 +842,8 @@ static void uvd_v5_0_get_clockgating_state(void *handle, u32 *flags) mutex_lock(&adev->pm.mutex); - if (adev->uvd.is_powergated) { + if (RREG32_SMC(ixCURRENT_PG_STATUS) & + CURRENT_PG_STATUS__UVD_PG_STATUS_MASK) { DRM_INFO("Cannot get clockgating state when UVD is powergated.\n"); goto out; } diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index af83ab8c1250..18a6de4e1512 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -1051,12 +1051,10 @@ static int uvd_v6_0_set_powergating_state(void *handle, if (state == AMD_PG_STATE_GATE) { uvd_v6_0_stop(adev); - adev->uvd.is_powergated = true; } else { ret = uvd_v6_0_start(adev); if (ret) goto out; - adev->uvd.is_powergated = false; } out: @@ -1070,7 +1068,8 @@ static void uvd_v6_0_get_clockgating_state(void *handle, u32 *flags) mutex_lock(&adev->pm.mutex); - if (adev->uvd.is_powergated) { + if (RREG32_SMC(ixCURRENT_PG_STATUS) & + CURRENT_PG_STATUS__UVD_PG_STATUS_MASK) { DRM_INFO("Cannot get clockgating state when UVD is powergated.\n"); goto out; } diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index a8c40eebdd78..93ec8815bb13 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -510,6 +510,8 @@ static void vce_v3_0_mc_resume(struct amdgpu_device *adev, int idx) WREG32(mmVCE_LMI_SWAP_CNTL, 0); WREG32(mmVCE_LMI_SWAP_CNTL1, 0); WREG32(mmVCE_LMI_VM_CTRL, 0); + WREG32_OR(mmVCE_VCPU_CNTL, 0x00100000); + if (adev->asic_type >= CHIP_STONEY) { WREG32(mmVCE_LMI_VCPU_CACHE_40BIT_BAR0, (adev->vce.gpu_addr >> 8)); WREG32(mmVCE_LMI_VCPU_CACHE_40BIT_BAR1, (adev->vce.gpu_addr >> 8)); @@ -766,12 +768,10 @@ static int vce_v3_0_set_powergating_state(void *handle, ret = vce_v3_0_stop(adev); if (ret) goto out; - adev->vce.is_powergated = true; } else { ret = vce_v3_0_start(adev); if (ret) goto out; - adev->vce.is_powergated = false; } out: @@ -785,7 +785,8 @@ static void vce_v3_0_get_clockgating_state(void *handle, u32 *flags) mutex_lock(&adev->pm.mutex); - if (adev->vce.is_powergated) { + if (RREG32_SMC(ixCURRENT_PG_STATUS) & + CURRENT_PG_STATUS__VCE_PG_STATUS_MASK) { DRM_INFO("Cannot get clockgating state when VCE is powergated.\n"); goto out; } diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 4922fff08c3c..50bdb24ef8d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -721,6 +721,7 @@ static int vi_gpu_pci_config_reset(struct amdgpu_device *adev) if (RREG32(mmCONFIG_MEMSIZE) != 0xffffffff) { /* enable BM */ pci_set_master(adev->pdev); + adev->has_hw_reset = true; return 0; } udelay(1); diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h index 25882a4dea5d..34c6ff52710e 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_0_1_sh_mask.h @@ -5452,5 +5452,7 @@ #define ROM_SW_DATA_63__ROM_SW_DATA__SHIFT 0x0 #define ROM_SW_DATA_64__ROM_SW_DATA_MASK 0xffffffff #define ROM_SW_DATA_64__ROM_SW_DATA__SHIFT 0x0 +#define CURRENT_PG_STATUS__VCE_PG_STATUS_MASK 0x00000002 +#define CURRENT_PG_STATUS__UVD_PG_STATUS_MASK 0x00000004 #endif /* SMU_7_0_1_SH_MASK_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_d.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_d.h index a9ef1562f43b..66597c64f525 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_d.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_d.h @@ -1121,5 +1121,6 @@ #define ixROM_SW_DATA_62 0xc060011c #define ixROM_SW_DATA_63 0xc0600120 #define ixROM_SW_DATA_64 0xc0600124 +#define ixCURRENT_PG_STATUS 0xc020029c #endif /* SMU_7_1_1_D_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_sh_mask.h index 2c997f7b5d13..fb06f2e2f6e6 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_1_sh_mask.h @@ -4860,5 +4860,7 @@ #define ROM_SW_DATA_63__ROM_SW_DATA__SHIFT 0x0 #define ROM_SW_DATA_64__ROM_SW_DATA_MASK 0xffffffff #define ROM_SW_DATA_64__ROM_SW_DATA__SHIFT 0x0 +#define CURRENT_PG_STATUS__VCE_PG_STATUS_MASK 0x00000002 +#define CURRENT_PG_STATUS__UVD_PG_STATUS_MASK 0x00000004 #endif /* SMU_7_1_1_SH_MASK_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h index 22dd4c2b7290..4446d43d2a8f 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_d.h @@ -1271,5 +1271,6 @@ #define ixROM_SW_DATA_62 0xc060011c #define ixROM_SW_DATA_63 0xc0600120 #define ixROM_SW_DATA_64 0xc0600124 +#define ixCURRENT_PG_STATUS 0xc020029c #endif /* SMU_7_1_2_D_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h index 518fd02e9d35..627906674fe8 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_2_sh_mask.h @@ -5830,5 +5830,7 @@ #define ROM_SW_DATA_63__ROM_SW_DATA__SHIFT 0x0 #define ROM_SW_DATA_64__ROM_SW_DATA_MASK 0xffffffff #define ROM_SW_DATA_64__ROM_SW_DATA__SHIFT 0x0 +#define CURRENT_PG_STATUS__VCE_PG_STATUS_MASK 0x00000002 +#define CURRENT_PG_STATUS__UVD_PG_STATUS_MASK 0x00000004 #endif /* SMU_7_1_2_SH_MASK_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h index eca2b851f25f..0333d880bc9e 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_d.h @@ -1244,5 +1244,5 @@ #define ixGC_CAC_ACC_CU14 0xc8 #define ixGC_CAC_ACC_CU15 0xc9 #define ixGC_CAC_OVRD_CU 0xe7 - +#define ixCURRENT_PG_STATUS 0xc020029c #endif /* SMU_7_1_3_D_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h index 1ede9e274714..654c1093d362 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/smu/smu_7_1_3_sh_mask.h @@ -6076,5 +6076,8 @@ #define GC_CAC_OVRD_CU__OVRRD_SELECT__SHIFT 0x0 #define GC_CAC_OVRD_CU__OVRRD_VALUE_MASK 0xffff0000 #define GC_CAC_OVRD_CU__OVRRD_VALUE__SHIFT 0x10 +#define CURRENT_PG_STATUS__VCE_PG_STATUS_MASK 0x00000002 +#define CURRENT_PG_STATUS__UVD_PG_STATUS_MASK 0x00000004 + #endif /* SMU_7_1_3_SH_MASK_H */ diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h index 1d26ae768147..17b9d41f3e87 100644 --- a/drivers/gpu/drm/amd/include/cgs_common.h +++ b/drivers/gpu/drm/amd/include/cgs_common.h @@ -171,6 +171,7 @@ struct cgs_firmware_info { uint32_t ucode_start_address; void *kptr; + bool is_kicker; }; struct cgs_mode_info { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c index 4b0a94cc995e..953e0c9ad7cd 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c @@ -1396,3 +1396,25 @@ int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, return 0; } + +int atomctrl_get_svi2_info(struct pp_hwmgr *hwmgr, uint8_t voltage_type, + uint8_t *svd_gpio_id, uint8_t *svc_gpio_id, + uint16_t *load_line) +{ + ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info = + (ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device); + + const ATOM_VOLTAGE_OBJECT_V3 *voltage_object; + + PP_ASSERT_WITH_CODE((NULL != voltage_info), + "Could not find Voltage Table in BIOS.", return -EINVAL); + + voltage_object = atomctrl_lookup_voltage_type_v3 + (voltage_info, voltage_type, VOLTAGE_OBJ_SVID2); + + *svd_gpio_id = voltage_object->asSVID2Obj.ucSVDGpioId; + *svc_gpio_id = voltage_object->asSVID2Obj.ucSVCGpioId; + *load_line = voltage_object->asSVID2Obj.usLoadLine_PSI; + + return 0; +} diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h index fc898afce002..e9fe2e84006b 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h @@ -311,5 +311,8 @@ extern int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_a extern int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl__avfs_parameters *param); +extern int atomctrl_get_svi2_info(struct pp_hwmgr *hwmgr, uint8_t voltage_type, + uint8_t *svd_gpio_id, uint8_t *svc_gpio_id, + uint16_t *load_line); #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index b1de9e8ccdbc..f75ee33ec5bb 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -1383,6 +1383,15 @@ static void smu7_init_dpm_defaults(struct pp_hwmgr *hwmgr) data->force_pcie_gen = PP_PCIEGenInvalid; data->ulv_supported = hwmgr->feature_mask & PP_ULV_MASK ? true : false; + if (hwmgr->chip_id == CHIP_POLARIS12 || hwmgr->smumgr->is_kicker) { + uint8_t tmp1, tmp2; + uint16_t tmp3 = 0; + atomctrl_get_svi2_info(hwmgr, VOLTAGE_TYPE_VDDC, &tmp1, &tmp2, + &tmp3); + tmp3 = (tmp3 >> 5) & 0x3; + data->vddc_phase_shed_control = ((tmp3 << 1) | (tmp3 >> 1)) & 0x3; + } + data->fast_watermark_threshold = 100; if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr, VOLTAGE_TYPE_VDDC, VOLTAGE_OBJ_SVID2)) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h index 27e7f76ad8a6..f221e17b67e7 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h @@ -268,7 +268,7 @@ struct smu7_hwmgr { uint32_t fast_watermark_threshold; /* ---- Phase Shedding ---- */ - bool vddc_phase_shed_control; + uint8_t vddc_phase_shed_control; /* ---- DI/DT ---- */ struct smu7_display_timing display_timing; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c index 3341c0fbd069..1dc31aa72781 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c @@ -477,6 +477,151 @@ static const struct gpu_pt_config_reg DIDTConfig_Polaris12[] = { { 0xFFFFFFFF } }; +static const struct gpu_pt_config_reg DIDTConfig_Polaris11_Kicker[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value Type + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_WEIGHT0_3, DIDT_SQ_WEIGHT0_3__WEIGHT0_MASK, DIDT_SQ_WEIGHT0_3__WEIGHT0__SHIFT, 0x004c, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT0_3, DIDT_SQ_WEIGHT0_3__WEIGHT1_MASK, DIDT_SQ_WEIGHT0_3__WEIGHT1__SHIFT, 0x00d0, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT0_3, DIDT_SQ_WEIGHT0_3__WEIGHT2_MASK, DIDT_SQ_WEIGHT0_3__WEIGHT2__SHIFT, 0x0069, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT0_3, DIDT_SQ_WEIGHT0_3__WEIGHT3_MASK, DIDT_SQ_WEIGHT0_3__WEIGHT3__SHIFT, 0x0048, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_WEIGHT4_7, DIDT_SQ_WEIGHT4_7__WEIGHT4_MASK, DIDT_SQ_WEIGHT4_7__WEIGHT4__SHIFT, 0x005f, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT4_7, DIDT_SQ_WEIGHT4_7__WEIGHT5_MASK, DIDT_SQ_WEIGHT4_7__WEIGHT5__SHIFT, 0x007a, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT4_7, DIDT_SQ_WEIGHT4_7__WEIGHT6_MASK, DIDT_SQ_WEIGHT4_7__WEIGHT6__SHIFT, 0x001f, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT4_7, DIDT_SQ_WEIGHT4_7__WEIGHT7_MASK, DIDT_SQ_WEIGHT4_7__WEIGHT7__SHIFT, 0x002d, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_WEIGHT8_11, DIDT_SQ_WEIGHT8_11__WEIGHT8_MASK, DIDT_SQ_WEIGHT8_11__WEIGHT8__SHIFT, 0x0088, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT8_11, DIDT_SQ_WEIGHT8_11__WEIGHT9_MASK, DIDT_SQ_WEIGHT8_11__WEIGHT9__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT8_11, DIDT_SQ_WEIGHT8_11__WEIGHT10_MASK, DIDT_SQ_WEIGHT8_11__WEIGHT10__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_WEIGHT8_11, DIDT_SQ_WEIGHT8_11__WEIGHT11_MASK, DIDT_SQ_WEIGHT8_11__WEIGHT11__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_CTRL1, DIDT_SQ_CTRL1__MIN_POWER_MASK, DIDT_SQ_CTRL1__MIN_POWER__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL1, DIDT_SQ_CTRL1__MAX_POWER_MASK, DIDT_SQ_CTRL1__MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_CTRL_OCP, DIDT_SQ_CTRL_OCP__UNUSED_0_MASK, DIDT_SQ_CTRL_OCP__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL_OCP, DIDT_SQ_CTRL_OCP__OCP_MAX_POWER_MASK, DIDT_SQ_CTRL_OCP__OCP_MAX_POWER__SHIFT, 0x00ff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__MAX_POWER_DELTA_MASK, DIDT_SQ_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3fff, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__UNUSED_0_MASK, DIDT_SQ_CTRL2__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x000f, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__UNUSED_1_MASK, DIDT_SQ_CTRL2__UNUSED_1__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__UNUSED_2_MASK, DIDT_SQ_CTRL2__UNUSED_2__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x01aa, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__UNUSED_0_MASK, DIDT_SQ_STALL_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK, DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x0dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x0dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK, DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK, DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__USE_REF_CLOCK_MASK, DIDT_SQ_CTRL0__USE_REF_CLOCK__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__PHASE_OFFSET_MASK, DIDT_SQ_CTRL0__PHASE_OFFSET__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CTRL_RST_MASK, DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x0008, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x0008, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__UNUSED_0_MASK, DIDT_SQ_CTRL0__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + /* DIDT_TD */ + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT0_MASK, DIDT_TD_WEIGHT0_3__WEIGHT0__SHIFT, 0x000a, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT1_MASK, DIDT_TD_WEIGHT0_3__WEIGHT1__SHIFT, 0x0010, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT2_MASK, DIDT_TD_WEIGHT0_3__WEIGHT2__SHIFT, 0x0017, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT3_MASK, DIDT_TD_WEIGHT0_3__WEIGHT3__SHIFT, 0x002f, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT4_MASK, DIDT_TD_WEIGHT4_7__WEIGHT4__SHIFT, 0x0046, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT5_MASK, DIDT_TD_WEIGHT4_7__WEIGHT5__SHIFT, 0x005d, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT6_MASK, DIDT_TD_WEIGHT4_7__WEIGHT6__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT7_MASK, DIDT_TD_WEIGHT4_7__WEIGHT7__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_CTRL1, DIDT_TD_CTRL1__MIN_POWER_MASK, DIDT_TD_CTRL1__MIN_POWER__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL1, DIDT_TD_CTRL1__MAX_POWER_MASK, DIDT_TD_CTRL1__MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_CTRL_OCP, DIDT_TD_CTRL_OCP__UNUSED_0_MASK, DIDT_TD_CTRL_OCP__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL_OCP, DIDT_TD_CTRL_OCP__OCP_MAX_POWER_MASK, DIDT_TD_CTRL_OCP__OCP_MAX_POWER__SHIFT, 0x00ff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__MAX_POWER_DELTA_MASK, DIDT_TD_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3fff, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__UNUSED_0_MASK, DIDT_TD_CTRL2__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x000f, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__UNUSED_1_MASK, DIDT_TD_CTRL2__UNUSED_1__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__UNUSED_2_MASK, DIDT_TD_CTRL2__UNUSED_2__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK, DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x01aa, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__UNUSED_0_MASK, DIDT_TD_STALL_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_TUNING_CTRL, DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK, DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_TUNING_CTRL, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x0dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_TUNING_CTRL, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x0dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_TUNING_CTRL, DIDT_TD_TUNING_CTRL__UNUSED_0_MASK, DIDT_TD_TUNING_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_CTRL_EN_MASK, DIDT_TD_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__USE_REF_CLOCK_MASK, DIDT_TD_CTRL0__USE_REF_CLOCK__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__PHASE_OFFSET_MASK, DIDT_TD_CTRL0__PHASE_OFFSET__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_CTRL_RST_MASK, DIDT_TD_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x0008, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x0008, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__UNUSED_0_MASK, DIDT_TD_CTRL0__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + /* DIDT_TCP */ + { ixDIDT_TCP_WEIGHT0_3, DIDT_TCP_WEIGHT0_3__WEIGHT0_MASK, DIDT_TCP_WEIGHT0_3__WEIGHT0__SHIFT, 0x0004, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_WEIGHT0_3, DIDT_TCP_WEIGHT0_3__WEIGHT1_MASK, DIDT_TCP_WEIGHT0_3__WEIGHT1__SHIFT, 0x0037, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_WEIGHT0_3, DIDT_TCP_WEIGHT0_3__WEIGHT2_MASK, DIDT_TCP_WEIGHT0_3__WEIGHT2__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_WEIGHT0_3, DIDT_TCP_WEIGHT0_3__WEIGHT3_MASK, DIDT_TCP_WEIGHT0_3__WEIGHT3__SHIFT, 0x00ff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_WEIGHT4_7, DIDT_TCP_WEIGHT4_7__WEIGHT4_MASK, DIDT_TCP_WEIGHT4_7__WEIGHT4__SHIFT, 0x0054, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_WEIGHT4_7, DIDT_TCP_WEIGHT4_7__WEIGHT5_MASK, DIDT_TCP_WEIGHT4_7__WEIGHT5__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_WEIGHT4_7, DIDT_TCP_WEIGHT4_7__WEIGHT6_MASK, DIDT_TCP_WEIGHT4_7__WEIGHT6__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_WEIGHT4_7, DIDT_TCP_WEIGHT4_7__WEIGHT7_MASK, DIDT_TCP_WEIGHT4_7__WEIGHT7__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_CTRL1, DIDT_TCP_CTRL1__MIN_POWER_MASK, DIDT_TCP_CTRL1__MIN_POWER__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL1, DIDT_TCP_CTRL1__MAX_POWER_MASK, DIDT_TCP_CTRL1__MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_CTRL_OCP, DIDT_TCP_CTRL_OCP__UNUSED_0_MASK, DIDT_TCP_CTRL_OCP__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL_OCP, DIDT_TCP_CTRL_OCP__OCP_MAX_POWER_MASK, DIDT_TCP_CTRL_OCP__OCP_MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__MAX_POWER_DELTA_MASK, DIDT_TCP_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__UNUSED_0_MASK, DIDT_TCP_CTRL2__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x0032, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__UNUSED_1_MASK, DIDT_TCP_CTRL2__UNUSED_1__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__UNUSED_2_MASK, DIDT_TCP_CTRL2__UNUSED_2__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK, DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT,0x01aa, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__UNUSED_0_MASK, DIDT_TCP_STALL_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_TUNING_CTRL, DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK, DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_TUNING_CTRL, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x3dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_TUNING_CTRL, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x3dde, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_TUNING_CTRL, DIDT_TCP_TUNING_CTRL__UNUSED_0_MASK, DIDT_TCP_TUNING_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_CTRL_EN_MASK, DIDT_TCP_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__USE_REF_CLOCK_MASK, DIDT_TCP_CTRL0__USE_REF_CLOCK__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__PHASE_OFFSET_MASK, DIDT_TCP_CTRL0__PHASE_OFFSET__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_CTRL_RST_MASK, DIDT_TCP_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x0010, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x0010, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__UNUSED_0_MASK, DIDT_TCP_CTRL0__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { 0xFFFFFFFF } /* End of list */ +}; static int smu7_enable_didt(struct pp_hwmgr *hwmgr, const bool enable) { @@ -630,7 +775,10 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr) } else if (hwmgr->chip_id == CHIP_POLARIS11) { result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11); PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result); - result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11); + if (hwmgr->smumgr->is_kicker) + result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11_Kicker); + else + result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11); PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result); } else if (hwmgr->chip_id == CHIP_POLARIS12) { result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11); diff --git a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h index 9b6531bd6350..7c318a95e0c2 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h @@ -137,6 +137,7 @@ struct pp_smumgr { uint32_t usec_timeout; bool reload_fw; const struct pp_smumgr_func *smumgr_funcs; + bool is_kicker; }; extern int smum_early_init(struct pp_instance *handle); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c index 0e26900e459e..80e2329a1b9e 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c @@ -494,6 +494,7 @@ static int polaris10_populate_ulv_level(struct pp_hwmgr *hwmgr, struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); struct phm_ppt_v1_information *table_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); + struct pp_smumgr *smumgr = hwmgr->smumgr; state->CcPwrDynRm = 0; state->CcPwrDynRm1 = 0; @@ -502,7 +503,10 @@ static int polaris10_populate_ulv_level(struct pp_hwmgr *hwmgr, state->VddcOffsetVid = (uint8_t)(table_info->us_ulv_voltage_offset * VOLTAGE_VID_OFFSET_SCALE2 / VOLTAGE_VID_OFFSET_SCALE1); - state->VddcPhase = (data->vddc_phase_shed_control) ? 0 : 1; + if (smumgr->chip_id == CHIP_POLARIS12 || smumgr->is_kicker) + state->VddcPhase = data->vddc_phase_shed_control ^ 0x3; + else + state->VddcPhase = (data->vddc_phase_shed_control) ? 0 : 1; CONVERT_FROM_HOST_TO_SMC_UL(state->CcPwrDynRm); CONVERT_FROM_HOST_TO_SMC_UL(state->CcPwrDynRm1); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c index 6749fbe26c74..35ac27681415 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c @@ -533,6 +533,8 @@ int smu7_upload_smu_firmware_image(struct pp_smumgr *smumgr) cgs_get_firmware_info(smumgr->device, smu7_convert_fw_type_to_cgs(UCODE_ID_SMU_SK), &info); + smumgr->is_kicker = info.is_kicker; + result = smu7_upload_smc_firmware_data(smumgr, info.image_size, (uint32_t *)info.kptr, SMU7_SMC_SIZE); return result; diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h index 1051181d8c0d..5a8fa1c85229 100644 --- a/drivers/gpu/drm/ast/ast_drv.h +++ b/drivers/gpu/drm/ast/ast_drv.h @@ -114,6 +114,7 @@ struct ast_private { struct ttm_bo_kmap_obj cache_kmap; int next_cursor; bool support_wide_screen; + bool DisableP2A; enum ast_tx_chip tx_chip_type; u8 dp501_maxclk; diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c index 5992ed2166ec..993909430736 100644 --- a/drivers/gpu/drm/ast/ast_main.c +++ b/drivers/gpu/drm/ast/ast_main.c @@ -124,6 +124,12 @@ static int ast_detect_chip(struct drm_device *dev, bool *need_post) } else *need_post = false; + /* Check P2A Access */ + ast->DisableP2A = true; + data = ast_read32(ast, 0xf004); + if (data != 0xFFFFFFFF) + ast->DisableP2A = false; + /* Check if we support wide screen */ switch (ast->chip) { case AST1180: @@ -140,15 +146,17 @@ static int ast_detect_chip(struct drm_device *dev, bool *need_post) ast->support_wide_screen = true; else { ast->support_wide_screen = false; - /* Read SCU7c (silicon revision register) */ - ast_write32(ast, 0xf004, 0x1e6e0000); - ast_write32(ast, 0xf000, 0x1); - data = ast_read32(ast, 0x1207c); - data &= 0x300; - if (ast->chip == AST2300 && data == 0x0) /* ast1300 */ - ast->support_wide_screen = true; - if (ast->chip == AST2400 && data == 0x100) /* ast1400 */ - ast->support_wide_screen = true; + if (ast->DisableP2A == false) { + /* Read SCU7c (silicon revision register) */ + ast_write32(ast, 0xf004, 0x1e6e0000); + ast_write32(ast, 0xf000, 0x1); + data = ast_read32(ast, 0x1207c); + data &= 0x300; + if (ast->chip == AST2300 && data == 0x0) /* ast1300 */ + ast->support_wide_screen = true; + if (ast->chip == AST2400 && data == 0x100) /* ast1400 */ + ast->support_wide_screen = true; + } } break; } @@ -216,80 +224,81 @@ static int ast_get_dram_info(struct drm_device *dev) uint32_t data, data2; uint32_t denum, num, div, ref_pll; - ast_write32(ast, 0xf004, 0x1e6e0000); - ast_write32(ast, 0xf000, 0x1); - - - ast_write32(ast, 0x10000, 0xfc600309); - - do { - if (pci_channel_offline(dev->pdev)) - return -EIO; - } while (ast_read32(ast, 0x10000) != 0x01); - data = ast_read32(ast, 0x10004); - - if (data & 0x40) + if (ast->DisableP2A) + { ast->dram_bus_width = 16; + ast->dram_type = AST_DRAM_1Gx16; + ast->mclk = 396; + } else - ast->dram_bus_width = 32; + { + ast_write32(ast, 0xf004, 0x1e6e0000); + ast_write32(ast, 0xf000, 0x1); + data = ast_read32(ast, 0x10004); - if (ast->chip == AST2300 || ast->chip == AST2400) { - switch (data & 0x03) { - case 0: - ast->dram_type = AST_DRAM_512Mx16; - break; - default: - case 1: - ast->dram_type = AST_DRAM_1Gx16; + if (data & 0x40) + ast->dram_bus_width = 16; + else + ast->dram_bus_width = 32; + + if (ast->chip == AST2300 || ast->chip == AST2400) { + switch (data & 0x03) { + case 0: + ast->dram_type = AST_DRAM_512Mx16; + break; + default: + case 1: + ast->dram_type = AST_DRAM_1Gx16; + break; + case 2: + ast->dram_type = AST_DRAM_2Gx16; + break; + case 3: + ast->dram_type = AST_DRAM_4Gx16; + break; + } + } else { + switch (data & 0x0c) { + case 0: + case 4: + ast->dram_type = AST_DRAM_512Mx16; + break; + case 8: + if (data & 0x40) + ast->dram_type = AST_DRAM_1Gx16; + else + ast->dram_type = AST_DRAM_512Mx32; + break; + case 0xc: + ast->dram_type = AST_DRAM_1Gx32; + break; + } + } + + data = ast_read32(ast, 0x10120); + data2 = ast_read32(ast, 0x10170); + if (data2 & 0x2000) + ref_pll = 14318; + else + ref_pll = 12000; + + denum = data & 0x1f; + num = (data & 0x3fe0) >> 5; + data = (data & 0xc000) >> 14; + switch (data) { + case 3: + div = 0x4; break; case 2: - ast->dram_type = AST_DRAM_2Gx16; + case 1: + div = 0x2; break; - case 3: - ast->dram_type = AST_DRAM_4Gx16; - break; - } - } else { - switch (data & 0x0c) { - case 0: - case 4: - ast->dram_type = AST_DRAM_512Mx16; - break; - case 8: - if (data & 0x40) - ast->dram_type = AST_DRAM_1Gx16; - else - ast->dram_type = AST_DRAM_512Mx32; - break; - case 0xc: - ast->dram_type = AST_DRAM_1Gx32; + default: + div = 0x1; break; } + ast->mclk = ref_pll * (num + 2) / (denum + 2) * (div * 1000); } - - data = ast_read32(ast, 0x10120); - data2 = ast_read32(ast, 0x10170); - if (data2 & 0x2000) - ref_pll = 14318; - else - ref_pll = 12000; - - denum = data & 0x1f; - num = (data & 0x3fe0) >> 5; - data = (data & 0xc000) >> 14; - switch (data) { - case 3: - div = 0x4; - break; - case 2: - case 1: - div = 0x2; - break; - default: - div = 0x1; - break; - } - ast->mclk = ref_pll * (num + 2) / (denum + 2) * (div * 1000); return 0; } diff --git a/drivers/gpu/drm/ast/ast_post.c b/drivers/gpu/drm/ast/ast_post.c index 810c51d92b99..5331ee1df086 100644 --- a/drivers/gpu/drm/ast/ast_post.c +++ b/drivers/gpu/drm/ast/ast_post.c @@ -379,12 +379,20 @@ void ast_post_gpu(struct drm_device *dev) ast_open_key(ast); ast_set_def_ext_reg(dev); - if (ast->chip == AST2300 || ast->chip == AST2400) - ast_init_dram_2300(dev); - else - ast_init_dram_reg(dev); + if (ast->DisableP2A == false) + { + if (ast->chip == AST2300 || ast->chip == AST2400) + ast_init_dram_2300(dev); + else + ast_init_dram_reg(dev); - ast_init_3rdtx(dev); + ast_init_3rdtx(dev); + } + else + { + if (ast->tx_chip_type != AST_TX_NONE) + ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xcf, 0x80); /* Enable DVO */ + } } /* AST 2300 DRAM settings */ diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index c97588a28216..a5673107db26 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -286,15 +286,15 @@ drm_atomic_get_crtc_state(struct drm_atomic_state *state, EXPORT_SYMBOL(drm_atomic_get_crtc_state); static void set_out_fence_for_crtc(struct drm_atomic_state *state, - struct drm_crtc *crtc, s64 __user *fence_ptr) + struct drm_crtc *crtc, s32 __user *fence_ptr) { state->crtcs[drm_crtc_index(crtc)].out_fence_ptr = fence_ptr; } -static s64 __user *get_out_fence_for_crtc(struct drm_atomic_state *state, +static s32 __user *get_out_fence_for_crtc(struct drm_atomic_state *state, struct drm_crtc *crtc) { - s64 __user *fence_ptr; + s32 __user *fence_ptr; fence_ptr = state->crtcs[drm_crtc_index(crtc)].out_fence_ptr; state->crtcs[drm_crtc_index(crtc)].out_fence_ptr = NULL; @@ -505,7 +505,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc, state->color_mgmt_changed |= replaced; return ret; } else if (property == config->prop_out_fence_ptr) { - s64 __user *fence_ptr = u64_to_user_ptr(val); + s32 __user *fence_ptr = u64_to_user_ptr(val); if (!fence_ptr) return 0; @@ -1902,7 +1902,7 @@ EXPORT_SYMBOL(drm_atomic_clean_old_fb); */ struct drm_out_fence_state { - s64 __user *out_fence_ptr; + s32 __user *out_fence_ptr; struct sync_file *sync_file; int fd; }; @@ -1939,7 +1939,7 @@ static int prepare_crtc_signaling(struct drm_device *dev, return 0; for_each_crtc_in_state(state, crtc, crtc_state, i) { - u64 __user *fence_ptr; + s32 __user *fence_ptr; fence_ptr = get_out_fence_for_crtc(crtc_state->state, crtc); @@ -2019,13 +2019,16 @@ static void complete_crtc_signaling(struct drm_device *dev, } for_each_crtc_in_state(state, crtc, crtc_state, i) { + struct drm_pending_vblank_event *event = crtc_state->event; /* - * TEST_ONLY and PAGE_FLIP_EVENT are mutually - * exclusive, if they weren't, this code should be - * called on success for TEST_ONLY too. + * Free the allocated event. drm_atomic_helper_setup_commit + * can allocate an event too, so only free it if it's ours + * to prevent a double free in drm_atomic_state_clear. */ - if (crtc_state->event) - drm_event_cancel_free(dev, &crtc_state->event->base); + if (event && (event->base.fence || event->base.file_priv)) { + drm_event_cancel_free(dev, &event->base); + crtc_state->event = NULL; + } } if (!fence_state) diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c index e4d2c8a49076..45464c8b797d 100644 --- a/drivers/gpu/drm/drm_connector.c +++ b/drivers/gpu/drm/drm_connector.c @@ -378,6 +378,9 @@ int drm_connector_register(struct drm_connector *connector) { int ret = 0; + if (!connector->dev->registered) + return 0; + mutex_lock(&connector->mutex); if (connector->registered) goto unlock; diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c index 45ce224688ce..b5c6bb46a425 100644 --- a/drivers/gpu/drm/drm_drv.c +++ b/drivers/gpu/drm/drm_drv.c @@ -776,6 +776,8 @@ int drm_dev_register(struct drm_device *dev, unsigned long flags) if (ret) goto err_minors; + dev->registered = true; + if (dev->driver->load) { ret = dev->driver->load(dev, flags); if (ret) @@ -823,6 +825,8 @@ void drm_dev_unregister(struct drm_device *dev) drm_lastclose(dev); + dev->registered = false; + if (drm_core_check_feature(dev, DRIVER_MODESET)) drm_modeset_unregister_all(dev); diff --git a/drivers/gpu/drm/drm_panel.c b/drivers/gpu/drm/drm_panel.c index 3dfe3c886502..308d442a531b 100644 --- a/drivers/gpu/drm/drm_panel.c +++ b/drivers/gpu/drm/drm_panel.c @@ -137,7 +137,7 @@ EXPORT_SYMBOL(drm_panel_detach); * Return: A pointer to the panel registered for the specified device tree * node or NULL if no panel matching the device tree node can be found. */ -struct drm_panel *of_drm_find_panel(struct device_node *np) +struct drm_panel *of_drm_find_panel(const struct device_node *np) { struct drm_panel *panel; diff --git a/drivers/gpu/drm/i915/gvt/aperture_gm.c b/drivers/gpu/drm/i915/gvt/aperture_gm.c index 7311aeab16f7..3b6caaca9751 100644 --- a/drivers/gpu/drm/i915/gvt/aperture_gm.c +++ b/drivers/gpu/drm/i915/gvt/aperture_gm.c @@ -49,20 +49,21 @@ static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm) if (high_gm) { node = &vgpu->gm.high_gm_node; size = vgpu_hidden_sz(vgpu); - start = gvt_hidden_gmadr_base(gvt); - end = gvt_hidden_gmadr_end(gvt); + start = ALIGN(gvt_hidden_gmadr_base(gvt), I915_GTT_PAGE_SIZE); + end = ALIGN(gvt_hidden_gmadr_end(gvt), I915_GTT_PAGE_SIZE); flags = PIN_HIGH; } else { node = &vgpu->gm.low_gm_node; size = vgpu_aperture_sz(vgpu); - start = gvt_aperture_gmadr_base(gvt); - end = gvt_aperture_gmadr_end(gvt); + start = ALIGN(gvt_aperture_gmadr_base(gvt), I915_GTT_PAGE_SIZE); + end = ALIGN(gvt_aperture_gmadr_end(gvt), I915_GTT_PAGE_SIZE); flags = PIN_MAPPABLE; } mutex_lock(&dev_priv->drm.struct_mutex); ret = i915_gem_gtt_insert(&dev_priv->ggtt.base, node, - size, 4096, I915_COLOR_UNEVICTABLE, + size, I915_GTT_PAGE_SIZE, + I915_COLOR_UNEVICTABLE, start, end, flags); mutex_unlock(&dev_priv->drm.struct_mutex); if (ret) @@ -254,7 +255,7 @@ static int alloc_resource(struct intel_vgpu *vgpu, if (request > avail) goto no_enough_resource; - vgpu_aperture_sz(vgpu) = request; + vgpu_aperture_sz(vgpu) = ALIGN(request, I915_GTT_PAGE_SIZE); item = "high GM space"; max = gvt_hidden_sz(gvt) - HOST_HIGH_GM_SIZE; @@ -265,7 +266,7 @@ static int alloc_resource(struct intel_vgpu *vgpu, if (request > avail) goto no_enough_resource; - vgpu_hidden_sz(vgpu) = request; + vgpu_hidden_sz(vgpu) = ALIGN(request, I915_GTT_PAGE_SIZE); item = "fence"; max = gvt_fence_sz(gvt) - HOST_FENCE; diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index c3d44c03157c..81076d8ec41a 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -481,7 +481,6 @@ struct parser_exec_state { (s->vgpu->gvt->device_info.gmadr_bytes_in_cmd >> 2) static unsigned long bypass_scan_mask = 0; -static bool bypass_batch_buffer_scan = true; /* ring ALL, type = 0 */ static struct sub_op_bits sub_op_mi[] = { @@ -1135,6 +1134,8 @@ static int skl_decode_mi_display_flip(struct parser_exec_state *s, u32 dword2 = cmd_val(s, 2); u32 plane = (dword0 & GENMASK(12, 8)) >> 8; + info->plane = PRIMARY_PLANE; + switch (plane) { case MI_DISPLAY_FLIP_SKL_PLANE_1_A: info->pipe = PIPE_A; @@ -1148,12 +1149,28 @@ static int skl_decode_mi_display_flip(struct parser_exec_state *s, info->pipe = PIPE_C; info->event = PRIMARY_C_FLIP_DONE; break; + + case MI_DISPLAY_FLIP_SKL_PLANE_2_A: + info->pipe = PIPE_A; + info->event = SPRITE_A_FLIP_DONE; + info->plane = SPRITE_PLANE; + break; + case MI_DISPLAY_FLIP_SKL_PLANE_2_B: + info->pipe = PIPE_B; + info->event = SPRITE_B_FLIP_DONE; + info->plane = SPRITE_PLANE; + break; + case MI_DISPLAY_FLIP_SKL_PLANE_2_C: + info->pipe = PIPE_C; + info->event = SPRITE_C_FLIP_DONE; + info->plane = SPRITE_PLANE; + break; + default: gvt_err("unknown plane code %d\n", plane); return -EINVAL; } - info->pipe = PRIMARY_PLANE; info->stride_val = (dword1 & GENMASK(15, 6)) >> 6; info->tile_val = (dword1 & GENMASK(2, 0)); info->surf_val = (dword2 & GENMASK(31, 12)) >> 12; @@ -1525,9 +1542,6 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s) { struct intel_gvt *gvt = s->vgpu->gvt; - if (bypass_batch_buffer_scan) - return 0; - if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) { /* BDW decides privilege based on address space */ if (cmd_val(s, 0) & (1 << 8)) diff --git a/drivers/gpu/drm/i915/gvt/display.c b/drivers/gpu/drm/i915/gvt/display.c index c0c884aeb30e..6d8fde880c39 100644 --- a/drivers/gpu/drm/i915/gvt/display.c +++ b/drivers/gpu/drm/i915/gvt/display.c @@ -83,7 +83,7 @@ static int pipe_is_enabled(struct intel_vgpu *vgpu, int pipe) return 0; } -/* EDID with 1024x768 as its resolution */ +/* EDID with 1920x1200 as its resolution */ static unsigned char virtual_dp_monitor_edid[] = { /*Header*/ 0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00, @@ -97,11 +97,16 @@ static unsigned char virtual_dp_monitor_edid[] = { 0xfc, 0x81, 0xa4, 0x55, 0x4d, 0x9d, 0x25, 0x12, 0x50, 0x54, /* Established Timings: maximum resolution is 1024x768 */ 0x21, 0x08, 0x00, - /* Standard Timings. All invalid */ - 0x00, 0xc0, 0x00, 0xc0, 0x00, 0x40, 0x00, 0x80, 0x00, 0x00, - 0x00, 0x40, 0x00, 0x00, 0x00, 0x01, - /* 18 Byte Data Blocks 1: invalid */ - 0x00, 0x00, 0x80, 0xa0, 0x70, 0xb0, + /* + * Standard Timings. + * below new resolutions can be supported: + * 1920x1080, 1280x720, 1280x960, 1280x1024, + * 1440x900, 1600x1200, 1680x1050 + */ + 0xd1, 0xc0, 0x81, 0xc0, 0x81, 0x40, 0x81, 0x80, 0x95, 0x00, + 0xa9, 0x40, 0xb3, 0x00, 0x01, 0x01, + /* 18 Byte Data Blocks 1: max resolution is 1920x1200 */ + 0x28, 0x3c, 0x80, 0xa0, 0x70, 0xb0, 0x23, 0x40, 0x30, 0x20, 0x36, 0x00, 0x06, 0x44, 0x21, 0x00, 0x00, 0x1a, /* 18 Byte Data Blocks 2: invalid */ 0x00, 0x00, 0x00, 0xfd, 0x00, 0x18, 0x3c, 0x18, 0x50, 0x11, 0x00, 0x0a, @@ -115,7 +120,7 @@ static unsigned char virtual_dp_monitor_edid[] = { /* Extension Block Count */ 0x00, /* Checksum */ - 0xef, + 0x45, }; #define DPCD_HEADER_SIZE 0xb @@ -328,3 +333,15 @@ int intel_vgpu_init_display(struct intel_vgpu *vgpu) else return setup_virtual_dp_monitor(vgpu, PORT_B, GVT_DP_B); } + +/** + * intel_vgpu_reset_display- reset vGPU virtual display emulation + * @vgpu: a vGPU + * + * This function is used to reset vGPU virtual display emulation stuffs + * + */ +void intel_vgpu_reset_display(struct intel_vgpu *vgpu) +{ + emulate_monitor_status_change(vgpu); +} diff --git a/drivers/gpu/drm/i915/gvt/display.h b/drivers/gpu/drm/i915/gvt/display.h index 7a60cb848268..8b234ea961f6 100644 --- a/drivers/gpu/drm/i915/gvt/display.h +++ b/drivers/gpu/drm/i915/gvt/display.h @@ -158,6 +158,7 @@ void intel_gvt_emulate_vblank(struct intel_gvt *gvt); void intel_gvt_check_vblank_emulation(struct intel_gvt *gvt); int intel_vgpu_init_display(struct intel_vgpu *vgpu); +void intel_vgpu_reset_display(struct intel_vgpu *vgpu); void intel_vgpu_clean_display(struct intel_vgpu *vgpu); #endif diff --git a/drivers/gpu/drm/i915/gvt/execlist.c b/drivers/gpu/drm/i915/gvt/execlist.c index f32bb6f6495c..46eb9fd3c03f 100644 --- a/drivers/gpu/drm/i915/gvt/execlist.c +++ b/drivers/gpu/drm/i915/gvt/execlist.c @@ -364,58 +364,30 @@ static void free_workload(struct intel_vgpu_workload *workload) #define get_desc_from_elsp_dwords(ed, i) \ ((struct execlist_ctx_descriptor_format *)&((ed)->data[i * 2])) - -#define BATCH_BUFFER_ADDR_MASK ((1UL << 32) - (1U << 2)) -#define BATCH_BUFFER_ADDR_HIGH_MASK ((1UL << 16) - (1U)) -static int set_gma_to_bb_cmd(struct intel_shadow_bb_entry *entry_obj, - unsigned long add, int gmadr_bytes) -{ - if (WARN_ON(gmadr_bytes != 4 && gmadr_bytes != 8)) - return -1; - - *((u32 *)(entry_obj->bb_start_cmd_va + (1 << 2))) = add & - BATCH_BUFFER_ADDR_MASK; - if (gmadr_bytes == 8) { - *((u32 *)(entry_obj->bb_start_cmd_va + (2 << 2))) = - add & BATCH_BUFFER_ADDR_HIGH_MASK; - } - - return 0; -} - static void prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload) { - int gmadr_bytes = workload->vgpu->gvt->device_info.gmadr_bytes_in_cmd; + const int gmadr_bytes = workload->vgpu->gvt->device_info.gmadr_bytes_in_cmd; + struct intel_shadow_bb_entry *entry_obj; /* pin the gem object to ggtt */ - if (!list_empty(&workload->shadow_bb)) { - struct intel_shadow_bb_entry *entry_obj = - list_first_entry(&workload->shadow_bb, - struct intel_shadow_bb_entry, - list); - struct intel_shadow_bb_entry *temp; + list_for_each_entry(entry_obj, &workload->shadow_bb, list) { + struct i915_vma *vma; - list_for_each_entry_safe(entry_obj, temp, &workload->shadow_bb, - list) { - struct i915_vma *vma; - - vma = i915_gem_object_ggtt_pin(entry_obj->obj, NULL, 0, - 4, 0); - if (IS_ERR(vma)) { - gvt_err("Cannot pin\n"); - return; - } - - /* FIXME: we are not tracking our pinned VMA leaving it - * up to the core to fix up the stray pin_count upon - * free. - */ - - /* update the relocate gma with shadow batch buffer*/ - set_gma_to_bb_cmd(entry_obj, - i915_ggtt_offset(vma), - gmadr_bytes); + vma = i915_gem_object_ggtt_pin(entry_obj->obj, NULL, 0, 4, 0); + if (IS_ERR(vma)) { + gvt_err("Cannot pin\n"); + return; } + + /* FIXME: we are not tracking our pinned VMA leaving it + * up to the core to fix up the stray pin_count upon + * free. + */ + + /* update the relocate gma with shadow batch buffer*/ + entry_obj->bb_start_cmd_va[1] = i915_ggtt_offset(vma); + if (gmadr_bytes == 8) + entry_obj->bb_start_cmd_va[2] = 0; } } @@ -515,7 +487,7 @@ static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload) static void release_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) { - if (wa_ctx->indirect_ctx.size == 0) + if (!wa_ctx->indirect_ctx.obj) return; i915_gem_object_unpin_map(wa_ctx->indirect_ctx.obj); @@ -826,7 +798,7 @@ int intel_vgpu_init_execlist(struct intel_vgpu *vgpu) INIT_LIST_HEAD(&vgpu->workload_q_head[i]); } - vgpu->workloads = kmem_cache_create("gvt-g vgpu workload", + vgpu->workloads = kmem_cache_create("gvt-g_vgpu_workload", sizeof(struct intel_vgpu_workload), 0, SLAB_HWCACHE_ALIGN, NULL); diff --git a/drivers/gpu/drm/i915/gvt/firmware.c b/drivers/gpu/drm/i915/gvt/firmware.c index 2fae2a2ca96f..1cb29b2d7dc6 100644 --- a/drivers/gpu/drm/i915/gvt/firmware.c +++ b/drivers/gpu/drm/i915/gvt/firmware.c @@ -48,31 +48,6 @@ struct gvt_firmware_header { unsigned char data[1]; }; -#define RD(offset) (readl(mmio + offset.reg)) -#define WR(v, offset) (writel(v, mmio + offset.reg)) - -static void bdw_forcewake_get(void __iomem *mmio) -{ - WR(_MASKED_BIT_DISABLE(0xffff), FORCEWAKE_MT); - - RD(ECOBUS); - - if (wait_for((RD(FORCEWAKE_ACK_HSW) & FORCEWAKE_KERNEL) == 0, 50)) - gvt_err("fail to wait forcewake idle\n"); - - WR(_MASKED_BIT_ENABLE(FORCEWAKE_KERNEL), FORCEWAKE_MT); - - if (wait_for((RD(FORCEWAKE_ACK_HSW) & FORCEWAKE_KERNEL), 50)) - gvt_err("fail to wait forcewake ack\n"); - - if (wait_for((RD(GEN6_GT_THREAD_STATUS_REG) & - GEN6_GT_THREAD_STATUS_CORE_MASK) == 0, 50)) - gvt_err("fail to wait c0 wake up\n"); -} - -#undef RD -#undef WR - #define dev_to_drm_minor(d) dev_get_drvdata((d)) static ssize_t @@ -91,9 +66,9 @@ static struct bin_attribute firmware_attr = { .mmap = NULL, }; -static int expose_firmware_sysfs(struct intel_gvt *gvt, - void __iomem *mmio) +static int expose_firmware_sysfs(struct intel_gvt *gvt) { + struct drm_i915_private *dev_priv = gvt->dev_priv; struct intel_gvt_device_info *info = &gvt->device_info; struct pci_dev *pdev = gvt->dev_priv->drm.pdev; struct intel_gvt_mmio_info *e; @@ -132,7 +107,7 @@ static int expose_firmware_sysfs(struct intel_gvt *gvt, for (j = 0; j < e->length; j += 4) *(u32 *)(p + e->offset + j) = - readl(mmio + e->offset + j); + I915_READ_NOTRACE(_MMIO(e->offset + j)); } memcpy(gvt->firmware.mmio, p, info->mmio_size); @@ -235,7 +210,6 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt) struct gvt_firmware_header *h; const struct firmware *fw; char *path; - void __iomem *mmio; void *mem; int ret; @@ -260,17 +234,6 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt) firmware->mmio = mem; - mmio = pci_iomap(pdev, info->mmio_bar, info->mmio_size); - if (!mmio) { - kfree(path); - kfree(firmware->cfg_space); - kfree(firmware->mmio); - return -EINVAL; - } - - if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) - bdw_forcewake_get(mmio); - sprintf(path, "%s/vid_0x%04x_did_0x%04x_rid_0x%04x.golden_hw_state", GVT_FIRMWARE_PATH, pdev->vendor, pdev->device, pdev->revision); @@ -300,13 +263,11 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt) release_firmware(fw); firmware->firmware_loaded = true; - pci_iounmap(pdev, mmio); return 0; out_free_fw: release_firmware(fw); expose_firmware: - expose_firmware_sysfs(gvt, mmio); - pci_iounmap(pdev, mmio); + expose_firmware_sysfs(gvt); return 0; } diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index 47dec4acf7ff..28c92346db0e 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -606,21 +606,33 @@ struct intel_vgpu_guest_page *intel_vgpu_find_guest_page( static inline int init_shadow_page(struct intel_vgpu *vgpu, struct intel_vgpu_shadow_page *p, int type) { + struct device *kdev = &vgpu->gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr; + + daddr = dma_map_page(kdev, p->page, 0, 4096, PCI_DMA_BIDIRECTIONAL); + if (dma_mapping_error(kdev, daddr)) { + gvt_err("fail to map dma addr\n"); + return -EINVAL; + } + p->vaddr = page_address(p->page); p->type = type; INIT_HLIST_NODE(&p->node); - p->mfn = intel_gvt_hypervisor_virt_to_mfn(p->vaddr); - if (p->mfn == INTEL_GVT_INVALID_ADDR) - return -EFAULT; - + p->mfn = daddr >> GTT_PAGE_SHIFT; hash_add(vgpu->gtt.shadow_page_hash_table, &p->node, p->mfn); return 0; } -static inline void clean_shadow_page(struct intel_vgpu_shadow_page *p) +static inline void clean_shadow_page(struct intel_vgpu *vgpu, + struct intel_vgpu_shadow_page *p) { + struct device *kdev = &vgpu->gvt->dev_priv->drm.pdev->dev; + + dma_unmap_page(kdev, p->mfn << GTT_PAGE_SHIFT, 4096, + PCI_DMA_BIDIRECTIONAL); + if (!hlist_unhashed(&p->node)) hash_del(&p->node); } @@ -670,7 +682,7 @@ static void ppgtt_free_shadow_page(struct intel_vgpu_ppgtt_spt *spt) { trace_spt_free(spt->vgpu->id, spt, spt->shadow_page.type); - clean_shadow_page(&spt->shadow_page); + clean_shadow_page(spt->vgpu, &spt->shadow_page); intel_vgpu_clean_guest_page(spt->vgpu, &spt->guest_page); list_del_init(&spt->post_shadow_list); @@ -1875,8 +1887,9 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu, int page_entry_num = GTT_PAGE_SIZE >> vgpu->gvt->device_info.gtt_entry_size_shift; void *scratch_pt; - unsigned long mfn; int i; + struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr; if (WARN_ON(type < GTT_TYPE_PPGTT_PTE_PT || type >= GTT_TYPE_MAX)) return -EINVAL; @@ -1887,16 +1900,18 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu, return -ENOMEM; } - mfn = intel_gvt_hypervisor_virt_to_mfn(scratch_pt); - if (mfn == INTEL_GVT_INVALID_ADDR) { - gvt_err("fail to translate vaddr:0x%lx\n", (unsigned long)scratch_pt); - free_page((unsigned long)scratch_pt); - return -EFAULT; + daddr = dma_map_page(dev, virt_to_page(scratch_pt), 0, + 4096, PCI_DMA_BIDIRECTIONAL); + if (dma_mapping_error(dev, daddr)) { + gvt_err("fail to dmamap scratch_pt\n"); + __free_page(virt_to_page(scratch_pt)); + return -ENOMEM; } - gtt->scratch_pt[type].page_mfn = mfn; + gtt->scratch_pt[type].page_mfn = + (unsigned long)(daddr >> GTT_PAGE_SHIFT); gtt->scratch_pt[type].page = virt_to_page(scratch_pt); gvt_dbg_mm("vgpu%d create scratch_pt: type %d mfn=0x%lx\n", - vgpu->id, type, mfn); + vgpu->id, type, gtt->scratch_pt[type].page_mfn); /* Build the tree by full filled the scratch pt with the entries which * point to the next level scratch pt or scratch page. The @@ -1930,9 +1945,14 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu, static int release_scratch_page_tree(struct intel_vgpu *vgpu) { int i; + struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr; for (i = GTT_TYPE_PPGTT_PTE_PT; i < GTT_TYPE_MAX; i++) { if (vgpu->gtt.scratch_pt[i].page != NULL) { + daddr = (dma_addr_t)(vgpu->gtt.scratch_pt[i].page_mfn << + GTT_PAGE_SHIFT); + dma_unmap_page(dev, daddr, 4096, PCI_DMA_BIDIRECTIONAL); __free_page(vgpu->gtt.scratch_pt[i].page); vgpu->gtt.scratch_pt[i].page = NULL; vgpu->gtt.scratch_pt[i].page_mfn = 0; @@ -2192,6 +2212,8 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt) { int ret; void *page; + struct device *dev = &gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr; gvt_dbg_core("init gtt\n"); @@ -2209,14 +2231,16 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt) gvt_err("fail to allocate scratch ggtt page\n"); return -ENOMEM; } - gvt->gtt.scratch_ggtt_page = virt_to_page(page); - gvt->gtt.scratch_ggtt_mfn = intel_gvt_hypervisor_virt_to_mfn(page); - if (gvt->gtt.scratch_ggtt_mfn == INTEL_GVT_INVALID_ADDR) { - gvt_err("fail to translate scratch ggtt page\n"); - __free_page(gvt->gtt.scratch_ggtt_page); - return -EFAULT; + daddr = dma_map_page(dev, virt_to_page(page), 0, + 4096, PCI_DMA_BIDIRECTIONAL); + if (dma_mapping_error(dev, daddr)) { + gvt_err("fail to dmamap scratch ggtt page\n"); + __free_page(virt_to_page(page)); + return -ENOMEM; } + gvt->gtt.scratch_ggtt_page = virt_to_page(page); + gvt->gtt.scratch_ggtt_mfn = (unsigned long)(daddr >> GTT_PAGE_SHIFT); if (enable_out_of_sync) { ret = setup_spt_oos(gvt); @@ -2239,6 +2263,12 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt) */ void intel_gvt_clean_gtt(struct intel_gvt *gvt) { + struct device *dev = &gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr = (dma_addr_t)(gvt->gtt.scratch_ggtt_mfn << + GTT_PAGE_SHIFT); + + dma_unmap_page(dev, daddr, 4096, PCI_DMA_BIDIRECTIONAL); + __free_page(gvt->gtt.scratch_ggtt_page); if (enable_out_of_sync) diff --git a/drivers/gpu/drm/i915/gvt/gvt.c b/drivers/gpu/drm/i915/gvt/gvt.c index e6bf5c533fbe..3b9d59e457ba 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.c +++ b/drivers/gpu/drm/i915/gvt/gvt.c @@ -68,8 +68,6 @@ static const struct intel_gvt_ops intel_gvt_ops = { */ int intel_gvt_init_host(void) { - int ret; - if (intel_gvt_host.initialized) return 0; @@ -96,11 +94,6 @@ int intel_gvt_init_host(void) if (!intel_gvt_host.mpt) return -EINVAL; - /* Try to detect if we're running in host instead of VM. */ - ret = intel_gvt_hypervisor_detect_host(); - if (ret) - return -ENODEV; - gvt_dbg_core("Running with hypervisor %s in host mode\n", supported_hypervisors[intel_gvt_host.hypervisor_type]); diff --git a/drivers/gpu/drm/i915/gvt/hypercall.h b/drivers/gpu/drm/i915/gvt/hypercall.h index 30e543f5a703..df7f33abd393 100644 --- a/drivers/gpu/drm/i915/gvt/hypercall.h +++ b/drivers/gpu/drm/i915/gvt/hypercall.h @@ -38,7 +38,6 @@ * both Xen and KVM by providing dedicated hypervisor-related MPT modules. */ struct intel_gvt_mpt { - int (*detect_host)(void); int (*host_init)(struct device *dev, void *gvt, const void *ops); void (*host_exit)(struct device *dev, void *gvt); int (*attach_vgpu)(void *vgpu, unsigned long *handle); diff --git a/drivers/gpu/drm/i915/gvt/interrupt.c b/drivers/gpu/drm/i915/gvt/interrupt.c index f7be02ac4be1..92bb247e3478 100644 --- a/drivers/gpu/drm/i915/gvt/interrupt.c +++ b/drivers/gpu/drm/i915/gvt/interrupt.c @@ -176,26 +176,15 @@ int intel_vgpu_reg_imr_handler(struct intel_vgpu *vgpu, { struct intel_gvt *gvt = vgpu->gvt; struct intel_gvt_irq_ops *ops = gvt->irq.ops; - u32 changed, masked, unmasked; u32 imr = *(u32 *)p_data; - gvt_dbg_irq("write IMR %x with val %x\n", - reg, imr); - - gvt_dbg_irq("old vIMR %x\n", vgpu_vreg(vgpu, reg)); - - /* figure out newly masked/unmasked bits */ - changed = vgpu_vreg(vgpu, reg) ^ imr; - masked = (vgpu_vreg(vgpu, reg) & changed) ^ changed; - unmasked = masked ^ changed; - - gvt_dbg_irq("changed %x, masked %x, unmasked %x\n", - changed, masked, unmasked); + gvt_dbg_irq("write IMR %x, new %08x, old %08x, changed %08x\n", + reg, imr, vgpu_vreg(vgpu, reg), vgpu_vreg(vgpu, reg) ^ imr); vgpu_vreg(vgpu, reg) = imr; ops->check_pending_irq(vgpu); - gvt_dbg_irq("IRQ: new vIMR %x\n", vgpu_vreg(vgpu, reg)); + return 0; } @@ -217,14 +206,11 @@ int intel_vgpu_reg_master_irq_handler(struct intel_vgpu *vgpu, { struct intel_gvt *gvt = vgpu->gvt; struct intel_gvt_irq_ops *ops = gvt->irq.ops; - u32 changed, enabled, disabled; u32 ier = *(u32 *)p_data; u32 virtual_ier = vgpu_vreg(vgpu, reg); - gvt_dbg_irq("write master irq reg %x with val %x\n", - reg, ier); - - gvt_dbg_irq("old vreg %x\n", vgpu_vreg(vgpu, reg)); + gvt_dbg_irq("write MASTER_IRQ %x, new %08x, old %08x, changed %08x\n", + reg, ier, virtual_ier, virtual_ier ^ ier); /* * GEN8_MASTER_IRQ is a special irq register, @@ -236,16 +222,8 @@ int intel_vgpu_reg_master_irq_handler(struct intel_vgpu *vgpu, vgpu_vreg(vgpu, reg) &= ~GEN8_MASTER_IRQ_CONTROL; vgpu_vreg(vgpu, reg) |= ier; - /* figure out newly enabled/disable bits */ - changed = virtual_ier ^ ier; - enabled = (virtual_ier & changed) ^ changed; - disabled = enabled ^ changed; - - gvt_dbg_irq("changed %x, enabled %x, disabled %x\n", - changed, enabled, disabled); - ops->check_pending_irq(vgpu); - gvt_dbg_irq("new vreg %x\n", vgpu_vreg(vgpu, reg)); + return 0; } @@ -268,21 +246,11 @@ int intel_vgpu_reg_ier_handler(struct intel_vgpu *vgpu, struct intel_gvt *gvt = vgpu->gvt; struct intel_gvt_irq_ops *ops = gvt->irq.ops; struct intel_gvt_irq_info *info; - u32 changed, enabled, disabled; u32 ier = *(u32 *)p_data; - gvt_dbg_irq("write IER %x with val %x\n", - reg, ier); + gvt_dbg_irq("write IER %x, new %08x, old %08x, changed %08x\n", + reg, ier, vgpu_vreg(vgpu, reg), vgpu_vreg(vgpu, reg) ^ ier); - gvt_dbg_irq("old vIER %x\n", vgpu_vreg(vgpu, reg)); - - /* figure out newly enabled/disable bits */ - changed = vgpu_vreg(vgpu, reg) ^ ier; - enabled = (vgpu_vreg(vgpu, reg) & changed) ^ changed; - disabled = enabled ^ changed; - - gvt_dbg_irq("changed %x, enabled %x, disabled %x\n", - changed, enabled, disabled); vgpu_vreg(vgpu, reg) = ier; info = regbase_to_irq_info(gvt, ier_to_regbase(reg)); @@ -293,7 +261,7 @@ int intel_vgpu_reg_ier_handler(struct intel_vgpu *vgpu, update_upstream_irq(vgpu, info); ops->check_pending_irq(vgpu); - gvt_dbg_irq("new vIER %x\n", vgpu_vreg(vgpu, reg)); + return 0; } @@ -317,7 +285,8 @@ int intel_vgpu_reg_iir_handler(struct intel_vgpu *vgpu, unsigned int reg, iir_to_regbase(reg)); u32 iir = *(u32 *)p_data; - gvt_dbg_irq("write IIR %x with val %x\n", reg, iir); + gvt_dbg_irq("write IIR %x, new %08x, old %08x, changed %08x\n", + reg, iir, vgpu_vreg(vgpu, reg), vgpu_vreg(vgpu, reg) ^ iir); if (WARN_ON(!info)) return -EINVAL; @@ -619,6 +588,10 @@ static void gen8_init_irq( SET_BIT_INFO(irq, 3, PRIMARY_A_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_A); SET_BIT_INFO(irq, 3, PRIMARY_B_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_B); SET_BIT_INFO(irq, 3, PRIMARY_C_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_C); + + SET_BIT_INFO(irq, 4, SPRITE_A_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_A); + SET_BIT_INFO(irq, 4, SPRITE_B_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_B); + SET_BIT_INFO(irq, 4, SPRITE_C_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_C); } /* GEN8 interrupt PCU events */ diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c index 0c9234a87a20..0f7f5d97f582 100644 --- a/drivers/gpu/drm/i915/gvt/kvmgt.c +++ b/drivers/gpu/drm/i915/gvt/kvmgt.c @@ -77,7 +77,7 @@ struct kvmgt_guest_info { struct gvt_dma { struct rb_node node; gfn_t gfn; - kvm_pfn_t pfn; + unsigned long iova; }; static inline bool handle_valid(unsigned long handle) @@ -89,6 +89,35 @@ static int kvmgt_guest_init(struct mdev_device *mdev); static void intel_vgpu_release_work(struct work_struct *work); static bool kvmgt_guest_exit(struct kvmgt_guest_info *info); +static int gvt_dma_map_iova(struct intel_vgpu *vgpu, kvm_pfn_t pfn, + unsigned long *iova) +{ + struct page *page; + struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr; + + page = pfn_to_page(pfn); + if (is_error_page(page)) + return -EFAULT; + + daddr = dma_map_page(dev, page, 0, PAGE_SIZE, + PCI_DMA_BIDIRECTIONAL); + if (dma_mapping_error(dev, daddr)) + return -ENOMEM; + + *iova = (unsigned long)(daddr >> PAGE_SHIFT); + return 0; +} + +static void gvt_dma_unmap_iova(struct intel_vgpu *vgpu, unsigned long iova) +{ + struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev; + dma_addr_t daddr; + + daddr = (dma_addr_t)(iova << PAGE_SHIFT); + dma_unmap_page(dev, daddr, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); +} + static struct gvt_dma *__gvt_cache_find(struct intel_vgpu *vgpu, gfn_t gfn) { struct rb_node *node = vgpu->vdev.cache.rb_node; @@ -111,21 +140,22 @@ static struct gvt_dma *__gvt_cache_find(struct intel_vgpu *vgpu, gfn_t gfn) return ret; } -static kvm_pfn_t gvt_cache_find(struct intel_vgpu *vgpu, gfn_t gfn) +static unsigned long gvt_cache_find(struct intel_vgpu *vgpu, gfn_t gfn) { struct gvt_dma *entry; - kvm_pfn_t pfn; + unsigned long iova; mutex_lock(&vgpu->vdev.cache_lock); entry = __gvt_cache_find(vgpu, gfn); - pfn = (entry == NULL) ? 0 : entry->pfn; + iova = (entry == NULL) ? INTEL_GVT_INVALID_ADDR : entry->iova; mutex_unlock(&vgpu->vdev.cache_lock); - return pfn; + return iova; } -static void gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn, kvm_pfn_t pfn) +static void gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn, + unsigned long iova) { struct gvt_dma *new, *itr; struct rb_node **link = &vgpu->vdev.cache.rb_node, *parent = NULL; @@ -135,7 +165,7 @@ static void gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn, kvm_pfn_t pfn) return; new->gfn = gfn; - new->pfn = pfn; + new->iova = iova; mutex_lock(&vgpu->vdev.cache_lock); while (*link) { @@ -182,6 +212,7 @@ static void gvt_cache_remove(struct intel_vgpu *vgpu, gfn_t gfn) } g1 = gfn; + gvt_dma_unmap_iova(vgpu, this->iova); rc = vfio_unpin_pages(dev, &g1, 1); WARN_ON(rc != 1); __gvt_cache_remove_entry(vgpu, this); @@ -204,6 +235,7 @@ static void gvt_cache_destroy(struct intel_vgpu *vgpu) mutex_lock(&vgpu->vdev.cache_lock); while ((node = rb_first(&vgpu->vdev.cache))) { dma = rb_entry(node, struct gvt_dma, node); + gvt_dma_unmap_iova(vgpu, dma->iova); gfn = dma->gfn; vfio_unpin_pages(dev, &gfn, 1); @@ -230,8 +262,8 @@ static struct intel_vgpu_type *intel_gvt_find_vgpu_type(struct intel_gvt *gvt, return NULL; } -static ssize_t available_instance_show(struct kobject *kobj, struct device *dev, - char *buf) +static ssize_t available_instances_show(struct kobject *kobj, + struct device *dev, char *buf) { struct intel_vgpu_type *type; unsigned int num = 0; @@ -269,12 +301,12 @@ static ssize_t description_show(struct kobject *kobj, struct device *dev, type->fence); } -static MDEV_TYPE_ATTR_RO(available_instance); +static MDEV_TYPE_ATTR_RO(available_instances); static MDEV_TYPE_ATTR_RO(device_api); static MDEV_TYPE_ATTR_RO(description); static struct attribute *type_attrs[] = { - &mdev_type_attr_available_instance.attr, + &mdev_type_attr_available_instances.attr, &mdev_type_attr_device_api.attr, &mdev_type_attr_description.attr, NULL, @@ -965,11 +997,6 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd, sparse->areas[0].offset = PAGE_ALIGN(vgpu_aperture_offset(vgpu)); sparse->areas[0].size = vgpu_aperture_sz(vgpu); - if (!caps.buf) { - kfree(caps.buf); - caps.buf = NULL; - caps.size = 0; - } break; case VFIO_PCI_BAR3_REGION_INDEX ... VFIO_PCI_BAR5_REGION_INDEX: @@ -1248,43 +1275,6 @@ static void kvmgt_page_track_flush_slot(struct kvm *kvm, spin_unlock(&kvm->mmu_lock); } -static bool kvmgt_check_guest(void) -{ - unsigned int eax, ebx, ecx, edx; - char s[12]; - unsigned int *i; - - eax = KVM_CPUID_SIGNATURE; - ebx = ecx = edx = 0; - - asm volatile ("cpuid" - : "+a"(eax), "=b"(ebx), "=c"(ecx), "=d"(edx) - : - : "cc", "memory"); - i = (unsigned int *)s; - i[0] = ebx; - i[1] = ecx; - i[2] = edx; - - return !strncmp(s, "KVMKVMKVM", strlen("KVMKVMKVM")); -} - -/** - * NOTE: - * It's actually impossible to check if we are running in KVM host, - * since the "KVM host" is simply native. So we only dectect guest here. - */ -static int kvmgt_detect_host(void) -{ -#ifdef CONFIG_INTEL_IOMMU - if (intel_iommu_gfx_mapped) { - gvt_err("Hardware IOMMU compatibility not yet supported, try to boot with intel_iommu=igfx_off\n"); - return -ENODEV; - } -#endif - return kvmgt_check_guest() ? -ENODEV : 0; -} - static bool __kvmgt_vgpu_exist(struct intel_vgpu *vgpu, struct kvm *kvm) { struct intel_vgpu *itr; @@ -1390,7 +1380,7 @@ static int kvmgt_inject_msi(unsigned long handle, u32 addr, u16 data) static unsigned long kvmgt_gfn_to_pfn(unsigned long handle, unsigned long gfn) { - unsigned long pfn; + unsigned long iova, pfn; struct kvmgt_guest_info *info; struct device *dev; int rc; @@ -1399,9 +1389,9 @@ static unsigned long kvmgt_gfn_to_pfn(unsigned long handle, unsigned long gfn) return INTEL_GVT_INVALID_ADDR; info = (struct kvmgt_guest_info *)handle; - pfn = gvt_cache_find(info->vgpu, gfn); - if (pfn != 0) - return pfn; + iova = gvt_cache_find(info->vgpu, gfn); + if (iova != INTEL_GVT_INVALID_ADDR) + return iova; pfn = INTEL_GVT_INVALID_ADDR; dev = mdev_dev(info->vgpu->vdev.mdev); @@ -1410,9 +1400,16 @@ static unsigned long kvmgt_gfn_to_pfn(unsigned long handle, unsigned long gfn) gvt_err("vfio_pin_pages failed for gfn 0x%lx: %d\n", gfn, rc); return INTEL_GVT_INVALID_ADDR; } + /* transfer to host iova for GFX to use DMA */ + rc = gvt_dma_map_iova(info->vgpu, pfn, &iova); + if (rc) { + gvt_err("gvt_dma_map_iova failed for gfn: 0x%lx\n", gfn); + vfio_unpin_pages(dev, &gfn, 1); + return INTEL_GVT_INVALID_ADDR; + } - gvt_cache_add(info->vgpu, gfn, pfn); - return pfn; + gvt_cache_add(info->vgpu, gfn, iova); + return iova; } static int kvmgt_rw_gpa(unsigned long handle, unsigned long gpa, @@ -1459,7 +1456,6 @@ static unsigned long kvmgt_virt_to_pfn(void *addr) } struct intel_gvt_mpt kvmgt_mpt = { - .detect_host = kvmgt_detect_host, .host_init = kvmgt_host_init, .host_exit = kvmgt_host_exit, .attach_vgpu = kvmgt_attach_vgpu, diff --git a/drivers/gpu/drm/i915/gvt/mpt.h b/drivers/gpu/drm/i915/gvt/mpt.h index 1af5830c0a56..419353624c5a 100644 --- a/drivers/gpu/drm/i915/gvt/mpt.h +++ b/drivers/gpu/drm/i915/gvt/mpt.h @@ -43,18 +43,6 @@ * hypervisor. */ -/** - * intel_gvt_hypervisor_detect_host - check if GVT-g is running within - * hypervisor host/privilged domain - * - * Returns: - * Zero on success, -ENODEV if current kernel is running inside a VM - */ -static inline int intel_gvt_hypervisor_detect_host(void) -{ - return intel_gvt_host.mpt->detect_host(); -} - /** * intel_gvt_hypervisor_host_init - init GVT-g host side * diff --git a/drivers/gpu/drm/i915/gvt/render.c b/drivers/gpu/drm/i915/gvt/render.c index 44136b1f3aab..2b3a642284b6 100644 --- a/drivers/gpu/drm/i915/gvt/render.c +++ b/drivers/gpu/drm/i915/gvt/render.c @@ -236,12 +236,18 @@ static void restore_mocs(struct intel_vgpu *vgpu, int ring_id) } } +#define CTX_CONTEXT_CONTROL_VAL 0x03 + void intel_gvt_load_render_mmio(struct intel_vgpu *vgpu, int ring_id) { struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv; struct render_mmio *mmio; u32 v; int i, array_size; + u32 *reg_state = vgpu->shadow_ctx->engine[ring_id].lrc_reg_state; + u32 ctx_ctrl = reg_state[CTX_CONTEXT_CONTROL_VAL]; + u32 inhibit_mask = + _MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT); if (IS_SKYLAKE(vgpu->gvt->dev_priv)) { mmio = gen9_render_mmio_list; @@ -257,6 +263,17 @@ void intel_gvt_load_render_mmio(struct intel_vgpu *vgpu, int ring_id) continue; mmio->value = I915_READ(mmio->reg); + + /* + * if it is an inhibit context, load in_context mmio + * into HW by mmio write. If it is not, skip this mmio + * write. + */ + if (mmio->in_context && + ((ctx_ctrl & inhibit_mask) != inhibit_mask) && + i915.enable_execlists) + continue; + if (mmio->mask) v = vgpu_vreg(vgpu, mmio->reg) | (mmio->mask << 16); else diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.c b/drivers/gpu/drm/i915/gvt/sched_policy.c index 678b0be85376..06c9584ac5f0 100644 --- a/drivers/gpu/drm/i915/gvt/sched_policy.c +++ b/drivers/gpu/drm/i915/gvt/sched_policy.c @@ -125,7 +125,6 @@ static void tbs_sched_func(struct work_struct *work) vgpu_data = scheduler->current_vgpu->sched_data; head = &vgpu_data->list; } else { - gvt_dbg_sched("no current vgpu search from q head\n"); head = &sched_data->runq_head; } diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 7ea68a75dc46..d6b6d0efdd1a 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -169,7 +169,8 @@ static int dispatch_workload(struct intel_vgpu_workload *workload) gvt_dbg_sched("ring id %d prepare to dispatch workload %p\n", ring_id, workload); - shadow_ctx->desc_template = workload->ctx_desc.addressing_mode << + shadow_ctx->desc_template &= ~(0x3 << GEN8_CTX_ADDRESSING_MODE_SHIFT); + shadow_ctx->desc_template |= workload->ctx_desc.addressing_mode << GEN8_CTX_ADDRESSING_MODE_SHIFT; mutex_lock(&dev_priv->drm.struct_mutex); @@ -456,7 +457,7 @@ static int workload_thread(void *priv) } complete: - gvt_dbg_sched("will complete workload %p\n, status: %d\n", + gvt_dbg_sched("will complete workload %p, status: %d\n", workload, workload->status); if (workload->req) diff --git a/drivers/gpu/drm/i915/gvt/scheduler.h b/drivers/gpu/drm/i915/gvt/scheduler.h index 3b30c28bff51..2833dfa8c9ae 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.h +++ b/drivers/gpu/drm/i915/gvt/scheduler.h @@ -113,7 +113,7 @@ struct intel_shadow_bb_entry { struct drm_i915_gem_object *obj; void *va; unsigned long len; - void *bb_start_cmd_va; + u32 *bb_start_cmd_va; }; #define workload_q_head(vgpu, ring_id) \ diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c index 7295bc8e12fb..95a97aa0051e 100644 --- a/drivers/gpu/drm/i915/gvt/vgpu.c +++ b/drivers/gpu/drm/i915/gvt/vgpu.c @@ -74,7 +74,7 @@ void populate_pvinfo_page(struct intel_vgpu *vgpu) int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) { unsigned int num_types; - unsigned int i, low_avail; + unsigned int i, low_avail, high_avail; unsigned int min_low; /* vGPU type name is defined as GVTg_Vx_y which contains @@ -89,9 +89,9 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) * to indicate how many vGPU instance can be created for this * type. * - * Currently use static size here as we init type earlier.. */ - low_avail = MB_TO_BYTES(256) - HOST_LOW_GM_SIZE; + low_avail = gvt_aperture_sz(gvt) - HOST_LOW_GM_SIZE; + high_avail = gvt_hidden_sz(gvt) - HOST_HIGH_GM_SIZE; num_types = 4; gvt->types = kzalloc(num_types * sizeof(struct intel_vgpu_type), @@ -106,7 +106,8 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) gvt->types[i].low_gm_size = min_low; gvt->types[i].high_gm_size = max((min_low<<3), MB_TO_BYTES(384U)); gvt->types[i].fence = 4; - gvt->types[i].max_instance = low_avail / min_low; + gvt->types[i].max_instance = min(low_avail / min_low, + high_avail / gvt->types[i].high_gm_size); gvt->types[i].avail_instance = gvt->types[i].max_instance; if (IS_GEN8(gvt->dev_priv)) @@ -142,9 +143,9 @@ static void intel_gvt_update_vgpu_types(struct intel_gvt *gvt) /* Need to depend on maxium hw resource size but keep on * static config for now. */ - low_gm_avail = MB_TO_BYTES(256) - HOST_LOW_GM_SIZE - + low_gm_avail = gvt_aperture_sz(gvt) - HOST_LOW_GM_SIZE - gvt->gm.vgpu_allocated_low_gm_size; - high_gm_avail = MB_TO_BYTES(256) * 8UL - HOST_HIGH_GM_SIZE - + high_gm_avail = gvt_hidden_sz(gvt) - HOST_HIGH_GM_SIZE - gvt->gm.vgpu_allocated_high_gm_size; fence_avail = gvt_fence_sz(gvt) - HOST_FENCE - gvt->fence.vgpu_allocated_fence_num; @@ -384,6 +385,7 @@ void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr, intel_vgpu_reset_resource(vgpu); intel_vgpu_reset_mmio(vgpu); populate_pvinfo_page(vgpu); + intel_vgpu_reset_display(vgpu); if (dmlr) intel_vgpu_reset_cfg_space(vgpu); diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 51241dec3f83..9be9b9b7f9cb 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -845,10 +845,6 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv, if (ret < 0) goto err_engines; - ret = intel_gvt_init(dev_priv); - if (ret < 0) - goto err_workqueues; - /* This must be called before any calls to HAS_PCH_* */ intel_detect_pch(dev_priv); @@ -862,7 +858,7 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv, intel_init_audio_hooks(dev_priv); ret = i915_gem_load_init(dev_priv); if (ret < 0) - goto err_gvt; + goto err_workqueues; intel_display_crc_init(dev_priv); @@ -874,8 +870,6 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv, return 0; -err_gvt: - intel_gvt_cleanup(dev_priv); err_workqueues: i915_workqueues_cleanup(dev_priv); err_engines: @@ -1102,6 +1096,10 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) DRM_DEBUG_DRIVER("can't enable MSI"); } + ret = intel_gvt_init(dev_priv); + if (ret) + goto out_ggtt; + return 0; out_ggtt: @@ -1340,6 +1338,8 @@ void i915_driver_unload(struct drm_device *dev) drm_modeset_drop_locks(&ctx); drm_modeset_acquire_fini(&ctx); + intel_gvt_cleanup(dev_priv); + i915_driver_unregister(dev_priv); drm_vblank_cleanup(dev); diff --git a/drivers/gpu/drm/i915/intel_gvt.c b/drivers/gpu/drm/i915/intel_gvt.c index 290384e86c63..d23c0fcff751 100644 --- a/drivers/gpu/drm/i915/intel_gvt.c +++ b/drivers/gpu/drm/i915/intel_gvt.c @@ -67,6 +67,11 @@ int intel_gvt_init(struct drm_i915_private *dev_priv) return 0; } + if (intel_vgpu_active(dev_priv)) { + DRM_DEBUG_DRIVER("GVT-g is disabled for guest\n"); + goto bail; + } + if (!is_supported_device(dev_priv)) { DRM_DEBUG_DRIVER("Unsupported device. GVT-g is disabled\n"); goto bail; diff --git a/drivers/gpu/drm/imx/imx-drm-core.c b/drivers/gpu/drm/imx/imx-drm-core.c index 5ae48836652e..f562cb7964b0 100644 --- a/drivers/gpu/drm/imx/imx-drm-core.c +++ b/drivers/gpu/drm/imx/imx-drm-core.c @@ -357,8 +357,8 @@ static int imx_drm_bind(struct device *dev) * this value would be used to check framebuffer size limitation * at drm_mode_addfb(). */ - drm->mode_config.min_width = 64; - drm->mode_config.min_height = 64; + drm->mode_config.min_width = 1; + drm->mode_config.min_height = 1; drm->mode_config.max_width = 4096; drm->mode_config.max_height = 4096; drm->mode_config.funcs = &imx_drm_mode_config_funcs; diff --git a/drivers/gpu/drm/imx/imx-tve.c b/drivers/gpu/drm/imx/imx-tve.c index 8f8aa4a63122..4826bb781723 100644 --- a/drivers/gpu/drm/imx/imx-tve.c +++ b/drivers/gpu/drm/imx/imx-tve.c @@ -98,6 +98,8 @@ /* TVE_TST_MODE_REG */ #define TVE_TVDAC_TEST_MODE_MASK (0x7 << 0) +#define IMX_TVE_DAC_VOLTAGE 2750000 + enum { TVE_MODE_TVOUT, TVE_MODE_VGA, @@ -616,9 +618,8 @@ static int imx_tve_bind(struct device *dev, struct device *master, void *data) tve->dac_reg = devm_regulator_get(dev, "dac"); if (!IS_ERR(tve->dac_reg)) { - ret = regulator_set_voltage(tve->dac_reg, 2750000, 2750000); - if (ret) - return ret; + if (regulator_get_voltage(tve->dac_reg) != IMX_TVE_DAC_VOLTAGE) + dev_warn(dev, "dac voltage is not %d uV\n", IMX_TVE_DAC_VOLTAGE); ret = regulator_enable(tve->dac_reg); if (ret) return ret; diff --git a/drivers/gpu/drm/nouveau/dispnv04/arb.c b/drivers/gpu/drm/nouveau/dispnv04/arb.c index a555681c3096..90075b676256 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/arb.c +++ b/drivers/gpu/drm/nouveau/dispnv04/arb.c @@ -198,7 +198,7 @@ nv04_update_arb(struct drm_device *dev, int VClk, int bpp, int *burst, int *lwm) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; struct nv_fifo_info fifo_data; struct nv_sim_state sim_data; int MClk = nouveau_hw_get_clock(dev, PLL_MEMORY); @@ -227,7 +227,7 @@ nv04_update_arb(struct drm_device *dev, int VClk, int bpp, sim_data.mem_page_miss = ((cfg1 >> 4) & 0xf) + ((cfg1 >> 31) & 0x1); } - if (drm->device.info.family == NV_DEVICE_INFO_V0_TNT) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_TNT) nv04_calc_arb(&fifo_data, &sim_data); else nv10_calc_arb(&fifo_data, &sim_data); @@ -254,7 +254,7 @@ nouveau_calc_arb(struct drm_device *dev, int vclk, int bpp, int *burst, int *lwm { struct nouveau_drm *drm = nouveau_drm(dev); - if (drm->device.info.family < NV_DEVICE_INFO_V0_KELVIN) + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_KELVIN) nv04_update_arb(dev, vclk, bpp, burst, lwm); else if ((dev->pdev->device & 0xfff0) == 0x0240 /*CHIPSET_C51*/ || (dev->pdev->device & 0xfff0) == 0x03d0 /*CHIPSET_C512*/) { diff --git a/drivers/gpu/drm/nouveau/dispnv04/crtc.c b/drivers/gpu/drm/nouveau/dispnv04/crtc.c index a72754d73c84..ab7b69c11d40 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/crtc.c +++ b/drivers/gpu/drm/nouveau/dispnv04/crtc.c @@ -113,8 +113,8 @@ static void nv_crtc_calc_state_ext(struct drm_crtc *crtc, struct drm_display_mod { struct drm_device *dev = crtc->dev; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_bios *bios = nvxx_bios(&drm->device); - struct nvkm_clk *clk = nvxx_clk(&drm->device); + struct nvkm_bios *bios = nvxx_bios(&drm->client.device); + struct nvkm_clk *clk = nvxx_clk(&drm->client.device); struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc); struct nv04_mode_state *state = &nv04_display(dev)->mode_reg; struct nv04_crtc_reg *regp = &state->crtc_reg[nv_crtc->index]; @@ -138,7 +138,7 @@ static void nv_crtc_calc_state_ext(struct drm_crtc *crtc, struct drm_display_mod * has yet been observed in allowing the use a single stage pll on all * nv43 however. the behaviour of single stage use is untested on nv40 */ - if (drm->device.info.chipset > 0x40 && dot_clock <= (pll_lim.vco1.max_freq / 2)) + if (drm->client.device.info.chipset > 0x40 && dot_clock <= (pll_lim.vco1.max_freq / 2)) memset(&pll_lim.vco2, 0, sizeof(pll_lim.vco2)); @@ -148,10 +148,10 @@ static void nv_crtc_calc_state_ext(struct drm_crtc *crtc, struct drm_display_mod state->pllsel &= PLLSEL_VPLL1_MASK | PLLSEL_VPLL2_MASK | PLLSEL_TV_MASK; /* The blob uses this always, so let's do the same */ - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) state->pllsel |= NV_PRAMDAC_PLL_COEFF_SELECT_USE_VPLL2_TRUE; /* again nv40 and some nv43 act more like nv3x as described above */ - if (drm->device.info.chipset < 0x41) + if (drm->client.device.info.chipset < 0x41) state->pllsel |= NV_PRAMDAC_PLL_COEFF_SELECT_SOURCE_PROG_MPLL | NV_PRAMDAC_PLL_COEFF_SELECT_SOURCE_PROG_NVPLL; state->pllsel |= nv_crtc->index ? PLLSEL_VPLL2_MASK : PLLSEL_VPLL1_MASK; @@ -270,7 +270,7 @@ nv_crtc_mode_set_vga(struct drm_crtc *crtc, struct drm_display_mode *mode) horizEnd = horizTotal - 2; horizBlankEnd = horizTotal + 4; #if 0 - if (dev->overlayAdaptor && drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) + if (dev->overlayAdaptor && drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) /* This reportedly works around some video overlay bandwidth problems */ horizTotal += 2; #endif @@ -505,7 +505,7 @@ nv_crtc_mode_set_regs(struct drm_crtc *crtc, struct drm_display_mode * mode) regp->cursor_cfg = NV_PCRTC_CURSOR_CONFIG_CUR_LINES_64 | NV_PCRTC_CURSOR_CONFIG_CUR_PIXELS_64 | NV_PCRTC_CURSOR_CONFIG_ADDRESS_SPACE_PNVM; - if (drm->device.info.chipset >= 0x11) + if (drm->client.device.info.chipset >= 0x11) regp->cursor_cfg |= NV_PCRTC_CURSOR_CONFIG_CUR_BPP_32; if (mode->flags & DRM_MODE_FLAG_DBLSCAN) regp->cursor_cfg |= NV_PCRTC_CURSOR_CONFIG_DOUBLE_SCAN_ENABLE; @@ -546,26 +546,26 @@ nv_crtc_mode_set_regs(struct drm_crtc *crtc, struct drm_display_mode * mode) * 1 << 30 on 0x60.830), for no apparent reason */ regp->CRTC[NV_CIO_CRE_59] = off_chip_digital; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) regp->CRTC[0x9f] = off_chip_digital ? 