linux_old1/drivers/ata/sata_nv.c

2520 lines
68 KiB
C
Raw Normal View History

/*
* sata_nv.c - NVIDIA nForce SATA
*
* Copyright 2004 NVIDIA Corp. All rights reserved.
* Copyright 2004 Andrew Chew
*
*
* 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, or (at your option)
* any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; see the file COPYING. If not, write to
* the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
*
*
* libata documentation is available via 'make {ps|pdf}docs',
* as Documentation/driver-api/libata.rst
*
* No hardware documentation available outside of NVIDIA.
* This driver programs the NVIDIA SATA controller in a similar
* fashion as with other PCI IDE BMDMA controllers, with a few
* NV-specific details such as register offsets, SATA phy location,
* hotplug info, etc.
*
* CK804/MCP04 controllers support an alternate programming interface
* similar to the ADMA specification (with some modifications).
* This allows the use of NCQ. Non-DMA-mapped ATA commands are still
* sent through the legacy interface.
*
*/
#include <linux/kernel.h>
#include <linux/module.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/gfp.h>
#include <linux/pci.h>
#include <linux/blkdev.h>
#include <linux/delay.h>
#include <linux/interrupt.h>
#include <linux/device.h>
#include <scsi/scsi_host.h>
#include <scsi/scsi_device.h>
#include <linux/libata.h>
#define DRV_NAME "sata_nv"
#define DRV_VERSION "3.5"
#define NV_ADMA_DMA_BOUNDARY 0xffffffffUL
enum {
NV_MMIO_BAR = 5,
NV_PORTS = 2,
NV_PIO_MASK = ATA_PIO4,
NV_MWDMA_MASK = ATA_MWDMA2,
NV_UDMA_MASK = ATA_UDMA6,
NV_PORT0_SCR_REG_OFFSET = 0x00,
NV_PORT1_SCR_REG_OFFSET = 0x40,
/* INT_STATUS/ENABLE */
NV_INT_STATUS = 0x10,
NV_INT_ENABLE = 0x11,
NV_INT_STATUS_CK804 = 0x440,
NV_INT_ENABLE_CK804 = 0x441,
/* INT_STATUS/ENABLE bits */
NV_INT_DEV = 0x01,
NV_INT_PM = 0x02,
NV_INT_ADDED = 0x04,
NV_INT_REMOVED = 0x08,
NV_INT_PORT_SHIFT = 4, /* each port occupies 4 bits */
NV_INT_ALL = 0x0f,
NV_INT_MASK = NV_INT_DEV |
NV_INT_ADDED | NV_INT_REMOVED,
/* INT_CONFIG */
NV_INT_CONFIG = 0x12,
NV_INT_CONFIG_METHD = 0x01, // 0 = INT, 1 = SMI
// For PCI config register 20
NV_MCP_SATA_CFG_20 = 0x50,
NV_MCP_SATA_CFG_20_SATA_SPACE_EN = 0x04,
NV_MCP_SATA_CFG_20_PORT0_EN = (1 << 17),
NV_MCP_SATA_CFG_20_PORT1_EN = (1 << 16),
NV_MCP_SATA_CFG_20_PORT0_PWB_EN = (1 << 14),
NV_MCP_SATA_CFG_20_PORT1_PWB_EN = (1 << 12),
NV_ADMA_MAX_CPBS = 32,
NV_ADMA_CPB_SZ = 128,
NV_ADMA_APRD_SZ = 16,
NV_ADMA_SGTBL_LEN = (1024 - NV_ADMA_CPB_SZ) /
NV_ADMA_APRD_SZ,
NV_ADMA_SGTBL_TOTAL_LEN = NV_ADMA_SGTBL_LEN + 5,
NV_ADMA_SGTBL_SZ = NV_ADMA_SGTBL_LEN * NV_ADMA_APRD_SZ,
NV_ADMA_PORT_PRIV_DMA_SZ = NV_ADMA_MAX_CPBS *
(NV_ADMA_CPB_SZ + NV_ADMA_SGTBL_SZ),
/* BAR5 offset to ADMA general registers */
NV_ADMA_GEN = 0x400,
NV_ADMA_GEN_CTL = 0x00,
NV_ADMA_NOTIFIER_CLEAR = 0x30,
/* BAR5 offset to ADMA ports */
NV_ADMA_PORT = 0x480,
/* size of ADMA port register space */
NV_ADMA_PORT_SIZE = 0x100,
/* ADMA port registers */
NV_ADMA_CTL = 0x40,
NV_ADMA_CPB_COUNT = 0x42,
NV_ADMA_NEXT_CPB_IDX = 0x43,
NV_ADMA_STAT = 0x44,
NV_ADMA_CPB_BASE_LOW = 0x48,
NV_ADMA_CPB_BASE_HIGH = 0x4C,
NV_ADMA_APPEND = 0x50,
NV_ADMA_NOTIFIER = 0x68,
NV_ADMA_NOTIFIER_ERROR = 0x6C,
/* NV_ADMA_CTL register bits */
NV_ADMA_CTL_HOTPLUG_IEN = (1 << 0),
NV_ADMA_CTL_CHANNEL_RESET = (1 << 5),
NV_ADMA_CTL_GO = (1 << 7),
NV_ADMA_CTL_AIEN = (1 << 8),
NV_ADMA_CTL_READ_NON_COHERENT = (1 << 11),
NV_ADMA_CTL_WRITE_NON_COHERENT = (1 << 12),
/* CPB response flag bits */
NV_CPB_RESP_DONE = (1 << 0),
NV_CPB_RESP_ATA_ERR = (1 << 3),
NV_CPB_RESP_CMD_ERR = (1 << 4),
NV_CPB_RESP_CPB_ERR = (1 << 7),
/* CPB control flag bits */
NV_CPB_CTL_CPB_VALID = (1 << 0),
NV_CPB_CTL_QUEUE = (1 << 1),
NV_CPB_CTL_APRD_VALID = (1 << 2),
NV_CPB_CTL_IEN = (1 << 3),
NV_CPB_CTL_FPDMA = (1 << 4),
/* APRD flags */
NV_APRD_WRITE = (1 << 1),
NV_APRD_END = (1 << 2),
NV_APRD_CONT = (1 << 3),
/* NV_ADMA_STAT flags */
NV_ADMA_STAT_TIMEOUT = (1 << 0),
NV_ADMA_STAT_HOTUNPLUG = (1 << 1),
NV_ADMA_STAT_HOTPLUG = (1 << 2),
NV_ADMA_STAT_CPBERR = (1 << 4),
NV_ADMA_STAT_SERROR = (1 << 5),
NV_ADMA_STAT_CMD_COMPLETE = (1 << 6),
NV_ADMA_STAT_IDLE = (1 << 8),
NV_ADMA_STAT_LEGACY = (1 << 9),
NV_ADMA_STAT_STOPPED = (1 << 10),
NV_ADMA_STAT_DONE = (1 << 12),
NV_ADMA_STAT_ERR = NV_ADMA_STAT_CPBERR |
NV_ADMA_STAT_TIMEOUT,
/* port flags */
NV_ADMA_PORT_REGISTER_MODE = (1 << 0),
NV_ADMA_ATAPI_SETUP_COMPLETE = (1 << 1),
/* MCP55 reg offset */
NV_CTL_MCP55 = 0x400,
NV_INT_STATUS_MCP55 = 0x440,
NV_INT_ENABLE_MCP55 = 0x444,
NV_NCQ_REG_MCP55 = 0x448,
/* MCP55 */
NV_INT_ALL_MCP55 = 0xffff,
NV_INT_PORT_SHIFT_MCP55 = 16, /* each port occupies 16 bits */
NV_INT_MASK_MCP55 = NV_INT_ALL_MCP55 & 0xfffd,
/* SWNCQ ENABLE BITS*/
NV_CTL_PRI_SWNCQ = 0x02,
NV_CTL_SEC_SWNCQ = 0x04,
/* SW NCQ status bits*/
NV_SWNCQ_IRQ_DEV = (1 << 0),
NV_SWNCQ_IRQ_PM = (1 << 1),
NV_SWNCQ_IRQ_ADDED = (1 << 2),
NV_SWNCQ_IRQ_REMOVED = (1 << 3),
NV_SWNCQ_IRQ_BACKOUT = (1 << 4),
NV_SWNCQ_IRQ_SDBFIS = (1 << 5),
NV_SWNCQ_IRQ_DHREGFIS = (1 << 6),
NV_SWNCQ_IRQ_DMASETUP = (1 << 7),
NV_SWNCQ_IRQ_HOTPLUG = NV_SWNCQ_IRQ_ADDED |
NV_SWNCQ_IRQ_REMOVED,
};
/* ADMA Physical Region Descriptor - one SG segment */
struct nv_adma_prd {
__le64 addr;
__le32 len;
u8 flags;
u8 packet_len;
__le16 reserved;
};
enum nv_adma_regbits {
CMDEND = (1 << 15), /* end of command list */
WNB = (1 << 14), /* wait-not-BSY */
IGN = (1 << 13), /* ignore this entry */
CS1n = (1 << (4 + 8)), /* std. PATA signals follow... */
DA2 = (1 << (2 + 8)),
DA1 = (1 << (1 + 8)),
DA0 = (1 << (0 + 8)),
};
/* ADMA Command Parameter Block
The first 5 SG segments are stored inside the Command Parameter Block itself.
If there are more than 5 segments the remainder are stored in a separate
memory area indicated by next_aprd. */
struct nv_adma_cpb {
u8 resp_flags; /* 0 */
u8 reserved1; /* 1 */
u8 ctl_flags; /* 2 */
/* len is length of taskfile in 64 bit words */
u8 len; /* 3 */
u8 tag; /* 4 */
u8 next_cpb_idx; /* 5 */
__le16 reserved2; /* 6-7 */
__le16 tf[12]; /* 8-31 */
struct nv_adma_prd aprd[5]; /* 32-111 */
__le64 next_aprd; /* 112-119 */
__le64 reserved3; /* 120-127 */
};
struct nv_adma_port_priv {
struct nv_adma_cpb *cpb;
dma_addr_t cpb_dma;
struct nv_adma_prd *aprd;
dma_addr_t aprd_dma;
void __iomem *ctl_block;
void __iomem *gen_block;
void __iomem *notifier_clear_block;
u64 adma_dma_mask;
u8 flags;
int last_issue_ncq;
};
struct nv_host_priv {
unsigned long type;
};
struct defer_queue {
u32 defer_bits;
unsigned int head;
unsigned int tail;
unsigned int tag[ATA_MAX_QUEUE];
};
enum ncq_saw_flag_list {
ncq_saw_d2h = (1U << 0),
ncq_saw_dmas = (1U << 1),
ncq_saw_sdb = (1U << 2),
ncq_saw_backout = (1U << 3),
};
struct nv_swncq_port_priv {
struct ata_bmdma_prd *prd; /* our SG list */
dma_addr_t prd_dma; /* and its DMA mapping */
void __iomem *sactive_block;
void __iomem *irq_block;
void __iomem *tag_block;
u32 qc_active;
unsigned int last_issue_tag;
/* fifo circular queue to store deferral command */
struct defer_queue defer_queue;
/* for NCQ interrupt analysis */
u32 dhfis_bits;
u32 dmafis_bits;
u32 sdbfis_bits;
unsigned int ncq_flags;
};
#define NV_ADMA_CHECK_INTR(GCTL, PORT) ((GCTL) & (1 << (19 + (12 * (PORT)))))
static int nv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
#ifdef CONFIG_PM_SLEEP
static int nv_pci_device_resume(struct pci_dev *pdev);
#endif
static void nv_ck804_host_stop(struct ata_host *host);
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance);
static irqreturn_t nv_nf2_interrupt(int irq, void *dev_instance);
static irqreturn_t nv_ck804_interrupt(int irq, void *dev_instance);
static int nv_scr_read(struct ata_link *link, unsigned int sc_reg, u32 *val);
static int nv_scr_write(struct ata_link *link, unsigned int sc_reg, u32 val);
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
static int nv_hardreset(struct ata_link *link, unsigned int *class,
unsigned long deadline);
static void nv_nf2_freeze(struct ata_port *ap);
static void nv_nf2_thaw(struct ata_port *ap);
static void nv_ck804_freeze(struct ata_port *ap);
static void nv_ck804_thaw(struct ata_port *ap);
static int nv_adma_slave_config(struct scsi_device *sdev);
static int nv_adma_check_atapi_dma(struct ata_queued_cmd *qc);
static void nv_adma_qc_prep(struct ata_queued_cmd *qc);
static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc);
static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance);
static void nv_adma_irq_clear(struct ata_port *ap);
static int nv_adma_port_start(struct ata_port *ap);
static void nv_adma_port_stop(struct ata_port *ap);
#ifdef CONFIG_PM
static int nv_adma_port_suspend(struct ata_port *ap, pm_message_t mesg);
static int nv_adma_port_resume(struct ata_port *ap);
#endif
static void nv_adma_freeze(struct ata_port *ap);
static void nv_adma_thaw(struct ata_port *ap);
static void nv_adma_error_handler(struct ata_port *ap);
static void nv_adma_host_stop(struct ata_host *host);
static void nv_adma_post_internal_cmd(struct ata_queued_cmd *qc);
static void nv_adma_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
static void nv_mcp55_thaw(struct ata_port *ap);
static void nv_mcp55_freeze(struct ata_port *ap);
static void nv_swncq_error_handler(struct ata_port *ap);
static int nv_swncq_slave_config(struct scsi_device *sdev);
static int nv_swncq_port_start(struct ata_port *ap);
static void nv_swncq_qc_prep(struct ata_queued_cmd *qc);
static void nv_swncq_fill_sg(struct ata_queued_cmd *qc);
static unsigned int nv_swncq_qc_issue(struct ata_queued_cmd *qc);
static void nv_swncq_irq_clear(struct ata_port *ap, u16 fis);
static irqreturn_t nv_swncq_interrupt(int irq, void *dev_instance);
#ifdef CONFIG_PM
static int nv_swncq_port_suspend(struct ata_port *ap, pm_message_t mesg);
static int nv_swncq_port_resume(struct ata_port *ap);
#endif
enum nv_host_type
{
GENERIC,
NFORCE2,
NFORCE3 = NFORCE2, /* NF2 == NF3 as far as sata_nv is concerned */
CK804,
ADMA,
MCP5x,
SWNCQ,
};
static const struct pci_device_id nv_pci_tbl[] = {
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE2S_SATA), NFORCE2 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE3S_SATA), NFORCE3 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE3S_SATA2), NFORCE3 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_CK804_SATA), CK804 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_CK804_SATA2), CK804 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_SATA), CK804 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_SATA2), CK804 },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA), MCP5x },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA2), MCP5x },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA), MCP5x },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA2), MCP5x },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA), GENERIC },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA2), GENERIC },
{ PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA3), GENERIC },
{ } /* terminate list */
};
static struct pci_driver nv_pci_driver = {
.name = DRV_NAME,
.id_table = nv_pci_tbl,
.probe = nv_init_one,
#ifdef CONFIG_PM_SLEEP
.suspend = ata_pci_device_suspend,
.resume = nv_pci_device_resume,
#endif
.remove = ata_pci_remove_one,
};
static struct scsi_host_template nv_sht = {
ATA_BMDMA_SHT(DRV_NAME),
};
static struct scsi_host_template nv_adma_sht = {
ATA_NCQ_SHT(DRV_NAME),
.can_queue = NV_ADMA_MAX_CPBS,
.sg_tablesize = NV_ADMA_SGTBL_TOTAL_LEN,
.dma_boundary = NV_ADMA_DMA_BOUNDARY,
.slave_configure = nv_adma_slave_config,
};
static struct scsi_host_template nv_swncq_sht = {
ATA_NCQ_SHT(DRV_NAME),
.can_queue = ATA_MAX_QUEUE,
.sg_tablesize = LIBATA_MAX_PRD,
.dma_boundary = ATA_DMA_BOUNDARY,
.slave_configure = nv_swncq_slave_config,
};
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
/*
* NV SATA controllers have various different problems with hardreset
* protocol depending on the specific controller and device.
