Currently the Ice Lake and Tiger Lake NHI (host controller) LC (link
controller) mailbox command processing checks for the completion of
command every 100 msecs. These controllers are found to complete this in
the order of 1 ms or so. Since this delay is in suspend path, surplus
delay is effectively affecting runtime PM suspend flows.
Optimize this so that we do the wait for 1 ms after reading the mailbox
register. This should make Ice Lake and Tiger Lake runtime suspend take
less time to complete.
Reported-by: Dana Alkattan <dana.alkattan@intel.com>
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Currently the "Force Power" logic uses 10 retries, each with a delay of
250 ms. Thunderbolt controllers in Ice Lake and Tiger Lake platforms are
found to complete this in the order of 3 ms or so. Since this delay
is in resume path, surplus delay is effectively affecting runtime PM
resume flows.
Decrease the granularity of the delay to 3 ms and increase the number of
retries so we wait maximum of ~1 s which is the recommended timeout.
This should make runtime resume a bit faster.
Reported-by: Dana Alkattan <dana.alkattan@intel.com>
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Doesn't really matter for an individual driver, but it may
get coppied to lots more. I consider it's a little tidy up.
Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
This includes two fixes, one that fixes a regression around reboot and
other that uses a correct link rate when USB3 bandwidth is reclaimed
when the link is not up.
Both have been in linux-next with no reported issues.
-----BEGIN PGP SIGNATURE-----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=vxwa
-----END PGP SIGNATURE-----
Merge tag 'thunderbolt-for-v5.9-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt into usb-linus
Mika writes:
thunderbolt: Fixes for v5.9-rc4
This includes two fixes, one that fixes a regression around reboot and
other that uses a correct link rate when USB3 bandwidth is reclaimed
when the link is not up.
Both have been in linux-next with no reported issues.
* tag 'thunderbolt-for-v5.9-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt:
thunderbolt: Use maximum USB3 link rate when reclaiming if link is not up
thunderbolt: Disable ports that are not implemented
If the USB3 link is not up the actual link rate is 0 so when reclaiming
bandwidth we should look at maximum supported link rate instead.
Cc: stable@vger.kernel.org
Fixes: 0bd680cd90 ("thunderbolt: Add USB3 bandwidth management")
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Commit 4caf2511ec ("thunderbolt: Add trivial .shutdown") exposes a bug
in the Thunderbolt driver, that frees an unallocated id, resulting in the
following spinlock bad magic bug.
[ 20.633803] BUG: spinlock bad magic on CPU#4, halt/3313
[ 20.640030] lock: 0xffff92e6ad5c97e0, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
[ 20.672139] Call Trace:
[ 20.675032] dump_stack+0x97/0xdb
[ 20.678950] ? spin_bug+0xa5/0xb0
[ 20.682865] do_raw_spin_lock+0x68/0x98
[ 20.687397] _raw_spin_lock_irqsave+0x3f/0x5d
[ 20.692535] ida_destroy+0x4f/0x124
[ 20.696657] tb_switch_release+0x6d/0xfd
[ 20.701295] device_release+0x2c/0x7d
[ 20.705622] kobject_put+0x8e/0xac
[ 20.709637] tb_stop+0x55/0x66
[ 20.713243] tb_domain_remove+0x36/0x62
[ 20.717774] nhi_remove+0x4d/0x58
Fix the issue by disabling ports that are enabled as per the EEPROM, but
not implemented. While at it, update the kernel doc for the disabled
field, to reflect this.
Cc: stable@vger.kernel.org
Fixes: 4caf2511ec ("thunderbolt: Add trivial .shutdown")
Reported-by: Srikanth Nandamuri <srikanth.nandamuri@intel.com>
Signed-off-by: Nikunj A. Dadhania <nikunj.dadhania@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Some external devices can support completing thunderbolt authentication
when they are unplugged. For this to work though, the link controller must
remain operational.
The only device known to support this right now is the Dell WD19TB, so add
a quirk for this.
Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
This allows userspace to have a shorter period of time that the device
is unusable and to call it at a more convenient time.
For example flushing the image may happen while the user is using the
machine and authenticating/rebooting may happen while logging out.
Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The 32 bit int value 512 is being left shifted and then used in a context
that expects the expression to be a larger unsigned long. There may be
a potential integer overflow, so make 512 a UL before shift to avoid
any such issues.
Addresses-Coverity: ("Uninintentional integer overflow")
Fixes: 3b1d8d577c ("thunderbolt: Implement USB3 bandwidth negotiation routines")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB4 spec specifies standard access to retimers (both on-board and
cable) through USB4 port sideband access. This makes it possible to
upgrade their firmware in the same way than we already do with the
routers.
This enumerates on-board retimers under each USB4 port when the link
comes up and adds them to the bus under the router the retimer belongs
to. Retimers are exposed in sysfs with name like <device>:<port>.<index>
where device is the router the retimer belongs to, port is the USB4 port
the retimer is connected to and index is the retimer index under that
port (starting from 1). This applies to the upstream USB4 port as well
so if there is on-board retimer between the port and the router it is
also added accordingly.
At this time we do not add cable retimers but there is no techincal
restriction to do so in the future if needed. It is not clear whether it
makes sense to upgrade their firmwares and at least Thunderbolt 3 cables
it has not been done outside of lab environments.
The sysfs interface is made to follow the router NVM upgrade to make it
easy to extend the existing userspace (fwupd) to handle these as well.
Signed-off-by: Kranthi Kuntala <kranthi.kuntala@intel.com>
Co-developed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB4 spec specifies standard set of sideband operations that are send
over the low speed link to access either retimers on the link or the
link parter (the other router). The USB4 retimer spec extends these and
adds operations for retimer NVM upgrade.
This implements the retimer access and NVM upgrade USB4 port sideband
operations which we need for retimer support in the patch that follows.
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Co-developed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Especially when accessing retimers over USB4 sideband operations the
possibility to get read errors seems to be higher so make the
usb4_do_read_data() retry a couple of times if it sees any other error
than -ENODEV (device is gone). We can only do this for read side because
it carries the offset as part of metadata (as opposed to writes).
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Currently these functions operate on struct tb_switch but we are going
to need the same functionality with retimers as well so make the two
functions work with an arbitrary object that gets passed as parameter to
the callbacks.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
We are going to reuse some of this functionality to implement retimer
NVM upgrade so move common NVM functionality into its own file. We also
rename the structure from tb_switch_nvm to tb_nvm to make it clear that
it is not just for switches.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
With USB4 Intel is also using its USB-IF ID (0x8087) with the new
devices. The NVM format is the same. Add this to the driver so NVM
upgrade is possible with these devices as well.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
We can test some parts of tunneling, like path allocation without access
to test hardware so add KUnit tests for PCIe, DP and USB3 tunneling.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB3 supports both isochronous and non-isochronous traffic. The former
requires guaranteed bandwidth and can take up to 90% of the total
bandwidth. With USB4 USB3 is tunneled over USB4 fabric which means that
we need to make sure there is enough bandwidth allocated for the USB3
tunnels in addition to DisplayPort tunnels.
Whereas DisplayPort bandwidth management is static and done before the
DP tunnel is established, the USB3 bandwidth management is dynamic and
allows increasing and decreasing the allocated bandwidth according to
what is currently consumed. This is done through host router USB3
downstream adapter registers.
This adds USB3 bandwidth management to the software connection manager
so that we always try to allocate maximum bandwidth for DP tunnels and
what is left is allocated for USB3.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
We need to call this from tb.c when we improve the bandwidth management
to take USB3 into account.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Each host router USB3 downstream adapter has a set of registers that are
used to negotiate bandwidth between the connection manager and the
internal xHCI controller. These registers allow dynamic bandwidth
management for USB3 isochronous traffic based on what is actually
consumed vs. allocated at any given time.