0x11 : 0x1; regp->crtc_830 = mode->crtc_vdisplay - 3; regp->crtc_834 = mode->crtc_vdisplay - 1; - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) /* This is what the blob does */ regp->crtc_850 = NVReadCRTC(dev, 0, NV_PCRTC_850); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) regp->gpio_ext = NVReadCRTC(dev, 0, NV_PCRTC_GPIO_EXT); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) regp->crtc_cfg = NV10_PCRTC_CONFIG_START_ADDRESS_HSYNC; else regp->crtc_cfg = NV04_PCRTC_CONFIG_START_ADDRESS_HSYNC; /* Some misc regs */ - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) { regp->CRTC[NV_CIO_CRE_85] = 0xFF; regp->CRTC[NV_CIO_CRE_86] = 0x1; } @@ -577,7 +577,7 @@ nv_crtc_mode_set_regs(struct drm_crtc *crtc, struct drm_display_mode * mode) /* Generic PRAMDAC regs */ - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) /* Only bit that bios and blob set. */ regp->nv10_cursync = (1 << 25); @@ -586,7 +586,7 @@ nv_crtc_mode_set_regs(struct drm_crtc *crtc, struct drm_display_mode * mode) NV_PRAMDAC_GENERAL_CONTROL_PIXMIX_ON; if (fb->format->depth == 16) regp->ramdac_gen_ctrl |= NV_PRAMDAC_GENERAL_CONTROL_ALT_MODE_SEL; - if (drm->device.info.chipset >= 0x11) + if (drm->client.device.info.chipset >= 0x11) regp->ramdac_gen_ctrl |= NV_PRAMDAC_GENERAL_CONTROL_PIPE_LONG; regp->ramdac_630 = 0; /* turn off green mode (tv test pattern?) */ @@ -649,7 +649,7 @@ nv_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode, nv_crtc_mode_set_vga(crtc, adjusted_mode); /* calculated in nv04_dfp_prepare, nv40 needs it written before calculating PLLs */ - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) NVWriteRAMDAC(dev, 0, NV_PRAMDAC_SEL_CLK, nv04_display(dev)->mode_reg.sel_clk); nv_crtc_mode_set_regs(crtc, adjusted_mode); nv_crtc_calc_state_ext(crtc, mode, adjusted_mode->clock); @@ -710,7 +710,7 @@ static void nv_crtc_prepare(struct drm_crtc *crtc) /* Some more preparation. */ NVWriteCRTC(dev, nv_crtc->index, NV_PCRTC_CONFIG, NV_PCRTC_CONFIG_START_ADDRESS_NON_VGA); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) { uint32_t reg900 = NVReadRAMDAC(dev, nv_crtc->index, NV_PRAMDAC_900); NVWriteRAMDAC(dev, nv_crtc->index, NV_PRAMDAC_900, reg900 & ~0x10000); } @@ -886,7 +886,7 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc, crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_FF_INDEX); crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_FFLWM__INDEX); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_KELVIN) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_KELVIN) { regp->CRTC[NV_CIO_CRE_47] = arb_lwm >> 8; crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_47); } @@ -967,7 +967,7 @@ static void nv11_cursor_upload(struct drm_device *dev, struct nouveau_bo *src, { struct nouveau_drm *drm = nouveau_drm(dev); - if (drm->device.info.chipset == 0x11) { + if (drm->client.device.info.chipset == 0x11) { pixel = ((pixel & 0x000000ff) << 24) | ((pixel & 0x0000ff00) << 8) | ((pixel & 0x00ff0000) >> 8) | @@ -1008,7 +1008,7 @@ nv04_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv, if (ret) goto out; - if (drm->device.info.chipset >= 0x11) + if (drm->client.device.info.chipset >= 0x11) nv11_cursor_upload(dev, cursor, nv_crtc->cursor.nvbo); else nv04_cursor_upload(dev, cursor, nv_crtc->cursor.nvbo); @@ -1124,8 +1124,9 @@ nv04_crtc_create(struct drm_device *dev, int crtc_num) drm_crtc_helper_add(&nv_crtc->base, &nv04_crtc_helper_funcs); drm_mode_crtc_set_gamma_size(&nv_crtc->base, 256); - ret = nouveau_bo_new(dev, 64*64*4, 0x100, TTM_PL_FLAG_VRAM, - 0, 0x0000, NULL, NULL, &nv_crtc->cursor.nvbo); + ret = nouveau_bo_new(&nouveau_drm(dev)->client, 64*64*4, 0x100, + TTM_PL_FLAG_VRAM, 0, 0x0000, NULL, NULL, + &nv_crtc->cursor.nvbo); if (!ret) { ret = nouveau_bo_pin(nv_crtc->cursor.nvbo, TTM_PL_FLAG_VRAM, false); if (!ret) { diff --git a/drivers/gpu/drm/nouveau/dispnv04/cursor.c b/drivers/gpu/drm/nouveau/dispnv04/cursor.c index c83116a308a4..f26e44ea7389 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/cursor.c +++ b/drivers/gpu/drm/nouveau/dispnv04/cursor.c @@ -55,7 +55,7 @@ nv04_cursor_set_offset(struct nouveau_crtc *nv_crtc, uint32_t offset) crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_HCUR_ADDR0_INDEX); crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_HCUR_ADDR1_INDEX); crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_HCUR_ADDR2_INDEX); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) nv_fix_nv40_hw_cursor(dev, nv_crtc->index); } diff --git a/drivers/gpu/drm/nouveau/dispnv04/dac.c b/drivers/gpu/drm/nouveau/dispnv04/dac.c index b6cc7766e6f7..4feab0a5419d 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/dac.c +++ b/drivers/gpu/drm/nouveau/dispnv04/dac.c @@ -66,7 +66,7 @@ int nv04_dac_output_offset(struct drm_encoder *encoder) static int sample_load_twice(struct drm_device *dev, bool sense[2]) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int i; for (i = 0; i < 2; i++) { @@ -80,19 +80,19 @@ static int sample_load_twice(struct drm_device *dev, bool sense[2]) * use a 10ms timeout (guards against crtc being inactive, in * which case blank state would never change) */ - if (nvif_msec(&drm->device, 10, + if (nvif_msec(&drm->client.device, 10, if (!(nvif_rd32(device, NV_PRMCIO_INP0__COLOR) & 1)) break; ) < 0) return -EBUSY; - if (nvif_msec(&drm->device, 10, + if (nvif_msec(&drm->client.device, 10, if ( (nvif_rd32(device, NV_PRMCIO_INP0__COLOR) & 1)) break; ) < 0) return -EBUSY; - if (nvif_msec(&drm->device, 10, + if (nvif_msec(&drm->client.device, 10, if (!(nvif_rd32(device, NV_PRMCIO_INP0__COLOR) & 1)) break; ) < 0) @@ -133,7 +133,7 @@ static enum drm_connector_status nv04_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector) { struct drm_device *dev = encoder->dev; - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; struct nouveau_drm *drm = nouveau_drm(dev); uint8_t saved_seq1, saved_pi, saved_rpc1, saved_cr_mode; uint8_t saved_palette0[3], saved_palette_mask; @@ -236,8 +236,8 @@ uint32_t nv17_dac_sample_load(struct drm_encoder *encoder) { struct drm_device *dev = encoder->dev; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &nouveau_drm(dev)->device.object; - struct nvkm_gpio *gpio = nvxx_gpio(&drm->device); + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; + struct nvkm_gpio *gpio = nvxx_gpio(&drm->client.device); struct dcb_output *dcb = nouveau_encoder(encoder)->dcb; uint32_t sample, testval, regoffset = nv04_dac_output_offset(encoder); uint32_t saved_powerctrl_2 = 0, saved_powerctrl_4 = 0, saved_routput, @@ -288,7 +288,7 @@ uint32_t nv17_dac_sample_load(struct drm_encoder *encoder) /* nv driver and nv31 use 0xfffffeee, nv34 and 6600 use 0xfffffece */ routput = (saved_routput & 0xfffffece) | head << 8; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CURIE) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CURIE) { if (dcb->type == DCB_OUTPUT_TV) routput |= 0x1a << 16; else @@ -403,7 +403,7 @@ static void nv04_dac_mode_set(struct drm_encoder *encoder, } /* This could use refinement for flatpanels, but it should work this way */ - if (drm->device.info.chipset < 0x44) + if (drm->client.device.info.chipset < 0x44) NVWriteRAMDAC(dev, 0, NV_PRAMDAC_TEST_CONTROL + nv04_dac_output_offset(encoder), 0xf0000000); else NVWriteRAMDAC(dev, 0, NV_PRAMDAC_TEST_CONTROL + nv04_dac_output_offset(encoder), 0x00100000); diff --git a/drivers/gpu/drm/nouveau/dispnv04/dfp.c b/drivers/gpu/drm/nouveau/dispnv04/dfp.c index 2e5bb2afda7c..9805d2cdc1a1 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/dfp.c +++ b/drivers/gpu/drm/nouveau/dispnv04/dfp.c @@ -281,7 +281,7 @@ static void nv04_dfp_mode_set(struct drm_encoder *encoder, struct drm_display_mode *adjusted_mode) { struct drm_device *dev = encoder->dev; - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; struct nouveau_drm *drm = nouveau_drm(dev); struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc); struct nv04_crtc_reg *regp = &nv04_display(dev)->mode_reg.crtc_reg[nv_crtc->index]; @@ -417,7 +417,7 @@ static void nv04_dfp_mode_set(struct drm_encoder *encoder, if ((nv_connector->dithering_mode == DITHERING_MODE_ON) || (nv_connector->dithering_mode == DITHERING_MODE_AUTO && fb->format->depth > connector->display_info.bpc * 3)) { - if (drm->device.info.chipset == 0x11) + if (drm->client.device.info.chipset == 0x11) regp->dither = savep->dither | 0x00010000; else { int i; @@ -428,7 +428,7 @@ static void nv04_dfp_mode_set(struct drm_encoder *encoder, } } } else { - if (drm->device.info.chipset != 0x11) { + if (drm->client.device.info.chipset != 0x11) { /* reset them */ int i; for (i = 0; i < 3; i++) { @@ -464,7 +464,7 @@ static void nv04_dfp_commit(struct drm_encoder *encoder) NVReadRAMDAC(dev, head, NV_PRAMDAC_FP_TG_CONTROL); /* This could use refinement for flatpanels, but it should work this way */ - if (drm->device.info.chipset < 0x44) + if (drm->client.device.info.chipset < 0x44) NVWriteRAMDAC(dev, 0, NV_PRAMDAC_TEST_CONTROL + nv04_dac_output_offset(encoder), 0xf0000000); else NVWriteRAMDAC(dev, 0, NV_PRAMDAC_TEST_CONTROL + nv04_dac_output_offset(encoder), 0x00100000); @@ -486,7 +486,7 @@ static void nv04_dfp_update_backlight(struct drm_encoder *encoder, int mode) { #ifdef __powerpc__ struct drm_device *dev = encoder->dev; - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; /* BIOS scripts usually take care of the backlight, thanks * Apple for your consistency. @@ -624,7 +624,7 @@ static void nv04_tmds_slave_init(struct drm_encoder *encoder) struct drm_device *dev = encoder->dev; struct dcb_output *dcb = nouveau_encoder(encoder)->dcb; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct nvkm_i2c_bus *bus = nvkm_i2c_bus_find(i2c, NVKM_I2C_BUS_PRI); struct nvkm_i2c_bus_probe info[] = { { diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.c b/drivers/gpu/drm/nouveau/dispnv04/disp.c index 34c0f2f67548..5b9d549aa791 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/disp.c +++ b/drivers/gpu/drm/nouveau/dispnv04/disp.c @@ -35,7 +35,7 @@ int nv04_display_create(struct drm_device *dev) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct dcb_table *dcb = &drm->vbios.dcb; struct drm_connector *connector, *ct; struct drm_encoder *encoder; @@ -48,7 +48,7 @@ nv04_display_create(struct drm_device *dev) if (!disp) return -ENOMEM; - nvif_object_map(&drm->device.object); + nvif_object_map(&drm->client.device.object); nouveau_display(dev)->priv = disp; nouveau_display(dev)->dtor = nv04_display_destroy; @@ -139,7 +139,7 @@ nv04_display_destroy(struct drm_device *dev) nouveau_display(dev)->priv = NULL; kfree(disp); - nvif_object_unmap(&drm->device.object); + nvif_object_unmap(&drm->client.device.object); } int diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.h b/drivers/gpu/drm/nouveau/dispnv04/disp.h index 7030307d2d48..bea4543554ba 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/disp.h +++ b/drivers/gpu/drm/nouveau/dispnv04/disp.h @@ -129,7 +129,7 @@ nv_two_heads(struct drm_device *dev) struct nouveau_drm *drm = nouveau_drm(dev); const int impl = dev->pdev->device & 0x0ff0; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS && impl != 0x0100 && + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS && impl != 0x0100 && impl != 0x0150 && impl != 0x01a0 && impl != 0x0200) return true; @@ -148,7 +148,7 @@ nv_two_reg_pll(struct drm_device *dev) struct nouveau_drm *drm = nouveau_drm(dev); const int impl = dev->pdev->device & 0x0ff0; - if (impl == 0x0310 || impl == 0x0340 || drm->device.info.family >= NV_DEVICE_INFO_V0_CURIE) + if (impl == 0x0310 || impl == 0x0340 || drm->client.device.info.family >= NV_DEVICE_INFO_V0_CURIE) return true; return false; } @@ -170,7 +170,7 @@ nouveau_bios_run_init_table(struct drm_device *dev, u16 table, struct dcb_output *outp, int crtc) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_bios *bios = nvxx_bios(&drm->device); + struct nvkm_bios *bios = nvxx_bios(&drm->client.device); struct nvbios_init init = { .subdev = &bios->subdev, .bios = bios, diff --git a/drivers/gpu/drm/nouveau/dispnv04/hw.c b/drivers/gpu/drm/nouveau/dispnv04/hw.c index 74856a8b8f35..b98599002831 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/hw.c +++ b/drivers/gpu/drm/nouveau/dispnv04/hw.c @@ -89,7 +89,7 @@ NVSetOwner(struct drm_device *dev, int owner) if (owner == 1) owner *= 3; - if (drm->device.info.chipset == 0x11) { + if (drm->client.device.info.chipset == 0x11) { /* This might seem stupid, but the blob does it and * omitting it often locks the system up. */ @@ -100,7 +100,7 @@ NVSetOwner(struct drm_device *dev, int owner) /* CR44 is always changed on CRTC0 */ NVWriteVgaCrtc(dev, 0, NV_CIO_CRE_44, owner); - if (drm->device.info.chipset == 0x11) { /* set me harder */ + if (drm->client.device.info.chipset == 0x11) { /* set me harder */ NVWriteVgaCrtc(dev, 0, NV_CIO_CRE_2E, owner); NVWriteVgaCrtc(dev, 0, NV_CIO_CRE_2E, owner); } @@ -149,7 +149,7 @@ nouveau_hw_decode_pll(struct drm_device *dev, uint32_t reg1, uint32_t pll1, pllvals->NM1 = pll1 & 0xffff; if (nv_two_reg_pll(dev) && pll2 & NV31_RAMDAC_ENABLE_VCO2) pllvals->NM2 = pll2 & 0xffff; - else if (drm->device.info.chipset == 0x30 || drm->device.info.chipset == 0x35) { + else if (drm->client.device.info.chipset == 0x30 || drm->client.device.info.chipset == 0x35) { pllvals->M1 &= 0xf; /* only 4 bits */ if (pll1 & NV30_RAMDAC_ENABLE_VCO2) { pllvals->M2 = (pll1 >> 4) & 0x7; @@ -165,8 +165,8 @@ nouveau_hw_get_pllvals(struct drm_device *dev, enum nvbios_pll_type plltype, struct nvkm_pll_vals *pllvals) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; - struct nvkm_bios *bios = nvxx_bios(&drm->device); + struct nvif_object *device = &drm->client.device.object; + struct nvkm_bios *bios = nvxx_bios(&drm->client.device); uint32_t reg1, pll1, pll2 = 0; struct nvbios_pll pll_lim; int ret; @@ -184,7 +184,7 @@ nouveau_hw_get_pllvals(struct drm_device *dev, enum nvbios_pll_type plltype, pll2 = nvif_rd32(device, reg2); } - if (drm->device.info.family == NV_DEVICE_INFO_V0_CELSIUS && reg1 >= NV_PRAMDAC_VPLL_COEFF) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CELSIUS && reg1 >= NV_PRAMDAC_VPLL_COEFF) { uint32_t ramdac580 = NVReadRAMDAC(dev, 0, NV_PRAMDAC_580); /* check whether vpll has been forced into single stage mode */ @@ -222,6 +222,7 @@ nouveau_hw_get_clock(struct drm_device *dev, enum nvbios_pll_type plltype) uint32_t mpllP; pci_read_config_dword(pci_get_bus_and_slot(0, 3), 0x6c, &mpllP); + mpllP = (mpllP >> 8) & 0xf; if (!mpllP) mpllP = 4; @@ -232,7 +233,7 @@ nouveau_hw_get_clock(struct drm_device *dev, enum nvbios_pll_type plltype) uint32_t clock; pci_read_config_dword(pci_get_bus_and_slot(0, 5), 0x4c, &clock); - return clock; + return clock / 1000; } ret = nouveau_hw_get_pllvals(dev, plltype, &pllvals); @@ -252,7 +253,7 @@ nouveau_hw_fix_bad_vpll(struct drm_device *dev, int head) */ struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; struct nvkm_clk *clk = nvxx_clk(device); struct nvkm_bios *bios = nvxx_bios(device); struct nvbios_pll pll_lim; @@ -391,21 +392,21 @@ nv_save_state_ramdac(struct drm_device *dev, int head, struct nv04_crtc_reg *regp = &state->crtc_reg[head]; int i; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) regp->nv10_cursync = NVReadRAMDAC(dev, head, NV_RAMDAC_NV10_CURSYNC); nouveau_hw_get_pllvals(dev, head ? PLL_VPLL1 : PLL_VPLL0, ®p->pllvals); state->pllsel = NVReadRAMDAC(dev, 0, NV_PRAMDAC_PLL_COEFF_SELECT); if (nv_two_heads(dev)) state->sel_clk = NVReadRAMDAC(dev, 0, NV_PRAMDAC_SEL_CLK); - if (drm->device.info.chipset == 0x11) + if (drm->client.device.info.chipset == 0x11) regp->dither = NVReadRAMDAC(dev, head, NV_RAMDAC_DITHER_NV11); regp->ramdac_gen_ctrl = NVReadRAMDAC(dev, head, NV_PRAMDAC_GENERAL_CONTROL); if (nv_gf4_disp_arch(dev)) regp->ramdac_630 = NVReadRAMDAC(dev, head, NV_PRAMDAC_630); - if (drm->device.info.chipset >= 0x30) + if (drm->client.device.info.chipset >= 0x30) regp->ramdac_634 = NVReadRAMDAC(dev, head, NV_PRAMDAC_634); regp->tv_setup = NVReadRAMDAC(dev, head, NV_PRAMDAC_TV_SETUP); @@ -447,7 +448,7 @@ nv_save_state_ramdac(struct drm_device *dev, int head, if (nv_gf4_disp_arch(dev)) regp->ramdac_8c0 = NVReadRAMDAC(dev, head, NV_PRAMDAC_8C0); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) { regp->ramdac_a20 = NVReadRAMDAC(dev, head, NV_PRAMDAC_A20); regp->ramdac_a24 = NVReadRAMDAC(dev, head, NV_PRAMDAC_A24); regp->ramdac_a34 = NVReadRAMDAC(dev, head, NV_PRAMDAC_A34); @@ -463,26 +464,26 @@ nv_load_state_ramdac(struct drm_device *dev, int head, struct nv04_mode_state *state) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_clk *clk = nvxx_clk(&drm->device); + struct nvkm_clk *clk = nvxx_clk(&drm->client.device); struct nv04_crtc_reg *regp = &state->crtc_reg[head]; uint32_t pllreg = head ? NV_RAMDAC_VPLL2 : NV_PRAMDAC_VPLL_COEFF; int i; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) NVWriteRAMDAC(dev, head, NV_RAMDAC_NV10_CURSYNC, regp->nv10_cursync); clk->pll_prog(clk, pllreg, ®p->pllvals); NVWriteRAMDAC(dev, 0, NV_PRAMDAC_PLL_COEFF_SELECT, state->pllsel); if (nv_two_heads(dev)) NVWriteRAMDAC(dev, 0, NV_PRAMDAC_SEL_CLK, state->sel_clk); - if (drm->device.info.chipset == 0x11) + if (drm->client.device.info.chipset == 0x11) NVWriteRAMDAC(dev, head, NV_RAMDAC_DITHER_NV11, regp->dither); NVWriteRAMDAC(dev, head, NV_PRAMDAC_GENERAL_CONTROL, regp->ramdac_gen_ctrl); if (nv_gf4_disp_arch(dev)) NVWriteRAMDAC(dev, head, NV_PRAMDAC_630, regp->ramdac_630); - if (drm->device.info.chipset >= 0x30) + if (drm->client.device.info.chipset >= 0x30) NVWriteRAMDAC(dev, head, NV_PRAMDAC_634, regp->ramdac_634); NVWriteRAMDAC(dev, head, NV_PRAMDAC_TV_SETUP, regp->tv_setup); @@ -519,7 +520,7 @@ nv_load_state_ramdac(struct drm_device *dev, int head, if (nv_gf4_disp_arch(dev)) NVWriteRAMDAC(dev, head, NV_PRAMDAC_8C0, regp->ramdac_8c0); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) { NVWriteRAMDAC(dev, head, NV_PRAMDAC_A20, regp->ramdac_a20); NVWriteRAMDAC(dev, head, NV_PRAMDAC_A24, regp->ramdac_a24); NVWriteRAMDAC(dev, head, NV_PRAMDAC_A34, regp->ramdac_a34); @@ -600,10 +601,10 @@ nv_save_state_ext(struct drm_device *dev, int head, rd_cio_state(dev, head, regp, NV_CIO_CRE_FFLWM__INDEX); rd_cio_state(dev, head, regp, NV_CIO_CRE_21); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_KELVIN) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_KELVIN) rd_cio_state(dev, head, regp, NV_CIO_CRE_47); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) rd_cio_state(dev, head, regp, 0x9f); rd_cio_state(dev, head, regp, NV_CIO_CRE_49); @@ -612,14 +613,14 @@ nv_save_state_ext(struct drm_device *dev, int head, rd_cio_state(dev, head, regp, NV_CIO_CRE_HCUR_ADDR2_INDEX); rd_cio_state(dev, head, regp, NV_CIO_CRE_ILACE__INDEX); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { regp->crtc_830 = NVReadCRTC(dev, head, NV_PCRTC_830); regp->crtc_834 = NVReadCRTC(dev, head, NV_PCRTC_834); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) regp->gpio_ext = NVReadCRTC(dev, head, NV_PCRTC_GPIO_EXT); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) regp->crtc_850 = NVReadCRTC(dev, head, NV_PCRTC_850); if (nv_two_heads(dev)) @@ -631,7 +632,7 @@ nv_save_state_ext(struct drm_device *dev, int head, rd_cio_state(dev, head, regp, NV_CIO_CRE_SCRATCH3__INDEX); rd_cio_state(dev, head, regp, NV_CIO_CRE_SCRATCH4__INDEX); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { rd_cio_state(dev, head, regp, NV_CIO_CRE_EBR_INDEX); rd_cio_state(dev, head, regp, NV_CIO_CRE_CSB); rd_cio_state(dev, head, regp, NV_CIO_CRE_4B); @@ -660,12 +661,12 @@ nv_load_state_ext(struct drm_device *dev, int head, struct nv04_mode_state *state) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; struct nv04_crtc_reg *regp = &state->crtc_reg[head]; uint32_t reg900; int i; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { if (nv_two_heads(dev)) /* setting ENGINE_CTRL (EC) *must* come before * CIO_CRE_LCD, as writing CRE_LCD sets bits 16 & 17 in @@ -677,20 +678,20 @@ nv_load_state_ext(struct drm_device *dev, int head, nvif_wr32(device, NV_PVIDEO_INTR_EN, 0); nvif_wr32(device, NV_PVIDEO_OFFSET_BUFF(0), 0); nvif_wr32(device, NV_PVIDEO_OFFSET_BUFF(1), 0); - nvif_wr32(device, NV_PVIDEO_LIMIT(0), drm->device.info.ram_size - 1); - nvif_wr32(device, NV_PVIDEO_LIMIT(1), drm->device.info.ram_size - 1); - nvif_wr32(device, NV_PVIDEO_UVPLANE_LIMIT(0), drm->device.info.ram_size - 1); - nvif_wr32(device, NV_PVIDEO_UVPLANE_LIMIT(1), drm->device.info.ram_size - 1); + nvif_wr32(device, NV_PVIDEO_LIMIT(0), drm->client.device.info.ram_size - 1); + nvif_wr32(device, NV_PVIDEO_LIMIT(1), drm->client.device.info.ram_size - 1); + nvif_wr32(device, NV_PVIDEO_UVPLANE_LIMIT(0), drm->client.device.info.ram_size - 1); + nvif_wr32(device, NV_PVIDEO_UVPLANE_LIMIT(1), drm->client.device.info.ram_size - 1); nvif_wr32(device, NV_PBUS_POWERCTRL_2, 0); NVWriteCRTC(dev, head, NV_PCRTC_CURSOR_CONFIG, regp->cursor_cfg); NVWriteCRTC(dev, head, NV_PCRTC_830, regp->crtc_830); NVWriteCRTC(dev, head, NV_PCRTC_834, regp->crtc_834); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) NVWriteCRTC(dev, head, NV_PCRTC_GPIO_EXT, regp->gpio_ext); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) { NVWriteCRTC(dev, head, NV_PCRTC_850, regp->crtc_850); reg900 = NVReadRAMDAC(dev, head, NV_PRAMDAC_900); @@ -713,23 +714,23 @@ nv_load_state_ext(struct drm_device *dev, int head, wr_cio_state(dev, head, regp, NV_CIO_CRE_FF_INDEX); wr_cio_state(dev, head, regp, NV_CIO_CRE_FFLWM__INDEX); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_KELVIN) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_KELVIN) wr_cio_state(dev, head, regp, NV_CIO_CRE_47); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) wr_cio_state(dev, head, regp, 0x9f); wr_cio_state(dev, head, regp, NV_CIO_CRE_49); wr_cio_state(dev, head, regp, NV_CIO_CRE_HCUR_ADDR0_INDEX); wr_cio_state(dev, head, regp, NV_CIO_CRE_HCUR_ADDR1_INDEX); wr_cio_state(dev, head, regp, NV_CIO_CRE_HCUR_ADDR2_INDEX); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) nv_fix_nv40_hw_cursor(dev, head); wr_cio_state(dev, head, regp, NV_CIO_CRE_ILACE__INDEX); wr_cio_state(dev, head, regp, NV_CIO_CRE_SCRATCH3__INDEX); wr_cio_state(dev, head, regp, NV_CIO_CRE_SCRATCH4__INDEX); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { wr_cio_state(dev, head, regp, NV_CIO_CRE_EBR_INDEX); wr_cio_state(dev, head, regp, NV_CIO_CRE_CSB); wr_cio_state(dev, head, regp, NV_CIO_CRE_4B); @@ -737,14 +738,14 @@ nv_load_state_ext(struct drm_device *dev, int head, } /* NV11 and NV20 stop at 0x52. */ if (nv_gf4_disp_arch(dev)) { - if (drm->device.info.family < NV_DEVICE_INFO_V0_KELVIN) { + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_KELVIN) { /* Not waiting for vertical retrace before modifying CRE_53/CRE_54 causes lockups. */ - nvif_msec(&drm->device, 650, + nvif_msec(&drm->client.device, 650, if ( (nvif_rd32(device, NV_PRMCIO_INP0__COLOR) & 8)) break; ); - nvif_msec(&drm->device, 650, + nvif_msec(&drm->client.device, 650, if (!(nvif_rd32(device, NV_PRMCIO_INP0__COLOR) & 8)) break; ); @@ -770,7 +771,7 @@ static void nv_save_state_palette(struct drm_device *dev, int head, struct nv04_mode_state *state) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; int head_offset = head * NV_PRMDIO_SIZE, i; nvif_wr08(device, NV_PRMDIO_PIXEL_MASK + head_offset, @@ -789,7 +790,7 @@ void nouveau_hw_load_state_palette(struct drm_device *dev, int head, struct nv04_mode_state *state) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; int head_offset = head * NV_PRMDIO_SIZE, i; nvif_wr08(device, NV_PRMDIO_PIXEL_MASK + head_offset, @@ -809,7 +810,7 @@ void nouveau_hw_save_state(struct drm_device *dev, int head, { struct nouveau_drm *drm = nouveau_drm(dev); - if (drm->device.info.chipset == 0x11) + if (drm->client.device.info.chipset == 0x11) /* NB: no attempt is made to restore the bad pll later on */ nouveau_hw_fix_bad_vpll(dev, head); nv_save_state_ramdac(dev, head, state); diff --git a/drivers/gpu/drm/nouveau/dispnv04/hw.h b/drivers/gpu/drm/nouveau/dispnv04/hw.h index 3bded60c5596..3a2be47fb4f1 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/hw.h +++ b/drivers/gpu/drm/nouveau/dispnv04/hw.h @@ -60,7 +60,7 @@ extern void nouveau_calc_arb(struct drm_device *, int vclk, int bpp, static inline uint32_t NVReadCRTC(struct drm_device *dev, int head, uint32_t reg) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; uint32_t val; if (head) reg += NV_PCRTC0_SIZE; @@ -71,7 +71,7 @@ static inline uint32_t NVReadCRTC(struct drm_device *dev, static inline void NVWriteCRTC(struct drm_device *dev, int head, uint32_t reg, uint32_t val) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; if (head) reg += NV_PCRTC0_SIZE; nvif_wr32(device, reg, val); @@ -80,7 +80,7 @@ static inline void NVWriteCRTC(struct drm_device *dev, static inline uint32_t NVReadRAMDAC(struct drm_device *dev, int head, uint32_t reg) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; uint32_t val; if (head) reg += NV_PRAMDAC0_SIZE; @@ -91,7 +91,7 @@ static inline uint32_t NVReadRAMDAC(struct drm_device *dev, static inline void NVWriteRAMDAC(struct drm_device *dev, int head, uint32_t reg, uint32_t val) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; if (head) reg += NV_PRAMDAC0_SIZE; nvif_wr32(device, reg, val); @@ -120,7 +120,7 @@ static inline void nv_write_tmds(struct drm_device *dev, static inline void NVWriteVgaCrtc(struct drm_device *dev, int head, uint8_t index, uint8_t value) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; nvif_wr08(device, NV_PRMCIO_CRX__COLOR + head * NV_PRMCIO_SIZE, index); nvif_wr08(device, NV_PRMCIO_CR__COLOR + head * NV_PRMCIO_SIZE, value); } @@ -128,7 +128,7 @@ static inline void NVWriteVgaCrtc(struct drm_device *dev, static inline uint8_t NVReadVgaCrtc(struct drm_device *dev, int head, uint8_t index) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; uint8_t val; nvif_wr08(device, NV_PRMCIO_CRX__COLOR + head * NV_PRMCIO_SIZE, index); val = nvif_rd08(device, NV_PRMCIO_CR__COLOR + head * NV_PRMCIO_SIZE); @@ -165,13 +165,13 @@ static inline uint8_t NVReadVgaCrtc5758(struct drm_device *dev, int head, uint8_ static inline uint8_t NVReadPRMVIO(struct drm_device *dev, int head, uint32_t reg) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; struct nouveau_drm *drm = nouveau_drm(dev); uint8_t val; /* Only NV4x have two pvio ranges; other twoHeads cards MUST call * NVSetOwner for the relevant head to be programmed */ - if (head && drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (head && drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) reg += NV_PRMVIO_SIZE; val = nvif_rd08(device, reg); @@ -181,12 +181,12 @@ static inline uint8_t NVReadPRMVIO(struct drm_device *dev, static inline void NVWritePRMVIO(struct drm_device *dev, int head, uint32_t reg, uint8_t value) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; struct nouveau_drm *drm = nouveau_drm(dev); /* Only NV4x have two pvio ranges; other twoHeads cards MUST call * NVSetOwner for the relevant head to be programmed */ - if (head && drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (head && drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) reg += NV_PRMVIO_SIZE; nvif_wr08(device, reg, value); @@ -194,14 +194,14 @@ static inline void NVWritePRMVIO(struct drm_device *dev, static inline void NVSetEnablePalette(struct drm_device *dev, int head, bool enable) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; nvif_rd08(device, NV_PRMCIO_INP0__COLOR + head * NV_PRMCIO_SIZE); nvif_wr08(device, NV_PRMCIO_ARX + head * NV_PRMCIO_SIZE, enable ? 0 : 0x20); } static inline bool NVGetEnablePalette(struct drm_device *dev, int head) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; nvif_rd08(device, NV_PRMCIO_INP0__COLOR + head * NV_PRMCIO_SIZE); return !(nvif_rd08(device, NV_PRMCIO_ARX + head * NV_PRMCIO_SIZE) & 0x20); } @@ -209,7 +209,7 @@ static inline bool NVGetEnablePalette(struct drm_device *dev, int head) static inline void NVWriteVgaAttr(struct drm_device *dev, int head, uint8_t index, uint8_t value) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; if (NVGetEnablePalette(dev, head)) index &= ~0x20; else @@ -223,7 +223,7 @@ static inline void NVWriteVgaAttr(struct drm_device *dev, static inline uint8_t NVReadVgaAttr(struct drm_device *dev, int head, uint8_t index) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; uint8_t val; if (NVGetEnablePalette(dev, head)) index &= ~0x20; @@ -259,10 +259,10 @@ static inline void NVVgaProtect(struct drm_device *dev, int head, bool protect) static inline bool nv_heads_tied(struct drm_device *dev) { - struct nvif_object *device = &nouveau_drm(dev)->device.object; + struct nvif_object *device = &nouveau_drm(dev)->client.device.object; struct nouveau_drm *drm = nouveau_drm(dev); - if (drm->device.info.chipset == 0x11) + if (drm->client.device.info.chipset == 0x11) return !!(nvif_rd32(device, NV_PBUS_DEBUG_1) & (1 << 28)); return NVReadVgaCrtc(dev, 0, NV_CIO_CRE_44) & 0x4; @@ -318,7 +318,7 @@ NVLockVgaCrtcs(struct drm_device *dev, bool lock) NVWriteVgaCrtc(dev, 0, NV_CIO_SR_LOCK_INDEX, lock ? NV_CIO_SR_LOCK_VALUE : NV_CIO_SR_UNLOCK_RW_VALUE); /* NV11 has independently lockable extended crtcs, except when tied */ - if (drm->device.info.chipset == 0x11 && !nv_heads_tied(dev)) + if (drm->client.device.info.chipset == 0x11 && !nv_heads_tied(dev)) NVWriteVgaCrtc(dev, 1, NV_CIO_SR_LOCK_INDEX, lock ? NV_CIO_SR_LOCK_VALUE : NV_CIO_SR_UNLOCK_RW_VALUE); @@ -335,7 +335,7 @@ static inline int nv_cursor_width(struct drm_device *dev) { struct nouveau_drm *drm = nouveau_drm(dev); - return drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS ? NV10_CURSOR_SIZE : NV04_CURSOR_SIZE; + return drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS ? NV10_CURSOR_SIZE : NV04_CURSOR_SIZE; } static inline void @@ -357,7 +357,7 @@ nv_set_crtc_base(struct drm_device *dev, int head, uint32_t offset) NVWriteCRTC(dev, head, NV_PCRTC_START, offset); - if (drm->device.info.family == NV_DEVICE_INFO_V0_TNT) { + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_TNT) { /* * Hilarious, the 24th bit doesn't want to stick to * PCRTC_START... @@ -382,7 +382,7 @@ nv_show_cursor(struct drm_device *dev, int head, bool show) *curctl1 &= ~MASK(NV_CIO_CRE_HCUR_ADDR1_ENABLE); NVWriteVgaCrtc(dev, head, NV_CIO_CRE_HCUR_ADDR1_INDEX, *curctl1); - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) nv_fix_nv40_hw_cursor(dev, head); } @@ -398,7 +398,7 @@ nv_pitch_align(struct drm_device *dev, uint32_t width, int bpp) bpp = 8; /* Alignment requirements taken from the Haiku driver */ - if (drm->device.info.family == NV_DEVICE_INFO_V0_TNT) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_TNT) mask = 128 / bpp - 1; else mask = 512 / bpp - 1; diff --git a/drivers/gpu/drm/nouveau/dispnv04/overlay.c b/drivers/gpu/drm/nouveau/dispnv04/overlay.c index 6275c270df25..5319f2a7f24d 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/overlay.c +++ b/drivers/gpu/drm/nouveau/dispnv04/overlay.c @@ -97,7 +97,7 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, uint32_t src_w, uint32_t src_h) { struct nouveau_drm *drm = nouveau_drm(plane->dev); - struct nvif_object *dev = &drm->device.object; + struct nvif_object *dev = &drm->client.device.object; struct nouveau_plane *nv_plane = container_of(plane, struct nouveau_plane, base); struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb); @@ -119,7 +119,7 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, if (format > 0xffff) return -ERANGE; - if (drm->device.info.chipset >= 0x30) { + if (drm->client.device.info.chipset >= 0x30) { if (crtc_w < (src_w >> 1) || crtc_h < (src_h >> 1)) return -ERANGE; } else { @@ -174,7 +174,7 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, static int nv10_disable_plane(struct drm_plane *plane) { - struct nvif_object *dev = &nouveau_drm(plane->dev)->device.object; + struct nvif_object *dev = &nouveau_drm(plane->dev)->client.device.object; struct nouveau_plane *nv_plane = container_of(plane, struct nouveau_plane, base); @@ -198,7 +198,7 @@ nv_destroy_plane(struct drm_plane *plane) static void nv10_set_params(struct nouveau_plane *plane) { - struct nvif_object *dev = &nouveau_drm(plane->base.dev)->device.object; + struct nvif_object *dev = &nouveau_drm(plane->base.dev)->client.device.object; u32 luma = (plane->brightness - 512) << 16 | plane->contrast; u32 chroma = ((sin_mul(plane->hue, plane->saturation) & 0xffff) << 16) | (cos_mul(plane->hue, plane->saturation) & 0xffff); @@ -268,7 +268,7 @@ nv10_overlay_init(struct drm_device *device) if (!plane) return; - switch (drm->device.info.chipset) { + switch (drm->client.device.info.chipset) { case 0x10: case 0x11: case 0x15: @@ -347,7 +347,7 @@ nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, uint32_t src_x, uint32_t src_y, uint32_t src_w, uint32_t src_h) { - struct nvif_object *dev = &nouveau_drm(plane->dev)->device.object; + struct nvif_object *dev = &nouveau_drm(plane->dev)->client.device.object; struct nouveau_plane *nv_plane = container_of(plane, struct nouveau_plane, base); struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb); @@ -427,7 +427,7 @@ nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, static int nv04_disable_plane(struct drm_plane *plane) { - struct nvif_object *dev = &nouveau_drm(plane->dev)->device.object; + struct nvif_object *dev = &nouveau_drm(plane->dev)->client.device.object; struct nouveau_plane *nv_plane = container_of(plane, struct nouveau_plane, base); @@ -495,7 +495,7 @@ nv04_overlay_init(struct drm_device *device) void nouveau_overlay_init(struct drm_device *device) { - struct nvif_device *dev = &nouveau_drm(device)->device; + struct nvif_device *dev = &nouveau_drm(device)->client.device; if (dev->info.chipset < 0x10) nv04_overlay_init(device); else if (dev->info.chipset <= 0x40) diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c b/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c index 477a8d072af4..01664357d3e1 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c +++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c @@ -54,7 +54,7 @@ static struct nvkm_i2c_bus_probe nv04_tv_encoder_info[] = { int nv04_tv_identify(struct drm_device *dev, int i2c_index) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct nvkm_i2c_bus *bus = nvkm_i2c_bus_find(i2c, i2c_index); if (bus) { return nvkm_i2c_bus_probe(bus, "TV encoder", @@ -206,7 +206,7 @@ nv04_tv_create(struct drm_connector *connector, struct dcb_output *entry) struct drm_encoder *encoder; struct drm_device *dev = connector->dev; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct nvkm_i2c_bus *bus = nvkm_i2c_bus_find(i2c, entry->i2c_index); int type, ret; diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c index 434d1e29f279..6d99f11fee4e 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c +++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c @@ -46,7 +46,7 @@ static uint32_t nv42_tv_sample_load(struct drm_encoder *encoder) { struct drm_device *dev = encoder->dev; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_gpio *gpio = nvxx_gpio(&drm->device); + struct nvkm_gpio *gpio = nvxx_gpio(&drm->client.device); uint32_t testval, regoffset = nv04_dac_output_offset(encoder); uint32_t gpio0, gpio1, fp_htotal, fp_hsync_start, fp_hsync_end, fp_control, test_ctrl, dacclk, ctv_14, ctv_1c, ctv_6c; @@ -130,7 +130,7 @@ static bool get_tv_detect_quirks(struct drm_device *dev, uint32_t *pin_mask) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); if (device->quirk && device->quirk->tv_pin_mask) { *pin_mask = device->quirk->tv_pin_mask; @@ -154,8 +154,8 @@ nv17_tv_detect(struct drm_encoder *encoder, struct drm_connector *connector) return connector_status_disconnected; if (reliable) { - if (drm->device.info.chipset == 0x42 || - drm->device.info.chipset == 0x43) + if (drm->client.device.info.chipset == 0x42 || + drm->client.device.info.chipset == 0x43) tv_enc->pin_mask = nv42_tv_sample_load(encoder) >> 28 & 0xe; else @@ -362,7 +362,7 @@ static void nv17_tv_dpms(struct drm_encoder *encoder, int mode) { struct drm_device *dev = encoder->dev; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_gpio *gpio = nvxx_gpio(&drm->device); + struct nvkm_gpio *gpio = nvxx_gpio(&drm->client.device); struct nv17_tv_state *regs = &to_tv_enc(encoder)->state; struct nv17_tv_norm_params *tv_norm = get_tv_norm(encoder); @@ -435,7 +435,7 @@ static void nv17_tv_prepare(struct drm_encoder *encoder) /* Set the DACCLK register */ dacclk = (NVReadRAMDAC(dev, 0, dacclk_off) & ~0x30) | 0x1; - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE) dacclk |= 0x1a << 16; if (tv_norm->kind == CTV_ENC_MODE) { @@ -492,7 +492,7 @@ static void nv17_tv_mode_set(struct drm_encoder *encoder, tv_regs->ptv_614 = 0x13; } - if (drm->device.info.family >= NV_DEVICE_INFO_V0_RANKINE) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_RANKINE) { tv_regs->ptv_500 = 0xe8e0; tv_regs->ptv_504 = 0x1710; tv_regs->ptv_604 = 0x0; @@ -587,7 +587,7 @@ static void nv17_tv_commit(struct drm_encoder *encoder) nv17_tv_state_load(dev, &to_tv_enc(encoder)->state); /* This could use refinement for flatpanels, but it should work */ - if (drm->device.info.chipset < 0x44) + if (drm->client.device.info.chipset < 0x44) NVWriteRAMDAC(dev, 0, NV_PRAMDAC_TEST_CONTROL + nv04_dac_output_offset(encoder), 0xf0000000); diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.h b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.h index 1b07521cde0d..29773b325bd9 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.h +++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.h @@ -130,13 +130,13 @@ void nv17_ctv_update_rescaler(struct drm_encoder *encoder); static inline void nv_write_ptv(struct drm_device *dev, uint32_t reg, uint32_t val) { - struct nvif_device *device = &nouveau_drm(dev)->device; + struct nvif_device *device = &nouveau_drm(dev)->client.device; nvif_wr32(&device->object, reg, val); } static inline uint32_t nv_read_ptv(struct drm_device *dev, uint32_t reg) { - struct nvif_device *device = &nouveau_drm(dev)->device; + struct nvif_device *device = &nouveau_drm(dev)->client.device; return nvif_rd32(&device->object, reg); } diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl826e.h b/drivers/gpu/drm/nouveau/include/nvif/cl826e.h index 05e6ef7cd190..91e33db21a2f 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/cl826e.h +++ b/drivers/gpu/drm/nouveau/include/nvif/cl826e.h @@ -10,5 +10,5 @@ struct g82_channel_dma_v0 { __u64 offset; }; -#define G82_CHANNEL_DMA_V0_NTFY_UEVENT 0x00 +#define NV826E_V0_NTFY_NON_STALL_INTERRUPT 0x00 #endif diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl826f.h b/drivers/gpu/drm/nouveau/include/nvif/cl826f.h index cecafcb1e954..e34efd4ec537 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/cl826f.h +++ b/drivers/gpu/drm/nouveau/include/nvif/cl826f.h @@ -11,5 +11,5 @@ struct g82_channel_gpfifo_v0 { __u64 vm; }; -#define G82_CHANNEL_GPFIFO_V0_NTFY_UEVENT 0x00 +#define NV826F_V0_NTFY_NON_STALL_INTERRUPT 0x00 #endif diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl906f.h b/drivers/gpu/drm/nouveau/include/nvif/cl906f.h index 2caf0838fcfd..a2d5410a491b 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/cl906f.h +++ b/drivers/gpu/drm/nouveau/include/nvif/cl906f.h @@ -10,5 +10,6 @@ struct fermi_channel_gpfifo_v0 { __u64 vm; }; -#define FERMI_CHANNEL_GPFIFO_V0_NTFY_UEVENT 0x00 +#define NV906F_V0_NTFY_NON_STALL_INTERRUPT 0x00 +#define NV906F_V0_NTFY_KILLED 0x01 #endif diff --git a/drivers/gpu/drm/nouveau/include/nvif/cla06f.h b/drivers/gpu/drm/nouveau/include/nvif/cla06f.h index 46301ec018ce..2efa3d048bb9 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/cla06f.h +++ b/drivers/gpu/drm/nouveau/include/nvif/cla06f.h @@ -25,5 +25,6 @@ struct kepler_channel_gpfifo_a_v0 { __u64 vm; }; -#define NVA06F_V0_NTFY_UEVENT 0x00 +#define NVA06F_V0_NTFY_NON_STALL_INTERRUPT 0x00 +#define NVA06F_V0_NTFY_KILLED 0x01 #endif diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h index 82235f30277c..3a2c0137d4b4 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/class.h +++ b/drivers/gpu/drm/nouveau/include/nvif/class.h @@ -2,23 +2,31 @@ #define __NVIF_CLASS_H__ /* these class numbers are made up by us, and not nvidia-assigned */ -#define NVIF_CLASS_CONTROL /* if0001.h */ -1 -#define NVIF_CLASS_PERFMON /* if0002.h */ -2 -#define NVIF_CLASS_PERFDOM /* if0003.h */ -3 -#define NVIF_CLASS_SW_NV04 /* if0004.h */ -4 -#define NVIF_CLASS_SW_NV10 /* if0005.h */ -5 -#define NVIF_CLASS_SW_NV50 /* if0005.h */ -6 -#define NVIF_CLASS_SW_GF100 /* if0005.h */ -7 +#define NVIF_CLASS_CLIENT /* if0000.h */ -0x00000000 + +#define NVIF_CLASS_CONTROL /* if0001.h */ -0x00000001 + +#define NVIF_CLASS_PERFMON /* if0002.h */ -0x00000002 +#define NVIF_CLASS_PERFDOM /* if0003.h */ -0x00000003 + +#define NVIF_CLASS_SW_NV04 /* if0004.h */ -0x00000004 +#define NVIF_CLASS_SW_NV10 /* if0005.h */ -0x00000005 +#define NVIF_CLASS_SW_NV50 /* if0005.h */ -0x00000006 +#define NVIF_CLASS_SW_GF100 /* if0005.h */ -0x00000007 /* the below match nvidia-assigned (either in hw, or sw) class numbers */ +#define NV_NULL_CLASS 0x00000030 + #define NV_DEVICE /* cl0080.h */ 0x00000080 #define NV_DMA_FROM_MEMORY /* cl0002.h */ 0x00000002 #define NV_DMA_TO_MEMORY /* cl0002.h */ 0x00000003 #define NV_DMA_IN_MEMORY /* cl0002.h */ 0x0000003d +#define NV50_TWOD 0x0000502d #define FERMI_TWOD_A 0x0000902d +#define NV50_MEMORY_TO_MEMORY_FORMAT 0x00005039 #define FERMI_MEMORY_TO_MEMORY_FORMAT_A 0x00009039 #define KEPLER_INLINE_TO_MEMORY_A 0x0000a040 @@ -99,6 +107,12 @@ #define GF110_DISP_OVERLAY_CONTROL_DMA /* cl507e.h */ 0x0000907e #define GK104_DISP_OVERLAY_CONTROL_DMA /* cl507e.h */ 0x0000917e +#define NV50_TESLA 0x00005097 +#define G82_TESLA 0x00008297 +#define GT200_TESLA 0x00008397 +#define GT214_TESLA 0x00008597 +#define GT21A_TESLA 0x00008697 + #define FERMI_A /* cl9097.h */ 0x00009097 #define FERMI_B /* cl9097.h */ 0x00009197 #define FERMI_C /* cl9097.h */ 0x00009297 @@ -140,6 +154,8 @@ #define FERMI_DECOMPRESS 0x000090b8 +#define NV50_COMPUTE 0x000050c0 +#define GT214_COMPUTE 0x000085c0 #define FERMI_COMPUTE_A 0x000090c0 #define FERMI_COMPUTE_B 0x000091c0 #define KEPLER_COMPUTE_A 0x0000a0c0 diff --git a/drivers/gpu/drm/nouveau/include/nvif/client.h b/drivers/gpu/drm/nouveau/include/nvif/client.h index 4a7f6f7b836d..b52a8eadce01 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/client.h +++ b/drivers/gpu/drm/nouveau/include/nvif/client.h @@ -11,8 +11,7 @@ struct nvif_client { bool super; }; -int nvif_client_init(const char *drv, const char *name, u64 device, - const char *cfg, const char *dbg, +int nvif_client_init(struct nvif_client *parent, const char *name, u64 device, struct nvif_client *); void nvif_client_fini(struct nvif_client *); int nvif_client_ioctl(struct nvif_client *, void *, u32); diff --git a/drivers/gpu/drm/nouveau/include/nvif/driver.h b/drivers/gpu/drm/nouveau/include/nvif/driver.h index 8bd39e69229c..0c6f48d8140a 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/driver.h +++ b/drivers/gpu/drm/nouveau/include/nvif/driver.h @@ -1,5 +1,7 @@ #ifndef __NVIF_DRIVER_H__ #define __NVIF_DRIVER_H__ +#include +struct nvif_client; struct nvif_driver { const char *name; @@ -14,9 +16,11 @@ struct nvif_driver { bool keep; }; +int nvif_driver_init(const char *drv, const char *cfg, const char *dbg, + const char *name, u64 device, struct nvif_client *); + extern const struct nvif_driver nvif_driver_nvkm; extern const struct nvif_driver nvif_driver_drm; extern const struct nvif_driver nvif_driver_lib; extern const struct nvif_driver nvif_driver_null; - #endif diff --git a/drivers/gpu/drm/nouveau/include/nvif/if0000.h b/drivers/gpu/drm/nouveau/include/nvif/if0000.h index 85c44e8a1201..c2c0fc41e017 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/if0000.h +++ b/drivers/gpu/drm/nouveau/include/nvif/if0000.h @@ -1,9 +1,16 @@ #ifndef __NVIF_IF0000_H__ #define __NVIF_IF0000_H__ -#define NV_CLIENT_DEVLIST 0x00 +struct nvif_client_v0 { + __u8 version; + __u8 pad01[7]; + __u64 device; + char name[32]; +}; -struct nv_client_devlist_v0 { +#define NVIF_CLIENT_V0_DEVLIST 0x00 + +struct nvif_client_devlist_v0 { __u8 version; __u8 count; __u8 pad02[6]; diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/client.h b/drivers/gpu/drm/nouveau/include/nvkm/core/client.h index eaf5905a87a3..e876634da10a 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/client.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/client.h @@ -1,5 +1,6 @@ #ifndef __NVKM_CLIENT_H__ #define __NVKM_CLIENT_H__ +#define nvkm_client(p) container_of((p), struct nvkm_client, object) #include struct nvkm_client { @@ -8,9 +9,8 @@ struct nvkm_client { u64 device; u32 debug; - struct nvkm_client_notify *notify[16]; + struct nvkm_client_notify *notify[32]; struct rb_root objroot; - struct rb_root dmaroot; bool super; void *data; @@ -19,15 +19,11 @@ struct nvkm_client { struct nvkm_vm *vm; }; -bool nvkm_client_insert(struct nvkm_client *, struct nvkm_object *); -void nvkm_client_remove(struct nvkm_client *, struct nvkm_object *); -struct nvkm_object *nvkm_client_search(struct nvkm_client *, u64 object); - int nvkm_client_new(const char *name, u64 device, const char *cfg, - const char *dbg, struct nvkm_client **); -void nvkm_client_del(struct nvkm_client **); -int nvkm_client_init(struct nvkm_client *); -int nvkm_client_fini(struct nvkm_client *, bool suspend); + const char *dbg, + int (*)(const void *, u32, const void *, u32), + struct nvkm_client **); +struct nvkm_client *nvkm_client_search(struct nvkm_client *, u64 handle); int nvkm_client_notify_new(struct nvkm_object *, struct nvkm_event *, void *data, u32 size); @@ -37,8 +33,8 @@ int nvkm_client_notify_put(struct nvkm_client *, int index); /* logging for client-facing objects */ #define nvif_printk(o,l,p,f,a...) do { \ - struct nvkm_object *_object = (o); \ - struct nvkm_client *_client = _object->client; \ + const struct nvkm_object *_object = (o); \ + const struct nvkm_client *_client = _object->client; \ if (_client->debug >= NV_DBG_##l) \ printk(KERN_##p "nouveau: %s:%08x:%08x: "f, _client->name, \ _object->handle, _object->oclass, ##a); \ diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h index 6bc712f32c8b..d426b86e2712 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h @@ -262,7 +262,7 @@ extern const struct nvkm_sclass nvkm_udevice_sclass; /* device logging */ #define nvdev_printk_(d,l,p,f,a...) do { \ - struct nvkm_device *_device = (d); \ + const struct nvkm_device *_device = (d); \ if (_device->debug >= (l)) \ dev_##p(_device->dev, f, ##a); \ } while(0) diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h b/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h index 9ebfd8782366..d4cd2fbfde88 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h @@ -20,6 +20,7 @@ struct nvkm_engine_func { int (*fini)(struct nvkm_engine *, bool suspend); void (*intr)(struct nvkm_engine *); void (*tile)(struct nvkm_engine *, int region, struct nvkm_fb_tile *); + bool (*chsw_load)(struct nvkm_engine *); struct { int (*sclass)(struct nvkm_oclass *, int index, @@ -44,4 +45,5 @@ int nvkm_engine_new_(const struct nvkm_engine_func *, struct nvkm_device *, struct nvkm_engine *nvkm_engine_ref(struct nvkm_engine *); void nvkm_engine_unref(struct nvkm_engine **); void nvkm_engine_tile(struct nvkm_engine *, int region); +bool nvkm_engine_chsw_load(struct nvkm_engine *); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/memory.h b/drivers/gpu/drm/nouveau/include/nvkm/core/memory.h index 9363b839a9da..33ca6769266a 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/memory.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/memory.h @@ -6,9 +6,10 @@ struct nvkm_vma; struct nvkm_vm; enum nvkm_memory_target { - NVKM_MEM_TARGET_INST, - NVKM_MEM_TARGET_VRAM, - NVKM_MEM_TARGET_HOST, + NVKM_MEM_TARGET_INST, /* instance memory */ + NVKM_MEM_TARGET_VRAM, /* video memory */ + NVKM_MEM_TARGET_HOST, /* coherent system memory */ + NVKM_MEM_TARGET_NCOH, /* non-coherent system memory */ }; struct nvkm_memory { diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/mm.h b/drivers/gpu/drm/nouveau/include/nvkm/core/mm.h index d92fd41e4056..7bd4897a8a2a 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/mm.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/mm.h @@ -5,7 +5,7 @@ struct nvkm_mm_node { struct list_head nl_entry; struct list_head fl_entry; - struct list_head rl_entry; + struct nvkm_mm_node *next; #define NVKM_MM_HEAP_ANY 0x00 u8 heap; @@ -38,4 +38,10 @@ int nvkm_mm_tail(struct nvkm_mm *, u8 heap, u8 type, u32 size_max, u32 size_min, u32 align, struct nvkm_mm_node **); void nvkm_mm_free(struct nvkm_mm *, struct nvkm_mm_node **); void nvkm_mm_dump(struct nvkm_mm *, const char *); + +static inline bool +nvkm_mm_contiguous(struct nvkm_mm_node *node) +{ + return !node->next; +} #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/object.h b/drivers/gpu/drm/nouveau/include/nvkm/core/object.h index dcd048b91fac..96dda350ada3 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/object.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/object.h @@ -62,6 +62,11 @@ int nvkm_object_wr32(struct nvkm_object *, u64 addr, u32 data); int nvkm_object_bind(struct nvkm_object *, struct nvkm_gpuobj *, int align, struct nvkm_gpuobj **); +bool nvkm_object_insert(struct nvkm_object *); +void nvkm_object_remove(struct nvkm_object *); +struct nvkm_object *nvkm_object_search(struct nvkm_client *, u64 object, + const struct nvkm_object_func *); + struct nvkm_sclass { int minver; int maxver; diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h b/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h index 57adefa8b08e..ca9ed3d68f44 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h @@ -32,7 +32,7 @@ void nvkm_subdev_intr(struct nvkm_subdev *); /* subdev logging */ #define nvkm_printk_(s,l,p,f,a...) do { \ - struct nvkm_subdev *_subdev = (s); \ + const struct nvkm_subdev *_subdev = (s); \ if (_subdev->debug >= (l)) { \ dev_##p(_subdev->device->dev, "%s: "f, \ nvkm_subdev_name[_subdev->index], ##a); \ diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h index 114bfb737a81..d2a6532ce3b9 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h @@ -12,9 +12,6 @@ struct nvkm_dmaobj { u32 access; u64 start; u64 limit; - - struct rb_node rb; - u64 handle; /*XXX HANDLE MERGE */ }; struct nvkm_dma { @@ -22,8 +19,7 @@ struct nvkm_dma { struct nvkm_engine engine; }; -struct nvkm_dmaobj * -nvkm_dma_search(struct nvkm_dma *, struct nvkm_client *, u64 object); +struct nvkm_dmaobj *nvkm_dmaobj_search(struct nvkm_client *, u64 object); int nv04_dma_new(struct nvkm_device *, int, struct nvkm_dma **); int nv50_dma_new(struct nvkm_device *, int, struct nvkm_dma **); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/falcon.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/falcon.h index e6baf039c269..7e498e65b1e8 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/falcon.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/falcon.h @@ -4,13 +4,26 @@ #include struct nvkm_fifo_chan; +enum nvkm_falcon_dmaidx { + FALCON_DMAIDX_UCODE = 0, + FALCON_DMAIDX_VIRT = 1, + FALCON_DMAIDX_PHYS_VID = 2, + FALCON_DMAIDX_PHYS_SYS_COH = 3, + FALCON_DMAIDX_PHYS_SYS_NCOH = 4, +}; + struct nvkm_falcon { const struct nvkm_falcon_func *func; - struct nvkm_engine engine; - + const struct nvkm_subdev *owner; + const char *name; u32 addr; - u8 version; - u8 secret; + + struct mutex mutex; + const struct nvkm_subdev *user; + + u8 version; + u8 secret; + bool debug; struct nvkm_memory *core; bool external; @@ -19,15 +32,25 @@ struct nvkm_falcon { u32 limit; u32 *data; u32 size; + u8 ports; } code; struct { u32 limit; u32 *data; u32 size; + u8 ports; } data; + + struct nvkm_engine engine; }; +int nvkm_falcon_v1_new(struct nvkm_subdev *owner, const char *name, u32 addr, + struct nvkm_falcon **); +void nvkm_falcon_del(struct nvkm_falcon **); +int nvkm_falcon_get(struct nvkm_falcon *, const struct nvkm_subdev *); +void nvkm_falcon_put(struct nvkm_falcon *, const struct nvkm_subdev *); + int nvkm_falcon_new_(const struct nvkm_falcon_func *, struct nvkm_device *, int index, bool enable, u32 addr, struct nvkm_engine **); @@ -42,6 +65,51 @@ struct nvkm_falcon_func { } data; void (*init)(struct nvkm_falcon *); void (*intr)(struct nvkm_falcon *, struct nvkm_fifo_chan *); + void (*load_imem)(struct nvkm_falcon *, void *, u32, u32, u16, u8, bool); + void (*load_dmem)(struct nvkm_falcon *, void *, u32, u32, u8); + void (*read_dmem)(struct nvkm_falcon *, u32, u32, u8, void *); + void (*bind_context)(struct nvkm_falcon *, struct nvkm_gpuobj *); + int (*wait_for_halt)(struct nvkm_falcon *, u32); + int (*clear_interrupt)(struct nvkm_falcon *, u32); + void (*set_start_addr)(struct nvkm_falcon *, u32 start_addr); + void (*start)(struct nvkm_falcon *); + int (*enable)(struct nvkm_falcon *falcon); + void (*disable)(struct nvkm_falcon *falcon); + struct nvkm_sclass sclass[]; }; + +static inline u32 +nvkm_falcon_rd32(struct nvkm_falcon *falcon, u32 addr) +{ + return nvkm_rd32(falcon->owner->device, falcon->addr + addr); +} + +static inline void +nvkm_falcon_wr32(struct nvkm_falcon *falcon, u32 addr, u32 data) +{ + nvkm_wr32(falcon->owner->device, falcon->addr + addr, data); +} + +static inline u32 +nvkm_falcon_mask(struct nvkm_falcon *falcon, u32 addr, u32 mask, u32 val) +{ + struct nvkm_device *device = falcon->owner->device; + + return nvkm_mask(device, falcon->addr + addr, mask, val); +} + +void nvkm_falcon_load_imem(struct nvkm_falcon *, void *, u32, u32, u16, u8, + bool); +void nvkm_falcon_load_dmem(struct nvkm_falcon *, void *, u32, u32, u8); +void nvkm_falcon_read_dmem(struct nvkm_falcon *, u32, u32, u8, void *); +void nvkm_falcon_bind_context(struct nvkm_falcon *, struct nvkm_gpuobj *); +void nvkm_falcon_set_start_addr(struct nvkm_falcon *, u32); +void nvkm_falcon_start(struct nvkm_falcon *); +int nvkm_falcon_wait_for_halt(struct nvkm_falcon *, u32); +int nvkm_falcon_clear_interrupt(struct nvkm_falcon *, u32); +int nvkm_falcon_enable(struct nvkm_falcon *); +void nvkm_falcon_disable(struct nvkm_falcon *); +int nvkm_falcon_reset(struct nvkm_falcon *); + #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h index ed92fec5292c..24efa900d8ca 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h @@ -40,6 +40,7 @@ struct nvkm_fifo { struct nvkm_event uevent; /* async user trigger */ struct nvkm_event cevent; /* channel creation event */ + struct nvkm_event kevent; /* channel killed */ }; void nvkm_fifo_pause(struct nvkm_fifo *, unsigned long *); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/power_budget.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/power_budget.h new file mode 100644 index 000000000000..f5f4a14c4030 --- /dev/null +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/power_budget.h @@ -0,0 +1,26 @@ +#ifndef __NVBIOS_POWER_BUDGET_H__ +#define __NVBIOS_POWER_BUDGET_H__ + +#include + +struct nvbios_power_budget_entry { + u32 min_w; + u32 avg_w; + u32 max_w; +}; + +struct nvbios_power_budget { + u32 offset; + u8 ver; + u8 hlen; + u8 elen; + u8 ecount; + u8 cap_entry; +}; + +int nvbios_power_budget_header(struct nvkm_bios *, + struct nvbios_power_budget *); +int nvbios_power_budget_entry(struct nvkm_bios *, struct nvbios_power_budget *, + u8 idx, struct nvbios_power_budget_entry *); + +#endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h index 794e432578b2..0b26a4c860ec 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h @@ -29,7 +29,7 @@ struct nvkm_mem { u8 page_shift; struct nvkm_mm_node *tag; - struct list_head regions; + struct nvkm_mm_node *mem; dma_addr_t *pages; u32 memtype; u64 offset; diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/iccsense.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/iccsense.h index 3c2ddd975273..b7a9b041e130 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/iccsense.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/iccsense.h @@ -8,6 +8,9 @@ struct nvkm_iccsense { bool data_valid; struct list_head sensors; struct list_head rails; + + u32 power_w_max; + u32 power_w_crit; }; int gf100_iccsense_new(struct nvkm_device *, int index, struct nvkm_iccsense **); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h index 27d25b18d85c..e68ba636741b 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h @@ -9,6 +9,7 @@ struct nvkm_mc { void nvkm_mc_enable(struct nvkm_device *, enum nvkm_devidx); void nvkm_mc_disable(struct nvkm_device *, enum nvkm_devidx); +bool nvkm_mc_enabled(struct nvkm_device *, enum nvkm_devidx); void nvkm_mc_reset(struct nvkm_device *, enum nvkm_devidx); void nvkm_mc_intr(struct nvkm_device *, bool *handled); void nvkm_mc_intr_unarm(struct nvkm_device *); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h index e6523e2cea9f..ac2a695963c1 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h @@ -43,6 +43,7 @@ int nv40_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int nv46_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int nv4c_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int g84_pci_new(struct nvkm_device *, int, struct nvkm_pci **); +int g92_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int g94_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int gf100_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int gf106_pci_new(struct nvkm_device *, int, struct nvkm_pci **); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/pmu.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/pmu.h index f37538eb1fe5..179b6ed3f595 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/pmu.