*
* GENERIC:
*
* bko11195 reports that link doesn't come online after hardreset on
* generic nv's and there have been several other similar reports on
* linux-ide.
*
* bko12351#c23 reports that warmplug on MCP61 doesn't work with
* softreset.
*
* NF2/3:
*
* bko3352 reports nf2/3 controllers can't determine device signature
* reliably after hardreset. The following thread reports detection
* failure on cold boot with the standard debouncing timing.
*
* http://thread.gmane.org/gmane.linux.ide/34098
*
* bko12176 reports that hardreset fails to bring up the link during
* boot on nf2.
*
* CK804:
*
* For initial probing after boot and hot plugging, hardreset mostly
* works fine on CK804 but curiously, reprobing on the initial port
* by rescanning or rmmod/insmod fails to acquire the initial D2H Reg
* FIS in somewhat undeterministic way.
*
* SWNCQ:
*
* bko12351 reports that when SWNCQ is enabled, for hotplug to work,
* hardreset should be used and hardreset can't report proper
* signature, which suggests that mcp5x is closer to nf2 as long as
* reset quirkiness is concerned.
*
* bko12703 reports that boot probing fails for intel SSD with
* hardreset. Link fails to come online. Softreset works fine.
*
* The failures are varied but the following patterns seem true for
* all flavors.
*
* - Softreset during boot always works.
*
* - Hardreset during boot sometimes fails to bring up the link on
* certain comibnations and device signature acquisition is
* unreliable.
*
* - Hardreset is often necessary after hotplug.
*
* So, preferring softreset for boot probing and error handling (as
* hardreset might bring down the link) but using hardreset for
* post-boot probing should work around the above issues in most
* cases. Define nv_hardreset() which only kicks in for post-boot
* probing and use it for all variants.
*/
static struct ata_port_operations nv_generic_ops = {
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
.inherits = &ata_bmdma_port_ops,
.lost_interrupt = ATA_OP_NULL,
.scr_read = nv_scr_read,
.scr_write = nv_scr_write,
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
.hardreset = nv_hardreset,
};
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
static struct ata_port_operations nv_nf2_ops = {
.inherits = &nv_generic_ops,
.freeze = nv_nf2_freeze,
.thaw = nv_nf2_thaw,
};
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
static struct ata_port_operations nv_ck804_ops = {
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
.inherits = &nv_generic_ops,
.freeze = nv_ck804_freeze,
.thaw = nv_ck804_thaw,
.host_stop = nv_ck804_host_stop,
};
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
static struct ata_port_operations nv_adma_ops = {
.inherits = &nv_ck804_ops,
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
.check_atapi_dma = nv_adma_check_atapi_dma,
.sff_tf_read = nv_adma_tf_read,
.qc_defer = ata_std_qc_defer,
.qc_prep = nv_adma_qc_prep,
.qc_issue = nv_adma_qc_issue,
.sff_irq_clear = nv_adma_irq_clear,
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
.freeze = nv_adma_freeze,
.thaw = nv_adma_thaw,
.error_handler = nv_adma_error_handler,
.post_internal_cmd = nv_adma_post_internal_cmd,
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
.port_start = nv_adma_port_start,
.port_stop = nv_adma_port_stop,
#ifdef CONFIG_PM
.port_suspend = nv_adma_port_suspend,
.port_resume = nv_adma_port_resume,
#endif
.host_stop = nv_adma_host_stop,
};
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
static struct ata_port_operations nv_swncq_ops = {
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
.inherits = &nv_generic_ops,
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
.qc_defer = ata_std_qc_defer,
.qc_prep = nv_swncq_qc_prep,
.qc_issue = nv_swncq_qc_issue,
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
.freeze = nv_mcp55_freeze,
.thaw = nv_mcp55_thaw,
.error_handler = nv_swncq_error_handler,
libata: implement and use ops inheritance libata lets low level drivers build ata_port_operations table and register it with libata core layer. This allows low level drivers high level of flexibility but also burdens them with lots of boilerplate entries. This becomes worse for drivers which support related similar controllers which differ slightly. They share most of the operations except for a few. However, the driver still needs to list all operations for each variant. This results in large number of duplicate entries, which is not only inefficient but also error-prone as it becomes very difficult to tell what the actual differences are. This duplicate boilerplates all over the low level drivers also make updating the core layer exteremely difficult and error-prone. When compounded with multi-branched development model, it ends up accumulating inconsistencies over time. Some of those inconsistencies cause immediate problems and fixed. Others just remain there dormant making maintenance increasingly difficult. To rectify the problem, this patch implements ata_port_operations inheritance. To allow LLDs to easily re-use their own ops tables overriding only specific methods, this patch implements poor man's class inheritance. An ops table has ->inherits field which can be set to any ops table as long as it doesn't create a loop. When the host is started, the inheritance chain is followed and any operation which isn't specified is taken from the nearest ancestor which has it specified. This operation is called finalization and done only once per an ops table and the LLD doesn't have to do anything special about it other than making the ops table non-const such that libata can update it. libata provides four base ops tables lower drivers can inherit from - base, sata, pmp, sff and bmdma. To avoid overriding these ops accidentaly, these ops are declared const and LLDs should always inherit these instead of using them directly. After finalization, all the ops table are identical before and after the patch except for setting .irq_handler to ata_interrupt in drivers which didn't use to. The .irq_handler doesn't have any actual effect and the field will soon be removed by later patch. * sata_sx4 is still using old style EH and currently doesn't take advantage of ops inheritance. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-03-25 11:22:49 +08:00
#ifdef CONFIG_PM
.port_suspend = nv_swncq_port_suspend,
.port_resume = nv_swncq_port_resume,
#endif
.port_start = nv_swncq_port_start,
};
struct nv_pi_priv {
irq_handler_t irq_handler;
struct scsi_host_template *sht;
};
#define NV_PI_PRIV(_irq_handler, _sht) \
&(struct nv_pi_priv){ .irq_handler = _irq_handler, .sht = _sht }
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 18:43:58 +08:00
static const struct ata_port_info nv_port_info[] = {
/* generic */
{
.flags = ATA_FLAG_SATA,
.pio_mask = NV_PIO_MASK,
.mwdma_mask = NV_MWDMA_MASK,
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_generic_ops,
.private_data = NV_PI_PRIV(nv_generic_interrupt, &nv_sht),
},
/* nforce2/3 */
{
.flags = ATA_FLAG_SATA,
.pio_mask = NV_PIO_MASK,
.mwdma_mask = NV_MWDMA_MASK,
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_nf2_ops,
.private_data = NV_PI_PRIV(nv_nf2_interrupt, &nv_sht),
},
/* ck804 */
{
.flags = ATA_FLAG_SATA,
.pio_mask = NV_PIO_MASK,
.mwdma_mask = NV_MWDMA_MASK,
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_ck804_ops,
.private_data = NV_PI_PRIV(nv_ck804_interrupt, &nv_sht),
},
/* ADMA */
{
.flags = ATA_FLAG_SATA | ATA_FLAG_NCQ,
.pio_mask = NV_PIO_MASK,
.mwdma_mask = NV_MWDMA_MASK,
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_adma_ops,
.private_data = NV_PI_PRIV(nv_adma_interrupt, &nv_adma_sht),
},
/* MCP5x */
{
.flags = ATA_FLAG_SATA,
.pio_mask = NV_PIO_MASK,
.mwdma_mask = NV_MWDMA_MASK,
.udma_mask = NV_UDMA_MASK,
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
.port_ops = &nv_generic_ops,
.private_data = NV_PI_PRIV(nv_generic_interrupt, &nv_sht),
},
/* SWNCQ */
{
.flags = ATA_FLAG_SATA | ATA_FLAG_NCQ,
.pio_mask = NV_PIO_MASK,
.mwdma_mask = NV_MWDMA_MASK,
.udma_mask = NV_UDMA_MASK,
.port_ops = &nv_swncq_ops,
.private_data = NV_PI_PRIV(nv_swncq_interrupt, &nv_swncq_sht),
},
};
MODULE_AUTHOR("NVIDIA");
MODULE_DESCRIPTION("low-level driver for NVIDIA nForce SATA controller");
MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, nv_pci_tbl);
MODULE_VERSION(DRV_VERSION);
static bool adma_enabled;
static bool swncq_enabled = true;
static bool msi_enabled;
static void nv_adma_register_mode(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u16 tmp, status;
int count = 0;
if (pp->flags & NV_ADMA_PORT_REGISTER_MODE)
return;
status = readw(mmio + NV_ADMA_STAT);
while (!(status & NV_ADMA_STAT_IDLE) && count < 20) {
ndelay(50);
status = readw(mmio + NV_ADMA_STAT);
count++;
}
if (count == 20)
ata_port_warn(ap, "timeout waiting for ADMA IDLE, stat=0x%hx\n",
status);
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp & ~NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
count = 0;
status = readw(mmio + NV_ADMA_STAT);
while (!(status & NV_ADMA_STAT_LEGACY) && count < 20) {
ndelay(50);
status = readw(mmio + NV_ADMA_STAT);
count++;
}
if (count == 20)
ata_port_warn(ap,
"timeout waiting for ADMA LEGACY, stat=0x%hx\n",
status);
pp->flags |= NV_ADMA_PORT_REGISTER_MODE;
}
static void nv_adma_mode(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u16 tmp, status;
int count = 0;
if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE))
return;
WARN_ON(pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE);
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp | NV_ADMA_CTL_GO, mmio + NV_ADMA_CTL);
status = readw(mmio + NV_ADMA_STAT);
while (((status & NV_ADMA_STAT_LEGACY) ||
!(status & NV_ADMA_STAT_IDLE)) && count < 20) {
ndelay(50);
status = readw(mmio + NV_ADMA_STAT);
count++;
}
if (count == 20)
ata_port_warn(ap,
"timeout waiting for ADMA LEGACY clear and IDLE, stat=0x%hx\n",
status);
pp->flags &= ~NV_ADMA_PORT_REGISTER_MODE;
}
static int nv_adma_slave_config(struct scsi_device *sdev)
{
struct ata_port *ap = ata_shost_to_port(sdev->host);
struct nv_adma_port_priv *pp = ap->private_data;
struct nv_adma_port_priv *port0, *port1;
struct scsi_device *sdev0, *sdev1;
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
unsigned long segment_boundary, flags;
unsigned short sg_tablesize;
int rc;
int adma_enable;
u32 current_reg, new_reg, config_mask;
rc = ata_scsi_slave_config(sdev);
if (sdev->id >= ATA_MAX_DEVICES || sdev->channel || sdev->lun)
/* Not a proper libata device, ignore */
return rc;
spin_lock_irqsave(ap->lock, flags);
if (ap->link.device[sdev->id].class == ATA_DEV_ATAPI) {
/*
* NVIDIA reports that ADMA mode does not support ATAPI commands.
* Therefore ATAPI commands are sent through the legacy interface.
* However, the legacy interface only supports 32-bit DMA.
* Restrict DMA parameters as required by the legacy interface
* when an ATAPI device is connected.
*/
segment_boundary = ATA_DMA_BOUNDARY;
/* Subtract 1 since an extra entry may be needed for padding, see
libata-scsi.c */
sg_tablesize = LIBATA_MAX_PRD - 1;
/* Since the legacy DMA engine is in use, we need to disable ADMA
on the port. */
adma_enable = 0;
nv_adma_register_mode(ap);
} else {
segment_boundary = NV_ADMA_DMA_BOUNDARY;
sg_tablesize = NV_ADMA_SGTBL_TOTAL_LEN;
adma_enable = 1;
}
pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &current_reg);
if (ap->port_no == 1)
config_mask = NV_MCP_SATA_CFG_20_PORT1_EN |
NV_MCP_SATA_CFG_20_PORT1_PWB_EN;
else
config_mask = NV_MCP_SATA_CFG_20_PORT0_EN |
NV_MCP_SATA_CFG_20_PORT0_PWB_EN;
if (adma_enable) {
new_reg = current_reg | config_mask;
pp->flags &= ~NV_ADMA_ATAPI_SETUP_COMPLETE;
} else {
new_reg = current_reg & ~config_mask;
pp->flags |= NV_ADMA_ATAPI_SETUP_COMPLETE;
}
if (current_reg != new_reg)
pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, new_reg);
port0 = ap->host->ports[0]->private_data;
port1 = ap->host->ports[1]->private_data;
sdev0 = ap->host->ports[0]->link.device[0].sdev;
sdev1 = ap->host->ports[1]->link.device[0].sdev;
if ((port0->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) ||
(port1->flags & NV_ADMA_ATAPI_SETUP_COMPLETE)) {
/** We have to set the DMA mask to 32-bit if either port is in
ATAPI mode, since they are on the same PCI device which is
used for DMA mapping. If we set the mask we also need to set
the bounce limit on both ports to ensure that the block
layer doesn't feed addresses that cause DMA mapping to
choke. If either SCSI device is not allocated yet, it's OK
since that port will discover its correct setting when it
does get allocated.