Implement these USB3 bandwidth negotiation routines to allow the
software connection manager take advantage of these.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Sometimes it takes longer for DPRX to be set so increase the timeout to
cope with this.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Whereas DisplayPort bandwidth is consumed only in one direction (from DP
IN adapter to DP OUT adapter), USB3 adds separate bandwidth for both
upstream and downstream directions.
For this reason extend the tunnel consumed bandwidth routines to support
both directions and implement this for DP.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Just for symmetry with the usb4_switch_map_usb3_down() make this one
also return ports that are enabled.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
We need to call this on enabled ports in order to find the mapping from
host router USB4 port to a USB 3.x downstream adapter, so make the
function return enabled ports as well.
While there fix parameter alignment in tb_find_usb3_down().
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB3 tunneling is possible only over USB4 link so don't create USB3
tunnels if that's not the case.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB4 spec allows DP tunneling from any router that has DP IN adapter,
not just from host router. The driver currently only added the DP IN
resources for the host router because Thunderbolt 1, 2 and 3 devices do
not have DP IN adapters. However, USB4 allows device routers to have DP
IN adapter as well so update the driver to add DP IN resources for each
device that has one. One example would be an eGPU enclosure where the
eGPU output is forwarded to DP IN port and then tunneled over the USB4
fabric.
Only limitation we add now is that the DP IN and DP OUT that gets paired
for tunnel creation should both be under the same topology starting from
host router downstream port. In other words we do not create DP tunnels
across host router at this time even though that is possible as well but
it complicates the bandwidth management and there is no real use-case
for this anyway.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
This adds KUnit tests for path walking which is only dependent on
software structures, so no hardware is needed to run these.
We make these available only when both KUnit and the driver itself are
built into the kernel image. The reason for this is that KUnit adds its
own module_init() call in kunit_test_suite() which generates linker
error because the driver does the same in nhi.c. This should be fine for
now because these tests are only meant to run by developers anyway.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Currently we have only supported paths that follow daisy-chain topology
but USB4 also allows to build trees of devices. For this reason increase
maximum path length we use for discovery to be from the lowest level to
the host router and back to the same level.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
If the path is not complete when we do discovery the number of hops may
be less than with the full path. As an example when this can happen is
that user unloads the driver, disconnects the topology, and loads the
driver back. If there is PCIe or USB3 tunnel involved this may happen.
Take this into account in tb_pcie_init_path() and tb_usb3_init_path()
and prevent potential access over array limits.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Add sanity check that given src and dst ports are reachable through path
walk before allocating a path. If they are not then bail out early.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
With USB4, topologies are not limited to daisy-chains anymore so when
calculating how many hops are between two ports we need to walk the
whole path instead.
Add helper function tb_for_each_port_on_path() that can be used to walk
over each port on a path and make tb_path_alloc() to use it.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB4 makes it possible to have tree topology of devices connected in the
same way than USB3. This was actually possible in Thunderbolt 1, 2 and 3
as well but all the available devices only had two ports which allows
building only daisy-chains of devices.
With USB4 it is possible for example that there is DP IN adapter as part
of eGPU device router and that should be tunneled over the tree topology
to a DP OUT adapter. This updates the tb_next_port_on_path() to support
such topologies.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The USB3 discovery used wrong indices when tunnel is discovered. It
should use TB_USB3_PATH_DOWN for path that flows downstream and
TB_USB3_PATH_UP when it flows upstream. This should not affect the
functionality but better to fix it.
Fixes: e6f8185857 ("thunderbolt: Add support for USB 3.x tunnels")
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: stable@vger.kernel.org # v5.6+
The end-to-end (E2E) workaround is needed for Falcon Ridge (TBT 2)
controller when E2E is enabled for both ends of the host-to-host
connection. However, we never supported full E2E in the first place so
this code is not necessary at the moment. Further this allows us to use
all available rings for data except ring 0 which is reserved for the
control path.