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/pmu.h @@ -1,10 +1,12 @@ #ifndef __NVKM_PMU_H__ #define __NVKM_PMU_H__ #include +#include struct nvkm_pmu { const struct nvkm_pmu_func *func; struct nvkm_subdev subdev; + struct nvkm_falcon *falcon; struct { u32 base; @@ -35,6 +37,7 @@ int gk110_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); int gk208_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); int gk20a_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); int gm107_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); +int gm20b_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); int gp100_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); int gp102_pmu_new(struct nvkm_device *, int, struct nvkm_pmu **); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h index b04c38c07761..5dbd8aa4f8c2 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h @@ -26,7 +26,7 @@ #include enum nvkm_secboot_falcon { - NVKM_SECBOOT_FALCON_PMU = 0, + NVKM_SECBOOT_FALCON_PMU = 0, NVKM_SECBOOT_FALCON_RESERVED = 1, NVKM_SECBOOT_FALCON_FECS = 2, NVKM_SECBOOT_FALCON_GPCCS = 3, @@ -35,22 +35,23 @@ enum nvkm_secboot_falcon { }; /** - * @base: base IO address of the falcon performing secure boot - * @irq_mask: IRQ mask of the falcon performing secure boot - * @enable_mask: enable mask of the falcon performing secure boot + * @wpr_set: whether the WPR region is currently set */ struct nvkm_secboot { const struct nvkm_secboot_func *func; + struct nvkm_acr *acr; struct nvkm_subdev subdev; + struct nvkm_falcon *boot_falcon; - enum nvkm_devidx devidx; - u32 base; + u64 wpr_addr; + u32 wpr_size; + + bool wpr_set; }; #define nvkm_secboot(p) container_of((p), struct nvkm_secboot, subdev) bool nvkm_secboot_is_managed(struct nvkm_secboot *, enum nvkm_secboot_falcon); -int nvkm_secboot_reset(struct nvkm_secboot *, u32 falcon); -int nvkm_secboot_start(struct nvkm_secboot *, u32 falcon); +int nvkm_secboot_reset(struct nvkm_secboot *, enum nvkm_secboot_falcon); int gm200_secboot_new(struct nvkm_device *, int, struct nvkm_secboot **); int gm20b_secboot_new(struct nvkm_device *, int, struct nvkm_secboot **); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/timer.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/timer.h index 82d3e28918fd..6a567fe347b3 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/timer.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/timer.h @@ -48,10 +48,8 @@ void nvkm_timer_alarm_cancel(struct nvkm_timer *, struct nvkm_alarm *); } while (_taken = nvkm_timer_read(_tmr) - _time0, _taken < _nsecs); \ \ if (_taken >= _nsecs) { \ - if (_warn) { \ - dev_warn(_device->dev, "timeout at %s:%d/%s()!\n", \ - __FILE__, __LINE__, __func__); \ - } \ + if (_warn) \ + dev_WARN(_device->dev, "timeout\n"); \ _taken = -ETIMEDOUT; \ } \ _taken; \ diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h index 71ebbfd4484f..d23209b62c25 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h @@ -11,6 +11,7 @@ struct nvkm_top { u32 nvkm_top_reset(struct nvkm_device *, enum nvkm_devidx); u32 nvkm_top_intr(struct nvkm_device *, u32 intr, u64 *subdevs); u32 nvkm_top_intr_mask(struct nvkm_device *, enum nvkm_devidx); +int nvkm_top_fault_id(struct nvkm_device *, enum nvkm_devidx); enum nvkm_devidx nvkm_top_fault(struct nvkm_device *, int fault); enum nvkm_devidx nvkm_top_engine(struct nvkm_device *, int, int *runl, int *engn); diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c index 4df4f6ed4886..f98f800cc011 100644 --- a/drivers/gpu/drm/nouveau/nouveau_abi16.c +++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c @@ -87,7 +87,7 @@ nouveau_abi16_put(struct nouveau_abi16 *abi16, int ret) s32 nouveau_abi16_swclass(struct nouveau_drm *drm) { - switch (drm->device.info.family) { + switch (drm->client.device.info.family) { case NV_DEVICE_INFO_V0_TNT: return NVIF_CLASS_SW_NV04; case NV_DEVICE_INFO_V0_CELSIUS: @@ -175,7 +175,7 @@ nouveau_abi16_ioctl_getparam(ABI16_IOCTL_ARGS) { struct nouveau_cli *cli = nouveau_cli(file_priv); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; struct nvkm_gr *gr = nvxx_gr(device); struct drm_nouveau_getparam *getparam = data; @@ -321,7 +321,7 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS) } /* Named memory object area */ - ret = nouveau_gem_new(dev, PAGE_SIZE, 0, NOUVEAU_GEM_DOMAIN_GART, + ret = nouveau_gem_new(cli, PAGE_SIZE, 0, NOUVEAU_GEM_DOMAIN_GART, 0, 0, &chan->ntfy); if (ret == 0) ret = nouveau_bo_pin(chan->ntfy, TTM_PL_FLAG_TT, false); diff --git a/drivers/gpu/drm/nouveau/nouveau_backlight.c b/drivers/gpu/drm/nouveau/nouveau_backlight.c index 8b1ca4add2ed..380f340204e8 100644 --- a/drivers/gpu/drm/nouveau/nouveau_backlight.c +++ b/drivers/gpu/drm/nouveau/nouveau_backlight.c @@ -65,7 +65,7 @@ static int nv40_get_intensity(struct backlight_device *bd) { struct nouveau_drm *drm = bl_get_data(bd); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int val = (nvif_rd32(device, NV40_PMC_BACKLIGHT) & NV40_PMC_BACKLIGHT_MASK) >> 16; @@ -76,7 +76,7 @@ static int nv40_set_intensity(struct backlight_device *bd) { struct nouveau_drm *drm = bl_get_data(bd); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int val = bd->props.brightness; int reg = nvif_rd32(device, NV40_PMC_BACKLIGHT); @@ -96,7 +96,7 @@ static int nv40_backlight_init(struct drm_connector *connector) { struct nouveau_drm *drm = nouveau_drm(connector->dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; struct backlight_properties props; struct backlight_device *bd; struct backlight_connector bl_connector; @@ -133,7 +133,7 @@ nv50_get_intensity(struct backlight_device *bd) { struct nouveau_encoder *nv_encoder = bl_get_data(bd); struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int or = nv_encoder->or; u32 div = 1025; u32 val; @@ -148,7 +148,7 @@ nv50_set_intensity(struct backlight_device *bd) { struct nouveau_encoder *nv_encoder = bl_get_data(bd); struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int or = nv_encoder->or; u32 div = 1025; u32 val = (bd->props.brightness * div) / 100; @@ -169,7 +169,7 @@ nva3_get_intensity(struct backlight_device *bd) { struct nouveau_encoder *nv_encoder = bl_get_data(bd); struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int or = nv_encoder->or; u32 div, val; @@ -187,7 +187,7 @@ nva3_set_intensity(struct backlight_device *bd) { struct nouveau_encoder *nv_encoder = bl_get_data(bd); struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; int or = nv_encoder->or; u32 div, val; @@ -213,7 +213,7 @@ static int nv50_backlight_init(struct drm_connector *connector) { struct nouveau_drm *drm = nouveau_drm(connector->dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; struct nouveau_encoder *nv_encoder; struct backlight_properties props; struct backlight_device *bd; @@ -231,9 +231,9 @@ nv50_backlight_init(struct drm_connector *connector) if (!nvif_rd32(device, NV50_PDISP_SOR_PWM_CTL(nv_encoder->or))) return 0; - if (drm->device.info.chipset <= 0xa0 || - drm->device.info.chipset == 0xaa || - drm->device.info.chipset == 0xac) + if (drm->client.device.info.chipset <= 0xa0 || + drm->client.device.info.chipset == 0xaa || + drm->client.device.info.chipset == 0xac) ops = &nv50_bl_ops; else ops = &nva3_bl_ops; @@ -265,7 +265,7 @@ int nouveau_backlight_init(struct drm_device *dev) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; struct drm_connector *connector; if (apple_gmux_present()) { diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c index 23ffe8571a99..9a0772ad495a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bios.c +++ b/drivers/gpu/drm/nouveau/nouveau_bios.c @@ -215,7 +215,7 @@ int call_lvds_script(struct drm_device *dev, struct dcb_output *dcbent, int head */ struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; struct nvbios *bios = &drm->vbios; uint8_t lvds_ver = bios->data[bios->fp.lvdsmanufacturerpointer]; uint32_t sel_clk_binding, sel_clk; @@ -319,7 +319,7 @@ static int get_fp_strap(struct drm_device *dev, struct nvbios *bios) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; /* * The fp strap is normally dictated by the "User Strap" in @@ -333,10 +333,10 @@ get_fp_strap(struct drm_device *dev, struct nvbios *bios) if (bios->major_version < 5 && bios->data[0x48] & 0x4) return NVReadVgaCrtc5758(dev, 0, 0xf) & 0xf; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_MAXWELL) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_MAXWELL) return nvif_rd32(device, 0x001800) & 0x0000000f; else - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) return (nvif_rd32(device, NV_PEXTDEV_BOOT_0) >> 24) & 0xf; else return (nvif_rd32(device, NV_PEXTDEV_BOOT_0) >> 16) & 0xf; @@ -638,7 +638,7 @@ int run_tmds_table(struct drm_device *dev, struct dcb_output *dcbent, int head, */ struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; struct nvbios *bios = &drm->vbios; int cv = bios->chip_version; uint16_t clktable = 0, scriptptr; @@ -1255,7 +1255,7 @@ olddcb_table(struct drm_device *dev) struct nouveau_drm *drm = nouveau_drm(dev); u8 *dcb = NULL; - if (drm->device.info.family > NV_DEVICE_INFO_V0_TNT) + if (drm->client.device.info.family > NV_DEVICE_INFO_V0_TNT) dcb = ROMPTR(dev, drm->vbios.data[0x36]); if (!dcb) { NV_WARN(drm, "No DCB data found in VBIOS\n"); @@ -1918,7 +1918,7 @@ static int load_nv17_hwsq_ucode_entry(struct drm_device *dev, struct nvbios *bio */ struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; uint8_t bytes_to_write; uint16_t hwsq_entry_offset; int i; @@ -2012,7 +2012,7 @@ uint8_t *nouveau_bios_embedded_edid(struct drm_device *dev) static bool NVInitVBIOS(struct drm_device *dev) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_bios *bios = nvxx_bios(&drm->device); + struct nvkm_bios *bios = nvxx_bios(&drm->client.device); struct nvbios *legacy = &drm->vbios; memset(legacy, 0, sizeof(struct nvbios)); @@ -2064,7 +2064,7 @@ nouveau_bios_posted(struct drm_device *dev) struct nouveau_drm *drm = nouveau_drm(dev); unsigned htotal; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) return true; htotal = NVReadVgaCrtc(dev, 0, 0x06); diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 8a528ebe30f3..548f36d33924 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -48,7 +48,7 @@ nv10_bo_update_tile_region(struct drm_device *dev, struct nouveau_drm_tile *reg, { struct nouveau_drm *drm = nouveau_drm(dev); int i = reg - drm->tile.reg; - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); struct nvkm_fb *fb = device->fb; struct nvkm_fb_tile *tile = &fb->tile.region[i]; @@ -100,7 +100,7 @@ nv10_bo_set_tiling(struct drm_device *dev, u32 addr, u32 size, u32 pitch, u32 flags) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_fb *fb = nvxx_fb(&drm->device); + struct nvkm_fb *fb = nvxx_fb(&drm->client.device); struct nouveau_drm_tile *tile, *found = NULL; int i; @@ -139,60 +139,62 @@ nouveau_bo_del_ttm(struct ttm_buffer_object *bo) kfree(nvbo); } +static inline u64 +roundup_64(u64 x, u32 y) +{ + x += y - 1; + do_div(x, y); + return x * y; +} + static void nouveau_bo_fixup_align(struct nouveau_bo *nvbo, u32 flags, - int *align, int *size) + int *align, u64 *size) { struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; if (device->info.family < NV_DEVICE_INFO_V0_TESLA) { if (nvbo->tile_mode) { if (device->info.chipset >= 0x40) { *align = 65536; - *size = roundup(*size, 64 * nvbo->tile_mode); + *size = roundup_64(*size, 64 * nvbo->tile_mode); } else if (device->info.chipset >= 0x30) { *align = 32768; - *size = roundup(*size, 64 * nvbo->tile_mode); + *size = roundup_64(*size, 64 * nvbo->tile_mode); } else if (device->info.chipset >= 0x20) { *align = 16384; - *size = roundup(*size, 64 * nvbo->tile_mode); + *size = roundup_64(*size, 64 * nvbo->tile_mode); } else if (device->info.chipset >= 0x10) { *align = 16384; - *size = roundup(*size, 32 * nvbo->tile_mode); + *size = roundup_64(*size, 32 * nvbo->tile_mode); } } } else { - *size = roundup(*size, (1 << nvbo->page_shift)); + *size = roundup_64(*size, (1 << nvbo->page_shift)); *align = max((1 << nvbo->page_shift), *align); } - *size = roundup(*size, PAGE_SIZE); + *size = roundup_64(*size, PAGE_SIZE); } int -nouveau_bo_new(struct drm_device *dev, int size, int align, +nouveau_bo_new(struct nouveau_cli *cli, u64 size, int align, uint32_t flags, uint32_t tile_mode, uint32_t tile_flags, struct sg_table *sg, struct reservation_object *robj, struct nouveau_bo **pnvbo) { - struct nouveau_drm *drm = nouveau_drm(dev); + struct nouveau_drm *drm = nouveau_drm(cli->dev); struct nouveau_bo *nvbo; size_t acc_size; int ret; int type = ttm_bo_type_device; - int lpg_shift = 12; - int max_size; - if (drm->client.vm) - lpg_shift = drm->client.vm->mmu->lpg_shift; - max_size = INT_MAX & ~((1 << lpg_shift) - 1); - - if (size <= 0 || size > max_size) { - NV_WARN(drm, "skipped size %x\n", (u32)size); + if (!size) { + NV_WARN(drm, "skipped size %016llx\n", size); return -EINVAL; } @@ -208,8 +210,9 @@ nouveau_bo_new(struct drm_device *dev, int size, int align, nvbo->tile_mode = tile_mode; nvbo->tile_flags = tile_flags; nvbo->bo.bdev = &drm->ttm.bdev; + nvbo->cli = cli; - if (!nvxx_device(&drm->device)->func->cpu_coherent) + if (!nvxx_device(&drm->client.device)->func->cpu_coherent) nvbo->force_coherent = flags & TTM_PL_FLAG_UNCACHED; nvbo->page_shift = 12; @@ -255,10 +258,10 @@ static void set_placement_range(struct nouveau_bo *nvbo, uint32_t type) { struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev); - u32 vram_pages = drm->device.info.ram_size >> PAGE_SHIFT; + u32 vram_pages = drm->client.device.info.ram_size >> PAGE_SHIFT; unsigned i, fpfn, lpfn; - if (drm->device.info.family == NV_DEVICE_INFO_V0_CELSIUS && + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CELSIUS && nvbo->tile_mode && (type & TTM_PL_FLAG_VRAM) && nvbo->bo.mem.num_pages < vram_pages / 4) { /* @@ -316,12 +319,12 @@ nouveau_bo_pin(struct nouveau_bo *nvbo, uint32_t memtype, bool contig) if (ret) return ret; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA && + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA && memtype == TTM_PL_FLAG_VRAM && contig) { if (nvbo->tile_flags & NOUVEAU_GEM_TILE_NONCONTIG) { if (bo->mem.mem_type == TTM_PL_VRAM) { struct nvkm_mem *mem = bo->mem.mm_node; - if (!list_is_singular(&mem->regions)) + if (!nvkm_mm_contiguous(mem->mem)) evict = true; } nvbo->tile_flags &= ~NOUVEAU_GEM_TILE_NONCONTIG; @@ -443,7 +446,7 @@ void nouveau_bo_sync_for_device(struct nouveau_bo *nvbo) { struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev); - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); struct ttm_dma_tt *ttm_dma = (struct ttm_dma_tt *)nvbo->bo.ttm; int i; @@ -463,7 +466,7 @@ void nouveau_bo_sync_for_cpu(struct nouveau_bo *nvbo) { struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev); - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); struct ttm_dma_tt *ttm_dma = (struct ttm_dma_tt *)nvbo->bo.ttm; int i; @@ -579,9 +582,9 @@ nouveau_bo_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, TTM_PL_FLAG_WC; man->default_caching = TTM_PL_FLAG_WC; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) { /* Some BARs do not support being ioremapped WC */ - if (nvxx_bar(&drm->device)->iomap_uncached) { + if (nvxx_bar(&drm->client.device)->iomap_uncached) { man->available_caching = TTM_PL_FLAG_UNCACHED; man->default_caching = TTM_PL_FLAG_UNCACHED; } @@ -594,7 +597,7 @@ nouveau_bo_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, } break; case TTM_PL_TT: - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) man->func = &nouveau_gart_manager; else if (!drm->agp.bridge) @@ -654,20 +657,20 @@ nve0_bo_move_init(struct nouveau_channel *chan, u32 handle) static int nve0_bo_move_copy(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; + struct nvkm_mem *mem = old_reg->mm_node; int ret = RING_SPACE(chan, 10); if (ret == 0) { BEGIN_NVC0(chan, NvSubCopy, 0x0400, 8); - OUT_RING (chan, upper_32_bits(node->vma[0].offset)); - OUT_RING (chan, lower_32_bits(node->vma[0].offset)); - OUT_RING (chan, upper_32_bits(node->vma[1].offset)); - OUT_RING (chan, lower_32_bits(node->vma[1].offset)); + OUT_RING (chan, upper_32_bits(mem->vma[0].offset)); + OUT_RING (chan, lower_32_bits(mem->vma[0].offset)); + OUT_RING (chan, upper_32_bits(mem->vma[1].offset)); + OUT_RING (chan, lower_32_bits(mem->vma[1].offset)); OUT_RING (chan, PAGE_SIZE); OUT_RING (chan, PAGE_SIZE); OUT_RING (chan, PAGE_SIZE); - OUT_RING (chan, new_mem->num_pages); + OUT_RING (chan, new_reg->num_pages); BEGIN_IMC0(chan, NvSubCopy, 0x0300, 0x0386); } return ret; @@ -686,15 +689,15 @@ nvc0_bo_move_init(struct nouveau_channel *chan, u32 handle) static int nvc0_bo_move_copy(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; - u64 src_offset = node->vma[0].offset; - u64 dst_offset = node->vma[1].offset; - u32 page_count = new_mem->num_pages; + struct nvkm_mem *mem = old_reg->mm_node; + u64 src_offset = mem->vma[0].offset; + u64 dst_offset = mem->vma[1].offset; + u32 page_count = new_reg->num_pages; int ret; - page_count = new_mem->num_pages; + page_count = new_reg->num_pages; while (page_count) { int line_count = (page_count > 8191) ? 8191 : page_count; @@ -724,15 +727,15 @@ nvc0_bo_move_copy(struct nouveau_channel *chan, struct ttm_buffer_object *bo, static int nvc0_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; - u64 src_offset = node->vma[0].offset; - u64 dst_offset = node->vma[1].offset; - u32 page_count = new_mem->num_pages; + struct nvkm_mem *mem = old_reg->mm_node; + u64 src_offset = mem->vma[0].offset; + u64 dst_offset = mem->vma[1].offset; + u32 page_count = new_reg->num_pages; int ret; - page_count = new_mem->num_pages; + page_count = new_reg->num_pages; while (page_count) { int line_count = (page_count > 2047) ? 2047 : page_count; @@ -763,15 +766,15 @@ nvc0_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo, static int nva3_bo_move_copy(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; - u64 src_offset = node->vma[0].offset; - u64 dst_offset = node->vma[1].offset; - u32 page_count = new_mem->num_pages; + struct nvkm_mem *mem = old_reg->mm_node; + u64 src_offset = mem->vma[0].offset; + u64 dst_offset = mem->vma[1].offset; + u32 page_count = new_reg->num_pages; int ret; - page_count = new_mem->num_pages; + page_count = new_reg->num_pages; while (page_count) { int line_count = (page_count > 8191) ? 8191 : page_count; @@ -801,35 +804,35 @@ nva3_bo_move_copy(struct nouveau_channel *chan, struct ttm_buffer_object *bo, static int nv98_bo_move_exec(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; + struct nvkm_mem *mem = old_reg->mm_node; int ret = RING_SPACE(chan, 7); if (ret == 0) { BEGIN_NV04(chan, NvSubCopy, 0x0320, 6); - OUT_RING (chan, upper_32_bits(node->vma[0].offset)); - OUT_RING (chan, lower_32_bits(node->vma[0].offset)); - OUT_RING (chan, upper_32_bits(node->vma[1].offset)); - OUT_RING (chan, lower_32_bits(node->vma[1].offset)); + OUT_RING (chan, upper_32_bits(mem->vma[0].offset)); + OUT_RING (chan, lower_32_bits(mem->vma[0].offset)); + OUT_RING (chan, upper_32_bits(mem->vma[1].offset)); + OUT_RING (chan, lower_32_bits(mem->vma[1].offset)); OUT_RING (chan, 0x00000000 /* COPY */); - OUT_RING (chan, new_mem->num_pages << PAGE_SHIFT); + OUT_RING (chan, new_reg->num_pages << PAGE_SHIFT); } return ret; } static int nv84_bo_move_exec(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; + struct nvkm_mem *mem = old_reg->mm_node; int ret = RING_SPACE(chan, 7); if (ret == 0) { BEGIN_NV04(chan, NvSubCopy, 0x0304, 6); - OUT_RING (chan, new_mem->num_pages << PAGE_SHIFT); - OUT_RING (chan, upper_32_bits(node->vma[0].offset)); - OUT_RING (chan, lower_32_bits(node->vma[0].offset)); - OUT_RING (chan, upper_32_bits(node->vma[1].offset)); - OUT_RING (chan, lower_32_bits(node->vma[1].offset)); + OUT_RING (chan, new_reg->num_pages << PAGE_SHIFT); + OUT_RING (chan, upper_32_bits(mem->vma[0].offset)); + OUT_RING (chan, lower_32_bits(mem->vma[0].offset)); + OUT_RING (chan, upper_32_bits(mem->vma[1].offset)); + OUT_RING (chan, lower_32_bits(mem->vma[1].offset)); OUT_RING (chan, 0x00000000 /* MODE_COPY, QUERY_NONE */); } return ret; @@ -853,14 +856,14 @@ nv50_bo_move_init(struct nouveau_channel *chan, u32 handle) static int nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - struct nvkm_mem *node = old_mem->mm_node; - u64 length = (new_mem->num_pages << PAGE_SHIFT); - u64 src_offset = node->vma[0].offset; - u64 dst_offset = node->vma[1].offset; - int src_tiled = !!node->memtype; - int dst_tiled = !!((struct nvkm_mem *)new_mem->mm_node)->memtype; + struct nvkm_mem *mem = old_reg->mm_node; + u64 length = (new_reg->num_pages << PAGE_SHIFT); + u64 src_offset = mem->vma[0].offset; + u64 dst_offset = mem->vma[1].offset; + int src_tiled = !!mem->memtype; + int dst_tiled = !!((struct nvkm_mem *)new_reg->mm_node)->memtype; int ret; while (length) { @@ -940,20 +943,20 @@ nv04_bo_move_init(struct nouveau_channel *chan, u32 handle) static inline uint32_t nouveau_bo_mem_ctxdma(struct ttm_buffer_object *bo, - struct nouveau_channel *chan, struct ttm_mem_reg *mem) + struct nouveau_channel *chan, struct ttm_mem_reg *reg) { - if (mem->mem_type == TTM_PL_TT) + if (reg->mem_type == TTM_PL_TT) return NvDmaTT; return chan->vram.handle; } static int nv04_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo, - struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *old_reg, struct ttm_mem_reg *new_reg) { - u32 src_offset = old_mem->start << PAGE_SHIFT; - u32 dst_offset = new_mem->start << PAGE_SHIFT; - u32 page_count = new_mem->num_pages; + u32 src_offset = old_reg->start << PAGE_SHIFT; + u32 dst_offset = new_reg->start << PAGE_SHIFT; + u32 page_count = new_reg->num_pages; int ret; ret = RING_SPACE(chan, 3); @@ -961,10 +964,10 @@ nv04_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo, return ret; BEGIN_NV04(chan, NvSubCopy, NV_MEMORY_TO_MEMORY_FORMAT_DMA_SOURCE, 2); - OUT_RING (chan, nouveau_bo_mem_ctxdma(bo, chan, old_mem)); - OUT_RING (chan, nouveau_bo_mem_ctxdma(bo, chan, new_mem)); + OUT_RING (chan, nouveau_bo_mem_ctxdma(bo, chan, old_reg)); + OUT_RING (chan, nouveau_bo_mem_ctxdma(bo, chan, new_reg)); - page_count = new_mem->num_pages; + page_count = new_reg->num_pages; while (page_count) { int line_count = (page_count > 2047) ? 2047 : page_count; @@ -995,33 +998,33 @@ nv04_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo, static int nouveau_bo_move_prep(struct nouveau_drm *drm, struct ttm_buffer_object *bo, - struct ttm_mem_reg *mem) + struct ttm_mem_reg *reg) { - struct nvkm_mem *old_node = bo->mem.mm_node; - struct nvkm_mem *new_node = mem->mm_node; - u64 size = (u64)mem->num_pages << PAGE_SHIFT; + struct nvkm_mem *old_mem = bo->mem.mm_node; + struct nvkm_mem *new_mem = reg->mm_node; + u64 size = (u64)reg->num_pages << PAGE_SHIFT; int ret; - ret = nvkm_vm_get(drm->client.vm, size, old_node->page_shift, - NV_MEM_ACCESS_RW, &old_node->vma[0]); + ret = nvkm_vm_get(drm->client.vm, size, old_mem->page_shift, + NV_MEM_ACCESS_RW, &old_mem->vma[0]); if (ret) return ret; - ret = nvkm_vm_get(drm->client.vm, size, new_node->page_shift, - NV_MEM_ACCESS_RW, &old_node->vma[1]); + ret = nvkm_vm_get(drm->client.vm, size, new_mem->page_shift, + NV_MEM_ACCESS_RW, &old_mem->vma[1]); if (ret) { - nvkm_vm_put(&old_node->vma[0]); + nvkm_vm_put(&old_mem->vma[0]); return ret; } - nvkm_vm_map(&old_node->vma[0], old_node); - nvkm_vm_map(&old_node->vma[1], new_node); + nvkm_vm_map(&old_mem->vma[0], old_mem); + nvkm_vm_map(&old_mem->vma[1], new_mem); return 0; } static int nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr, - bool no_wait_gpu, struct ttm_mem_reg *new_mem) + bool no_wait_gpu, struct ttm_mem_reg *new_reg) { struct nouveau_drm *drm = nouveau_bdev(bo->bdev); struct nouveau_channel *chan = drm->ttm.chan; @@ -1033,8 +1036,8 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr, * old nvkm_mem node, these will get cleaned up after ttm has * destroyed the ttm_mem_reg */ - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { - ret = nouveau_bo_move_prep(drm, bo, new_mem); + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + ret = nouveau_bo_move_prep(drm, bo, new_reg); if (ret) return ret; } @@ -1042,14 +1045,14 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr, mutex_lock_nested(&cli->mutex, SINGLE_DEPTH_NESTING); ret = nouveau_fence_sync(nouveau_bo(bo), chan, true, intr); if (ret == 0) { - ret = drm->ttm.move(chan, bo, &bo->mem, new_mem); + ret = drm->ttm.move(chan, bo, &bo->mem, new_reg); if (ret == 0) { ret = nouveau_fence_new(chan, false, &fence); if (ret == 0) { ret = ttm_bo_move_accel_cleanup(bo, &fence->base, evict, - new_mem); + new_reg); nouveau_fence_unref(&fence); } } @@ -1124,7 +1127,7 @@ nouveau_bo_move_init(struct nouveau_drm *drm) static int nouveau_bo_move_flipd(struct ttm_buffer_object *bo, bool evict, bool intr, - bool no_wait_gpu, struct ttm_mem_reg *new_mem) + bool no_wait_gpu, struct ttm_mem_reg *new_reg) { struct ttm_place placement_memtype = { .fpfn = 0, @@ -1132,35 +1135,35 @@ nouveau_bo_move_flipd(struct ttm_buffer_object *bo, bool evict, bool intr, .flags = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING }; struct ttm_placement placement; - struct ttm_mem_reg tmp_mem; + struct ttm_mem_reg tmp_reg; int ret; placement.num_placement = placement.num_busy_placement = 1; placement.placement = placement.busy_placement = &placement_memtype; - tmp_mem = *new_mem; - tmp_mem.mm_node = NULL; - ret = ttm_bo_mem_space(bo, &placement, &tmp_mem, intr, no_wait_gpu); + tmp_reg = *new_reg; + tmp_reg.mm_node = NULL; + ret = ttm_bo_mem_space(bo, &placement, &tmp_reg, intr, no_wait_gpu); if (ret) return ret; - ret = ttm_tt_bind(bo->ttm, &tmp_mem); + ret = ttm_tt_bind(bo->ttm, &tmp_reg); if (ret) goto out; - ret = nouveau_bo_move_m2mf(bo, true, intr, no_wait_gpu, &tmp_mem); + ret = nouveau_bo_move_m2mf(bo, true, intr, no_wait_gpu, &tmp_reg); if (ret) goto out; - ret = ttm_bo_move_ttm(bo, intr, no_wait_gpu, new_mem); + ret = ttm_bo_move_ttm(bo, intr, no_wait_gpu, new_reg); out: - ttm_bo_mem_put(bo, &tmp_mem); + ttm_bo_mem_put(bo, &tmp_reg); return ret; } static int nouveau_bo_move_flips(struct ttm_buffer_object *bo, bool evict, bool intr, - bool no_wait_gpu, struct ttm_mem_reg *new_mem) + bool no_wait_gpu, struct ttm_mem_reg *new_reg) { struct ttm_place placement_memtype = { .fpfn = 0, @@ -1168,34 +1171,34 @@ nouveau_bo_move_flips(struct ttm_buffer_object *bo, bool evict, bool intr, .flags = TTM_PL_FLAG_TT | TTM_PL_MASK_CACHING }; struct ttm_placement placement; - struct ttm_mem_reg tmp_mem; + struct ttm_mem_reg tmp_reg; int ret; placement.num_placement = placement.num_busy_placement = 1; placement.placement = placement.busy_placement = &placement_memtype; - tmp_mem = *new_mem; - tmp_mem.mm_node = NULL; - ret = ttm_bo_mem_space(bo, &placement, &tmp_mem, intr, no_wait_gpu); + tmp_reg = *new_reg; + tmp_reg.mm_node = NULL; + ret = ttm_bo_mem_space(bo, &placement, &tmp_reg, intr, no_wait_gpu); if (ret) return ret; - ret = ttm_bo_move_ttm(bo, intr, no_wait_gpu, &tmp_mem); + ret = ttm_bo_move_ttm(bo, intr, no_wait_gpu, &tmp_reg); if (ret) goto out; - ret = nouveau_bo_move_m2mf(bo, true, intr, no_wait_gpu, new_mem); + ret = nouveau_bo_move_m2mf(bo, true, intr, no_wait_gpu, new_reg); if (ret) goto out; out: - ttm_bo_mem_put(bo, &tmp_mem); + ttm_bo_mem_put(bo, &tmp_reg); return ret; } static void nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, bool evict, - struct ttm_mem_reg *new_mem) + struct ttm_mem_reg *new_reg) { struct nouveau_bo *nvbo = nouveau_bo(bo); struct nvkm_vma *vma; @@ -1205,10 +1208,10 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, bool evict, return; list_for_each_entry(vma, &nvbo->vma_list, head) { - if (new_mem && new_mem->mem_type != TTM_PL_SYSTEM && - (new_mem->mem_type == TTM_PL_VRAM || + if (new_reg && new_reg->mem_type != TTM_PL_SYSTEM && + (new_reg->mem_type == TTM_PL_VRAM || nvbo->page_shift != vma->vm->mmu->lpg_shift)) { - nvkm_vm_map(vma, new_mem->mm_node); + nvkm_vm_map(vma, new_reg->mm_node); } else { WARN_ON(ttm_bo_wait(bo, false, false)); nvkm_vm_unmap(vma); @@ -1217,20 +1220,20 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, bool evict, } static int -nouveau_bo_vm_bind(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem, +nouveau_bo_vm_bind(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_reg, struct nouveau_drm_tile **new_tile) { struct nouveau_drm *drm = nouveau_bdev(bo->bdev); struct drm_device *dev = drm->dev; struct nouveau_bo *nvbo = nouveau_bo(bo); - u64 offset = new_mem->start << PAGE_SHIFT; + u64 offset = new_reg->start << PAGE_SHIFT; *new_tile = NULL; - if (new_mem->mem_type != TTM_PL_VRAM) + if (new_reg->mem_type != TTM_PL_VRAM) return 0; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { - *new_tile = nv10_bo_set_tiling(dev, offset, new_mem->size, + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_CELSIUS) { + *new_tile = nv10_bo_set_tiling(dev, offset, new_reg->size, nvbo->tile_mode, nvbo->tile_flags); } @@ -1253,11 +1256,11 @@ nouveau_bo_vm_cleanup(struct ttm_buffer_object *bo, static int nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr, - bool no_wait_gpu, struct ttm_mem_reg *new_mem) + bool no_wait_gpu, struct ttm_mem_reg *new_reg) { struct nouveau_drm *drm = nouveau_bdev(bo->bdev); struct nouveau_bo *nvbo = nouveau_bo(bo); - struct ttm_mem_reg *old_mem = &bo->mem; + struct ttm_mem_reg *old_reg = &bo->mem; struct nouveau_drm_tile *new_tile = NULL; int ret = 0; @@ -1268,31 +1271,31 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr, if (nvbo->pin_refcnt) NV_WARN(drm, "Moving pinned object %p!\n", nvbo); - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) { - ret = nouveau_bo_vm_bind(bo, new_mem, &new_tile); + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA) { + ret = nouveau_bo_vm_bind(bo, new_reg, &new_tile); if (ret) return ret; } /* Fake bo copy. */ - if (old_mem->mem_type == TTM_PL_SYSTEM && !bo->ttm) { + if (old_reg->mem_type == TTM_PL_SYSTEM && !bo->ttm) { BUG_ON(bo->mem.mm_node != NULL); - bo->mem = *new_mem; - new_mem->mm_node = NULL; + bo->mem = *new_reg; + new_reg->mm_node = NULL; goto out; } /* Hardware assisted copy. */ if (drm->ttm.move) { - if (new_mem->mem_type == TTM_PL_SYSTEM) + if (new_reg->mem_type == TTM_PL_SYSTEM) ret = nouveau_bo_move_flipd(bo, evict, intr, - no_wait_gpu, new_mem); - else if (old_mem->mem_type == TTM_PL_SYSTEM) + no_wait_gpu, new_reg); + else if (old_reg->mem_type == TTM_PL_SYSTEM) ret = nouveau_bo_move_flips(bo, evict, intr, - no_wait_gpu, new_mem); + no_wait_gpu, new_reg); else ret = nouveau_bo_move_m2mf(bo, evict, intr, - no_wait_gpu, new_mem); + no_wait_gpu, new_reg); if (!ret) goto out; } @@ -1300,10 +1303,10 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr, /* Fallback to software copy. */ ret = ttm_bo_wait(bo, intr, no_wait_gpu); if (ret == 0) - ret = ttm_bo_move_memcpy(bo, intr, no_wait_gpu, new_mem); + ret = ttm_bo_move_memcpy(bo, intr, no_wait_gpu, new_reg); out: - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) { + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA) { if (ret) nouveau_bo_vm_cleanup(bo, NULL, &new_tile); else @@ -1323,54 +1326,54 @@ nouveau_bo_verify_access(struct ttm_buffer_object *bo, struct file *filp) } static int -nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem) +nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) { - struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type]; + struct ttm_mem_type_manager *man = &bdev->man[reg->mem_type]; struct nouveau_drm *drm = nouveau_bdev(bdev); - struct nvkm_device *device = nvxx_device(&drm->device); - struct nvkm_mem *node = mem->mm_node; + struct nvkm_device *device = nvxx_device(&drm->client.device); + struct nvkm_mem *mem = reg->mm_node; int ret; - mem->bus.addr = NULL; - mem->bus.offset = 0; - mem->bus.size = mem->num_pages << PAGE_SHIFT; - mem->bus.base = 0; - mem->bus.is_iomem = false; + reg->bus.addr = NULL; + reg->bus.offset = 0; + reg->bus.size = reg->num_pages << PAGE_SHIFT; + reg->bus.base = 0; + reg->bus.is_iomem = false; if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE)) return -EINVAL; - switch (mem->mem_type) { + switch (reg->mem_type) { case TTM_PL_SYSTEM: /* System memory */ return 0; case TTM_PL_TT: #if IS_ENABLED(CONFIG_AGP) if (drm->agp.bridge) { - mem->bus.offset = mem->start << PAGE_SHIFT; - mem->bus.base = drm->agp.base; - mem->bus.is_iomem = !drm->agp.cma; + reg->bus.offset = reg->start << PAGE_SHIFT; + reg->bus.base = drm->agp.base; + reg->bus.is_iomem = !drm->agp.cma; } #endif - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA || !node->memtype) + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA || !mem->memtype) /* untiled */ break; /* fallthrough, tiled memory */ case TTM_PL_VRAM: - mem->bus.offset = mem->start << PAGE_SHIFT; - mem->bus.base = device->func->resource_addr(device, 1); - mem->bus.is_iomem = true; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { - struct nvkm_bar *bar = nvxx_bar(&drm->device); + reg->bus.offset = reg->start << PAGE_SHIFT; + reg->bus.base = device->func->resource_addr(device, 1); + reg->bus.is_iomem = true; + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + struct nvkm_bar *bar = nvxx_bar(&drm->client.device); int page_shift = 12; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_FERMI) - page_shift = node->page_shift; + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_FERMI) + page_shift = mem->page_shift; - ret = nvkm_bar_umap(bar, node->size << 12, page_shift, - &node->bar_vma); + ret = nvkm_bar_umap(bar, mem->size << 12, page_shift, + &mem->bar_vma); if (ret) return ret; - nvkm_vm_map(&node->bar_vma, node); - mem->bus.offset = node->bar_vma.offset; + nvkm_vm_map(&mem->bar_vma, mem); + reg->bus.offset = mem->bar_vma.offset; } break; default: @@ -1380,15 +1383,15 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem) } static void -nouveau_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem) +nouveau_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) { - struct nvkm_mem *node = mem->mm_node; + struct nvkm_mem *mem = reg->mm_node; - if (!node->bar_vma.node) + if (!mem->bar_vma.node) return; - nvkm_vm_unmap(&node->bar_vma); - nvkm_vm_put(&node->bar_vma); + nvkm_vm_unmap(&mem->bar_vma); + nvkm_vm_put(&mem->bar_vma); } static int @@ -1396,7 +1399,7 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo) { struct nouveau_drm *drm = nouveau_bdev(bo->bdev); struct nouveau_bo *nvbo = nouveau_bo(bo); - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); u32 mappable = device->func->resource_size(device, 1) >> PAGE_SHIFT; int i, ret; @@ -1404,7 +1407,7 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo) * nothing to do here. */ if (bo->mem.mem_type != TTM_PL_VRAM) { - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA || + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA || !nouveau_bo_tile_layout(nvbo)) return 0; @@ -1419,7 +1422,7 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo) } /* make sure bo is in mappable vram */ - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA || + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA || bo->mem.start + bo->mem.num_pages < mappable) return 0; @@ -1461,7 +1464,7 @@ nouveau_ttm_tt_populate(struct ttm_tt *ttm) } drm = nouveau_bdev(ttm->bdev); - device = nvxx_device(&drm->device); + device = nvxx_device(&drm->client.device); dev = drm->dev; pdev = device->dev; @@ -1518,7 +1521,7 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm) return; drm = nouveau_bdev(ttm->bdev); - device = nvxx_device(&drm->device); + device = nvxx_device(&drm->client.device); dev = drm->dev; pdev = device->dev; diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h index e42360983229..b06a5385d6dd 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.h +++ b/drivers/gpu/drm/nouveau/nouveau_bo.h @@ -26,6 +26,8 @@ struct nouveau_bo { struct list_head vma_list; unsigned page_shift; + struct nouveau_cli *cli; + u32 tile_mode; u32 tile_flags; struct nouveau_drm_tile *tile; @@ -69,7 +71,7 @@ nouveau_bo_ref(struct nouveau_bo *ref, struct nouveau_bo **pnvbo) extern struct ttm_bo_driver nouveau_bo_driver; void nouveau_bo_move_init(struct nouveau_drm *); -int nouveau_bo_new(struct drm_device *, int size, int align, u32 flags, +int nouveau_bo_new(struct nouveau_cli *, u64 size, int align, u32 flags, u32 tile_mode, u32 tile_flags, struct sg_table *sg, struct reservation_object *robj, struct nouveau_bo **); diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c index f9b3c811187e..dbc41fa86ee8 100644 --- a/drivers/gpu/drm/nouveau/nouveau_chan.c +++ b/drivers/gpu/drm/nouveau/nouveau_chan.c @@ -45,10 +45,20 @@ MODULE_PARM_DESC(vram_pushbuf, "Create DMA push buffers in VRAM"); int nouveau_vram_pushbuf; module_param_named(vram_pushbuf, nouveau_vram_pushbuf, int, 0400); +static int +nouveau_channel_killed(struct nvif_notify *ntfy) +{ + struct nouveau_channel *chan = container_of(ntfy, typeof(*chan), kill); + struct nouveau_cli *cli = (void *)chan->user.client; + NV_PRINTK(warn, cli, "channel %d killed!\n", chan->chid); + atomic_set(&chan->killed, 1); + return NVIF_NOTIFY_DROP; +} + int nouveau_channel_idle(struct nouveau_channel *chan) { - if (likely(chan && chan->fence)) { + if (likely(chan && chan->fence && !atomic_read(&chan->killed))) { struct nouveau_cli *cli = (void *)chan->user.client; struct nouveau_fence *fence = NULL; int ret; @@ -78,6 +88,7 @@ nouveau_channel_del(struct nouveau_channel **pchan) nvif_object_fini(&chan->nvsw); nvif_object_fini(&chan->gart); nvif_object_fini(&chan->vram); + nvif_notify_fini(&chan->kill); nvif_object_fini(&chan->user); nvif_object_fini(&chan->push.ctxdma); nouveau_bo_vma_del(chan->push.buffer, &chan->push.vma); @@ -107,13 +118,14 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device, chan->device = device; chan->drm = drm; + atomic_set(&chan->killed, 0); /* allocate memory for dma push buffer */ target = TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED; if (nouveau_vram_pushbuf) target = TTM_PL_FLAG_VRAM; - ret = nouveau_bo_new(drm->dev, size, 0, target, 0, 0, NULL, NULL, + ret = nouveau_bo_new(cli, size, 0, target, 0, 0, NULL, NULL, &chan->push.buffer); if (ret == 0) { ret = nouveau_bo_pin(chan->push.buffer, target, false); @@ -301,12 +313,26 @@ nouveau_channel_init(struct nouveau_channel *chan, u32 vram, u32 gart) { struct nvif_device *device = chan->device; struct nouveau_cli *cli = (void *)chan->user.client; + struct nouveau_drm *drm = chan->drm; struct nvkm_mmu *mmu = nvxx_mmu(device); struct nv_dma_v0 args = {}; int ret, i; nvif_object_map(&chan->user); + if (chan->user.oclass >= FERMI_CHANNEL_GPFIFO) { + ret = nvif_notify_init(&chan->user, nouveau_channel_killed, + true, NV906F_V0_NTFY_KILLED, + NULL, 0, 0, &chan->kill); + if (ret == 0) + ret = nvif_notify_get(&chan->kill); + if (ret) { + NV_ERROR(drm, "Failed to request channel kill " + "notification: %d\n", ret); + return ret; + } + } + /* allocate dma objects to cover all allowed vram, and gart */ if (device->info.family < NV_DEVICE_INFO_V0_FERMI) { if (device->info.family >= NV_DEVICE_INFO_V0_TESLA) { diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.h b/drivers/gpu/drm/nouveau/nouveau_chan.h index 48062c94f36d..46b947ba1cf4 100644 --- a/drivers/gpu/drm/nouveau/nouveau_chan.h +++ b/drivers/gpu/drm/nouveau/nouveau_chan.h @@ -1,7 +1,7 @@ #ifndef __NOUVEAU_CHAN_H__ #define __NOUVEAU_CHAN_H__ - #include +#include struct nvif_device; struct nouveau_channel { @@ -38,6 +38,9 @@ struct nouveau_channel { u32 user_put; struct nvif_object user; + + struct nvif_notify kill; + atomic_t killed; }; diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c index 966d20ab4de4..f5add64c093f 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.c +++ b/drivers/gpu/drm/nouveau/nouveau_connector.c @@ -419,7 +419,7 @@ nouveau_connector_ddc_detect(struct drm_connector *connector) struct drm_device *dev = connector->dev; struct nouveau_connector *nv_connector = nouveau_connector(connector); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_gpio *gpio = nvxx_gpio(&drm->device); + struct nvkm_gpio *gpio = nvxx_gpio(&drm->client.device); struct nouveau_encoder *nv_encoder; struct drm_encoder *encoder; int i, panel = -ENODEV; @@ -521,7 +521,7 @@ nouveau_connector_set_encoder(struct drm_connector *connector, return; nv_connector->detected_encoder = nv_encoder; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) { connector->interlace_allowed = true; connector->doublescan_allowed = true; } else @@ -531,8 +531,8 @@ nouveau_connector_set_encoder(struct drm_connector *connector, connector->interlace_allowed = false; } else { connector->doublescan_allowed = true; - if (drm->device.info.family == NV_DEVICE_INFO_V0_KELVIN || - (drm->device.info.family == NV_DEVICE_INFO_V0_CELSIUS && + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_KELVIN || + (drm->client.device.info.family == NV_DEVICE_INFO_V0_CELSIUS && (dev->pdev->device & 0x0ff0) != 0x0100 && (dev->pdev->device & 0x0ff0) != 0x0150)) /* HW is broken */ @@ -984,17 +984,17 @@ get_tmds_link_bandwidth(struct drm_connector *connector, bool hdmi) /* Note: these limits are conservative, some Fermi's * can do 297 MHz. Unclear how this can be determined. */ - if (drm->device.info.family >= NV_DEVICE_INFO_V0_KEPLER) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_KEPLER) return 297000; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_FERMI) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_FERMI) return 225000; } if (dcb->location != DCB_LOC_ON_CHIP || - drm->device.info.chipset >= 0x46) + drm->client.device.info.chipset >= 0x46) return 165000; - else if (drm->device.info.chipset >= 0x40) + else if (drm->client.device.info.chipset >= 0x40) return 155000; - else if (drm->device.info.chipset >= 0x18) + else if (drm->client.device.info.chipset >= 0x18) return 135000; else return 112000; @@ -1041,7 +1041,7 @@ nouveau_connector_mode_valid(struct drm_connector *connector, clock = clock * (connector->display_info.bpc * 3) / 10; break; default: - BUG_ON(1); + BUG(); return MODE_BAD; } diff --git a/drivers/gpu/drm/nouveau/nouveau_debugfs.c b/drivers/gpu/drm/nouveau/nouveau_debugfs.c index 411c12cdb249..fd64dfdc7d4f 100644 --- a/drivers/gpu/drm/nouveau/nouveau_debugfs.c +++ b/drivers/gpu/drm/nouveau/nouveau_debugfs.c @@ -259,8 +259,9 @@ nouveau_debugfs_init(struct nouveau_drm *drm) if (!drm->debugfs) return -ENOMEM; - ret = nvif_object_init(&drm->device.object, 0, NVIF_CLASS_CONTROL, - NULL, 0, &drm->debugfs->ctrl); + ret = nvif_object_init(&drm->client.device.object, 0, + NVIF_CLASS_CONTROL, NULL, 0, + &drm->debugfs->ctrl); if (ret) return ret; diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c index 6b570079d185..72fdba1a1c5d 100644 --- a/drivers/gpu/drm/nouveau/nouveau_display.c +++ b/drivers/gpu/drm/nouveau/nouveau_display.c @@ -414,7 +414,8 @@ nouveau_display_init(struct drm_device *dev) return ret; /* enable polling for external displays */ - drm_kms_helper_poll_enable(dev); + if (!dev->mode_config.poll_enabled) + drm_kms_helper_poll_enable(dev); /* enable hotplug interrupts */ list_for_each_entry(connector, &dev->mode_config.connector_list, head) { @@ -495,7 +496,7 @@ int nouveau_display_create(struct drm_device *dev) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); struct nouveau_display *disp; int ret; @@ -512,15 +513,15 @@ nouveau_display_create(struct drm_device *dev) dev->mode_config.min_width = 0; dev->mode_config.min_height = 0; - if (drm->device.info.family < NV_DEVICE_INFO_V0_CELSIUS) { + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_CELSIUS) { dev->mode_config.max_width = 2048; dev->mode_config.max_height = 2048; } else - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) { + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA) { dev->mode_config.max_width = 4096; dev->mode_config.max_height = 4096; } else - if (drm->device.info.family < NV_DEVICE_INFO_V0_FERMI) { + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_FERMI) { dev->mode_config.max_width = 8192; dev->mode_config.max_height = 8192; } else { @@ -531,7 +532,7 @@ nouveau_display_create(struct drm_device *dev) dev->mode_config.preferred_depth = 24; dev->mode_config.prefer_shadow = 1; - if (drm->device.info.chipset < 0x11) + if (drm->client.device.info.chipset < 0x11) dev->mode_config.async_page_flip = false; else dev->mode_config.async_page_flip = true; @@ -558,7 +559,7 @@ nouveau_display_create(struct drm_device *dev) int i; for (i = 0, ret = -ENODEV; ret && i < ARRAY_SIZE(oclass); i++) { - ret = nvif_object_init(&drm->device.object, 0, + ret = nvif_object_init(&drm->client.device.object, 0, oclass[i], NULL, 0, &disp->disp); } @@ -1057,6 +1058,7 @@ int nouveau_display_dumb_create(struct drm_file *file_priv, struct drm_device *dev, struct drm_mode_create_dumb *args) { + struct nouveau_cli *cli = nouveau_cli(file_priv); struct nouveau_bo *bo; uint32_t domain; int ret; @@ -1066,12 +1068,12 @@ nouveau_display_dumb_create(struct drm_file *file_priv, struct drm_device *dev, args->size = roundup(args->size, PAGE_SIZE); /* Use VRAM if there is any ; otherwise fallback to system memory */ - if (nouveau_drm(dev)->device.info.ram_size != 0) + if (nouveau_drm(dev)->client.device.info.ram_size != 0) domain = NOUVEAU_GEM_DOMAIN_VRAM; else domain = NOUVEAU_GEM_DOMAIN_GART; - ret = nouveau_gem_new(dev, args->size, 0, domain, 0, 0, &bo); + ret = nouveau_gem_new(cli, args->size, 0, domain, 0, 0, &bo); if (ret) return ret; diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index dd7b52ab505a..468ed1d3bb26 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -37,6 +37,8 @@ #include #include +#include + #include #include #include @@ -109,35 +111,53 @@ nouveau_name(struct drm_device *dev) return nouveau_platform_name(dev->platformdev); } -static int -nouveau_cli_create(struct drm_device *dev, const char *sname, - int size, void **pcli) -{ - struct nouveau_cli *cli = *pcli = kzalloc(size, GFP_KERNEL); - int ret; - if (cli) { - snprintf(cli->name, sizeof(cli->name), "%s", sname); - cli->dev = dev; - - ret = nvif_client_init(NULL, cli->name, nouveau_name(dev), - nouveau_config, nouveau_debug, - &cli->base); - if (ret == 0) { - mutex_init(&cli->mutex); - usif_client_init(cli); - } - return ret; - } - return -ENOMEM; -} - static void -nouveau_cli_destroy(struct nouveau_cli *cli) +nouveau_cli_fini(struct nouveau_cli *cli) { nvkm_vm_ref(NULL, &nvxx_client(&cli->base)->vm, NULL); - nvif_client_fini(&cli->base); usif_client_fini(cli); - kfree(cli); + nvif_device_fini(&cli->device); + nvif_client_fini(&cli->base); +} + +static int +nouveau_cli_init(struct nouveau_drm *drm, const char *sname, + struct nouveau_cli *cli) +{ + u64 device = nouveau_name(drm->dev); + int ret; + + snprintf(cli->name, sizeof(cli->name), "%s", sname); + cli->dev = drm->dev; + mutex_init(&cli->mutex); + usif_client_init(cli); + + if (cli == &drm->client) { + ret = nvif_driver_init(NULL, nouveau_config, nouveau_debug, + cli->name, device, &cli->base); + } else { + ret = nvif_client_init(&drm->client.base, cli->name, device, + &cli->base); + } + if (ret) { + NV_ERROR(drm, "Client allocation failed: %d\n", ret); + goto done; + } + + ret = nvif_device_init(&cli->base.object, 0, NV_DEVICE, + &(struct nv_device_v0) { + .device = ~0, + }, sizeof(struct nv_device_v0), + &cli->device); + if (ret) { + NV_ERROR(drm, "Device allocation failed: %d\n", ret); + goto done; + } + +done: + if (ret) + nouveau_cli_fini(cli); + return ret; } static void @@ -161,7 +181,7 @@ nouveau_accel_fini(struct nouveau_drm *drm) static void nouveau_accel_init(struct nouveau_drm *drm) { - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; struct nvif_sclass *sclass; u32 arg0, arg1; int ret, i, n; @@ -215,7 +235,7 @@ nouveau_accel_init(struct nouveau_drm *drm) } if (device->info.family >= NV_DEVICE_INFO_V0_KEPLER) { - ret = nouveau_channel_new(drm, &drm->device, + ret = nouveau_channel_new(drm, &drm->client.device, NVA06F_V0_ENGINE_CE0 | NVA06F_V0_ENGINE_CE1, 0, &drm->cechan); @@ -228,7 +248,7 @@ nouveau_accel_init(struct nouveau_drm *drm) if (device->info.chipset >= 0xa3 && device->info.chipset != 0xaa && device->info.chipset != 0xac) { - ret = nouveau_channel_new(drm, &drm->device, + ret = nouveau_channel_new(drm, &drm->client.device, NvDmaFB, NvDmaTT, &drm->cechan); if (ret) NV_ERROR(drm, "failed to create ce channel, %d\n", ret); @@ -240,7 +260,8 @@ nouveau_accel_init(struct nouveau_drm *drm) arg1 = NvDmaTT; } - ret = nouveau_channel_new(drm, &drm->device, arg0, arg1, &drm->channel); + ret = nouveau_channel_new(drm, &drm->client.device, + arg0, arg1, &drm->channel); if (ret) { NV_ERROR(drm, "failed to create kernel channel, %d\n", ret); nouveau_accel_fini(drm); @@ -280,8 +301,8 @@ nouveau_accel_init(struct nouveau_drm *drm) } if (device->info.family < NV_DEVICE_INFO_V0_FERMI) { - ret = nvkm_gpuobj_new(nvxx_device(&drm->device), 32, 0, false, - NULL, &drm->notify); + ret = nvkm_gpuobj_new(nvxx_device(&drm->client.device), 32, 0, + false, NULL, &drm->notify); if (ret) { NV_ERROR(drm, "failed to allocate notifier, %d\n", ret); nouveau_accel_fini(drm); @@ -407,12 +428,17 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags) struct nouveau_drm *drm; int ret; - ret = nouveau_cli_create(dev, "DRM", sizeof(*drm), (void **)&drm); + if (!(drm = kzalloc(sizeof(*drm), GFP_KERNEL))) + return -ENOMEM; + dev->dev_private = drm; + drm->dev = dev; + + ret = nouveau_cli_init(drm, "DRM", &drm->client); if (ret) return ret; - dev->dev_private = drm; - drm->dev = dev; + dev->irq_enabled = true; + nvxx_client(&drm->client.base)->debug = nvkm_dbgopt(nouveau_debug, "DRM"); @@ -421,33 +447,24 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags) nouveau_get_hdmi_dev(drm); - ret = nvif_device_init(&drm->client.base.object, 0, NV_DEVICE, - &(struct nv_device_v0) { - .device = ~0, - }, sizeof(struct nv_device_v0), - &drm->device); - if (ret) - goto fail_device; - - dev->irq_enabled = true; - /* workaround an odd issue on nvc1 by disabling the device's * nosnoop capability. hopefully won't cause issues until a * better fix is found - assuming there is one... */ - if (drm->device.info.chipset == 0xc1) - nvif_mask(&drm->device.object, 0x00088080, 0x00000800, 0x00000000); + if (drm->client.device.info.chipset == 0xc1) + nvif_mask(&drm->client.device.object, 0x00088080, 0x00000800, 0x00000000); nouveau_vga_init(drm); - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { - if (!nvxx_device(&drm->device)->mmu) { + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + if (!nvxx_device(&drm->client.device)->mmu) { ret = -ENOSYS; goto fail_device; } - ret = nvkm_vm_new(nvxx_device(&drm->device), 0, (1ULL << 40), - 0x1000, NULL, &drm->client.vm); + ret = nvkm_vm_new(nvxx_device(&drm->client.device), + 0, (1ULL << 40), 0x1000, NULL, + &drm->client.vm); if (ret) goto fail_device; @@ -497,8 +514,8 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags) fail_ttm: nouveau_vga_fini(drm); fail_device: - nvif_device_fini(&drm->device); - nouveau_cli_destroy(&drm->client); + nouveau_cli_fini(&drm->client); + kfree(drm); return ret; } @@ -527,10 +544,10 @@ nouveau_drm_unload(struct drm_device *dev) nouveau_ttm_fini(drm); nouveau_vga_fini(drm); - nvif_device_fini(&drm->device); if (drm->hdmi_device) pci_dev_put(drm->hdmi_device); - nouveau_cli_destroy(&drm->client); + nouveau_cli_fini(&drm->client); + kfree(drm); } void @@ -560,7 +577,6 @@ static int nouveau_do_suspend(struct drm_device *dev, bool runtime) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nouveau_cli *cli; int ret; nouveau_led_suspend(dev); @@ -590,7 +606,7 @@ nouveau_do_suspend(struct drm_device *dev, bool runtime) goto fail_display; } - NV_INFO(drm, "suspending client object trees...\n"); + NV_INFO(drm, "suspending fence...\n"); if (drm->fence && nouveau_fence(drm)->suspend) { if (!nouveau_fence(drm)->suspend(drm)) { ret = -ENOMEM; @@ -598,13 +614,7 @@ nouveau_do_suspend(struct drm_device *dev, bool runtime) } } - list_for_each_entry(cli, &drm->clients, head) { - ret = nvif_client_suspend(&cli->base); - if (ret) - goto fail_client; - } - - NV_INFO(drm, "suspending kernel object tree...\n"); + NV_INFO(drm, "suspending object tree...\n"); ret = nvif_client_suspend(&drm->client.base); if (ret) goto fail_client; @@ -612,10 +622,6 @@ nouveau_do_suspend(struct drm_device *dev, bool runtime) return 0; fail_client: - list_for_each_entry_continue_reverse(cli, &drm->clients, head) { - nvif_client_resume(&cli->base); - } - if (drm->fence && nouveau_fence(drm)->resume) nouveau_fence(drm)->resume(drm); @@ -631,19 +637,14 @@ static int nouveau_do_resume(struct drm_device *dev, bool runtime) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nouveau_cli *cli; - NV_INFO(drm, "resuming kernel object tree...\n"); + NV_INFO(drm, "resuming object tree...\n"); nvif_client_resume(&drm->client.base); - NV_INFO(drm, "resuming client object trees...\n"); + NV_INFO(drm, "resuming fence...\n"); if (drm->fence && nouveau_fence(drm)->resume) nouveau_fence(drm)->resume(drm); - list_for_each_entry(cli, &drm->clients, head) { - nvif_client_resume(&cli->base); - } - nouveau_run_vbios_init(dev); if (dev->mode_config.num_crtc) { @@ -758,7 +759,7 @@ nouveau_pmops_runtime_resume(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct drm_device *drm_dev = pci_get_drvdata(pdev); - struct nvif_device *device = &nouveau_drm(drm_dev)->device; + struct nvif_device *device = &nouveau_drm(drm_dev)->client.device; int ret; if (nouveau_runtime_pm == 0) @@ -772,7 +773,10 @@ nouveau_pmops_runtime_resume(struct device *dev) pci_set_master(pdev); ret = nouveau_do_resume(drm_dev, true); - drm_kms_helper_poll_enable(drm_dev); + + if (!drm_dev->mode_config.poll_enabled) + drm_kms_helper_poll_enable(drm_dev); + /* do magic */ nvif_mask(&device->object, 0x088488, (1 << 25), (1 << 25)); vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON); @@ -841,20 +845,20 @@ nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv) get_task_comm(tmpname, current); snprintf(name, sizeof(name), "%s[%d]", tmpname, pid_nr(fpriv->pid)); - ret = nouveau_cli_create(dev, name, sizeof(*cli), (void **)&cli); + if (!(cli = kzalloc(sizeof(*cli), GFP_KERNEL))) + return ret; + ret = nouveau_cli_init(drm, name, cli); if (ret) - goto out_suspend; + goto done; cli->base.super = false; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { - ret = nvkm_vm_new(nvxx_device(&drm->device), 0, (1ULL << 40), - 0x1000, NULL, &cli->vm); - if (ret) { - nouveau_cli_destroy(cli); - goto out_suspend; - } + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + ret = nvkm_vm_new(nvxx_device(&drm->client.device), 0, + (1ULL << 40), 0x1000, NULL, &cli->vm); + if (ret) + goto done; nvxx_client(&cli->base)->vm = cli->vm; } @@ -865,10 +869,14 @@ nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv) list_add(&cli->head, &drm->clients); mutex_unlock(&drm->client.mutex); -out_suspend: +done: + if (ret && cli) { + nouveau_cli_fini(cli); + kfree(cli); + } + pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); - return ret; } @@ -895,7 +903,8 @@ static void nouveau_drm_postclose(struct drm_device *dev, struct drm_file *fpriv) { struct nouveau_cli *cli = nouveau_cli(fpriv); - nouveau_cli_destroy(cli); + nouveau_cli_fini(cli); + kfree(cli); pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); } diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index 8d5ed5bfdacb..eadec2f49ad3 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -86,14 +86,17 @@ enum nouveau_drm_handle { struct nouveau_cli { struct nvif_client base; + struct drm_device *dev; + struct mutex mutex; + + struct nvif_device device; + struct nvkm_vm *vm; /*XXX*/ struct list_head head; - struct mutex mutex; void *abi16; struct list_head objects; struct list_head notifys; char name[32]; - struct drm_device *dev; }; static inline struct nouveau_cli * @@ -111,7 +114,6 @@ struct nouveau_drm { struct nouveau_cli client; struct drm_device *dev; - struct nvif_device device; struct list_head clients; struct { @@ -165,6 +167,8 @@ struct nouveau_drm { struct backlight_device *backlight; struct list_head bl_connectors; struct work_struct hpd_work; + struct work_struct fbcon_work; + int fbcon_new_state; #ifdef CONFIG_ACPI struct notifier_block acpi_nb; #endif diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c index 971c147a3984..442e25c17383 100644 --- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c +++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c @@ -60,7 +60,7 @@ nouveau_fbcon_fillrect(struct fb_info *info, const struct fb_fillrect *rect) { struct nouveau_fbdev *fbcon = info->par; struct nouveau_drm *drm = nouveau_drm(fbcon->helper.dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; int ret; if (info->state != FBINFO_STATE_RUNNING) @@ -92,7 +92,7 @@ nouveau_fbcon_copyarea(struct fb_info *info, const struct fb_copyarea *image) { struct nouveau_fbdev *fbcon = info->par; struct nouveau_drm *drm = nouveau_drm(fbcon->helper.dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; int ret; if (info->state != FBINFO_STATE_RUNNING) @@ -124,7 +124,7 @@ nouveau_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) { struct nouveau_fbdev *fbcon = info->par; struct nouveau_drm *drm = nouveau_drm(fbcon->helper.dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; int ret; if (info->state != FBINFO_STATE_RUNNING) @@ -266,10 +266,10 @@ nouveau_fbcon_accel_init(struct drm_device *dev) struct fb_info *info = fbcon->helper.fbdev; int ret; - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA) ret = nv04_fbcon_accel_init(info); else - if (drm->device.info.family < NV_DEVICE_INFO_V0_FERMI) + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_FERMI) ret = nv50_fbcon_accel_init(info); else ret = nvc0_fbcon_accel_init(info); @@ -324,7 +324,7 @@ nouveau_fbcon_create(struct drm_fb_helper *helper, container_of(helper, struct nouveau_fbdev, helper); struct drm_device *dev = fbcon->helper.dev; struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; struct fb_info *info; struct nouveau_framebuffer *fb; struct nouveau_channel *chan; @@ -341,8 +341,9 @@ nouveau_fbcon_create(struct drm_fb_helper *helper, mode_cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp, sizes->surface_depth); - ret = nouveau_gem_new(dev, mode_cmd.pitches[0] * mode_cmd.height, - 0, NOUVEAU_GEM_DOMAIN_VRAM, 0, 0x0000, &nvbo); + ret = nouveau_gem_new(&drm->client, mode_cmd.pitches[0] * + mode_cmd.height, 0, NOUVEAU_GEM_DOMAIN_VRAM, + 0, 0x0000, &nvbo); if (ret) { NV_ERROR(drm, "failed to allocate framebuffer\n"); goto out; @@ -472,19 +473,43 @@ static const struct drm_fb_helper_funcs nouveau_fbcon_helper_funcs = { .fb_probe = nouveau_fbcon_create, }; +static void +nouveau_fbcon_set_suspend_work(struct work_struct *work) +{ + struct nouveau_drm *drm = container_of(work, typeof(*drm), fbcon_work); + int state = READ_ONCE(drm->fbcon_new_state); + + if (state == FBINFO_STATE_RUNNING) + pm_runtime_get_sync(drm->dev->dev); + + console_lock(); + if (state == FBINFO_STATE_RUNNING) + nouveau_fbcon_accel_restore(drm->dev); + drm_fb_helper_set_suspend(&drm->fbcon->helper, state); + if (state != FBINFO_STATE_RUNNING) + nouveau_fbcon_accel_save_disable(drm->dev); + console_unlock(); + + if (state == FBINFO_STATE_RUNNING) { + pm_runtime_mark_last_busy(drm->dev->dev); + pm_runtime_put_sync(drm->dev->dev); + } +} + void nouveau_fbcon_set_suspend(struct drm_device *dev, int state) { struct nouveau_drm *drm = nouveau_drm(dev); - if (drm->fbcon) { - console_lock(); - if (state == FBINFO_STATE_RUNNING) - nouveau_fbcon_accel_restore(dev); - drm_fb_helper_set_suspend(&drm->fbcon->helper, state); - if (state != FBINFO_STATE_RUNNING) - nouveau_fbcon_accel_save_disable(dev); - console_unlock(); - } + + if (!drm->fbcon) + return; + + drm->fbcon_new_state = state; + /* Since runtime resume can happen as a result of a sysfs operation, + * it's possible we already have the console locked. So handle fbcon + * init/deinit from a seperate work thread + */ + schedule_work(&drm->fbcon_work); } int @@ -504,6 +529,7 @@ nouveau_fbcon_init(struct drm_device *dev) return -ENOMEM; drm->fbcon = fbcon; + INIT_WORK(&drm->fbcon_work, nouveau_fbcon_set_suspend_work); drm_fb_helper_prepare(dev, &fbcon->helper, &nouveau_fbcon_helper_funcs); @@ -515,10 +541,10 @@ nouveau_fbcon_init(struct drm_device *dev) if (ret) goto fini; - if (drm->device.info.ram_size <= 32 * 1024 * 1024) + if (drm->client.device.info.ram_size <= 32 * 1024 * 1024) preferred_bpp = 8; else - if (drm->device.info.ram_size <= 64 * 1024 * 1024) + if (drm->client.device.info.ram_size <= 64 * 1024 * 1024) preferred_bpp = 16; else preferred_bpp = 32; diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c index a6126c93f215..f3e551f1aa46 100644 --- a/drivers/gpu/drm/nouveau/nouveau_fence.c +++ b/drivers/gpu/drm/nouveau/nouveau_fence.c @@ -190,7 +190,7 @@ nouveau_fence_context_new(struct nouveau_channel *chan, struct nouveau_fence_cha return; ret = nvif_notify_init(&chan->user, nouveau_fence_wait_uevent_handler, - false, G82_CHANNEL_DMA_V0_NTFY_UEVENT, + false, NV826E_V0_NTFY_NON_STALL_INTERRUPT, &(struct nvif_notify_uevent_req) { }, sizeof(struct nvif_notify_uevent_req), sizeof(struct nvif_notify_uevent_rep), diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.h b/drivers/gpu/drm/nouveau/nouveau_fence.h index ccdce1b4eec4..d5e58a38f160 100644 --- a/drivers/gpu/drm/nouveau/nouveau_fence.h +++ b/drivers/gpu/drm/nouveau/nouveau_fence.h @@ -99,6 +99,7 @@ struct nv84_fence_priv { struct nouveau_bo *bo; struct nouveau_bo *bo_gart; u32 *suspend; + struct mutex mutex; }; int nv84_fence_context_new(struct nouveau_channel *); diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c index 201b52b750dd..ca5397beb357 100644 --- a/drivers/gpu/drm/nouveau/nouveau_gem.c +++ b/drivers/gpu/drm/nouveau/nouveau_gem.c @@ -175,11 +175,11 @@ nouveau_gem_object_close(struct drm_gem_object *gem, struct drm_file *file_priv) } int -nouveau_gem_new(struct drm_device *dev, int size, int align, uint32_t domain, +nouveau_gem_new(struct nouveau_cli *cli, u64 size, int align, uint32_t domain, uint32_t tile_mode, uint32_t tile_flags, struct nouveau_bo **pnvbo) { - struct nouveau_drm *drm = nouveau_drm(dev); + struct nouveau_drm *drm = nouveau_drm(cli->dev); struct nouveau_bo *nvbo; u32 flags = 0; int ret; @@ -194,7 +194,7 @@ nouveau_gem_new(struct drm_device *dev, int size, int align, uint32_t domain, if (domain & NOUVEAU_GEM_DOMAIN_COHERENT) flags |= TTM_PL_FLAG_UNCACHED; - ret = nouveau_bo_new(dev, size, align, flags, tile_mode, + ret = nouveau_bo_new(cli, size, align, flags, tile_mode, tile_flags, NULL, NULL, pnvbo); if (ret) return ret; @@ -206,12 +206,12 @@ nouveau_gem_new(struct drm_device *dev, int size, int align, uint32_t domain, */ nvbo->valid_domains = NOUVEAU_GEM_DOMAIN_VRAM | NOUVEAU_GEM_DOMAIN_GART; - if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) + if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) nvbo->valid_domains &= domain; /* Initialize the embedded gem-object. We return a single gem-reference * to the caller, instead of a normal nouveau_bo ttm reference. */ - ret = drm_gem_object_init(dev, &nvbo->gem, nvbo->bo.mem.size); + ret = drm_gem_object_init(drm->dev, &nvbo->gem, nvbo->bo.mem.size); if (ret) { nouveau_bo_ref(NULL, pnvbo); return -ENOMEM; @@ -257,7 +257,7 @@ nouveau_gem_ioctl_new(struct drm_device *dev, void *data, { struct nouveau_drm *drm = nouveau_drm(dev); struct nouveau_cli *cli = nouveau_cli(file_priv); - struct nvkm_fb *fb = nvxx_fb(&drm->device); + struct nvkm_fb *fb = nvxx_fb(&drm->client.device); struct drm_nouveau_gem_new *req = data; struct nouveau_bo *nvbo = NULL; int ret = 0; @@ -267,7 +267,7 @@ nouveau_gem_ioctl_new(struct drm_device *dev, void *data, return -EINVAL; } - ret = nouveau_gem_new(dev, req->info.size, req->align, + ret = nouveau_gem_new(cli, req->info.size, req->align, req->info.domain, req->info.tile_mode, req->info.tile_flags, &nvbo); if (ret) @@ -496,7 +496,7 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli, return ret; } - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) { + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA) { if (nvbo->bo.offset == b->presumed.offset && ((nvbo->bo.mem.mem_type == TTM_PL_VRAM && b->presumed.domain & NOUVEAU_GEM_DOMAIN_VRAM) || @@ -767,7 +767,7 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data, push[i].length); } } else - if (drm->device.info.chipset >= 0x25) { + if (drm->client.device.info.chipset >= 0x25) { ret = RING_SPACE(chan, req->nr_push * 2); if (ret) { NV_PRINTK(err, cli, "cal_space: %d\n", ret); @@ -840,7 +840,7 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data, req->suffix0 = 0x00000000; req->suffix1 = 0x00000000; } else - if (drm->device.info.chipset >= 0x25) { + if (drm->client.device.info.chipset >= 0x25) { req->suffix0 = 0x00020000; req->suffix1 = 0x00000000; } else { diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.h b/drivers/gpu/drm/nouveau/nouveau_gem.h index 7e32da2e037a..8fa6ed9ddd3a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_gem.h +++ b/drivers/gpu/drm/nouveau/nouveau_gem.h @@ -16,7 +16,7 @@ nouveau_gem_object(struct drm_gem_object *gem) } /* nouveau_gem.c */ -extern int nouveau_gem_new(struct drm_device *, int size, int align, +extern int nouveau_gem_new(struct nouveau_cli *, u64 size, int align, uint32_t domain, uint32_t tile_mode, uint32_t tile_flags, struct nouveau_bo **); extern void nouveau_gem_object_del(struct drm_gem_object *); diff --git a/drivers/gpu/drm/nouveau/nouveau_hwmon.c b/drivers/gpu/drm/nouveau/nouveau_hwmon.c index 71f764bf4cc6..23b1670c1c2f 100644 --- a/drivers/gpu/drm/nouveau/nouveau_hwmon.c +++ b/drivers/gpu/drm/nouveau/nouveau_hwmon.c @@ -43,7 +43,7 @@ nouveau_hwmon_show_temp(struct device *d, struct device_attribute *a, char *buf) { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); int temp = nvkm_therm_temp_get(therm); if (temp < 0) @@ -69,7 +69,7 @@ nouveau_hwmon_temp1_auto_point1_temp(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_FAN_BOOST) * 1000); @@ -81,7 +81,7 @@ nouveau_hwmon_set_temp1_auto_point1_temp(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -102,7 +102,7 @@ nouveau_hwmon_temp1_auto_point1_temp_hyst(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_FAN_BOOST_HYST) * 1000); @@ -114,7 +114,7 @@ nouveau_hwmon_set_temp1_auto_point1_temp_hyst(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -134,7 +134,7 @@ nouveau_hwmon_max_temp(struct device *d, struct device_attribute *a, char *buf) { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_DOWN_CLK) * 1000); @@ -145,7 +145,7 @@ nouveau_hwmon_set_max_temp(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -165,7 +165,7 @@ nouveau_hwmon_max_temp_hyst(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_DOWN_CLK_HYST) * 1000); @@ -176,7 +176,7 @@ nouveau_hwmon_set_max_temp_hyst(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -197,7 +197,7 @@ nouveau_hwmon_critical_temp(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_CRITICAL) * 1000); @@ -209,7 +209,7 @@ nouveau_hwmon_set_critical_temp(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -230,7 +230,7 @@ nouveau_hwmon_critical_temp_hyst(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_CRITICAL_HYST) * 1000); @@ -243,7 +243,7 @@ nouveau_hwmon_set_critical_temp_hyst(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -263,7 +263,7 @@ nouveau_hwmon_emergency_temp(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_SHUTDOWN) * 1000); @@ -275,7 +275,7 @@ nouveau_hwmon_set_emergency_temp(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -296,7 +296,7 @@ nouveau_hwmon_emergency_temp_hyst(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", therm->attr_get(therm, NVKM_THERM_ATTR_THRS_SHUTDOWN_HYST) * 1000); @@ -309,7 +309,7 @@ nouveau_hwmon_set_emergency_temp_hyst(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; if (kstrtol(buf, 10, &value) == -EINVAL) @@ -349,7 +349,7 @@ nouveau_hwmon_show_fan1_input(struct device *d, struct device_attribute *attr, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); return snprintf(buf, PAGE_SIZE, "%d\n", nvkm_therm_fan_sense(therm)); } @@ -362,7 +362,7 @@ nouveau_hwmon_get_pwm1_enable(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); int ret; ret = therm->attr_get(therm, NVKM_THERM_ATTR_FAN_MODE); @@ -378,7 +378,7 @@ nouveau_hwmon_set_pwm1_enable(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; int ret; @@ -401,7 +401,7 @@ nouveau_hwmon_get_pwm1(struct device *d, struct device_attribute *a, char *buf) { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); int ret; ret = therm->fan_get(therm); @@ -417,7 +417,7 @@ nouveau_hwmon_set_pwm1(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); int ret = -ENODEV; long value; @@ -441,7 +441,7 @@ nouveau_hwmon_get_pwm1_min(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); int ret; ret = therm->attr_get(therm, NVKM_THERM_ATTR_FAN_MIN_DUTY); @@ -457,7 +457,7 @@ nouveau_hwmon_set_pwm1_min(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; int ret; @@ -481,7 +481,7 @@ nouveau_hwmon_get_pwm1_max(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); int ret; ret = therm->attr_get(therm, NVKM_THERM_ATTR_FAN_MAX_DUTY); @@ -497,7 +497,7 @@ nouveau_hwmon_set_pwm1_max(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); long value; int ret; @@ -521,7 +521,7 @@ nouveau_hwmon_get_in0_input(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_volt *volt = nvxx_volt(&drm->device); + struct nvkm_volt *volt = nvxx_volt(&drm->client.device); int ret; ret = nvkm_volt_get(volt); @@ -540,7 +540,7 @@ nouveau_hwmon_get_in0_min(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_volt *volt = nvxx_volt(&drm->device); + struct nvkm_volt *volt = nvxx_volt(&drm->client.device); if (!volt || !volt->min_uv) return -ENODEV; @@ -557,7 +557,7 @@ nouveau_hwmon_get_in0_max(struct device *d, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_volt *volt = nvxx_volt(&drm->device); + struct nvkm_volt *volt = nvxx_volt(&drm->client.device); if (!volt || !volt->max_uv) return -ENODEV; @@ -584,7 +584,7 @@ nouveau_hwmon_get_power1_input(struct device *d, struct device_attribute *a, { struct drm_device *dev = dev_get_drvdata(d); struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->device); + struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->client.device); int result = nvkm_iccsense_read_all(iccsense); if (result < 0) @@ -596,6 +596,32 @@ nouveau_hwmon_get_power1_input(struct device *d, struct device_attribute *a, static SENSOR_DEVICE_ATTR(power1_input, S_IRUGO, nouveau_hwmon_get_power1_input, NULL, 0); +static ssize_t +nouveau_hwmon_get_power1_max(struct device *d, struct device_attribute *a, + char *buf) +{ + struct drm_device *dev = dev_get_drvdata(d); + struct nouveau_drm *drm = nouveau_drm(dev); + struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->client.device); + return sprintf(buf, "%i\n", iccsense->power_w_max); +} + +static SENSOR_DEVICE_ATTR(power1_max, S_IRUGO, + nouveau_hwmon_get_power1_max, NULL, 0); + +static ssize_t +nouveau_hwmon_get_power1_crit(struct device *d, struct device_attribute *a, + char *buf) +{ + struct drm_device *dev = dev_get_drvdata(d); + struct nouveau_drm *drm = nouveau_drm(dev); + struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->client.device); + return sprintf(buf, "%i\n", iccsense->power_w_crit); +} + +static SENSOR_DEVICE_ATTR(power1_crit, S_IRUGO, + nouveau_hwmon_get_power1_crit, NULL, 0); + static struct attribute *hwmon_default_attributes[] = { &sensor_dev_attr_name.dev_attr.attr, &sensor_dev_attr_update_rate.dev_attr.attr, @@ -639,6 +665,12 @@ static struct attribute *hwmon_power_attributes[] = { NULL }; +static struct attribute *hwmon_power_caps_attributes[] = { + &sensor_dev_attr_power1_max.dev_attr.attr, + &sensor_dev_attr_power1_crit.dev_attr.attr, + NULL +}; + static const struct attribute_group hwmon_default_attrgroup = { .attrs = hwmon_default_attributes, }; @@ -657,6 +689,9 @@ static const struct attribute_group hwmon_in0_attrgroup = { static const struct attribute_group hwmon_power_attrgroup = { .attrs = hwmon_power_attributes, }; +static const struct attribute_group hwmon_power_caps_attrgroup = { + .attrs = hwmon_power_caps_attributes, +}; #endif int @@ -664,9 +699,9 @@ nouveau_hwmon_init(struct drm_device *dev) { #if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE)) struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_therm *therm = nvxx_therm(&drm->device); - struct nvkm_volt *volt = nvxx_volt(&drm->device); - struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->device); + struct nvkm_therm *therm = nvxx_therm(&drm->client.device); + struct nvkm_volt *volt = nvxx_volt(&drm->client.device); + struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->client.device); struct nouveau_hwmon *hwmon; struct device *hwmon_dev; int ret = 0; @@ -728,8 +763,16 @@ nouveau_hwmon_init(struct drm_device *dev) if (iccsense && iccsense->data_valid && !list_empty(&iccsense->rails)) { ret = sysfs_create_group(&hwmon_dev->kobj, &hwmon_power_attrgroup); + if (ret) goto error; + + if (iccsense->power_w_max && iccsense->power_w_crit) { + ret = sysfs_create_group(&hwmon_dev->kobj, + &hwmon_power_caps_attrgroup); + if (ret) + goto error; + } } hwmon->hwmon = hwmon_dev; @@ -759,6 +802,7 @@ nouveau_hwmon_fini(struct drm_device *dev) sysfs_remove_group(&hwmon->hwmon->kobj, &hwmon_fan_rpm_attrgroup); sysfs_remove_group(&hwmon->hwmon->kobj, &hwmon_in0_attrgroup); sysfs_remove_group(&hwmon->hwmon->kobj, &hwmon_power_attrgroup); + sysfs_remove_group(&hwmon->hwmon->kobj, &hwmon_power_caps_attrgroup); hwmon_device_unregister(hwmon->hwmon); } diff --git a/drivers/gpu/drm/nouveau/nouveau_led.c b/drivers/gpu/drm/nouveau/nouveau_led.c index 3e2f1b6cd4df..2c5e0628da12 100644 --- a/drivers/gpu/drm/nouveau/nouveau_led.c +++ b/drivers/gpu/drm/nouveau/nouveau_led.c @@ -38,7 +38,7 @@ nouveau_led_get_brightness(struct led_classdev *led) { struct drm_device *drm_dev = container_of(led, struct nouveau_led, led)->dev; struct nouveau_drm *drm = nouveau_drm(drm_dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; u32 div, duty; div = nvif_rd32(device, 0x61c880) & 0x00ffffff; @@ -55,7 +55,7 @@ nouveau_led_set_brightness(struct led_classdev *led, enum led_brightness value) { struct drm_device *drm_dev = container_of(led, struct nouveau_led, led)->dev; struct nouveau_drm *drm = nouveau_drm(drm_dev); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; u32 input_clk = 27e6; /* PDISPLAY.SOR[1].PWM is connected to the crystal */ u32 freq = 100; /* this is what nvidia uses and it should be good-enough */ @@ -78,7 +78,7 @@ int nouveau_led_init(struct drm_device *dev) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvkm_gpio *gpio = nvxx_gpio(&drm->device); + struct nvkm_gpio *gpio = nvxx_gpio(&drm->client.device); struct dcb_gpio_func logo_led; int ret; @@ -102,6 +102,7 @@ nouveau_led_init(struct drm_device *dev) ret = led_classdev_register(dev->dev, &drm->led->led); if (ret) { kfree(drm->led); + drm->led = NULL; return ret; } diff --git a/drivers/gpu/drm/nouveau/nouveau_led.h b/drivers/gpu/drm/nouveau/nouveau_led.h index 187ecdb82002..21a5775028cc 100644 --- a/drivers/gpu/drm/nouveau/nouveau_led.h +++ b/drivers/gpu/drm/nouveau/nouveau_led.h @@ -42,7 +42,7 @@ nouveau_led(struct drm_device *dev) } /* nouveau_led.c */ -#if IS_ENABLED(CONFIG_LEDS_CLASS) +#if IS_REACHABLE(CONFIG_LEDS_CLASS) int nouveau_led_init(struct drm_device *dev); void nouveau_led_suspend(struct drm_device *dev); void nouveau_led_resume(struct drm_device *dev); diff --git a/drivers/gpu/drm/nouveau/nouveau_nvif.c b/drivers/gpu/drm/nouveau/nouveau_nvif.c index 15f0925ea13b..b3f29b1ce9ea 100644 --- a/drivers/gpu/drm/nouveau/nouveau_nvif.c +++ b/drivers/gpu/drm/nouveau/nouveau_nvif.c @@ -60,20 +60,15 @@ nvkm_client_ioctl(void *priv, bool super, void *data, u32 size, void **hack) static int nvkm_client_resume(void *priv) { - return nvkm_client_init(priv); + struct nvkm_client *client = priv; + return nvkm_object_init(&client->object); } static int nvkm_client_suspend(void *priv) -{ - return nvkm_client_fini(priv, true); -} - -static void -nvkm_client_driver_fini(void *priv) { struct nvkm_client *client = priv; - nvkm_client_del(&client); + return nvkm_object_fini(&client->object, true); } static int @@ -108,23 +103,14 @@ static int nvkm_client_driver_init(const char *name, u64 device, const char *cfg, const char *dbg, void **ppriv) { - struct nvkm_client *client; - int ret; - - ret = nvkm_client_new(name, device, cfg, dbg, &client); - *ppriv = client; - if (ret) - return ret; - - client->ntfy = nvkm_client_ntfy; - return 0; + return nvkm_client_new(name, device, cfg, dbg, nvkm_client_ntfy, + (struct nvkm_client **)ppriv); } const struct nvif_driver nvif_driver_nvkm = { .name = "nvkm", .init = nvkm_client_driver_init, - .fini = nvkm_client_driver_fini, .suspend = nvkm_client_suspend, .resume = nvkm_client_resume, .ioctl = nvkm_client_ioctl, diff --git a/drivers/gpu/drm/nouveau/nouveau_prime.c b/drivers/gpu/drm/nouveau/nouveau_prime.c index a0a9704cfe2b..1fefc93af1d7 100644 --- a/drivers/gpu/drm/nouveau/nouveau_prime.c +++ b/drivers/gpu/drm/nouveau/nouveau_prime.c @@ -60,6 +60,7 @@ struct drm_gem_object *nouveau_gem_prime_import_sg_table(struct drm_device *dev, struct dma_buf_attachment *attach, struct sg_table *sg) { + struct nouveau_drm *drm = nouveau_drm(dev); struct nouveau_bo *nvbo; struct reservation_object *robj = attach->dmabuf->resv; u32 flags = 0; @@ -68,7 +69,7 @@ struct drm_gem_object *nouveau_gem_prime_import_sg_table(struct drm_device *dev, flags = TTM_PL_FLAG_TT; ww_mutex_lock(&robj->lock, NULL); - ret = nouveau_bo_new(dev, attach->dmabuf->size, 0, flags, 0, 0, + ret = nouveau_bo_new(&drm->client, attach->dmabuf->size, 0, flags, 0, 0, sg, robj, &nvbo); ww_mutex_unlock(&robj->lock); if (ret) diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c index db35ab5883ac..b7ab268f7d6f 100644 --- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c +++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c @@ -24,10 +24,10 @@ nouveau_sgdma_destroy(struct ttm_tt *ttm) } static int -nv04_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem) +nv04_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *reg) { struct nouveau_sgdma_be *nvbe = (struct nouveau_sgdma_be *)ttm; - struct nvkm_mem *node = mem->mm_node; + struct nvkm_mem *node = reg->mm_node; if (ttm->sg) { node->sg = ttm->sg; @@ -36,7 +36,7 @@ nv04_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem) node->sg = NULL; node->pages = nvbe->ttm.dma_address; } - node->size = (mem->num_pages << PAGE_SHIFT) >> 12; + node->size = (reg->num_pages << PAGE_SHIFT) >> 12; nvkm_vm_map(&node->vma[0], node); nvbe->node = node; @@ -58,10 +58,10 @@ static struct ttm_backend_func nv04_sgdma_backend = { }; static int -nv50_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem) +nv50_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *reg) { struct nouveau_sgdma_be *nvbe = (struct nouveau_sgdma_be *)ttm; - struct nvkm_mem *node = mem->mm_node; + struct nvkm_mem *node = reg->mm_node; /* noop: bound in move_notify() */ if (ttm->sg) { @@ -71,7 +71,7 @@ nv50_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem) node->sg = NULL; node->pages = nvbe->ttm.dma_address; } - node->size = (mem->num_pages << PAGE_SHIFT) >> 12; + node->size = (reg->num_pages << PAGE_SHIFT) >> 12; return 0; } @@ -100,7 +100,7 @@ nouveau_sgdma_create_ttm(struct ttm_bo_device *bdev, if (!nvbe) return NULL; - if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) + if (drm->client.device.info.family < NV_DEVICE_INFO_V0_TESLA) nvbe->ttm.ttm.func = &nv04_sgdma_backend; else nvbe->ttm.ttm.func = &nv50_sgdma_backend; diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index ec4668a41e01..13e5cc5f07fe 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -36,7 +36,7 @@ static int nouveau_vram_manager_init(struct ttm_mem_type_manager *man, unsigned long psize) { struct nouveau_drm *drm = nouveau_bdev(man->bdev); - struct nvkm_fb *fb = nvxx_fb(&drm->device); + struct nvkm_fb *fb = nvxx_fb(&drm->client.device); man->priv = fb; return 0; } @@ -64,45 +64,45 @@ nvkm_mem_node_cleanup(struct nvkm_mem *node) static void nouveau_vram_manager_del(struct ttm_mem_type_manager *man, - struct ttm_mem_reg *mem) + struct ttm_mem_reg *reg) { struct nouveau_drm *drm = nouveau_bdev(man->bdev); - struct nvkm_ram *ram = nvxx_fb(&drm->device)->ram; - nvkm_mem_node_cleanup(mem->mm_node); - ram->func->put(ram, (struct nvkm_mem **)&mem->mm_node); + struct nvkm_ram *ram = nvxx_fb(&drm->client.device)->ram; + nvkm_mem_node_cleanup(reg->mm_node); + ram->func->put(ram, (struct nvkm_mem **)®->mm_node); } static int nouveau_vram_manager_new(struct ttm_mem_type_manager *man, struct ttm_buffer_object *bo, const struct ttm_place *place, - struct ttm_mem_reg *mem) + struct ttm_mem_reg *reg) { struct nouveau_drm *drm = nouveau_bdev(man->bdev); - struct nvkm_ram *ram = nvxx_fb(&drm->device)->ram; + struct nvkm_ram *ram = nvxx_fb(&drm->client.device)->ram; struct nouveau_bo *nvbo = nouveau_bo(bo); struct nvkm_mem *node; u32 size_nc = 0; int ret; - if (drm->device.info.ram_size == 0) + if (drm->client.device.info.ram_size == 0) return -ENOMEM; if (nvbo->tile_flags & NOUVEAU_GEM_TILE_NONCONTIG) size_nc = 1 << nvbo->page_shift; - ret = ram->func->get(ram, mem->num_pages << PAGE_SHIFT, - mem->page_alignment << PAGE_SHIFT, size_nc, + ret = ram->func->get(ram, reg->num_pages << PAGE_SHIFT, + reg->page_alignment << PAGE_SHIFT, size_nc, (nvbo->tile_flags >> 8) & 0x3ff, &node); if (ret) { - mem->mm_node = NULL; + reg->mm_node = NULL; return (ret == -ENOSPC) ? 0 : ret; } node->page_shift = nvbo->page_shift; - mem->mm_node = node; - mem->start = node->offset >> PAGE_SHIFT; + reg->mm_node = node; + reg->start = node->offset >> PAGE_SHIFT; return 0; } @@ -127,18 +127,18 @@ nouveau_gart_manager_fini(struct ttm_mem_type_manager *man) static void nouveau_gart_manager_del(struct ttm_mem_type_manager *man, - struct ttm_mem_reg *mem) + struct ttm_mem_reg *reg) { - nvkm_mem_node_cleanup(mem->mm_node); - kfree(mem->mm_node); - mem->mm_node = NULL; + nvkm_mem_node_cleanup(reg->mm_node); + kfree(reg->mm_node); + reg->mm_node = NULL; } static int nouveau_gart_manager_new(struct ttm_mem_type_manager *man, struct ttm_buffer_object *bo, const struct ttm_place *place, - struct ttm_mem_reg *mem) + struct ttm_mem_reg *reg) { struct nouveau_drm *drm = nouveau_bdev(bo->bdev); struct nouveau_bo *nvbo = nouveau_bo(bo); @@ -150,7 +150,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man, node->page_shift = 12; - switch (drm->device.info.family) { + switch (drm->client.device.info.family) { case NV_DEVICE_INFO_V0_TNT: case NV_DEVICE_INFO_V0_CELSIUS: case NV_DEVICE_INFO_V0_KELVIN: @@ -158,7 +158,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man, case NV_DEVICE_INFO_V0_CURIE: break; case NV_DEVICE_INFO_V0_TESLA: - if (drm->device.info.chipset != 0x50) + if (drm->client.device.info.chipset != 0x50) node->memtype = (nvbo->tile_flags & 0x7f00) >> 8; break; case NV_DEVICE_INFO_V0_FERMI: @@ -169,12 +169,12 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man, break; default: NV_WARN(drm, "%s: unhandled family type %x\n", __func__, - drm->device.info.family); + drm->client.device.info.family); break; } - mem->mm_node = node; - mem->start = 0; + reg->mm_node = node; + reg->start = 0; return 0; } @@ -197,7 +197,7 @@ static int nv04_gart_manager_init(struct ttm_mem_type_manager *man, unsigned long psize) { struct nouveau_drm *drm = nouveau_bdev(man->bdev); - struct nvkm_mmu *mmu = nvxx_mmu(&drm->device); + struct nvkm_mmu *mmu = nvxx_mmu(&drm->client.device); struct nv04_mmu *priv = (void *)mmu; struct nvkm_vm *vm = NULL; nvkm_vm_ref(priv->vm, &vm, NULL); @@ -215,20 +215,20 @@ nv04_gart_manager_fini(struct ttm_mem_type_manager *man) } static void -nv04_gart_manager_del(struct ttm_mem_type_manager *man, struct ttm_mem_reg *mem) +nv04_gart_manager_del(struct ttm_mem_type_manager *man, struct ttm_mem_reg *reg) { - struct nvkm_mem *node = mem->mm_node; + struct nvkm_mem *node = reg->mm_node; if (node->vma[0].node) nvkm_vm_put(&node->vma[0]); - kfree(mem->mm_node); - mem->mm_node = NULL; + kfree(reg->mm_node); + reg->mm_node = NULL; } static int nv04_gart_manager_new(struct ttm_mem_type_manager *man, struct ttm_buffer_object *bo, const struct ttm_place *place, - struct ttm_mem_reg *mem) + struct ttm_mem_reg *reg) { struct nvkm_mem *node; int ret; @@ -239,15 +239,15 @@ nv04_gart_manager_new(struct ttm_mem_type_manager *man, node->page_shift = 12; - ret = nvkm_vm_get(man->priv, mem->num_pages << 12, node->page_shift, + ret = nvkm_vm_get(man->priv, reg->num_pages << 12, node->page_shift, NV_MEM_ACCESS_RW, &node->vma[0]); if (ret) { kfree(node); return ret; } - mem->mm_node = node; - mem->start = node->vma[0].offset >> PAGE_SHIFT; + reg->mm_node = node; + reg->start = node->vma[0].offset >> PAGE_SHIFT; return 0; } @@ -339,7 +339,7 @@ nouveau_ttm_global_release(struct nouveau_drm *drm) int nouveau_ttm_init(struct nouveau_drm *drm) { - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); struct nvkm_pci *pci = device->pci; struct drm_device *dev = drm->dev; u8 bits; @@ -352,8 +352,8 @@ nouveau_ttm_init(struct nouveau_drm *drm) drm->agp.cma = pci->agp.cma; } - bits = nvxx_mmu(&drm->device)->dma_bits; - if (nvxx_device(&drm->device)->func->pci) { + bits = nvxx_mmu(&drm->client.device)->dma_bits; + if (nvxx_device(&drm->client.device)->func->pci) { if (drm->agp.bridge) bits = 32; } else if (device->func->tegra) { @@ -396,7 +396,7 @@ nouveau_ttm_init(struct nouveau_drm *drm) } /* VRAM init */ - drm->gem.vram_available = drm->device.info.ram_user; + drm->gem.vram_available = drm->client.device.info.ram_user; arch_io_reserve_memtype_wc(device->func->resource_addr(device, 1), device->func->resource_size(device, 1)); @@ -413,7 +413,7 @@ nouveau_ttm_init(struct nouveau_drm *drm) /* GART init */ if (!drm->agp.bridge) { - drm->gem.gart_available = nvxx_mmu(&drm->device)->limit; + drm->gem.gart_available = nvxx_mmu(&drm->client.device)->limit; } else { drm->gem.gart_available = drm->agp.size; } @@ -433,7 +433,7 @@ nouveau_ttm_init(struct nouveau_drm *drm) void nouveau_ttm_fini(struct nouveau_drm *drm) { - struct nvkm_device *device = nvxx_device(&drm->device); + struct nvkm_device *device = nvxx_device(&drm->client.device); ttm_bo_clean_mm(&drm->ttm.bdev, TTM_PL_VRAM); ttm_bo_clean_mm(&drm->ttm.bdev, TTM_PL_TT); diff --git a/drivers/gpu/drm/nouveau/nouveau_usif.c b/drivers/gpu/drm/nouveau/nouveau_usif.c index 08f9c6fa0f7f..afbdbed1a690 100644 --- a/drivers/gpu/drm/nouveau/nouveau_usif.c +++ b/drivers/gpu/drm/nouveau/nouveau_usif.c @@ -103,7 +103,7 @@ usif_notify(const void *header, u32 length, const void *data, u32 size) } break; default: - BUG_ON(1); + BUG(); break; } @@ -313,7 +313,8 @@ usif_ioctl(struct drm_file *filp, void __user *user, u32 argc) if (!(ret = nvif_unpack(-ENOSYS, &data, &size, argv->v0, 0, 0, true))) { /* block access to objects not created via this interface */ owner = argv->v0.owner; - if (argv->v0.object == 0ULL) + if (argv->v0.object == 0ULL && + argv->v0.type != NVIF_IOCTL_V0_DEL) argv->v0.owner = NVDRM_OBJECT_ANY; /* except client */ else argv->v0.owner = NVDRM_OBJECT_USIF; diff --git a/drivers/gpu/drm/nouveau/nouveau_vga.c b/drivers/gpu/drm/nouveau/nouveau_vga.c index c6a180a0c284..eef22c6b9665 100644 --- a/drivers/gpu/drm/nouveau/nouveau_vga.c +++ b/drivers/gpu/drm/nouveau/nouveau_vga.c @@ -13,13 +13,13 @@ static unsigned int nouveau_vga_set_decode(void *priv, bool state) { struct nouveau_drm *drm = nouveau_drm(priv); - struct nvif_object *device = &drm->device.object; + struct nvif_object *device = &drm->client.device.object; - if (drm->device.info.family == NV_DEVICE_INFO_V0_CURIE && - drm->device.info.chipset >= 0x4c) + if (drm->client.device.info.family == NV_DEVICE_INFO_V0_CURIE && + drm->client.device.info.chipset >= 0x4c) nvif_wr32(device, 0x088060, state); else - if (drm->device.info.chipset >= 0x40) + if (drm->client.device.info.chipset >= 0x40) nvif_wr32(device, 0x088054, state); else nvif_wr32(device, 0x001854, state); diff --git a/drivers/gpu/drm/nouveau/nv04_fbcon.c b/drivers/gpu/drm/nouveau/nv04_fbcon.c index 6a2b187e3c3b..01731dbeb3d8 100644 --- a/drivers/gpu/drm/nouveau/nv04_fbcon.c +++ b/drivers/gpu/drm/nouveau/nv04_fbcon.c @@ -136,7 +136,7 @@ nv04_fbcon_accel_init(struct fb_info *info) struct drm_device *dev = nfbdev->helper.dev; struct nouveau_drm *drm = nouveau_drm(dev); struct nouveau_channel *chan = drm->channel; - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; int surface_fmt, pattern_fmt, rect_fmt; int ret; diff --git a/drivers/gpu/drm/nouveau/nv17_fence.c b/drivers/gpu/drm/nouveau/nv17_fence.c index 79bc01111351..6477b7069e14 100644 --- a/drivers/gpu/drm/nouveau/nv17_fence.c +++ b/drivers/gpu/drm/nouveau/nv17_fence.c @@ -76,9 +76,9 @@ nv17_fence_context_new(struct nouveau_channel *chan) { struct nv10_fence_priv *priv = chan->drm->fence; struct nv10_fence_chan *fctx; - struct ttm_mem_reg *mem = &priv->bo->bo.mem; - u32 start = mem->start * PAGE_SIZE; - u32 limit = start + mem->size - 1; + struct ttm_mem_reg *reg = &priv->bo->bo.mem; + u32 start = reg->start * PAGE_SIZE; + u32 limit = start + reg->size - 1; int ret = 0; fctx = chan->fence = kzalloc(sizeof(*fctx), GFP_KERNEL); @@ -129,7 +129,7 @@ nv17_fence_create(struct nouveau_drm *drm) priv->base.context_base = dma_fence_context_alloc(priv->base.contexts); spin_lock_init(&priv->lock); - ret = nouveau_bo_new(drm->dev, 4096, 0x1000, TTM_PL_FLAG_VRAM, + ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM, 0, 0x0000, NULL, NULL, &priv->bo); if (!ret) { ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM, false); diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c index 452da483ca01..0b4440ffbeae 100644 --- a/drivers/gpu/drm/nouveau/nv50_display.c +++ b/drivers/gpu/drm/nouveau/nv50_display.c @@ -447,18 +447,18 @@ nv50_dmac_ctxdma_new(struct nv50_dmac *dmac, struct nouveau_framebuffer *fb) args.base.target = NV_DMA_V0_TARGET_VRAM; args.base.access = NV_DMA_V0_ACCESS_RDWR; args.base.start = 0; - args.base.limit = drm->device.info.ram_user - 1; + args.base.limit = drm->client.device.info.ram_user - 1; - if (drm->device.info.chipset < 0x80) { + if (drm->client.device.info.chipset < 0x80) { args.nv50.part = NV50_DMA_V0_PART_256; argc += sizeof(args.nv50); } else - if (drm->device.info.chipset < 0xc0) { + if (drm->client.device.info.chipset < 0xc0) { args.nv50.part = NV50_DMA_V0_PART_256; args.nv50.kind = kind; argc += sizeof(args.nv50); } else - if (drm->device.info.chipset < 0xd0) { + if (drm->client.device.info.chipset < 0xd0) { args.gf100.kind = kind; argc += sizeof(args.gf100); } else { @@ -848,7 +848,7 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, asyw->image.kind = (fb->nvbo->tile_flags & 0x0000ff00) >> 8; if (asyw->image.kind) { asyw->image.layout = 0; - if (drm->device.info.chipset >= 0xc0) + if (drm->client.device.info.chipset >= 0xc0) asyw->image.block = fb->nvbo->tile_mode >> 4; else asyw->image.block = fb->nvbo->tile_mode; @@ -1397,7 +1397,7 @@ nv50_base_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw) { struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev); struct nv50_disp *disp = nv50_disp(wndw->plane.dev); - if (nvif_msec(&drm->device, 2000ULL, + if (nvif_msec(&drm->client.device, 2000ULL, u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4); if ((data & 0xc0000000) == 0x40000000) break; @@ -1522,7 +1522,7 @@ nv50_base_new(struct nouveau_drm *drm, struct nv50_head *head, return ret; } - ret = nv50_base_create(&drm->device, disp->disp, base->id, + ret = nv50_base_create(&drm->client.device, disp->disp, base->id, disp->sync->bo.offset, &base->chan); if (ret) return ret; @@ -2394,7 +2394,7 @@ static int nv50_head_create(struct drm_device *dev, int index) { struct nouveau_drm *drm = nouveau_drm(dev); - struct nvif_device *device = &drm->device; + struct nvif_device *device = &drm->client.device; struct nv50_disp *disp = nv50_disp(dev); struct nv50_head *head; struct nv50_base *base; @@ -2428,7 +2428,7 @@ nv50_head_create(struct drm_device *dev, int index) drm_crtc_helper_add(crtc, &nv50_head_help); drm_mode_crtc_set_gamma_size(crtc, 256); - ret = nouveau_bo_new(dev, 8192, 0x100, TTM_PL_FLAG_VRAM, + ret = nouveau_bo_new(&drm->client, 8192, 0x100, TTM_PL_FLAG_VRAM, 0, 0x0000, NULL, NULL, &head->base.lut.nvbo); if (!ret) { ret = nouveau_bo_pin(head->base.lut.nvbo, TTM_PL_FLAG_VRAM, true); @@ -2667,7 +2667,7 @@ static int nv50_dac_create(struct drm_connector *connector, struct dcb_output *dcbe) { struct nouveau_drm *drm = nouveau_drm(connector->dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct nvkm_i2c_bus *bus; struct nouveau_encoder *nv_encoder; struct drm_encoder *encoder; @@ -3623,7 +3623,7 @@ nv50_sor_enable(struct drm_encoder *encoder) nv50_audio_enable(encoder, mode); break; default: - BUG_ON(1); + BUG(); break; } @@ -3657,7 +3657,7 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe) { struct nouveau_connector *nv_connector = nouveau_connector(connector); struct nouveau_drm *drm = nouveau_drm(connector->dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct nouveau_encoder *nv_encoder; struct drm_encoder *encoder; int type, ret; @@ -3796,7 +3796,7 @@ nv50_pior_enable(struct drm_encoder *encoder) proto = 0x0; break; default: - BUG_ON(1); + BUG(); break; } @@ -3842,7 +3842,7 @@ static int nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe) { struct nouveau_drm *drm = nouveau_drm(connector->dev); - struct nvkm_i2c *i2c = nvxx_i2c(&drm->device); + struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device); struct nvkm_i2c_bus *bus = NULL; struct nvkm_i2c_aux *aux = NULL; struct i2c_adapter *ddc; @@ -3915,7 +3915,7 @@ nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock) evo_data(push, 0x00000000); nouveau_bo_wr32(disp->sync, 0, 0x00000000); evo_kick(push, core); - if (nvif_msec(&drm->device, 2000ULL, + if (nvif_msec(&drm->client.device, 2000ULL, if (nouveau_bo_rd32(disp->sync, 0)) break; usleep_range(1, 2); @@ -4050,6 +4050,11 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state) } } + for_each_crtc_in_state(state, crtc, crtc_state, i) { + if (crtc->state->event) + drm_crtc_vblank_get(crtc); + } + /* Update plane(s). */ for_each_plane_in_state(state, plane, plane_state, i) { struct nv50_wndw_atom *asyw = nv50_wndw_atom(plane->state); @@ -4099,6 +4104,7 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state) drm_crtc_send_vblank_event(crtc, crtc->state->event); spin_unlock_irqrestore(&crtc->dev->event_lock, flags); crtc->state->event = NULL; + drm_crtc_vblank_put(crtc); } } @@ -4427,7 +4433,7 @@ module_param_named(atomic, nouveau_atomic, int, 0400); int nv50_display_create(struct drm_device *dev) { - struct nvif_device *device = &nouveau_drm(dev)->device; + struct nvif_device *device = &nouveau_drm(dev)->client.device; struct nouveau_drm *drm = nouveau_drm(dev); struct dcb_table *dcb = &drm->vbios.dcb; struct drm_connector *connector, *tmp; @@ -4451,7 +4457,7 @@ nv50_display_create(struct drm_device *dev) dev->driver->driver_features |= DRIVER_ATOMIC; /* small shared memory area we use for notifiers and semaphores */ - ret = nouveau_bo_new(dev, 4096, 0x1000, TTM_PL_FLAG_VRAM, + ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM, 0, 0x0000, NULL, NULL, &disp->sync); if (!ret) { ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM, true); diff --git a/drivers/gpu/drm/nouveau/nv50_fence.c b/drivers/gpu/drm/nouveau/nv50_fence.c index f68c7054fd53..a369d978e267 100644 --- a/drivers/gpu/drm/nouveau/nv50_fence.c +++ b/drivers/gpu/drm/nouveau/nv50_fence.c @@ -37,9 +37,9 @@ nv50_fence_context_new(struct nouveau_channel *chan) { struct nv10_fence_priv *priv = chan->drm->fence; struct nv10_fence_chan *fctx; - struct ttm_mem_reg *mem = &priv->bo->bo.mem; - u32 start = mem->start * PAGE_SIZE; - u32 limit = start + mem->size - 1; + struct ttm_mem_reg *reg = &priv->bo->bo.mem; + u32 start = reg->start * PAGE_SIZE; + u32 limit = start + reg->size - 1; int ret; fctx = chan->fence = kzalloc(sizeof(*fctx), GFP_KERNEL); @@ -82,7 +82,7 @@ nv50_fence_create(struct nouveau_drm *drm) priv->base.context_base = dma_fence_context_alloc(priv->base.contexts); spin_lock_init(&priv->lock); - ret = nouveau_bo_new(drm->dev, 4096, 0x1000, TTM_PL_FLAG_VRAM, + ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM, 0, 0x0000, NULL, NULL, &priv->bo); if (!ret) { ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM, false); diff --git a/drivers/gpu/drm/nouveau/nv84_fence.c b/drivers/gpu/drm/nouveau/nv84_fence.c index 52b87ae83e7b..bd7a8a1e4ad9 100644 --- a/drivers/gpu/drm/nouveau/nv84_fence.c +++ b/drivers/gpu/drm/nouveau/nv84_fence.c @@ -107,8 +107,10 @@ nv84_fence_context_del(struct nouveau_channel *chan) struct nv84_fence_chan *fctx = chan->fence; nouveau_bo_wr32(priv->bo, chan->chid * 16 / 4, fctx->base.sequence); + mutex_lock(&priv->mutex); nouveau_bo_vma_del(priv->bo, &fctx->vma_gart); nouveau_bo_vma_del(priv->bo, &fctx->vma); + mutex_unlock(&priv->mutex); nouveau_fence_context_del(&fctx->base); chan->fence = NULL; nouveau_fence_context_free(&fctx->base); @@ -134,11 +136,13 @@ nv84_fence_context_new(struct nouveau_channel *chan) fctx->base.sync32 = nv84_fence_sync32; fctx->base.sequence = nv84_fence_read(chan); + mutex_lock(&priv->mutex); ret = nouveau_bo_vma_add(priv->bo, cli->vm, &fctx->vma); if (ret == 0) { ret = nouveau_bo_vma_add(priv->bo_gart, cli->vm, &fctx->vma_gart); } + mutex_unlock(&priv->mutex); if (ret) nv84_fence_context_del(chan); @@ -193,7 +197,7 @@ nv84_fence_destroy(struct nouveau_drm *drm) int nv84_fence_create(struct nouveau_drm *drm) { - struct nvkm_fifo *fifo = nvxx_fifo(&drm->device); + struct nvkm_fifo *fifo = nvxx_fifo(&drm->client.device); struct nv84_fence_priv *priv; u32 domain; int ret; @@ -212,15 +216,17 @@ nv84_fence_create(struct nouveau_drm *drm) priv->base.context_base = dma_fence_context_alloc(priv->base.contexts); priv->base.uevent = true; + mutex_init(&priv->mutex); + /* Use VRAM if there is any ; otherwise fallback to system memory */ - domain = drm->device.info.ram_size != 0 ? TTM_PL_FLAG_VRAM : + domain = drm->client.device.info.ram_size != 0 ? TTM_PL_FLAG_VRAM : /* * fences created in sysmem must be non-cached or we * will lose CPU/GPU coherency! */ TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED; - ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0, domain, 0, - 0, NULL, NULL, &priv->bo); + ret = nouveau_bo_new(&drm->client, 16 * priv->base.contexts, 0, + domain, 0, 0, NULL, NULL, &priv->bo); if (ret == 0) { ret = nouveau_bo_pin(priv->bo, domain, false); if (ret == 0) { @@ -233,7 +239,7 @@ nv84_fence_create(struct nouveau_drm *drm) } if (ret == 0) - ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0, + ret = nouveau_bo_new(&drm->client, 16 * priv->base.contexts, 0, TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED, 0, 0, NULL, NULL, &priv->bo_gart); if (ret == 0) { diff --git a/drivers/gpu/drm/nouveau/nvif/Kbuild b/drivers/gpu/drm/nouveau/nvif/Kbuild index ff8ed3a04d06..067b5e9f5ec1 100644 --- a/drivers/gpu/drm/nouveau/nvif/Kbuild +++ b/drivers/gpu/drm/nouveau/nvif/Kbuild @@ -1,4 +1,5 @@ nvif-y := nvif/object.o nvif-y += nvif/client.o nvif-y += nvif/device.o +nvif-y += nvif/driver.o nvif-y += nvif/notify.o diff --git a/drivers/gpu/drm/nouveau/nvif/client.c b/drivers/gpu/drm/nouveau/nvif/client.c index 29c20dfd894d..12db54965c20 100644 --- a/drivers/gpu/drm/nouveau/nvif/client.c +++ b/drivers/gpu/drm/nouveau/nvif/client.c @@ -26,6 +26,9 @@ #include #include +#include +#include + int nvif_client_ioctl(struct nvif_client *client, void *data, u32 size) { @@ -47,37 +50,29 @@ nvif_client_resume(struct nvif_client *client) void nvif_client_fini(struct nvif_client *client) { + nvif_object_fini(&client->object); if (client->driver) { - client->driver->fini(client->object.priv); + if (client->driver->fini) + client->driver->fini(client->object.priv); client->driver = NULL; - client->object.client = NULL; - nvif_object_fini(&client->object); } } -static const struct nvif_driver * -nvif_drivers[] = { -#ifdef __KERNEL__ - &nvif_driver_nvkm, -#else - &nvif_driver_drm, - &nvif_driver_lib, - &nvif_driver_null, -#endif - NULL -}; - int -nvif_client_init(const char *driver, const char *name, u64 device, - const char *cfg, const char *dbg, struct nvif_client *client) +nvif_client_init(struct nvif_client *parent, const char *name, u64 device, + struct nvif_client *client) { + struct nvif_client_v0 args = { .device = device }; struct { struct nvif_ioctl_v0 ioctl; struct nvif_ioctl_nop_v0 nop; - } args = {}; - int ret, i; + } nop = {}; + int ret; - ret = nvif_object_init(NULL, 0, 0, NULL, 0, &client->object); + strncpy(args.name, name, sizeof(args.name)); + ret = nvif_object_init(parent != client ? &parent->object : NULL, + 0, NVIF_CLASS_CLIENT, &args, sizeof(args), + &client->object); if (ret) return ret; @@ -85,19 +80,11 @@ nvif_client_init(const char *driver, const char *name, u64 device, client->object.handle = ~0; client->route = NVIF_IOCTL_V0_ROUTE_NVIF; client->super = true; - - for (i = 0, ret = -EINVAL; (client->driver = nvif_drivers[i]); i++) { - if (!driver || !strcmp(client->driver->name, driver)) { - ret = client->driver->init(name, device, cfg, dbg, - &client->object.priv); - if (!ret || driver) - break; - } - } + client->driver = parent->driver; if (ret == 0) { - ret = nvif_client_ioctl(client, &args, sizeof(args)); - client->version = args.nop.version; + ret = nvif_client_ioctl(client, &nop, sizeof(nop)); + client->version = nop.nop.version; } if (ret) diff --git a/drivers/gpu/drm/nouveau/nvif/driver.c b/drivers/gpu/drm/nouveau/nvif/driver.c new file mode 100644 index 000000000000..701330956e33 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvif/driver.c @@ -0,0 +1,58 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include +#include + +static const struct nvif_driver * +nvif_driver[] = { +#ifdef __KERNEL__ + &nvif_driver_nvkm, +#else + &nvif_driver_drm, + &nvif_driver_lib, + &nvif_driver_null, +#endif + NULL +}; + +int +nvif_driver_init(const char *drv, const char *cfg, const char *dbg, + const char *name, u64 device, struct nvif_client *client) +{ + int ret = -EINVAL, i; + + for (i = 0; (client->driver = nvif_driver[i]); i++) { + if (!drv || !strcmp(client->driver->name, drv)) { + ret = client->driver->init(name, device, cfg, dbg, + &client->object.priv); + if (ret == 0) + break; + client->driver->fini(client->object.priv); + } + } + + if (ret == 0) + ret = nvif_client_init(client, name, device, client); + return ret; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/Kbuild b/drivers/gpu/drm/nouveau/nvkm/Kbuild index 2832147b676c..e664378f6eda 100644 --- a/drivers/gpu/drm/nouveau/nvkm/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/Kbuild @@ -1,3 +1,4 @@ include $(src)/nvkm/core/Kbuild +include $(src)/nvkm/falcon/Kbuild include $(src)/nvkm/subdev/Kbuild include $(src)/nvkm/engine/Kbuild diff --git a/drivers/gpu/drm/nouveau/nvkm/core/client.c b/drivers/gpu/drm/nouveau/nvkm/core/client.c index e1943910858e..0d3a896892b4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/client.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/client.c @@ -31,6 +31,43 @@ #include #include +static int +nvkm_uclient_new(const struct nvkm_oclass *oclass, void *argv, u32 argc, + struct nvkm_object **pobject) +{ + union { + struct nvif_client_v0 v0; + } *args = argv; + struct nvkm_client *client; + int ret = -ENOSYS; + + if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))){ + args->v0.name[sizeof(args->v0.name) - 1] = 0; + ret = nvkm_client_new(args->v0.name, args->v0.device, NULL, + NULL, oclass->client->ntfy, &client); + if (ret) + return ret; + } else + return ret; + + client->object.client = oclass->client; + client->object.handle = oclass->handle; + client->object.route = oclass->route; + client->object.token = oclass->token; + client->object.object = oclass->object; + client->debug = oclass->client->debug; + *pobject = &client->object; + return 0; +} + +const struct nvkm_sclass +nvkm_uclient_sclass = { + .oclass = NVIF_CLASS_CLIENT, + .minver = 0, + .maxver = 0, + .ctor = nvkm_uclient_new, +}; + struct nvkm_client_notify { struct nvkm_client *client; struct nvkm_notify n; @@ -138,17 +175,30 @@ nvkm_client_notify_new(struct nvkm_object *object, return ret; } +static const struct nvkm_object_func nvkm_client; +struct nvkm_client * +nvkm_client_search(struct nvkm_client *client, u64 handle) +{ + struct nvkm_object *object; + + object = nvkm_object_search(client, handle, &nvkm_client); + if (IS_ERR(object)) + return (void *)object; + + return nvkm_client(object); +} + static int -nvkm_client_mthd_devlist(struct nvkm_object *object, void *data, u32 size) +nvkm_client_mthd_devlist(struct nvkm_client *client, void *data, u32 size) { union { - struct nv_client_devlist_v0 v0; + struct nvif_client_devlist_v0 v0; } *args = data; int ret = -ENOSYS; - nvif_ioctl(object, "client devlist size %d\n", size); + nvif_ioctl(&client->object, "client devlist size %d\n", size); if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, true))) { - nvif_ioctl(object, "client devlist vers %d count %d\n", + nvif_ioctl(&client->object, "client devlist vers %d count %d\n", args->v0.version, args->v0.count); if (size == sizeof(args->v0.device[0]) * args->v0.count) { ret = nvkm_device_list(args->v0.device, args->v0.count); @@ -167,9 +217,10 @@ nvkm_client_mthd_devlist(struct nvkm_object *object, void *data, u32 size) static int nvkm_client_mthd(struct nvkm_object *object, u32 mthd, void *data, u32 size) { + struct nvkm_client *client = nvkm_client(object); switch (mthd) { - case NV_CLIENT_DEVLIST: - return nvkm_client_mthd_devlist(object, data, size); + case NVIF_CLIENT_V0_DEVLIST: + return nvkm_client_mthd_devlist(client, data, size); default: break; } @@ -190,7 +241,8 @@ nvkm_client_child_get(struct nvkm_object *object, int index, const struct nvkm_sclass *sclass; switch (index) { - case 0: sclass = &nvkm_udevice_sclass; break; + case 0: sclass = &nvkm_uclient_sclass; break; + case 1: sclass = &nvkm_udevice_sclass; break; default: return -EINVAL; } @@ -200,110 +252,54 @@ nvkm_client_child_get(struct nvkm_object *object, int index, return 0; } -static const struct nvkm_object_func -nvkm_client_object_func = { - .mthd = nvkm_client_mthd, - .sclass = nvkm_client_child_get, -}; - -void -nvkm_client_remove(struct nvkm_client *client, struct nvkm_object *object) +static int +nvkm_client_fini(struct nvkm_object *object, bool suspend) { - if (!RB_EMPTY_NODE(&object->node)) - rb_erase(&object->node, &client->objroot); -} - -bool -nvkm_client_insert(struct nvkm_client *client, struct nvkm_object *object) -{ - struct rb_node **ptr = &client->objroot.rb_node; - struct rb_node *parent = NULL; - - while (*ptr) { - struct nvkm_object *this = - container_of(*ptr, typeof(*this), node); - parent = *ptr; - if (object->object < this->object) - ptr = &parent->rb_left; - else - if (object->object > this->object) - ptr = &parent->rb_right; - else - return false; - } - - rb_link_node(&object->node, parent, ptr); - rb_insert_color(&object->node, &client->objroot); - return true; -} - -struct nvkm_object * -nvkm_client_search(struct nvkm_client *client, u64 handle) -{ - struct rb_node *node = client->objroot.rb_node; - while (node) { - struct nvkm_object *object = - container_of(node, typeof(*object), node); - if (handle < object->object) - node = node->rb_left; - else - if (handle > object->object) - node = node->rb_right; - else - return object; - } - return NULL; -} - -int -nvkm_client_fini(struct nvkm_client *client, bool suspend) -{ - struct nvkm_object *object = &client->object; + struct nvkm_client *client = nvkm_client(object); const char *name[2] = { "fini", "suspend" }; int i; nvif_debug(object, "%s notify\n", name[suspend]); for (i = 0; i < ARRAY_SIZE(client->notify); i++) nvkm_client_notify_put(client, i); - return nvkm_object_fini(&client->object, suspend); + return 0; } -int -nvkm_client_init(struct nvkm_client *client) +static void * +nvkm_client_dtor(struct nvkm_object *object) { - return nvkm_object_init(&client->object); -} - -void -nvkm_client_del(struct nvkm_client **pclient) -{ - struct nvkm_client *client = *pclient; + struct nvkm_client *client = nvkm_client(object); int i; - if (client) { - nvkm_client_fini(client, false); - for (i = 0; i < ARRAY_SIZE(client->notify); i++) - nvkm_client_notify_del(client, i); - nvkm_object_dtor(&client->object); - kfree(*pclient); - *pclient = NULL; - } + for (i = 0; i < ARRAY_SIZE(client->notify); i++) + nvkm_client_notify_del(client, i); + return client; } +static const struct nvkm_object_func +nvkm_client = { + .dtor = nvkm_client_dtor, + .fini = nvkm_client_fini, + .mthd = nvkm_client_mthd, + .sclass = nvkm_client_child_get, +}; + int nvkm_client_new(const char *name, u64 device, const char *cfg, - const char *dbg, struct nvkm_client **pclient) + const char *dbg, + int (*ntfy)(const void *, u32, const void *, u32), + struct nvkm_client **pclient) { - struct nvkm_oclass oclass = {}; + struct nvkm_oclass oclass = { .base = nvkm_uclient_sclass }; struct nvkm_client *client; if (!(client = *pclient = kzalloc(sizeof(*client), GFP_KERNEL))) return -ENOMEM; oclass.client = client; - nvkm_object_ctor(&nvkm_client_object_func, &oclass, &client->object); + nvkm_object_ctor(&nvkm_client, &oclass, &client->object); snprintf(client->name, sizeof(client->name), "%s", name); client->device = device; client->debug = nvkm_dbgopt(dbg, "CLIENT"); client->objroot = RB_ROOT; - client->dmaroot = RB_ROOT; + client->ntfy = ntfy; return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/core/engine.c b/drivers/gpu/drm/nouveau/nvkm/core/engine.c index ee8e5831fe37..b6c916954a10 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/engine.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/engine.c @@ -27,6 +27,14 @@ #include +bool +nvkm_engine_chsw_load(struct nvkm_engine *engine) +{ + if (engine->func->chsw_load) + return engine->func->chsw_load(engine); + return false; +} + void nvkm_engine_unref(struct nvkm_engine **pengine) { diff --git a/drivers/gpu/drm/nouveau/nvkm/core/ioctl.c b/drivers/gpu/drm/nouveau/nvkm/core/ioctl.c index b0db51847c36..be19bbe56bba 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/ioctl.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/ioctl.c @@ -29,7 +29,8 @@ #include static int -nvkm_ioctl_nop(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_nop(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_nop_v0 v0; @@ -46,7 +47,8 @@ nvkm_ioctl_nop(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_sclass(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_sclass(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_sclass_v0 v0; @@ -78,12 +80,12 @@ nvkm_ioctl_sclass(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_new(struct nvkm_object *parent, void *data, u32 size) +nvkm_ioctl_new(struct nvkm_client *client, + struct nvkm_object *parent, void *data, u32 size) { union { struct nvif_ioctl_new_v0 v0; } *args = data; - struct nvkm_client *client = parent->client; struct nvkm_object *object = NULL; struct nvkm_oclass oclass; int ret = -ENOSYS, i = 0; @@ -104,9 +106,11 @@ nvkm_ioctl_new(struct nvkm_object *parent, void *data, u32 size) do { memset(&oclass, 0x00, sizeof(oclass)); - oclass.client = client; oclass.handle = args->v0.handle; + oclass.route = args->v0.route; + oclass.token = args->v0.token; oclass.object = args->v0.object; + oclass.client = client; oclass.parent = parent; ret = parent->func->sclass(parent, i++, &oclass); if (ret) @@ -125,10 +129,7 @@ nvkm_ioctl_new(struct nvkm_object *parent, void *data, u32 size) ret = nvkm_object_init(object); if (ret == 0) { list_add(&object->head, &parent->tree); - object->route = args->v0.route; - object->token = args->v0.token; - object->object = args->v0.object; - if (nvkm_client_insert(client, object)) { + if (nvkm_object_insert(object)) { client->data = object; return 0; } @@ -142,7 +143,8 @@ nvkm_ioctl_new(struct nvkm_object *parent, void *data, u32 size) } static int -nvkm_ioctl_del(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_del(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_del none; @@ -156,11 +158,12 @@ nvkm_ioctl_del(struct nvkm_object *object, void *data, u32 size) nvkm_object_del(&object); } - return ret; + return ret ? ret : 1; } static int -nvkm_ioctl_mthd(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_mthd(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_mthd_v0 v0; @@ -179,7 +182,8 @@ nvkm_ioctl_mthd(struct nvkm_object *object, void *data, u32 size) static int -nvkm_ioctl_rd(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_rd(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_rd_v0 v0; @@ -218,7 +222,8 @@ nvkm_ioctl_rd(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_wr(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_wr(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_wr_v0 v0; @@ -246,7 +251,8 @@ nvkm_ioctl_wr(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_map(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_map(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_map_v0 v0; @@ -264,7 +270,8 @@ nvkm_ioctl_map(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_unmap(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_unmap(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_unmap none; @@ -280,7 +287,8 @@ nvkm_ioctl_unmap(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_ntfy_new(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_ntfy_new(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { union { struct nvif_ioctl_ntfy_new_v0 v0; @@ -306,9 +314,9 @@ nvkm_ioctl_ntfy_new(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_ntfy_del(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_ntfy_del(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { - struct nvkm_client *client = object->client; union { struct nvif_ioctl_ntfy_del_v0 v0; } *args = data; @@ -325,9 +333,9 @@ nvkm_ioctl_ntfy_del(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_ntfy_get(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_ntfy_get(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { - struct nvkm_client *client = object->client; union { struct nvif_ioctl_ntfy_get_v0 v0; } *args = data; @@ -344,9 +352,9 @@ nvkm_ioctl_ntfy_get(struct nvkm_object *object, void *data, u32 size) } static int -nvkm_ioctl_ntfy_put(struct nvkm_object *object, void *data, u32 size) +nvkm_ioctl_ntfy_put(struct nvkm_client *client, + struct nvkm_object *object, void *data, u32 size) { - struct nvkm_client *client = object->client; union { struct nvif_ioctl_ntfy_put_v0 v0; } *args = data; @@ -364,7 +372,7 @@ nvkm_ioctl_ntfy_put(struct nvkm_object *object, void *data, u32 size) static struct { int version; - int (*func)(struct nvkm_object *, void *, u32); + int (*func)(struct nvkm_client *, struct nvkm_object *, void *, u32); } nvkm_ioctl_v0[] = { { 0x00, nvkm_ioctl_nop }, @@ -389,13 +397,10 @@ nvkm_ioctl_path(struct nvkm_client *client, u64 handle, u32 type, struct nvkm_object *object; int ret; - if (handle) - object = nvkm_client_search(client, handle); - else - object = &client->object; - if (unlikely(!object)) { + object = nvkm_object_search(client, handle, NULL); + if (IS_ERR(object)) { nvif_ioctl(&client->object, "object not found\n"); - return -ENOENT; + return PTR_ERR(object); } if (owner != NVIF_IOCTL_V0_OWNER_ANY && owner != object->route) { @@ -407,7 +412,7 @@ nvkm_ioctl_path(struct nvkm_client *client, u64 handle, u32 type, if (ret = -EINVAL, type < ARRAY_SIZE(nvkm_ioctl_v0)) { if (nvkm_ioctl_v0[type].version == 0) - ret = nvkm_ioctl_v0[type].func(object, data, size); + ret = nvkm_ioctl_v0[type].func(client, object, data, size); } return ret; @@ -436,12 +441,13 @@ nvkm_ioctl(struct nvkm_client *client, bool supervisor, &args->v0.route, &args->v0.token); } - nvif_ioctl(object, "return %d\n", ret); - if (hack) { - *hack = client->data; - client->data = NULL; + if (ret != 1) { + nvif_ioctl(object, "return %d\n", ret); + if (hack) { + *hack = client->data; + client->data = NULL; + } } - client->super = false; return ret; } diff --git a/drivers/gpu/drm/nouveau/nvkm/core/mm.c b/drivers/gpu/drm/nouveau/nvkm/core/mm.c index 09a1eee8fd33..fd19d652a7ab 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/mm.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/mm.c @@ -147,6 +147,7 @@ nvkm_mm_head(struct nvkm_mm *mm, u8 heap, u8 type, u32 size_max, u32 size_min, if (!this) return -ENOMEM; + this->next = NULL; this->type = type; list_del(&this->fl_entry); *pnode = this; @@ -225,6 +226,7 @@ nvkm_mm_tail(struct nvkm_mm *mm, u8 heap, u8 type, u32 size_max, u32 size_min, if (!this) return -ENOMEM; + this->next = NULL; this->type = type; list_del(&this->fl_entry); *pnode = this; diff --git a/drivers/gpu/drm/nouveau/nvkm/core/object.c b/drivers/gpu/drm/nouveau/nvkm/core/object.c index 67aa7223dcd7..89d2e9da11c7 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/object.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/object.c @@ -25,6 +25,65 @@ #include #include +struct nvkm_object * +nvkm_object_search(struct nvkm_client *client, u64 handle, + const struct nvkm_object_func *func) +{ + struct nvkm_object *object; + + if (handle) { + struct rb_node *node = client->objroot.rb_node; + while (node) { + object = rb_entry(node, typeof(*object), node); + if (handle < object->object) + node = node->rb_left; + else + if (handle > object->object) + node = node->rb_right; + else + goto done; + } + return ERR_PTR(-ENOENT); + } else { + object = &client->object; + } + +done: + if (unlikely(func && object->func != func)) + return ERR_PTR(-EINVAL); + return object; +} + +void +nvkm_object_remove(struct nvkm_object *object) +{ + if (!RB_EMPTY_NODE(&object->node)) + rb_erase(&object->node, &object->client->objroot); +} + +bool +nvkm_object_insert(struct nvkm_object *object) +{ + struct rb_node **ptr = &object->client->objroot.rb_node; + struct rb_node *parent = NULL; + + while (*ptr) { + struct nvkm_object *this = rb_entry(*ptr, typeof(*this), node); + parent = *ptr; + if (object->object < this->object) + ptr = &parent->rb_left; + else + if (object->object > this->object) + ptr = &parent->rb_right; + else + return false; + } + + rb_link_node(&object->node, parent, ptr); + rb_insert_color(&object->node, &object->client->objroot); + return true; +} + int nvkm_object_mthd(struct nvkm_object *object, u32 mthd, void *data, u32 size) { @@ -214,7 +273,7 @@ nvkm_object_del(struct nvkm_object **pobject) struct nvkm_object *object = *pobject; if (object && !WARN_ON(!object->func)) { *pobject = nvkm_object_dtor(object); - nvkm_client_remove(object->client, object); + nvkm_object_remove(object); list_del(&object->head); kfree(*pobject); *pobject = NULL; @@ -230,6 +289,9 @@ nvkm_object_ctor(const struct nvkm_object_func *func, object->engine = nvkm_engine_ref(oclass->engine); object->oclass = oclass->base.oclass; object->handle = oclass->handle; + object->route = oclass->route; + object->token = oclass->token; + object->object = oclass->object; INIT_LIST_HEAD(&object->head); INIT_LIST_HEAD(&object->tree); RB_CLEAR_NODE(&object->node); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c index cceda959b47c..273562dd6bbd 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c @@ -993,7 +993,7 @@ nv92_chipset = { .mc = g84_mc_new, .mmu = nv50_mmu_new, .mxm = nv50_mxm_new, - .pci = g84_pci_new, + .pci = g92_pci_new, .therm = g84_therm_new, .timer = nv41_timer_new, .volt = nv40_volt_new, @@ -2138,6 +2138,7 @@ nv12b_chipset = { .ltc = gm200_ltc_new, .mc = gk20a_mc_new, .mmu = gf100_mmu_new, + .pmu = gm20b_pmu_new, .secboot = gm20b_secboot_new, .timer = gk20a_timer_new, .top = gk104_top_new, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c index 0a1381a84552..070ec5e18fdb 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c @@ -137,7 +137,6 @@ nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func, const struct nvkm_oclass *oclass, struct nvkm_object **pobject) { - struct nvkm_device *device = root->disp->base.engine.subdev.device; struct nvkm_client *client = oclass->client; struct nvkm_dmaobj *dmaobj; struct nv50_disp_dmac *chan; @@ -153,9 +152,9 @@ nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func, if (ret) return ret; - dmaobj = nvkm_dma_search(device->dma, client, push); - if (!dmaobj) - return -ENOENT; + dmaobj = nvkm_dmaobj_search(client, push); + if (IS_ERR(dmaobj)) + return PTR_ERR(dmaobj); if (dmaobj->limit - dmaobj->start != 0xfff) return -EINVAL; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdagt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdagt215.c index 6f0436df0219..f8f2f16c22a2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdagt215.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdagt215.c @@ -59,7 +59,7 @@ gt215_hda_eld(NV50_DISP_MTHD_V1) ); } for (i = 0; i < size; i++) - nvkm_wr32(device, 0x61c440 + soff, (i << 8) | args->v0.data[0]); + nvkm_wr32(device, 0x61c440 + soff, (i << 8) | args->v0.data[i]); for (; i < 0x60; i++) nvkm_wr32(device, 0x61c440 + soff, (i << 8)); nvkm_mask(device, 0x61c448 + soff, 0x80000003, 0x80000003); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c index 567466f93cd5..0db8efbf1c2e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c @@ -433,8 +433,6 @@ nv50_disp_dptmds_war(struct nvkm_device *device) case 0x94: case 0x96: case 0x98: - case 0xaa: - case 0xac: return true; default: break; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c index 4510cb6e10a8..627b9ee1ddd2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c @@ -38,13 +38,6 @@ g94_sor_loff(struct nvkm_output_dp *outp) return g94_sor_soff(outp) + !(outp->base.info.sorconf.link & 1) * 0x80; } -/******************************************************************************* - * TMDS/LVDS - ******************************************************************************/ -static const struct nvkm_output_func -g94_sor_output_func = { -}; - /******************************************************************************* * DisplayPort ******************************************************************************/ diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/dma/base.c index f11ebdd16c77..11b7b8fd5dda 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/dma/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/base.c @@ -28,24 +28,6 @@ #include -struct nvkm_dmaobj * -nvkm_dma_search(struct nvkm_dma *dma, struct nvkm_client *client, u64 object) -{ - struct rb_node *node = client->dmaroot.rb_node; - while (node) { - struct nvkm_dmaobj *dmaobj = - container_of(node, typeof(*dmaobj), rb); - if (object < dmaobj->handle) - node = node->rb_left; - else - if (object > dmaobj->handle) - node = node->rb_right; - else - return dmaobj; - } - return NULL; -} - static int nvkm_dma_oclass_new(struct nvkm_device *device, const struct nvkm_oclass *oclass, void *data, u32 size, @@ -53,34 +35,12 @@ nvkm_dma_oclass_new(struct nvkm_device *device, { struct nvkm_dma *dma = nvkm_dma(oclass->engine); struct nvkm_dmaobj *dmaobj = NULL; - struct nvkm_client *client = oclass->client; - struct rb_node **ptr = &client->dmaroot.rb_node; - struct rb_node *parent = NULL; int ret; ret = dma->func->class_new(dma, oclass, data, size, &dmaobj); if (dmaobj) *pobject = &dmaobj->object; - if (ret) - return ret; - - dmaobj->handle = oclass->object; - - while (*ptr) { - struct nvkm_dmaobj *obj = container_of(*ptr, typeof(*obj), rb); - parent = *ptr; - if (dmaobj->handle < obj->handle) - ptr = &parent->rb_left; - else - if (dmaobj->handle > obj->handle) - ptr = &parent->rb_right; - else - return -EEXIST; - } - - rb_link_node(&dmaobj->rb, parent, ptr); - rb_insert_color(&dmaobj->rb, &client->dmaroot); - return 0; + return ret; } static const struct nvkm_device_oclass diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.c index 13c661b1ef14..d20cc0681a88 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.c @@ -31,6 +31,19 @@ #include #include +static const struct nvkm_object_func nvkm_dmaobj_func; +struct nvkm_dmaobj * +nvkm_dmaobj_search(struct nvkm_client *client, u64 handle) +{ + struct nvkm_object *object; + + object = nvkm_object_search(client, handle, &nvkm_dmaobj_func); + if (IS_ERR(object)) + return (void *)object; + + return nvkm_dmaobj(object); +} + static int nvkm_dmaobj_bind(struct nvkm_object *base, struct nvkm_gpuobj *gpuobj, int align, struct nvkm_gpuobj **pgpuobj) @@ -42,10 +55,7 @@ nvkm_dmaobj_bind(struct nvkm_object *base, struct nvkm_gpuobj *gpuobj, static void * nvkm_dmaobj_dtor(struct nvkm_object *base) { - struct nvkm_dmaobj *dmaobj = nvkm_dmaobj(base); - if (!RB_EMPTY_NODE(&dmaobj->rb)) - rb_erase(&dmaobj->rb, &dmaobj->object.client->dmaroot); - return dmaobj; + return nvkm_dmaobj(base); } static const struct nvkm_object_func @@ -74,7 +84,6 @@ nvkm_dmaobj_ctor(const struct nvkm_dmaobj_func *func, struct nvkm_dma *dma, nvkm_object_ctor(&nvkm_dmaobj_func, oclass, &dmaobj->object); dmaobj->func = func; dmaobj->dma = dma; - RB_CLEAR_NODE(&dmaobj->rb); nvif_ioctl(parent, "create dma size %d\n", *psize); if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, true))) { diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c index 1c9682ae3a6b..660ca7aa95ea 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c @@ -31,6 +31,17 @@ #include #include +void +nvkm_fifo_recover_chan(struct nvkm_fifo *fifo, int chid) +{ + unsigned long flags; + if (WARN_ON(!fifo->func->recover_chan)) + return; + spin_lock_irqsave(&fifo->lock, flags); + fifo->func->recover_chan(fifo, chid); + spin_unlock_irqrestore(&fifo->lock, flags); +} + void nvkm_fifo_pause(struct nvkm_fifo *fifo, unsigned long *flags) { @@ -54,19 +65,29 @@ nvkm_fifo_chan_put(struct nvkm_fifo *fifo, unsigned long flags, } } +struct nvkm_fifo_chan * +nvkm_fifo_chan_inst_locked(struct nvkm_fifo *fifo, u64 inst) +{ + struct nvkm_fifo_chan *chan; + list_for_each_entry(chan, &fifo->chan, head) { + if (chan->inst->addr == inst) { + list_del(&chan->head); + list_add(&chan->head, &fifo->chan); + return chan; + } + } + return NULL; +} + struct nvkm_fifo_chan * nvkm_fifo_chan_inst(struct nvkm_fifo *fifo, u64 inst, unsigned long *rflags) { struct nvkm_fifo_chan *chan; unsigned long flags; spin_lock_irqsave(&fifo->lock, flags); - list_for_each_entry(chan, &fifo->chan, head) { - if (chan->inst->addr == inst) { - list_del(&chan->head); - list_add(&chan->head, &fifo->chan); - *rflags = flags; - return chan; - } + if ((chan = nvkm_fifo_chan_inst_locked(fifo, inst))) { + *rflags = flags; + return chan; } spin_unlock_irqrestore(&fifo->lock, flags); return NULL; @@ -90,9 +111,34 @@ nvkm_fifo_chan_chid(struct nvkm_fifo *fifo, int chid, unsigned long *rflags) return NULL; } +void +nvkm_fifo_kevent(struct nvkm_fifo *fifo, int chid) +{ + nvkm_event_send(&fifo->kevent, 1, chid, NULL, 0); +} + static int -nvkm_fifo_event_ctor(struct nvkm_object *object, void *data, u32 size, - struct nvkm_notify *notify) +nvkm_fifo_kevent_ctor(struct nvkm_object *object, void *data, u32 size, + struct nvkm_notify *notify) +{ + struct nvkm_fifo_chan *chan = nvkm_fifo_chan(object); + if (size == 0) { + notify->size = 0; + notify->types = 1; + notify->index = chan->chid; + return 0; + } + return -ENOSYS; +} + +static const struct nvkm_event_func +nvkm_fifo_kevent_func = { + .ctor = nvkm_fifo_kevent_ctor, +}; + +static int +nvkm_fifo_cevent_ctor(struct nvkm_object *object, void *data, u32 size, + struct nvkm_notify *notify) { if (size == 0) { notify->size = 0; @@ -104,10 +150,16 @@ nvkm_fifo_event_ctor(struct nvkm_object *object, void *data, u32 size, } static const struct nvkm_event_func -nvkm_fifo_event_func = { - .ctor = nvkm_fifo_event_ctor, +nvkm_fifo_cevent_func = { + .ctor = nvkm_fifo_cevent_ctor, }; +void +nvkm_fifo_cevent(struct nvkm_fifo *fifo) +{ + nvkm_event_send(&fifo->cevent, 1, 0, NULL, 0); +} + static void nvkm_fifo_uevent_fini(struct nvkm_event *event, int type, int index) { @@ -241,6 +293,7 @@ nvkm_fifo_dtor(struct nvkm_engine *engine) void *data = fifo; if (fifo->func->dtor) data = fifo->func->dtor(fifo); + nvkm_event_fini(&fifo->kevent); nvkm_event_fini(&fifo->cevent); nvkm_event_fini(&fifo->uevent); return data; @@ -283,5 +336,9 @@ nvkm_fifo_ctor(const struct nvkm_fifo_func *func, struct nvkm_device *device, return ret; } - return nvkm_event_init(&nvkm_fifo_event_func, 1, 1, &fifo->cevent); + ret = nvkm_event_init(&nvkm_fifo_cevent_func, 1, 1, &fifo->cevent); + if (ret) + return ret; + + return nvkm_event_init(&nvkm_fifo_kevent_func, 1, nr, &fifo->kevent); } diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.c index dc6d4678f228..fab760ae922f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.c @@ -371,9 +371,9 @@ nvkm_fifo_chan_ctor(const struct nvkm_fifo_chan_func *func, /* allocate push buffer ctxdma instance */ if (push) { - dmaobj = nvkm_dma_search(device->dma, oclass->client, push); - if (!dmaobj) - return -ENOENT; + dmaobj = nvkm_dmaobj_search(client, push); + if (IS_ERR(dmaobj)) + return PTR_ERR(dmaobj); ret = nvkm_object_bind(&dmaobj->object, chan->inst, -16, &chan->push); @@ -410,6 +410,6 @@ nvkm_fifo_chan_ctor(const struct nvkm_fifo_chan_func *func, base + user * chan->chid; chan->size = user; - nvkm_event_send(&fifo->cevent, 1, 0, NULL, 0); + nvkm_fifo_cevent(fifo); return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.h index 55dc415c5c08..d8019bdacd61 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chan.h @@ -29,5 +29,5 @@ struct nvkm_fifo_chan_oclass { struct nvkm_sclass base; }; -int g84_fifo_chan_ntfy(struct nvkm_fifo_chan *, u32, struct nvkm_event **); +int gf100_fifo_chan_ntfy(struct nvkm_fifo_chan *, u32, struct nvkm_event **); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chang84.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chang84.c index 15a992b3580a..61797c4dd07a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chang84.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/chang84.c @@ -30,12 +30,12 @@ #include -int +static int g84_fifo_chan_ntfy(struct nvkm_fifo_chan *chan, u32 type, struct nvkm_event **pevent) { switch (type) { - case G82_CHANNEL_DMA_V0_NTFY_UEVENT: + case NV826E_V0_NTFY_NON_STALL_INTERRUPT: *pevent = &chan->fifo->uevent; return 0; default: diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gf100.c index ec68ea9747d5..cd468ab1db12 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gf100.c @@ -68,7 +68,14 @@ gf100_fifo_runlist_commit(struct gf100_fifo *fifo) } nvkm_done(cur); - target = (nvkm_memory_target(cur) == NVKM_MEM_TARGET_HOST) ? 