Note: Setting 32-bit mask should not fail. */
if (sdev0)
blk_queue_bounce_limit(sdev0->request_queue,
ATA_DMA_MASK);
if (sdev1)
blk_queue_bounce_limit(sdev1->request_queue,
ATA_DMA_MASK);
dma_set_mask(&pdev->dev, ATA_DMA_MASK);
} else {
/** This shouldn't fail as it was set to this value before */
dma_set_mask(&pdev->dev, pp->adma_dma_mask);
if (sdev0)
blk_queue_bounce_limit(sdev0->request_queue,
pp->adma_dma_mask);
if (sdev1)
blk_queue_bounce_limit(sdev1->request_queue,
pp->adma_dma_mask);
}
blk_queue_segment_boundary(sdev->request_queue, segment_boundary);
blk_queue_max_segments(sdev->request_queue, sg_tablesize);
ata_port_info(ap,
"DMA mask 0x%llX, segment boundary 0x%lX, hw segs %hu\n",
(unsigned long long)*ap->host->dev->dma_mask,
segment_boundary, sg_tablesize);
spin_unlock_irqrestore(ap->lock, flags);
return rc;
}
static int nv_adma_check_atapi_dma(struct ata_queued_cmd *qc)
{
struct nv_adma_port_priv *pp = qc->ap->private_data;
return !(pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE);
}
static void nv_adma_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
{
/* Other than when internal or pass-through commands are executed,
the only time this function will be called in ADMA mode will be
if a command fails. In the failure case we don't care about going
into register mode with ADMA commands pending, as the commands will
all shortly be aborted anyway. We assume that NCQ commands are not
issued via passthrough, which is the only way that switching into
ADMA mode could abort outstanding commands. */
nv_adma_register_mode(ap);
ata_sff_tf_read(ap, tf);
}
static unsigned int nv_adma_tf_to_cpb(struct ata_taskfile *tf, __le16 *cpb)
{
unsigned int idx = 0;
if (tf->flags & ATA_TFLAG_ISADDR) {
if (tf->flags & ATA_TFLAG_LBA48) {
cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->hob_feature | WNB);
cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->hob_nsect);
cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->hob_lbal);
cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->hob_lbam);
cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->hob_lbah);
cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->feature);
} else
cpb[idx++] = cpu_to_le16((ATA_REG_ERR << 8) | tf->feature | WNB);
cpb[idx++] = cpu_to_le16((ATA_REG_NSECT << 8) | tf->nsect);
cpb[idx++] = cpu_to_le16((ATA_REG_LBAL << 8) | tf->lbal);
cpb[idx++] = cpu_to_le16((ATA_REG_LBAM << 8) | tf->lbam);
cpb[idx++] = cpu_to_le16((ATA_REG_LBAH << 8) | tf->lbah);
}
if (tf->flags & ATA_TFLAG_DEVICE)
cpb[idx++] = cpu_to_le16((ATA_REG_DEVICE << 8) | tf->device);
cpb[idx++] = cpu_to_le16((ATA_REG_CMD << 8) | tf->command | CMDEND);
while (idx < 12)
cpb[idx++] = cpu_to_le16(IGN);
return idx;
}
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
static int nv_adma_check_cpb(struct ata_port *ap, int cpb_num, int force_err)
{
struct nv_adma_port_priv *pp = ap->private_data;
u8 flags = pp->cpb[cpb_num].resp_flags;
VPRINTK("CPB %d, flags=0x%x\n", cpb_num, flags);
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
if (unlikely((force_err ||
flags & (NV_CPB_RESP_ATA_ERR |
NV_CPB_RESP_CMD_ERR |
NV_CPB_RESP_CPB_ERR)))) {
struct ata_eh_info *ehi = &ap->link.eh_info;
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
int freeze = 0;
ata_ehi_clear_desc(ehi);
__ata_ehi_push_desc(ehi, "CPB resp_flags 0x%x: ", flags);
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
if (flags & NV_CPB_RESP_ATA_ERR) {
ata_ehi_push_desc(ehi, "ATA error");
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
ehi->err_mask |= AC_ERR_DEV;
} else if (flags & NV_CPB_RESP_CMD_ERR) {
ata_ehi_push_desc(ehi, "CMD error");
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
ehi->err_mask |= AC_ERR_DEV;
} else if (flags & NV_CPB_RESP_CPB_ERR) {
ata_ehi_push_desc(ehi, "CPB error");
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
ehi->err_mask |= AC_ERR_SYSTEM;
freeze = 1;
} else {
/* notifier error, but no error in CPB flags? */
ata_ehi_push_desc(ehi, "unknown");
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
ehi->err_mask |= AC_ERR_OTHER;
freeze = 1;
}
/* Kill all commands. EH will determine what actually failed. */
if (freeze)
ata_port_freeze(ap);
else
ata_port_abort(ap);
libata: always use ata_qc_complete_multiple() for NCQ command completions Currently, sata_fsl, mv and nv call ata_qc_complete() multiple times from their interrupt handlers to indicate completion of NCQ commands. This limits the visibility the libata core layer has into how commands are being executed and completed, which is necessary to support IRQ expecting in generic way. libata already has an interface to complete multiple commands at once - ata_qc_complete_multiple() which ahci and sata_sil24 already use. This patch updates the three drivers to use ata_qc_complete_multiple() too and updates comments on ata_qc_complete[_multiple]() regarding their usages with NCQ completions. This change not only provides better visibility into command execution to the core layer but also simplifies low level drivers. * sata_fsl: It already builds done_mask. Conversion is straight forward. * sata_mv: mv_process_crpb_response() no longer checks for illegal completions, it just returns whether the tag is completed or not. mv_process_crpb_entries() builds done_mask from it and passes it to ata_qc_complete_multiple() which will check for illegal completions. * sata_nv adma: Similar to sata_mv. nv_adma_check_cpb() now just returns the tag status and nv_adma_interrupt() builds done_mask from it and passes it to ata_qc_complete_multiple(). * sata_nv swncq: It already builds done_mask. Drop unnecessary illegal transition checks and call ata_qc_complete_multiple(). In the long run, it might be a good idea to make ata_qc_complete() whine if called when multiple NCQ commands are in flight. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Ashish Kalra <ashish.kalra@freescale.com> Cc: Saeed Bishara <saeed@marvell.com> Cc: Mark Lord <liml@rtr.ca> Cc: Robert Hancock <hancockr@shaw.ca> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-06-25 21:03:34 +08:00
return -1;
}
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
libata: always use ata_qc_complete_multiple() for NCQ command completions Currently, sata_fsl, mv and nv call ata_qc_complete() multiple times from their interrupt handlers to indicate completion of NCQ commands. This limits the visibility the libata core layer has into how commands are being executed and completed, which is necessary to support IRQ expecting in generic way. libata already has an interface to complete multiple commands at once - ata_qc_complete_multiple() which ahci and sata_sil24 already use. This patch updates the three drivers to use ata_qc_complete_multiple() too and updates comments on ata_qc_complete[_multiple]() regarding their usages with NCQ completions. This change not only provides better visibility into command execution to the core layer but also simplifies low level drivers. * sata_fsl: It already builds done_mask. Conversion is straight forward. * sata_mv: mv_process_crpb_response() no longer checks for illegal completions, it just returns whether the tag is completed or not. mv_process_crpb_entries() builds done_mask from it and passes it to ata_qc_complete_multiple() which will check for illegal completions. * sata_nv adma: Similar to sata_mv. nv_adma_check_cpb() now just returns the tag status and nv_adma_interrupt() builds done_mask from it and passes it to ata_qc_complete_multiple(). * sata_nv swncq: It already builds done_mask. Drop unnecessary illegal transition checks and call ata_qc_complete_multiple(). In the long run, it might be a good idea to make ata_qc_complete() whine if called when multiple NCQ commands are in flight. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Ashish Kalra <ashish.kalra@freescale.com> Cc: Saeed Bishara <saeed@marvell.com> Cc: Mark Lord <liml@rtr.ca> Cc: Robert Hancock <hancockr@shaw.ca> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-06-25 21:03:34 +08:00
if (likely(flags & NV_CPB_RESP_DONE))
return 1;
sata_nv: cleanup ADMA error handling This cleans up a few issues with the error handling in sata_nv in ADMA mode to make it more consistent with other NCQ-capable drivers like ahci and sata_sil24: - When a command failed, we would effectively set AC_ERR_DEV on the queued command always. In the case of NCQ commands this prevents libata from doing a log page query to determine the details of the failed command, since it thinks we've already analyzed. Just set flags in the port ehi->err_mask, then freeze or abort and let libata figure out what went wrong. - The code handled NV_ADMA_STAT_CPBERR as a "really bad error" which caused it to set error flags on every queued command. I don't know exactly what this flag means (no docs, grr!) but from what I can guess from the standard ADMA spec, it just means that one or more of the CPBs had an error, so we just need to go through and do our normal checks in this case. - In the error_handler function the code would always dump the state of all the CPBs. This output seems redundant at this point since libata already dumps the state of all active commands on errors (and it also triggers at times when it shouldn't, like when suspending). Take this out. [akpm@osdl.org: many coding-style fixes] Signed-off-by: Robert Hancock <hancockr@shaw.ca> Cc: Jeff Garzik <jeff@garzik.org> Cc: Tejun Heo <htejun@gmail.com> Cc: Allen Martin <AMartin@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-06 08:26:01 +08:00
return 0;
}
static int nv_host_intr(struct ata_port *ap, u8 irq_stat)
{
struct ata_queued_cmd *qc = ata_qc_from_tag(ap, ap->link.active_tag);
/* freeze if hotplugged */
if (unlikely(irq_stat & (NV_INT_ADDED | NV_INT_REMOVED))) {
ata_port_freeze(ap);
return 1;
}
/* bail out if not our interrupt */
if (!(irq_stat & NV_INT_DEV))
return 0;
/* DEV interrupt w/ no active qc? */
if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) {
ata_sff_check_status(ap);
return 1;
}
/* handle interrupt */
return ata_bmdma_port_intr(ap, qc);
}
static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
int i, handled = 0;
u32 notifier_clears[2];
spin_lock(&host->lock);
for (i = 0; i < host->n_ports; i++) {
struct ata_port *ap = host->ports[i];
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u16 status;
u32 gen_ctl;
u32 notifier, notifier_error;
notifier_clears[i] = 0;
/* if ADMA is disabled, use standard ata interrupt handler */
if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) {
u8 irq_stat = readb(host->iomap[NV_MMIO_BAR] + NV_INT_STATUS_CK804)
>> (NV_INT_PORT_SHIFT * i);
handled += nv_host_intr(ap, irq_stat);
continue;
}
/* if in ATA register mode, check for standard interrupts */
if (pp->flags & NV_ADMA_PORT_REGISTER_MODE) {
u8 irq_stat = readb(host->iomap[NV_MMIO_BAR] + NV_INT_STATUS_CK804)
>> (NV_INT_PORT_SHIFT * i);
if (ata_tag_valid(ap->link.active_tag))
/** NV_INT_DEV indication seems unreliable
at times at least in ADMA mode. Force it
on always when a command is active, to
prevent losing interrupts. */
irq_stat |= NV_INT_DEV;
handled += nv_host_intr(ap, irq_stat);
}
notifier = readl(mmio + NV_ADMA_NOTIFIER);
notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
notifier_clears[i] = notifier | notifier_error;
gen_ctl = readl(pp->gen_block + NV_ADMA_GEN_CTL);
if (!NV_ADMA_CHECK_INTR(gen_ctl, ap->port_no) && !notifier &&
!notifier_error)
/* Nothing to do */
continue;
status = readw(mmio + NV_ADMA_STAT);
/*
* Clear status. Ensure the controller sees the
* clearing before we start looking at any of the CPB
* statuses, so that any CPB completions after this
* point in the handler will raise another interrupt.