The complete E2E flow control is explained in the USB4 spec so we may
add it back later if needed but at least the networking driver seems to
work fine without, and the higher level stack, like TCP will retransmit
lost packets anyway.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
NHI (The host interface adapter) is allowed to use HopIDs 1-7 as well so
relax the restriction in tb_port_alloc_hopid() to support this.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
While Intel hardware typically has hop_count (Total Paths in the spec)
12 the USB4 spec allows this to be anything between 1 and 21 so no need
to warn about this. Simply log number of paths at debug level.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
On a systems where the Thunderbolt controller is present all the time
the kernel nodename may not yet set by the userspace when the driver is
loaded. This means when another host is connected it may see the default
"(none)" hostname instead of the system real hostname.
For this reason build the initial XDomain property block only upon first
connect. This should make sure the userspace has had chance to set it up.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Fix the spelling of "specification", and add a missing "the" article.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Here are the large set of USB and PHY driver updates for 5.8-rc1.
Nothing huge, just lots of little things:
- USB gadget fixes and additions all over the place
- new PHY drivers
- PHY driver fixes and updates
- XHCI driver updates
- musb driver updates
- more USB-serial driver ids added
- various USB quirks added
- thunderbolt minor updates and fixes
- typec updates and additions
Full details are in the shortlog.
All of these have been in linux-next for a while with no reported
issues.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
iG0EABECAC0WIQT0tgzFv3jCIUoxPcsxR9QN2y37KQUCXtzqVA8cZ3JlZ0Brcm9h
aC5jb20ACgkQMUfUDdst+ynftwCfeanyI6TR5AdfJVZN50B6/ySvVwcAn07i9VRX
tnt2kz0UqReYpLt0wyJ7
=YP7o
-----END PGP SIGNATURE-----
Merge tag 'usb-5.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB/PHY driver updates from Greg KH:
"Here are the large set of USB and PHY driver updates for 5.8-rc1.
Nothing huge, just lots of little things:
- USB gadget fixes and additions all over the place
- new PHY drivers
- PHY driver fixes and updates
- XHCI driver updates
- musb driver updates
- more USB-serial driver ids added
- various USB quirks added
- thunderbolt minor updates and fixes
- typec updates and additions
All of these have been in linux-next for a while with no reported
issues"
* tag 'usb-5.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (245 commits)
usb: dwc3: meson-g12a: fix USB2 PHY initialization on G12A and A1 SoCs
usb: dwc3: meson-g12a: fix error path when fetching the reset line fails
Revert "dt-bindings: usb: qcom,dwc3: Convert USB DWC3 bindings"
Revert "dt-bindings: usb: qcom,dwc3: Add compatible for SC7180"
Revert "dt-bindings: usb: qcom,dwc3: Introduce interconnect properties for Qualcomm DWC3 driver"
USB: serial: ch341: fix lockup of devices with limited prescaler
USB: serial: ch341: add basis for quirk detection
CDC-ACM: heed quirk also in error handling
USB: serial: option: add Telit LE910C1-EUX compositions
usb: musb: Fix runtime PM imbalance on error
usb: musb: jz4740: Prevent lockup when CONFIG_SMP is set
usb: musb: mediatek: add reset FADDR to zero in reset interrupt handle
usb: musb: use true for 'use_dma'
usb: musb: start session in resume for host port
usb: musb: return -ESHUTDOWN in urb when three-strikes error happened
USB: serial: qcserial: add DW5816e QDL support
thunderbolt: Add trivial .shutdown
usb: dwc3: keystone: Turn on USB3 PHY before controller
dt-bindings: usb: ti,keystone-dwc3.yaml: Add USB3.0 PHY property
dt-bindings: usb: convert keystone-usb.txt to YAML
...
This adds support for Intel Tiger Lake Thunderbolt controller using
firmware based connection manager. In addition the driver can now be
built on non-x86 architectures as well. Then there are a couple of
commits that make the driver work across kexec, replace a zero length
array with flexible one, and revert one change that is not needed
anymore because of NVMem subsystem improvements.