0x3 : 0x0; + switch (nvkm_memory_target(cur)) { + case NVKM_MEM_TARGET_VRAM: target = 0; break; + case NVKM_MEM_TARGET_NCOH: target = 3; break; + default: + mutex_unlock(&subdev->mutex); + WARN_ON(1); + return; + } nvkm_wr32(device, 0x002270, (nvkm_memory_addr(cur) >> 12) | (target << 28)); @@ -183,6 +190,7 @@ gf100_fifo_recover(struct gf100_fifo *fifo, struct nvkm_engine *engine, if (engine != &fifo->base.engine) fifo->recover.mask |= 1ULL << engine->subdev.index; schedule_work(&fifo->recover.work); + nvkm_fifo_kevent(&fifo->base, chid); } static const struct nvkm_enum diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c index 38c0910722c0..3a24788c3185 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c @@ -27,11 +27,71 @@ #include #include #include +#include #include #include #include +struct gk104_fifo_engine_status { + bool busy; + bool faulted; + bool chsw; + bool save; + bool load; + struct { + bool tsg; + u32 id; + } prev, next, *chan; +}; + +static void +gk104_fifo_engine_status(struct gk104_fifo *fifo, int engn, + struct gk104_fifo_engine_status *status) +{ + struct nvkm_engine *engine = fifo->engine[engn].engine; + struct nvkm_subdev *subdev = &fifo->base.engine.subdev; + struct nvkm_device *device = subdev->device; + u32 stat = nvkm_rd32(device, 0x002640 + (engn * 0x08)); + + status->busy = !!(stat & 0x80000000); + status->faulted = !!(stat & 0x40000000); + status->next.tsg = !!(stat & 0x10000000); + status->next.id = (stat & 0x0fff0000) >> 16; + status->chsw = !!(stat & 0x00008000); + status->save = !!(stat & 0x00004000); + status->load = !!(stat & 0x00002000); + status->prev.tsg = !!(stat & 0x00001000); + status->prev.id = (stat & 0x00000fff); + status->chan = NULL; + + if (status->busy && status->chsw) { + if (status->load && status->save) { + if (engine && nvkm_engine_chsw_load(engine)) + status->chan = &status->next; + else + status->chan = &status->prev; + } else + if (status->load) { + status->chan = &status->next; + } else { + status->chan = &status->prev; + } + } else + if (status->load) { + status->chan = &status->prev; + } + + nvkm_debug(subdev, "engine %02d: busy %d faulted %d chsw %d " + "save %d load %d %sid %d%s-> %sid %d%s\n", + engn, status->busy, status->faulted, + status->chsw, status->save, status->load, + status->prev.tsg ? "tsg" : "ch", status->prev.id, + status->chan == &status->prev ? "*" : " ", + status->next.tsg ? "tsg" : "ch", status->next.id, + status->chan == &status->next ? "*" : " "); +} + static int gk104_fifo_class_get(struct nvkm_fifo *base, int index, const struct nvkm_fifo_chan_oclass **psclass) @@ -83,10 +143,13 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl) } nvkm_done(mem); - if (nvkm_memory_target(mem) == NVKM_MEM_TARGET_VRAM) - target = 0; - else - target = 3; + switch (nvkm_memory_target(mem)) { + case NVKM_MEM_TARGET_VRAM: target = 0; break; + case NVKM_MEM_TARGET_NCOH: target = 3; break; + default: + WARN_ON(1); + return; + } nvkm_wr32(device, 0x002270, (nvkm_memory_addr(mem) >> 12) | (target << 28)); @@ -149,31 +212,137 @@ gk104_fifo_recover_work(struct work_struct *w) nvkm_mask(device, 0x002630, runm, 0x00000000); } +static void gk104_fifo_recover_engn(struct gk104_fifo *fifo, int engn); + static void -gk104_fifo_recover(struct gk104_fifo *fifo, struct nvkm_engine *engine, - struct gk104_fifo_chan *chan) +gk104_fifo_recover_runl(struct gk104_fifo *fifo, int runl) { struct nvkm_subdev *subdev = &fifo->base.engine.subdev; struct nvkm_device *device = subdev->device; - u32 chid = chan->base.chid; - int engn; + const u32 runm = BIT(runl); - nvkm_error(subdev, "%s engine fault on channel %d, recovering...\n", - nvkm_subdev_name[engine->subdev.index], chid); assert_spin_locked(&fifo->base.lock); + if (fifo->recover.runm & runm) + return; + fifo->recover.runm |= runm; - nvkm_mask(device, 0x800004 + (chid * 0x08), 0x00000800, 0x00000800); - list_del_init(&chan->head); - chan->killed = true; + /* Block runlist to prevent channel assignment(s) from changing. */ + nvkm_mask(device, 0x002630, runm, runm); - for (engn = 0; engn < fifo->engine_nr; engn++) { - if (fifo->engine[engn].engine == engine) { - fifo->recover.engm |= BIT(engn); + /* Schedule recovery. */ + nvkm_warn(subdev, "runlist %d: scheduled for recovery\n", runl); + schedule_work(&fifo->recover.work); +} + +static void +gk104_fifo_recover_chan(struct nvkm_fifo *base, int chid) +{ + struct gk104_fifo *fifo = gk104_fifo(base); + struct nvkm_subdev *subdev = &fifo->base.engine.subdev; + struct nvkm_device *device = subdev->device; + const u32 stat = nvkm_rd32(device, 0x800004 + (chid * 0x08)); + const u32 runl = (stat & 0x000f0000) >> 16; + const bool used = (stat & 0x00000001); + unsigned long engn, engm = fifo->runlist[runl].engm; + struct gk104_fifo_chan *chan; + + assert_spin_locked(&fifo->base.lock); + if (!used) + return; + + /* Lookup SW state for channel, and mark it as dead. */ + list_for_each_entry(chan, &fifo->runlist[runl].chan, head) { + if (chan->base.chid == chid) { + list_del_init(&chan->head); + chan->killed = true; + nvkm_fifo_kevent(&fifo->base, chid); break; } } - fifo->recover.runm |= BIT(chan->runl); + /* Disable channel. */ + nvkm_wr32(device, 0x800004 + (chid * 0x08), stat | 0x00000800); + nvkm_warn(subdev, "channel %d: killed\n", chid); + + /* Block channel assignments from changing during recovery. */ + gk104_fifo_recover_runl(fifo, runl); + + /* Schedule recovery for any engines the channel is on. */ + for_each_set_bit(engn, &engm, fifo->engine_nr) { + struct gk104_fifo_engine_status status; + gk104_fifo_engine_status(fifo, engn, &status); + if (!status.chan || status.chan->id != chid) + continue; + gk104_fifo_recover_engn(fifo, engn); + } +} + +static void +gk104_fifo_recover_engn(struct gk104_fifo *fifo, int engn) +{ + struct nvkm_engine *engine = fifo->engine[engn].engine; + struct nvkm_subdev *subdev = &fifo->base.engine.subdev; + struct nvkm_device *device = subdev->device; + const u32 runl = fifo->engine[engn].runl; + const u32 engm = BIT(engn); + struct gk104_fifo_engine_status status; + int mmui = -1; + + assert_spin_locked(&fifo->base.lock); + if (fifo->recover.engm & engm) + return; + fifo->recover.engm |= engm; + + /* Block channel assignments from changing during recovery. */ + gk104_fifo_recover_runl(fifo, runl); + + /* Determine which channel (if any) is currently on the engine. */ + gk104_fifo_engine_status(fifo, engn, &status); + if (status.chan) { + /* The channel is not longer viable, kill it. */ + gk104_fifo_recover_chan(&fifo->base, status.chan->id); + } + + /* Determine MMU fault ID for the engine, if we're not being + * called from the fault handler already. + */ + if (!status.faulted && engine) { + mmui = nvkm_top_fault_id(device, engine->subdev.index); + if (mmui < 0) { + const struct nvkm_enum *en = fifo->func->fault.engine; + for (; en && en->name; en++) { + if (en->data2 == engine->subdev.index) { + mmui = en->value; + break; + } + } + } + WARN_ON(mmui < 0); + } + + /* Trigger a MMU fault for the engine. + * + * No good idea why this is needed, but nvgpu does something similar, + * and it makes recovery from CTXSW_TIMEOUT a lot more reliable. + */ + if (mmui >= 0) { + nvkm_wr32(device, 0x002a30 + (engn * 0x04), 0x00000100 | mmui); + + /* Wait for fault to trigger. */ + nvkm_msec(device, 2000, + gk104_fifo_engine_status(fifo, engn, &status); + if (status.faulted) + break; + ); + + /* Release MMU fault trigger, and ACK the fault. */ + nvkm_wr32(device, 0x002a30 + (engn * 0x04), 0x00000000); + nvkm_wr32(device, 0x00259c, BIT(mmui)); + nvkm_wr32(device, 0x002100, 0x10000000); + } + + /* Schedule recovery. */ + nvkm_warn(subdev, "engine %d: scheduled for recovery\n", engn); schedule_work(&fifo->recover.work); } @@ -211,34 +380,30 @@ static void gk104_fifo_intr_sched_ctxsw(struct gk104_fifo *fifo) { struct nvkm_device *device = fifo->base.engine.subdev.device; - struct gk104_fifo_chan *chan; - unsigned long flags; + unsigned long flags, engm = 0; u32 engn; + /* We need to ACK the SCHED_ERROR here, and prevent it reasserting, + * as MMU_FAULT cannot be triggered while it's pending. + */ spin_lock_irqsave(&fifo->base.lock, flags); - for (engn = 0; engn < fifo->engine_nr; engn++) { - struct nvkm_engine *engine = fifo->engine[engn].engine; - int runl = fifo->engine[engn].runl; - u32 stat = nvkm_rd32(device, 0x002640 + (engn * 0x08)); - u32 busy = (stat & 0x80000000); - u32 next = (stat & 0x0fff0000) >> 16; - u32 chsw = (stat & 0x00008000); - u32 save = (stat & 0x00004000); - u32 load = (stat & 0x00002000); - u32 prev = (stat & 0x00000fff); - u32 chid = load ? next : prev; - (void)save; + nvkm_mask(device, 0x002140, 0x00000100, 0x00000000); + nvkm_wr32(device, 0x002100, 0x00000100); - if (!busy || !chsw) + for (engn = 0; engn < fifo->engine_nr; engn++) { + struct gk104_fifo_engine_status status; + + gk104_fifo_engine_status(fifo, engn, &status); + if (!status.busy || !status.chsw) continue; - list_for_each_entry(chan, &fifo->runlist[runl].chan, head) { - if (chan->base.chid == chid && engine) { - gk104_fifo_recover(fifo, engine, chan); - break; - } - } + engm |= BIT(engn); } + + for_each_set_bit(engn, &engm, fifo->engine_nr) + gk104_fifo_recover_engn(fifo, engn); + + nvkm_mask(device, 0x002140, 0x00000100, 0x00000100); spin_unlock_irqrestore(&fifo->base.lock, flags); } @@ -301,6 +466,7 @@ gk104_fifo_intr_fault(struct gk104_fifo *fifo, int unit) struct nvkm_fifo_chan *chan; unsigned long flags; char gpcid[8] = "", en[16] = ""; + int engn; er = nvkm_enum_find(fifo->func->fault.reason, reason); eu = nvkm_enum_find(fifo->func->fault.engine, unit); @@ -342,7 +508,8 @@ gk104_fifo_intr_fault(struct gk104_fifo *fifo, int unit) snprintf(en, sizeof(en), "%s", eu->name); } - chan = nvkm_fifo_chan_inst(&fifo->base, (u64)inst << 12, &flags); + spin_lock_irqsave(&fifo->base.lock, flags); + chan = nvkm_fifo_chan_inst_locked(&fifo->base, (u64)inst << 12); nvkm_error(subdev, "%s fault at %010llx engine %02x [%s] client %02x [%s%s] " @@ -353,9 +520,23 @@ gk104_fifo_intr_fault(struct gk104_fifo *fifo, int unit) (u64)inst << 12, chan ? chan->object.client->name : "unknown"); - if (engine && chan) - gk104_fifo_recover(fifo, engine, (void *)chan); - nvkm_fifo_chan_put(&fifo->base, flags, &chan); + + /* Kill the channel that caused the fault. */ + if (chan) + gk104_fifo_recover_chan(&fifo->base, chan->chid); + + /* Channel recovery will probably have already done this for the + * correct engine(s), but just in case we can't find the channel + * information... + */ + for (engn = 0; engn < fifo->engine_nr && engine; engn++) { + if (fifo->engine[engn].engine == engine) { + gk104_fifo_recover_engn(fifo, engn); + break; + } + } + + spin_unlock_irqrestore(&fifo->base.lock, flags); } static const struct nvkm_bitfield gk104_fifo_pbdma_intr_0[] = { @@ -716,6 +897,7 @@ gk104_fifo_ = { .intr = gk104_fifo_intr, .uevent_init = gk104_fifo_uevent_init, .uevent_fini = gk104_fifo_uevent_fini, + .recover_chan = gk104_fifo_recover_chan, .class_get = gk104_fifo_class_get, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogf100.c index 12d964260a29..f9e0377d3d24 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogf100.c @@ -32,6 +32,23 @@ #include #include +int +gf100_fifo_chan_ntfy(struct nvkm_fifo_chan *chan, u32 type, + struct nvkm_event **pevent) +{ + switch (type) { + case NV906F_V0_NTFY_NON_STALL_INTERRUPT: + *pevent = &chan->fifo->uevent; + return 0; + case NV906F_V0_NTFY_KILLED: + *pevent = &chan->fifo->kevent; + return 0; + default: + break; + } + return -EINVAL; +} + static u32 gf100_fifo_gpfifo_engine_addr(struct nvkm_engine *engine) { @@ -184,7 +201,7 @@ gf100_fifo_gpfifo_func = { .dtor = gf100_fifo_gpfifo_dtor, .init = gf100_fifo_gpfifo_init, .fini = gf100_fifo_gpfifo_fini, - .ntfy = g84_fifo_chan_ntfy, + .ntfy = gf100_fifo_chan_ntfy, .engine_ctor = gf100_fifo_gpfifo_engine_ctor, .engine_dtor = gf100_fifo_gpfifo_engine_dtor, .engine_init = gf100_fifo_gpfifo_engine_init, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c index a2df4f3e7763..8abf6f8ef445 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c @@ -50,6 +50,7 @@ gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan) ) < 0) { nvkm_error(subdev, "channel %d [%s] kick timeout\n", chan->base.chid, client->name); + nvkm_fifo_recover_chan(&fifo->base, chan->base.chid); ret = -ETIMEDOUT; } mutex_unlock(&subdev->mutex); @@ -213,7 +214,7 @@ gk104_fifo_gpfifo_func = { .dtor = gk104_fifo_gpfifo_dtor, .init = gk104_fifo_gpfifo_init, .fini = gk104_fifo_gpfifo_fini, - .ntfy = g84_fifo_chan_ntfy, + .ntfy = gf100_fifo_chan_ntfy, .engine_ctor = gk104_fifo_gpfifo_engine_ctor, .engine_dtor = gk104_fifo_gpfifo_engine_dtor, .engine_init = gk104_fifo_gpfifo_engine_init, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h index f6dfb37d9429..f889b13b5e41 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h @@ -6,6 +6,12 @@ int nvkm_fifo_ctor(const struct nvkm_fifo_func *, struct nvkm_device *, int index, int nr, struct nvkm_fifo *); void nvkm_fifo_uevent(struct nvkm_fifo *); +void nvkm_fifo_cevent(struct nvkm_fifo *); +void nvkm_fifo_kevent(struct nvkm_fifo *, int chid); +void nvkm_fifo_recover_chan(struct nvkm_fifo *, int chid); + +struct nvkm_fifo_chan * +nvkm_fifo_chan_inst_locked(struct nvkm_fifo *, u64 inst); struct nvkm_fifo_chan_oclass; struct nvkm_fifo_func { @@ -18,6 +24,7 @@ struct nvkm_fifo_func { void (*start)(struct nvkm_fifo *, unsigned long *); void (*uevent_init)(struct nvkm_fifo *); void (*uevent_fini)(struct nvkm_fifo *); + void (*recover_chan)(struct nvkm_fifo *, int chid); int (*class_get)(struct nvkm_fifo *, int index, const struct nvkm_fifo_chan_oclass **); const struct nvkm_fifo_chan_oclass *chan[]; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/base.c index 467065d1b4e6..cd8cf6f7024c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/base.c @@ -25,6 +25,15 @@ #include +static bool +nvkm_gr_chsw_load(struct nvkm_engine *engine) +{ + struct nvkm_gr *gr = nvkm_gr(engine); + if (gr->func->chsw_load) + return gr->func->chsw_load(gr); + return false; +} + static void nvkm_gr_tile(struct nvkm_engine *engine, int region, struct nvkm_fb_tile *tile) { @@ -106,6 +115,15 @@ nvkm_gr_init(struct nvkm_engine *engine) return gr->func->init(gr); } +static int +nvkm_gr_fini(struct nvkm_engine *engine, bool suspend) +{ + struct nvkm_gr *gr = nvkm_gr(engine); + if (gr->func->fini) + return gr->func->fini(gr, suspend); + return 0; +} + static void * nvkm_gr_dtor(struct nvkm_engine *engine) { @@ -120,8 +138,10 @@ nvkm_gr = { .dtor = nvkm_gr_dtor, .oneinit = nvkm_gr_oneinit, .init = nvkm_gr_init, + .fini = nvkm_gr_fini, .intr = nvkm_gr_intr, .tile = nvkm_gr_tile, + .chsw_load = nvkm_gr_chsw_load, .fifo.cclass = nvkm_gr_cclass_new, .fifo.sclass = nvkm_gr_oclass_get, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/g84.c index ce913300539f..da1ba74682b4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/g84.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/g84.c @@ -25,6 +25,8 @@ #include +#include + static const struct nvkm_bitfield nv50_gr_status[] = { { 0x00000001, "BUSY" }, /* set when any bit is set */ { 0x00000002, "DISPATCH" }, @@ -180,11 +182,11 @@ g84_gr = { .tlb_flush = g84_gr_tlb_flush, .units = nv50_gr_units, .sclass = { - { -1, -1, 0x0030, &nv50_gr_object }, - { -1, -1, 0x502d, &nv50_gr_object }, - { -1, -1, 0x5039, &nv50_gr_object }, - { -1, -1, 0x50c0, &nv50_gr_object }, - { -1, -1, 0x8297, &nv50_gr_object }, + { -1, -1, NV_NULL_CLASS, &nv50_gr_object }, + { -1, -1, NV50_TWOD, &nv50_gr_object }, + { -1, -1, NV50_MEMORY_TO_MEMORY_FORMAT, &nv50_gr_object }, + { -1, -1, NV50_COMPUTE, &nv50_gr_object }, + { -1, -1, G82_TESLA, &nv50_gr_object }, {} } }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c index f65a5b0a1a4d..f9acb8a944d2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c @@ -702,6 +702,22 @@ gf100_gr_pack_mmio[] = { * PGRAPH engine/subdev functions ******************************************************************************/ +static bool +gf100_gr_chsw_load(struct nvkm_gr *base) +{ + struct gf100_gr *gr = gf100_gr(base); + if (!gr->firmware) { + u32 trace = nvkm_rd32(gr->base.engine.subdev.device, 0x40981c); + if (trace & 0x00000040) + return true; + } else { + u32 mthd = nvkm_rd32(gr->base.engine.subdev.device, 0x409808); + if (mthd & 0x00080000) + return true; + } + return false; +} + int gf100_gr_rops(struct gf100_gr *gr) { @@ -1136,7 +1152,7 @@ gf100_gr_trap_intr(struct gf100_gr *gr) if (trap & 0x00000008) { u32 stat = nvkm_rd32(device, 0x408030); - nvkm_snprintbf(error, sizeof(error), gf100_m2mf_error, + nvkm_snprintbf(error, sizeof(error), gf100_ccache_error, stat & 0x3fffffff); nvkm_error(subdev, "CCACHE %08x [%s]\n", stat, error); nvkm_wr32(device, 0x408030, 0xc0000000); @@ -1391,26 +1407,11 @@ gf100_gr_intr(struct nvkm_gr *base) } static void -gf100_gr_init_fw(struct gf100_gr *gr, u32 fuc_base, +gf100_gr_init_fw(struct nvkm_falcon *falcon, struct gf100_gr_fuc *code, struct gf100_gr_fuc *data) { - struct nvkm_device *device = gr->base.engine.subdev.device; - int i; - - nvkm_wr32(device, fuc_base + 0x01c0, 0x01000000); - for (i = 0; i < data->size / 4; i++) - nvkm_wr32(device, fuc_base + 0x01c4, data->data[i]); - - nvkm_wr32(device, fuc_base + 0x0180, 0x01000000); - for (i = 0; i < code->size / 4; i++) { - if ((i & 0x3f) == 0) - nvkm_wr32(device, fuc_base + 0x0188, i >> 6); - nvkm_wr32(device, fuc_base + 0x0184, code->data[i]); - } - - /* code must be padded to 0x40 words */ - for (; i & 0x3f; i++) - nvkm_wr32(device, fuc_base + 0x0184, 0); + nvkm_falcon_load_dmem(falcon, data->data, 0x0, data->size, 0); + nvkm_falcon_load_imem(falcon, code->data, 0x0, code->size, 0, 0, false); } static void @@ -1455,162 +1456,149 @@ gf100_gr_init_csdata(struct gf100_gr *gr, nvkm_wr32(device, falcon + 0x01c4, star + 4); } -int -gf100_gr_init_ctxctl(struct gf100_gr *gr) +/* Initialize context from an external (secure or not) firmware */ +static int +gf100_gr_init_ctxctl_ext(struct gf100_gr *gr) +{ + struct nvkm_subdev *subdev = &gr->base.engine.subdev; + struct nvkm_device *device = subdev->device; + struct nvkm_secboot *sb = device->secboot; + int ret = 0; + + /* load fuc microcode */ + nvkm_mc_unk260(device, 0); + + /* securely-managed falcons must be reset using secure boot */ + if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_FECS)) + ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_FECS); + else + gf100_gr_init_fw(gr->fecs, &gr->fuc409c, &gr->fuc409d); + if (ret) + return ret; + + if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_GPCCS)) + ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_GPCCS); + else + gf100_gr_init_fw(gr->gpccs, &gr->fuc41ac, &gr->fuc41ad); + if (ret) + return ret; + + nvkm_mc_unk260(device, 1); + + /* start both of them running */ + nvkm_wr32(device, 0x409840, 0xffffffff); + nvkm_wr32(device, 0x41a10c, 0x00000000); + nvkm_wr32(device, 0x40910c, 0x00000000); + + nvkm_falcon_start(gr->gpccs); + nvkm_falcon_start(gr->fecs); + + if (nvkm_msec(device, 2000, + if (nvkm_rd32(device, 0x409800) & 0x00000001) + break; + ) < 0) + return -EBUSY; + + nvkm_wr32(device, 0x409840, 0xffffffff); + nvkm_wr32(device, 0x409500, 0x7fffffff); + nvkm_wr32(device, 0x409504, 0x00000021); + + nvkm_wr32(device, 0x409840, 0xffffffff); + nvkm_wr32(device, 0x409500, 0x00000000); + nvkm_wr32(device, 0x409504, 0x00000010); + if (nvkm_msec(device, 2000, + if ((gr->size = nvkm_rd32(device, 0x409800))) + break; + ) < 0) + return -EBUSY; + + nvkm_wr32(device, 0x409840, 0xffffffff); + nvkm_wr32(device, 0x409500, 0x00000000); + nvkm_wr32(device, 0x409504, 0x00000016); + if (nvkm_msec(device, 2000, + if (nvkm_rd32(device, 0x409800)) + break; + ) < 0) + return -EBUSY; + + nvkm_wr32(device, 0x409840, 0xffffffff); + nvkm_wr32(device, 0x409500, 0x00000000); + nvkm_wr32(device, 0x409504, 0x00000025); + if (nvkm_msec(device, 2000, + if (nvkm_rd32(device, 0x409800)) + break; + ) < 0) + return -EBUSY; + + if (device->chipset >= 0xe0) { + nvkm_wr32(device, 0x409800, 0x00000000); + nvkm_wr32(device, 0x409500, 0x00000001); + nvkm_wr32(device, 0x409504, 0x00000030); + if (nvkm_msec(device, 2000, + if (nvkm_rd32(device, 0x409800)) + break; + ) < 0) + return -EBUSY; + + nvkm_wr32(device, 0x409810, 0xb00095c8); + nvkm_wr32(device, 0x409800, 0x00000000); + nvkm_wr32(device, 0x409500, 0x00000001); + nvkm_wr32(device, 0x409504, 0x00000031); + if (nvkm_msec(device, 2000, + if (nvkm_rd32(device, 0x409800)) + break; + ) < 0) + return -EBUSY; + + nvkm_wr32(device, 0x409810, 0x00080420); + nvkm_wr32(device, 0x409800, 0x00000000); + nvkm_wr32(device, 0x409500, 0x00000001); + nvkm_wr32(device, 0x409504, 0x00000032); + if (nvkm_msec(device, 2000, + if (nvkm_rd32(device, 0x409800)) + break; + ) < 0) + return -EBUSY; + + nvkm_wr32(device, 0x409614, 0x00000070); + nvkm_wr32(device, 0x409614, 0x00000770); + nvkm_wr32(device, 0x40802c, 0x00000001); + } + + if (gr->data == NULL) { + int ret = gf100_grctx_generate(gr); + if (ret) { + nvkm_error(subdev, "failed to construct context\n"); + return ret; + } + } + + return 0; +} + +static int +gf100_gr_init_ctxctl_int(struct gf100_gr *gr) { const struct gf100_grctx_func *grctx = gr->func->grctx; struct nvkm_subdev *subdev = &gr->base.engine.subdev; struct nvkm_device *device = subdev->device; - struct nvkm_secboot *sb = device->secboot; - int i; - int ret = 0; - if (gr->firmware) { - /* load fuc microcode */ - nvkm_mc_unk260(device, 0); - - /* securely-managed falcons must be reset using secure boot */ - if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_FECS)) - ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_FECS); - else - gf100_gr_init_fw(gr, 0x409000, &gr->fuc409c, - &gr->fuc409d); - if (ret) - return ret; - - if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_GPCCS)) - ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_GPCCS); - else - gf100_gr_init_fw(gr, 0x41a000, &gr->fuc41ac, - &gr->fuc41ad); - if (ret) - return ret; - - nvkm_mc_unk260(device, 1); - - /* start both of them running */ - nvkm_wr32(device, 0x409840, 0xffffffff); - nvkm_wr32(device, 0x41a10c, 0x00000000); - nvkm_wr32(device, 0x40910c, 0x00000000); - - if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_GPCCS)) - nvkm_secboot_start(sb, NVKM_SECBOOT_FALCON_GPCCS); - else - nvkm_wr32(device, 0x41a100, 0x00000002); - if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_FECS)) - nvkm_secboot_start(sb, NVKM_SECBOOT_FALCON_FECS); - else - nvkm_wr32(device, 0x409100, 0x00000002); - if (nvkm_msec(device, 2000, - if (nvkm_rd32(device, 0x409800) & 0x00000001) - break; - ) < 0) - return -EBUSY; - - nvkm_wr32(device, 0x409840, 0xffffffff); - nvkm_wr32(device, 0x409500, 0x7fffffff); - nvkm_wr32(device, 0x409504, 0x00000021); - - nvkm_wr32(device, 0x409840, 0xffffffff); - nvkm_wr32(device, 0x409500, 0x00000000); - nvkm_wr32(device, 0x409504, 0x00000010); - if (nvkm_msec(device, 2000, - if ((gr->size = nvkm_rd32(device, 0x409800))) - break; - ) < 0) - return -EBUSY; - - nvkm_wr32(device, 0x409840, 0xffffffff); - nvkm_wr32(device, 0x409500, 0x00000000); - nvkm_wr32(device, 0x409504, 0x00000016); - if (nvkm_msec(device, 2000, - if (nvkm_rd32(device, 0x409800)) - break; - ) < 0) - return -EBUSY; - - nvkm_wr32(device, 0x409840, 0xffffffff); - nvkm_wr32(device, 0x409500, 0x00000000); - nvkm_wr32(device, 0x409504, 0x00000025); - if (nvkm_msec(device, 2000, - if (nvkm_rd32(device, 0x409800)) - break; - ) < 0) - return -EBUSY; - - if (device->chipset >= 0xe0) { - nvkm_wr32(device, 0x409800, 0x00000000); - nvkm_wr32(device, 0x409500, 0x00000001); - nvkm_wr32(device, 0x409504, 0x00000030); - if (nvkm_msec(device, 2000, - if (nvkm_rd32(device, 0x409800)) - break; - ) < 0) - return -EBUSY; - - nvkm_wr32(device, 0x409810, 0xb00095c8); - nvkm_wr32(device, 0x409800, 0x00000000); - nvkm_wr32(device, 0x409500, 0x00000001); - nvkm_wr32(device, 0x409504, 0x00000031); - if (nvkm_msec(device, 2000, - if (nvkm_rd32(device, 0x409800)) - break; - ) < 0) - return -EBUSY; - - nvkm_wr32(device, 0x409810, 0x00080420); - nvkm_wr32(device, 0x409800, 0x00000000); - nvkm_wr32(device, 0x409500, 0x00000001); - nvkm_wr32(device, 0x409504, 0x00000032); - if (nvkm_msec(device, 2000, - if (nvkm_rd32(device, 0x409800)) - break; - ) < 0) - return -EBUSY; - - nvkm_wr32(device, 0x409614, 0x00000070); - nvkm_wr32(device, 0x409614, 0x00000770); - nvkm_wr32(device, 0x40802c, 0x00000001); - } - - if (gr->data == NULL) { - int ret = gf100_grctx_generate(gr); - if (ret) { - nvkm_error(subdev, "failed to construct context\n"); - return ret; - } - } - - return 0; - } else if (!gr->func->fecs.ucode) { return -ENOSYS; } /* load HUB microcode */ nvkm_mc_unk260(device, 0); - nvkm_wr32(device, 0x4091c0, 0x01000000); - for (i = 0; i < gr->func->fecs.ucode->data.size / 4; i++) - nvkm_wr32(device, 0x4091c4, gr->func->fecs.ucode->data.data[i]); - - nvkm_wr32(device, 0x409180, 0x01000000); - for (i = 0; i < gr->func->fecs.ucode->code.size / 4; i++) { - if ((i & 0x3f) == 0) - nvkm_wr32(device, 0x409188, i >> 6); - nvkm_wr32(device, 0x409184, gr->func->fecs.ucode->code.data[i]); - } + nvkm_falcon_load_dmem(gr->fecs, gr->func->fecs.ucode->data.data, 0x0, + gr->func->fecs.ucode->data.size, 0); + nvkm_falcon_load_imem(gr->fecs, gr->func->fecs.ucode->code.data, 0x0, + gr->func->fecs.ucode->code.size, 0, 0, false); /* load GPC microcode */ - nvkm_wr32(device, 0x41a1c0, 0x01000000); - for (i = 0; i < gr->func->gpccs.ucode->data.size / 4; i++) - nvkm_wr32(device, 0x41a1c4, gr->func->gpccs.ucode->data.data[i]); - - nvkm_wr32(device, 0x41a180, 0x01000000); - for (i = 0; i < gr->func->gpccs.ucode->code.size / 4; i++) { - if ((i & 0x3f) == 0) - nvkm_wr32(device, 0x41a188, i >> 6); - nvkm_wr32(device, 0x41a184, gr->func->gpccs.ucode->code.data[i]); - } + nvkm_falcon_load_dmem(gr->gpccs, gr->func->gpccs.ucode->data.data, 0x0, + gr->func->gpccs.ucode->data.size, 0); + nvkm_falcon_load_imem(gr->gpccs, gr->func->gpccs.ucode->code.data, 0x0, + gr->func->gpccs.ucode->code.size, 0, 0, false); nvkm_mc_unk260(device, 1); /* load register lists */ @@ -1642,6 +1630,19 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr) return 0; } +int +gf100_gr_init_ctxctl(struct gf100_gr *gr) +{ + int ret; + + if (gr->firmware) + ret = gf100_gr_init_ctxctl_ext(gr); + else + ret = gf100_gr_init_ctxctl_int(gr); + + return ret; +} + static int gf100_gr_oneinit(struct nvkm_gr *base) { @@ -1711,10 +1712,32 @@ static int gf100_gr_init_(struct nvkm_gr *base) { struct gf100_gr *gr = gf100_gr(base); + struct nvkm_subdev *subdev = &base->engine.subdev; + u32 ret; + nvkm_pmu_pgob(gr->base.engine.subdev.device->pmu, false); + + ret = nvkm_falcon_get(gr->fecs, subdev); + if (ret) + return ret; + + ret = nvkm_falcon_get(gr->gpccs, subdev); + if (ret) + return ret; + return gr->func->init(gr); } +static int +gf100_gr_fini_(struct nvkm_gr *base, bool suspend) +{ + struct gf100_gr *gr = gf100_gr(base); + struct nvkm_subdev *subdev = &gr->base.engine.subdev; + nvkm_falcon_put(gr->gpccs, subdev); + nvkm_falcon_put(gr->fecs, subdev); + return 0; +} + void gf100_gr_dtor_fw(struct gf100_gr_fuc *fuc) { @@ -1737,6 +1760,9 @@ gf100_gr_dtor(struct nvkm_gr *base) gr->func->dtor(gr); kfree(gr->data); + nvkm_falcon_del(&gr->gpccs); + nvkm_falcon_del(&gr->fecs); + gf100_gr_dtor_fw(&gr->fuc409c); gf100_gr_dtor_fw(&gr->fuc409d); gf100_gr_dtor_fw(&gr->fuc41ac); @@ -1755,10 +1781,12 @@ gf100_gr_ = { .dtor = gf100_gr_dtor, .oneinit = gf100_gr_oneinit, .init = gf100_gr_init_, + .fini = gf100_gr_fini_, .intr = gf100_gr_intr, .units = gf100_gr_units, .chan_new = gf100_gr_chan_new, .object_get = gf100_gr_object_get, + .chsw_load = gf100_gr_chsw_load, }; int @@ -1828,6 +1856,7 @@ int gf100_gr_ctor(const struct gf100_gr_func *func, struct nvkm_device *device, int index, struct gf100_gr *gr) { + struct nvkm_subdev *subdev = &gr->base.engine.subdev; int ret; gr->func = func; @@ -1840,7 +1869,11 @@ gf100_gr_ctor(const struct gf100_gr_func *func, struct nvkm_device *device, if (ret) return ret; - return 0; + ret = nvkm_falcon_v1_new(subdev, "FECS", 0x409000, &gr->fecs); + if (ret) + return ret; + + return nvkm_falcon_v1_new(subdev, "GPCCS", 0x41a000, &gr->gpccs); } int diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h index 268b8d60ff73..db6ee3b06841 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h @@ -29,6 +29,7 @@ #include #include #include +#include #define GPC_MAX 32 #define TPC_MAX_PER_GPC 8 @@ -75,6 +76,8 @@ struct gf100_gr { const struct gf100_gr_func *func; struct nvkm_gr base; + struct nvkm_falcon *fecs; + struct nvkm_falcon *gpccs; struct gf100_gr_fuc fuc409c; struct gf100_gr_fuc fuc409d; struct gf100_gr_fuc fuc41ac; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt200.c index 2e68919f00b2..c711a55ce392 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt200.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt200.c @@ -23,6 +23,8 @@ */ #include "nv50.h" +#include + static const struct nvkm_gr_func gt200_gr = { .init = nv50_gr_init, @@ -31,11 +33,11 @@ gt200_gr = { .tlb_flush = g84_gr_tlb_flush, .units = nv50_gr_units, .sclass = { - { -1, -1, 0x0030, &nv50_gr_object }, - { -1, -1, 0x502d, &nv50_gr_object }, - { -1, -1, 0x5039, &nv50_gr_object }, - { -1, -1, 0x50c0, &nv50_gr_object }, - { -1, -1, 0x8397, &nv50_gr_object }, + { -1, -1, NV_NULL_CLASS, &nv50_gr_object }, + { -1, -1, NV50_TWOD, &nv50_gr_object }, + { -1, -1, NV50_MEMORY_TO_MEMORY_FORMAT, &nv50_gr_object }, + { -1, -1, NV50_COMPUTE, &nv50_gr_object }, + { -1, -1, GT200_TESLA, &nv50_gr_object }, {} } }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt215.c index 2bf7aac360cc..fa103df32ec7 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt215.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gt215.c @@ -23,6 +23,8 @@ */ #include "nv50.h" +#include + static const struct nvkm_gr_func gt215_gr = { .init = nv50_gr_init, @@ -31,12 +33,12 @@ gt215_gr = { .tlb_flush = g84_gr_tlb_flush, .units = nv50_gr_units, .sclass = { - { -1, -1, 0x0030, &nv50_gr_object }, - { -1, -1, 0x502d, &nv50_gr_object }, - { -1, -1, 0x5039, &nv50_gr_object }, - { -1, -1, 0x50c0, &nv50_gr_object }, - { -1, -1, 0x8597, &nv50_gr_object }, - { -1, -1, 0x85c0, &nv50_gr_object }, + { -1, -1, NV_NULL_CLASS, &nv50_gr_object }, + { -1, -1, NV50_TWOD, &nv50_gr_object }, + { -1, -1, NV50_MEMORY_TO_MEMORY_FORMAT, &nv50_gr_object }, + { -1, -1, NV50_COMPUTE, &nv50_gr_object }, + { -1, -1, GT214_TESLA, &nv50_gr_object }, + { -1, -1, GT214_COMPUTE, &nv50_gr_object }, {} } }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp79.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp79.c index 95d5219faf93..eb1a90644752 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp79.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp79.c @@ -23,6 +23,8 @@ */ #include "nv50.h" +#include + static const struct nvkm_gr_func mcp79_gr = { .init = nv50_gr_init, @@ -30,11 +32,11 @@ mcp79_gr = { .chan_new = nv50_gr_chan_new, .units = nv50_gr_units, .sclass = { - { -1, -1, 0x0030, &nv50_gr_object }, - { -1, -1, 0x502d, &nv50_gr_object }, - { -1, -1, 0x5039, &nv50_gr_object }, - { -1, -1, 0x50c0, &nv50_gr_object }, - { -1, -1, 0x8397, &nv50_gr_object }, + { -1, -1, NV_NULL_CLASS, &nv50_gr_object }, + { -1, -1, NV50_TWOD, &nv50_gr_object }, + { -1, -1, NV50_MEMORY_TO_MEMORY_FORMAT, &nv50_gr_object }, + { -1, -1, NV50_COMPUTE, &nv50_gr_object }, + { -1, -1, GT200_TESLA, &nv50_gr_object }, {} } }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp89.c index 027b58e5976b..c91eb56e9327 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp89.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/mcp89.c @@ -23,6 +23,8 @@ */ #include "nv50.h" +#include + static const struct nvkm_gr_func mcp89_gr = { .init = nv50_gr_init, @@ -31,12 +33,12 @@ mcp89_gr = { .tlb_flush = g84_gr_tlb_flush, .units = nv50_gr_units, .sclass = { - { -1, -1, 0x0030, &nv50_gr_object }, - { -1, -1, 0x502d, &nv50_gr_object }, - { -1, -1, 0x5039, &nv50_gr_object }, - { -1, -1, 0x50c0, &nv50_gr_object }, - { -1, -1, 0x85c0, &nv50_gr_object }, - { -1, -1, 0x8697, &nv50_gr_object }, + { -1, -1, NV_NULL_CLASS, &nv50_gr_object }, + { -1, -1, NV50_TWOD, &nv50_gr_object }, + { -1, -1, NV50_MEMORY_TO_MEMORY_FORMAT, &nv50_gr_object }, + { -1, -1, NV50_COMPUTE, &nv50_gr_object }, + { -1, -1, GT214_COMPUTE, &nv50_gr_object }, + { -1, -1, GT21A_TESLA, &nv50_gr_object }, {} } }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv50.c index fca67de43f2b..df16ffda1749 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv50.c @@ -27,6 +27,8 @@ #include #include +#include + u64 nv50_gr_units(struct nvkm_gr *gr) { @@ -778,11 +780,11 @@ nv50_gr = { .chan_new = nv50_gr_chan_new, .units = nv50_gr_units, .sclass = { - { -1, -1, 0x0030, &nv50_gr_object }, - { -1, -1, 0x502d, &nv50_gr_object }, - { -1, -1, 0x5039, &nv50_gr_object }, - { -1, -1, 0x5097, &nv50_gr_object }, - { -1, -1, 0x50c0, &nv50_gr_object }, + { -1, -1, NV_NULL_CLASS, &nv50_gr_object }, + { -1, -1, NV50_TWOD, &nv50_gr_object }, + { -1, -1, NV50_MEMORY_TO_MEMORY_FORMAT, &nv50_gr_object }, + { -1, -1, NV50_TESLA, &nv50_gr_object }, + { -1, -1, NV50_COMPUTE, &nv50_gr_object }, {} } }; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/priv.h index d8adcdf6985a..2a52d9f026ec 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/priv.h @@ -15,6 +15,7 @@ struct nvkm_gr_func { void *(*dtor)(struct nvkm_gr *); int (*oneinit)(struct nvkm_gr *); int (*init)(struct nvkm_gr *); + int (*fini)(struct nvkm_gr *, bool); void (*intr)(struct nvkm_gr *); void (*tile)(struct nvkm_gr *, int region, struct nvkm_fb_tile *); int (*tlb_flush)(struct nvkm_gr *); @@ -24,6 +25,7 @@ struct nvkm_gr_func { /* Returns chipset-specific counts of units packed into an u64. */ u64 (*units)(struct nvkm_gr *); + bool (*chsw_load)(struct nvkm_gr *); struct nvkm_sclass sclass[]; }; diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/Kbuild b/drivers/gpu/drm/nouveau/nvkm/falcon/Kbuild new file mode 100644 index 000000000000..584863db9bfc --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/Kbuild @@ -0,0 +1,2 @@ +nvkm-y += nvkm/falcon/base.o +nvkm-y += nvkm/falcon/v1.o diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/base.c b/drivers/gpu/drm/nouveau/nvkm/falcon/base.c new file mode 100644 index 000000000000..4852f313762f --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/base.c @@ -0,0 +1,191 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#include "priv.h" + +#include + +void +nvkm_falcon_load_imem(struct nvkm_falcon *falcon, void *data, u32 start, + u32 size, u16 tag, u8 port, bool secure) +{ + if (secure && !falcon->secret) { + nvkm_warn(falcon->user, + "writing with secure tag on a non-secure falcon!\n"); + return; + } + + falcon->func->load_imem(falcon, data, start, size, tag, port, + secure); +} + +void +nvkm_falcon_load_dmem(struct nvkm_falcon *falcon, void *data, u32 start, + u32 size, u8 port) +{ + falcon->func->load_dmem(falcon, data, start, size, port); +} + +void +nvkm_falcon_read_dmem(struct nvkm_falcon *falcon, u32 start, u32 size, u8 port, + void *data) +{ + falcon->func->read_dmem(falcon, start, size, port, data); +} + +void +nvkm_falcon_bind_context(struct nvkm_falcon *falcon, struct nvkm_gpuobj *inst) +{ + if (!falcon->func->bind_context) { + nvkm_error(falcon->user, + "Context binding not supported on this falcon!\n"); + return; + } + + falcon->func->bind_context(falcon, inst); +} + +void +nvkm_falcon_set_start_addr(struct nvkm_falcon *falcon, u32 start_addr) +{ + falcon->func->set_start_addr(falcon, start_addr); +} + +void +nvkm_falcon_start(struct nvkm_falcon *falcon) +{ + falcon->func->start(falcon); +} + +int +nvkm_falcon_enable(struct nvkm_falcon *falcon) +{ + struct nvkm_device *device = falcon->owner->device; + enum nvkm_devidx id = falcon->owner->index; + int ret; + + nvkm_mc_enable(device, id); + ret = falcon->func->enable(falcon); + if (ret) { + nvkm_mc_disable(device, id); + return ret; + } + + return 0; +} + +void +nvkm_falcon_disable(struct nvkm_falcon *falcon) +{ + struct nvkm_device *device = falcon->owner->device; + enum nvkm_devidx id = falcon->owner->index; + + /* already disabled, return or wait_idle will timeout */ + if (!nvkm_mc_enabled(device, id)) + return; + + falcon->func->disable(falcon); + + nvkm_mc_disable(device, id); +} + +int +nvkm_falcon_reset(struct nvkm_falcon *falcon) +{ + nvkm_falcon_disable(falcon); + return nvkm_falcon_enable(falcon); +} + +int +nvkm_falcon_wait_for_halt(struct nvkm_falcon *falcon, u32 ms) +{ + return falcon->func->wait_for_halt(falcon, ms); +} + +int +nvkm_falcon_clear_interrupt(struct nvkm_falcon *falcon, u32 mask) +{ + return falcon->func->clear_interrupt(falcon, mask); +} + +void +nvkm_falcon_put(struct nvkm_falcon *falcon, const struct nvkm_subdev *user) +{ + mutex_lock(&falcon->mutex); + if (falcon->user == user) { + nvkm_debug(falcon->user, "released %s falcon\n", falcon->name); + falcon->user = NULL; + } + mutex_unlock(&falcon->mutex); +} + +int +nvkm_falcon_get(struct nvkm_falcon *falcon, const struct nvkm_subdev *user) +{ + mutex_lock(&falcon->mutex); + if (falcon->user) { + nvkm_error(user, "%s falcon already acquired by %s!\n", + falcon->name, nvkm_subdev_name[falcon->user->index]); + mutex_unlock(&falcon->mutex); + return -EBUSY; + } + + nvkm_debug(user, "acquired %s falcon\n", falcon->name); + falcon->user = user; + mutex_unlock(&falcon->mutex); + return 0; +} + +void +nvkm_falcon_ctor(const struct nvkm_falcon_func *func, + struct nvkm_subdev *subdev, const char *name, u32 addr, + struct nvkm_falcon *falcon) +{ + u32 reg; + + falcon->func = func; + falcon->owner = subdev; + falcon->name = name; + falcon->addr = addr; + mutex_init(&falcon->mutex); + + reg = nvkm_falcon_rd32(falcon, 0x12c); + falcon->version = reg & 0xf; + falcon->secret = (reg >> 4) & 0x3; + falcon->code.ports = (reg >> 8) & 0xf; + falcon->data.ports = (reg >> 12) & 0xf; + + reg = nvkm_falcon_rd32(falcon, 0x108); + falcon->code.limit = (reg & 0x1ff) << 8; + falcon->data.limit = (reg & 0x3fe00) >> 1; + + reg = nvkm_falcon_rd32(falcon, 0xc08); + falcon->debug = (reg >> 20) & 0x1; +} + +void +nvkm_falcon_del(struct nvkm_falcon **pfalcon) +{ + if (*pfalcon) { + kfree(*pfalcon); + *pfalcon = NULL; + } +} diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/priv.h b/drivers/gpu/drm/nouveau/nvkm/falcon/priv.h new file mode 100644 index 000000000000..97b56f759d0b --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/priv.h @@ -0,0 +1,8 @@ +#ifndef __NVKM_FALCON_PRIV_H__ +#define __NVKM_FALCON_PRIV_H__ +#include + +void +nvkm_falcon_ctor(const struct nvkm_falcon_func *, struct nvkm_subdev *, + const char *, u32, struct nvkm_falcon *); +#endif diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/v1.c b/drivers/gpu/drm/nouveau/nvkm/falcon/v1.c new file mode 100644 index 000000000000..b537f111f39c --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/v1.c @@ -0,0 +1,266 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#include "priv.h" + +#include +#include +#include + +static void +nvkm_falcon_v1_load_imem(struct nvkm_falcon *falcon, void *data, u32 start, + u32 size, u16 tag, u8 port, bool secure) +{ + u8 rem = size % 4; + u32 reg; + int i; + + size -= rem; + + reg = start | BIT(24) | (secure ? BIT(28) : 0); + nvkm_falcon_wr32(falcon, 0x180 + (port * 16), reg); + for (i = 0; i < size / 4; i++) { + /* write new tag every 256B */ + if ((i & 0x3f) == 0) + nvkm_falcon_wr32(falcon, 0x188, tag++); + nvkm_falcon_wr32(falcon, 0x184, ((u32 *)data)[i]); + } + + /* + * If size is not a multiple of 4, mask the last work to ensure garbage + * does not get written + */ + if (rem) { + u32 extra = ((u32 *)data)[i]; + + /* write new tag every 256B */ + if ((i & 0x3f) == 0) + nvkm_falcon_wr32(falcon, 0x188, tag++); + nvkm_falcon_wr32(falcon, 0x184, extra & (BIT(rem * 8) - 1)); + ++i; + } + + /* code must be padded to 0x40 words */ + for (; i & 0x3f; i++) + nvkm_falcon_wr32(falcon, 0x184, 0); +} + +static void +nvkm_falcon_v1_load_dmem(struct nvkm_falcon *falcon, void *data, u32 start, + u32 size, u8 port) +{ + u8 rem = size % 4; + int i; + + size -= rem; + + nvkm_falcon_wr32(falcon, 0x1c0 + (port * 16), start | (0x1 << 24)); + for (i = 0; i < size / 4; i++) + nvkm_falcon_wr32(falcon, 0x1c4, ((u32 *)data)[i]); + + /* + * If size is not a multiple of 4, mask the last work to ensure garbage + * does not get read + */ + if (rem) { + u32 extra = ((u32 *)data)[i]; + + nvkm_falcon_wr32(falcon, 0x1c4, extra & (BIT(rem * 8) - 1)); + } +} + +static void +nvkm_falcon_v1_read_dmem(struct nvkm_falcon *falcon, u32 start, u32 size, + u8 port, void *data) +{ + u8 rem = size % 4; + int i; + + size -= rem; + + nvkm_falcon_wr32(falcon, 0x1c0 + (port * 16), start | (0x1 << 25)); + for (i = 0; i < size / 4; i++) + ((u32 *)data)[i] = nvkm_falcon_rd32(falcon, 0x1c4); + + /* + * If size is not a multiple of 4, mask the last work to ensure garbage + * does not get read + */ + if (rem) { + u32 extra = nvkm_falcon_rd32(falcon, 0x1c4); + + for (i = size; i < size + rem; i++) { + ((u8 *)data)[i] = (u8)(extra & 0xff); + extra >>= 8; + } + } +} + +static void +nvkm_falcon_v1_bind_context(struct nvkm_falcon *falcon, struct nvkm_gpuobj *ctx) +{ + u32 inst_loc; + + /* disable instance block binding */ + if (ctx == NULL) { + nvkm_falcon_wr32(falcon, 0x10c, 0x0); + return; + } + + nvkm_falcon_wr32(falcon, 0x10c, 0x1); + + /* setup apertures - virtual */ + nvkm_falcon_wr32(falcon, 0xe00 + 4 * FALCON_DMAIDX_UCODE, 0x4); + nvkm_falcon_wr32(falcon, 0xe00 + 4 * FALCON_DMAIDX_VIRT, 0x0); + /* setup apertures - physical */ + nvkm_falcon_wr32(falcon, 0xe00 + 4 * FALCON_DMAIDX_PHYS_VID, 0x4); + nvkm_falcon_wr32(falcon, 0xe00 + 4 * FALCON_DMAIDX_PHYS_SYS_COH, 0x5); + nvkm_falcon_wr32(falcon, 0xe00 + 4 * FALCON_DMAIDX_PHYS_SYS_NCOH, 0x6); + + /* Set context */ + switch (nvkm_memory_target(ctx->memory)) { + case NVKM_MEM_TARGET_VRAM: inst_loc = 0; break; + case NVKM_MEM_TARGET_NCOH: inst_loc = 3; break; + default: + WARN_ON(1); + return; + } + + /* Enable context */ + nvkm_falcon_mask(falcon, 0x048, 0x1, 0x1); + nvkm_falcon_wr32(falcon, 0x480, + ((ctx->addr >> 12) & 0xfffffff) | + (inst_loc << 28) | (1 << 30)); +} + +static void +nvkm_falcon_v1_set_start_addr(struct nvkm_falcon *falcon, u32 start_addr) +{ + nvkm_falcon_wr32(falcon, 0x104, start_addr); +} + +static void +nvkm_falcon_v1_start(struct nvkm_falcon *falcon) +{ + u32 reg = nvkm_falcon_rd32(falcon, 0x100); + + if (reg & BIT(6)) + nvkm_falcon_wr32(falcon, 0x130, 0x2); + else + nvkm_falcon_wr32(falcon, 0x100, 0x2); +} + +static int +nvkm_falcon_v1_wait_for_halt(struct nvkm_falcon *falcon, u32 ms) +{ + struct nvkm_device *device = falcon->owner->device; + int ret; + + ret = nvkm_wait_msec(device, ms, falcon->addr + 0x100, 0x10, 0x10); + if (ret < 0) + return ret; + + return 0; +} + +static int +nvkm_falcon_v1_clear_interrupt(struct nvkm_falcon *falcon, u32 mask) +{ + struct nvkm_device *device = falcon->owner->device; + int ret; + + /* clear interrupt(s) */ + nvkm_falcon_mask(falcon, 0x004, mask, mask); + /* wait until interrupts are cleared */ + ret = nvkm_wait_msec(device, 10, falcon->addr + 0x008, mask, 0x0); + if (ret < 0) + return ret; + + return 0; +} + +static int +falcon_v1_wait_idle(struct nvkm_falcon *falcon) +{ + struct nvkm_device *device = falcon->owner->device; + int ret; + + ret = nvkm_wait_msec(device, 10, falcon->addr + 0x04c, 0xffff, 0x0); + if (ret < 0) + return ret; + + return 0; +} + +static int +nvkm_falcon_v1_enable(struct nvkm_falcon *falcon) +{ + struct nvkm_device *device = falcon->owner->device; + int ret; + + ret = nvkm_wait_msec(device, 10, falcon->addr + 0x10c, 0x6, 0x0); + if (ret < 0) { + nvkm_error(falcon->user, "Falcon mem scrubbing timeout\n"); + return ret; + } + + ret = falcon_v1_wait_idle(falcon); + if (ret) + return ret; + + /* enable IRQs */ + nvkm_falcon_wr32(falcon, 0x010, 0xff); + + return 0; +} + +static void +nvkm_falcon_v1_disable(struct nvkm_falcon *falcon) +{ + /* disable IRQs and wait for any previous code to complete */ + nvkm_falcon_wr32(falcon, 0x014, 0xff); + falcon_v1_wait_idle(falcon); +} + +static const struct nvkm_falcon_func +nvkm_falcon_v1 = { + .load_imem = nvkm_falcon_v1_load_imem, + .load_dmem = nvkm_falcon_v1_load_dmem, + .read_dmem = nvkm_falcon_v1_read_dmem, + .bind_context = nvkm_falcon_v1_bind_context, + .start = nvkm_falcon_v1_start, + .wait_for_halt = nvkm_falcon_v1_wait_for_halt, + .clear_interrupt = nvkm_falcon_v1_clear_interrupt, + .enable = nvkm_falcon_v1_enable, + .disable = nvkm_falcon_v1_disable, + .set_start_addr = nvkm_falcon_v1_set_start_addr, +}; + +int +nvkm_falcon_v1_new(struct nvkm_subdev *owner, const char *name, u32 addr, + struct nvkm_falcon **pfalcon) +{ + struct nvkm_falcon *falcon; + if (!(falcon = *pfalcon = kzalloc(sizeof(*falcon), GFP_KERNEL))) + return -ENOMEM; + nvkm_falcon_ctor(&nvkm_falcon_v1, owner, name, addr, falcon); + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/Kbuild index be57220a2e01..6b4f1e06a38f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/Kbuild @@ -19,6 +19,7 @@ nvkm-y += nvkm/subdev/bios/pcir.o nvkm-y += nvkm/subdev/bios/perf.o nvkm-y += nvkm/subdev/bios/pll.o nvkm-y += nvkm/subdev/bios/pmu.o +nvkm-y += nvkm/subdev/bios/power_budget.o nvkm-y += nvkm/subdev/bios/ramcfg.o nvkm-y += nvkm/subdev/bios/rammap.o nvkm-y += nvkm/subdev/bios/shadow.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/power_budget.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/power_budget.c new file mode 100644 index 000000000000..617bfffce4ad --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/power_budget.c @@ -0,0 +1,126 @@ +/* + * Copyright 2016 Karol Herbst + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Karol Herbst + */ +#include +#include +#include + +static u32 +nvbios_power_budget_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, + u8 *len) +{ + struct bit_entry bit_P; + u32 power_budget; + + if (bit_entry(bios, 'P', &bit_P) || bit_P.version != 2 || + bit_P.length < 0x2c) + return 0; + + power_budget = nvbios_rd32(bios, bit_P.offset + 0x2c); + if (!power_budget) + return 0; + + *ver = nvbios_rd08(bios, power_budget); + switch (*ver) { + case 0x20: + case 0x30: + *hdr = nvbios_rd08(bios, power_budget + 0x1); + *len = nvbios_rd08(bios, power_budget + 0x2); + *cnt = nvbios_rd08(bios, power_budget + 0x3); + return power_budget; + default: + break; + } + + return 0; +} + +int +nvbios_power_budget_header(struct nvkm_bios *bios, + struct nvbios_power_budget *budget) +{ + struct nvkm_subdev *subdev = &bios->subdev; + u8 ver, hdr, cnt, len, cap_entry; + u32 header; + + if (!bios || !budget) + return -EINVAL; + + header = nvbios_power_budget_table(bios, &ver, &hdr, &cnt, &len); + if (!header || !cnt) + return -ENODEV; + + switch (ver) { + case 0x20: + cap_entry = nvbios_rd08(bios, header + 0x9); + break; + case 0x30: + cap_entry = nvbios_rd08(bios, header + 0xa); + break; + default: + cap_entry = 0xff; + } + + if (cap_entry >= cnt && cap_entry != 0xff) { + nvkm_warn(subdev, + "invalid cap_entry in power budget table found\n"); + budget->cap_entry = 0xff; + return -EINVAL; + } + + budget->offset = header; + budget->ver = ver; + budget->hlen = hdr; + budget->elen = len; + budget->ecount = cnt; + + budget->cap_entry = cap_entry; + + return 0; +} + +int +nvbios_power_budget_entry(struct nvkm_bios *bios, + struct nvbios_power_budget *budget, + u8 idx, struct nvbios_power_budget_entry *entry) +{ + u32 entry_offset; + + if (!bios || !budget || !budget->offset || idx >= budget->ecount + || !entry) + return -EINVAL; + + entry_offset = budget->offset + budget->hlen + idx * budget->elen; + + if (budget->ver >= 0x20) { + entry->min_w = nvbios_rd32(bios, entry_offset + 0x2); + entry->avg_w = nvbios_rd32(bios, entry_offset + 0x6); + entry->max_w = nvbios_rd32(bios, entry_offset + 0xa); + } else { + entry->min_w = 0; + entry->max_w = nvbios_rd32(bios, entry_offset + 0x2); + entry->avg_w = entry->max_w; + } + + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/nv50.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/nv50.c index 5841f297973c..da1770e47490 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/nv50.c @@ -112,7 +112,7 @@ read_pll_src(struct nv50_clk *clk, u32 base) M = (coef & 0x000000ff) >> 0; break; default: - BUG_ON(1); + BUG(); } if (M) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.c b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.c index c714b097719c..59362f8dee22 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.c @@ -50,7 +50,7 @@ nv50_devinit_pll_set(struct nvkm_devinit *init, u32 type, u32 freq) ret = nv04_pll_calc(subdev, &info, freq, &N1, &M1, &N2, &M2, &P); if (!ret) { nvkm_error(subdev, "failed pll calculation\n"); - return ret; + return -EINVAL; } switch (info.type) { diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index 093223d1df4f..6758da93a3a1 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c @@ -445,7 +445,7 @@ gf100_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, { struct nvkm_ltc *ltc = ram->fb->subdev.device->ltc; struct nvkm_mm *mm = &ram->vram; - struct nvkm_mm_node *r; + struct nvkm_mm_node **node, *r; struct nvkm_mem *mem; int type = (memtype & 0x0ff); int back = (memtype & 0x800); @@ -462,7 +462,6 @@ gf100_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, if (!mem) return -ENOMEM; - INIT_LIST_HEAD(&mem->regions); mem->size = size; mutex_lock(&ram->fb->subdev.mutex); @@ -478,6 +477,7 @@ gf100_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, } mem->memtype = type; + node = &mem->mem; do { if (back) ret = nvkm_mm_tail(mm, 0, 1, size, ncmin, align, &r); @@ -489,13 +489,13 @@ gf100_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, return ret; } - list_add_tail(&r->rl_entry, &mem->regions); + *node = r; + node = &r->next; size -= r->length; } while (size); mutex_unlock(&ram->fb->subdev.mutex); - r = list_first_entry(&mem->regions, struct nvkm_mm_node, rl_entry); - mem->offset = (u64)r->offset << NVKM_RAM_MM_SHIFT; + mem->offset = (u64)mem->mem->offset << NVKM_RAM_MM_SHIFT; *pmem = mem; return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c index 7904fa41acef..fb8a1239743d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c @@ -989,7 +989,7 @@ gk104_pll_calc_hiclk(int target_khz, int crystal, int *N1, int *fN1, int *M1, int *P1, int *N2, int *M2, int *P2) { - int best_clk = 0, best_err = target_khz, p_ref, n_ref; + int best_err = target_khz, p_ref, n_ref; bool upper = false; *M1 = 1; @@ -1010,7 +1010,6 @@ gk104_pll_calc_hiclk(int target_khz, int crystal, /* we found a better combination */ if (cur_err < best_err) { best_err = cur_err; - best_clk = cur_clk; *N2 = cur_N; *N1 = n_ref; *P1 = p_ref; @@ -1022,7 +1021,6 @@ gk104_pll_calc_hiclk(int target_khz, int crystal, - target_khz; if (cur_err < best_err) { best_err = cur_err; - best_clk = cur_clk; *N2 = cur_N; *N1 = n_ref; *P1 = p_ref; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/rammcp77.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/rammcp77.c index 0a0e44b75577..017a91de74a0 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/rammcp77.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/rammcp77.c @@ -39,7 +39,7 @@ mcp77_ram_init(struct nvkm_ram *base) u32 flush = ((ram->base.size - (ram->poller_base + 0x40)) >> 5) - 1; /* Enable NISO poller for various clients and set their associated - * read address, only for MCP77/78 and MCP79/7A. (fd#25701) + * read address, only for MCP77/78 and MCP79/7A. (fd#27501) */ nvkm_wr32(device, 0x100c18, dniso); nvkm_mask(device, 0x100c14, 0x00000000, 0x00000001); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv50.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv50.c index 87bde8ff2d6b..6549b0588309 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv50.c @@ -496,15 +496,12 @@ nv50_ram_tidy(struct nvkm_ram *base) void __nv50_ram_put(struct nvkm_ram *ram, struct nvkm_mem *mem) { - struct nvkm_mm_node *this; - - while (!list_empty(&mem->regions)) { - this = list_first_entry(&mem->regions, typeof(*this), rl_entry); - - list_del(&this->rl_entry); - nvkm_mm_free(&ram->vram, &this); + struct nvkm_mm_node *next = mem->mem; + struct nvkm_mm_node *node; + while ((node = next)) { + next = node->next; + nvkm_mm_free(&ram->vram, &node); } - nvkm_mm_free(&ram->tags, &mem->tag); } @@ -530,7 +527,7 @@ nv50_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, { struct nvkm_mm *heap = &ram->vram; struct nvkm_mm *tags = &ram->tags; - struct nvkm_mm_node *r; + struct nvkm_mm_node **node, *r; struct nvkm_mem *mem; int comp = (memtype & 0x300) >> 8; int type = (memtype & 0x07f); @@ -559,11 +556,11 @@ nv50_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, comp = 0; } - INIT_LIST_HEAD(&mem->regions); mem->memtype = (comp << 7) | type; mem->size = max; type = nv50_fb_memtype[type]; + node = &mem->mem; do { if (back) ret = nvkm_mm_tail(heap, 0, type, max, min, align, &r); @@ -575,13 +572,13 @@ nv50_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin, return ret; } - list_add_tail(&r->rl_entry, &mem->regions); + *node = r; + node = &r->next; max -= r->length; } while (max); mutex_unlock(&ram->fb->subdev.mutex); - r = list_first_entry(&mem->regions, struct nvkm_mm_node, rl_entry); - mem->offset = (u64)r->offset << NVKM_RAM_MM_SHIFT; + mem->offset = (u64)mem->mem->offset << NVKM_RAM_MM_SHIFT; *pmem = mem; return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c index f0af2a381eea..fecfa6afcf54 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c @@ -26,6 +26,7 @@ #include #include #include +#include #include static bool @@ -216,10 +217,25 @@ nvkm_iccsense_oneinit(struct nvkm_subdev *subdev) { struct nvkm_iccsense *iccsense = nvkm_iccsense(subdev); struct nvkm_bios *bios = subdev->device->bios; + struct nvbios_power_budget budget; struct nvbios_iccsense stbl; - int i; + int i, ret; - if (!bios || nvbios_iccsense_parse(bios, &stbl) || !stbl.nr_entry) + if (!bios) + return 0; + + ret = nvbios_power_budget_header(bios, &budget); + if (!ret && budget.cap_entry != 0xff) { + struct nvbios_power_budget_entry entry; + ret = nvbios_power_budget_entry(bios, &budget, + budget.cap_entry, &entry); + if (!ret) { + iccsense->power_w_max = entry.avg_w; + iccsense->power_w_crit = entry.max_w; + } + } + + if (nvbios_iccsense_parse(bios, &stbl) || !stbl.nr_entry) return 0; iccsense->data_valid = true; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c index a6a7fa0d7679..9dec58ec3d9f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c @@ -116,7 +116,7 @@ struct gk20a_instmem { static enum nvkm_memory_target gk20a_instobj_target(struct nvkm_memory *memory) { - return NVKM_MEM_TARGET_HOST; + return NVKM_MEM_TARGET_NCOH; } static u64 @@ -305,11 +305,11 @@ gk20a_instobj_dtor_iommu(struct nvkm_memory *memory) struct gk20a_instobj_iommu *node = gk20a_instobj_iommu(memory); struct gk20a_instmem *imem = node->base.imem; struct device *dev = imem->base.subdev.device->dev; - struct nvkm_mm_node *r; + struct nvkm_mm_node *r = node->base.mem.mem; unsigned long flags; int i; - if (unlikely(list_empty(&node->base.mem.regions))) + if (unlikely(!r)) goto out; spin_lock_irqsave(&imem->lock, flags); @@ -320,9 +320,6 @@ gk20a_instobj_dtor_iommu(struct nvkm_memory *memory) spin_unlock_irqrestore(&imem->lock, flags); - r = list_first_entry(&node->base.mem.regions, struct nvkm_mm_node, - rl_entry); - /* clear IOMMU bit to unmap pages */ r->offset &= ~BIT(imem->iommu_bit - imem->iommu_pgshift); @@ -404,10 +401,7 @@ gk20a_instobj_ctor_dma(struct gk20a_instmem *imem, u32 npages, u32 align, node->r.length = (npages << PAGE_SHIFT) >> 12; node->base.mem.offset = node->handle; - - INIT_LIST_HEAD(&node->base.mem.regions); - list_add_tail(&node->r.rl_entry, &node->base.mem.regions); - + node->base.mem.mem = &node->r; return 0; } @@ -484,10 +478,7 @@ gk20a_instobj_ctor_iommu(struct gk20a_instmem *imem, u32 npages, u32 align, r->offset |= BIT(imem->iommu_bit - imem->iommu_pgshift); node->base.mem.offset = ((u64)r->offset) << imem->iommu_pgshift; - - INIT_LIST_HEAD(&node->base.mem.regions); - list_add_tail(&r->rl_entry, &node->base.mem.regions); - + node->base.mem.mem = r; return 0; release_area: diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c index 6b25e25f9eba..09f669ac6630 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c @@ -161,6 +161,16 @@ nvkm_mc_enable(struct nvkm_device *device, enum nvkm_devidx devidx) } } +bool +nvkm_mc_enabled(struct nvkm_device *device, enum nvkm_devidx devidx) +{ + u64 pmc_enable = nvkm_mc_reset_mask(device, false, devidx); + + return (pmc_enable != 0) && + ((nvkm_rd32(device, 0x000200) & pmc_enable) == pmc_enable); +} + + static int nvkm_mc_fini(struct nvkm_subdev *subdev, bool suspend) { diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/base.c index 5df9669ea39c..d06ad2c372bf 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/base.c @@ -31,7 +31,7 @@ nvkm_vm_map_at(struct nvkm_vma *vma, u64 delta, struct nvkm_mem *node) { struct nvkm_vm *vm = vma->vm; struct nvkm_mmu *mmu = vm->mmu; - struct nvkm_mm_node *r; + struct nvkm_mm_node *r = node->mem; int big = vma->node->type != mmu->func->spg_shift; u32 offset = vma->node->offset + (delta >> 12); u32 bits = vma->node->type - 12; @@ -41,7 +41,7 @@ nvkm_vm_map_at(struct nvkm_vma *vma, u64 delta, struct nvkm_mem *node) u32 end, len; delta = 0; - list_for_each_entry(r, &node->regions, rl_entry) { + while (r) { u64 phys = (u64)r->offset << 12; u32 num = r->length >> bits; @@ -65,7 +65,8 @@ nvkm_vm_map_at(struct nvkm_vma *vma, u64 delta, struct nvkm_mem *node) delta += (u64)len << vma->node->type; } - } + r = r->next; + }; mmu->func->flush(vm); } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild index 2a31b7d66a6d..87bf41cef0c6 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild @@ -6,6 +6,7 @@ nvkm-y += nvkm/subdev/pci/nv40.o nvkm-y += nvkm/subdev/pci/nv46.o nvkm-y += nvkm/subdev/pci/nv4c.o nvkm-y += nvkm/subdev/pci/g84.o +nvkm-y += nvkm/subdev/pci/g92.o nvkm-y += nvkm/subdev/pci/g94.o nvkm-y += nvkm/subdev/pci/gf100.o nvkm-y += nvkm/subdev/pci/gf106.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g92.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g92.c new file mode 100644 index 000000000000..48874359d5f6 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g92.c @@ -0,0 +1,57 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "priv.h" + +int +g92_pcie_version_supported(struct nvkm_pci *pci) +{ + if ((nvkm_pci_rd32(pci, 0x460) & 0x200) == 0x200) + return 2; + return 1; +} + +static const struct nvkm_pci_func +g92_pci_func = { + .init = g84_pci_init, + .rd32 = nv40_pci_rd32, + .wr08 = nv40_pci_wr08, + .wr32 = nv40_pci_wr32, + .msi_rearm = nv46_pci_msi_rearm, + + .pcie.init = g84_pcie_init, + .pcie.set_link = g84_pcie_set_link, + + .pcie.max_speed = g84_pcie_max_speed, + .pcie.cur_speed = g84_pcie_cur_speed, + + .pcie.set_version = g84_pcie_set_version, + .pcie.version = g84_pcie_version, + .pcie.version_supported = g92_pcie_version_supported, +}; + +int +g92_pci_new(struct nvkm_device *device, int index, struct nvkm_pci **ppci) +{ + return nvkm_pci_new_(&g92_pci_func, device, index, ppci); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g94.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g94.c index 43444123bc04..09adb37a5664 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g94.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/g94.c @@ -23,14 +23,6 @@ */ #include "priv.h" -int -g94_pcie_version_supported(struct nvkm_pci *pci) -{ - if ((nvkm_pci_rd32(pci, 0x460) & 0x200) == 0x200) - return 2; - return 1; -} - static const struct nvkm_pci_func g94_pci_func = { .init = g84_pci_init, @@ -47,7 +39,7 @@ g94_pci_func = { .pcie.set_version = g84_pcie_set_version, .pcie.version = g84_pcie_version, - .pcie.version_supported = g94_pcie_version_supported, + .pcie.version_supported = g92_pcie_version_supported, }; int diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf100.c index e30ea676baf6..00a5e7d3ee9d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf100.c @@ -92,7 +92,7 @@ gf100_pci_func = { .pcie.set_version = gf100_pcie_set_version, .pcie.version = gf100_pcie_version, - .pcie.version_supported = g94_pcie_version_supported, + .pcie.version_supported = g92_pcie_version_supported, }; int diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf106.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf106.c index c3b798c5c6dd..11bf419afe3f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf106.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gf106.c @@ -39,7 +39,7 @@ gf106_pci_func = { .pcie.set_version = gf100_pcie_set_version, .pcie.version = gf100_pcie_version, - .pcie.version_supported = g94_pcie_version_supported, + .pcie.version_supported = g92_pcie_version_supported, }; int diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/priv.h index 23de3180aae5..86921ec962d6 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/priv.h @@ -44,7 +44,7 @@ enum nvkm_pcie_speed g84_pcie_max_speed(struct nvkm_pci *); int g84_pcie_init(struct nvkm_pci *); int g84_pcie_set_link(struct nvkm_pci *, enum nvkm_pcie_speed, u8); -int g94_pcie_version_supported(struct nvkm_pci *); +int g92_pcie_version_supported(struct nvkm_pci *); void gf100_pcie_set_version(struct nvkm_pci *, u8); int gf100_pcie_version(struct nvkm_pci *); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/Kbuild index 51fb4bf94a44..ca57c1e491b0 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/Kbuild @@ -8,5 +8,6 @@ nvkm-y += nvkm/subdev/pmu/gk110.o nvkm-y += nvkm/subdev/pmu/gk208.o nvkm-y += nvkm/subdev/pmu/gk20a.o nvkm-y += nvkm/subdev/pmu/gm107.o +nvkm-y += nvkm/subdev/pmu/gm20b.o nvkm-y += nvkm/subdev/pmu/gp100.o nvkm-y += nvkm/subdev/pmu/gp102.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c index e611ce80f8ef..a73f690eb4b5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c @@ -116,6 +116,8 @@ nvkm_pmu_init(struct nvkm_subdev *subdev) static void * nvkm_pmu_dtor(struct nvkm_subdev *subdev) { + struct nvkm_pmu *pmu = nvkm_pmu(subdev); + nvkm_falcon_del(&pmu->falcon); return nvkm_pmu(subdev); } @@ -128,6 +130,17 @@ nvkm_pmu = { .intr = nvkm_pmu_intr, }; +int +nvkm_pmu_ctor(const struct nvkm_pmu_func *func, struct nvkm_device *device, + int index, struct nvkm_pmu *pmu) +{ + nvkm_subdev_ctor(&nvkm_pmu, device, index, &pmu->subdev); + pmu->func = func; + INIT_WORK(&pmu->recv.work, nvkm_pmu_recv); + init_waitqueue_head(&pmu->recv.wait); + return nvkm_falcon_v1_new(&pmu->subdev, "PMU", 0x10a000, &pmu->falcon); +} + int nvkm_pmu_new_(const struct nvkm_pmu_func *func, struct nvkm_device *device, int index, struct nvkm_pmu **ppmu) @@ -135,9 +148,5 @@ nvkm_pmu_new_(const struct nvkm_pmu_func *func, struct nvkm_device *device, struct nvkm_pmu *pmu; if (!(pmu = *ppmu = kzalloc(sizeof(*pmu), GFP_KERNEL))) return -ENOMEM; - nvkm_subdev_ctor(&nvkm_pmu, device, index, &pmu->subdev); - pmu->func = func; - INIT_WORK(&pmu->recv.work, nvkm_pmu_recv); - init_waitqueue_head(&pmu->recv.wait); - return 0; + return nvkm_pmu_ctor(func, device, index, *ppmu); } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gk20a.c index f996d90c9f0d..9ca0db796cbe 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gk20a.c @@ -19,7 +19,7 @@ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER * DEALINGS IN THE SOFTWARE. */ -#define gk20a_pmu(p) container_of((p), struct gk20a_pmu, base.subdev) +#define gk20a_pmu(p) container_of((p), struct gk20a_pmu, base) #include "priv.h" #include @@ -43,9 +43,8 @@ struct gk20a_pmu { }; struct gk20a_pmu_dvfs_dev_status { - unsigned long total; - unsigned long busy; - int cur_state; + u32 total; + u32 busy; }; static int @@ -56,13 +55,12 @@ gk20a_pmu_dvfs_target(struct gk20a_pmu *pmu, int *state) return nvkm_clk_astate(clk, *state, 0, false); } -static int +static void gk20a_pmu_dvfs_get_cur_state(struct gk20a_pmu *pmu, int *state) { struct nvkm_clk *clk = pmu->base.subdev.device->clk; *state = clk->pstate; - return 0; } static int @@ -90,28 +88,26 @@ gk20a_pmu_dvfs_get_target_state(struct gk20a_pmu *pmu, *state = level; - if (level == cur_level) - return 0; - else - return 1; + return (level != cur_level); } -static int +static void gk20a_pmu_dvfs_get_dev_status(struct gk20a_pmu *pmu, struct gk20a_pmu_dvfs_dev_status *status) { - struct nvkm_device *device = pmu->base.subdev.device; - status->busy = nvkm_rd32(device, 0x10a508 + (BUSY_SLOT * 0x10)); - status->total= nvkm_rd32(device, 0x10a508 + (CLK_SLOT * 0x10)); - return 0; + struct nvkm_falcon *falcon = pmu->base.falcon; + + status->busy = nvkm_falcon_rd32(falcon, 0x508 + (BUSY_SLOT * 0x10)); + status->total= nvkm_falcon_rd32(falcon, 0x508 + (CLK_SLOT * 0x10)); } static void gk20a_pmu_dvfs_reset_dev_status(struct gk20a_pmu *pmu) { - struct nvkm_device *device = pmu->base.subdev.device; - nvkm_wr32(device, 0x10a508 + (BUSY_SLOT * 0x10), 0x80000000); - nvkm_wr32(device, 0x10a508 + (CLK_SLOT * 0x10), 0x80000000); + struct nvkm_falcon *falcon = pmu->base.falcon; + + nvkm_falcon_wr32(falcon, 0x508 + (BUSY_SLOT * 0x10), 0x80000000); + nvkm_falcon_wr32(falcon, 0x508 + (CLK_SLOT * 0x10), 0x80000000); } static void @@ -127,7 +123,7 @@ gk20a_pmu_dvfs_work(struct nvkm_alarm *alarm) struct nvkm_timer *tmr = device->timer; struct nvkm_volt *volt = device->volt; u32 utilization = 0; - int state, ret; + int state; /* * The PMU is initialized before CLK and VOLT, so we have to make sure the @@ -136,11 +132,7 @@ gk20a_pmu_dvfs_work(struct nvkm_alarm *alarm) if (!clk || !volt) goto resched; - ret = gk20a_pmu_dvfs_get_dev_status(pmu, &status); - if (ret) { - nvkm_warn(subdev, "failed to get device status\n"); - goto resched; - } + gk20a_pmu_dvfs_get_dev_status(pmu, &status); if (status.total) utilization = div_u64((u64)status.busy * 100, status.total); @@ -150,11 +142,7 @@ gk20a_pmu_dvfs_work(struct nvkm_alarm *alarm) nvkm_trace(subdev, "utilization = %d %%, avg_load = %d %%\n", utilization, data->avg_load); - ret = gk20a_pmu_dvfs_get_cur_state(pmu, &state); - if (ret) { - nvkm_warn(subdev, "failed to get current state\n"); - goto resched; - } + gk20a_pmu_dvfs_get_cur_state(pmu, &state); if (gk20a_pmu_dvfs_get_target_state(pmu, &state, data->avg_load)) { nvkm_trace(subdev, "set new state to %d\n", state); @@ -166,32 +154,36 @@ gk20a_pmu_dvfs_work(struct nvkm_alarm *alarm) nvkm_timer_alarm(tmr, 100000000, alarm); } -static int -gk20a_pmu_fini(struct nvkm_subdev *subdev, bool suspend) +static void +gk20a_pmu_fini(struct nvkm_pmu *pmu) { - struct gk20a_pmu *pmu = gk20a_pmu(subdev); - nvkm_timer_alarm_cancel(subdev->device->timer, &pmu->alarm); - return 0; -} + struct gk20a_pmu *gpmu = gk20a_pmu(pmu); + nvkm_timer_alarm_cancel(pmu->subdev.device->timer, &gpmu->alarm); -static void * -gk20a_pmu_dtor(struct nvkm_subdev *subdev) -{ - return gk20a_pmu(subdev); + nvkm_falcon_put(pmu->falcon, &pmu->subdev); } static int -gk20a_pmu_init(struct nvkm_subdev *subdev) +gk20a_pmu_init(struct nvkm_pmu *pmu) { - struct gk20a_pmu *pmu = gk20a_pmu(subdev); - struct nvkm_device *device = pmu->base.subdev.device; + struct gk20a_pmu *gpmu = gk20a_pmu(pmu); + struct nvkm_subdev *subdev = &pmu->subdev; + struct nvkm_device *device = pmu->subdev.device; + struct nvkm_falcon *falcon = pmu->falcon; + int ret; + + ret = nvkm_falcon_get(falcon, subdev); + if (ret) { + nvkm_error(subdev, "cannot acquire %s falcon!