*/
writew(status, mmio + NV_ADMA_STAT);
readw(mmio + NV_ADMA_STAT); /* flush posted write */
rmb();
handled++; /* irq handled if we got here */
/* freeze if hotplugged or controller error */
if (unlikely(status & (NV_ADMA_STAT_HOTPLUG |
NV_ADMA_STAT_HOTUNPLUG |
NV_ADMA_STAT_TIMEOUT |
NV_ADMA_STAT_SERROR))) {
struct ata_eh_info *ehi = &ap->link.eh_info;
ata_ehi_clear_desc(ehi);
__ata_ehi_push_desc(ehi, "ADMA status 0x%08x: ", status);
if (status & NV_ADMA_STAT_TIMEOUT) {
ehi->err_mask |= AC_ERR_SYSTEM;
ata_ehi_push_desc(ehi, "timeout");
} else if (status & NV_ADMA_STAT_HOTPLUG) {
ata_ehi_hotplugged(ehi);
ata_ehi_push_desc(ehi, "hotplug");
} else if (status & NV_ADMA_STAT_HOTUNPLUG) {
ata_ehi_hotplugged(ehi);
ata_ehi_push_desc(ehi, "hot unplug");
} else if (status & NV_ADMA_STAT_SERROR) {
/* let EH analyze SError and figure out cause */
ata_ehi_push_desc(ehi, "SError");
} else
ata_ehi_push_desc(ehi, "unknown");
ata_port_freeze(ap);
continue;
}
if (status & (NV_ADMA_STAT_DONE |
NV_ADMA_STAT_CPBERR |
NV_ADMA_STAT_CMD_COMPLETE)) {
u32 check_commands = notifier_clears[i];
libata: always use ata_qc_complete_multiple() for NCQ command completions Currently, sata_fsl, mv and nv call ata_qc_complete() multiple times from their interrupt handlers to indicate completion of NCQ commands. This limits the visibility the libata core layer has into how commands are being executed and completed, which is necessary to support IRQ expecting in generic way. libata already has an interface to complete multiple commands at once - ata_qc_complete_multiple() which ahci and sata_sil24 already use. This patch updates the three drivers to use ata_qc_complete_multiple() too and updates comments on ata_qc_complete[_multiple]() regarding their usages with NCQ completions. This change not only provides better visibility into command execution to the core layer but also simplifies low level drivers. * sata_fsl: It already builds done_mask. Conversion is straight forward. * sata_mv: mv_process_crpb_response() no longer checks for illegal completions, it just returns whether the tag is completed or not. mv_process_crpb_entries() builds done_mask from it and passes it to ata_qc_complete_multiple() which will check for illegal completions. * sata_nv adma: Similar to sata_mv. nv_adma_check_cpb() now just returns the tag status and nv_adma_interrupt() builds done_mask from it and passes it to ata_qc_complete_multiple(). * sata_nv swncq: It already builds done_mask. Drop unnecessary illegal transition checks and call ata_qc_complete_multiple(). In the long run, it might be a good idea to make ata_qc_complete() whine if called when multiple NCQ commands are in flight. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Ashish Kalra <ashish.kalra@freescale.com> Cc: Saeed Bishara <saeed@marvell.com> Cc: Mark Lord <liml@rtr.ca> Cc: Robert Hancock <hancockr@shaw.ca> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-06-25 21:03:34 +08:00
u32 done_mask = 0;
int pos, rc;
if (status & NV_ADMA_STAT_CPBERR) {
/* check all active commands */
if (ata_tag_valid(ap->link.active_tag))
check_commands = 1 <<
ap->link.active_tag;
else
check_commands = ap->link.sactive;
}
/* check CPBs for completed commands */
while ((pos = ffs(check_commands))) {
pos--;
rc = nv_adma_check_cpb(ap, pos,
notifier_error & (1 << pos));
libata: always use ata_qc_complete_multiple() for NCQ command completions Currently, sata_fsl, mv and nv call ata_qc_complete() multiple times from their interrupt handlers to indicate completion of NCQ commands. This limits the visibility the libata core layer has into how commands are being executed and completed, which is necessary to support IRQ expecting in generic way. libata already has an interface to complete multiple commands at once - ata_qc_complete_multiple() which ahci and sata_sil24 already use. This patch updates the three drivers to use ata_qc_complete_multiple() too and updates comments on ata_qc_complete[_multiple]() regarding their usages with NCQ completions. This change not only provides better visibility into command execution to the core layer but also simplifies low level drivers. * sata_fsl: It already builds done_mask. Conversion is straight forward. * sata_mv: mv_process_crpb_response() no longer checks for illegal completions, it just returns whether the tag is completed or not. mv_process_crpb_entries() builds done_mask from it and passes it to ata_qc_complete_multiple() which will check for illegal completions. * sata_nv adma: Similar to sata_mv. nv_adma_check_cpb() now just returns the tag status and nv_adma_interrupt() builds done_mask from it and passes it to ata_qc_complete_multiple(). * sata_nv swncq: It already builds done_mask. Drop unnecessary illegal transition checks and call ata_qc_complete_multiple(). In the long run, it might be a good idea to make ata_qc_complete() whine if called when multiple NCQ commands are in flight. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Ashish Kalra <ashish.kalra@freescale.com> Cc: Saeed Bishara <saeed@marvell.com> Cc: Mark Lord <liml@rtr.ca> Cc: Robert Hancock <hancockr@shaw.ca> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-06-25 21:03:34 +08:00
if (rc > 0)
done_mask |= 1 << pos;
else if (unlikely(rc < 0))
check_commands = 0;
check_commands &= ~(1 << pos);
}
libata: always use ata_qc_complete_multiple() for NCQ command completions Currently, sata_fsl, mv and nv call ata_qc_complete() multiple times from their interrupt handlers to indicate completion of NCQ commands. This limits the visibility the libata core layer has into how commands are being executed and completed, which is necessary to support IRQ expecting in generic way. libata already has an interface to complete multiple commands at once - ata_qc_complete_multiple() which ahci and sata_sil24 already use. This patch updates the three drivers to use ata_qc_complete_multiple() too and updates comments on ata_qc_complete[_multiple]() regarding their usages with NCQ completions. This change not only provides better visibility into command execution to the core layer but also simplifies low level drivers. * sata_fsl: It already builds done_mask. Conversion is straight forward. * sata_mv: mv_process_crpb_response() no longer checks for illegal completions, it just returns whether the tag is completed or not. mv_process_crpb_entries() builds done_mask from it and passes it to ata_qc_complete_multiple() which will check for illegal completions. * sata_nv adma: Similar to sata_mv. nv_adma_check_cpb() now just returns the tag status and nv_adma_interrupt() builds done_mask from it and passes it to ata_qc_complete_multiple(). * sata_nv swncq: It already builds done_mask. Drop unnecessary illegal transition checks and call ata_qc_complete_multiple(). In the long run, it might be a good idea to make ata_qc_complete() whine if called when multiple NCQ commands are in flight. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Ashish Kalra <ashish.kalra@freescale.com> Cc: Saeed Bishara <saeed@marvell.com> Cc: Mark Lord <liml@rtr.ca> Cc: Robert Hancock <hancockr@shaw.ca> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-06-25 21:03:34 +08:00
ata_qc_complete_multiple(ap, ap->qc_active ^ done_mask);
}
}
if (notifier_clears[0] || notifier_clears[1]) {
/* Note: Both notifier clear registers must be written
if either is set, even if one is zero, according to NVIDIA. */
struct nv_adma_port_priv *pp = host->ports[0]->private_data;
writel(notifier_clears[0], pp->notifier_clear_block);
pp = host->ports[1]->private_data;
writel(notifier_clears[1], pp->notifier_clear_block);
}
spin_unlock(&host->lock);
return IRQ_RETVAL(handled);
}
static void nv_adma_freeze(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u16 tmp;
nv_ck804_freeze(ap);
if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE)
return;
/* clear any outstanding CK804 notifications */
writeb(NV_INT_ALL << (ap->port_no * NV_INT_PORT_SHIFT),
ap->host->iomap[NV_MMIO_BAR] + NV_INT_STATUS_CK804);
/* Disable interrupt */
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp & ~(NV_ADMA_CTL_AIEN | NV_ADMA_CTL_HOTPLUG_IEN),
mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
}
static void nv_adma_thaw(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u16 tmp;
nv_ck804_thaw(ap);
if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE)
return;
/* Enable interrupt */
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp | (NV_ADMA_CTL_AIEN | NV_ADMA_CTL_HOTPLUG_IEN),
mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
}
static void nv_adma_irq_clear(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u32 notifier_clears[2];
if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) {
ata_bmdma_irq_clear(ap);
return;
}
/* clear any outstanding CK804 notifications */
writeb(NV_INT_ALL << (ap->port_no * NV_INT_PORT_SHIFT),
ap->host->iomap[NV_MMIO_BAR] + NV_INT_STATUS_CK804);
/* clear ADMA status */
writew(0xffff, mmio + NV_ADMA_STAT);
/* clear notifiers - note both ports need to be written with
something even though we are only clearing on one */
if (ap->port_no == 0) {
notifier_clears[0] = 0xFFFFFFFF;
notifier_clears[1] = 0;
} else {
notifier_clears[0] = 0;
notifier_clears[1] = 0xFFFFFFFF;
}
pp = ap->host->ports[0]->private_data;
writel(notifier_clears[0], pp->notifier_clear_block);
pp = ap->host->ports[1]->private_data;
writel(notifier_clears[1], pp->notifier_clear_block);
}
static void nv_adma_post_internal_cmd(struct ata_queued_cmd *qc)
{
struct nv_adma_port_priv *pp = qc->ap->private_data;
if (pp->flags & NV_ADMA_PORT_REGISTER_MODE)
ata_bmdma_post_internal_cmd(qc);
}
static int nv_adma_port_start(struct ata_port *ap)
{
struct device *dev = ap->host->dev;
struct nv_adma_port_priv *pp;
int rc;
void *mem;
dma_addr_t mem_dma;
void __iomem *mmio;
struct pci_dev *pdev = to_pci_dev(dev);
u16 tmp;
VPRINTK("ENTER\n");
/* Ensure DMA mask is set to 32-bit before allocating legacy PRD and
pad buffers */
rc = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (rc)
return rc;
rc = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
if (rc)
return rc;
libata-sff: clean up BMDMA initialization When BMDMA initialization failed or BMDMA was not available for whatever reason, bmdma_addr was left at zero and used as an indication that BMDMA shouldn't be used. This leads to the following problems. p1. For BMDMA drivers which don't use traditional BMDMA register, ata_bmdma_mode_filter() incorrectly inhibits DMA modes. Those drivers either have to inherit from ata_sff_port_ops or clear ->mode_filter explicitly. p2. non-BMDMA drivers call into BMDMA PRD table allocation. It doesn't actually allocate PRD table if bmdma_addr is not initialized but is still confusing. p3. For BMDMA drivers which don't use traditional BMDMA register, some methods might not be invoked as expected (e.g. bmdma_stop from ata_sff_post_internal_cmd()). p4. SFF drivers w/ custom DMA interface implement noop BMDMA ops worrying libata core might call into one of them. These problems are caused by the muddy line between SFF and BMDMA and the assumption that all BMDMA controllers initialize bmdma_addr. This patch fixes p1 and p2 by removing the bmdma_addr assumption and moving prd allocation to BMDMA port start. Later patches will fix the remaining issues. This patch improves BMDMA initialization such that * When BMDMA register initialization fails, falls back to PIO instead of failing. ata_pci_bmdma_init() never fails now. * When ata_pci_bmdma_init() falls back to PIO, it clears ap->mwdma_mask and udma_mask instead of depending on ata_bmdma_mode_filter(). This makes ata_bmdma_mode_filter() unnecessary thus resolving p1. * ata_port_start() which actually is BMDMA specific is moved to ata_bmdma_port_start(). ata_port_start() and ata_sff_port_start() are killed. * ata_sff_port_start32() is moved and renamed to ata_bmdma_port_start32(). Drivers which no longer call into PRD table allocation are... pdc_adma, sata_inic162x, sata_qstor, sata_sx4, pata_cmd640 and all drivers which inherit from ata_sff_port_ops. pata_icside sets ->port_start to ATA_OP_NULL as it doesn't need PRD but is a BMDMA controller and doesn't have custom port_start like other such controllers. Note that with the previous patch which makes all and only BMDMA drivers inherit from ata_bmdma_port_ops, this change doesn't break drivers which need PRD table. Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-05-11 03:41:34 +08:00
/* we might fallback to bmdma, allocate bmdma resources */
rc = ata_bmdma_port_start(ap);
if (rc)
return rc;
pp = devm_kzalloc(dev, sizeof(*pp), GFP_KERNEL);
if (!pp)
return -ENOMEM;
mmio = ap->host->iomap[NV_MMIO_BAR] + NV_ADMA_PORT +
ap->port_no * NV_ADMA_PORT_SIZE;
pp->ctl_block = mmio;
pp->gen_block = ap->host->iomap[NV_MMIO_BAR] + NV_ADMA_GEN;
pp->notifier_clear_block = pp->gen_block +
NV_ADMA_NOTIFIER_CLEAR + (4 * ap->port_no);
/* Now that the legacy PRD and padding buffer are allocated we can
safely raise the DMA mask to allocate the CPB/APRD table.