-----BEGIN PGP SIGNATURE-----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=qZRi
-----END PGP SIGNATURE-----
Merge tag 'thunderbolt-for-v5.8' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt into usb-next
Mika writes:
thunderbolt: Changes for v5.8 merge window
This adds support for Intel Tiger Lake Thunderbolt controller using
firmware based connection manager. In addition the driver can now be
built on non-x86 architectures as well. Then there are a couple of
commits that make the driver work across kexec, replace a zero length
array with flexible one, and revert one change that is not needed
anymore because of NVMem subsystem improvements.
* tag 'thunderbolt-for-v5.8' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt:
thunderbolt: Add trivial .shutdown
thunderbolt: Update Kconfig to allow building on other architectures.
thunderbolt: Replace zero-length array with flexible-array
thunderbolt: Add support for Intel Tiger Lake
Revert "thunderbolt: Prevent crash if non-active NVMem file is read"
On my machine, a kexec with this driver loaded in the old kernel causes
a very long delay on boot in the kexec'ed kernel, most likely due to
unclean shutdown prior to that.
Unloading thunderbolt driver prior to kexec allows kexec to work as fast
as regular kernel boot, as well as adding this .shutdown pointer.
Shutting a device prior to the shutdown completely is always a good idea
IMHO to help with kexec, and this one-liner patch implements it.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Previously we used pcie_find_root_port() to find a Root Port from a PCIe
device and pci_find_pcie_root_port() to find a Root Port from a
Conventional PCI device.
Unify the two functions and use pcie_find_root_port() to find a Root Port
from either a Conventional PCI device or a PCIe device. Then there is no
need to distinguish the type of the device.
Link: https://lore.kernel.org/r/1589019568-5216-1-git-send-email-yangyicong@hisilicon.com
Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Kalle Valo <kvalo@codeaurora.org> # wireless
Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com> # thunderbolt
Thunderbolt 3 and USB4 shouldn't be x86 only.
Tested on a SolidRun HoneyComb (ARM Cortex-A72) with a
Gigabyte Titan Ridge Thunderbolt 3 PCIe card (JHL7540).
Signed-off-by: David Manouchehri <david.manouchehri@riseup.net>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The function misses checking return value of tb_sw_read() before it
accesses the value that was read. Fix this by checking the return value
first.
Fixes: b04079837b ("thunderbolt: Add initial support for USB4")
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <yehezkelshb@gmail.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tiger Lake integrated Thunderbolt/USB4 controller is quite close to
Intel Ice Lake. By default it is still using firmware based connection
manager so we can use most of the Ice Lake flows in Tiger Lake as well.
We check if the firmware connection manager is running and in that case
use it, otherwise use the software based connection manager.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: Yehezkel Bernat <yehezkelshb@gmail.com>
This reverts commit 03cd45d2e2.
Commit 664f054938 ("nvmem: core: use is_bin_visible for permissions")
incidentally adds support for write-only nvmem. Hence, this workaround
is no longer required, so drop it.
Signed-off-by: Nicholas Johnson <nicholas.johnson-opensource@outlook.com.au>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
- A couple of commits that make the driver to use flexible-array member
instead of zero-length array extension. This allows compiler to issue a
warning if the flexible array is not the last member of a structure.
- Use scnprintf() instead of snprintf() to avoid potential buffer
overflow.
-----BEGIN PGP SIGNATURE-----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=/p8i
-----END PGP SIGNATURE-----
Merge tag 'thunderbolt-for-v5.7' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt into usb-next
Mika writes:
thunderbolt: Changes for v5.7 merge window
- A couple of commits that make the driver to use flexible-array member
instead of zero-length array extension. This allows compiler to issue a
warning if the flexible array is not the last member of a structure.
- Use scnprintf() instead of snprintf() to avoid potential buffer
overflow.