\n", falcon->name); + return ret; + } /* init pwr perf counter */ - nvkm_wr32(device, 0x10a504 + (BUSY_SLOT * 0x10), 0x00200001); - nvkm_wr32(device, 0x10a50c + (BUSY_SLOT * 0x10), 0x00000002); - nvkm_wr32(device, 0x10a50c + (CLK_SLOT * 0x10), 0x00000003); + nvkm_falcon_wr32(falcon, 0x504 + (BUSY_SLOT * 0x10), 0x00200001); + nvkm_falcon_wr32(falcon, 0x50c + (BUSY_SLOT * 0x10), 0x00000002); + nvkm_falcon_wr32(falcon, 0x50c + (CLK_SLOT * 0x10), 0x00000003); - nvkm_timer_alarm(device->timer, 2000000000, &pmu->alarm); + nvkm_timer_alarm(device->timer, 2000000000, &gpmu->alarm); return 0; } @@ -202,26 +194,26 @@ gk20a_dvfs_data= { .p_smooth = 1, }; -static const struct nvkm_subdev_func +static const struct nvkm_pmu_func gk20a_pmu = { .init = gk20a_pmu_init, .fini = gk20a_pmu_fini, - .dtor = gk20a_pmu_dtor, + .reset = gt215_pmu_reset, }; int gk20a_pmu_new(struct nvkm_device *device, int index, struct nvkm_pmu **ppmu) { - static const struct nvkm_pmu_func func = {}; struct gk20a_pmu *pmu; if (!(pmu = kzalloc(sizeof(*pmu), GFP_KERNEL))) return -ENOMEM; - pmu->base.func = &func; *ppmu = &pmu->base; - nvkm_subdev_ctor(&gk20a_pmu, device, index, &pmu->base.subdev); + nvkm_pmu_ctor(&gk20a_pmu, device, index, &pmu->base); + pmu->data = &gk20a_dvfs_data; nvkm_alarm_init(&pmu->alarm, gk20a_pmu_dvfs_work); + return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gm20b.c new file mode 100644 index 000000000000..0b8a1cc4a0ee --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/gm20b.c @@ -0,0 +1,34 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#include "priv.h" + +static const struct nvkm_pmu_func +gm20b_pmu = { + .reset = gt215_pmu_reset, +}; + +int +gm20b_pmu_new(struct nvkm_device *device, int index, struct nvkm_pmu **ppmu) +{ + return nvkm_pmu_new_(&gm20b_pmu, device, index, ppmu); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/priv.h index 2e2179a4ad17..096cba069f72 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/priv.h @@ -4,6 +4,8 @@ #include #include +int nvkm_pmu_ctor(const struct nvkm_pmu_func *, struct nvkm_device *, + int index, struct nvkm_pmu *); int nvkm_pmu_new_(const struct nvkm_pmu_func *, struct nvkm_device *, int index, struct nvkm_pmu **); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild index b02b868a6589..5076d1500f47 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild @@ -1,3 +1,7 @@ nvkm-y += nvkm/subdev/secboot/base.o +nvkm-y += nvkm/subdev/secboot/ls_ucode_gr.o +nvkm-y += nvkm/subdev/secboot/acr.o +nvkm-y += nvkm/subdev/secboot/acr_r352.o +nvkm-y += nvkm/subdev/secboot/acr_r361.o nvkm-y += nvkm/subdev/secboot/gm200.o nvkm-y += nvkm/subdev/secboot/gm20b.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.c new file mode 100644 index 000000000000..75dc06557877 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.c @@ -0,0 +1,54 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#include "acr.h" + +#include + +/** + * Convenience function to duplicate a firmware file in memory and check that + * it has the required minimum size. + */ +void * +nvkm_acr_load_firmware(const struct nvkm_subdev *subdev, const char *name, + size_t min_size) +{ + const struct firmware *fw; + void *blob; + int ret; + + ret = nvkm_firmware_get(subdev->device, name, &fw); + if (ret) + return ERR_PTR(ret); + if (fw->size < min_size) { + nvkm_error(subdev, "%s is smaller than expected size %zu\n", + name, min_size); + nvkm_firmware_put(fw); + return ERR_PTR(-EINVAL); + } + blob = kmemdup(fw->data, fw->size, GFP_KERNEL); + nvkm_firmware_put(fw); + if (!blob) + return ERR_PTR(-ENOMEM); + + return blob; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h new file mode 100644 index 000000000000..97795b342b6f --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h @@ -0,0 +1,69 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#ifndef __NVKM_SECBOOT_ACR_H__ +#define __NVKM_SECBOOT_ACR_H__ + +#include "priv.h" + +struct nvkm_acr; + +/** + * struct nvkm_acr_func - properties and functions specific to an ACR + * + * @load: make the ACR ready to run on the given secboot device + * @reset: reset the specified falcon + * @start: start the specified falcon (assumed to have been reset) + */ +struct nvkm_acr_func { + void (*dtor)(struct nvkm_acr *); + int (*oneinit)(struct nvkm_acr *, struct nvkm_secboot *); + int (*fini)(struct nvkm_acr *, struct nvkm_secboot *, bool); + int (*load)(struct nvkm_acr *, struct nvkm_secboot *, + struct nvkm_gpuobj *, u64); + int (*reset)(struct nvkm_acr *, struct nvkm_secboot *, + enum nvkm_secboot_falcon); + int (*start)(struct nvkm_acr *, struct nvkm_secboot *, + enum nvkm_secboot_falcon); +}; + +/** + * struct nvkm_acr - instance of an ACR + * + * @boot_falcon: ID of the falcon that will perform secure boot + * @managed_falcons: bitfield of falcons managed by this ACR + * @start_address: virtual start address of the HS bootloader + */ +struct nvkm_acr { + const struct nvkm_acr_func *func; + const struct nvkm_subdev *subdev; + + enum nvkm_secboot_falcon boot_falcon; + unsigned long managed_falcons; + u32 start_address; +}; + +void *nvkm_acr_load_firmware(const struct nvkm_subdev *, const char *, size_t); + +struct nvkm_acr *acr_r352_new(unsigned long); +struct nvkm_acr *acr_r361_new(unsigned long); + +#endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c new file mode 100644 index 000000000000..1aa37ea18580 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c @@ -0,0 +1,936 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#include "acr_r352.h" + +#include +#include +#include + +/** + * struct hsf_fw_header - HS firmware descriptor + * @sig_dbg_offset: offset of the debug signature + * @sig_dbg_size: size of the debug signature + * @sig_prod_offset: offset of the production signature + * @sig_prod_size: size of the production signature + * @patch_loc: offset of the offset (sic) of where the signature is + * @patch_sig: offset of the offset (sic) to add to sig_*_offset + * @hdr_offset: offset of the load header (see struct hs_load_header) + * @hdr_size: size of above header + * + * This structure is embedded in the HS firmware image at + * hs_bin_hdr.header_offset. + */ +struct hsf_fw_header { + u32 sig_dbg_offset; + u32 sig_dbg_size; + u32 sig_prod_offset; + u32 sig_prod_size; + u32 patch_loc; + u32 patch_sig; + u32 hdr_offset; + u32 hdr_size; +}; + +/** + * struct acr_r352_flcn_bl_desc - DMEM bootloader descriptor + * @signature: 16B signature for secure code. 0s if no secure code + * @ctx_dma: DMA context to be used by BL while loading code/data + * @code_dma_base: 256B-aligned Physical FB Address where code is located + * (falcon's $xcbase register) + * @non_sec_code_off: offset from code_dma_base where the non-secure code is + * located. The offset must be multiple of 256 to help perf + * @non_sec_code_size: the size of the nonSecure code part. + * @sec_code_off: offset from code_dma_base where the secure code is + * located. The offset must be multiple of 256 to help perf + * @sec_code_size: offset from code_dma_base where the secure code is + * located. The offset must be multiple of 256 to help perf + * @code_entry_point: code entry point which will be invoked by BL after + * code is loaded. + * @data_dma_base: 256B aligned Physical FB Address where data is located. + * (falcon's $xdbase register) + * @data_size: size of data block. Should be multiple of 256B + * + * Structure used by the bootloader to load the rest of the code. This has + * to be filled by host and copied into DMEM at offset provided in the + * hsflcn_bl_desc.bl_desc_dmem_load_off. + */ +struct acr_r352_flcn_bl_desc { + u32 reserved[4]; + u32 signature[4]; + u32 ctx_dma; + u32 code_dma_base; + u32 non_sec_code_off; + u32 non_sec_code_size; + u32 sec_code_off; + u32 sec_code_size; + u32 code_entry_point; + u32 data_dma_base; + u32 data_size; + u32 code_dma_base1; + u32 data_dma_base1; +}; + +/** + * acr_r352_generate_flcn_bl_desc - generate generic BL descriptor for LS image + */ +static void +acr_r352_generate_flcn_bl_desc(const struct nvkm_acr *acr, + const struct ls_ucode_img *_img, u64 wpr_addr, + void *_desc) +{ + struct ls_ucode_img_r352 *img = ls_ucode_img_r352(_img); + struct acr_r352_flcn_bl_desc *desc = _desc; + const struct ls_ucode_img_desc *pdesc = &_img->ucode_desc; + u64 base, addr_code, addr_data; + + base = wpr_addr + img->lsb_header.ucode_off + pdesc->app_start_offset; + addr_code = (base + pdesc->app_resident_code_offset) >> 8; + addr_data = (base + pdesc->app_resident_data_offset) >> 8; + + desc->ctx_dma = FALCON_DMAIDX_UCODE; + desc->code_dma_base = lower_32_bits(addr_code); + desc->code_dma_base1 = upper_32_bits(addr_code); + desc->non_sec_code_off = pdesc->app_resident_code_offset; + desc->non_sec_code_size = pdesc->app_resident_code_size; + desc->code_entry_point = pdesc->app_imem_entry; + desc->data_dma_base = lower_32_bits(addr_data); + desc->data_dma_base1 = upper_32_bits(addr_data); + desc->data_size = pdesc->app_resident_data_size; +} + + +/** + * struct hsflcn_acr_desc - data section of the HS firmware + * + * This header is to be copied at the beginning of DMEM by the HS bootloader. + * + * @signature: signature of ACR ucode + * @wpr_region_id: region ID holding the WPR header and its details + * @wpr_offset: offset from the WPR region holding the wpr header + * @regions: region descriptors + * @nonwpr_ucode_blob_size: size of LS blob + * @nonwpr_ucode_blob_start: FB location of LS blob is + */ +struct hsflcn_acr_desc { + union { + u8 reserved_dmem[0x200]; + u32 signatures[4]; + } ucode_reserved_space; + u32 wpr_region_id; + u32 wpr_offset; + u32 mmu_mem_range; +#define FLCN_ACR_MAX_REGIONS 2 + struct { + u32 no_regions; + struct { + u32 start_addr; + u32 end_addr; + u32 region_id; + u32 read_mask; + u32 write_mask; + u32 client_mask; + } region_props[FLCN_ACR_MAX_REGIONS]; + } regions; + u32 ucode_blob_size; + u64 ucode_blob_base __aligned(8); + struct { + u32 vpr_enabled; + u32 vpr_start; + u32 vpr_end; + u32 hdcp_policies; + } vpr_desc; +}; + + +/* + * Low-secure blob creation + */ + +/** + * ls_ucode_img_load() - create a lsf_ucode_img and load it + */ +struct ls_ucode_img * +acr_r352_ls_ucode_img_load(const struct acr_r352 *acr, + enum nvkm_secboot_falcon falcon_id) +{ + const struct nvkm_subdev *subdev = acr->base.subdev; + struct ls_ucode_img_r352 *img; + int ret; + + img = kzalloc(sizeof(*img), GFP_KERNEL); + if (!img) + return ERR_PTR(-ENOMEM); + + img->base.falcon_id = falcon_id; + + ret = acr->func->ls_func[falcon_id]->load(subdev, &img->base); + + if (ret) { + kfree(img->base.ucode_data); + kfree(img->base.sig); + kfree(img); + return ERR_PTR(ret); + } + + /* Check that the signature size matches our expectations... */ + if (img->base.sig_size != sizeof(img->lsb_header.signature)) { + nvkm_error(subdev, "invalid signature size for %s falcon!\n", + nvkm_secboot_falcon_name[falcon_id]); + return ERR_PTR(-EINVAL); + } + + /* Copy signature to the right place */ + memcpy(&img->lsb_header.signature, img->base.sig, img->base.sig_size); + + /* not needed? the signature should already have the right value */ + img->lsb_header.signature.falcon_id = falcon_id; + + return &img->base; +} + +#define LSF_LSB_HEADER_ALIGN 256 +#define LSF_BL_DATA_ALIGN 256 +#define LSF_BL_DATA_SIZE_ALIGN 256 +#define LSF_BL_CODE_SIZE_ALIGN 256 +#define LSF_UCODE_DATA_ALIGN 4096 + +/** + * acr_r352_ls_img_fill_headers - fill the WPR and LSB headers of an image + * @acr: ACR to use + * @img: image to generate for + * @offset: offset in the WPR region where this image starts + * + * Allocate space in the WPR area from offset and write the WPR and LSB headers + * accordingly. + * + * Return: offset at the end of this image. + */ +static u32 +acr_r352_ls_img_fill_headers(struct acr_r352 *acr, + struct ls_ucode_img_r352 *img, u32 offset) +{ + struct ls_ucode_img *_img = &img->base; + struct acr_r352_lsf_wpr_header *whdr = &img->wpr_header; + struct acr_r352_lsf_lsb_header *lhdr = &img->lsb_header; + struct ls_ucode_img_desc *desc = &_img->ucode_desc; + const struct acr_r352_ls_func *func = + acr->func->ls_func[_img->falcon_id]; + + /* Fill WPR header */ + whdr->falcon_id = _img->falcon_id; + whdr->bootstrap_owner = acr->base.boot_falcon; + whdr->status = LSF_IMAGE_STATUS_COPY; + + /* Skip bootstrapping falcons started by someone else than ACR */ + if (acr->lazy_bootstrap & BIT(_img->falcon_id)) + whdr->lazy_bootstrap = 1; + + /* Align, save off, and include an LSB header size */ + offset = ALIGN(offset, LSF_LSB_HEADER_ALIGN); + whdr->lsb_offset = offset; + offset += sizeof(*lhdr); + + /* + * Align, save off, and include the original (static) ucode + * image size + */ + offset = ALIGN(offset, LSF_UCODE_DATA_ALIGN); + lhdr->ucode_off = offset; + offset += _img->ucode_size; + + /* + * For falcons that use a boot loader (BL), we append a loader + * desc structure on the end of the ucode image and consider + * this the boot loader data. The host will then copy the loader + * desc args to this space within the WPR region (before locking + * down) and the HS bin will then copy them to DMEM 0 for the + * loader. + */ + lhdr->bl_code_size = ALIGN(desc->bootloader_size, + LSF_BL_CODE_SIZE_ALIGN); + lhdr->ucode_size = ALIGN(desc->app_resident_data_offset, + LSF_BL_CODE_SIZE_ALIGN) + lhdr->bl_code_size; + lhdr->data_size = ALIGN(desc->app_size, LSF_BL_CODE_SIZE_ALIGN) + + lhdr->bl_code_size - lhdr->ucode_size; + /* + * Though the BL is located at 0th offset of the image, the VA + * is different to make sure that it doesn't collide the actual + * OS VA range + */ + lhdr->bl_imem_off = desc->bootloader_imem_offset; + lhdr->app_code_off = desc->app_start_offset + + desc->app_resident_code_offset; + lhdr->app_code_size = desc->app_resident_code_size; + lhdr->app_data_off = desc->app_start_offset + + desc->app_resident_data_offset; + lhdr->app_data_size = desc->app_resident_data_size; + + lhdr->flags = func->lhdr_flags; + if (_img->falcon_id == acr->base.boot_falcon) + lhdr->flags |= LSF_FLAG_DMACTL_REQ_CTX; + + /* Align and save off BL descriptor size */ + lhdr->bl_data_size = ALIGN(func->bl_desc_size, LSF_BL_DATA_SIZE_ALIGN); + + /* + * Align, save off, and include the additional BL data + */ + offset = ALIGN(offset, LSF_BL_DATA_ALIGN); + lhdr->bl_data_off = offset; + offset += lhdr->bl_data_size; + + return offset; +} + +/** + * acr_r352_ls_fill_headers - fill WPR and LSB headers of all managed images + */ +int +acr_r352_ls_fill_headers(struct acr_r352 *acr, struct list_head *imgs) +{ + struct ls_ucode_img_r352 *img; + struct list_head *l; + u32 count = 0; + u32 offset; + + /* Count the number of images to manage */ + list_for_each(l, imgs) + count++; + + /* + * Start with an array of WPR headers at the base of the WPR. + * The expectation here is that the secure falcon will do a single DMA + * read of this array and cache it internally so it's ok to pack these. + * Also, we add 1 to the falcon count to indicate the end of the array. + */ + offset = sizeof(img->wpr_header) * (count + 1); + + /* + * Walk the managed falcons, accounting for the LSB structs + * as well as the ucode images. + */ + list_for_each_entry(img, imgs, base.node) { + offset = acr_r352_ls_img_fill_headers(acr, img, offset); + } + + return offset; +} + +/** + * acr_r352_ls_write_wpr - write the WPR blob contents + */ +int +acr_r352_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs, + struct nvkm_gpuobj *wpr_blob, u32 wpr_addr) +{ + struct ls_ucode_img *_img; + u32 pos = 0; + + nvkm_kmap(wpr_blob); + + list_for_each_entry(_img, imgs, node) { + struct ls_ucode_img_r352 *img = ls_ucode_img_r352(_img); + const struct acr_r352_ls_func *ls_func = + acr->func->ls_func[_img->falcon_id]; + u8 gdesc[ls_func->bl_desc_size]; + + nvkm_gpuobj_memcpy_to(wpr_blob, pos, &img->wpr_header, + sizeof(img->wpr_header)); + + nvkm_gpuobj_memcpy_to(wpr_blob, img->wpr_header.lsb_offset, + &img->lsb_header, sizeof(img->lsb_header)); + + /* Generate and write BL descriptor */ + memset(gdesc, 0, ls_func->bl_desc_size); + ls_func->generate_bl_desc(&acr->base, _img, wpr_addr, gdesc); + + nvkm_gpuobj_memcpy_to(wpr_blob, img->lsb_header.bl_data_off, + gdesc, ls_func->bl_desc_size); + + /* Copy ucode */ + nvkm_gpuobj_memcpy_to(wpr_blob, img->lsb_header.ucode_off, + _img->ucode_data, _img->ucode_size); + + pos += sizeof(img->wpr_header); + } + + nvkm_wo32(wpr_blob, pos, NVKM_SECBOOT_FALCON_INVALID); + + nvkm_done(wpr_blob); + + return 0; +} + +/* Both size and address of WPR need to be 128K-aligned */ +#define WPR_ALIGNMENT 0x20000 +/** + * acr_r352_prepare_ls_blob() - prepare the LS blob + * + * For each securely managed falcon, load the FW, signatures and bootloaders and + * prepare a ucode blob. Then, compute the offsets in the WPR region for each + * blob, and finally write the headers and ucode blobs into a GPU object that + * will be copied into the WPR region by the HS firmware. + */ +static int +acr_r352_prepare_ls_blob(struct acr_r352 *acr, u64 wpr_addr, u32 wpr_size) +{ + const struct nvkm_subdev *subdev = acr->base.subdev; + struct list_head imgs; + struct ls_ucode_img *img, *t; + unsigned long managed_falcons = acr->base.managed_falcons; + int managed_count = 0; + u32 image_wpr_size; + int falcon_id; + int ret; + + INIT_LIST_HEAD(&imgs); + + /* Load all LS blobs */ + for_each_set_bit(falcon_id, &managed_falcons, NVKM_SECBOOT_FALCON_END) { + struct ls_ucode_img *img; + + img = acr->func->ls_ucode_img_load(acr, falcon_id); + if (IS_ERR(img)) { + ret = PTR_ERR(img); + goto cleanup; + } + + list_add_tail(&img->node, &imgs); + managed_count++; + } + + /* + * Fill the WPR and LSF headers with the right offsets and compute + * required WPR size + */ + image_wpr_size = acr->func->ls_fill_headers(acr, &imgs); + image_wpr_size = ALIGN(image_wpr_size, WPR_ALIGNMENT); + + /* Allocate GPU object that will contain the WPR region */ + ret = nvkm_gpuobj_new(subdev->device, image_wpr_size, WPR_ALIGNMENT, + false, NULL, &acr->ls_blob); + if (ret) + goto cleanup; + + nvkm_debug(subdev, "%d managed LS falcons, WPR size is %d bytes\n", + managed_count, image_wpr_size); + + /* If WPR address and size are not fixed, set them to fit the LS blob */ + if (wpr_size == 0) { + wpr_addr = acr->ls_blob->addr; + wpr_size = image_wpr_size; + /* + * But if the WPR region is set by the bootloader, it is illegal for + * the HS blob to be larger than this region. + */ + } else if (image_wpr_size > wpr_size) { + nvkm_error(subdev, "WPR region too small for FW blob!\n"); + nvkm_error(subdev, "required: %dB\n", image_wpr_size); + nvkm_error(subdev, "available: %dB\n", wpr_size); + ret = -ENOSPC; + goto cleanup; + } + + /* Write LS blob */ + ret = acr->func->ls_write_wpr(acr, &imgs, acr->ls_blob, wpr_addr); + if (ret) + nvkm_gpuobj_del(&acr->ls_blob); + +cleanup: + list_for_each_entry_safe(img, t, &imgs, node) { + kfree(img->ucode_data); + kfree(img->sig); + kfree(img); + } + + return ret; +} + + + + +/** + * acr_r352_hsf_patch_signature() - patch HS blob with correct signature + */ +static void +acr_r352_hsf_patch_signature(struct nvkm_secboot *sb, void *acr_image) +{ + struct fw_bin_header *hsbin_hdr = acr_image; + struct hsf_fw_header *fw_hdr = acr_image + hsbin_hdr->header_offset; + void *hs_data = acr_image + hsbin_hdr->data_offset; + void *sig; + u32 sig_size; + + /* Falcon in debug or production mode? */ + if (sb->boot_falcon->debug) { + sig = acr_image + fw_hdr->sig_dbg_offset; + sig_size = fw_hdr->sig_dbg_size; + } else { + sig = acr_image + fw_hdr->sig_prod_offset; + sig_size = fw_hdr->sig_prod_size; + } + + /* Patch signature */ + memcpy(hs_data + fw_hdr->patch_loc, sig + fw_hdr->patch_sig, sig_size); +} + +static void +acr_r352_fixup_hs_desc(struct acr_r352 *acr, struct nvkm_secboot *sb, + struct hsflcn_acr_desc *desc) +{ + struct nvkm_gpuobj *ls_blob = acr->ls_blob; + + /* WPR region information if WPR is not fixed */ + if (sb->wpr_size == 0) { + u32 wpr_start = ls_blob->addr; + u32 wpr_end = wpr_start + ls_blob->size; + + desc->wpr_region_id = 1; + desc->regions.no_regions = 2; + desc->regions.region_props[0].start_addr = wpr_start >> 8; + desc->regions.region_props[0].end_addr = wpr_end >> 8; + desc->regions.region_props[0].region_id = 1; + desc->regions.region_props[0].read_mask = 0xf; + desc->regions.region_props[0].write_mask = 0xc; + desc->regions.region_props[0].client_mask = 0x2; + } else { + desc->ucode_blob_base = ls_blob->addr; + desc->ucode_blob_size = ls_blob->size; + } +} + +static void +acr_r352_generate_hs_bl_desc(const struct hsf_load_header *hdr, void *_bl_desc, + u64 offset) +{ + struct acr_r352_flcn_bl_desc *bl_desc = _bl_desc; + u64 addr_code, addr_data; + + addr_code = offset >> 8; + addr_data = (offset + hdr->data_dma_base) >> 8; + + bl_desc->ctx_dma = FALCON_DMAIDX_VIRT; + bl_desc->code_dma_base = lower_32_bits(addr_code); + bl_desc->non_sec_code_off = hdr->non_sec_code_off; + bl_desc->non_sec_code_size = hdr->non_sec_code_size; + bl_desc->sec_code_off = hdr->app[0].sec_code_off; + bl_desc->sec_code_size = hdr->app[0].sec_code_size; + bl_desc->code_entry_point = 0; + bl_desc->data_dma_base = lower_32_bits(addr_data); + bl_desc->data_size = hdr->data_size; +} + +/** + * acr_r352_prepare_hs_blob - load and prepare a HS blob and BL descriptor + * + * @sb secure boot instance to prepare for + * @fw name of the HS firmware to load + * @blob pointer to gpuobj that will be allocated to receive the HS FW payload + * @bl_desc pointer to the BL descriptor to write for this firmware + * @patch whether we should patch the HS descriptor (only for HS loaders) + */ +static int +acr_r352_prepare_hs_blob(struct acr_r352 *acr, struct nvkm_secboot *sb, + const char *fw, struct nvkm_gpuobj **blob, + struct hsf_load_header *load_header, bool patch) +{ + struct nvkm_subdev *subdev = &sb->subdev; + void *acr_image; + struct fw_bin_header *hsbin_hdr; + struct hsf_fw_header *fw_hdr; + struct hsf_load_header *load_hdr; + void *acr_data; + int ret; + + acr_image = nvkm_acr_load_firmware(subdev, fw, 0); + if (IS_ERR(acr_image)) + return PTR_ERR(acr_image); + + hsbin_hdr = acr_image; + fw_hdr = acr_image + hsbin_hdr->header_offset; + load_hdr = acr_image + fw_hdr->hdr_offset; + acr_data = acr_image + hsbin_hdr->data_offset; + + /* Patch signature */ + acr_r352_hsf_patch_signature(sb, acr_image); + + /* Patch descriptor with WPR information? */ + if (patch) { + struct hsflcn_acr_desc *desc; + + desc = acr_data + load_hdr->data_dma_base; + acr_r352_fixup_hs_desc(acr, sb, desc); + } + + if (load_hdr->num_apps > ACR_R352_MAX_APPS) { + nvkm_error(subdev, "more apps (%d) than supported (%d)!", + load_hdr->num_apps, ACR_R352_MAX_APPS); + ret = -EINVAL; + goto cleanup; + } + memcpy(load_header, load_hdr, sizeof(*load_header) + + (sizeof(load_hdr->app[0]) * load_hdr->num_apps)); + + /* Create ACR blob and copy HS data to it */ + ret = nvkm_gpuobj_new(subdev->device, ALIGN(hsbin_hdr->data_size, 256), + 0x1000, false, NULL, blob); + if (ret) + goto cleanup; + + nvkm_kmap(*blob); + nvkm_gpuobj_memcpy_to(*blob, 0, acr_data, hsbin_hdr->data_size); + nvkm_done(*blob); + +cleanup: + kfree(acr_image); + + return ret; +} + +static int +acr_r352_prepare_hsbl_blob(struct acr_r352 *acr) +{ + const struct nvkm_subdev *subdev = acr->base.subdev; + struct fw_bin_header *hdr; + struct fw_bl_desc *hsbl_desc; + + acr->hsbl_blob = nvkm_acr_load_firmware(subdev, "acr/bl", 0); + if (IS_ERR(acr->hsbl_blob)) { + int ret = PTR_ERR(acr->hsbl_blob); + + acr->hsbl_blob = NULL; + return ret; + } + + hdr = acr->hsbl_blob; + hsbl_desc = acr->hsbl_blob + hdr->header_offset; + + /* virtual start address for boot vector */ + acr->base.start_address = hsbl_desc->start_tag << 8; + + return 0; +} + +/** + * acr_r352_load_blobs - load blobs common to all ACR V1 versions. + * + * This includes the LS blob, HS ucode loading blob, and HS bootloader. + * + * The HS ucode unload blob is only used on dGPU if the WPR region is variable. + */ +int +acr_r352_load_blobs(struct acr_r352 *acr, struct nvkm_secboot *sb) +{ + int ret; + + /* Firmware already loaded? */ + if (acr->firmware_ok) + return 0; + + /* Load and prepare the managed falcon's firmwares */ + ret = acr_r352_prepare_ls_blob(acr, sb->wpr_addr, sb->wpr_size); + if (ret) + return ret; + + /* Load the HS firmware that will load the LS firmwares */ + if (!acr->load_blob) { + ret = acr_r352_prepare_hs_blob(acr, sb, "acr/ucode_load", + &acr->load_blob, + &acr->load_bl_header, true); + if (ret) + return ret; + } + + /* If the ACR region is dynamically programmed, we need an unload FW */ + if (sb->wpr_size == 0) { + ret = acr_r352_prepare_hs_blob(acr, sb, "acr/ucode_unload", + &acr->unload_blob, + &acr->unload_bl_header, false); + if (ret) + return ret; + } + + /* Load the HS firmware bootloader */ + if (!acr->hsbl_blob) { + ret = acr_r352_prepare_hsbl_blob(acr); + if (ret) + return ret; + } + + acr->firmware_ok = true; + nvkm_debug(&sb->subdev, "LS blob successfully created\n"); + + return 0; +} + +/** + * acr_r352_load() - prepare HS falcon to run the specified blob, mapped + * at GPU address offset. + */ +static int +acr_r352_load(struct nvkm_acr *_acr, struct nvkm_secboot *sb, + struct nvkm_gpuobj *blob, u64 offset) +{ + struct acr_r352 *acr = acr_r352(_acr); + struct nvkm_falcon *falcon = sb->boot_falcon; + struct fw_bin_header *hdr = acr->hsbl_blob; + struct fw_bl_desc *hsbl_desc = acr->hsbl_blob + hdr->header_offset; + void *blob_data = acr->hsbl_blob + hdr->data_offset; + void *hsbl_code = blob_data + hsbl_desc->code_off; + void *hsbl_data = blob_data + hsbl_desc->data_off; + u32 code_size = ALIGN(hsbl_desc->code_size, 256); + const struct hsf_load_header *load_hdr; + const u32 bl_desc_size = acr->func->hs_bl_desc_size; + u8 bl_desc[bl_desc_size]; + + /* Find the bootloader descriptor for our blob and copy it */ + if (blob == acr->load_blob) { + load_hdr = &acr->load_bl_header; + } else if (blob == acr->unload_blob) { + load_hdr = &acr->unload_bl_header; + } else { + nvkm_error(_acr->subdev, "invalid secure boot blob!\n"); + return -EINVAL; + } + + /* + * Copy HS bootloader data + */ + nvkm_falcon_load_dmem(falcon, hsbl_data, 0x0, hsbl_desc->data_size, 0); + + /* Copy HS bootloader code to end of IMEM */ + nvkm_falcon_load_imem(falcon, hsbl_code, falcon->code.limit - code_size, + code_size, hsbl_desc->start_tag, 0, false); + + /* Generate the BL header */ + memset(bl_desc, 0, bl_desc_size); + acr->func->generate_hs_bl_desc(load_hdr, bl_desc, offset); + + /* + * Copy HS BL header where the HS descriptor expects it to be + */ + nvkm_falcon_load_dmem(falcon, bl_desc, hsbl_desc->dmem_load_off, + bl_desc_size, 0); + + return 0; +} + +static int +acr_r352_shutdown(struct acr_r352 *acr, struct nvkm_secboot *sb) +{ + int i; + + /* Run the unload blob to unprotect the WPR region */ + if (acr->unload_blob && sb->wpr_set) { + int ret; + + nvkm_debug(&sb->subdev, "running HS unload blob\n"); + ret = sb->func->run_blob(sb, acr->unload_blob); + if (ret) + return ret; + nvkm_debug(&sb->subdev, "HS unload blob completed\n"); + } + + for (i = 0; i < NVKM_SECBOOT_FALCON_END; i++) + acr->falcon_state[i] = NON_SECURE; + + sb->wpr_set = false; + + return 0; +} + +static int +acr_r352_bootstrap(struct acr_r352 *acr, struct nvkm_secboot *sb) +{ + int ret; + + if (sb->wpr_set) + return 0; + + /* Make sure all blobs are ready */ + ret = acr_r352_load_blobs(acr, sb); + if (ret) + return ret; + + nvkm_debug(&sb->subdev, "running HS load blob\n"); + ret = sb->func->run_blob(sb, acr->load_blob); + /* clear halt interrupt */ + nvkm_falcon_clear_interrupt(sb->boot_falcon, 0x10); + if (ret) + return ret; + nvkm_debug(&sb->subdev, "HS load blob completed\n"); + + sb->wpr_set = true; + + return 0; +} + +/* + * acr_r352_reset() - execute secure boot from the prepared state + * + * Load the HS bootloader and ask the falcon to run it. This will in turn + * load the HS firmware and run it, so once the falcon stops all the managed + * falcons should have their LS firmware loaded and be ready to run. + */ +static int +acr_r352_reset(struct nvkm_acr *_acr, struct nvkm_secboot *sb, + enum nvkm_secboot_falcon falcon) +{ + struct acr_r352 *acr = acr_r352(_acr); + int ret; + + /* + * Dummy GM200 implementation: perform secure boot each time we are + * called on FECS. Since only FECS and GPCCS are managed and started + * together, this ought to be safe. + * + * Once we have proper PMU firmware and support, this will be changed + * to a proper call to the PMU method. + */ + if (falcon != NVKM_SECBOOT_FALCON_FECS) + goto end; + + ret = acr_r352_shutdown(acr, sb); + if (ret) + return ret; + + acr_r352_bootstrap(acr, sb); + if (ret) + return ret; + +end: + acr->falcon_state[falcon] = RESET; + return 0; +} + +static int +acr_r352_start(struct nvkm_acr *_acr, struct nvkm_secboot *sb, + enum nvkm_secboot_falcon falcon) +{ + struct acr_r352 *acr = acr_r352(_acr); + const struct nvkm_subdev *subdev = &sb->subdev; + int base; + + switch (falcon) { + case NVKM_SECBOOT_FALCON_FECS: + base = 0x409000; + break; + case NVKM_SECBOOT_FALCON_GPCCS: + base = 0x41a000; + break; + default: + nvkm_error(subdev, "cannot start unhandled falcon!\n"); + return -EINVAL; + } + + nvkm_wr32(subdev->device, base + 0x130, 0x00000002); + acr->falcon_state[falcon] = RUNNING; + + return 0; +} + +static int +acr_r352_fini(struct nvkm_acr *_acr, struct nvkm_secboot *sb, bool suspend) +{ + struct acr_r352 *acr = acr_r352(_acr); + + return acr_r352_shutdown(acr, sb); +} + +static void +acr_r352_dtor(struct nvkm_acr *_acr) +{ + struct acr_r352 *acr = acr_r352(_acr); + + nvkm_gpuobj_del(&acr->unload_blob); + + kfree(acr->hsbl_blob); + nvkm_gpuobj_del(&acr->load_blob); + nvkm_gpuobj_del(&acr->ls_blob); + + kfree(acr); +} + +const struct acr_r352_ls_func +acr_r352_ls_fecs_func = { + .load = acr_ls_ucode_load_fecs, + .generate_bl_desc = acr_r352_generate_flcn_bl_desc, + .bl_desc_size = sizeof(struct acr_r352_flcn_bl_desc), +}; + +const struct acr_r352_ls_func +acr_r352_ls_gpccs_func = { + .load = acr_ls_ucode_load_gpccs, + .generate_bl_desc = acr_r352_generate_flcn_bl_desc, + .bl_desc_size = sizeof(struct acr_r352_flcn_bl_desc), + /* GPCCS will be loaded using PRI */ + .lhdr_flags = LSF_FLAG_FORCE_PRIV_LOAD, +}; + +const struct acr_r352_func +acr_r352_func = { + .generate_hs_bl_desc = acr_r352_generate_hs_bl_desc, + .hs_bl_desc_size = sizeof(struct acr_r352_flcn_bl_desc), + .ls_ucode_img_load = acr_r352_ls_ucode_img_load, + .ls_fill_headers = acr_r352_ls_fill_headers, + .ls_write_wpr = acr_r352_ls_write_wpr, + .ls_func = { + [NVKM_SECBOOT_FALCON_FECS] = &acr_r352_ls_fecs_func, + [NVKM_SECBOOT_FALCON_GPCCS] = &acr_r352_ls_gpccs_func, + }, +}; + +static const struct nvkm_acr_func +acr_r352_base_func = { + .dtor = acr_r352_dtor, + .fini = acr_r352_fini, + .load = acr_r352_load, + .reset = acr_r352_reset, + .start = acr_r352_start, +}; + +struct nvkm_acr * +acr_r352_new_(const struct acr_r352_func *func, + enum nvkm_secboot_falcon boot_falcon, + unsigned long managed_falcons) +{ + struct acr_r352 *acr; + + acr = kzalloc(sizeof(*acr), GFP_KERNEL); + if (!acr) + return ERR_PTR(-ENOMEM); + + acr->base.boot_falcon = boot_falcon; + acr->base.managed_falcons = managed_falcons; + acr->base.func = &acr_r352_base_func; + acr->func = func; + + return &acr->base; +} + +struct nvkm_acr * +acr_r352_new(unsigned long managed_falcons) +{ + return acr_r352_new_(&acr_r352_func, NVKM_SECBOOT_FALCON_PMU, + managed_falcons); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h new file mode 100644 index 000000000000..ad5923b0fd3c --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h @@ -0,0 +1,250 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#ifndef __NVKM_SECBOOT_ACR_R352_H__ +#define __NVKM_SECBOOT_ACR_R352_H__ + +#include "acr.h" +#include "ls_ucode.h" + +struct ls_ucode_img; + +#define ACR_R352_MAX_APPS 8 + +/* + * + * LS blob structures + * + */ + +/** + * struct acr_r352_lsf_lsb_header - LS firmware header + * @signature: signature to verify the firmware against + * @ucode_off: offset of the ucode blob in the WPR region. The ucode + * blob contains the bootloader, code and data of the + * LS falcon + * @ucode_size: size of the ucode blob, including bootloader + * @data_size: size of the ucode blob data + * @bl_code_size: size of the bootloader code + * @bl_imem_off: offset in imem of the bootloader + * @bl_data_off: offset of the bootloader data in WPR region + * @bl_data_size: size of the bootloader data + * @app_code_off: offset of the app code relative to ucode_off + * @app_code_size: size of the app code + * @app_data_off: offset of the app data relative to ucode_off + * @app_data_size: size of the app data + * @flags: flags for the secure bootloader + * + * This structure is written into the WPR region for each managed falcon. Each + * instance is referenced by the lsb_offset member of the corresponding + * lsf_wpr_header. + */ +struct acr_r352_lsf_lsb_header { + /** + * LS falcon signatures + * @prd_keys: signature to use in production mode + * @dgb_keys: signature to use in debug mode + * @b_prd_present: whether the production key is present + * @b_dgb_present: whether the debug key is present + * @falcon_id: ID of the falcon the ucode applies to + */ + struct { + u8 prd_keys[2][16]; + u8 dbg_keys[2][16]; + u32 b_prd_present; + u32 b_dbg_present; + u32 falcon_id; + } signature; + u32 ucode_off; + u32 ucode_size; + u32 data_size; + u32 bl_code_size; + u32 bl_imem_off; + u32 bl_data_off; + u32 bl_data_size; + u32 app_code_off; + u32 app_code_size; + u32 app_data_off; + u32 app_data_size; + u32 flags; +#define LSF_FLAG_LOAD_CODE_AT_0 1 +#define LSF_FLAG_DMACTL_REQ_CTX 4 +#define LSF_FLAG_FORCE_PRIV_LOAD 8 +}; + +/** + * struct acr_r352_lsf_wpr_header - LS blob WPR Header + * @falcon_id: LS falcon ID + * @lsb_offset: offset of the lsb_lsf_header in the WPR region + * @bootstrap_owner: secure falcon reponsible for bootstrapping the LS falcon + * @lazy_bootstrap: skip bootstrapping by ACR + * @status: bootstrapping status + * + * An array of these is written at the beginning of the WPR region, one for + * each managed falcon. The array is terminated by an instance which falcon_id + * is LSF_FALCON_ID_INVALID. + */ +struct acr_r352_lsf_wpr_header { + u32 falcon_id; + u32 lsb_offset; + u32 bootstrap_owner; + u32 lazy_bootstrap; + u32 status; +#define LSF_IMAGE_STATUS_NONE 0 +#define LSF_IMAGE_STATUS_COPY 1 +#define LSF_IMAGE_STATUS_VALIDATION_CODE_FAILED 2 +#define LSF_IMAGE_STATUS_VALIDATION_DATA_FAILED 3 +#define LSF_IMAGE_STATUS_VALIDATION_DONE 4 +#define LSF_IMAGE_STATUS_VALIDATION_SKIPPED 5 +#define LSF_IMAGE_STATUS_BOOTSTRAP_READY 6 +}; + +/** + * struct ls_ucode_img_r352 - ucode image augmented with r352 headers + */ +struct ls_ucode_img_r352 { + struct ls_ucode_img base; + + struct acr_r352_lsf_wpr_header wpr_header; + struct acr_r352_lsf_lsb_header lsb_header; +}; +#define ls_ucode_img_r352(i) container_of(i, struct ls_ucode_img_r352, base) + + +/* + * HS blob structures + */ + +struct hsf_load_header_app { + u32 sec_code_off; + u32 sec_code_size; +}; + +/** + * struct hsf_load_header - HS firmware load header + */ +struct hsf_load_header { + u32 non_sec_code_off; + u32 non_sec_code_size; + u32 data_dma_base; + u32 data_size; + u32 num_apps; + struct hsf_load_header_app app[0]; +}; + +/** + * struct acr_r352_ls_func - manages a single LS firmware + * + * @load: load the external firmware into a ls_ucode_img + * @generate_bl_desc: function called on a block of bl_desc_size to generate the + * proper bootloader descriptor for this LS firmware + * @bl_desc_size: size of the bootloader descriptor + * @lhdr_flags: LS flags + */ +struct acr_r352_ls_func { + int (*load)(const struct nvkm_subdev *, struct ls_ucode_img *); + void (*generate_bl_desc)(const struct nvkm_acr *, + const struct ls_ucode_img *, u64, void *); + u32 bl_desc_size; + u32 lhdr_flags; +}; + +struct acr_r352; + +/** + * struct acr_r352_func - manages nuances between ACR versions + * + * @generate_hs_bl_desc: function called on a block of bl_desc_size to generate + * the proper HS bootloader descriptor + * @hs_bl_desc_size: size of the HS bootloader descriptor + */ +struct acr_r352_func { + void (*generate_hs_bl_desc)(const struct hsf_load_header *, void *, + u64); + u32 hs_bl_desc_size; + + struct ls_ucode_img *(*ls_ucode_img_load)(const struct acr_r352 *, + enum nvkm_secboot_falcon); + int (*ls_fill_headers)(struct acr_r352 *, struct list_head *); + int (*ls_write_wpr)(struct acr_r352 *, struct list_head *, + struct nvkm_gpuobj *, u32); + + const struct acr_r352_ls_func *ls_func[NVKM_SECBOOT_FALCON_END]; +}; + +/** + * struct acr_r352 - ACR data for driver release 352 (and beyond) + */ +struct acr_r352 { + struct nvkm_acr base; + const struct acr_r352_func *func; + + /* + * HS FW - lock WPR region (dGPU only) and load LS FWs + * on Tegra the HS FW copies the LS blob into the fixed WPR instead + */ + struct nvkm_gpuobj *load_blob; + struct { + struct hsf_load_header load_bl_header; + struct hsf_load_header_app __load_apps[ACR_R352_MAX_APPS]; + }; + + /* HS FW - unlock WPR region (dGPU only) */ + struct nvkm_gpuobj *unload_blob; + struct { + struct hsf_load_header unload_bl_header; + struct hsf_load_header_app __unload_apps[ACR_R352_MAX_APPS]; + }; + + /* HS bootloader */ + void *hsbl_blob; + + /* LS FWs, to be loaded by the HS ACR */ + struct nvkm_gpuobj *ls_blob; + + /* Firmware already loaded? */ + bool firmware_ok; + + /* Falcons to lazy-bootstrap */ + u32 lazy_bootstrap; + + /* To keep track of the state of all managed falcons */ + enum { + /* In non-secure state, no firmware loaded, no privileges*/ + NON_SECURE = 0, + /* In low-secure mode and ready to be started */ + RESET, + /* In low-secure mode and running */ + RUNNING, + } falcon_state[NVKM_SECBOOT_FALCON_END]; +}; +#define acr_r352(acr) container_of(acr, struct acr_r352, base) + +struct nvkm_acr *acr_r352_new_(const struct acr_r352_func *, + enum nvkm_secboot_falcon, unsigned long); + +struct ls_ucode_img *acr_r352_ls_ucode_img_load(const struct acr_r352 *, + enum nvkm_secboot_falcon); +int acr_r352_ls_fill_headers(struct acr_r352 *, struct list_head *); +int acr_r352_ls_write_wpr(struct acr_r352 *, struct list_head *, + struct nvkm_gpuobj *, u32); + +#endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r361.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r361.c new file mode 100644 index 000000000000..f0aff1d98474 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r361.c @@ -0,0 +1,138 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#include "acr_r352.h" + +#include + +/** + * struct acr_r361_flcn_bl_desc - DMEM bootloader descriptor + * @signature: 16B signature for secure code. 0s if no secure code + * @ctx_dma: DMA context to be used by BL while loading code/data + * @code_dma_base: 256B-aligned Physical FB Address where code is located + * (falcon's $xcbase register) + * @non_sec_code_off: offset from code_dma_base where the non-secure code is + * located. The offset must be multiple of 256 to help perf + * @non_sec_code_size: the size of the nonSecure code part. + * @sec_code_off: offset from code_dma_base where the secure code is + * located. The offset must be multiple of 256 to help perf + * @sec_code_size: offset from code_dma_base where the secure code is + * located. The offset must be multiple of 256 to help perf + * @code_entry_point: code entry point which will be invoked by BL after + * code is loaded. + * @data_dma_base: 256B aligned Physical FB Address where data is located. + * (falcon's $xdbase register) + * @data_size: size of data block. Should be multiple of 256B + * + * Structure used by the bootloader to load the rest of the code. This has + * to be filled by host and copied into DMEM at offset provided in the + * hsflcn_bl_desc.bl_desc_dmem_load_off. + */ +struct acr_r361_flcn_bl_desc { + u32 reserved[4]; + u32 signature[4]; + u32 ctx_dma; + struct flcn_u64 code_dma_base; + u32 non_sec_code_off; + u32 non_sec_code_size; + u32 sec_code_off; + u32 sec_code_size; + u32 code_entry_point; + struct flcn_u64 data_dma_base; + u32 data_size; +}; + +static void +acr_r361_generate_flcn_bl_desc(const struct nvkm_acr *acr, + const struct ls_ucode_img *_img, u64 wpr_addr, + void *_desc) +{ + struct ls_ucode_img_r352 *img = ls_ucode_img_r352(_img); + struct acr_r361_flcn_bl_desc *desc = _desc; + const struct ls_ucode_img_desc *pdesc = &img->base.ucode_desc; + u64 base, addr_code, addr_data; + + base = wpr_addr + img->lsb_header.ucode_off + pdesc->app_start_offset; + addr_code = base + pdesc->app_resident_code_offset; + addr_data = base + pdesc->app_resident_data_offset; + + desc->ctx_dma = FALCON_DMAIDX_UCODE; + desc->code_dma_base = u64_to_flcn64(addr_code); + desc->non_sec_code_off = pdesc->app_resident_code_offset; + desc->non_sec_code_size = pdesc->app_resident_code_size; + desc->code_entry_point = pdesc->app_imem_entry; + desc->data_dma_base = u64_to_flcn64(addr_data); + desc->data_size = pdesc->app_resident_data_size; +} + +static void +acr_r361_generate_hs_bl_desc(const struct hsf_load_header *hdr, void *_bl_desc, + u64 offset) +{ + struct acr_r361_flcn_bl_desc *bl_desc = _bl_desc; + + bl_desc->ctx_dma = FALCON_DMAIDX_VIRT; + bl_desc->code_dma_base = u64_to_flcn64(offset); + bl_desc->non_sec_code_off = hdr->non_sec_code_off; + bl_desc->non_sec_code_size = hdr->non_sec_code_size; + bl_desc->sec_code_off = hdr->app[0].sec_code_off; + bl_desc->sec_code_size = hdr->app[0].sec_code_size; + bl_desc->code_entry_point = 0; + bl_desc->data_dma_base = u64_to_flcn64(offset + hdr->data_dma_base); + bl_desc->data_size = hdr->data_size; +} + +const struct acr_r352_ls_func +acr_r361_ls_fecs_func = { + .load = acr_ls_ucode_load_fecs, + .generate_bl_desc = acr_r361_generate_flcn_bl_desc, + .bl_desc_size = sizeof(struct acr_r361_flcn_bl_desc), +}; + +const struct acr_r352_ls_func +acr_r361_ls_gpccs_func = { + .load = acr_ls_ucode_load_gpccs, + .generate_bl_desc = acr_r361_generate_flcn_bl_desc, + .bl_desc_size = sizeof(struct acr_r361_flcn_bl_desc), + /* GPCCS will be loaded using PRI */ + .lhdr_flags = LSF_FLAG_FORCE_PRIV_LOAD, +}; + +const struct acr_r352_func +acr_r361_func = { + .generate_hs_bl_desc = acr_r361_generate_hs_bl_desc, + .hs_bl_desc_size = sizeof(struct acr_r361_flcn_bl_desc), + .ls_ucode_img_load = acr_r352_ls_ucode_img_load, + .ls_fill_headers = acr_r352_ls_fill_headers, + .ls_write_wpr = acr_r352_ls_write_wpr, + .ls_func = { + [NVKM_SECBOOT_FALCON_FECS] = &acr_r361_ls_fecs_func, + [NVKM_SECBOOT_FALCON_GPCCS] = &acr_r361_ls_gpccs_func, + }, +}; + +struct nvkm_acr * +acr_r361_new(unsigned long managed_falcons) +{ + return acr_r352_new_(&acr_r361_func, NVKM_SECBOOT_FALCON_PMU, + managed_falcons); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c index 314be2192b7d..27c9dfffb9a6 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c @@ -19,184 +19,108 @@ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER * DEALINGS IN THE SOFTWARE. */ + +/* + * Secure boot is the process by which NVIDIA-signed firmware is loaded into + * some of the falcons of a GPU. For production devices this is the only way + * for the firmware to access useful (but sensitive) registers. + * + * A Falcon microprocessor supporting advanced security modes can run in one of + * three modes: + * + * - Non-secure (NS). In this mode, functionality is similar to Falcon + * architectures before security modes were introduced (pre-Maxwell), but + * capability is restricted. In particular, certain registers may be + * inaccessible for reads and/or writes, and physical memory access may be + * disabled (on certain Falcon instances). This is the only possible mode that + * can be used if you don't have microcode cryptographically signed by NVIDIA. + * + * - Heavy Secure (HS). In this mode, the microprocessor is a black box - it's + * not possible to read or write any Falcon internal state or Falcon registers + * from outside the Falcon (for example, from the host system). The only way + * to enable this mode is by loading microcode that has been signed by NVIDIA. + * (The loading process involves tagging the IMEM block as secure, writing the + * signature into a Falcon register, and starting execution. The hardware will + * validate the signature, and if valid, grant HS privileges.) + * + * - Light Secure (LS). In this mode, the microprocessor has more privileges + * than NS but fewer than HS. Some of the microprocessor state is visible to + * host software to ease debugging. The only way to enable this mode is by HS + * microcode enabling LS mode. Some privileges available to HS mode are not + * available here. LS mode is introduced in GM20x. + * + * Secure boot consists in temporarily switching a HS-capable falcon (typically + * PMU) into HS mode in order to validate the LS firmwares of managed falcons, + * load them, and switch managed falcons into LS mode. Once secure boot + * completes, no falcon remains in HS mode. + * + * Secure boot requires a write-protected memory region (WPR) which can only be + * written by the secure falcon. On dGPU, the driver sets up the WPR region in + * video memory. On Tegra, it is set up by the bootloader and its location and + * size written into memory controller registers. + * + * The secure boot process takes place as follows: + * + * 1) A LS blob is constructed that contains all the LS firmwares we want to + * load, along with their signatures and bootloaders. + * + * 2) A HS blob (also called ACR) is created that contains the signed HS + * firmware in charge of loading the LS firmwares into their respective + * falcons. + * + * 3) The HS blob is loaded (via its own bootloader) and executed on the + * HS-capable falcon. It authenticates itself, switches the secure falcon to + * HS mode and setup the WPR region around the LS blob (dGPU) or copies the + * LS blob into the WPR region (Tegra). + * + * 4) The LS blob is now secure from all external tampering. The HS falcon + * checks the signatures of the LS firmwares and, if valid, switches the + * managed falcons to LS mode and makes them ready to run the LS firmware. + * + * 5) The managed falcons remain in LS mode and can be started. + * + */ + #include "priv.h" +#include "acr.h" #include #include +#include -static const char * -managed_falcons_names[] = { +const char * +nvkm_secboot_falcon_name[] = { [NVKM_SECBOOT_FALCON_PMU] = "PMU", [NVKM_SECBOOT_FALCON_RESERVED] = "", [NVKM_SECBOOT_FALCON_FECS] = "FECS", [NVKM_SECBOOT_FALCON_GPCCS] = "GPCCS", [NVKM_SECBOOT_FALCON_END] = "", }; - -/* - * Helper falcon functions - */ - -static int -falcon_clear_halt_interrupt(struct nvkm_device *device, u32 base) -{ - int ret; - - /* clear halt interrupt */ - nvkm_mask(device, base + 0x004, 0x10, 0x10); - /* wait until halt interrupt is cleared */ - ret = nvkm_wait_msec(device, 10, base + 0x008, 0x10, 0x0); - if (ret < 0) - return ret; - - return 0; -} - -static int -falcon_wait_idle(struct nvkm_device *device, u32 base) -{ - int ret; - - ret = nvkm_wait_msec(device, 10, base + 0x04c, 0xffff, 0x0); - if (ret < 0) - return ret; - - return 0; -} - -static int -nvkm_secboot_falcon_enable(struct nvkm_secboot *sb) -{ - struct nvkm_device *device = sb->subdev.device; - int ret; - - /* enable engine */ - nvkm_mc_enable(device, sb->devidx); - ret = nvkm_wait_msec(device, 10, sb->base + 0x10c, 0x6, 0x0); - if (ret < 0) { - nvkm_error(&sb->subdev, "Falcon mem scrubbing timeout\n"); - nvkm_mc_disable(device, sb->devidx); - return ret; - } - - ret = falcon_wait_idle(device, sb->base); - if (ret) - return ret; - - /* enable IRQs */ - nvkm_wr32(device, sb->base + 0x010, 0xff); - nvkm_mc_intr_mask(device, sb->devidx, true); - - return 0; -} - -static int -nvkm_secboot_falcon_disable(struct nvkm_secboot *sb) -{ - struct nvkm_device *device = sb->subdev.device; - - /* disable IRQs and wait for any previous code to complete */ - nvkm_mc_intr_mask(device, sb->devidx, false); - nvkm_wr32(device, sb->base + 0x014, 0xff); - - falcon_wait_idle(device, sb->base); - - /* disable engine */ - nvkm_mc_disable(device, sb->devidx); - - return 0; -} - -int -nvkm_secboot_falcon_reset(struct nvkm_secboot *sb) -{ - int ret; - - ret = nvkm_secboot_falcon_disable(sb); - if (ret) - return ret; - - ret = nvkm_secboot_falcon_enable(sb); - if (ret) - return ret; - - return 0; -} - -/** - * nvkm_secboot_falcon_run - run the falcon that will perform secure boot - * - * This function is to be called after all chip-specific preparations have - * been completed. It will start the falcon to perform secure boot, wait for - * it to halt, and report if an error occurred. - */ -int -nvkm_secboot_falcon_run(struct nvkm_secboot *sb) -{ - struct nvkm_device *device = sb->subdev.device; - int ret; - - /* Start falcon */ - nvkm_wr32(device, sb->base + 0x100, 0x2); - - /* Wait for falcon halt */ - ret = nvkm_wait_msec(device, 100, sb->base + 0x100, 0x10, 0x10); - if (ret < 0) - return ret; - - /* If mailbox register contains an error code, then ACR has failed */ - ret = nvkm_rd32(device, sb->base + 0x040); - if (ret) { - nvkm_error(&sb->subdev, "ACR boot failed, ret 0x%08x", ret); - falcon_clear_halt_interrupt(device, sb->base); - return -EINVAL; - } - - return 0; -} - - /** * nvkm_secboot_reset() - reset specified falcon */ int -nvkm_secboot_reset(struct nvkm_secboot *sb, u32 falcon) +nvkm_secboot_reset(struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon) { /* Unmanaged falcon? */ - if (!(BIT(falcon) & sb->func->managed_falcons)) { + if (!(BIT(falcon) & sb->acr->managed_falcons)) { nvkm_error(&sb->subdev, "cannot reset unmanaged falcon!\n"); return -EINVAL; } - return sb->func->reset(sb, falcon); -} - -/** - * nvkm_secboot_start() - start specified falcon - */ -int -nvkm_secboot_start(struct nvkm_secboot *sb, u32 falcon) -{ - /* Unmanaged falcon? */ - if (!(BIT(falcon) & sb->func->managed_falcons)) { - nvkm_error(&sb->subdev, "cannot start unmanaged falcon!\n"); - return -EINVAL; - } - - return sb->func->start(sb, falcon); + return sb->acr->func->reset(sb->acr, sb, falcon); } /** * nvkm_secboot_is_managed() - check whether a given falcon is securely-managed */ bool -nvkm_secboot_is_managed(struct nvkm_secboot *secboot, - enum nvkm_secboot_falcon fid) +nvkm_secboot_is_managed(struct nvkm_secboot *sb, enum nvkm_secboot_falcon fid) { - if (!secboot) + if (!sb) return false; - return secboot->func->managed_falcons & BIT(fid); + return sb->acr->managed_falcons & BIT(fid); } static int @@ -205,9 +129,19 @@ nvkm_secboot_oneinit(struct nvkm_subdev *subdev) struct nvkm_secboot *sb = nvkm_secboot(subdev); int ret = 0; + switch (sb->acr->boot_falcon) { + case NVKM_SECBOOT_FALCON_PMU: + sb->boot_falcon = subdev->device->pmu->falcon; + break; + default: + nvkm_error(subdev, "Unmanaged boot falcon %s!\n", + nvkm_secboot_falcon_name[sb->acr->boot_falcon]); + return -EINVAL; + } + /* Call chip-specific init function */ - if (sb->func->init) - ret = sb->func->init(sb); + if (sb->func->oneinit) + ret = sb->func->oneinit(sb); if (ret) { nvkm_error(subdev, "Secure Boot initialization failed: %d\n", ret); @@ -249,7 +183,7 @@ nvkm_secboot = { }; int -nvkm_secboot_ctor(const struct nvkm_secboot_func *func, +nvkm_secboot_ctor(const struct nvkm_secboot_func *func, struct nvkm_acr *acr, struct nvkm_device *device, int index, struct nvkm_secboot *sb) { @@ -257,22 +191,14 @@ nvkm_secboot_ctor(const struct nvkm_secboot_func *func, nvkm_subdev_ctor(&nvkm_secboot, device, index, &sb->subdev); sb->func = func; - - /* setup the performing falcon's base address and masks */ - switch (func->boot_falcon) { - case NVKM_SECBOOT_FALCON_PMU: - sb->devidx = NVKM_SUBDEV_PMU; - sb->base = 0x10a000; - break; - default: - nvkm_error(&sb->subdev, "invalid secure boot falcon\n"); - return -EINVAL; - }; + sb->acr = acr; + acr->subdev = &sb->subdev; nvkm_debug(&sb->subdev, "securely managed falcons:\n"); - for_each_set_bit(fid, &sb->func->managed_falcons, + for_each_set_bit(fid, &sb->acr->managed_falcons, NVKM_SECBOOT_FALCON_END) - nvkm_debug(&sb->subdev, "- %s\n", managed_falcons_names[fid]); + nvkm_debug(&sb->subdev, "- %s\n", + nvkm_secboot_falcon_name[fid]); return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c index ec48e4ace37a..813c4eb0b25f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c @@ -20,1313 +20,84 @@ * DEALINGS IN THE SOFTWARE. */ -/* - * Secure boot is the process by which NVIDIA-signed firmware is loaded into - * some of the falcons of a GPU. For production devices this is the only way - * for the firmware to access useful (but sensitive) registers. - * - * A Falcon microprocessor supporting advanced security modes can run in one of - * three modes: - * - * - Non-secure (NS). In this mode, functionality is similar to Falcon - * architectures before security modes were introduced (pre-Maxwell), but - * capability is restricted. In particular, certain registers may be - * inaccessible for reads and/or writes, and physical memory access may be - * disabled (on certain Falcon instances). This is the only possible mode that - * can be used if you don't have microcode cryptographically signed by NVIDIA. - * - * - Heavy Secure (HS). In this mode, the microprocessor is a black box - it's - * not possible to read or write any Falcon internal state or Falcon registers - * from outside the Falcon (for example, from the host system). The only way - * to enable this mode is by loading microcode that has been signed by NVIDIA. - * (The loading process involves tagging the IMEM block as secure, writing the - * signature into a Falcon register, and starting execution. The hardware will - * validate the signature, and if valid, grant HS privileges.) - * - * - Light Secure (LS). In this mode, the microprocessor has more privileges - * than NS but fewer than HS. Some of the microprocessor state is visible to - * host software to ease debugging. The only way to enable this mode is by HS - * microcode enabling LS mode. Some privileges available to HS mode are not - * available here. LS mode is introduced in GM20x. - * - * Secure boot consists in temporarily switching a HS-capable falcon (typically - * PMU) into HS mode in order to validate the LS firmwares of managed falcons, - * load them, and switch managed falcons into LS mode. Once secure boot - * completes, no falcon remains in HS mode. - * - * Secure boot requires a write-protected memory region (WPR) which can only be - * written by the secure falcon. On dGPU, the driver sets up the WPR region in - * video memory. On Tegra, it is set up by the bootloader and its location and - * size written into memory controller registers. - * - * The secure boot process takes place as follows: - * - * 1) A LS blob is constructed that contains all the LS firmwares we want to - * load, along with their signatures and bootloaders. - * - * 2) A HS blob (also called ACR) is created that contains the signed HS - * firmware in charge of loading the LS firmwares into their respective - * falcons. - * - * 3) The HS blob is loaded (via its own bootloader) and executed on the - * HS-capable falcon. It authenticates itself, switches the secure falcon to - * HS mode and setup the WPR region around the LS blob (dGPU) or copies the - * LS blob into the WPR region (Tegra). - * - * 4) The LS blob is now secure from all external tampering. The HS falcon - * checks the signatures of the LS firmwares and, if valid, switches the - * managed falcons to LS mode and makes them ready to run the LS firmware. - * - * 5) The managed falcons remain in LS mode and can be started. - * - */ -#include "priv.h" +#include "acr.h" +#include "gm200.h" #include -#include #include - -enum { - FALCON_DMAIDX_UCODE = 0, - FALCON_DMAIDX_VIRT = 1, - FALCON_DMAIDX_PHYS_VID = 2, - FALCON_DMAIDX_PHYS_SYS_COH = 3, - FALCON_DMAIDX_PHYS_SYS_NCOH = 4, -}; +#include +#include /** - * struct fw_bin_header - header of firmware files - * @bin_magic: always 0x3b1d14f0 - * @bin_ver: version of the bin format - * @bin_size: entire image size including this header - * @header_offset: offset of the firmware/bootloader header in the file - * @data_offset: offset of the firmware/bootloader payload in the file - * @data_size: size of the payload + * gm200_secboot_run_blob() - run the given high-secure blob * - * This header is located at the beginning of the HS firmware and HS bootloader - * files, to describe where the headers and data can be found. - */ -struct fw_bin_header { - u32 bin_magic; - u32 bin_ver; - u32 bin_size; - u32 header_offset; - u32 data_offset; - u32 data_size; -}; - -/** - * struct fw_bl_desc - firmware bootloader descriptor - * @start_tag: starting tag of bootloader - * @desc_dmem_load_off: DMEM offset of flcn_bl_dmem_desc - * @code_off: offset of code section - * @code_size: size of code section - * @data_off: offset of data section - * @data_size: size of data section - * - * This structure is embedded in bootloader firmware files at to describe the - * IMEM and DMEM layout expected by the bootloader. - */ -struct fw_bl_desc { - u32 start_tag; - u32 dmem_load_off; - u32 code_off; - u32 code_size; - u32 data_off; - u32 data_size; -}; - - -/* - * - * LS blob structures - * - */ - -/** - * struct lsf_ucode_desc - LS falcon signatures - * @prd_keys: signature to use when the GPU is in production mode - * @dgb_keys: signature to use when the GPU is in debug mode - * @b_prd_present: whether the production key is present - * @b_dgb_present: whether the debug key is present - * @falcon_id: ID of the falcon the ucode applies to - * - * Directly loaded from a signature file. - */ -struct lsf_ucode_desc { - u8 prd_keys[2][16]; - u8 dbg_keys[2][16]; - u32 b_prd_present; - u32 b_dbg_present; - u32 falcon_id; -}; - -/** - * struct lsf_lsb_header - LS firmware header - * @signature: signature to verify the firmware against - * @ucode_off: offset of the ucode blob in the WPR region. The ucode - * blob contains the bootloader, code and data of the - * LS falcon - * @ucode_size: size of the ucode blob, including bootloader - * @data_size: size of the ucode blob data - * @bl_code_size: size of the bootloader code - * @bl_imem_off: offset in imem of the bootloader - * @bl_data_off: offset of the bootloader data in WPR region - * @bl_data_size: size of the bootloader data - * @app_code_off: offset of the app code relative to ucode_off - * @app_code_size: size of the app code - * @app_data_off: offset of the app data relative to ucode_off - * @app_data_size: size of the app data - * @flags: flags for the secure bootloader - * - * This structure is written into the WPR region for each managed falcon. Each - * instance is referenced by the lsb_offset member of the corresponding - * lsf_wpr_header. - */ -struct lsf_lsb_header { - struct lsf_ucode_desc signature; - u32 ucode_off; - u32 ucode_size; - u32 data_size; - u32 bl_code_size; - u32 bl_imem_off; - u32 bl_data_off; - u32 bl_data_size; - u32 app_code_off; - u32 app_code_size; - u32 app_data_off; - u32 app_data_size; - u32 flags; -#define LSF_FLAG_LOAD_CODE_AT_0 1 -#define LSF_FLAG_DMACTL_REQ_CTX 4 -#define LSF_FLAG_FORCE_PRIV_LOAD 8 -}; - -/** - * struct lsf_wpr_header - LS blob WPR Header - * @falcon_id: LS falcon ID - * @lsb_offset: offset of the lsb_lsf_header in the WPR region - * @bootstrap_owner: secure falcon reponsible for bootstrapping the LS falcon - * @lazy_bootstrap: skip bootstrapping by ACR - * @status: bootstrapping status - * - * An array of these is written at the beginning of the WPR region, one for - * each managed falcon. The array is terminated by an instance which falcon_id - * is LSF_FALCON_ID_INVALID. - */ -struct lsf_wpr_header { - u32 falcon_id; - u32 lsb_offset; - u32 bootstrap_owner; - u32 lazy_bootstrap; - u32 status; -#define LSF_IMAGE_STATUS_NONE 0 -#define LSF_IMAGE_STATUS_COPY 1 -#define LSF_IMAGE_STATUS_VALIDATION_CODE_FAILED 2 -#define LSF_IMAGE_STATUS_VALIDATION_DATA_FAILED 3 -#define LSF_IMAGE_STATUS_VALIDATION_DONE 4 -#define LSF_IMAGE_STATUS_VALIDATION_SKIPPED 5 -#define LSF_IMAGE_STATUS_BOOTSTRAP_READY 6 -}; - - -/** - * struct ls_ucode_img_desc - descriptor of firmware image - * @descriptor_size: size of this descriptor - * @image_size: size of the whole image - * @bootloader_start_offset: start offset of the bootloader in ucode image - * @bootloader_size: size of the bootloader - * @bootloader_imem_offset: start off set of the bootloader in IMEM - * @bootloader_entry_point: entry point of the bootloader in IMEM - * @app_start_offset: start offset of the LS firmware - * @app_size: size of the LS firmware's code and data - * @app_imem_offset: offset of the app in IMEM - * @app_imem_entry: entry point of the app in IMEM - * @app_dmem_offset: offset of the data in DMEM - * @app_resident_code_offset: offset of app code from app_start_offset - * @app_resident_code_size: size of the code - * @app_resident_data_offset: offset of data from app_start_offset - * @app_resident_data_size: size of data - * - * A firmware image contains the code, data, and bootloader of a given LS - * falcon in a single blob. This structure describes where everything is. - * - * This can be generated from a (bootloader, code, data) set if they have - * been loaded separately, or come directly from a file. - */ -struct ls_ucode_img_desc { - u32 descriptor_size; - u32 image_size; - u32 tools_version; - u32 app_version; - char date[64]; - u32 bootloader_start_offset; - u32 bootloader_size; - u32 bootloader_imem_offset; - u32 bootloader_entry_point; - u32 app_start_offset; - u32 app_size; - u32 app_imem_offset; - u32 app_imem_entry; - u32 app_dmem_offset; - u32 app_resident_code_offset; - u32 app_resident_code_size; - u32 app_resident_data_offset; - u32 app_resident_data_size; - u32 nb_overlays; - struct {u32 start; u32 size; } load_ovl[64]; - u32 compressed; -}; - -/** - * struct ls_ucode_img - temporary storage for loaded LS firmwares - * @node: to link within lsf_ucode_mgr - * @falcon_id: ID of the falcon this LS firmware is for - * @ucode_desc: loaded or generated map of ucode_data - * @ucode_header: header of the firmware - * @ucode_data: firmware payload (code and data) - * @ucode_size: size in bytes of data in ucode_data - * @wpr_header: WPR header to be written to the LS blob - * @lsb_header: LSB header to be written to the LS blob - * - * Preparing the WPR LS blob requires information about all the LS firmwares - * (size, etc) to be known. This structure contains all the data of one LS - * firmware. - */ -struct ls_ucode_img { - struct list_head node; - enum nvkm_secboot_falcon falcon_id; - - struct ls_ucode_img_desc ucode_desc; - u32 *ucode_header; - u8 *ucode_data; - u32 ucode_size; - - struct lsf_wpr_header wpr_header; - struct lsf_lsb_header lsb_header; -}; - -/** - * struct ls_ucode_mgr - manager for all LS falcon firmwares - * @count: number of managed LS falcons - * @wpr_size: size of the required WPR region in bytes - * @img_list: linked list of lsf_ucode_img - */ -struct ls_ucode_mgr { - u16 count; - u32 wpr_size; - struct list_head img_list; -}; - - -/* - * - * HS blob structures - * - */ - -/** - * struct hsf_fw_header - HS firmware descriptor - * @sig_dbg_offset: offset of the debug signature - * @sig_dbg_size: size of the debug signature - * @sig_prod_offset: offset of the production signature - * @sig_prod_size: size of the production signature - * @patch_loc: offset of the offset (sic) of where the signature is - * @patch_sig: offset of the offset (sic) to add to sig_*_offset - * @hdr_offset: offset of the load header (see struct hs_load_header) - * @hdr_size: size of above header - * - * This structure is embedded in the HS firmware image at - * hs_bin_hdr.header_offset. - */ -struct hsf_fw_header { - u32 sig_dbg_offset; - u32 sig_dbg_size; - u32 sig_prod_offset; - u32 sig_prod_size; - u32 patch_loc; - u32 patch_sig; - u32 hdr_offset; - u32 hdr_size; -}; - -/** - * struct hsf_load_header - HS firmware load header - */ -struct hsf_load_header { - u32 non_sec_code_off; - u32 non_sec_code_size; - u32 data_dma_base; - u32 data_size; - u32 num_apps; - struct { - u32 sec_code_off; - u32 sec_code_size; - } app[0]; -}; - -/** - * Convenience function to duplicate a firmware file in memory and check that - * it has the required minimum size. - */ -static void * -gm200_secboot_load_firmware(struct nvkm_subdev *subdev, const char *name, - size_t min_size) -{ - const struct firmware *fw; - void *blob; - int ret; - - ret = nvkm_firmware_get(subdev->device, name, &fw); - if (ret) - return ERR_PTR(ret); - if (fw->size < min_size) { - nvkm_error(subdev, "%s is smaller than expected size %zu\n", - name, min_size); - nvkm_firmware_put(fw); - return ERR_PTR(-EINVAL); - } - blob = kmemdup(fw->data, fw->size, GFP_KERNEL); - nvkm_firmware_put(fw); - if (!blob) - return ERR_PTR(-ENOMEM); - - return blob; -} - - -/* - * Low-secure blob creation - */ - -#define BL_DESC_BLK_SIZE 256 -/** - * Build a ucode image and descriptor from provided bootloader, code and data. - * - * @bl: bootloader image, including 16-bytes descriptor - * @code: LS firmware code segment - * @data: LS firmware data segment - * @desc: ucode descriptor to be written - * - * Return: allocated ucode image with corresponding descriptor information. desc - * is also updated to contain the right offsets within returned image. - */ -static void * -ls_ucode_img_build(const struct firmware *bl, const struct firmware *code, - const struct firmware *data, struct ls_ucode_img_desc *desc) -{ - struct fw_bin_header *bin_hdr = (void *)bl->data; - struct fw_bl_desc *bl_desc = (void *)bl->data + bin_hdr->header_offset; - void *bl_data = (void *)bl->data + bin_hdr->data_offset; - u32 pos = 0; - void *image; - - desc->bootloader_start_offset = pos; - desc->bootloader_size = ALIGN(bl_desc->code_size, sizeof(u32)); - desc->bootloader_imem_offset = bl_desc->start_tag * 256; - desc->bootloader_entry_point = bl_desc->start_tag * 256; - - pos = ALIGN(pos + desc->bootloader_size, BL_DESC_BLK_SIZE); - desc->app_start_offset = pos; - desc->app_size = ALIGN(code->size, BL_DESC_BLK_SIZE) + - ALIGN(data->size, BL_DESC_BLK_SIZE); - desc->app_imem_offset = 0; - desc->app_imem_entry = 0; - desc->app_dmem_offset = 0; - desc->app_resident_code_offset = 0; - desc->app_resident_code_size = ALIGN(code->size, BL_DESC_BLK_SIZE); - - pos = ALIGN(pos + desc->app_resident_code_size, BL_DESC_BLK_SIZE); - desc->app_resident_data_offset = pos - desc->app_start_offset; - desc->app_resident_data_size = ALIGN(data->size, BL_DESC_BLK_SIZE); - - desc->image_size = ALIGN(bl_desc->code_size, BL_DESC_BLK_SIZE) + - desc->app_size; - - image = kzalloc(desc->image_size, GFP_KERNEL); - if (!image) - return ERR_PTR(-ENOMEM); - - memcpy(image + desc->bootloader_start_offset, bl_data, - bl_desc->code_size); - memcpy(image + desc->app_start_offset, code->data, code->size); - memcpy(image + desc->app_start_offset + desc->app_resident_data_offset, - data->data, data->size); - - return image; -} - -/** - * ls_ucode_img_load_generic() - load and prepare a LS ucode image - * - * Load the LS microcode, bootloader and signature and pack them into a single - * blob. Also generate the corresponding ucode descriptor. - */ -static int -ls_ucode_img_load_generic(struct nvkm_subdev *subdev, - struct ls_ucode_img *img, const char *falcon_name, - const u32 falcon_id) -{ - const struct firmware *bl, *code, *data; - struct lsf_ucode_desc *lsf_desc; - char f[64]; - int ret; - - img->ucode_header = NULL; - - snprintf(f, sizeof(f), "gr/%s_bl", falcon_name); - ret = nvkm_firmware_get(subdev->device, f, &bl); - if (ret) - goto error; - - snprintf(f, sizeof(f), "gr/%s_inst", falcon_name); - ret = nvkm_firmware_get(subdev->device, f, &code); - if (ret) - goto free_bl; - - snprintf(f, sizeof(f), "gr/%s_data", falcon_name); - ret = nvkm_firmware_get(subdev->device, f, &data); - if (ret) - goto free_inst; - - img->ucode_data = ls_ucode_img_build(bl, code, data, - &img->ucode_desc); - if (IS_ERR(img->ucode_data)) { - ret = PTR_ERR(img->ucode_data); - goto free_data; - } - img->ucode_size = img->ucode_desc.image_size; - - snprintf(f, sizeof(f), "gr/%s_sig", falcon_name); - lsf_desc = gm200_secboot_load_firmware(subdev, f, sizeof(*lsf_desc)); - if (IS_ERR(lsf_desc)) { - ret = PTR_ERR(lsf_desc); - goto free_image; - } - /* not needed? the signature should already have the right value */ - lsf_desc->falcon_id = falcon_id; - memcpy(&img->lsb_header.signature, lsf_desc, sizeof(*lsf_desc)); - img->falcon_id = lsf_desc->falcon_id; - kfree(lsf_desc); - - /* success path - only free requested firmware files */ - goto free_data; - -free_image: - kfree(img->ucode_data); -free_data: - nvkm_firmware_put(data); -free_inst: - nvkm_firmware_put(code); -free_bl: - nvkm_firmware_put(bl); -error: - return ret; -} - -typedef int (*lsf_load_func)(struct nvkm_subdev *, struct ls_ucode_img *); - -static int -ls_ucode_img_load_fecs(struct nvkm_subdev *subdev, struct ls_ucode_img *img) -{ - return ls_ucode_img_load_generic(subdev, img, "fecs", - NVKM_SECBOOT_FALCON_FECS); -} - -static int -ls_ucode_img_load_gpccs(struct nvkm_subdev *subdev, struct ls_ucode_img *img) -{ - return ls_ucode_img_load_generic(subdev, img, "gpccs", - NVKM_SECBOOT_FALCON_GPCCS); -} - -/** - * ls_ucode_img_load() - create a lsf_ucode_img and load it - */ -static struct ls_ucode_img * -ls_ucode_img_load(struct nvkm_subdev *subdev, lsf_load_func load_func) -{ - struct ls_ucode_img *img; - int ret; - - img = kzalloc(sizeof(*img), GFP_KERNEL); - if (!img) - return ERR_PTR(-ENOMEM); - - ret = load_func(subdev, img); - if (ret) { - kfree(img); - return ERR_PTR(ret); - } - - return img; -} - -static const lsf_load_func lsf_load_funcs[] = { - [NVKM_SECBOOT_FALCON_END] = NULL, /* reserve enough space */ - [NVKM_SECBOOT_FALCON_FECS] = ls_ucode_img_load_fecs, - [NVKM_SECBOOT_FALCON_GPCCS] = ls_ucode_img_load_gpccs, -}; - -/** - * ls_ucode_img_populate_bl_desc() - populate a DMEM BL descriptor for LS image - * @img: ucode image to generate against - * @desc: descriptor to populate - * @sb: secure boot state to use for base addresses - * - * Populate the DMEM BL descriptor with the information contained in a - * ls_ucode_desc. - * - */ -static void -ls_ucode_img_populate_bl_desc(struct ls_ucode_img *img, u64 wpr_addr, - struct gm200_flcn_bl_desc *desc) -{ - struct ls_ucode_img_desc *pdesc = &img->ucode_desc; - u64 addr_base; - - addr_base = wpr_addr + img->lsb_header.ucode_off + - pdesc->app_start_offset; - - memset(desc, 0, sizeof(*desc)); - desc->ctx_dma = FALCON_DMAIDX_UCODE; - desc->code_dma_base.lo = lower_32_bits( - (addr_base + pdesc->app_resident_code_offset)); - desc->code_dma_base.hi = upper_32_bits( - (addr_base + pdesc->app_resident_code_offset)); - desc->non_sec_code_size = pdesc->app_resident_code_size; - desc->data_dma_base.lo = lower_32_bits( - (addr_base + pdesc->app_resident_data_offset)); - desc->data_dma_base.hi = upper_32_bits( - (addr_base + pdesc->app_resident_data_offset)); - desc->data_size = pdesc->app_resident_data_size; - desc->code_entry_point = pdesc->app_imem_entry; -} - -#define LSF_LSB_HEADER_ALIGN 256 -#define LSF_BL_DATA_ALIGN 256 -#define LSF_BL_DATA_SIZE_ALIGN 256 -#define LSF_BL_CODE_SIZE_ALIGN 256 -#define LSF_UCODE_DATA_ALIGN 4096 - -/** - * ls_ucode_img_fill_headers - fill the WPR and LSB headers of an image - * @gsb: secure boot device used - * @img: image to generate for - * @offset: offset in the WPR region where this image starts - * - * Allocate space in the WPR area from offset and write the WPR and LSB headers - * accordingly. - * - * Return: offset at the end of this image. - */ -static u32 -ls_ucode_img_fill_headers(struct gm200_secboot *gsb, struct ls_ucode_img *img, - u32 offset) -{ - struct lsf_wpr_header *whdr = &img->wpr_header; - struct lsf_lsb_header *lhdr = &img->lsb_header; - struct ls_ucode_img_desc *desc = &img->ucode_desc; - - if (img->ucode_header) { - nvkm_fatal(&gsb->base.subdev, - "images withough loader are not supported yet!