These are allowed to fail since we store the value that ends up
being used to set as the bounce limit in slave_config later if
needed. */
dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
pp->adma_dma_mask = *dev->dma_mask;
mem = dmam_alloc_coherent(dev, NV_ADMA_PORT_PRIV_DMA_SZ,
&mem_dma, GFP_KERNEL);
if (!mem)
return -ENOMEM;
memset(mem, 0, NV_ADMA_PORT_PRIV_DMA_SZ);
/*
* First item in chunk of DMA memory:
* 128-byte command parameter block (CPB)
* one for each command tag
*/
pp->cpb = mem;
pp->cpb_dma = mem_dma;
writel(mem_dma & 0xFFFFFFFF, mmio + NV_ADMA_CPB_BASE_LOW);
writel((mem_dma >> 16) >> 16, mmio + NV_ADMA_CPB_BASE_HIGH);
mem += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
mem_dma += NV_ADMA_MAX_CPBS * NV_ADMA_CPB_SZ;
/*
* Second item: block of ADMA_SGTBL_LEN s/g entries
*/
pp->aprd = mem;
pp->aprd_dma = mem_dma;
ap->private_data = pp;
/* clear any outstanding interrupt conditions */
writew(0xffff, mmio + NV_ADMA_STAT);
/* initialize port variables */
pp->flags = NV_ADMA_PORT_REGISTER_MODE;
/* clear CPB fetch count */
writew(0, mmio + NV_ADMA_CPB_COUNT);
/* clear GO for register mode, enable interrupt */
tmp = readw(mmio + NV_ADMA_CTL);
writew((tmp & ~NV_ADMA_CTL_GO) | NV_ADMA_CTL_AIEN |
NV_ADMA_CTL_HOTPLUG_IEN, mmio + NV_ADMA_CTL);
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
udelay(1);
writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
return 0;
}
static void nv_adma_port_stop(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
VPRINTK("ENTER\n");
writew(0, mmio + NV_ADMA_CTL);
}
#ifdef CONFIG_PM
static int nv_adma_port_suspend(struct ata_port *ap, pm_message_t mesg)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
/* Go to register mode - clears GO */
nv_adma_register_mode(ap);
/* clear CPB fetch count */
writew(0, mmio + NV_ADMA_CPB_COUNT);
/* disable interrupt, shut down port */
writew(0, mmio + NV_ADMA_CTL);
return 0;
}
static int nv_adma_port_resume(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
void __iomem *mmio = pp->ctl_block;
u16 tmp;
/* set CPB block location */
writel(pp->cpb_dma & 0xFFFFFFFF, mmio + NV_ADMA_CPB_BASE_LOW);
writel((pp->cpb_dma >> 16) >> 16, mmio + NV_ADMA_CPB_BASE_HIGH);
/* clear any outstanding interrupt conditions */
writew(0xffff, mmio + NV_ADMA_STAT);
/* initialize port variables */
pp->flags |= NV_ADMA_PORT_REGISTER_MODE;
/* clear CPB fetch count */
writew(0, mmio + NV_ADMA_CPB_COUNT);
/* clear GO for register mode, enable interrupt */
tmp = readw(mmio + NV_ADMA_CTL);
writew((tmp & ~NV_ADMA_CTL_GO) | NV_ADMA_CTL_AIEN |
NV_ADMA_CTL_HOTPLUG_IEN, mmio + NV_ADMA_CTL);
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
udelay(1);
writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
return 0;
}
#endif
static void nv_adma_setup_port(struct ata_port *ap)
{
void __iomem *mmio = ap->host->iomap[NV_MMIO_BAR];
struct ata_ioports *ioport = &ap->ioaddr;
VPRINTK("ENTER\n");
mmio += NV_ADMA_PORT + ap->port_no * NV_ADMA_PORT_SIZE;
ioport->cmd_addr = mmio;
ioport->data_addr = mmio + (ATA_REG_DATA * 4);
ioport->error_addr =
ioport->feature_addr = mmio + (ATA_REG_ERR * 4);
ioport->nsect_addr = mmio + (ATA_REG_NSECT * 4);
ioport->lbal_addr = mmio + (ATA_REG_LBAL * 4);
ioport->lbam_addr = mmio + (ATA_REG_LBAM * 4);
ioport->lbah_addr = mmio + (ATA_REG_LBAH * 4);
ioport->device_addr = mmio + (ATA_REG_DEVICE * 4);
ioport->status_addr =
ioport->command_addr = mmio + (ATA_REG_STATUS * 4);
ioport->altstatus_addr =
ioport->ctl_addr = mmio + 0x20;
}
static int nv_adma_host_init(struct ata_host *host)
{
struct pci_dev *pdev = to_pci_dev(host->dev);
unsigned int i;
u32 tmp32;
VPRINTK("ENTER\n");
/* enable ADMA on the ports */
pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
tmp32 |= NV_MCP_SATA_CFG_20_PORT0_EN |
NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
NV_MCP_SATA_CFG_20_PORT1_EN |
NV_MCP_SATA_CFG_20_PORT1_PWB_EN;
pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
for (i = 0; i < host->n_ports; i++)
nv_adma_setup_port(host->ports[i]);
return 0;
}
static void nv_adma_fill_aprd(struct ata_queued_cmd *qc,
struct scatterlist *sg,
int idx,
struct nv_adma_prd *aprd)
{
u8 flags = 0;
if (qc->tf.flags & ATA_TFLAG_WRITE)
flags |= NV_APRD_WRITE;
if (idx == qc->n_elem - 1)
flags |= NV_APRD_END;
else if (idx != 4)
flags |= NV_APRD_CONT;
aprd->addr = cpu_to_le64(((u64)sg_dma_address(sg)));
aprd->len = cpu_to_le32(((u32)sg_dma_len(sg))); /* len in bytes */
aprd->flags = flags;
aprd->packet_len = 0;
}
static void nv_adma_fill_sg(struct ata_queued_cmd *qc, struct nv_adma_cpb *cpb)
{
struct nv_adma_port_priv *pp = qc->ap->private_data;
struct nv_adma_prd *aprd;
struct scatterlist *sg;
unsigned int si;
VPRINTK("ENTER\n");
for_each_sg(qc->sg, sg, qc->n_elem, si) {
aprd = (si < 5) ? &cpb->aprd[si] :
&pp->aprd[NV_ADMA_SGTBL_LEN * qc->tag + (si-5)];
nv_adma_fill_aprd(qc, sg, si, aprd);
}
if (si > 5)
cpb->next_aprd = cpu_to_le64(((u64)(pp->aprd_dma + NV_ADMA_SGTBL_SZ * qc->tag)));
else
cpb->next_aprd = cpu_to_le64(0);
}
static int nv_adma_use_reg_mode(struct ata_queued_cmd *qc)
{
struct nv_adma_port_priv *pp = qc->ap->private_data;
/* ADMA engine can only be used for non-ATAPI DMA commands,
or interrupt-driven no-data commands. */
if ((pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) ||
(qc->tf.flags & ATA_TFLAG_POLLING))
return 1;
if ((qc->flags & ATA_QCFLAG_DMAMAP) ||
(qc->tf.protocol == ATA_PROT_NODATA))
return 0;
return 1;
}
static void nv_adma_qc_prep(struct ata_queued_cmd *qc)
{
struct nv_adma_port_priv *pp = qc->ap->private_data;
struct nv_adma_cpb *cpb = &pp->cpb[qc->tag];
u8 ctl_flags = NV_CPB_CTL_CPB_VALID |
NV_CPB_CTL_IEN;
if (nv_adma_use_reg_mode(qc)) {
BUG_ON(!(pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) &&
(qc->flags & ATA_QCFLAG_DMAMAP));
nv_adma_register_mode(qc->ap);
ata_bmdma_qc_prep(qc);
return;
}
cpb->resp_flags = NV_CPB_RESP_DONE;
wmb();
cpb->ctl_flags = 0;
wmb();
cpb->len = 3;
cpb->tag = qc->tag;
cpb->next_cpb_idx = 0;
/* turn on NCQ flags for NCQ commands */
if (qc->tf.protocol == ATA_PROT_NCQ)
ctl_flags |= NV_CPB_CTL_QUEUE | NV_CPB_CTL_FPDMA;
VPRINTK("qc->flags = 0x%lx\n", qc->flags);
nv_adma_tf_to_cpb(&qc->tf, cpb->tf);
if (qc->flags & ATA_QCFLAG_DMAMAP) {
nv_adma_fill_sg(qc, cpb);
ctl_flags |= NV_CPB_CTL_APRD_VALID;
} else
memset(&cpb->aprd[0], 0, sizeof(struct nv_adma_prd) * 5);
/* Be paranoid and don't let the device see NV_CPB_CTL_CPB_VALID
until we are finished filling in all of the contents */
wmb();
cpb->ctl_flags = ctl_flags;
wmb();
cpb->resp_flags = 0;
}
static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc)
{
struct nv_adma_port_priv *pp = qc->ap->private_data;
void __iomem *mmio = pp->ctl_block;
int curr_ncq = (qc->tf.protocol == ATA_PROT_NCQ);
VPRINTK("ENTER\n");
/* We can't handle result taskfile with NCQ commands, since
retrieving the taskfile switches us out of ADMA mode and would abort
existing commands. */
if (unlikely(qc->tf.protocol == ATA_PROT_NCQ &&
(qc->flags & ATA_QCFLAG_RESULT_TF))) {
ata_dev_err(qc->dev, "NCQ w/ RESULT_TF not allowed\n");
return AC_ERR_SYSTEM;
}
if (nv_adma_use_reg_mode(qc)) {
/* use ATA register mode */
VPRINTK("using ATA register mode: 0x%lx\n", qc->flags);
BUG_ON(!(pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) &&
(qc->flags & ATA_QCFLAG_DMAMAP));
nv_adma_register_mode(qc->ap);
return ata_bmdma_qc_issue(qc);
} else
nv_adma_mode(qc->ap);
/* write append register, command tag in lower 8 bits
and (number of cpbs to append -1) in top 8 bits */
wmb();
if (curr_ncq != pp->last_issue_ncq) {
/* Seems to need some delay before switching between NCQ and
non-NCQ commands, else we get command timeouts and such. */
udelay(20);
pp->last_issue_ncq = curr_ncq;
}
writew(qc->tag, mmio + NV_ADMA_APPEND);
DPRINTK("Issued tag %u\n", qc->tag);
return 0;
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
unsigned int i;
unsigned int handled = 0;
unsigned long flags;
spin_lock_irqsave(&host->lock, flags);
for (i = 0; i < host->n_ports; i++) {
struct ata_port *ap = host->ports[i];
struct ata_queued_cmd *qc;
qc = ata_qc_from_tag(ap, ap->link.active_tag);
if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING))) {
handled += ata_bmdma_port_intr(ap, qc);
} else {
/*
* No request pending? Clear interrupt status
* anyway, in case there's one pending.
*/
ap->ops->sff_check_status(ap);
}
}
spin_unlock_irqrestore(&host->lock, flags);
return IRQ_RETVAL(handled);
}
static irqreturn_t nv_do_interrupt(struct ata_host *host, u8 irq_stat)
{
int i, handled = 0;
for (i = 0; i < host->n_ports; i++) {
handled += nv_host_intr(host->ports[i], irq_stat);
irq_stat >>= NV_INT_PORT_SHIFT;
}
return IRQ_RETVAL(handled);
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t nv_nf2_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
u8 irq_stat;
irqreturn_t ret;
spin_lock(&host->lock);
irq_stat = ioread8(host->ports[0]->ioaddr.scr_addr + NV_INT_STATUS);
ret = nv_do_interrupt(host, irq_stat);
spin_unlock(&host->lock);
return ret;
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t nv_ck804_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
u8 irq_stat;
irqreturn_t ret;
spin_lock(&host->lock);
irq_stat = readb(host->iomap[NV_MMIO_BAR] + NV_INT_STATUS_CK804);
ret = nv_do_interrupt(host, irq_stat);
spin_unlock(&host->lock);
return ret;
}
static int nv_scr_read(struct ata_link *link, unsigned int sc_reg, u32 *val)
{
if (sc_reg > SCR_CONTROL)
return -EINVAL;
*val = ioread32(link->ap->ioaddr.scr_addr + (sc_reg * 4));
return 0;
}
static int nv_scr_write(struct ata_link *link, unsigned int sc_reg, u32 val)
{
if (sc_reg > SCR_CONTROL)
return -EINVAL;
iowrite32(val, link->ap->ioaddr.scr_addr + (sc_reg * 4));
return 0;
}
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
static int nv_hardreset(struct ata_link *link, unsigned int *class,
unsigned long deadline)
{
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
struct ata_eh_context *ehc = &link->eh_context;
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
/* Do hardreset iff it's post-boot probing, please read the
* comment above port ops for details.
*/
if (!(link->ap->pflags & ATA_PFLAG_LOADING) &&
!ata_dev_enabled(link->device))
sata_link_hardreset(link, sata_deb_timing_hotplug, deadline,
NULL, NULL);
else {
const unsigned long *timing = sata_ehc_deb_timing(ehc);
int rc;
if (!(ehc->i.flags & ATA_EHI_QUIET))
ata_link_info(link,
"nv: skipping hardreset on occupied port\n");
/* make sure the link is online */
rc = sata_link_resume(link, timing, deadline);
/* whine about phy resume failure but proceed */
if (rc && rc != -EOPNOTSUPP)
ata_link_warn(link, "failed to resume link (errno=%d)\n",
rc);
}
sata_nv: use hardreset only for post-boot probing When I thought it was finally defeated, it came back with vengeance. The failure cases are ever more convoluted. Now there is a single combination which fails boot probing - MCP5x + Intel SSD and there are two hotplug failure reports on different flavors where softreset fails to bring up the device. Through the many bug reports after the switch to hardreset, the following patterns emerged. - Softreset during boot always works. - Hardreset during boot sometimes fails to bring up the link on certain comibnations and device signature acquisition is unreliable. - Hardreset is often necessary after hotplug. It looks like the old behavior of preferring softreset was somehow pretty close to the working reset protocol although it could have lost a device during phy error handling by issuing hardreset. This patch implements nv_hardreset() which kicks in only for post-boot (!LOADING) device probing resets. This should be able to work around all known problem cases. This isn't perfect but given the various hardreset quirks on these controllers, I think this is as good as it can get. Tested on mcp5x (swncq), nf3 and ck804 for all both boot, warm and hot probing cases. Kudos to all the bug reporters and their painful hours with these damn controllers. ;-) Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Robert Hancock <hancockr@shaw.ca> Reported-by: David Lang <david@lang.hm> Reported-by: Samo Vodopivec <lament.email.si@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2009-06-10 15:29:07 +08:00
/* device signature acquisition is unreliable */
return -EAGAIN;
}
static void nv_nf2_freeze(struct ata_port *ap)
{
void __iomem *scr_addr = ap->host->ports[0]->ioaddr.scr_addr;
int shift = ap->port_no * NV_INT_PORT_SHIFT;
u8 mask;
mask = ioread8(scr_addr + NV_INT_ENABLE);
mask &= ~(NV_INT_ALL << shift);
iowrite8(mask, scr_addr + NV_INT_ENABLE);
}
static void nv_nf2_thaw(struct ata_port *ap)
{
void __iomem *scr_addr = ap->host->ports[0]->ioaddr.scr_addr;
int shift = ap->port_no * NV_INT_PORT_SHIFT;
u8 mask;
iowrite8(NV_INT_ALL << shift, scr_addr + NV_INT_STATUS);
mask = ioread8(scr_addr + NV_INT_ENABLE);
mask |= (NV_INT_MASK << shift);
iowrite8(mask, scr_addr + NV_INT_ENABLE);
}
static void nv_ck804_freeze(struct ata_port *ap)
{
void __iomem *mmio_base = ap->host->iomap[NV_MMIO_BAR];
int shift = ap->port_no * NV_INT_PORT_SHIFT;
u8 mask;
mask = readb(mmio_base + NV_INT_ENABLE_CK804);
mask &= ~(NV_INT_ALL << shift);
writeb(mask, mmio_base + NV_INT_ENABLE_CK804);
}
static void nv_ck804_thaw(struct ata_port *ap)
{
void __iomem *mmio_base = ap->host->iomap[NV_MMIO_BAR];
int shift = ap->port_no * NV_INT_PORT_SHIFT;
u8 mask;
writeb(NV_INT_ALL << shift, mmio_base + NV_INT_STATUS_CK804);
mask = readb(mmio_base + NV_INT_ENABLE_CK804);
mask |= (NV_INT_MASK << shift);
writeb(mask, mmio_base + NV_INT_ENABLE_CK804);
}
static void nv_mcp55_freeze(struct ata_port *ap)
{
void __iomem *mmio_base = ap->host->iomap[NV_MMIO_BAR];
int shift = ap->port_no * NV_INT_PORT_SHIFT_MCP55;
u32 mask;
writel(NV_INT_ALL_MCP55 << shift, mmio_base + NV_INT_STATUS_MCP55);
mask = readl(mmio_base + NV_INT_ENABLE_MCP55);
mask &= ~(NV_INT_ALL_MCP55 << shift);
writel(mask, mmio_base + NV_INT_ENABLE_MCP55);
}
static void nv_mcp55_thaw(struct ata_port *ap)
{
void __iomem *mmio_base = ap->host->iomap[NV_MMIO_BAR];
int shift = ap->port_no * NV_INT_PORT_SHIFT_MCP55;
u32 mask;
writel(NV_INT_ALL_MCP55 << shift, mmio_base + NV_INT_STATUS_MCP55);
mask = readl(mmio_base + NV_INT_ENABLE_MCP55);
mask |= (NV_INT_MASK_MCP55 << shift);
writel(mask, mmio_base + NV_INT_ENABLE_MCP55);
}
static void nv_adma_error_handler(struct ata_port *ap)
{
struct nv_adma_port_priv *pp = ap->private_data;
if (!(pp->flags & NV_ADMA_PORT_REGISTER_MODE)) {
void __iomem *mmio = pp->ctl_block;
int i;
u16 tmp;
if (ata_tag_valid(ap->link.active_tag) || ap->link.sactive) {
u32 notifier = readl(mmio + NV_ADMA_NOTIFIER);
u32 notifier_error = readl(mmio + NV_ADMA_NOTIFIER_ERROR);
u32 gen_ctl = readl(pp->gen_block + NV_ADMA_GEN_CTL);
u32 status = readw(mmio + NV_ADMA_STAT);
u8 cpb_count = readb(mmio + NV_ADMA_CPB_COUNT);
u8 next_cpb_idx = readb(mmio + NV_ADMA_NEXT_CPB_IDX);
ata_port_err(ap,
"EH in ADMA mode, notifier 0x%X "
"notifier_error 0x%X gen_ctl 0x%X status 0x%X "
"next cpb count 0x%X next cpb idx 0x%x\n",
notifier, notifier_error, gen_ctl, status,
cpb_count, next_cpb_idx);
for (i = 0; i < NV_ADMA_MAX_CPBS; i++) {
struct nv_adma_cpb *cpb = &pp->cpb[i];
if ((ata_tag_valid(ap->link.active_tag) && i == ap->link.active_tag) ||
ap->link.sactive & (1 << i))
ata_port_err(ap,
"CPB %d: ctl_flags 0x%x, resp_flags 0x%x\n",
i, cpb->ctl_flags, cpb->resp_flags);
}
}
/* Push us back into port register mode for error handling. */
nv_adma_register_mode(ap);
/* Mark all of the CPBs as invalid to prevent them from
being executed */
for (i = 0; i < NV_ADMA_MAX_CPBS; i++)
pp->cpb[i].ctl_flags &= ~NV_CPB_CTL_CPB_VALID;
/* clear CPB fetch count */
writew(0, mmio + NV_ADMA_CPB_COUNT);
/* Reset channel */
tmp = readw(mmio + NV_ADMA_CTL);
writew(tmp | NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
udelay(1);
writew(tmp & ~NV_ADMA_CTL_CHANNEL_RESET, mmio + NV_ADMA_CTL);
readw(mmio + NV_ADMA_CTL); /* flush posted write */
}
ata_bmdma_error_handler(ap);
}
static void nv_swncq_qc_to_dq(struct ata_port *ap, struct ata_queued_cmd *qc)
{
struct nv_swncq_port_priv *pp = ap->private_data;
struct defer_queue *dq = &pp->defer_queue;
/* queue is full */
WARN_ON(dq->tail - dq->head == ATA_MAX_QUEUE);
dq->defer_bits |= (1 << qc->tag);
dq->tag[dq->tail++ & (ATA_MAX_QUEUE - 1)] = qc->tag;
}
static struct ata_queued_cmd *nv_swncq_qc_from_dq(struct ata_port *ap)
{
struct nv_swncq_port_priv *pp = ap->private_data;
struct defer_queue *dq = &pp->defer_queue;
unsigned int tag;
if (dq->head == dq->tail) /* null queue */
return NULL;
tag = dq->tag[dq->head & (ATA_MAX_QUEUE - 1)];
dq->tag[dq->head++ & (ATA_MAX_QUEUE - 1)] = ATA_TAG_POISON;
WARN_ON(!(dq->defer_bits & (1 << tag)));
dq->defer_bits &= ~(1 << tag);
return ata_qc_from_tag(ap, tag);
}
static void nv_swncq_fis_reinit(struct ata_port *ap)
{
struct nv_swncq_port_priv *pp = ap->private_data;
pp->dhfis_bits = 0;
pp->dmafis_bits = 0;
pp->sdbfis_bits = 0;
pp->ncq_flags = 0;
}
static void nv_swncq_pp_reinit(struct ata_port *ap)
{
struct nv_swncq_port_priv *pp = ap->private_data;
struct defer_queue *dq = &pp->defer_queue;
dq->head = 0;
dq->tail = 0;
dq->defer_bits = 0;
pp->qc_active = 0;
pp->last_issue_tag = ATA_TAG_POISON;
nv_swncq_fis_reinit(ap);
}
static void nv_swncq_irq_clear(struct ata_port *ap, u16 fis)
{
struct nv_swncq_port_priv *pp = ap->private_data;
writew(fis, pp->irq_block);
}
static void __ata_bmdma_stop(struct ata_port *ap)
{
struct ata_queued_cmd qc;
qc.ap = ap;
ata_bmdma_stop(&qc);
}
static void nv_swncq_ncq_stop(struct ata_port *ap)
{
struct nv_swncq_port_priv *pp = ap->private_data;
unsigned int i;
u32 sactive;
u32 done_mask;
ata_port_err(ap, "EH in SWNCQ mode,QC:qc_active 0x%X sactive 0x%X\n",
ap->qc_active, ap->link.sactive);
ata_port_err(ap,
"SWNCQ:qc_active 0x%X defer_bits 0x%X last_issue_tag 0x%x\n "
"dhfis 0x%X dmafis 0x%X sdbfis 0x%X\n",
pp->qc_active, pp->defer_queue.defer_bits, pp->last_issue_tag,
pp->dhfis_bits, pp->dmafis_bits, pp->sdbfis_bits);
ata_port_err(ap, "ATA_REG 0x%X ERR_REG 0x%X\n",
ap->ops->sff_check_status(ap),
ioread8(ap->ioaddr.error_addr));
sactive = readl(pp->sactive_block);
done_mask = pp->qc_active ^ sactive;
ata_port_err(ap, "tag : dhfis dmafis sdbfis sactive\n");
for (i = 0; i < ATA_MAX_QUEUE; i++) {
u8 err = 0;
if (pp->qc_active & (1 << i))
err = 0;
else if (done_mask & (1 << i))
err = 1;
else
continue;
ata_port_err(ap,
"tag 0x%x: %01x %01x %01x %01x %s\n", i,
(pp->dhfis_bits >> i) & 0x1,
(pp->dmafis_bits >> i) & 0x1,
(pp->sdbfis_bits >> i) & 0x1,
(sactive >> i) & 0x1,
(err ? "error! tag doesn't exit" : " "));
}
nv_swncq_pp_reinit(ap);
ap->ops->sff_irq_clear(ap);
__ata_bmdma_stop(ap);
nv_swncq_irq_clear(ap, 0xffff);
}
static void nv_swncq_error_handler(struct ata_port *ap)
{
struct ata_eh_context *ehc = &ap->link.eh_context;
if (ap->link.sactive) {
nv_swncq_ncq_stop(ap);
libata: prefer hardreset When both soft and hard resets are available, libata preferred softreset till now. The logic behind it was to be softer to devices; however, this doesn't really help much. Rationales for the change: * BIOS may freeze lock certain things during boot and softreset can't unlock those. This by itself is okay but during operation PHY event or other error conditions can trigger hardreset and the device may end up with different configuration. For example, after a hardreset, previously unlockable HPA can be unlocked resulting in different device size and thus revalidation failure. Similar condition can occur during or after resume. * Certain ATAPI devices require hardreset to recover after certain error conditions. On PATA, this is done by issuing the DEVICE RESET command. On SATA, COMRESET has equivalent effect. The problem is that DEVICE RESET needs its own execution protocol. For SFF controllers with bare TF access, it can be easily implemented but more advanced controllers (e.g. ahci and sata_sil24) require specialized implementations. Simply using hardreset solves the problem nicely. * COMRESET initialization sequence is the norm in SATA land and many SATA devices don't work properly if only SRST is used. For example, some PMPs behave this way and libata works around by always issuing hardreset if the host supports PMP. Like the above example, libata has developed a number of mechanisms aiming to promote softreset to hardreset if softreset is not going to work. This approach is time consuming and error prone. Also, note that, dependingon how you read the specs, it could be argued that PMP fan-out ports require COMRESET to start operation. In fact, all the PMPs on the market except one don't work properly if COMRESET is not issued to fan-out ports after PMP reset. * COMRESET is an integral part of SATA connection and any working device should be able to handle COMRESET properly. After all, it's the way to signal hardreset during reboot. This is the most used and recommended (at least by the ahci spec) method of resetting devices. So, this patch makes libata prefer hardreset over softreset by making the following changes. * Rename ATA_EH_RESET_MASK to ATA_EH_RESET and use it whereever ATA_EH_{SOFT|HARD}RESET used to be used. ATA_EH_{SOFT|HARD}RESET is now only used to tell prereset whether soft or hard reset will be issued. * Strip out now unneeded promote-to-hardreset logics from ata_eh_reset(), ata_std_prereset(), sata_pmp_std_prereset() and other places. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-01-23 23:05:14 +08:00
ehc->i.action |= ATA_EH_RESET;
}
ata_bmdma_error_handler(ap);
}
#ifdef CONFIG_PM
static int nv_swncq_port_suspend(struct ata_port *ap, pm_message_t mesg)
{
void __iomem *mmio = ap->host->iomap[NV_MMIO_BAR];
u32 tmp;
/* clear irq */
writel(~0, mmio + NV_INT_STATUS_MCP55);
/* disable irq */
writel(0, mmio + NV_INT_ENABLE_MCP55);
/* disable swncq */
tmp = readl(mmio + NV_CTL_MCP55);
tmp &= ~(NV_CTL_PRI_SWNCQ | NV_CTL_SEC_SWNCQ);
writel(tmp, mmio + NV_CTL_MCP55);
return 0;
}
static int nv_swncq_port_resume(struct ata_port *ap)
{
void __iomem *mmio = ap->host->iomap[NV_MMIO_BAR];
u32 tmp;
/* clear irq */
writel(~0, mmio + NV_INT_STATUS_MCP55);
/* enable irq */
writel(0x00fd00fd, mmio + NV_INT_ENABLE_MCP55);
/* enable swncq */
tmp = readl(mmio + NV_CTL_MCP55);
writel(tmp | NV_CTL_PRI_SWNCQ | NV_CTL_SEC_SWNCQ, mmio + NV_CTL_MCP55);
return 0;
}
#endif
static void nv_swncq_host_init(struct ata_host *host)
{
u32 tmp;
void __iomem *mmio = host->iomap[NV_MMIO_BAR];
struct pci_dev *pdev = to_pci_dev(host->dev);
u8 regval;
/* disable ECO 398 */
pci_read_config_byte(pdev, 0x7f, &regval);
regval &= ~(1 << 7);
pci_write_config_byte(pdev, 0x7f, regval);
/* enable swncq */
tmp = readl(mmio + NV_CTL_MCP55);
VPRINTK("HOST_CTL:0x%X\n", tmp);
writel(tmp | NV_CTL_PRI_SWNCQ | NV_CTL_SEC_SWNCQ, mmio + NV_CTL_MCP55);
/* enable irq intr */
tmp = readl(mmio + NV_INT_ENABLE_MCP55);
VPRINTK("HOST_ENABLE:0x%X\n", tmp);
writel(tmp | 0x00fd00fd, mmio + NV_INT_ENABLE_MCP55);
/* clear port irq */
writel(~0x0, mmio + NV_INT_STATUS_MCP55);
}
static int nv_swncq_slave_config(struct scsi_device *sdev)
{
struct ata_port *ap = ata_shost_to_port(sdev->host);
struct pci_dev *pdev = to_pci_dev(ap->host->dev);
struct ata_device *dev;
int rc;
u8 rev;
u8 check_maxtor = 0;
unsigned char model_num[ATA_ID_PROD_LEN + 1];
rc = ata_scsi_slave_config(sdev);
if (sdev->id >= ATA_MAX_DEVICES || sdev->channel || sdev->lun)
/* Not a proper libata device, ignore */
return rc;
dev = &ap->link.device[sdev->id];
if (!(ap->flags & ATA_FLAG_NCQ) || dev->class == ATA_DEV_ATAPI)
return rc;
/* if MCP51 and Maxtor, then disable ncq */
if (pdev->device == PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA ||
pdev->device == PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA2)
check_maxtor = 1;
/* if MCP55 and rev <= a2 and Maxtor, then disable ncq */
if (pdev->device == PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA ||
pdev->device == PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA2) {
pci_read_config_byte(pdev, 0x8, &rev);
if (rev <= 0xa2)
check_maxtor = 1;
}
if (!check_maxtor)
return rc;
ata_id_c_string(dev->id, model_num, ATA_ID_PROD, sizeof(model_num));
if (strncmp(model_num, "Maxtor", 6) == 0) {
ata_scsi_change_queue_depth(sdev, 1);
ata_dev_notice(dev, "Disabling SWNCQ mode (depth %x)\n",
sdev->queue_depth);
}
return rc;
}
static int nv_swncq_port_start(struct ata_port *ap)
{
struct device *dev = ap->host->dev;
void __iomem *mmio = ap->host->iomap[NV_MMIO_BAR];
struct nv_swncq_port_priv *pp;
int rc;
libata-sff: clean up BMDMA initialization When BMDMA initialization failed or BMDMA was not available for whatever reason, bmdma_addr was left at zero and used as an indication that BMDMA shouldn't be used. This leads to the following problems. p1. For BMDMA drivers which don't use traditional BMDMA register, ata_bmdma_mode_filter() incorrectly inhibits DMA modes. Those drivers either have to inherit from ata_sff_port_ops or clear ->mode_filter explicitly. p2. non-BMDMA drivers call into BMDMA PRD table allocation. It doesn't actually allocate PRD table if bmdma_addr is not initialized but is still confusing. p3. For BMDMA drivers which don't use traditional BMDMA register, some methods might not be invoked as expected (e.g. bmdma_stop from ata_sff_post_internal_cmd()). p4. SFF drivers w/ custom DMA interface implement noop BMDMA ops worrying libata core might call into one of them. These problems are caused by the muddy line between SFF and BMDMA and the assumption that all BMDMA controllers initialize bmdma_addr. This patch fixes p1 and p2 by removing the bmdma_addr assumption and moving prd allocation to BMDMA port start. Later patches will fix the remaining issues. This patch improves BMDMA initialization such that * When BMDMA register initialization fails, falls back to PIO instead of failing. ata_pci_bmdma_init() never fails now. * When ata_pci_bmdma_init() falls back to PIO, it clears ap->mwdma_mask and udma_mask instead of depending on ata_bmdma_mode_filter(). This makes ata_bmdma_mode_filter() unnecessary thus resolving p1. * ata_port_start() which actually is BMDMA specific is moved to ata_bmdma_port_start(). ata_port_start() and ata_sff_port_start() are killed. * ata_sff_port_start32() is moved and renamed to ata_bmdma_port_start32(). Drivers which no longer call into PRD table allocation are... pdc_adma, sata_inic162x, sata_qstor, sata_sx4, pata_cmd640 and all drivers which inherit from ata_sff_port_ops. pata_icside sets ->port_start to ATA_OP_NULL as it doesn't need PRD but is a BMDMA controller and doesn't have custom port_start like other such controllers. Note that with the previous patch which makes all and only BMDMA drivers inherit from ata_bmdma_port_ops, this change doesn't break drivers which need PRD table. Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-05-11 03:41:34 +08:00
/* we might fallback to bmdma, allocate bmdma resources */