\n"); - return offset; - } - - /* Fill WPR header */ - whdr->falcon_id = img->falcon_id; - whdr->bootstrap_owner = gsb->base.func->boot_falcon; - whdr->status = LSF_IMAGE_STATUS_COPY; - - /* Align, save off, and include an LSB header size */ - offset = ALIGN(offset, LSF_LSB_HEADER_ALIGN); - whdr->lsb_offset = offset; - offset += sizeof(struct lsf_lsb_header); - - /* - * Align, save off, and include the original (static) ucode - * image size - */ - offset = ALIGN(offset, LSF_UCODE_DATA_ALIGN); - lhdr->ucode_off = offset; - offset += img->ucode_size; - - /* - * For falcons that use a boot loader (BL), we append a loader - * desc structure on the end of the ucode image and consider - * this the boot loader data. The host will then copy the loader - * desc args to this space within the WPR region (before locking - * down) and the HS bin will then copy them to DMEM 0 for the - * loader. - */ - lhdr->bl_code_size = ALIGN(desc->bootloader_size, - LSF_BL_CODE_SIZE_ALIGN); - lhdr->ucode_size = ALIGN(desc->app_resident_data_offset, - LSF_BL_CODE_SIZE_ALIGN) + lhdr->bl_code_size; - lhdr->data_size = ALIGN(desc->app_size, LSF_BL_CODE_SIZE_ALIGN) + - lhdr->bl_code_size - lhdr->ucode_size; - /* - * Though the BL is located at 0th offset of the image, the VA - * is different to make sure that it doesn't collide the actual - * OS VA range - */ - lhdr->bl_imem_off = desc->bootloader_imem_offset; - lhdr->app_code_off = desc->app_start_offset + - desc->app_resident_code_offset; - lhdr->app_code_size = desc->app_resident_code_size; - lhdr->app_data_off = desc->app_start_offset + - desc->app_resident_data_offset; - lhdr->app_data_size = desc->app_resident_data_size; - - lhdr->flags = 0; - if (img->falcon_id == gsb->base.func->boot_falcon) - lhdr->flags = LSF_FLAG_DMACTL_REQ_CTX; - - /* GPCCS will be loaded using PRI */ - if (img->falcon_id == NVKM_SECBOOT_FALCON_GPCCS) - lhdr->flags |= LSF_FLAG_FORCE_PRIV_LOAD; - - /* Align (size bloat) and save off BL descriptor size */ - lhdr->bl_data_size = ALIGN(sizeof(struct gm200_flcn_bl_desc), - LSF_BL_DATA_SIZE_ALIGN); - /* - * Align, save off, and include the additional BL data - */ - offset = ALIGN(offset, LSF_BL_DATA_ALIGN); - lhdr->bl_data_off = offset; - offset += lhdr->bl_data_size; - - return offset; -} - -static void -ls_ucode_mgr_init(struct ls_ucode_mgr *mgr) -{ - memset(mgr, 0, sizeof(*mgr)); - INIT_LIST_HEAD(&mgr->img_list); -} - -static void -ls_ucode_mgr_cleanup(struct ls_ucode_mgr *mgr) -{ - struct ls_ucode_img *img, *t; - - list_for_each_entry_safe(img, t, &mgr->img_list, node) { - kfree(img->ucode_data); - kfree(img->ucode_header); - kfree(img); - } -} - -static void -ls_ucode_mgr_add_img(struct ls_ucode_mgr *mgr, struct ls_ucode_img *img) -{ - mgr->count++; - list_add_tail(&img->node, &mgr->img_list); -} - -/** - * ls_ucode_mgr_fill_headers - fill WPR and LSB headers of all managed images - */ -static void -ls_ucode_mgr_fill_headers(struct gm200_secboot *gsb, struct ls_ucode_mgr *mgr) -{ - struct ls_ucode_img *img; - u32 offset; - - /* - * Start with an array of WPR headers at the base of the WPR. - * The expectation here is that the secure falcon will do a single DMA - * read of this array and cache it internally so it's ok to pack these. - * Also, we add 1 to the falcon count to indicate the end of the array. - */ - offset = sizeof(struct lsf_wpr_header) * (mgr->count + 1); - - /* - * Walk the managed falcons, accounting for the LSB structs - * as well as the ucode images. - */ - list_for_each_entry(img, &mgr->img_list, node) { - offset = ls_ucode_img_fill_headers(gsb, img, offset); - } - - mgr->wpr_size = offset; -} - -/** - * ls_ucode_mgr_write_wpr - write the WPR blob contents - */ -static int -ls_ucode_mgr_write_wpr(struct gm200_secboot *gsb, struct ls_ucode_mgr *mgr, - struct nvkm_gpuobj *wpr_blob) -{ - struct ls_ucode_img *img; - u32 pos = 0; - - nvkm_kmap(wpr_blob); - - list_for_each_entry(img, &mgr->img_list, node) { - nvkm_gpuobj_memcpy_to(wpr_blob, pos, &img->wpr_header, - sizeof(img->wpr_header)); - - nvkm_gpuobj_memcpy_to(wpr_blob, img->wpr_header.lsb_offset, - &img->lsb_header, sizeof(img->lsb_header)); - - /* Generate and write BL descriptor */ - if (!img->ucode_header) { - u8 desc[gsb->func->bl_desc_size]; - struct gm200_flcn_bl_desc gdesc; - - ls_ucode_img_populate_bl_desc(img, gsb->wpr_addr, - &gdesc); - gsb->func->fixup_bl_desc(&gdesc, &desc); - nvkm_gpuobj_memcpy_to(wpr_blob, - img->lsb_header.bl_data_off, - &desc, gsb->func->bl_desc_size); - } - - /* Copy ucode */ - nvkm_gpuobj_memcpy_to(wpr_blob, img->lsb_header.ucode_off, - img->ucode_data, img->ucode_size); - - pos += sizeof(img->wpr_header); - } - - nvkm_wo32(wpr_blob, pos, NVKM_SECBOOT_FALCON_INVALID); - - nvkm_done(wpr_blob); - - return 0; -} - -/* Both size and address of WPR need to be 128K-aligned */ -#define WPR_ALIGNMENT 0x20000 -/** - * gm200_secboot_prepare_ls_blob() - prepare the LS blob - * - * For each securely managed falcon, load the FW, signatures and bootloaders and - * prepare a ucode blob. Then, compute the offsets in the WPR region for each - * blob, and finally write the headers and ucode blobs into a GPU object that - * will be copied into the WPR region by the HS firmware. - */ -static int -gm200_secboot_prepare_ls_blob(struct gm200_secboot *gsb) -{ - struct nvkm_secboot *sb = &gsb->base; - struct nvkm_device *device = sb->subdev.device; - struct ls_ucode_mgr mgr; - int falcon_id; - int ret; - - ls_ucode_mgr_init(&mgr); - - /* Load all LS blobs */ - for_each_set_bit(falcon_id, &gsb->base.func->managed_falcons, - NVKM_SECBOOT_FALCON_END) { - struct ls_ucode_img *img; - - img = ls_ucode_img_load(&sb->subdev, lsf_load_funcs[falcon_id]); - - if (IS_ERR(img)) { - ret = PTR_ERR(img); - goto cleanup; - } - ls_ucode_mgr_add_img(&mgr, img); - } - - /* - * Fill the WPR and LSF headers with the right offsets and compute - * required WPR size - */ - ls_ucode_mgr_fill_headers(gsb, &mgr); - mgr.wpr_size = ALIGN(mgr.wpr_size, WPR_ALIGNMENT); - - /* Allocate GPU object that will contain the WPR region */ - ret = nvkm_gpuobj_new(device, mgr.wpr_size, WPR_ALIGNMENT, false, NULL, - &gsb->ls_blob); - if (ret) - goto cleanup; - - nvkm_debug(&sb->subdev, "%d managed LS falcons, WPR size is %d bytes\n", - mgr.count, mgr.wpr_size); - - /* If WPR address and size are not fixed, set them to fit the LS blob */ - if (!gsb->wpr_size) { - gsb->wpr_addr = gsb->ls_blob->addr; - gsb->wpr_size = gsb->ls_blob->size; - } - - /* Write LS blob */ - ret = ls_ucode_mgr_write_wpr(gsb, &mgr, gsb->ls_blob); - if (ret) - nvkm_gpuobj_del(&gsb->ls_blob); - -cleanup: - ls_ucode_mgr_cleanup(&mgr); - - return ret; -} - -/* - * High-secure blob creation - */ - -/** - * gm200_secboot_hsf_patch_signature() - patch HS blob with correct signature - */ -static void -gm200_secboot_hsf_patch_signature(struct gm200_secboot *gsb, void *acr_image) -{ - struct nvkm_secboot *sb = &gsb->base; - struct fw_bin_header *hsbin_hdr = acr_image; - struct hsf_fw_header *fw_hdr = acr_image + hsbin_hdr->header_offset; - void *hs_data = acr_image + hsbin_hdr->data_offset; - void *sig; - u32 sig_size; - - /* Falcon in debug or production mode? */ - if ((nvkm_rd32(sb->subdev.device, sb->base + 0xc08) >> 20) & 0x1) { - sig = acr_image + fw_hdr->sig_dbg_offset; - sig_size = fw_hdr->sig_dbg_size; - } else { - sig = acr_image + fw_hdr->sig_prod_offset; - sig_size = fw_hdr->sig_prod_size; - } - - /* Patch signature */ - memcpy(hs_data + fw_hdr->patch_loc, sig + fw_hdr->patch_sig, sig_size); -} - -/** - * gm200_secboot_populate_hsf_bl_desc() - populate BL descriptor for HS image - */ -static void -gm200_secboot_populate_hsf_bl_desc(void *acr_image, - struct gm200_flcn_bl_desc *bl_desc) -{ - struct fw_bin_header *hsbin_hdr = acr_image; - struct hsf_fw_header *fw_hdr = acr_image + hsbin_hdr->header_offset; - struct hsf_load_header *load_hdr = acr_image + fw_hdr->hdr_offset; - - /* - * Descriptor for the bootloader that will load the ACR image into - * IMEM/DMEM memory. - */ - fw_hdr = acr_image + hsbin_hdr->header_offset; - load_hdr = acr_image + fw_hdr->hdr_offset; - memset(bl_desc, 0, sizeof(*bl_desc)); - bl_desc->ctx_dma = FALCON_DMAIDX_VIRT; - bl_desc->non_sec_code_off = load_hdr->non_sec_code_off; - bl_desc->non_sec_code_size = load_hdr->non_sec_code_size; - bl_desc->sec_code_off = load_hdr->app[0].sec_code_off; - bl_desc->sec_code_size = load_hdr->app[0].sec_code_size; - bl_desc->code_entry_point = 0; - /* - * We need to set code_dma_base to the virtual address of the acr_blob, - * and add this address to data_dma_base before writing it into DMEM - */ - bl_desc->code_dma_base.lo = 0; - bl_desc->data_dma_base.lo = load_hdr->data_dma_base; - bl_desc->data_size = load_hdr->data_size; -} - -/** - * gm200_secboot_prepare_hs_blob - load and prepare a HS blob and BL descriptor - * - * @gsb secure boot instance to prepare for - * @fw name of the HS firmware to load - * @blob pointer to gpuobj that will be allocated to receive the HS FW payload - * @bl_desc pointer to the BL descriptor to write for this firmware - * @patch whether we should patch the HS descriptor (only for HS loaders) - */ -static int -gm200_secboot_prepare_hs_blob(struct gm200_secboot *gsb, const char *fw, - struct nvkm_gpuobj **blob, - struct gm200_flcn_bl_desc *bl_desc, bool patch) -{ - struct nvkm_subdev *subdev = &gsb->base.subdev; - void *acr_image; - struct fw_bin_header *hsbin_hdr; - struct hsf_fw_header *fw_hdr; - void *acr_data; - struct hsf_load_header *load_hdr; - struct hsflcn_acr_desc *desc; - int ret; - - acr_image = gm200_secboot_load_firmware(subdev, fw, 0); - if (IS_ERR(acr_image)) - return PTR_ERR(acr_image); - hsbin_hdr = acr_image; - - /* Patch signature */ - gm200_secboot_hsf_patch_signature(gsb, acr_image); - - acr_data = acr_image + hsbin_hdr->data_offset; - - /* Patch descriptor? */ - if (patch) { - fw_hdr = acr_image + hsbin_hdr->header_offset; - load_hdr = acr_image + fw_hdr->hdr_offset; - desc = acr_data + load_hdr->data_dma_base; - gsb->func->fixup_hs_desc(gsb, desc); - } - - /* Generate HS BL descriptor */ - gm200_secboot_populate_hsf_bl_desc(acr_image, bl_desc); - - /* Create ACR blob and copy HS data to it */ - ret = nvkm_gpuobj_new(subdev->device, ALIGN(hsbin_hdr->data_size, 256), - 0x1000, false, NULL, blob); - if (ret) - goto cleanup; - - nvkm_kmap(*blob); - nvkm_gpuobj_memcpy_to(*blob, 0, acr_data, hsbin_hdr->data_size); - nvkm_done(*blob); - -cleanup: - kfree(acr_image); - - return ret; -} - -/* - * High-secure bootloader blob creation - */ - -static int -gm200_secboot_prepare_hsbl_blob(struct gm200_secboot *gsb) -{ - struct nvkm_subdev *subdev = &gsb->base.subdev; - - gsb->hsbl_blob = gm200_secboot_load_firmware(subdev, "acr/bl", 0); - if (IS_ERR(gsb->hsbl_blob)) { - int ret = PTR_ERR(gsb->hsbl_blob); - - gsb->hsbl_blob = NULL; - return ret; - } - - return 0; -} - -/** - * gm20x_secboot_prepare_blobs - load blobs common to all GM20X GPUs. - * - * This includes the LS blob, HS ucode loading blob, and HS bootloader. - * - * The HS ucode unload blob is only used on dGPU. */ int -gm20x_secboot_prepare_blobs(struct gm200_secboot *gsb) -{ - int ret; - - /* Load and prepare the managed falcon's firmwares */ - if (!gsb->ls_blob) { - ret = gm200_secboot_prepare_ls_blob(gsb); - if (ret) - return ret; - } - - /* Load the HS firmware that will load the LS firmwares */ - if (!gsb->acr_load_blob) { - ret = gm200_secboot_prepare_hs_blob(gsb, "acr/ucode_load", - &gsb->acr_load_blob, - &gsb->acr_load_bl_desc, true); - if (ret) - return ret; - } - - /* Load the HS firmware bootloader */ - if (!gsb->hsbl_blob) { - ret = gm200_secboot_prepare_hsbl_blob(gsb); - if (ret) - return ret; - } - - return 0; -} - -static int -gm200_secboot_prepare_blobs(struct gm200_secboot *gsb) -{ - int ret; - - ret = gm20x_secboot_prepare_blobs(gsb); - if (ret) - return ret; - - /* dGPU only: load the HS firmware that unprotects the WPR region */ - if (!gsb->acr_unload_blob) { - ret = gm200_secboot_prepare_hs_blob(gsb, "acr/ucode_unload", - &gsb->acr_unload_blob, - &gsb->acr_unload_bl_desc, false); - if (ret) - return ret; - } - - return 0; -} - -static int -gm200_secboot_blobs_ready(struct gm200_secboot *gsb) +gm200_secboot_run_blob(struct nvkm_secboot *sb, struct nvkm_gpuobj *blob) { + struct gm200_secboot *gsb = gm200_secboot(sb); struct nvkm_subdev *subdev = &gsb->base.subdev; + struct nvkm_falcon *falcon = gsb->base.boot_falcon; + struct nvkm_vma vma; int ret; - /* firmware already loaded, nothing to do... */ - if (gsb->firmware_ok) - return 0; - - ret = gsb->func->prepare_blobs(gsb); - if (ret) { - nvkm_error(subdev, "failed to load secure firmware\n"); - return ret; - } - - gsb->firmware_ok = true; - - return 0; -} - - -/* - * Secure Boot Execution - */ - -/** - * gm200_secboot_load_hs_bl() - load HS bootloader into DMEM and IMEM - */ -static void -gm200_secboot_load_hs_bl(struct gm200_secboot *gsb, void *data, u32 data_size) -{ - struct nvkm_device *device = gsb->base.subdev.device; - struct fw_bin_header *hdr = gsb->hsbl_blob; - struct fw_bl_desc *hsbl_desc = gsb->hsbl_blob + hdr->header_offset; - void *blob_data = gsb->hsbl_blob + hdr->data_offset; - void *hsbl_code = blob_data + hsbl_desc->code_off; - void *hsbl_data = blob_data + hsbl_desc->data_off; - u32 code_size = ALIGN(hsbl_desc->code_size, 256); - const u32 base = gsb->base.base; - u32 blk; - u32 tag; - int i; - - /* - * Copy HS bootloader data - */ - nvkm_wr32(device, base + 0x1c0, (0x00000000 | (0x1 << 24))); - for (i = 0; i < hsbl_desc->data_size / 4; i++) - nvkm_wr32(device, base + 0x1c4, ((u32 *)hsbl_data)[i]); - - /* - * Copy HS bootloader interface structure where the HS descriptor - * expects it to be - */ - nvkm_wr32(device, base + 0x1c0, - (hsbl_desc->dmem_load_off | (0x1 << 24))); - for (i = 0; i < data_size / 4; i++) - nvkm_wr32(device, base + 0x1c4, ((u32 *)data)[i]); - - /* Copy HS bootloader code to end of IMEM */ - blk = (nvkm_rd32(device, base + 0x108) & 0x1ff) - (code_size >> 8); - tag = hsbl_desc->start_tag; - nvkm_wr32(device, base + 0x180, ((blk & 0xff) << 8) | (0x1 << 24)); - for (i = 0; i < code_size / 4; i++) { - /* write new tag every 256B */ - if ((i & 0x3f) == 0) { - nvkm_wr32(device, base + 0x188, tag & 0xffff); - tag++; - } - nvkm_wr32(device, base + 0x184, ((u32 *)hsbl_code)[i]); - } - nvkm_wr32(device, base + 0x188, 0); -} - -/** - * gm200_secboot_setup_falcon() - set up the secure falcon for secure boot - */ -static int -gm200_secboot_setup_falcon(struct gm200_secboot *gsb) -{ - struct nvkm_device *device = gsb->base.subdev.device; - struct fw_bin_header *hdr = gsb->hsbl_blob; - struct fw_bl_desc *hsbl_desc = gsb->hsbl_blob + hdr->header_offset; - /* virtual start address for boot vector */ - u32 virt_addr = hsbl_desc->start_tag << 8; - const u32 base = gsb->base.base; - const u32 reg_base = base + 0xe00; - u32 inst_loc; - int ret; - - ret = nvkm_secboot_falcon_reset(&gsb->base); + ret = nvkm_falcon_get(falcon, subdev); if (ret) return ret; - /* setup apertures - virtual */ - nvkm_wr32(device, reg_base + 4 * (FALCON_DMAIDX_UCODE), 0x4); - nvkm_wr32(device, reg_base + 4 * (FALCON_DMAIDX_VIRT), 0x0); - /* setup apertures - physical */ - nvkm_wr32(device, reg_base + 4 * (FALCON_DMAIDX_PHYS_VID), 0x4); - nvkm_wr32(device, reg_base + 4 * (FALCON_DMAIDX_PHYS_SYS_COH), - 0x4 | 0x1); - nvkm_wr32(device, reg_base + 4 * (FALCON_DMAIDX_PHYS_SYS_NCOH), - 0x4 | 0x2); - - /* Set context */ - if (nvkm_memory_target(gsb->inst->memory) == NVKM_MEM_TARGET_VRAM) - inst_loc = 0x0; /* FB */ - else - inst_loc = 0x3; /* Non-coherent sysmem */ - - nvkm_mask(device, base + 0x048, 0x1, 0x1); - nvkm_wr32(device, base + 0x480, - ((gsb->inst->addr >> 12) & 0xfffffff) | - (inst_loc << 28) | (1 << 30)); - - /* Set boot vector to code's starting virtual address */ - nvkm_wr32(device, base + 0x104, virt_addr); - - return 0; -} - -/** - * gm200_secboot_run_hs_blob() - run the given high-secure blob - */ -static int -gm200_secboot_run_hs_blob(struct gm200_secboot *gsb, struct nvkm_gpuobj *blob, - struct gm200_flcn_bl_desc *desc) -{ - struct nvkm_vma vma; - u64 vma_addr; - const u32 bl_desc_size = gsb->func->bl_desc_size; - u8 bl_desc[bl_desc_size]; - int ret; - /* Map the HS firmware so the HS bootloader can see it */ ret = nvkm_gpuobj_map(blob, gsb->vm, NV_MEM_ACCESS_RW, &vma); - if (ret) + if (ret) { + nvkm_falcon_put(falcon, subdev); return ret; + } - /* Add the mapping address to the DMA bases */ - vma_addr = flcn64_to_u64(desc->code_dma_base) + vma.offset; - desc->code_dma_base.lo = lower_32_bits(vma_addr); - desc->code_dma_base.hi = upper_32_bits(vma_addr); - vma_addr = flcn64_to_u64(desc->data_dma_base) + vma.offset; - desc->data_dma_base.lo = lower_32_bits(vma_addr); - desc->data_dma_base.hi = upper_32_bits(vma_addr); - - /* Fixup the BL header */ - gsb->func->fixup_bl_desc(desc, &bl_desc); - - /* Reset the falcon and make it ready to run the HS bootloader */ - ret = gm200_secboot_setup_falcon(gsb); + /* Reset and set the falcon up */ + ret = nvkm_falcon_reset(falcon); if (ret) - goto done; + goto end; + nvkm_falcon_bind_context(falcon, gsb->inst); /* Load the HS bootloader into the falcon's IMEM/DMEM */ - gm200_secboot_load_hs_bl(gsb, &bl_desc, bl_desc_size); + ret = sb->acr->func->load(sb->acr, &gsb->base, blob, vma.offset); + if (ret) + goto end; + + /* Disable interrupts as we will poll for the HALT bit */ + nvkm_mc_intr_mask(sb->subdev.device, falcon->owner->index, false); + + /* Set default error value in mailbox register */ + nvkm_falcon_wr32(falcon, 0x040, 0xdeada5a5); /* Start the HS bootloader */ - ret = nvkm_secboot_falcon_run(&gsb->base); + nvkm_falcon_set_start_addr(falcon, sb->acr->start_address); + nvkm_falcon_start(falcon); + ret = nvkm_falcon_wait_for_halt(falcon, 100); if (ret) - goto done; + goto end; -done: - /* Restore the original DMA addresses */ - vma_addr = flcn64_to_u64(desc->code_dma_base) - vma.offset; - desc->code_dma_base.lo = lower_32_bits(vma_addr); - desc->code_dma_base.hi = upper_32_bits(vma_addr); - vma_addr = flcn64_to_u64(desc->data_dma_base) - vma.offset; - desc->data_dma_base.lo = lower_32_bits(vma_addr); - desc->data_dma_base.hi = upper_32_bits(vma_addr); + /* If mailbox register contains an error code, then ACR has failed */ + ret = nvkm_falcon_rd32(falcon, 0x040); + if (ret) { + nvkm_error(subdev, "ACR boot failed, ret 0x%08x", ret); + ret = -EINVAL; + goto end; + } + +end: + /* Reenable interrupts */ + nvkm_mc_intr_mask(sb->subdev.device, falcon->owner->index, true); /* We don't need the ACR firmware anymore */ nvkm_gpuobj_unmap(&vma); + nvkm_falcon_put(falcon, subdev); return ret; } -/* - * gm200_secboot_reset() - execute secure boot from the prepared state - * - * Load the HS bootloader and ask the falcon to run it. This will in turn - * load the HS firmware and run it, so once the falcon stops all the managed - * falcons should have their LS firmware loaded and be ready to run. - */ int -gm200_secboot_reset(struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon) -{ - struct gm200_secboot *gsb = gm200_secboot(sb); - int ret; - - /* Make sure all blobs are ready */ - ret = gm200_secboot_blobs_ready(gsb); - if (ret) - return ret; - - /* - * Dummy GM200 implementation: perform secure boot each time we are - * called on FECS. Since only FECS and GPCCS are managed and started - * together, this ought to be safe. - * - * Once we have proper PMU firmware and support, this will be changed - * to a proper call to the PMU method. - */ - if (falcon != NVKM_SECBOOT_FALCON_FECS) - goto end; - - /* If WPR is set and we have an unload blob, run it to unlock WPR */ - if (gsb->acr_unload_blob && - gsb->falcon_state[NVKM_SECBOOT_FALCON_FECS] != NON_SECURE) { - ret = gm200_secboot_run_hs_blob(gsb, gsb->acr_unload_blob, - &gsb->acr_unload_bl_desc); - if (ret) - return ret; - } - - /* Reload all managed falcons */ - ret = gm200_secboot_run_hs_blob(gsb, gsb->acr_load_blob, - &gsb->acr_load_bl_desc); - if (ret) - return ret; - -end: - gsb->falcon_state[falcon] = RESET; - return 0; -} - -int -gm200_secboot_start(struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon) -{ - struct gm200_secboot *gsb = gm200_secboot(sb); - int base; - - switch (falcon) { - case NVKM_SECBOOT_FALCON_FECS: - base = 0x409000; - break; - case NVKM_SECBOOT_FALCON_GPCCS: - base = 0x41a000; - break; - default: - nvkm_error(&sb->subdev, "cannot start unhandled falcon!\n"); - return -EINVAL; - } - - nvkm_wr32(sb->subdev.device, base + 0x130, 0x00000002); - gsb->falcon_state[falcon] = RUNNING; - - return 0; -} - - - -int -gm200_secboot_init(struct nvkm_secboot *sb) +gm200_secboot_oneinit(struct nvkm_secboot *sb) { struct gm200_secboot *gsb = gm200_secboot(sb); struct nvkm_device *device = sb->subdev.device; @@ -1361,24 +132,22 @@ gm200_secboot_init(struct nvkm_secboot *sb) nvkm_wo32(gsb->inst, 0x20c, upper_32_bits(vm_area_len - 1)); nvkm_done(gsb->inst); + if (sb->acr->func->oneinit) { + ret = sb->acr->func->oneinit(sb->acr, sb); + if (ret) + return ret; + } + return 0; } -static int +int gm200_secboot_fini(struct nvkm_secboot *sb, bool suspend) { - struct gm200_secboot *gsb = gm200_secboot(sb); int ret = 0; - int i; - /* Run the unload blob to unprotect the WPR region */ - if (gsb->acr_unload_blob && - gsb->falcon_state[NVKM_SECBOOT_FALCON_FECS] != NON_SECURE) - ret = gm200_secboot_run_hs_blob(gsb, gsb->acr_unload_blob, - &gsb->acr_unload_bl_desc); - - for (i = 0; i < NVKM_SECBOOT_FALCON_END; i++) - gsb->falcon_state[i] = NON_SECURE; + if (sb->acr->func->fini) + ret = sb->acr->func->fini(sb->acr, sb, suspend); return ret; } @@ -1388,11 +157,7 @@ gm200_secboot_dtor(struct nvkm_secboot *sb) { struct gm200_secboot *gsb = gm200_secboot(sb); - nvkm_gpuobj_del(&gsb->acr_unload_blob); - - kfree(gsb->hsbl_blob); - nvkm_gpuobj_del(&gsb->acr_load_blob); - nvkm_gpuobj_del(&gsb->ls_blob); + sb->acr->func->dtor(sb->acr); nvkm_vm_ref(NULL, &gsb->vm, gsb->pgd); nvkm_gpuobj_del(&gsb->pgd); @@ -1405,50 +170,9 @@ gm200_secboot_dtor(struct nvkm_secboot *sb) static const struct nvkm_secboot_func gm200_secboot = { .dtor = gm200_secboot_dtor, - .init = gm200_secboot_init, + .oneinit = gm200_secboot_oneinit, .fini = gm200_secboot_fini, - .reset = gm200_secboot_reset, - .start = gm200_secboot_start, - .managed_falcons = BIT(NVKM_SECBOOT_FALCON_FECS) | - BIT(NVKM_SECBOOT_FALCON_GPCCS), - .boot_falcon = NVKM_SECBOOT_FALCON_PMU, -}; - -/** - * gm200_fixup_bl_desc - just copy the BL descriptor - * - * Use the GM200 descriptor format by default. - */ -static void -gm200_secboot_fixup_bl_desc(const struct gm200_flcn_bl_desc *desc, void *ret) -{ - memcpy(ret, desc, sizeof(*desc)); -} - -static void -gm200_secboot_fixup_hs_desc(struct gm200_secboot *gsb, - struct hsflcn_acr_desc *desc) -{ - desc->ucode_blob_base = gsb->ls_blob->addr; - desc->ucode_blob_size = gsb->ls_blob->size; - - desc->wpr_offset = 0; - - /* WPR region information for the HS binary to set up */ - desc->wpr_region_id = 1; - desc->regions.no_regions = 1; - desc->regions.region_props[0].region_id = 1; - desc->regions.region_props[0].start_addr = gsb->wpr_addr >> 8; - desc->regions.region_props[0].end_addr = - (gsb->wpr_addr + gsb->wpr_size) >> 8; -} - -static const struct gm200_secboot_func -gm200_secboot_func = { - .bl_desc_size = sizeof(struct gm200_flcn_bl_desc), - .fixup_bl_desc = gm200_secboot_fixup_bl_desc, - .fixup_hs_desc = gm200_secboot_fixup_hs_desc, - .prepare_blobs = gm200_secboot_prepare_blobs, + .run_blob = gm200_secboot_run_blob, }; int @@ -1457,6 +181,12 @@ gm200_secboot_new(struct nvkm_device *device, int index, { int ret; struct gm200_secboot *gsb; + struct nvkm_acr *acr; + + acr = acr_r361_new(BIT(NVKM_SECBOOT_FALCON_FECS) | + BIT(NVKM_SECBOOT_FALCON_GPCCS)); + if (IS_ERR(acr)) + return PTR_ERR(acr); gsb = kzalloc(sizeof(*gsb), GFP_KERNEL); if (!gsb) { @@ -1465,15 +195,14 @@ gm200_secboot_new(struct nvkm_device *device, int index, } *psb = &gsb->base; - ret = nvkm_secboot_ctor(&gm200_secboot, device, index, &gsb->base); + ret = nvkm_secboot_ctor(&gm200_secboot, acr, device, index, &gsb->base); if (ret) return ret; - gsb->func = &gm200_secboot_func; - return 0; } + MODULE_FIRMWARE("nvidia/gm200/acr/bl.bin"); MODULE_FIRMWARE("nvidia/gm200/acr/ucode_load.bin"); MODULE_FIRMWARE("nvidia/gm200/acr/ucode_unload.bin"); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h new file mode 100644 index 000000000000..45adf1a3bc20 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h @@ -0,0 +1,43 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __NVKM_SECBOOT_GM200_H__ +#define __NVKM_SECBOOT_GM200_H__ + +#include "priv.h" + +struct gm200_secboot { + struct nvkm_secboot base; + + /* Instance block & address space used for HS FW execution */ + struct nvkm_gpuobj *inst; + struct nvkm_gpuobj *pgd; + struct nvkm_vm *vm; +}; +#define gm200_secboot(sb) container_of(sb, struct gm200_secboot, base) + +int gm200_secboot_oneinit(struct nvkm_secboot *); +int gm200_secboot_fini(struct nvkm_secboot *, bool); +void *gm200_secboot_dtor(struct nvkm_secboot *); +int gm200_secboot_run_blob(struct nvkm_secboot *, struct nvkm_gpuobj *); + +#endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c index d5395ebfe8d3..6707b8edc086 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c @@ -20,103 +20,8 @@ * DEALINGS IN THE SOFTWARE. */ -#include "priv.h" - -#include - -/* - * The BL header format used by GM20B's firmware is slightly different - * from the one of GM200. Fix the differences here. - */ -struct gm20b_flcn_bl_desc { - u32 reserved[4]; - u32 signature[4]; - u32 ctx_dma; - u32 code_dma_base; - u32 non_sec_code_off; - u32 non_sec_code_size; - u32 sec_code_off; - u32 sec_code_size; - u32 code_entry_point; - u32 data_dma_base; - u32 data_size; -}; - -static int -gm20b_secboot_prepare_blobs(struct gm200_secboot *gsb) -{ - struct nvkm_subdev *subdev = &gsb->base.subdev; - int acr_size; - int ret; - - ret = gm20x_secboot_prepare_blobs(gsb); - if (ret) - return ret; - - acr_size = gsb->acr_load_blob->size; - /* - * On Tegra the WPR region is set by the bootloader. It is illegal for - * the HS blob to be larger than this region. - */ - if (acr_size > gsb->wpr_size) { - nvkm_error(subdev, "WPR region too small for FW blob!\n"); - nvkm_error(subdev, "required: %dB\n", acr_size); - nvkm_error(subdev, "WPR size: %dB\n", gsb->wpr_size); - return -ENOSPC; - } - - return 0; -} - -/** - * gm20b_secboot_fixup_bl_desc - adapt BL descriptor to format used by GM20B FW - * - * There is only a slight format difference (DMA addresses being 32-bits and - * 256B-aligned) to address. - */ -static void -gm20b_secboot_fixup_bl_desc(const struct gm200_flcn_bl_desc *desc, void *ret) -{ - struct gm20b_flcn_bl_desc *gdesc = ret; - u64 addr; - - memcpy(gdesc->reserved, desc->reserved, sizeof(gdesc->reserved)); - memcpy(gdesc->signature, desc->signature, sizeof(gdesc->signature)); - gdesc->ctx_dma = desc->ctx_dma; - addr = desc->code_dma_base.hi; - addr <<= 32; - addr |= desc->code_dma_base.lo; - gdesc->code_dma_base = lower_32_bits(addr >> 8); - gdesc->non_sec_code_off = desc->non_sec_code_off; - gdesc->non_sec_code_size = desc->non_sec_code_size; - gdesc->sec_code_off = desc->sec_code_off; - gdesc->sec_code_size = desc->sec_code_size; - gdesc->code_entry_point = desc->code_entry_point; - addr = desc->data_dma_base.hi; - addr <<= 32; - addr |= desc->data_dma_base.lo; - gdesc->data_dma_base = lower_32_bits(addr >> 8); - gdesc->data_size = desc->data_size; -} - -static void -gm20b_secboot_fixup_hs_desc(struct gm200_secboot *gsb, - struct hsflcn_acr_desc *desc) -{ - desc->ucode_blob_base = gsb->ls_blob->addr; - desc->ucode_blob_size = gsb->ls_blob->size; - - desc->wpr_offset = 0; -} - -static const struct gm200_secboot_func -gm20b_secboot_func = { - .bl_desc_size = sizeof(struct gm20b_flcn_bl_desc), - .fixup_bl_desc = gm20b_secboot_fixup_bl_desc, - .fixup_hs_desc = gm20b_secboot_fixup_hs_desc, - .prepare_blobs = gm20b_secboot_prepare_blobs, -}; - +#include "acr.h" +#include "gm200.h" #ifdef CONFIG_ARCH_TEGRA #define TEGRA_MC_BASE 0x70019000 @@ -144,15 +49,15 @@ gm20b_tegra_read_wpr(struct gm200_secboot *gsb) nvkm_error(&sb->subdev, "Cannot map Tegra MC registers\n"); return PTR_ERR(mc); } - gsb->wpr_addr = ioread32_native(mc + MC_SECURITY_CARVEOUT2_BOM_0) | + sb->wpr_addr = ioread32_native(mc + MC_SECURITY_CARVEOUT2_BOM_0) | ((u64)ioread32_native(mc + MC_SECURITY_CARVEOUT2_BOM_HI_0) << 32); - gsb->wpr_size = ioread32_native(mc + MC_SECURITY_CARVEOUT2_SIZE_128K) + sb->wpr_size = ioread32_native(mc + MC_SECURITY_CARVEOUT2_SIZE_128K) << 17; cfg = ioread32_native(mc + MC_SECURITY_CARVEOUT2_CFG0); iounmap(mc); /* Check that WPR settings are valid */ - if (gsb->wpr_size == 0) { + if (sb->wpr_size == 0) { nvkm_error(&sb->subdev, "WPR region is empty\n"); return -EINVAL; } @@ -174,7 +79,7 @@ gm20b_tegra_read_wpr(struct gm200_secboot *gsb) #endif static int -gm20b_secboot_init(struct nvkm_secboot *sb) +gm20b_secboot_oneinit(struct nvkm_secboot *sb) { struct gm200_secboot *gsb = gm200_secboot(sb); int ret; @@ -183,17 +88,15 @@ gm20b_secboot_init(struct nvkm_secboot *sb) if (ret) return ret; - return gm200_secboot_init(sb); + return gm200_secboot_oneinit(sb); } static const struct nvkm_secboot_func gm20b_secboot = { .dtor = gm200_secboot_dtor, - .init = gm20b_secboot_init, - .reset = gm200_secboot_reset, - .start = gm200_secboot_start, - .managed_falcons = BIT(NVKM_SECBOOT_FALCON_FECS), - .boot_falcon = NVKM_SECBOOT_FALCON_PMU, + .oneinit = gm20b_secboot_oneinit, + .fini = gm200_secboot_fini, + .run_blob = gm200_secboot_run_blob, }; int @@ -202,6 +105,11 @@ gm20b_secboot_new(struct nvkm_device *device, int index, { int ret; struct gm200_secboot *gsb; + struct nvkm_acr *acr; + + acr = acr_r352_new(BIT(NVKM_SECBOOT_FALCON_FECS)); + if (IS_ERR(acr)) + return PTR_ERR(acr); gsb = kzalloc(sizeof(*gsb), GFP_KERNEL); if (!gsb) { @@ -210,12 +118,10 @@ gm20b_secboot_new(struct nvkm_device *device, int index, } *psb = &gsb->base; - ret = nvkm_secboot_ctor(&gm20b_secboot, device, index, &gsb->base); + ret = nvkm_secboot_ctor(&gm20b_secboot, acr, device, index, &gsb->base); if (ret) return ret; - gsb->func = &gm20b_secboot_func; - return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h new file mode 100644 index 000000000000..00886cee57eb --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h @@ -0,0 +1,151 @@ +/* + * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __NVKM_SECBOOT_LS_UCODE_H__ +#define __NVKM_SECBOOT_LS_UCODE_H__ + +#include +#include +#include + + +/** + * struct ls_ucode_img_desc - descriptor of firmware image + * @descriptor_size: size of this descriptor + * @image_size: size of the whole image + * @bootloader_start_offset: start offset of the bootloader in ucode image + * @bootloader_size: size of the bootloader + * @bootloader_imem_offset: start off set of the bootloader in IMEM + * @bootloader_entry_point: entry point of the bootloader in IMEM + * @app_start_offset: start offset of the LS firmware + * @app_size: size of the LS firmware's code and data + * @app_imem_offset: offset of the app in IMEM + * @app_imem_entry: entry point of the app in IMEM + * @app_dmem_offset: offset of the data in DMEM + * @app_resident_code_offset: offset of app code from app_start_offset + * @app_resident_code_size: size of the code + * @app_resident_data_offset: offset of data from app_start_offset + * @app_resident_data_size: size of data + * + * A firmware image contains the code, data, and bootloader of a given LS + * falcon in a single blob. This structure describes where everything is. + * + * This can be generated from a (bootloader, code, data) set if they have + * been loaded separately, or come directly from a file. + */ +struct ls_ucode_img_desc { + u32 descriptor_size; + u32 image_size; + u32 tools_version; + u32 app_version; + char date[64]; + u32 bootloader_start_offset; + u32 bootloader_size; + u32 bootloader_imem_offset; + u32 bootloader_entry_point; + u32 app_start_offset; + u32 app_size; + u32 app_imem_offset; + u32 app_imem_entry; + u32 app_dmem_offset; + u32 app_resident_code_offset; + u32 app_resident_code_size; + u32 app_resident_data_offset; + u32 app_resident_data_size; + u32 nb_overlays; + struct {u32 start; u32 size; } load_ovl[64]; + u32 compressed; +}; + +/** + * struct ls_ucode_img - temporary storage for loaded LS firmwares + * @node: to link within lsf_ucode_mgr + * @falcon_id: ID of the falcon this LS firmware is for + * @ucode_desc: loaded or generated map of ucode_data + * @ucode_data: firmware payload (code and data) + * @ucode_size: size in bytes of data in ucode_data + * @sig: signature for this firmware + * @sig:size: size of the signature in bytes + * + * Preparing the WPR LS blob requires information about all the LS firmwares + * (size, etc) to be known. This structure contains all the data of one LS + * firmware. + */ +struct ls_ucode_img { + struct list_head node; + enum nvkm_secboot_falcon falcon_id; + + struct ls_ucode_img_desc ucode_desc; + u8 *ucode_data; + u32 ucode_size; + + u8 *sig; + u32 sig_size; +}; + +/** + * struct fw_bin_header - header of firmware files + * @bin_magic: always 0x3b1d14f0 + * @bin_ver: version of the bin format + * @bin_size: entire image size including this header + * @header_offset: offset of the firmware/bootloader header in the file + * @data_offset: offset of the firmware/bootloader payload in the file + * @data_size: size of the payload + * + * This header is located at the beginning of the HS firmware and HS bootloader + * files, to describe where the headers and data can be found. + */ +struct fw_bin_header { + u32 bin_magic; + u32 bin_ver; + u32 bin_size; + u32 header_offset; + u32 data_offset; + u32 data_size; +}; + +/** + * struct fw_bl_desc - firmware bootloader descriptor + * @start_tag: starting tag of bootloader + * @desc_dmem_load_off: DMEM offset of flcn_bl_dmem_desc + * @code_off: offset of code section + * @code_size: size of code section + * @data_off: offset of data section + * @data_size: size of data section + * + * This structure is embedded in bootloader firmware files at to describe the + * IMEM and DMEM layout expected by the bootloader. + */ +struct fw_bl_desc { + u32 start_tag; + u32 dmem_load_off; + u32 code_off; + u32 code_size; + u32 data_off; + u32 data_size; +}; + +int acr_ls_ucode_load_fecs(const struct nvkm_subdev *, struct ls_ucode_img *); +int acr_ls_ucode_load_gpccs(const struct nvkm_subdev *, struct ls_ucode_img *); + + +#endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c new file mode 100644 index 000000000000..40a6df77bb8a --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c @@ -0,0 +1,158 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + + +#include "ls_ucode.h" +#include "acr.h" + +#include + +#define BL_DESC_BLK_SIZE 256 +/** + * Build a ucode image and descriptor from provided bootloader, code and data. + * + * @bl: bootloader image, including 16-bytes descriptor + * @code: LS firmware code segment + * @data: LS firmware data segment + * @desc: ucode descriptor to be written + * + * Return: allocated ucode image with corresponding descriptor information. desc + * is also updated to contain the right offsets within returned image. + */ +static void * +ls_ucode_img_build(const struct firmware *bl, const struct firmware *code, + const struct firmware *data, struct ls_ucode_img_desc *desc) +{ + struct fw_bin_header *bin_hdr = (void *)bl->data; + struct fw_bl_desc *bl_desc = (void *)bl->data + bin_hdr->header_offset; + void *bl_data = (void *)bl->data + bin_hdr->data_offset; + u32 pos = 0; + void *image; + + desc->bootloader_start_offset = pos; + desc->bootloader_size = ALIGN(bl_desc->code_size, sizeof(u32)); + desc->bootloader_imem_offset = bl_desc->start_tag * 256; + desc->bootloader_entry_point = bl_desc->start_tag * 256; + + pos = ALIGN(pos + desc->bootloader_size, BL_DESC_BLK_SIZE); + desc->app_start_offset = pos; + desc->app_size = ALIGN(code->size, BL_DESC_BLK_SIZE) + + ALIGN(data->size, BL_DESC_BLK_SIZE); + desc->app_imem_offset = 0; + desc->app_imem_entry = 0; + desc->app_dmem_offset = 0; + desc->app_resident_code_offset = 0; + desc->app_resident_code_size = ALIGN(code->size, BL_DESC_BLK_SIZE); + + pos = ALIGN(pos + desc->app_resident_code_size, BL_DESC_BLK_SIZE); + desc->app_resident_data_offset = pos - desc->app_start_offset; + desc->app_resident_data_size = ALIGN(data->size, BL_DESC_BLK_SIZE); + + desc->image_size = ALIGN(bl_desc->code_size, BL_DESC_BLK_SIZE) + + desc->app_size; + + image = kzalloc(desc->image_size, GFP_KERNEL); + if (!image) + return ERR_PTR(-ENOMEM); + + memcpy(image + desc->bootloader_start_offset, bl_data, + bl_desc->code_size); + memcpy(image + desc->app_start_offset, code->data, code->size); + memcpy(image + desc->app_start_offset + desc->app_resident_data_offset, + data->data, data->size); + + return image; +} + +/** + * ls_ucode_img_load_gr() - load and prepare a LS GR ucode image + * + * Load the LS microcode, bootloader and signature and pack them into a single + * blob. Also generate the corresponding ucode descriptor. + */ +static int +ls_ucode_img_load_gr(const struct nvkm_subdev *subdev, struct ls_ucode_img *img, + const char *falcon_name) +{ + const struct firmware *bl, *code, *data, *sig; + char f[64]; + int ret; + + snprintf(f, sizeof(f), "gr/%s_bl", falcon_name); + ret = nvkm_firmware_get(subdev->device, f, &bl); + if (ret) + goto error; + + snprintf(f, sizeof(f), "gr/%s_inst", falcon_name); + ret = nvkm_firmware_get(subdev->device, f, &code); + if (ret) + goto free_bl; + + snprintf(f, sizeof(f), "gr/%s_data", falcon_name); + ret = nvkm_firmware_get(subdev->device, f, &data); + if (ret) + goto free_inst; + + snprintf(f, sizeof(f), "gr/%s_sig", falcon_name); + ret = nvkm_firmware_get(subdev->device, f, &sig); + if (ret) + goto free_data; + img->sig = kmemdup(sig->data, sig->size, GFP_KERNEL); + if (!img->sig) { + ret = -ENOMEM; + goto free_sig; + } + img->sig_size = sig->size; + + img->ucode_data = ls_ucode_img_build(bl, code, data, + &img->ucode_desc); + if (IS_ERR(img->ucode_data)) { + ret = PTR_ERR(img->ucode_data); + goto free_data; + } + img->ucode_size = img->ucode_desc.image_size; + +free_sig: + nvkm_firmware_put(sig); +free_data: + nvkm_firmware_put(data); +free_inst: + nvkm_firmware_put(code); +free_bl: + nvkm_firmware_put(bl); +error: + return ret; +} + +int +acr_ls_ucode_load_fecs(const struct nvkm_subdev *subdev, + struct ls_ucode_img *img) +{ + return ls_ucode_img_load_gr(subdev, img, "fecs"); +} + +int +acr_ls_ucode_load_gpccs(const struct nvkm_subdev *subdev, + struct ls_ucode_img *img) +{ + return ls_ucode_img_load_gr(subdev, img, "gpccs"); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h index a9a8a0e1017e..936a65f5658c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h @@ -27,20 +27,16 @@ #include struct nvkm_secboot_func { - int (*init)(struct nvkm_secboot *); + int (*oneinit)(struct nvkm_secboot *); int (*fini)(struct nvkm_secboot *, bool suspend); void *(*dtor)(struct nvkm_secboot *); - int (*reset)(struct nvkm_secboot *, enum nvkm_secboot_falcon); - int (*start)(struct nvkm_secboot *, enum nvkm_secboot_falcon); - - /* ID of the falcon that will perform secure boot */ - enum nvkm_secboot_falcon boot_falcon; - /* Bit-mask of IDs of managed falcons */ - unsigned long managed_falcons; + int (*run_blob)(struct nvkm_secboot *, struct nvkm_gpuobj *); }; -int nvkm_secboot_ctor(const struct nvkm_secboot_func *, struct nvkm_device *, - int index, struct nvkm_secboot *); +extern const char *nvkm_secboot_falcon_name[]; + +int nvkm_secboot_ctor(const struct nvkm_secboot_func *, struct nvkm_acr *, + struct nvkm_device *, int, struct nvkm_secboot *); int nvkm_secboot_falcon_reset(struct nvkm_secboot *); int nvkm_secboot_falcon_run(struct nvkm_secboot *); @@ -48,187 +44,20 @@ struct flcn_u64 { u32 lo; u32 hi; }; + static inline u64 flcn64_to_u64(const struct flcn_u64 f) { return ((u64)f.hi) << 32 | f.lo; } -/** - * struct gm200_flcn_bl_desc - DMEM bootloader descriptor - * @signature: 16B signature for secure code. 0s if no secure code - * @ctx_dma: DMA context to be used by BL while loading code/data - * @code_dma_base: 256B-aligned Physical FB Address where code is located - * (falcon's $xcbase register) - * @non_sec_code_off: offset from code_dma_base where the non-secure code is - * located. The offset must be multiple of 256 to help perf - * @non_sec_code_size: the size of the nonSecure code part. - * @sec_code_off: offset from code_dma_base where the secure code is - * located. The offset must be multiple of 256 to help perf - * @sec_code_size: offset from code_dma_base where the secure code is - * located. The offset must be multiple of 256 to help perf - * @code_entry_point: code entry point which will be invoked by BL after - * code is loaded. - * @data_dma_base: 256B aligned Physical FB Address where data is located. - * (falcon's $xdbase register) - * @data_size: size of data block. Should be multiple of 256B - * - * Structure used by the bootloader to load the rest of the code. This has - * to be filled by host and copied into DMEM at offset provided in the - * hsflcn_bl_desc.bl_desc_dmem_load_off. - */ -struct gm200_flcn_bl_desc { - u32 reserved[4]; - u32 signature[4]; - u32 ctx_dma; - struct flcn_u64 code_dma_base; - u32 non_sec_code_off; - u32 non_sec_code_size; - u32 sec_code_off; - u32 sec_code_size; - u32 code_entry_point; - struct flcn_u64 data_dma_base; - u32 data_size; -}; +static inline struct flcn_u64 u64_to_flcn64(u64 u) +{ + struct flcn_u64 ret; -/** - * struct hsflcn_acr_desc - data section of the HS firmware - * - * This header is to be copied at the beginning of DMEM by the HS bootloader. - * - * @signature: signature of ACR ucode - * @wpr_region_id: region ID holding the WPR header and its details - * @wpr_offset: offset from the WPR region holding the wpr header - * @regions: region descriptors - * @nonwpr_ucode_blob_size: size of LS blob - * @nonwpr_ucode_blob_start: FB location of LS blob is - */ -struct hsflcn_acr_desc { - union { - u8 reserved_dmem[0x200]; - u32 signatures[4]; - } ucode_reserved_space; - u32 wpr_region_id; - u32 wpr_offset; - u32 mmu_mem_range; -#define FLCN_ACR_MAX_REGIONS 2 - struct { - u32 no_regions; - struct { - u32 start_addr; - u32 end_addr; - u32 region_id; - u32 read_mask; - u32 write_mask; - u32 client_mask; - } region_props[FLCN_ACR_MAX_REGIONS]; - } regions; - u32 ucode_blob_size; - u64 ucode_blob_base __aligned(8); - struct { - u32 vpr_enabled; - u32 vpr_start; - u32 vpr_end; - u32 hdcp_policies; - } vpr_desc; -}; + ret.hi = upper_32_bits(u); + ret.lo = lower_32_bits(u); -/** - * Contains the whole secure boot state, allowing it to be performed as needed - * @wpr_addr: physical address of the WPR region - * @wpr_size: size in bytes of the WPR region - * @ls_blob: LS blob of all the LS firmwares, signatures, bootloaders - * @ls_blob_size: size of the LS blob - * @ls_blob_nb_regions: number of LS firmwares that will be loaded - * @acr_blob: HS blob - * @acr_blob_vma: mapping of the HS blob into the secure falcon's VM - * @acr_bl_desc: bootloader descriptor of the HS blob - * @hsbl_blob: HS blob bootloader - * @inst: instance block for HS falcon - * @pgd: page directory for the HS falcon - * @vm: address space used by the HS falcon - * @falcon_state: current state of the managed falcons - * @firmware_ok: whether the firmware blobs have been created - */ -struct gm200_secboot { - struct nvkm_secboot base; - const struct gm200_secboot_func *func; - - /* - * Address and size of the WPR region. On dGPU this will be the - * address of the LS blob. On Tegra this is a fixed region set by the - * bootloader - */ - u64 wpr_addr; - u32 wpr_size; - - /* - * HS FW - lock WPR region (dGPU only) and load LS FWs - * on Tegra the HS FW copies the LS blob into the fixed WPR instead - */ - struct nvkm_gpuobj *acr_load_blob; - struct gm200_flcn_bl_desc acr_load_bl_desc; - - /* HS FW - unlock WPR region (dGPU only) */ - struct nvkm_gpuobj *acr_unload_blob; - struct gm200_flcn_bl_desc acr_unload_bl_desc; - - /* HS bootloader */ - void *hsbl_blob; - - /* LS FWs, to be loaded by the HS ACR */ - struct nvkm_gpuobj *ls_blob; - - /* Instance block & address space used for HS FW execution */ - struct nvkm_gpuobj *inst; - struct nvkm_gpuobj *pgd; - struct nvkm_vm *vm; - - /* To keep track of the state of all managed falcons */ - enum { - /* In non-secure state, no firmware loaded, no privileges*/ - NON_SECURE = 0, - /* In low-secure mode and ready to be started */ - RESET, - /* In low-secure mode and running */ - RUNNING, - } falcon_state[NVKM_SECBOOT_FALCON_END]; - - bool firmware_ok; -}; -#define gm200_secboot(sb) container_of(sb, struct gm200_secboot, base) - -/** - * Contains functions we wish to abstract between GM200-like implementations - * @bl_desc_size: size of the BL descriptor used by this chip. - * @fixup_bl_desc: hook that generates the proper BL descriptor format from - * the generic GM200 format into a data array of size - * bl_desc_size - * @fixup_hs_desc: hook that twiddles the HS descriptor before it is used - * @prepare_blobs: prepares the various blobs needed for secure booting - */ -struct gm200_secboot_func { - /* - * Size of the bootloader descriptor for this chip. A block of this - * size is allocated before booting a falcon and the fixup_bl_desc - * callback is called on it - */ - u32 bl_desc_size; - void (*fixup_bl_desc)(const struct gm200_flcn_bl_desc *, void *); - - /* - * Chip-specific modifications of the HS descriptor can be done here. - * On dGPU this is used to fill the information about the WPR region - * we want the HS FW to set up. - */ - void (*fixup_hs_desc)(struct gm200_secboot *, struct hsflcn_acr_desc *); - int (*prepare_blobs)(struct gm200_secboot *); -}; - -int gm200_secboot_init(struct nvkm_secboot *); -void *gm200_secboot_dtor(struct nvkm_secboot *); -int gm200_secboot_reset(struct nvkm_secboot *, u32); -int gm200_secboot_start(struct nvkm_secboot *, u32); - -int gm20x_secboot_prepare_blobs(struct gm200_secboot *); + return ret; +} #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c index 8894fee30cbc..df949fa7d05d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c @@ -64,10 +64,9 @@ nvkm_therm_update_trip(struct nvkm_therm *therm) } static int -nvkm_therm_update_linear(struct nvkm_therm *therm) +nvkm_therm_compute_linear_duty(struct nvkm_therm *therm, u8 linear_min_temp, + u8 linear_max_temp) { - u8 linear_min_temp = therm->fan->bios.linear_min_temp; - u8 linear_max_temp = therm->fan->bios.linear_max_temp; u8 temp = therm->func->temp_get(therm); u16 duty; @@ -85,6 +84,21 @@ nvkm_therm_update_linear(struct nvkm_therm *therm) return duty; } +static int +nvkm_therm_update_linear(struct nvkm_therm *therm) +{ + u8 min = therm->fan->bios.linear_min_temp; + u8 max = therm->fan->bios.linear_max_temp; + return nvkm_therm_compute_linear_duty(therm, min, max); +} + +static int +nvkm_therm_update_linear_fallback(struct nvkm_therm *therm) +{ + u8 max = therm->bios_sensor.thrs_fan_boost.temp; + return nvkm_therm_compute_linear_duty(therm, 30, max); +} + static void nvkm_therm_update(struct nvkm_therm *therm, int mode) { @@ -119,6 +133,8 @@ nvkm_therm_update(struct nvkm_therm *therm, int mode) case NVBIOS_THERM_FAN_OTHER: if (therm->cstate) duty = therm->cstate; + else + duty = nvkm_therm_update_linear_fallback(therm); poll = false; break; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c index fe063d5728e2..67ada1d9a28c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c @@ -95,6 +95,20 @@ nvkm_top_intr(struct nvkm_device *device, u32 intr, u64 *psubdevs) return intr & ~handled; } +int +nvkm_top_fault_id(struct nvkm_device *device, enum nvkm_devidx devidx) +{ + struct nvkm_top *top = device->top; + struct nvkm_top_device *info; + + list_for_each_entry(info, &top->device, head) { + if (info->index == devidx && info->fault >= 0) + return info->fault; + } + + return -ENOENT; +} + enum nvkm_devidx nvkm_top_fault(struct nvkm_device *device, int fault) { diff --git a/drivers/gpu/drm/omapdrm/dss/dispc.c b/drivers/gpu/drm/omapdrm/dss/dispc.c index 5554b72cf56a..d956e6266368 100644 --- a/drivers/gpu/drm/omapdrm/dss/dispc.c +++ b/drivers/gpu/drm/omapdrm/dss/dispc.c @@ -2506,6 +2506,25 @@ static int dispc_ovl_calc_scaling_44xx(unsigned long pclk, unsigned long lclk, return -EINVAL; } + if (*decim_x > 4 && color_mode != OMAP_DSS_COLOR_NV12) { + /* + * Let's disable all scaling that requires horizontal + * decimation with higher factor than 4, until we have + * better estimates of what we can and can not + * do. However, NV12 color format appears to work Ok + * with all decimation factors. + * + * When decimating horizontally by more that 4 the dss + * is not able to fetch the data in burst mode. When + * this happens it is hard to tell if there enough + * bandwidth. Despite what theory says this appears to + * be true also for 16-bit color formats. + */ + DSSERR("Not enough bandwidth, too much downscaling (x-decimation factor %d > 4)", *decim_x); + + return -EINVAL; + } + *core_clk = dispc.feat->calc_core_clk(pclk, in_width, in_height, out_width, out_height, mem_to_mem); return 0; diff --git a/drivers/gpu/drm/omapdrm/omap_crtc.c b/drivers/gpu/drm/omapdrm/omap_crtc.c index dd47dc191e6b..b68c70eb395f 100644 --- a/drivers/gpu/drm/omapdrm/omap_crtc.c +++ b/drivers/gpu/drm/omapdrm/omap_crtc.c @@ -410,13 +410,7 @@ static void omap_crtc_atomic_flush(struct drm_crtc *crtc, dispc_mgr_set_gamma(omap_crtc->channel, lut, length); } - /* - * Only flush the CRTC if it is currently enabled. CRTCs that require a - * mode set are disabled prior plane updates and enabled afterwards. - * They are thus not active (regardless of what their CRTC core state - * reports) and the DRM core could thus call this function even though - * the CRTC is currently disabled. Do nothing in that case. - */ + /* Only flush the CRTC if it is currently enabled. */ if (!omap_crtc->enabled) return; diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c index afe8f05b927b..3f2554235225 100644 --- a/drivers/gpu/drm/omapdrm/omap_drv.c +++ b/drivers/gpu/drm/omapdrm/omap_drv.c @@ -96,9 +96,22 @@ static void omap_atomic_complete(struct omap_atomic_state_commit *commit) dispc_runtime_get(); drm_atomic_helper_commit_modeset_disables(dev, old_state); - drm_atomic_helper_commit_planes(dev, old_state, - DRM_PLANE_COMMIT_ACTIVE_ONLY); + + /* With the current dss dispc implementation we have to enable + * the new modeset before we can commit planes. The dispc ovl + * configuration relies on the video mode configuration been + * written into the HW when the ovl configuration is + * calculated. + * + * This approach is not ideal because after a mode change the + * plane update is executed only after the first vblank + * interrupt. The dispc implementation should be fixed so that + * it is able use uncommitted drm state information. + */ drm_atomic_helper_commit_modeset_enables(dev, old_state); + omap_atomic_wait_for_completion(dev, old_state); + + drm_atomic_helper_commit_planes(dev, old_state, 0); omap_atomic_wait_for_completion(dev, old_state); diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c index 06aaf79de8c8..89eb0422821c 100644 --- a/drivers/gpu/drm/panel/panel-simple.c +++ b/drivers/gpu/drm/panel/panel-simple.c @@ -668,6 +668,48 @@ static const struct panel_desc avic_tm070ddh03 = { }, }; +static const struct drm_display_mode boe_nv101wxmn51_modes[] = { + { + .clock = 71900, + .hdisplay = 1280, + .hsync_start = 1280 + 48, + .hsync_end = 1280 + 48 + 32, + .htotal = 1280 + 48 + 32 + 80, + .vdisplay = 800, + .vsync_start = 800 + 3, + .vsync_end = 800 + 3 + 5, + .vtotal = 800 + 3 + 5 + 24, + .vrefresh = 60, + }, + { + .clock = 57500, + .hdisplay = 1280, + .hsync_start = 1280 + 48, + .hsync_end = 1280 + 48 + 32, + .htotal = 1280 + 48 + 32 + 80, + .vdisplay = 800, + .vsync_start = 800 + 3, + .vsync_end = 800 + 3 + 5, + .vtotal = 800 + 3 + 5 + 24, + .vrefresh = 48, + }, +}; + +static const struct panel_desc boe_nv101wxmn51 = { + .modes = boe_nv101wxmn51_modes, + .num_modes = ARRAY_SIZE(boe_nv101wxmn51_modes), + .bpc = 8, + .size = { + .width = 217, + .height = 136, + }, + .delay = { + .prepare = 210, + .enable = 50, + .unprepare = 160, + }, +}; + static const struct drm_display_mode chunghwa_claa070wp03xg_mode = { .clock = 66770, .hdisplay = 800, @@ -760,6 +802,8 @@ static const struct panel_desc edt_et057090dhu = { .width = 115, .height = 86, }, + .bus_format = MEDIA_BUS_FMT_RGB666_1X18, + .bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_NEGEDGE, }; static const struct drm_display_mode edt_etm0700g0dh6_mode = { @@ -784,6 +828,8 @@ static const struct panel_desc edt_etm0700g0dh6 = { .width = 152, .height = 91, }, + .bus_format = MEDIA_BUS_FMT_RGB666_1X18, + .bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_NEGEDGE, }; static const struct drm_display_mode foxlink_fl500wvr00_a0t_mode = { @@ -1277,6 +1323,29 @@ static const struct panel_desc nec_nl4827hc19_05b = { .bus_flags = DRM_BUS_FLAG_PIXDATA_POSEDGE, }; +static const struct drm_display_mode netron_dy_e231732_mode = { + .clock = 66000, + .hdisplay = 1024, + .hsync_start = 1024 + 160, + .hsync_end = 1024 + 160 + 70, + .htotal = 1024 + 160 + 70 + 90, + .vdisplay = 600, + .vsync_start = 600 + 127, + .vsync_end = 600 + 127 + 20, + .vtotal = 600 + 127 + 20 + 3, + .vrefresh = 60, +}; + +static const struct panel_desc netron_dy_e231732 = { + .modes = &netron_dy_e231732_mode, + .num_modes = 1, + .size = { + .width = 154, + .height = 87, + }, + .bus_format = MEDIA_BUS_FMT_RGB666_1X18, +}; + static const struct drm_display_mode nvd_9128_mode = { .clock = 29500, .hdisplay = 800, @@ -1632,6 +1701,30 @@ static const struct panel_desc starry_kr122ea0sra = { }, }; +static const struct display_timing tianma_tm070jdhg30_timing = { + .pixelclock = { 62600000, 68200000, 78100000 }, + .hactive = { 1280, 1280, 1280 }, + .hfront_porch = { 15, 64, 159 }, + .hback_porch = { 5, 5, 5 }, + .hsync_len = { 1, 1, 256 }, + .vactive = { 800, 800, 800 }, + .vfront_porch = { 3, 40, 99 }, + .vback_porch = { 2, 2, 2 }, + .vsync_len = { 1, 1, 128 }, + .flags = DISPLAY_FLAGS_DE_HIGH, +}; + +static const struct panel_desc tianma_tm070jdhg30 = { + .timings = &tianma_tm070jdhg30_timing, + .num_timings = 1, + .bpc = 8, + .size = { + .width = 151, + .height = 95, + }, + .bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG, +}; + static const struct drm_display_mode tpk_f07a_0102_mode = { .clock = 33260, .hdisplay = 800, @@ -1747,6 +1840,9 @@ static const struct of_device_id platform_of_match[] = { }, { .compatible = "avic,tm070ddh03", .data = &avic_tm070ddh03, + }, { + .compatible = "boe,nv101wxmn51", + .data = &boe_nv101wxmn51, }, { .compatible = "chunghwa,claa070wp03xg", .data = &chunghwa_claa070wp03xg, @@ -1825,6 +1921,9 @@ static const struct of_device_id platform_of_match[] = { }, { .compatible = "nec,nl4827hc19-05b", .data = &nec_nl4827hc19_05b, + }, { + .compatible = "netron-dy,e231732", + .data = &netron_dy_e231732, }, { .compatible = "nvd,9128", .data = &nvd_9128, @@ -1867,6 +1966,9 @@ static const struct of_device_id platform_of_match[] = { }, { .compatible = "starry,kr122ea0sra", .data = &starry_kr122ea0sra, + }, { + .compatible = "tianma,tm070jdhg30", + .data = &tianma_tm070jdhg30, }, { .compatible = "tpk,f07a-0102", .data = &tpk_f07a_0102, diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index af3bbe82fd48..956c425e639e 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -97,9 +97,10 @@ * 2.46.0 - Add PFP_SYNC_ME support on evergreen * 2.47.0 - Add UVD_NO_OP register support * 2.48.0 - TA_CS_BC_BASE_ADDR allowed on SI + * 2.49.0 - DRM_RADEON_GEM_INFO ioctl returns correct vram_size/visible values */ #define KMS_DRIVER_MAJOR 2 -#define KMS_DRIVER_MINOR 48 +#define KMS_DRIVER_MINOR 49 #define KMS_DRIVER_PATCHLEVEL 0 int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags); void radeon_driver_unload_kms(struct drm_device *dev); @@ -366,11 +367,10 @@ static void radeon_pci_shutdown(struct pci_dev *pdev) { /* if we are running in a VM, make sure the device - * torn down properly on reboot/shutdown. - * unfortunately we can't detect certain - * hypervisors so just do this all the time. + * torn down properly on reboot/shutdown */ - radeon_pci_remove(pdev); + if (radeon_device_is_virtual()) + radeon_pci_remove(pdev); } static int radeon_pmops_suspend(struct device *dev) diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index 0bcffd8a7bd3..96683f5b2b1b 