rc = ata_bmdma_port_start(ap);
if (rc)
return rc;
pp = devm_kzalloc(dev, sizeof(*pp), GFP_KERNEL);
if (!pp)
return -ENOMEM;
pp->prd = dmam_alloc_coherent(dev, ATA_PRD_TBL_SZ * ATA_MAX_QUEUE,
&pp->prd_dma, GFP_KERNEL);
if (!pp->prd)
return -ENOMEM;
memset(pp->prd, 0, ATA_PRD_TBL_SZ * ATA_MAX_QUEUE);
ap->private_data = pp;
pp->sactive_block = ap->ioaddr.scr_addr + 4 * SCR_ACTIVE;
pp->irq_block = mmio + NV_INT_STATUS_MCP55 + ap->port_no * 2;
pp->tag_block = mmio + NV_NCQ_REG_MCP55 + ap->port_no * 2;
return 0;
}
static void nv_swncq_qc_prep(struct ata_queued_cmd *qc)
{
if (qc->tf.protocol != ATA_PROT_NCQ) {
ata_bmdma_qc_prep(qc);
return;
}
if (!(qc->flags & ATA_QCFLAG_DMAMAP))
return;
nv_swncq_fill_sg(qc);
}
static void nv_swncq_fill_sg(struct ata_queued_cmd *qc)
{
struct ata_port *ap = qc->ap;
struct scatterlist *sg;
struct nv_swncq_port_priv *pp = ap->private_data;
struct ata_bmdma_prd *prd;
unsigned int si, idx;
prd = pp->prd + ATA_MAX_PRD * qc->tag;
idx = 0;
for_each_sg(qc->sg, sg, qc->n_elem, si) {
u32 addr, offset;
u32 sg_len, len;
addr = (u32)sg_dma_address(sg);
sg_len = sg_dma_len(sg);
while (sg_len) {
offset = addr & 0xffff;
len = sg_len;
if ((offset + sg_len) > 0x10000)
len = 0x10000 - offset;
prd[idx].addr = cpu_to_le32(addr);
prd[idx].flags_len = cpu_to_le32(len & 0xffff);
idx++;
sg_len -= len;
addr += len;
}
}
prd[idx - 1].flags_len |= cpu_to_le32(ATA_PRD_EOT);
}
static unsigned int nv_swncq_issue_atacmd(struct ata_port *ap,
struct ata_queued_cmd *qc)
{
struct nv_swncq_port_priv *pp = ap->private_data;
if (qc == NULL)
return 0;
DPRINTK("Enter\n");
writel((1 << qc->tag), pp->sactive_block);
pp->last_issue_tag = qc->tag;
pp->dhfis_bits &= ~(1 << qc->tag);
pp->dmafis_bits &= ~(1 << qc->tag);
pp->qc_active |= (0x1 << qc->tag);
ap->ops->sff_tf_load(ap, &qc->tf); /* load tf registers */
ap->ops->sff_exec_command(ap, &qc->tf);
DPRINTK("Issued tag %u\n", qc->tag);
return 0;
}
static unsigned int nv_swncq_qc_issue(struct ata_queued_cmd *qc)
{
struct ata_port *ap = qc->ap;
struct nv_swncq_port_priv *pp = ap->private_data;
if (qc->tf.protocol != ATA_PROT_NCQ)
return ata_bmdma_qc_issue(qc);
DPRINTK("Enter\n");
if (!pp->qc_active)
nv_swncq_issue_atacmd(ap, qc);
else
nv_swncq_qc_to_dq(ap, qc); /* add qc to defer queue */
return 0;
}
static void nv_swncq_hotplug(struct ata_port *ap, u32 fis)
{
u32 serror;
struct ata_eh_info *ehi = &ap->link.eh_info;
ata_ehi_clear_desc(ehi);
/* AHCI needs SError cleared; otherwise, it might lock up */
sata_scr_read(&ap->link, SCR_ERROR, &serror);
sata_scr_write(&ap->link, SCR_ERROR, serror);
/* analyze @irq_stat */
if (fis & NV_SWNCQ_IRQ_ADDED)
ata_ehi_push_desc(ehi, "hot plug");
else if (fis & NV_SWNCQ_IRQ_REMOVED)
ata_ehi_push_desc(ehi, "hot unplug");
ata_ehi_hotplugged(ehi);
/* okay, let's hand over to EH */
ehi->serror |= serror;
ata_port_freeze(ap);
}
static int nv_swncq_sdbfis(struct ata_port *ap)
{
struct ata_queued_cmd *qc;
struct nv_swncq_port_priv *pp = ap->private_data;
struct ata_eh_info *ehi = &ap->link.eh_info;
u32 sactive;
u32 done_mask;
u8 host_stat;
u8 lack_dhfis = 0;
host_stat = ap->ops->bmdma_status(ap);
if (unlikely(host_stat & ATA_DMA_ERR)) {
/* error when transferring data to/from memory */
ata_ehi_clear_desc(ehi);
ata_ehi_push_desc(ehi, "BMDMA stat 0x%x", host_stat);
ehi->err_mask |= AC_ERR_HOST_BUS;
libata: prefer hardreset When both soft and hard resets are available, libata preferred softreset till now. The logic behind it was to be softer to devices; however, this doesn't really help much. Rationales for the change: * BIOS may freeze lock certain things during boot and softreset can't unlock those. This by itself is okay but during operation PHY event or other error conditions can trigger hardreset and the device may end up with different configuration. For example, after a hardreset, previously unlockable HPA can be unlocked resulting in different device size and thus revalidation failure. Similar condition can occur during or after resume. * Certain ATAPI devices require hardreset to recover after certain error conditions. On PATA, this is done by issuing the DEVICE RESET command. On SATA, COMRESET has equivalent effect. The problem is that DEVICE RESET needs its own execution protocol. For SFF controllers with bare TF access, it can be easily implemented but more advanced controllers (e.g. ahci and sata_sil24) require specialized implementations. Simply using hardreset solves the problem nicely. * COMRESET initialization sequence is the norm in SATA land and many SATA devices don't work properly if only SRST is used. For example, some PMPs behave this way and libata works around by always issuing hardreset if the host supports PMP. Like the above example, libata has developed a number of mechanisms aiming to promote softreset to hardreset if softreset is not going to work. This approach is time consuming and error prone. Also, note that, dependingon how you read the specs, it could be argued that PMP fan-out ports require COMRESET to start operation. In fact, all the PMPs on the market except one don't work properly if COMRESET is not issued to fan-out ports after PMP reset. * COMRESET is an integral part of SATA connection and any working device should be able to handle COMRESET properly. After all, it's the way to signal hardreset during reboot. This is the most used and recommended (at least by the ahci spec) method of resetting devices. So, this patch makes libata prefer hardreset over softreset by making the following changes. * Rename ATA_EH_RESET_MASK to ATA_EH_RESET and use it whereever ATA_EH_{SOFT|HARD}RESET used to be used. ATA_EH_{SOFT|HARD}RESET is now only used to tell prereset whether soft or hard reset will be issued. * Strip out now unneeded promote-to-hardreset logics from ata_eh_reset(), ata_std_prereset(), sata_pmp_std_prereset() and other places. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-01-23 23:05:14 +08:00
ehi->action |= ATA_EH_RESET;
return -EINVAL;
}
ap->ops->sff_irq_clear(ap);
__ata_bmdma_stop(ap);
sactive = readl(pp->sactive_block);
done_mask = pp->qc_active ^ sactive;
libata: always use ata_qc_complete_multiple() for NCQ command completions Currently, sata_fsl, mv and nv call ata_qc_complete() multiple times from their interrupt handlers to indicate completion of NCQ commands. This limits the visibility the libata core layer has into how commands are being executed and completed, which is necessary to support IRQ expecting in generic way. libata already has an interface to complete multiple commands at once - ata_qc_complete_multiple() which ahci and sata_sil24 already use. This patch updates the three drivers to use ata_qc_complete_multiple() too and updates comments on ata_qc_complete[_multiple]() regarding their usages with NCQ completions. This change not only provides better visibility into command execution to the core layer but also simplifies low level drivers. * sata_fsl: It already builds done_mask. Conversion is straight forward. * sata_mv: mv_process_crpb_response() no longer checks for illegal completions, it just returns whether the tag is completed or not. mv_process_crpb_entries() builds done_mask from it and passes it to ata_qc_complete_multiple() which will check for illegal completions. * sata_nv adma: Similar to sata_mv. nv_adma_check_cpb() now just returns the tag status and nv_adma_interrupt() builds done_mask from it and passes it to ata_qc_complete_multiple(). * sata_nv swncq: It already builds done_mask. Drop unnecessary illegal transition checks and call ata_qc_complete_multiple(). In the long run, it might be a good idea to make ata_qc_complete() whine if called when multiple NCQ commands are in flight. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Ashish Kalra <ashish.kalra@freescale.com> Cc: Saeed Bishara <saeed@marvell.com> Cc: Mark Lord <liml@rtr.ca> Cc: Robert Hancock <hancockr@shaw.ca> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2010-06-25 21:03:34 +08:00
pp->qc_active &= ~done_mask;
pp->dhfis_bits &= ~done_mask;
pp->dmafis_bits &= ~done_mask;
pp->sdbfis_bits |= done_mask;
ata_qc_complete_multiple(ap, ap->qc_active ^ done_mask);
if (!ap->qc_active) {
DPRINTK("over\n");
nv_swncq_pp_reinit(ap);
return 0;
}
if (pp->qc_active & pp->dhfis_bits)
return 0;
if ((pp->ncq_flags & ncq_saw_backout) ||
(pp->qc_active ^ pp->dhfis_bits))
/* if the controller can't get a device to host register FIS,
* The driver needs to reissue the new command.
*/
lack_dhfis = 1;
DPRINTK("id 0x%x QC: qc_active 0x%x,"
"SWNCQ:qc_active 0x%X defer_bits %X "
"dhfis 0x%X dmafis 0x%X last_issue_tag %x\n",
ap->print_id, ap->qc_active, pp->qc_active,
pp->defer_queue.defer_bits, pp->dhfis_bits,
pp->dmafis_bits, pp->last_issue_tag);
nv_swncq_fis_reinit(ap);
if (lack_dhfis) {
qc = ata_qc_from_tag(ap, pp->last_issue_tag);
nv_swncq_issue_atacmd(ap, qc);
return 0;
}
if (pp->defer_queue.defer_bits) {
/* send deferral queue command */
qc = nv_swncq_qc_from_dq(ap);
WARN_ON(qc == NULL);
nv_swncq_issue_atacmd(ap, qc);
}
return 0;
}
static inline u32 nv_swncq_tag(struct ata_port *ap)
{
struct nv_swncq_port_priv *pp = ap->private_data;
u32 tag;
tag = readb(pp->tag_block) >> 2;
return (tag & 0x1f);
}
static void nv_swncq_dmafis(struct ata_port *ap)
{
struct ata_queued_cmd *qc;
unsigned int rw;
u8 dmactl;
u32 tag;
struct nv_swncq_port_priv *pp = ap->private_data;
__ata_bmdma_stop(ap);
tag = nv_swncq_tag(ap);
DPRINTK("dma setup tag 0x%x\n", tag);
qc = ata_qc_from_tag(ap, tag);
if (unlikely(!qc))
return;
rw = qc->tf.flags & ATA_TFLAG_WRITE;
/* load PRD table addr. */
iowrite32(pp->prd_dma + ATA_PRD_TBL_SZ * qc->tag,
ap->ioaddr.bmdma_addr + ATA_DMA_TABLE_OFS);
/* specify data direction, triple-check start bit is clear */
dmactl = ioread8(ap->ioaddr.bmdma_addr + ATA_DMA_CMD);
dmactl &= ~ATA_DMA_WR;
if (!rw)
dmactl |= ATA_DMA_WR;
iowrite8(dmactl | ATA_DMA_START, ap->ioaddr.bmdma_addr + ATA_DMA_CMD);
}
static void nv_swncq_host_interrupt(struct ata_port *ap, u16 fis)
{
struct nv_swncq_port_priv *pp = ap->private_data;
struct ata_queued_cmd *qc;
struct ata_eh_info *ehi = &ap->link.eh_info;
u32 serror;
u8 ata_stat;
ata_stat = ap->ops->sff_check_status(ap);
nv_swncq_irq_clear(ap, fis);
if (!fis)
return;
if (ap->pflags & ATA_PFLAG_FROZEN)
return;
if (fis & NV_SWNCQ_IRQ_HOTPLUG) {
nv_swncq_hotplug(ap, fis);
return;
}
if (!pp->qc_active)
return;
if (ap->ops->scr_read(&ap->link, SCR_ERROR, &serror))
return;
ap->ops->scr_write(&ap->link, SCR_ERROR, serror);
if (ata_stat & ATA_ERR) {
ata_ehi_clear_desc(ehi);
ata_ehi_push_desc(ehi, "Ata error. fis:0x%X", fis);
ehi->err_mask |= AC_ERR_DEV;
ehi->serror |= serror;
libata: prefer hardreset When both soft and hard resets are available, libata preferred softreset till now. The logic behind it was to be softer to devices; however, this doesn't really help much. Rationales for the change: * BIOS may freeze lock certain things during boot and softreset can't unlock those. This by itself is okay but during operation PHY event or other error conditions can trigger hardreset and the device may end up with different configuration. For example, after a hardreset, previously unlockable HPA can be unlocked resulting in different device size and thus revalidation failure. Similar condition can occur during or after resume. * Certain ATAPI devices require hardreset to recover after certain error conditions. On PATA, this is done by issuing the DEVICE RESET command. On SATA, COMRESET has equivalent effect. The problem is that DEVICE RESET needs its own execution protocol. For SFF controllers with bare TF access, it can be easily implemented but more advanced controllers (e.g. ahci and sata_sil24) require specialized implementations. Simply using hardreset solves the problem nicely. * COMRESET initialization sequence is the norm in SATA land and many SATA devices don't work properly if only SRST is used. For example, some PMPs behave this way and libata works around by always issuing hardreset if the host supports PMP. Like the above example, libata has developed a number of mechanisms aiming to promote softreset to hardreset if softreset is not going to work. This approach is time consuming and error prone. Also, note that, dependingon how you read the specs, it could be argued that PMP fan-out ports require COMRESET to start operation. In fact, all the PMPs on the market except one don't work properly if COMRESET is not issued to fan-out ports after PMP reset. * COMRESET is an integral part of SATA connection and any working device should be able to handle COMRESET properly. After all, it's the way to signal hardreset during reboot. This is the most used and recommended (at least by the ahci spec) method of resetting devices. So, this patch makes libata prefer hardreset over softreset by making the following changes. * Rename ATA_EH_RESET_MASK to ATA_EH_RESET and use it whereever ATA_EH_{SOFT|HARD}RESET used to be used. ATA_EH_{SOFT|HARD}RESET is now only used to tell prereset whether soft or hard reset will be issued. * Strip out now unneeded promote-to-hardreset logics from ata_eh_reset(), ata_std_prereset(), sata_pmp_std_prereset() and other places. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-01-23 23:05:14 +08:00
ehi->action |= ATA_EH_RESET;
ata_port_freeze(ap);
return;
}
if (fis & NV_SWNCQ_IRQ_BACKOUT) {
/* If the IRQ is backout, driver must issue
* the new command again some time later.
*/
pp->ncq_flags |= ncq_saw_backout;
}
if (fis & NV_SWNCQ_IRQ_SDBFIS) {
pp->ncq_flags |= ncq_saw_sdb;
DPRINTK("id 0x%x SWNCQ: qc_active 0x%X "
"dhfis 0x%X dmafis 0x%X sactive 0x%X\n",
ap->print_id, pp->qc_active, pp->dhfis_bits,
pp->dmafis_bits, readl(pp->sactive_block));
if (nv_swncq_sdbfis(ap) < 0)
goto irq_error;
}
if (fis & NV_SWNCQ_IRQ_DHREGFIS) {
/* The interrupt indicates the new command
* was transmitted correctly to the drive.
*/
pp->dhfis_bits |= (0x1 << pp->last_issue_tag);
pp->ncq_flags |= ncq_saw_d2h;
if (pp->ncq_flags & (ncq_saw_sdb | ncq_saw_backout)) {
ata_ehi_push_desc(ehi, "illegal fis transaction");
ehi->err_mask |= AC_ERR_HSM;
libata: prefer hardreset When both soft and hard resets are available, libata preferred softreset till now. The logic behind it was to be softer to devices; however, this doesn't really help much. Rationales for the change: * BIOS may freeze lock certain things during boot and softreset can't unlock those. This by itself is okay but during operation PHY event or other error conditions can trigger hardreset and the device may end up with different configuration. For example, after a hardreset, previously unlockable HPA can be unlocked resulting in different device size and thus revalidation failure. Similar condition can occur during or after resume. * Certain ATAPI devices require hardreset to recover after certain error conditions. On PATA, this is done by issuing the DEVICE RESET command. On SATA, COMRESET has equivalent effect. The problem is that DEVICE RESET needs its own execution protocol. For SFF controllers with bare TF access, it can be easily implemented but more advanced controllers (e.g. ahci and sata_sil24) require specialized implementations. Simply using hardreset solves the problem nicely. * COMRESET initialization sequence is the norm in SATA land and many SATA devices don't work properly if only SRST is used. For example, some PMPs behave this way and libata works around by always issuing hardreset if the host supports PMP. Like the above example, libata has developed a number of mechanisms aiming to promote softreset to hardreset if softreset is not going to work. This approach is time consuming and error prone. Also, note that, dependingon how you read the specs, it could be argued that PMP fan-out ports require COMRESET to start operation. In fact, all the PMPs on the market except one don't work properly if COMRESET is not issued to fan-out ports after PMP reset. * COMRESET is an integral part of SATA connection and any working device should be able to handle COMRESET properly. After all, it's the way to signal hardreset during reboot. This is the most used and recommended (at least by the ahci spec) method of resetting devices. So, this patch makes libata prefer hardreset over softreset by making the following changes. * Rename ATA_EH_RESET_MASK to ATA_EH_RESET and use it whereever ATA_EH_{SOFT|HARD}RESET used to be used. ATA_EH_{SOFT|HARD}RESET is now only used to tell prereset whether soft or hard reset will be issued. * Strip out now unneeded promote-to-hardreset logics from ata_eh_reset(), ata_std_prereset(), sata_pmp_std_prereset() and other places. Signed-off-by: Tejun Heo <htejun@gmail.com>
2008-01-23 23:05:14 +08:00
ehi->action |= ATA_EH_RESET;
goto irq_error;
}
if (!(fis & NV_SWNCQ_IRQ_DMASETUP) &&
!(pp->ncq_flags & ncq_saw_dmas)) {
ata_stat = ap->ops->sff_check_status(ap);
if (ata_stat & ATA_BUSY)
goto irq_exit;
if (pp->defer_queue.defer_bits) {
DPRINTK("send next command\n");
qc = nv_swncq_qc_from_dq(ap);
nv_swncq_issue_atacmd(ap, qc);
}
}
}
if (fis & NV_SWNCQ_IRQ_DMASETUP) {
/* program the dma controller with appropriate PRD buffers
* and start the DMA transfer for requested command.
*/
pp->dmafis_bits |= (0x1 << nv_swncq_tag(ap));
pp->ncq_flags |= ncq_saw_dmas;
nv_swncq_dmafis(ap);
}
irq_exit:
return;
irq_error:
ata_ehi_push_desc(ehi, "fis:0x%x", fis);
ata_port_freeze(ap);
return;
}
static irqreturn_t nv_swncq_interrupt(int irq, void *dev_instance)
{
struct ata_host *host = dev_instance;
unsigned int i;
unsigned int handled = 0;
unsigned long flags;
u32 irq_stat;
spin_lock_irqsave(&host->lock, flags);
irq_stat = readl(host->iomap[NV_MMIO_BAR] + NV_INT_STATUS_MCP55);
for (i = 0; i < host->n_ports; i++) {
struct ata_port *ap = host->ports[i];
if (ap->link.sactive) {
nv_swncq_host_interrupt(ap, (u16)irq_stat);
handled = 1;
} else {
if (irq_stat) /* reserve Hotplug */
nv_swncq_irq_clear(ap, 0xfff0);
handled += nv_host_intr(ap, (u8)irq_stat);
}
irq_stat >>= NV_INT_PORT_SHIFT_MCP55;
}
spin_unlock_irqrestore(&host->lock, flags);
return IRQ_RETVAL(handled);
}
static int nv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
{
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 18:43:58 +08:00
const struct ata_port_info *ppi[] = { NULL, NULL };
struct nv_pi_priv *ipriv;
struct ata_host *host;
struct nv_host_priv *hpriv;
int rc;
u32 bar;
void __iomem *base;
unsigned long type = ent->driver_data;
// Make sure this is a SATA controller by counting the number of bars
// (NVIDIA SATA controllers will always have six bars). Otherwise,
// it's an IDE controller and we ignore it.
for (bar = 0; bar < 6; bar++)
if (pci_resource_start(pdev, bar) == 0)
return -ENODEV;
ata_print_version_once(&pdev->dev, DRV_VERSION);
rc = pcim_enable_device(pdev);
if (rc)
return rc;
/* determine type and allocate host */
if (type == CK804 && adma_enabled) {
dev_notice(&pdev->dev, "Using ADMA mode\n");
type = ADMA;
} else if (type == MCP5x && swncq_enabled) {
dev_notice(&pdev->dev, "Using SWNCQ mode\n");
type = SWNCQ;
}
libata: clean up SFF init mess The intention of using port_mask in SFF init helpers was to eventually support exoctic configurations such as combination of legacy and native port on the same controller. This never became actually necessary and the related code always has been subtly broken one way or the other. Now that new init model is in place, there is no reason to make common helpers capable of handling all corner cases. Exotic cases can simply dealt within LLDs as necessary. This patch removes port_mask handling in SFF init helpers. SFF init helpers don't take n_ports argument and interpret it into port_mask anymore. All information is carried via port_info. n_ports argument is dropped and always two ports are allocated. LLD can tell SFF to skip certain port by marking it dummy. Note that SFF code has been treating unuvailable ports this way for a long time until recent breakage fix from Linus and is consistent with how other drivers handle with unavailable ports. This fixes 1-port legacy host handling still broken after the recent native mode fix and simplifies SFF init logic. The following changes are made... * ata_pci_init_native_host() and ata_init_legacy_host() both now try to initialized whatever they can and mark failed ports dummy. They return 0 if any port is successfully initialized. * ata_pci_prepare_native_host() and ata_pci_init_one() now doesn't take n_ports argument. All info should be specified via port_info array. Always two ports are allocated. * ata_pci_init_bmdma() exported to be used by LLDs in exotic cases. * port_info handling in all LLDs are standardized - all port_info arrays are const stack variable named ppi. Unless the second port is different from the first, its port_info is specified as NULL (tells libata that it's identical to the last non-NULL port_info). * pata_hpt37x/hpt3x2n: don't modify static variable directly. Make an on-stack copy instead as ata_piix does. * pata_uli: It has 4 ports instead of 2. Don't use ata_pci_prepare_native_host(). Allocate the host explicitly and use init helpers. It's simple enough. Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-04 18:43:58 +08:00
ppi[0] = &nv_port_info[type];
ipriv = ppi[0]->private_data;
rc = ata_pci_bmdma_prepare_host(pdev, ppi, &host);
if (rc)
return rc;
hpriv = devm_kzalloc(&pdev->dev, sizeof(*hpriv), GFP_KERNEL);
if (!hpriv)
return -ENOMEM;
hpriv->type = type;
host->private_data = hpriv;
/* request and iomap NV_MMIO_BAR */
rc = pcim_iomap_regions(pdev, 1 << NV_MMIO_BAR, DRV_NAME);
if (rc)
return rc;
/* configure SCR access */
base = host->iomap[NV_MMIO_BAR];
host->ports[0]->ioaddr.scr_addr = base + NV_PORT0_SCR_REG_OFFSET;
host->ports[1]->ioaddr.scr_addr = base + NV_PORT1_SCR_REG_OFFSET;
/* enable SATA space for CK804 */
if (type >= CK804) {
u8 regval;
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, &regval);
regval |= NV_MCP_SATA_CFG_20_SATA_SPACE_EN;
pci_write_config_byte(pdev, NV_MCP_SATA_CFG_20, regval);
}
/* init ADMA */
if (type == ADMA) {
rc = nv_adma_host_init(host);
if (rc)
return rc;
} else if (type == SWNCQ)
nv_swncq_host_init(host);
if (msi_enabled) {
dev_notice(&pdev->dev, "Using MSI\n");
pci_enable_msi(pdev);
}
pci_set_master(pdev);
return ata_pci_sff_activate_host(host, ipriv->irq_handler, ipriv->sht);
}
#ifdef CONFIG_PM_SLEEP
static int nv_pci_device_resume(struct pci_dev *pdev)
{
struct ata_host *host = pci_get_drvdata(pdev);
struct nv_host_priv *hpriv = host->private_data;
int rc;
rc = ata_pci_device_do_resume(pdev);
if (rc)
return rc;
if (pdev->dev.power.power_state.event == PM_EVENT_SUSPEND) {
if (hpriv->type >= CK804) {
u8 regval;
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, &regval);
regval |= NV_MCP_SATA_CFG_20_SATA_SPACE_EN;
pci_write_config_byte(pdev, NV_MCP_SATA_CFG_20, regval);
}
if (hpriv->type == ADMA) {
u32 tmp32;
struct nv_adma_port_priv *pp;
/* enable/disable ADMA on the ports appropriately */
pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
pp = host->ports[0]->private_data;
if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE)
tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT0_EN |
NV_MCP_SATA_CFG_20_PORT0_PWB_EN);
else
tmp32 |= (NV_MCP_SATA_CFG_20_PORT0_EN |
NV_MCP_SATA_CFG_20_PORT0_PWB_EN);
pp = host->ports[1]->private_data;
if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE)
tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT1_EN |
NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
else
tmp32 |= (NV_MCP_SATA_CFG_20_PORT1_EN |
NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
}
}
ata_host_resume(host);
return 0;
}
#endif
static void nv_ck804_host_stop(struct ata_host *host)
{
struct pci_dev *pdev = to_pci_dev(host->dev);
u8 regval;
/* disable SATA space for CK804 */
pci_read_config_byte(pdev, NV_MCP_SATA_CFG_20, &regval);
regval &= ~NV_MCP_SATA_CFG_20_SATA_SPACE_EN;
pci_write_config_byte(pdev, NV_MCP_SATA_CFG_20, regval);
}
static void nv_adma_host_stop(struct ata_host *host)
{
struct pci_dev *pdev = to_pci_dev(host->dev);
u32 tmp32;
/* disable ADMA on the ports */
pci_read_config_dword(pdev, NV_MCP_SATA_CFG_20, &tmp32);
tmp32 &= ~(NV_MCP_SATA_CFG_20_PORT0_EN |
NV_MCP_SATA_CFG_20_PORT0_PWB_EN |
NV_MCP_SATA_CFG_20_PORT1_EN |
NV_MCP_SATA_CFG_20_PORT1_PWB_EN);
pci_write_config_dword(pdev, NV_MCP_SATA_CFG_20, tmp32);
nv_ck804_host_stop(host);
}
module_pci_driver(nv_pci_driver);
module_param_named(adma, adma_enabled, bool, 0444);
MODULE_PARM_DESC(adma, "Enable use of ADMA (Default: false)");
module_param_named(swncq, swncq_enabled, bool, 0444);
MODULE_PARM_DESC(swncq, "Enable use of SWNCQ (Default: true)");
module_param_named(msi, msi_enabled, bool, 0444);
MODULE_PARM_DESC(msi, "Enable use of MSI (Default: false)");