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -220,8 +220,8 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data, man = &rdev->mman.bdev.man[TTM_PL_VRAM]; - args->vram_size = rdev->mc.real_vram_size; - args->vram_visible = (u64)man->size << PAGE_SHIFT; + args->vram_size = (u64)man->size << PAGE_SHIFT; + args->vram_visible = rdev->mc.visible_vram_size; args->vram_visible -= rdev->vram_pin_size; args->gart_size = rdev->mc.gtt_size; args->gart_size -= rdev->gart_pin_size; diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c index c30d649cb147..b360e6251836 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c +++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c @@ -14,19 +14,19 @@ * GNU General Public License for more details. */ -#include - #include #include #include #include #include #include +#include #include #include #include #include #include +#include #include "rockchip_drm_drv.h" #include "rockchip_drm_fb.h" @@ -50,28 +50,31 @@ static struct drm_driver rockchip_drm_driver; int rockchip_drm_dma_attach_device(struct drm_device *drm_dev, struct device *dev) { - struct dma_iommu_mapping *mapping = drm_dev->dev->archdata.mapping; + struct rockchip_drm_private *private = drm_dev->dev_private; int ret; if (!is_support_iommu) return 0; - ret = dma_set_coherent_mask(dev, DMA_BIT_MASK(32)); - if (ret) + ret = iommu_attach_device(private->domain, dev); + if (ret) { + dev_err(dev, "Failed to attach iommu device\n"); return ret; + } - dma_set_max_seg_size(dev, DMA_BIT_MASK(32)); - - return arm_iommu_attach_device(dev, mapping); + return 0; } void rockchip_drm_dma_detach_device(struct drm_device *drm_dev, struct device *dev) { + struct rockchip_drm_private *private = drm_dev->dev_private; + struct iommu_domain *domain = private->domain; + if (!is_support_iommu) return; - arm_iommu_detach_device(dev); + iommu_detach_device(domain, dev); } int rockchip_register_crtc_funcs(struct drm_crtc *crtc, @@ -123,11 +126,46 @@ static void rockchip_drm_crtc_disable_vblank(struct drm_device *dev, priv->crtc_funcs[pipe]->disable_vblank(crtc); } +static int rockchip_drm_init_iommu(struct drm_device *drm_dev) +{ + struct rockchip_drm_private *private = drm_dev->dev_private; + struct iommu_domain_geometry *geometry; + u64 start, end; + + if (!is_support_iommu) + return 0; + + private->domain = iommu_domain_alloc(&platform_bus_type); + if (!private->domain) + return -ENOMEM; + + geometry = &private->domain->geometry; + start = geometry->aperture_start; + end = geometry->aperture_end; + + DRM_DEBUG("IOMMU context initialized (aperture: %#llx-%#llx)\n", + start, end); + drm_mm_init(&private->mm, start, end - start + 1); + mutex_init(&private->mm_lock); + + return 0; +} + +static void rockchip_iommu_cleanup(struct drm_device *drm_dev) +{ + struct rockchip_drm_private *private = drm_dev->dev_private; + + if (!is_support_iommu) + return; + + drm_mm_takedown(&private->mm); + iommu_domain_free(private->domain); +} + static int rockchip_drm_bind(struct device *dev) { struct drm_device *drm_dev; struct rockchip_drm_private *private; - struct dma_iommu_mapping *mapping = NULL; int ret; drm_dev = drm_dev_alloc(&rockchip_drm_driver, dev); @@ -151,38 +189,14 @@ static int rockchip_drm_bind(struct device *dev) rockchip_drm_mode_config_init(drm_dev); - dev->dma_parms = devm_kzalloc(dev, sizeof(*dev->dma_parms), - GFP_KERNEL); - if (!dev->dma_parms) { - ret = -ENOMEM; + ret = rockchip_drm_init_iommu(drm_dev); + if (ret) goto err_config_cleanup; - } - - if (is_support_iommu) { - /* TODO(djkurtz): fetch the mapping start/size from somewhere */ - mapping = arm_iommu_create_mapping(&platform_bus_type, - 0x00000000, - SZ_2G); - if (IS_ERR(mapping)) { - ret = PTR_ERR(mapping); - goto err_config_cleanup; - } - - ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(32)); - if (ret) - goto err_release_mapping; - - dma_set_max_seg_size(dev, DMA_BIT_MASK(32)); - - ret = arm_iommu_attach_device(dev, mapping); - if (ret) - goto err_release_mapping; - } /* Try to bind all sub drivers. */ ret = component_bind_all(dev, drm_dev); if (ret) - goto err_detach_device; + goto err_iommu_cleanup; /* init kms poll for handling hpd */ drm_kms_helper_poll_init(drm_dev); @@ -207,8 +221,6 @@ static int rockchip_drm_bind(struct device *dev) if (ret) goto err_fbdev_fini; - if (is_support_iommu) - arm_iommu_release_mapping(mapping); return 0; err_fbdev_fini: rockchip_drm_fbdev_fini(drm_dev); @@ -217,12 +229,8 @@ static int rockchip_drm_bind(struct device *dev) err_kms_helper_poll_fini: drm_kms_helper_poll_fini(drm_dev); component_unbind_all(dev, drm_dev); -err_detach_device: - if (is_support_iommu) - arm_iommu_detach_device(dev); -err_release_mapping: - if (is_support_iommu) - arm_iommu_release_mapping(mapping); +err_iommu_cleanup: + rockchip_iommu_cleanup(drm_dev); err_config_cleanup: drm_mode_config_cleanup(drm_dev); drm_dev->dev_private = NULL; @@ -239,8 +247,7 @@ static void rockchip_drm_unbind(struct device *dev) drm_vblank_cleanup(drm_dev); drm_kms_helper_poll_fini(drm_dev); component_unbind_all(dev, drm_dev); - if (is_support_iommu) - arm_iommu_detach_device(dev); + rockchip_iommu_cleanup(drm_dev); drm_mode_config_cleanup(drm_dev); drm_dev->dev_private = NULL; drm_dev_unregister(drm_dev); diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h index fb6226cf84b7..adc39302bec5 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h +++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h @@ -30,6 +30,7 @@ struct drm_device; struct drm_connector; +struct iommu_domain; /* * Rockchip drm private crtc funcs. @@ -60,7 +61,10 @@ struct rockchip_drm_private { struct drm_gem_object *fbdev_bo; const struct rockchip_crtc_funcs *crtc_funcs[ROCKCHIP_MAX_CRTC]; struct drm_atomic_state *state; - + struct iommu_domain *domain; + /* protect drm_mm on multi-threads */ + struct mutex mm_lock; + struct drm_mm mm; struct list_head psr_list; spinlock_t psr_list_lock; }; diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c index b70f9423379c..df9e57064f19 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c +++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c @@ -16,11 +16,146 @@ #include #include #include +#include #include "rockchip_drm_drv.h" #include "rockchip_drm_gem.h" -static int rockchip_gem_alloc_buf(struct rockchip_gem_object *rk_obj, +static int rockchip_gem_iommu_map(struct rockchip_gem_object *rk_obj) +{ + struct drm_device *drm = rk_obj->base.dev; + struct rockchip_drm_private *private = drm->dev_private; + int prot = IOMMU_READ | IOMMU_WRITE; + ssize_t ret; + + mutex_lock(&private->mm_lock); + + ret = drm_mm_insert_node_generic(&private->mm, &rk_obj->mm, + rk_obj->base.size, PAGE_SIZE, + 0, 0); + + mutex_unlock(&private->mm_lock); + if (ret < 0) { + DRM_ERROR("out of I/O virtual memory: %zd\n", ret); + return ret; + } + + rk_obj->dma_addr = rk_obj->mm.start; + + ret = iommu_map_sg(private->domain, rk_obj->dma_addr, rk_obj->sgt->sgl, + rk_obj->sgt->nents, prot); + if (ret < rk_obj->base.size) { + DRM_ERROR("failed to map buffer: size=%zd request_size=%zd\n", + ret, rk_obj->base.size); + ret = -ENOMEM; + goto err_remove_node; + } + + rk_obj->size = ret; + + return 0; + +err_remove_node: + drm_mm_remove_node(&rk_obj->mm); + + return ret; +} + +static int rockchip_gem_iommu_unmap(struct rockchip_gem_object *rk_obj) +{ + struct drm_device *drm = rk_obj->base.dev; + struct rockchip_drm_private *private = drm->dev_private; + + iommu_unmap(private->domain, rk_obj->dma_addr, rk_obj->size); + + mutex_lock(&private->mm_lock); + + drm_mm_remove_node(&rk_obj->mm); + + mutex_unlock(&private->mm_lock); + + return 0; +} + +static int rockchip_gem_get_pages(struct rockchip_gem_object *rk_obj) +{ + struct drm_device *drm = rk_obj->base.dev; + int ret, i; + struct scatterlist *s; + + rk_obj->pages = drm_gem_get_pages(&rk_obj->base); + if (IS_ERR(rk_obj->pages)) + return PTR_ERR(rk_obj->pages); + + rk_obj->num_pages = rk_obj->base.size >> PAGE_SHIFT; + + rk_obj->sgt = drm_prime_pages_to_sg(rk_obj->pages, rk_obj->num_pages); + if (IS_ERR(rk_obj->sgt)) { + ret = PTR_ERR(rk_obj->sgt); + goto err_put_pages; + } + + /* + * Fake up the SG table so that dma_sync_sg_for_device() can be used + * to flush the pages associated with it. + * + * TODO: Replace this by drm_clflush_sg() once it can be implemented + * without relying on symbols that are not exported. + */ + for_each_sg(rk_obj->sgt->sgl, s, rk_obj->sgt->nents, i) + sg_dma_address(s) = sg_phys(s); + + dma_sync_sg_for_device(drm->dev, rk_obj->sgt->sgl, rk_obj->sgt->nents, + DMA_TO_DEVICE); + + return 0; + +err_put_pages: + drm_gem_put_pages(&rk_obj->base, rk_obj->pages, false, false); + return ret; +} + +static void rockchip_gem_put_pages(struct rockchip_gem_object *rk_obj) +{ + sg_free_table(rk_obj->sgt); + kfree(rk_obj->sgt); + drm_gem_put_pages(&rk_obj->base, rk_obj->pages, true, true); +} + +static int rockchip_gem_alloc_iommu(struct rockchip_gem_object *rk_obj, + bool alloc_kmap) +{ + int ret; + + ret = rockchip_gem_get_pages(rk_obj); + if (ret < 0) + return ret; + + ret = rockchip_gem_iommu_map(rk_obj); + if (ret < 0) + goto err_free; + + if (alloc_kmap) { + rk_obj->kvaddr = vmap(rk_obj->pages, rk_obj->num_pages, VM_MAP, + pgprot_writecombine(PAGE_KERNEL)); + if (!rk_obj->kvaddr) { + DRM_ERROR("failed to vmap() buffer\n"); + ret = -ENOMEM; + goto err_unmap; + } + } + + return 0; + +err_unmap: + rockchip_gem_iommu_unmap(rk_obj); +err_free: + rockchip_gem_put_pages(rk_obj); + + return ret; +} + +static int rockchip_gem_alloc_dma(struct rockchip_gem_object *rk_obj, bool alloc_kmap) { struct drm_gem_object *obj = &rk_obj->base; @@ -42,7 +177,27 @@ static int rockchip_gem_alloc_buf(struct rockchip_gem_object *rk_obj, return 0; } -static void rockchip_gem_free_buf(struct rockchip_gem_object *rk_obj) +static int rockchip_gem_alloc_buf(struct rockchip_gem_object *rk_obj, + bool alloc_kmap) +{ + struct drm_gem_object *obj = &rk_obj->base; + struct drm_device *drm = obj->dev; + struct rockchip_drm_private *private = drm->dev_private; + + if (private->domain) + return rockchip_gem_alloc_iommu(rk_obj, alloc_kmap); + else + return rockchip_gem_alloc_dma(rk_obj, alloc_kmap); +} + +static void rockchip_gem_free_iommu(struct rockchip_gem_object *rk_obj) +{ + vunmap(rk_obj->kvaddr); + rockchip_gem_iommu_unmap(rk_obj); + rockchip_gem_put_pages(rk_obj); +} + +static void rockchip_gem_free_dma(struct rockchip_gem_object *rk_obj) { struct drm_gem_object *obj = &rk_obj->base; struct drm_device *drm = obj->dev; @@ -51,23 +206,68 @@ static void rockchip_gem_free_buf(struct rockchip_gem_object *rk_obj) rk_obj->dma_attrs); } -static int rockchip_drm_gem_object_mmap(struct drm_gem_object *obj, - struct vm_area_struct *vma) - +static void rockchip_gem_free_buf(struct rockchip_gem_object *rk_obj) { + if (rk_obj->pages) + rockchip_gem_free_iommu(rk_obj); + else + rockchip_gem_free_dma(rk_obj); +} + +static int rockchip_drm_gem_object_mmap_iommu(struct drm_gem_object *obj, + struct vm_area_struct *vma) +{ + struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj); + unsigned int i, count = obj->size >> PAGE_SHIFT; + unsigned long user_count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT; + unsigned long uaddr = vma->vm_start; + unsigned long offset = vma->vm_pgoff; + unsigned long end = user_count + offset; int ret; + + if (user_count == 0) + return -ENXIO; + if (end > count) + return -ENXIO; + + for (i = offset; i < end; i++) { + ret = vm_insert_page(vma, uaddr, rk_obj->pages[i]); + if (ret) + return ret; + uaddr += PAGE_SIZE; + } + + return 0; +} + +static int rockchip_drm_gem_object_mmap_dma(struct drm_gem_object *obj, + struct vm_area_struct *vma) +{ struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj); struct drm_device *drm = obj->dev; + return dma_mmap_attrs(drm->dev, vma, rk_obj->kvaddr, rk_obj->dma_addr, + obj->size, rk_obj->dma_attrs); +} + +static int rockchip_drm_gem_object_mmap(struct drm_gem_object *obj, + struct vm_area_struct *vma) +{ + int ret; + struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj); + /* - * dma_alloc_attrs() allocated a struct page table for rk_obj, so clear + * We allocated a struct page table for rk_obj, so clear * VM_PFNMAP flag that was set by drm_gem_mmap_obj()/drm_gem_mmap(). */ vma->vm_flags &= ~VM_PFNMAP; vma->vm_pgoff = 0; - ret = dma_mmap_attrs(drm->dev, vma, rk_obj->kvaddr, rk_obj->dma_addr, - obj->size, rk_obj->dma_attrs); + if (rk_obj->pages) + ret = rockchip_drm_gem_object_mmap_iommu(obj, vma); + else + ret = rockchip_drm_gem_object_mmap_dma(obj, vma); + if (ret) drm_gem_vm_close(vma); @@ -101,6 +301,12 @@ int rockchip_gem_mmap(struct file *filp, struct vm_area_struct *vma) return rockchip_drm_gem_object_mmap(obj, vma); } +static void rockchip_gem_release_object(struct rockchip_gem_object *rk_obj) +{ + drm_gem_object_release(&rk_obj->base); + kfree(rk_obj); +} + struct rockchip_gem_object * rockchip_gem_create_object(struct drm_device *drm, unsigned int size, bool alloc_kmap) @@ -117,7 +323,7 @@ struct rockchip_gem_object * obj = &rk_obj->base; - drm_gem_private_object_init(drm, obj, size); + drm_gem_object_init(drm, obj, size); ret = rockchip_gem_alloc_buf(rk_obj, alloc_kmap); if (ret) @@ -126,7 +332,7 @@ struct rockchip_gem_object * return rk_obj; err_free_rk_obj: - kfree(rk_obj); + rockchip_gem_release_object(rk_obj); return ERR_PTR(ret); } @@ -138,13 +344,11 @@ void rockchip_gem_free_object(struct drm_gem_object *obj) { struct rockchip_gem_object *rk_obj; - drm_gem_free_mmap_offset(obj); - rk_obj = to_rockchip_obj(obj); rockchip_gem_free_buf(rk_obj); - kfree(rk_obj); + rockchip_gem_release_object(rk_obj); } /* @@ -253,6 +457,9 @@ struct sg_table *rockchip_gem_prime_get_sg_table(struct drm_gem_object *obj) struct sg_table *sgt; int ret; + if (rk_obj->pages) + return drm_prime_pages_to_sg(rk_obj->pages, rk_obj->num_pages); + sgt = kzalloc(sizeof(*sgt), GFP_KERNEL); if (!sgt) return ERR_PTR(-ENOMEM); @@ -273,6 +480,10 @@ void *rockchip_gem_prime_vmap(struct drm_gem_object *obj) { struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj); + if (rk_obj->pages) + return vmap(rk_obj->pages, rk_obj->num_pages, VM_MAP, + pgprot_writecombine(PAGE_KERNEL)); + if (rk_obj->dma_attrs & DMA_ATTR_NO_KERNEL_MAPPING) return NULL; @@ -281,5 +492,12 @@ void *rockchip_gem_prime_vmap(struct drm_gem_object *obj) void rockchip_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr) { - /* Nothing to do */ + struct rockchip_gem_object *rk_obj = to_rockchip_obj(obj); + + if (rk_obj->pages) { + vunmap(vaddr); + return; + } + + /* Nothing to do if allocated by DMA mapping API. */ } diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.h b/drivers/gpu/drm/rockchip/rockchip_drm_gem.h index 18b3488db4ec..3f6ea4d18a5c 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_gem.h +++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.h @@ -23,7 +23,15 @@ struct rockchip_gem_object { void *kvaddr; dma_addr_t dma_addr; + /* Used when IOMMU is disabled */ unsigned long dma_attrs; + + /* Used when IOMMU is enabled */ + struct drm_mm_node mm; + unsigned long num_pages; + struct page **pages; + struct sg_table *sgt; + size_t size; }; struct sg_table *rockchip_gem_prime_get_sg_table(struct drm_gem_object *obj); diff --git a/drivers/gpu/drm/sti/sti_crtc.c b/drivers/gpu/drm/sti/sti_crtc.c index e992bed98dcb..d45a4335df5d 100644 --- a/drivers/gpu/drm/sti/sti_crtc.c +++ b/drivers/gpu/drm/sti/sti_crtc.c @@ -134,21 +134,6 @@ sti_crtc_mode_set_nofb(struct drm_crtc *crtc) sti_crtc_mode_set(crtc, &crtc->state->adjusted_mode); } -static void sti_crtc_atomic_begin(struct drm_crtc *crtc, - struct drm_crtc_state *old_crtc_state) -{ - struct sti_mixer *mixer = to_sti_mixer(crtc); - - if (crtc->state->event) { - crtc->state->event->pipe = drm_crtc_index(crtc); - - WARN_ON(drm_crtc_vblank_get(crtc) != 0); - - mixer->pending_event = crtc->state->event; - crtc->state->event = NULL; - } -} - static void sti_crtc_atomic_flush(struct drm_crtc *crtc, struct drm_crtc_state *old_crtc_state) { @@ -156,6 +141,8 @@ static void sti_crtc_atomic_flush(struct drm_crtc *crtc, struct sti_mixer *mixer = to_sti_mixer(crtc); struct sti_compositor *compo = dev_get_drvdata(mixer->dev); struct drm_plane *p; + struct drm_pending_vblank_event *event; + unsigned long flags; DRM_DEBUG_DRIVER("\n"); @@ -220,13 +207,24 @@ static void sti_crtc_atomic_flush(struct drm_crtc *crtc, break; } } + + event = crtc->state->event; + if (event) { + crtc->state->event = NULL; + + spin_lock_irqsave(&crtc->dev->event_lock, flags); + if (drm_crtc_vblank_get(crtc) == 0) + drm_crtc_arm_vblank_event(crtc, event); + else + drm_crtc_send_vblank_event(crtc, event); + spin_unlock_irqrestore(&crtc->dev->event_lock, flags); + } } static const struct drm_crtc_helper_funcs sti_crtc_helper_funcs = { .enable = sti_crtc_enable, .disable = sti_crtc_disabling, .mode_set_nofb = sti_crtc_mode_set_nofb, - .atomic_begin = sti_crtc_atomic_begin, .atomic_flush = sti_crtc_atomic_flush, }; @@ -250,7 +248,6 @@ int sti_crtc_vblank_cb(struct notifier_block *nb, struct sti_compositor *compo; struct drm_crtc *crtc = data; struct sti_mixer *mixer; - unsigned long flags; struct sti_private *priv; unsigned int pipe; @@ -267,14 +264,6 @@ int sti_crtc_vblank_cb(struct notifier_block *nb, drm_crtc_handle_vblank(crtc); - spin_lock_irqsave(&crtc->dev->event_lock, flags); - if (mixer->pending_event) { - drm_crtc_send_vblank_event(crtc, mixer->pending_event); - drm_crtc_vblank_put(crtc); - mixer->pending_event = NULL; - } - spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - if (mixer->status == STI_MIXER_DISABLING) { struct drm_plane *p; @@ -317,19 +306,12 @@ void sti_crtc_disable_vblank(struct drm_device *drm_dev, unsigned int pipe) struct sti_private *priv = drm_dev->dev_private; struct sti_compositor *compo = priv->compo; struct notifier_block *vtg_vblank_nb = &compo->vtg_vblank_nb[pipe]; - struct drm_crtc *crtc = &compo->mixer[pipe]->drm_crtc; struct sti_vtg *vtg = compo->vtg[pipe]; DRM_DEBUG_DRIVER("\n"); if (sti_vtg_unregister_client(vtg, vtg_vblank_nb)) DRM_DEBUG_DRIVER("Warning: cannot unregister VTG notifier\n"); - - /* free the resources of the pending requests */ - if (compo->mixer[pipe]->pending_event) { - drm_crtc_vblank_put(crtc); - compo->mixer[pipe]->pending_event = NULL; - } } static int sti_crtc_late_register(struct drm_crtc *crtc) diff --git a/drivers/gpu/drm/sti/sti_drv.c b/drivers/gpu/drm/sti/sti_drv.c index 788feed208d7..20fc0fbfa849 100644 --- a/drivers/gpu/drm/sti/sti_drv.c +++ b/drivers/gpu/drm/sti/sti_drv.c @@ -58,7 +58,9 @@ static int sti_drm_fps_set(void *data, u64 val) list_for_each_entry(p, &drm_dev->mode_config.plane_list, head) { struct sti_plane *plane = to_sti_plane(p); + memset(&plane->fps_info, 0, sizeof(plane->fps_info)); plane->fps_info.output = (val >> i) & 1; + i++; } @@ -115,52 +117,6 @@ static int sti_drm_dbg_init(struct drm_minor *minor) return ret; } -static void sti_atomic_schedule(struct sti_private *private, - struct drm_atomic_state *state) -{ - private->commit.state = state; - schedule_work(&private->commit.work); -} - -static void sti_atomic_complete(struct sti_private *private, - struct drm_atomic_state *state) -{ - struct drm_device *drm = private->drm_dev; - - /* - * Everything below can be run asynchronously without the need to grab - * any modeset locks at all under one condition: It must be guaranteed - * that the asynchronous work has either been cancelled (if the driver - * supports it, which at least requires that the framebuffers get - * cleaned up with drm_atomic_helper_cleanup_planes()) or completed - * before the new state gets committed on the software side with - * drm_atomic_helper_swap_state(). - * - * This scheme allows new atomic state updates to be prepared and - * checked in parallel to the asynchronous completion of the previous - * update. Which is important since compositors need to figure out the - * composition of the next frame right after having submitted the - * current layout. - */ - - drm_atomic_helper_commit_modeset_disables(drm, state); - drm_atomic_helper_commit_planes(drm, state, 0); - drm_atomic_helper_commit_modeset_enables(drm, state); - - drm_atomic_helper_wait_for_vblanks(drm, state); - - drm_atomic_helper_cleanup_planes(drm, state); - drm_atomic_state_put(state); -} - -static void sti_atomic_work(struct work_struct *work) -{ - struct sti_private *private = container_of(work, - struct sti_private, commit.work); - - sti_atomic_complete(private, private->commit.state); -} - static int sti_atomic_check(struct drm_device *dev, struct drm_atomic_state *state) { @@ -181,38 +137,6 @@ static int sti_atomic_check(struct drm_device *dev, return ret; } -static int sti_atomic_commit(struct drm_device *drm, - struct drm_atomic_state *state, bool nonblock) -{ - struct sti_private *private = drm->dev_private; - int err; - - err = drm_atomic_helper_prepare_planes(drm, state); - if (err) - return err; - - /* serialize outstanding nonblocking commits */ - mutex_lock(&private->commit.lock); - flush_work(&private->commit.work); - - /* - * This is the point of no return - everything below never fails except - * when the hw goes bonghits. Which means we can commit the new state on - * the software side now. - */ - - drm_atomic_helper_swap_state(state, true); - - drm_atomic_state_get(state); - if (nonblock) - sti_atomic_schedule(private, state); - else - sti_atomic_complete(private, state); - - mutex_unlock(&private->commit.lock); - return 0; -} - static void sti_output_poll_changed(struct drm_device *ddev) { struct sti_private *private = ddev->dev_private; @@ -224,7 +148,7 @@ static const struct drm_mode_config_funcs sti_mode_config_funcs = { .fb_create = drm_fb_cma_create, .output_poll_changed = sti_output_poll_changed, .atomic_check = sti_atomic_check, - .atomic_commit = sti_atomic_commit, + .atomic_commit = drm_atomic_helper_commit, }; static void sti_mode_config_init(struct drm_device *dev) @@ -304,9 +228,6 @@ static int sti_init(struct drm_device *ddev) dev_set_drvdata(ddev->dev, ddev); private->drm_dev = ddev; - mutex_init(&private->commit.lock); - INIT_WORK(&private->commit.work, sti_atomic_work); - drm_mode_config_init(ddev); sti_mode_config_init(ddev); @@ -327,6 +248,7 @@ static void sti_cleanup(struct drm_device *ddev) drm_kms_helper_poll_fini(ddev); drm_vblank_cleanup(ddev); + component_unbind_all(ddev->dev, ddev); kfree(private); ddev->dev_private = NULL; } diff --git a/drivers/gpu/drm/sti/sti_drv.h b/drivers/gpu/drm/sti/sti_drv.h index 4c75845cc9ab..6502ed2d3351 100644 --- a/drivers/gpu/drm/sti/sti_drv.h +++ b/drivers/gpu/drm/sti/sti_drv.h @@ -25,12 +25,6 @@ struct sti_private { struct drm_property *plane_zorder_property; struct drm_device *drm_dev; struct drm_fbdev_cma *fbdev; - - struct { - struct drm_atomic_state *state; - struct work_struct work; - struct mutex lock; - } commit; }; extern struct platform_driver sti_tvout_driver; diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c index 877d053d86f4..86279f5022c2 100644 --- a/drivers/gpu/drm/sti/sti_gdp.c +++ b/drivers/gpu/drm/sti/sti_gdp.c @@ -610,7 +610,6 @@ static int sti_gdp_atomic_check(struct drm_plane *drm_plane, struct sti_plane *plane = to_sti_plane(drm_plane); struct sti_gdp *gdp = to_sti_gdp(plane); struct drm_crtc *crtc = state->crtc; - struct sti_compositor *compo = dev_get_drvdata(gdp->dev); struct drm_framebuffer *fb = state->fb; struct drm_crtc_state *crtc_state; struct sti_mixer *mixer; @@ -648,45 +647,30 @@ static int sti_gdp_atomic_check(struct drm_plane *drm_plane, return -EINVAL; } - if (!gdp->vtg) { - /* Register gdp callback */ - gdp->vtg = compo->vtg[mixer->id]; - if (sti_vtg_register_client(gdp->vtg, - &gdp->vtg_field_nb, crtc)) { - DRM_ERROR("Cannot register VTG notifier\n"); + /* Set gdp clock */ + if (mode->clock && gdp->clk_pix) { + struct clk *clkp; + int rate = mode->clock * 1000; + int res; + + /* + * According to the mixer used, the gdp pixel clock + * should have a different parent clock. + */ + if (mixer->id == STI_MIXER_MAIN) + clkp = gdp->clk_main_parent; + else + clkp = gdp->clk_aux_parent; + + if (clkp) + clk_set_parent(gdp->clk_pix, clkp); + + res = clk_set_rate(gdp->clk_pix, rate); + if (res < 0) { + DRM_ERROR("Cannot set rate (%dHz) for gdp\n", + rate); return -EINVAL; } - - /* Set and enable gdp clock */ - if (gdp->clk_pix) { - struct clk *clkp; - int rate = mode->clock * 1000; - int res; - - /* - * According to the mixer used, the gdp pixel clock - * should have a different parent clock. - */ - if (mixer->id == STI_MIXER_MAIN) - clkp = gdp->clk_main_parent; - else - clkp = gdp->clk_aux_parent; - - if (clkp) - clk_set_parent(gdp->clk_pix, clkp); - - res = clk_set_rate(gdp->clk_pix, rate); - if (res < 0) { - DRM_ERROR("Cannot set rate (%dHz) for gdp\n", - rate); - return -EINVAL; - } - - if (clk_prepare_enable(gdp->clk_pix)) { - DRM_ERROR("Failed to prepare/enable gdp\n"); - return -EINVAL; - } - } } DRM_DEBUG_KMS("CRTC:%d (%s) drm plane:%d (%s)\n", @@ -724,6 +708,31 @@ static void sti_gdp_atomic_update(struct drm_plane *drm_plane, if (!crtc || !fb) return; + if ((oldstate->fb == state->fb) && + (oldstate->crtc_x == state->crtc_x) && + (oldstate->crtc_y == state->crtc_y) && + (oldstate->crtc_w == state->crtc_w) && + (oldstate->crtc_h == state->crtc_h) && + (oldstate->src_x == state->src_x) && + (oldstate->src_y == state->src_y) && + (oldstate->src_w == state->src_w) && + (oldstate->src_h == state->src_h)) { + /* No change since last update, do not post cmd */ + DRM_DEBUG_DRIVER("No change, not posting cmd\n"); + plane->status = STI_PLANE_UPDATED; + return; + } + + if (!gdp->vtg) { + struct sti_compositor *compo = dev_get_drvdata(gdp->dev); + struct sti_mixer *mixer = to_sti_mixer(crtc); + + /* Register gdp callback */ + gdp->vtg = compo->vtg[mixer->id]; + sti_vtg_register_client(gdp->vtg, &gdp->vtg_field_nb, crtc); + clk_prepare_enable(gdp->clk_pix); + } + mode = &crtc->mode; dst_x = state->crtc_x; dst_y = state->crtc_y; diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c index c9151849d604..ce2dcba679d5 100644 --- a/drivers/gpu/drm/sti/sti_hdmi.c +++ b/drivers/gpu/drm/sti/sti_hdmi.c @@ -95,7 +95,6 @@ #define HDMI_CFG_HDCP_EN BIT(2) #define HDMI_CFG_ESS_NOT_OESS BIT(3) #define HDMI_CFG_H_SYNC_POL_NEG BIT(4) -#define HDMI_CFG_SINK_TERM_DET_EN BIT(5) #define HDMI_CFG_V_SYNC_POL_NEG BIT(6) #define HDMI_CFG_422_EN BIT(8) #define HDMI_CFG_FIFO_OVERRUN_CLR BIT(12) @@ -159,7 +158,6 @@ struct sti_hdmi_connector { struct drm_encoder *encoder; struct sti_hdmi *hdmi; struct drm_property *colorspace_property; - struct drm_property *hdmi_mode_property; }; #define to_sti_hdmi_connector(x) \ @@ -266,12 +264,9 @@ static void hdmi_config(struct sti_hdmi *hdmi) /* Select encryption type and the framing mode */ conf |= HDMI_CFG_ESS_NOT_OESS; - if (hdmi->hdmi_mode == HDMI_MODE_HDMI) + if (hdmi->hdmi_monitor) conf |= HDMI_CFG_HDMI_NOT_DVI; - /* Enable sink term detection */ - conf |= HDMI_CFG_SINK_TERM_DET_EN; - /* Set Hsync polarity */ if (hdmi->mode.flags & DRM_MODE_FLAG_NHSYNC) { DRM_DEBUG_DRIVER("H Sync Negative\n"); @@ -607,9 +602,6 @@ static void hdmi_dbg_cfg(struct seq_file *s, int val) tmp = val & HDMI_CFG_ESS_NOT_OESS; DBGFS_PRINT_STR("HDCP mode:", tmp ? "ESS enable" : "OESS enable"); seq_puts(s, "\t\t\t\t\t"); - tmp = val & HDMI_CFG_SINK_TERM_DET_EN; - DBGFS_PRINT_STR("Sink term detection:", tmp ? "enable" : "disable"); - seq_puts(s, "\t\t\t\t\t"); tmp = val & HDMI_CFG_H_SYNC_POL_NEG; DBGFS_PRINT_STR("Hsync polarity:", tmp ? "inverted" : "normal"); seq_puts(s, "\t\t\t\t\t"); @@ -977,6 +969,11 @@ static int sti_hdmi_connector_get_modes(struct drm_connector *connector) if (!edid) goto fail; + hdmi->hdmi_monitor = drm_detect_hdmi_monitor(edid); + DRM_DEBUG_KMS("%s : %dx%d cm\n", + (hdmi->hdmi_monitor ? "hdmi monitor" : "dvi monitor"), + edid->width_cm, edid->height_cm); + count = drm_add_edid_modes(connector, edid); drm_mode_connector_update_edid_property(connector, edid); drm_edid_to_eld(connector, edid); @@ -1060,19 +1057,6 @@ static void sti_hdmi_connector_init_property(struct drm_device *drm_dev, } hdmi_connector->colorspace_property = prop; drm_object_attach_property(&connector->base, prop, hdmi->colorspace); - - /* hdmi_mode property */ - hdmi->hdmi_mode = DEFAULT_HDMI_MODE; - prop = drm_property_create_enum(drm_dev, 0, "hdmi_mode", - hdmi_mode_names, - ARRAY_SIZE(hdmi_mode_names)); - if (!prop) { - DRM_ERROR("fails to create colorspace property\n"); - return; - } - hdmi_connector->hdmi_mode_property = prop; - drm_object_attach_property(&connector->base, prop, hdmi->hdmi_mode); - } static int @@ -1090,11 +1074,6 @@ sti_hdmi_connector_set_property(struct drm_connector *connector, return 0; } - if (property == hdmi_connector->hdmi_mode_property) { - hdmi->hdmi_mode = val; - return 0; - } - DRM_ERROR("failed to set hdmi connector property\n"); return -EINVAL; } @@ -1114,11 +1093,6 @@ sti_hdmi_connector_get_property(struct drm_connector *connector, return 0; } - if (property == hdmi_connector->hdmi_mode_property) { - *val = hdmi->hdmi_mode; - return 0; - } - DRM_ERROR("failed to get hdmi connector property\n"); return -EINVAL; } diff --git a/drivers/gpu/drm/sti/sti_hdmi.h b/drivers/gpu/drm/sti/sti_hdmi.h index 119bc3582ac7..407012350f1a 100644 --- a/drivers/gpu/drm/sti/sti_hdmi.h +++ b/drivers/gpu/drm/sti/sti_hdmi.h @@ -30,19 +30,6 @@ struct hdmi_audio_params { struct hdmi_audio_infoframe cea; }; -/* values for the framing mode property */ -enum sti_hdmi_modes { - HDMI_MODE_HDMI, - HDMI_MODE_DVI, -}; - -static const struct drm_prop_enum_list hdmi_mode_names[] = { - { HDMI_MODE_HDMI, "hdmi" }, - { HDMI_MODE_DVI, "dvi" }, -}; - -#define DEFAULT_HDMI_MODE HDMI_MODE_HDMI - static const struct drm_prop_enum_list colorspace_mode_names[] = { { HDMI_COLORSPACE_RGB, "rgb" }, { HDMI_COLORSPACE_YUV422, "yuv422" }, @@ -73,7 +60,7 @@ static const struct drm_prop_enum_list colorspace_mode_names[] = { * @reset: reset control of the hdmi phy * @ddc_adapt: i2c ddc adapter * @colorspace: current colorspace selected - * @hdmi_mode: select framing for HDMI or DVI + * @hdmi_monitor: true if HDMI monitor detected else DVI monitor assumed * @audio_pdev: ASoC hdmi-codec platform device * @audio: hdmi audio parameters. * @drm_connector: hdmi connector @@ -98,7 +85,7 @@ struct sti_hdmi { struct reset_control *reset; struct i2c_adapter *ddc_adapt; enum hdmi_colorspace colorspace; - enum sti_hdmi_modes hdmi_mode; + bool hdmi_monitor; struct platform_device *audio_pdev; struct hdmi_audio_params audio; struct drm_connector *drm_connector; diff --git a/drivers/gpu/drm/sti/sti_hqvdp.c b/drivers/gpu/drm/sti/sti_hqvdp.c index 4376fd8a8e52..66f843148ef7 100644 --- a/drivers/gpu/drm/sti/sti_hqvdp.c +++ b/drivers/gpu/drm/sti/sti_hqvdp.c @@ -1037,9 +1037,9 @@ static int sti_hqvdp_atomic_check(struct drm_plane *drm_plane, src_w = state->src_w >> 16; src_h = state->src_h >> 16; - if (!sti_hqvdp_check_hw_scaling(hqvdp, mode, - src_w, src_h, - dst_w, dst_h)) { + if (mode->clock && !sti_hqvdp_check_hw_scaling(hqvdp, mode, + src_w, src_h, + dst_w, dst_h)) { DRM_ERROR("Scaling beyond HW capabilities\n"); return -EINVAL; } diff --git a/drivers/gpu/drm/sti/sti_mixer.h b/drivers/gpu/drm/sti/sti_mixer.h index 830a3c42d886..e64a00e61049 100644 --- a/drivers/gpu/drm/sti/sti_mixer.h +++ b/drivers/gpu/drm/sti/sti_mixer.h @@ -28,7 +28,6 @@ enum sti_mixer_status { * @regs: mixer registers * @id: id of the mixer * @drm_crtc: crtc object link to the mixer - * @pending_event: set if a flip event is pending on crtc * @status: to know the status of the mixer */ struct sti_mixer { @@ -36,7 +35,6 @@ struct sti_mixer { void __iomem *regs; int id; struct drm_crtc drm_crtc; - struct drm_pending_vblank_event *pending_event; enum sti_mixer_status status; }; diff --git a/drivers/gpu/drm/sti/sti_vtg.c b/drivers/gpu/drm/sti/sti_vtg.c index c3d9c8ae14af..2dcba1d3a122 100644 --- a/drivers/gpu/drm/sti/sti_vtg.c +++ b/drivers/gpu/drm/sti/sti_vtg.c @@ -17,7 +17,6 @@ #include "sti_vtg.h" #define VTG_MODE_MASTER 0 -#define VTG_MODE_SLAVE_BY_EXT0 1 /* registers offset */ #define VTG_MODE 0x0000 @@ -132,7 +131,6 @@ struct sti_vtg_sync_params { * @irq_status: store the IRQ status value * @notifier_list: notifier callback * @crtc: the CRTC for vblank event - * @slave: slave vtg * @link: List node to link the structure in lookup list */ struct sti_vtg { @@ -144,7 +142,6 @@ struct sti_vtg { u32 irq_status; struct raw_notifier_head notifier_list; struct drm_crtc *crtc; - struct sti_vtg *slave; struct list_head link; }; @@ -166,10 +163,6 @@ struct sti_vtg *of_vtg_find(struct device_node *np) static void vtg_reset(struct sti_vtg *vtg) { - /* reset slave and then master */ - if (vtg->slave) - vtg_reset(vtg->slave); - writel(1, vtg->regs + VTG_DRST_AUTOC); } @@ -259,10 +252,6 @@ static void vtg_set_mode(struct sti_vtg *vtg, { unsigned int i; - if (vtg->slave) - vtg_set_mode(vtg->slave, VTG_MODE_SLAVE_BY_EXT0, - vtg->sync_params, mode); - /* Set the number of clock cycles per line */ writel(mode->htotal, vtg->regs + VTG_CLKLN); @@ -318,11 +307,7 @@ void sti_vtg_set_config(struct sti_vtg *vtg, vtg_reset(vtg); - /* enable irq for the vtg vblank synchro */ - if (vtg->slave) - vtg_enable_irq(vtg->slave); - else - vtg_enable_irq(vtg); + vtg_enable_irq(vtg); } /** @@ -365,18 +350,12 @@ u32 sti_vtg_get_pixel_number(struct drm_display_mode mode, int x) int sti_vtg_register_client(struct sti_vtg *vtg, struct notifier_block *nb, struct drm_crtc *crtc) { - if (vtg->slave) - return sti_vtg_register_client(vtg->slave, nb, crtc); - vtg->crtc = crtc; return raw_notifier_chain_register(&vtg->notifier_list, nb); } int sti_vtg_unregister_client(struct sti_vtg *vtg, struct notifier_block *nb) { - if (vtg->slave) - return sti_vtg_unregister_client(vtg->slave, nb); - return raw_notifier_chain_unregister(&vtg->notifier_list, nb); } @@ -410,7 +389,6 @@ static irqreturn_t vtg_irq(int irq, void *arg) static int vtg_probe(struct platform_device *pdev) { struct device *dev = &pdev->dev; - struct device_node *np; struct sti_vtg *vtg; struct resource *res; int ret; @@ -434,29 +412,20 @@ static int vtg_probe(struct platform_device *pdev) return -ENOMEM; } - np = of_parse_phandle(pdev->dev.of_node, "st,slave", 0); - if (np) { - vtg->slave = of_vtg_find(np); - of_node_put(np); + vtg->irq = platform_get_irq(pdev, 0); + if (vtg->irq < 0) { + DRM_ERROR("Failed to get VTG interrupt\n"); + return vtg->irq; + } - if (!vtg->slave) - return -EPROBE_DEFER; - } else { - vtg->irq = platform_get_irq(pdev, 0); - if (vtg->irq < 0) { - DRM_ERROR("Failed to get VTG interrupt\n"); - return vtg->irq; - } + RAW_INIT_NOTIFIER_HEAD(&vtg->notifier_list); - RAW_INIT_NOTIFIER_HEAD(&vtg->notifier_list); - - ret = devm_request_threaded_irq(dev, vtg->irq, vtg_irq, - vtg_irq_thread, IRQF_ONESHOT, - dev_name(dev), vtg); - if (ret < 0) { - DRM_ERROR("Failed to register VTG interrupt\n"); - return ret; - } + ret = devm_request_threaded_irq(dev, vtg->irq, vtg_irq, + vtg_irq_thread, IRQF_ONESHOT, + dev_name(dev), vtg); + if (ret < 0) { + DRM_ERROR("Failed to register VTG interrupt\n"); + return ret; } vtg_register(vtg); diff --git a/drivers/gpu/drm/tinydrm/Kconfig b/drivers/gpu/drm/tinydrm/Kconfig new file mode 100644 index 000000000000..3504c53846da --- /dev/null +++ b/drivers/gpu/drm/tinydrm/Kconfig @@ -0,0 +1,21 @@ +menuconfig DRM_TINYDRM + tristate "Support for simple displays" + depends on DRM + select DRM_KMS_HELPER + select DRM_KMS_CMA_HELPER + select BACKLIGHT_LCD_SUPPORT + select BACKLIGHT_CLASS_DEVICE + help + Choose this option if you have a tinydrm supported display. + If M is selected the module will be called tinydrm. + +config TINYDRM_MIPI_DBI + tristate + +config TINYDRM_MI0283QT + tristate "DRM support for MI0283QT" + depends on DRM_TINYDRM && SPI + select TINYDRM_MIPI_DBI + help + DRM driver for the Multi-Inno MI0283QT display panel + If M is selected the module will be called mi0283qt. diff --git a/drivers/gpu/drm/tinydrm/Makefile b/drivers/gpu/drm/tinydrm/Makefile new file mode 100644 index 000000000000..7a3604cf4fc2 --- /dev/null +++ b/drivers/gpu/drm/tinydrm/Makefile @@ -0,0 +1,7 @@ +obj-$(CONFIG_DRM_TINYDRM) += core/ + +# Controllers +obj-$(CONFIG_TINYDRM_MIPI_DBI) += mipi-dbi.o + +# Displays +obj-$(CONFIG_TINYDRM_MI0283QT) += mi0283qt.o diff --git a/drivers/gpu/drm/tinydrm/core/Makefile b/drivers/gpu/drm/tinydrm/core/Makefile new file mode 100644 index 000000000000..fb221e6f8885 --- /dev/null +++ b/drivers/gpu/drm/tinydrm/core/Makefile @@ -0,0 +1,3 @@ +tinydrm-y := tinydrm-core.o tinydrm-pipe.o tinydrm-helpers.o + +obj-$(CONFIG_DRM_TINYDRM) += tinydrm.o diff --git a/drivers/gpu/drm/tinydrm/core/tinydrm-core.c b/drivers/gpu/drm/tinydrm/core/tinydrm-core.c new file mode 100644 index 000000000000..6a257dd08ee0 --- /dev/null +++ b/drivers/gpu/drm/tinydrm/core/tinydrm-core.c @@ -0,0 +1,376 @@ +/* + * Copyright (C) 2016 Noralf Trønnes + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include + +/** + * DOC: overview + * + * This library provides driver helpers for very simple display hardware. + * + * It is based on &drm_simple_display_pipe coupled with a &drm_connector which + * has only one fixed &drm_display_mode. The framebuffers are backed by the + * cma helper and have support for framebuffer flushing (dirty). + * fbdev support is also included. + * + */ + +/** + * DOC: core + * + * The driver allocates &tinydrm_device, initializes it using + * devm_tinydrm_init(), sets up the pipeline using tinydrm_display_pipe_init() + * and registers the DRM device using devm_tinydrm_register(). + */ + +/** + * tinydrm_lastclose - DRM lastclose helper + * @drm: DRM device + * + * This function ensures that fbdev is restored when drm_lastclose() is called + * on the last drm_release(). Drivers can use this as their + * &drm_driver->lastclose callback. + */ +void tinydrm_lastclose(struct drm_device *drm) +{ + struct tinydrm_device *tdev = drm->dev_private; + + DRM_DEBUG_KMS("\n"); + drm_fbdev_cma_restore_mode(tdev->fbdev_cma); +} +EXPORT_SYMBOL(tinydrm_lastclose); + +/** + * tinydrm_gem_cma_prime_import_sg_table - Produce a CMA GEM object from + * another driver's scatter/gather table of pinned pages + * @drm: DRM device to import into + * @attach: DMA-BUF attachment + * @sgt: Scatter/gather table of pinned pages + * + * This function imports a scatter/gather table exported via DMA-BUF by + * another driver using drm_gem_cma_prime_import_sg_table(). It sets the + * kernel virtual address on the CMA object. Drivers should use this as their + * &drm_driver->gem_prime_import_sg_table callback if they need the virtual + * address. tinydrm_gem_cma_free_object() should be used in combination with + * this function. + * + * Returns: + * A pointer to a newly created GEM object or an ERR_PTR-encoded negative + * error code on failure. + */ +struct drm_gem_object * +tinydrm_gem_cma_prime_import_sg_table(struct drm_device *drm, + struct dma_buf_attachment *attach, + struct sg_table *sgt) +{ + struct drm_gem_cma_object *cma_obj; + struct drm_gem_object *obj; + void *vaddr; + + vaddr = dma_buf_vmap(attach->dmabuf); + if (!vaddr) { + DRM_ERROR("Failed to vmap PRIME buffer\n"); + return ERR_PTR(-ENOMEM); + } + + obj = drm_gem_cma_prime_import_sg_table(drm, attach, sgt); + if (IS_ERR(obj)) { + dma_buf_vunmap(attach->dmabuf, vaddr); + return obj; + } + + cma_obj = to_drm_gem_cma_obj(obj); + cma_obj->vaddr = vaddr; + + return obj; +} +EXPORT_SYMBOL(tinydrm_gem_cma_prime_import_sg_table); + +/** + * tinydrm_gem_cma_free_object - Free resources associated with a CMA GEM + * object + * @gem_obj: GEM object to free + * + * This function frees the backing memory of the CMA GEM object, cleans up the + * GEM object state and frees the memory used to store the object itself using + * drm_gem_cma_free_object(). It also handles PRIME buffers which has the kernel + * virtual address set by tinydrm_gem_cma_prime_import_sg_table(). Drivers + * can use this as their &drm_driver->gem_free_object callback. + */ +void tinydrm_gem_cma_free_object(struct drm_gem_object *gem_obj) +{ + if (gem_obj->import_attach) { + struct drm_gem_cma_object *cma_obj; + + cma_obj = to_drm_gem_cma_obj(gem_obj); + dma_buf_vunmap(gem_obj->import_attach->dmabuf, cma_obj->vaddr); + cma_obj->vaddr = NULL; + } + + drm_gem_cma_free_object(gem_obj); +} +EXPORT_SYMBOL_GPL(tinydrm_gem_cma_free_object); + +const struct file_operations tinydrm_fops = { + .owner = THIS_MODULE, + .open = drm_open, + .release = drm_release, + .unlocked_ioctl = drm_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = drm_compat_ioctl, +#endif + .poll = drm_poll, + .read = drm_read, + .llseek = no_llseek, + .mmap = drm_gem_cma_mmap, +}; +EXPORT_SYMBOL(tinydrm_fops); + +static struct drm_framebuffer * +tinydrm_fb_create(struct drm_device *drm, struct drm_file *file_priv, + const struct drm_mode_fb_cmd2 *mode_cmd) +{ + struct tinydrm_device *tdev = drm->dev_private; + + return drm_fb_cma_create_with_funcs(drm, file_priv, mode_cmd, + tdev->fb_funcs); +} + +static const struct drm_mode_config_funcs tinydrm_mode_config_funcs = { + .fb_create = tinydrm_fb_create, + .atomic_check = drm_atomic_helper_check, + .atomic_commit = drm_atomic_helper_commit, +}; + +static int tinydrm_init(struct device *parent, struct tinydrm_device *tdev, + const struct drm_framebuffer_funcs *fb_funcs, + struct drm_driver *driver) +{ + struct drm_device *drm; + + mutex_init(&tdev->dirty_lock); + tdev->fb_funcs = fb_funcs; + + /* + * We don't embed drm_device, because that prevent us from using + * devm_kzalloc() to allocate tinydrm_device in the driver since + * drm_dev_unref() frees the structure. The devm_ functions provide + * for easy error handling. + */ + drm = drm_dev_alloc(driver, parent); + if (IS_ERR(drm)) + return PTR_ERR(drm); + + tdev->drm = drm; + drm->dev_private = tdev; + drm_mode_config_init(drm); + drm->mode_config.funcs = &tinydrm_mode_config_funcs; + + return 0; +} + +static void tinydrm_fini(struct tinydrm_device *tdev) +{ + drm_mode_config_cleanup(tdev->drm); + mutex_destroy(&tdev->dirty_lock); + tdev->drm->dev_private = NULL; + drm_dev_unref(tdev->drm); +} + +static void devm_tinydrm_release(void *data) +{ + tinydrm_fini(data); +} + +/** + * devm_tinydrm_init - Initialize tinydrm device + * @parent: Parent device object + * @tdev: tinydrm device + * @fb_funcs: Framebuffer functions + * @driver: DRM driver + * + * This function initializes @tdev, the underlying DRM device and it's + * mode_config. Resources will be automatically freed on driver detach (devres) + * using drm_mode_config_cleanup() and drm_dev_unref(). + * + * Returns: + * Zero on success, negative error code on failure. + */ +int devm_tinydrm_init(struct device *parent, struct tinydrm_device *tdev, + const struct drm_framebuffer_funcs *fb_funcs, + struct drm_driver *driver) +{ + int ret; + + ret = tinydrm_init(parent, tdev, fb_funcs, driver); + if (ret) + return ret; + + ret = devm_add_action(parent, devm_tinydrm_release, tdev); + if (ret) + tinydrm_fini(tdev); + + return ret; +} +EXPORT_SYMBOL(devm_tinydrm_init); + +static int tinydrm_register(struct tinydrm_device *tdev) +{ + struct drm_device *drm = tdev->drm; + int bpp = drm->mode_config.preferred_depth; + struct drm_fbdev_cma *fbdev; + int ret; + + ret = drm_dev_register(tdev->drm, 0); + if (ret) + return ret; + + fbdev = drm_fbdev_cma_init_with_funcs(drm, bpp ? bpp : 32, + drm->mode_config.num_connector, + tdev->fb_funcs); + if (IS_ERR(fbdev)) + DRM_ERROR("Failed to initialize fbdev: %ld\n", PTR_ERR(fbdev)); + else + tdev->fbdev_cma = fbdev; + + return 0; +} + +static void tinydrm_unregister(struct tinydrm_device *tdev) +{ + struct drm_fbdev_cma *fbdev_cma = tdev->fbdev_cma; + + drm_crtc_force_disable_all(tdev->drm); + /* don't restore fbdev in lastclose, keep pipeline disabled */ + tdev->fbdev_cma = NULL; + drm_dev_unregister(tdev->drm); + if (fbdev_cma) + drm_fbdev_cma_fini(fbdev_cma); +} + +static void devm_tinydrm_register_release(void *data) +{ + tinydrm_unregister(data); +} + +/** + * devm_tinydrm_register - Register tinydrm device + * @tdev: tinydrm device + * + * This function registers the underlying DRM device and fbdev. + * These resources will be automatically unregistered on driver detach (devres) + * and the display pipeline will be disabled. + * + * Returns: + * Zero on success, negative error code on failure. + */ +int devm_tinydrm_register(struct tinydrm_device *tdev) +{ + struct device *dev = tdev->drm->dev; + int ret; + + ret = tinydrm_register(tdev); + if (ret) + return ret; + + ret = devm_add_action(dev, devm_tinydrm_register_release, tdev); + if (ret) + tinydrm_unregister(tdev); + + return ret; +} +EXPORT_SYMBOL(devm_tinydrm_register); + +/** + * tinydrm_shutdown - Shutdown tinydrm + * @tdev: tinydrm device + * + * This function makes sure that the display pipeline is disabled. + * Used by drivers in their shutdown callback to turn off the display + * on machine shutdown and reboot. + */ +void tinydrm_shutdown(struct tinydrm_device *tdev) +{ + drm_crtc_force_disable_all(tdev->drm); +} +EXPORT_SYMBOL(tinydrm_shutdown); + +/** + * tinydrm_suspend - Suspend tinydrm + * @tdev: tinydrm device + * + * Used in driver PM operations to suspend tinydrm. + * Suspends fbdev and DRM. + * Resume with tinydrm_resume(). + * + * Returns: + * Zero on success, negative error code on failure. + */ +int tinydrm_suspend(struct tinydrm_device *tdev) +{ + struct drm_atomic_state *state; + + if (tdev->suspend_state) { + DRM_ERROR("Failed to suspend: state already set\n"); + return -EINVAL; + } + + drm_fbdev_cma_set_suspend_unlocked(tdev->fbdev_cma, 1); + state = drm_atomic_helper_suspend(tdev->drm); + if (IS_ERR(state)) { + drm_fbdev_cma_set_suspend_unlocked(tdev->fbdev_cma, 0); + return PTR_ERR(state); + } + + tdev->suspend_state = state; + + return 0; +} +EXPORT_SYMBOL(tinydrm_suspend); + +/** + * tinydrm_resume - Resume tinydrm + * @tdev: tinydrm device + * + * Used in driver PM operations to resume tinydrm. + * Suspend with tinydrm_suspend(). + * + * Returns: + * Zero on success, negative error code on failure. + */ +int tinydrm_resume(struct tinydrm_device *tdev) +{ + struct drm_atomic_state *state = tdev->suspend_state; + int ret; + + if (!state) { + DRM_ERROR("Failed to resume: state is not set\n"); + return -EINVAL; + } + + tdev->suspend_state = NULL; + + ret = drm_atomic_helper_resume(tdev->drm, state); + if (ret) { + DRM_ERROR("Error resuming state: %d\n", ret); + return ret; + } + + drm_fbdev_cma_set_suspend_unlocked(tdev->fbdev_cma, 0); + + return 0; +} +EXPORT_SYMBOL(tinydrm_resume); + +MODULE_LICENSE("GPL"); diff --git a/drivers/gpu/drm/tinydrm/core/tinydrm-helpers.c b/drivers/gpu/drm/tinydrm/core/tinydrm-helpers.c new file mode 100644 index 000000000000..3ccda6c1e159 --- /dev/null +++ b/drivers/gpu/drm/tinydrm/core/tinydrm-helpers.c @@ -0,0 +1,460 @@ +/* + * Copyright (C) 2016 Noralf Trønnes + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include + +static unsigned int spi_max; +module_param(spi_max, uint, 0400); +MODULE_PARM_DESC(spi_max, "Set a lower SPI max transfer size"); + +/** + * tinydrm_merge_clips - Merge clip rectangles + * @dst: Destination clip rectangle + * @src: Source clip rectangle(s) + * @num_clips: Number of @src clip rectangles + * @flags: Dirty fb ioctl flags + * @max_width: Maximum width of @dst + * @max_height: Maximum height of @dst + * + * This function merges @src clip rectangle(s) into @dst. If @src is NULL, + * @max_width and @min_width is used to set a full @dst clip rectangle. + * + * Returns: + * true if it's a full clip, false otherwise + */ +bool tinydrm_merge_clips(struct drm_clip_rect *dst, + struct drm_clip_rect *src, unsigned int num_clips, + unsigned int flags, u32 max_width, u32 max_height) +{ + unsigned int i; + + if (!src || !num_clips) { + dst->x1 = 0; + dst->x2 = max_width; + dst->y1 = 0; + dst->y2 = max_height; + return true; + } + + dst->x1 = ~0; + dst->y1 = ~0; + dst->x2 = 0; + dst->y2 = 0; + + for (i = 0; i < num_clips; i++) { + if (flags & DRM_MODE_FB_DIRTY_ANNOTATE_COPY) + i++; + dst->x1 = min(dst->x1, src[i].x1); + dst->x2 = max(dst->x2, src[i].x2); + dst->y1 = min(dst->y1, src[i].y1); + dst->y2 = max(dst->y2, src[i].y2); + } + + if (dst->x2 > max_width || dst->y2 > max_height || + dst->x1 >= dst->x2 || dst->y1 >= dst->y2) { + DRM_DEBUG_KMS("Illegal clip: x1=%u, x2=%u, y1=%u, y2=%u\n", + dst->x1, dst->x2, dst->y1, dst->y2); + dst->x1 = 0; + dst->y1 = 0; + dst->x2 = max_width; + dst->y2 = max_height; + } + + return (dst->x2 - dst->x1) == max_width && + (dst->y2 - dst->y1) == max_height; +} +EXPORT_SYMBOL(tinydrm_merge_clips); + +/** + * tinydrm_memcpy - Copy clip buffer + * @dst: Destination buffer + * @vaddr: Source buffer + * @fb: DRM framebuffer + * @clip: Clip rectangle area to copy + */ +void tinydrm_memcpy(void *dst, void *vaddr, struct drm_framebuffer *fb, + struct drm_clip_rect *clip) +{ + unsigned int cpp = drm_format_plane_cpp(fb->format->format, 0); + unsigned int pitch = fb->pitches[0]; + void *src = vaddr + (clip->y1 * pitch) + (clip->x1 * cpp); + size_t len = (clip->x2 - clip->x1) * cpp; + unsigned int y; + + for (y = clip->y1; y < clip->y2; y++) { + memcpy(dst, src, len); + src += pitch; + dst += len; + } +} +EXPORT_SYMBOL(tinydrm_memcpy); + +/** + * tinydrm_swab16 - Swap bytes into clip buffer + * @dst: RGB565 destination buffer + * @vaddr: RGB565 source buffer + * @fb: DRM framebuffer + * @clip: Clip rectangle area to copy + */ +void tinydrm_swab16(u16 *dst, void *vaddr, struct drm_framebuffer *fb, + struct drm_clip_rect *clip) +{ + size_t len = (clip->x2 - clip->x1) * sizeof(u16); + unsigned int x, y; + u16 *src, *buf; + + /* + * The cma memory is write-combined so reads are uncached. + * Speed up by fetching one line at a time. + */ + buf = kmalloc(len, GFP_KERNEL); + if (!buf) + return; + + for (y = clip->y1; y < clip->y2; y++) { + src = vaddr + (y * fb->pitches[0]); + src += clip->x1; + memcpy(buf, src, len); + src = buf; + for (x = clip->x1; x < clip->x2; x++) + *dst++ = swab16(*src++); + } + + kfree(buf); +} +EXPORT_SYMBOL(tinydrm_swab16); + +/** + * tinydrm_xrgb8888_to_rgb565 - Convert XRGB8888 to RGB565 clip buffer + * @dst: RGB565 destination buffer + * @vaddr: XRGB8888 source buffer + * @fb: DRM framebuffer + * @clip: Clip rectangle area to copy + * @swap: Swap bytes + * + * Drivers can use this function for RGB565 devices that don't natively + * support XRGB8888. + */ +void tinydrm_xrgb8888_to_rgb565(u16 *dst, void *vaddr, + struct drm_framebuffer *fb, + struct drm_clip_rect *clip, bool swap) +{ + size_t len = (clip->x2 - clip->x1) * sizeof(u32); + unsigned int x, y; + u32 *src, *buf; + u16 val16; + + buf = kmalloc(len, GFP_KERNEL); + if (!buf) + return; + + for (y = clip->y1; y < clip->y2; y++) { + src = vaddr + (y * fb->pitches[0]); + src += clip->x1; + memcpy(buf, src, len); + src = buf; + for (x = clip->x1; x < clip->x2; x++) { + val16 = ((*src & 0x00F80000) >> 8) | + ((*src & 0x0000FC00) >> 5) | + ((*src & 0x000000F8) >> 3); + src++; + if (swap) + *dst++ = swab16(val16); + else + *dst++ = val16; + } + } + + kfree(buf); +} +EXPORT_SYMBOL(tinydrm_xrgb8888_to_rgb565); + +/** + * tinydrm_of_find_backlight - Find backlight device in device-tree + * @dev: Device + * + * This function looks for a DT node pointed to by a property named 'backlight' + * and uses of_find_backlight_by_node() to get the backlight device. + * Additionally if the brightness property is zero, it is set to + * max_brightness. + * + * Returns: + * NULL if there's no backlight property. + * Error pointer -EPROBE_DEFER if the DT node is found, but no backlight device + * is found. + * If the backlight device is found, a pointer to the structure is returned. + */ +struct backlight_device *tinydrm_of_find_backlight(struct device *dev) +{ + struct backlight_device *backlight; + struct device_node *np; + + np = of_parse_phandle(dev->of_node, "backlight", 0); + if (!np) + return NULL; + + backlight = of_find_backlight_by_node(np); + of_node_put(np); + + if (!backlight) + return ERR_PTR(-EPROBE_DEFER); + + if (!backlight->props.brightness) { + backlight->props.brightness = backlight->props.max_brightness; + DRM_DEBUG_KMS("Backlight brightness set to %d\n", + backlight->props.brightness); + } + + return backlight; +} +EXPORT_SYMBOL(tinydrm_of_find_backlight); + +/** + * tinydrm_enable_backlight - Enable backlight helper + * @backlight: Backlight device + * + * Returns: + * Zero on success, negative error code on failure. + */ +int tinydrm_enable_backlight(struct backlight_device *backlight) +{ + unsigned int old_state; + int ret; + + if (!backlight) + return 0; + + old_state = backlight->props.state; + backlight->props.state &= ~BL_CORE_FBBLANK; + DRM_DEBUG_KMS("Backlight state: 0x%x -> 0x%x\n", old_state, + backlight->props.state); + + ret = backlight_update_status(backlight); + if (ret) + DRM_ERROR("Failed to enable backlight %d\n", ret); + + return ret; +} +EXPORT_SYMBOL(tinydrm_enable_backlight); + +/** + * tinydrm_disable_backlight - Disable backlight helper + * @backlight: Backlight device + * + * Returns: + * Zero on success, negative error code on failure. + */ +int tinydrm_disable_backlight(struct backlight_device *backlight) +{ + unsigned int old_state; + int ret; + + if (!backlight) + return 0; + + old_state = backlight->props.state; + backlight->props.state |= BL_CORE_FBBLANK; + DRM_DEBUG_KMS("Backlight state: 0x%x -> 0x%x\n", old_state, + backlight->props.state); + ret = backlight_update_status(backlight); + if (ret) + DRM_ERROR("Failed to disable backlight %d\n", ret); + + return ret; +} +EXPORT_SYMBOL(tinydrm_disable_backlight); + +#if IS_ENABLED(CONFIG_SPI) + +/** + * tinydrm_spi_max_transfer_size - Determine max SPI transfer size + * @spi: SPI device + * @max_len: Maximum buffer size needed (optional) + * + * This function returns the maximum size to use for SPI transfers. It checks + * the SPI master, the optional @max_len and the module parameter spi_max and + * returns the smallest. + * + * Returns: + * Maximum size for SPI transfers + */ +size_t tinydrm_spi_max_transfer_size(struct spi_device *spi, size_t max_len) +{ + size_t ret; + + ret = min(spi_max_transfer_size(spi), spi->master->max_dma_len); + if (max_len) + ret = min(ret, max_len); + if (spi_max) + ret = min_t(size_t, ret, spi_max); + ret &= ~0x3; + if (ret < 4) + ret = 4; + + return ret; +} +EXPORT_SYMBOL(tinydrm_spi_max_transfer_size); + +/** + * tinydrm_spi_bpw_supported - Check if bits per word is supported + * @spi: SPI device + * @bpw: Bits per word + * + * This function checks to see if the SPI master driver supports @bpw. + * + * Returns: + * True if @bpw is supported, false otherwise. + */ +bool tinydrm_spi_bpw_supported(struct spi_device *spi, u8 bpw) +{ + u32 bpw_mask = spi->master->bits_per_word_mask; + + if (bpw == 8) + return true; + + if (!bpw_mask) { + dev_warn_once(&spi->dev, + "bits_per_word_mask not set, assume 8-bit only\n"); + return false; + } + + if (bpw_mask & SPI_BPW_MASK(bpw)) + return true; + + return false; +} +EXPORT_SYMBOL(tinydrm_spi_bpw_supported); + +static void +tinydrm_dbg_spi_print(struct spi_device *spi, struct spi_transfer *tr, + const void *buf, int idx, bool tx) +{ + u32 speed_hz = tr->speed_hz ? tr->speed_hz : spi->max_speed_hz; + char linebuf[3 * 32]; + + hex_dump_to_buffer(buf, tr->len, 16, + DIV_ROUND_UP(tr->bits_per_word, 8), + linebuf, sizeof(linebuf), false); + + printk(KERN_DEBUG + " tr(%i): speed=%u%s, bpw=%i, len=%u, %s_buf=[%s%s]\n", idx, + speed_hz > 1000000 ? speed_hz / 1000000 : speed_hz / 1000, + speed_hz > 1000000 ? "MHz" : "kHz", tr->bits_per_word, tr->len, + tx ? "tx" : "rx", linebuf, tr->len > 16 ? " ..." : ""); +} + +/* called through tinydrm_dbg_spi_message() */ +void _tinydrm_dbg_spi_message(struct spi_device *spi, struct spi_message *m) +{ + struct spi_transfer *tmp; + struct list_head *pos; + int i = 0; + + list_for_each(pos, &m->transfers) { + tmp = list_entry(pos, struct spi_transfer, transfer_list); + + if (tmp->tx_buf) + tinydrm_dbg_spi_print(spi, tmp, tmp->tx_buf, i, true); + if (tmp->rx_buf) + tinydrm_dbg_spi_print(spi, tmp, tmp->rx_buf, i, false); + i++; + } +} +EXPORT_SYMBOL(_tinydrm_dbg_spi_message); + +/** + * tinydrm_spi_transfer - SPI transfer helper + * @spi: SPI device + * @speed_hz: Override speed (optional) + * @header: Optional header transfer + * @bpw: Bits per word + * @buf: Buffer to transfer + * @len: Buffer length + * + * This SPI transfer helper breaks up the transfer of @buf into chunks which + * the SPI master driver can handle. If the machine is Little Endian and the + * SPI master driver doesn't support 16 bits per word, it swaps the bytes and + * does a 8-bit transfer. + * If @header is set, it is prepended to each SPI message. + * + * Returns: + * Zero on success, negative error code on failure. + */ +int tinydrm_spi_transfer(struct spi_device *spi, u32 speed_hz, + struct spi_transfer *header, u8 bpw, const void *buf, + size_t len) +{ + struct spi_transfer tr = { + .bits_per_word = bpw, + .speed_hz = speed_hz, + }; + struct spi_message m; + u16 *swap_buf = NULL; + size_t max_chunk; + size_t chunk; + int ret = 0; + + if (WARN_ON_ONCE(bpw != 8 && bpw != 16)) + return -EINVAL; + + max_chunk = tinydrm_spi_max_transfer_size(spi, 0); + + if (drm_debug & DRM_UT_DRIVER) + pr_debug("[drm:%s] bpw=%u, max_chunk=%zu, transfers:\n", + __func__, bpw, max_chunk); + + if (bpw == 16 && !tinydrm_spi_bpw_supported(spi, 16)) { + tr.bits_per_word = 8; + if (tinydrm_machine_little_endian()) { + swap_buf = kmalloc(min(len, max_chunk), GFP_KERNEL); + if (!swap_buf) + return -ENOMEM; + } + } + + spi_message_init(&m); + if (header) + spi_message_add_tail(header, &m); + spi_message_add_tail(&tr, &m); + + while (len) { + chunk = min(len, max_chunk); + + tr.tx_buf = buf; + tr.len = chunk; + + if (swap_buf) { + const u16 *buf16 = buf; + unsigned int i; + + for (i = 0; i < chunk / 2; i++) + swap_buf[i] = swab16(buf16[i]); + + tr.tx_buf = swap_buf; + } + + buf += chunk; + len -= chunk; + + tinydrm_dbg_spi_message(spi, &m); + ret = spi_sync(spi, &m); + if (ret) + return ret; + }; + + return 0; +} +EXPORT_SYMBOL(tinydrm_spi_transfer); + +#endif /* CONFIG_SPI */ diff --git a/drivers/gpu/drm/tinydrm/core/tinydrm-pipe.c b/drivers/gpu/drm/tinydrm/core/tinydrm-pipe.c new file mode 100644 index 000000000000..ec43fb7ad9e4 --- /dev/null +++ b/drivers/gpu/drm/tinydrm/core/tinydrm-pipe.c @@ -0,0 +1,234 @@ +/* + * Copyright (C) 2016 Noralf Trønnes + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include + +struct tinydrm_connector { + struct drm_connector base; + const struct drm_display_mode *mode; +}; + +static inline struct tinydrm_connector * +to_tinydrm_connector(struct drm_connector *connector) +{ + return container_of(connector, struct tinydrm_connector, base); +} + +static int tinydrm_connector_get_modes(struct drm_connector *connector) +{ + struct tinydrm_connector *tconn = to_tinydrm_connector(connector); + struct drm_display_mode *mode; + + mode = drm_mode_duplicate(connector->dev, tconn->mode); + if (!mode) { + DRM_ERROR("Failed to duplicate mode\n"); + return 0; + } + + if (mode->name[0] == '\0') + drm_mode_set_name(mode); + + mode->type |= DRM_MODE_TYPE_PREFERRED; + drm_mode_probed_add(connector, mode); + + if (mode->width_mm) { + connector->display_info.width_mm = mode->width_mm; + connector->display_info.height_mm = mode->height_mm; + } + + return 1; +} + +static const struct drm_connector_helper_funcs tinydrm_connector_hfuncs = { + .get_modes = tinydrm_connector_get_modes, + .best_encoder = drm_atomic_helper_best_encoder, +}; + +static enum drm_connector_status +tinydrm_connector_detect(struct drm_connector *connector, bool force) +{ + if (drm_device_is_unplugged(connector->dev)) + return connector_status_disconnected; + + return connector->status; +} + +static void tinydrm_connector_destroy(struct drm_connector *connector) +{ + struct tinydrm_connector *tconn = to_tinydrm_connector(connector); + + drm_connector_cleanup(connector); + kfree(tconn); +} + +static const struct drm_connector_funcs tinydrm_connector_funcs = { + .dpms = drm_atomic_helper_connector_dpms, + .reset = drm_atomic_helper_connector_reset, + .detect = tinydrm_connector_detect, + .fill_modes = drm_helper_probe_single_connector_modes, + .destroy = tinydrm_connector_destroy, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, +}; + +struct drm_connector * +tinydrm_connector_create(struct drm_device *drm, + const struct drm_display_mode *mode, + int connector_type) +{ + struct tinydrm_connector *tconn; + struct drm_connector *connector; + int ret; + + tconn = kzalloc(sizeof(*tconn), GFP_KERNEL); + if (!tconn) + return ERR_PTR(-ENOMEM); + + tconn->mode = mode; + connector = &tconn->base; + + drm_connector_helper_add(connector, &tinydrm_connector_hfuncs); + ret = drm_connector_init(drm, connector, &tinydrm_connector_funcs, + connector_type); + if (ret) { + kfree(tconn); + return ERR_PTR(ret); + } + + connector->status = connector_status_connected; + + return connector; +} + +/** + * tinydrm_display_pipe_update - Display pipe update helper + * @pipe: Simple display pipe + * @old_state: Old plane state + * + * This function does a full framebuffer flush if the plane framebuffer + * has changed. It also handles vblank events. Drivers can use this as their + * &drm_simple_display_pipe_funcs->update callback. + */ +void tinydrm_display_pipe_update(struct drm_simple_display_pipe *pipe, + struct drm_plane_state *old_state) +{ + struct tinydrm_device *tdev = pipe_to_tinydrm(pipe); + struct drm_framebuffer *fb = pipe->plane.state->fb; + struct drm_crtc *crtc = &tdev->pipe.crtc; + + if (fb && (fb != old_state->fb)) { + pipe->plane.fb = fb; + if (fb->funcs->dirty) + fb->funcs->dirty(fb, NULL, 0, 0, NULL, 0); + } + + if (crtc->state->event) { + spin_lock_irq(&crtc->dev->event_lock); + drm_crtc_send_vblank_event(crtc, crtc->state->event); + spin_unlock_irq(&crtc->dev->event_lock); + crtc->state->event = NULL; + } +} +EXPORT_SYMBOL(tinydrm_display_pipe_update); + +/** + * tinydrm_display_pipe_prepare_fb - Display pipe prepare_fb helper + * @pipe: Simple display pipe + * @plane_state: Plane state + * + * This function uses drm_fb_cma_prepare_fb() to check if the plane FB has an + * dma-buf attached, extracts the exclusive fence and attaches it to plane + * state for the atomic helper to wait on. Drivers can use this as their + * &drm_simple_display_pipe_funcs->prepare_fb callback. + */ +int tinydrm_display_pipe_prepare_fb(struct drm_simple_display_pipe *pipe, + struct drm_plane_state *plane_state) +{ + return drm_fb_cma_prepare_fb(&pipe->plane, plane_state); +} +EXPORT_SYMBOL(tinydrm_display_pipe_prepare_fb); + +static int tinydrm_rotate_mode(struct drm_display_mode *mode, + unsigned int rotation) +{ + if (rotation == 0 || rotation == 180) { + return 0; + } else if (rotation == 90 || rotation == 270) { + swap(mode->hdisplay, mode->vdisplay); + swap(mode->hsync_start, mode->vsync_start); + swap(mode->hsync_end, mode->vsync_end); + swap(mode->htotal, mode->vtotal); + swap(mode->width_mm, mode->height_mm); + return 0; + } else { + return -EINVAL; + } +} + +/** + * tinydrm_display_pipe_init - Initialize display pipe + * @tdev: tinydrm device + * @funcs: Display pipe functions + * @connector_type: Connector type + * @formats: Array of supported formats (DRM_FORMAT\_\*) + * @format_count: Number of elements in @formats + * @mode: Supported mode + * @rotation: Initial @mode rotation in degrees Counter Clock Wise + * + * This function sets up a &drm_simple_display_pipe with a &drm_connector that + * has one fixed &drm_display_mode which is rotated according to @rotation. + * + * Returns: + * Zero on success, negative error code on failure. + */ +int +tinydrm_display_pipe_init(struct tinydrm_device *tdev, + const struct drm_simple_display_pipe_funcs *funcs, + int connector_type, + const uint32_t *formats, + unsigned int format_count, + const struct drm_display_mode *mode, + unsigned int rotation) +{ + struct drm_device *drm = tdev->drm; + struct drm_display_mode *mode_copy; + struct drm_connector *connector; + int ret; + + mode_copy = devm_kmalloc(drm->dev, sizeof(*mode_copy), GFP_KERNEL); + if (!mode_copy) + return -ENOMEM; + + *mode_copy = *mode; + ret = tinydrm_rotate_mode(mode_copy, rotation); + if (ret) { + DRM_ERROR("Illegal rotation value %u\n", rotation); + return -EINVAL; + } + + drm->mode_config.min_width = mode_copy->hdisplay; + drm->mode_config.max_width = mode_copy->hdisplay; + drm->mode_config.min_height = mode_copy->vdisplay; + drm->mode_config.max_height = mode_copy->vdisplay; + + connector = tinydrm_connector_create(drm, mode_copy, connector_type); + if (IS_ERR(connector)) + return PTR_ERR(connector); + + ret = drm_simple_display_pipe_init(drm, &tdev->pipe, funcs, formats, + format_count, connector); + if (ret) + return ret; + + return 0; +} +EXPORT_SYMBOL(tinydrm_display_pipe_init); diff --git a/drivers/gpu/drm/tinydrm/mi0283qt.c b/drivers/gpu/drm/tinydrm/mi0283qt.c new file mode 100644 index 000000000000..b29fe86158f7 --- /dev/null +++ b/drivers/gpu/drm/tinydrm/mi0283qt.c @@ -0,0 +1,279 @@ +/* + * DRM driver for Multi-Inno MI0283QT panels + * + * Copyright 2016 Noralf Trønnes + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include