* tag 'thunderbolt-for-v5.7' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt:
thunderbolt: Use scnprintf() for avoiding potential buffer overflow
thunderbolt: icm: Replace zero-length array with flexible-array member
thunderbolt: eeprom: Replace zero-length array with flexible-array member
Since snprintf() returns the would-be-output size instead of the
actual output size, the succeeding calls may go beyond the given
buffer limit. Fix it by replacing with scnprintf().
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
This function is type bool, and it's supposed to return true on success.
Unfortunately, this path takes negative error codes and casts them to
bool (true) so it's treated as success instead of failure.
Fixes: 91c0c12080 ("thunderbolt: Add support for lane bonding")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertenly introduced[3] to the codebase from now on.
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertenly introduced[3] to the codebase from now on.
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The driver does not populate .reg_read callback for the non-active NVMem
because the file is supposed to be write-only. However, it turns out
NVMem subsystem does not yet support this and expects that the .reg_read
callback is provided. If user reads the binary attribute it triggers
NULL pointer dereference like this one:
BUG: kernel NULL pointer dereference, address: 0000000000000000
...
Call Trace:
bin_attr_nvmem_read+0x64/0x80
kernfs_fop_read+0xa7/0x180
vfs_read+0xbd/0x170
ksys_read+0x5a/0xd0
do_syscall_64+0x43/0x150
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Fix this in the driver by providing .reg_read callback that always
returns an error.
Reported-by: Nicholas Johnson <nicholas.johnson-opensource@outlook.com.au>
Fixes: e6b245ccd5 ("thunderbolt: Add support for host and device NVM firmware upgrade")
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200213095604.1074-1-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the case where the call tb_switch_exceeds_max_depth is true
the error reurn path leaks memory in sw. Fix this by setting
the return error code to -EADDRNOTAVAIL and returning via the
error exit path err_free_sw_ports to free sw. sw has been kzalloc'd
so the free of the NULL sw->ports is fine.
Addresses-Coverity: ("Resource leak")
Fixes: b04079837b ("thunderbolt: Add initial support for USB4")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191220220526.11307-1-colin.king@canonical.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code tried to check whether xhci variable has ROUTER_CS_6_HCI bit
set but since xhci type is bool and it already holds true or false based
on that very bit, fix the check to use the variable directly.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Fixes: b04079837b ("thunderbolt: Add initial support for USB4")
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20200108125317.36444-2-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
USB4 added a capability to tunnel USB 3.x protocol over the USB4
fabric. USB4 device routers may include integrated SuperSpeed HUB or a
function or both. USB tunneling follows PCIe so that the tunnel is
created between the parent and the child router from USB3 downstream
adapter port to USB3 upstream adapter port over a single USB4 link.
This adds support for USB 3.x tunneling and also capability to discover
existing USB 3.x tunnels (for example created by connection manager in
boot firmware).
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Co-developed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-9-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Time Management Unit (TMU) is included in each USB4 router. It is used
to synchronize time across the USB4 fabric. By default when USB4 router
is plugged to the domain, its TMU is turned off. This differs from
Thunderbolt (1, 2 and 3) devices whose TMU is by default configured to
bi-directional HiFi mode. Since time synchronization is needed for
proper Display Port tunneling this means we need to configure the TMU on
USB4 compliant devices.
The USB4 spec allows some flexibility on how the TMU can be configured.
This makes it possible to enable link power management states (CLx) in
certain topologies, where for example DP tunneling is not used. TMU can
also be re-configured dynamicaly depending on types of tunnels created
over the USB4 fabric.
In this patch we simply configure the TMU to be in bi-directional HiFi
mode. This way we can tunnel any kind of traffic without need to perform
complex steps to re-configure the domain dynamically. We can add more
fine-grained TMU configuration later on when we start enabling CLx
states.
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Co-developed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-8-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We need to find switch capabilities in order to implement TMU support so
make it available to other files as well.
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-7-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since the driver now supports USB4 which is the standard going forward,
update the Kconfig entry to mention this and rename the entry from
CONFIG_THUNDERBOLT to CONFIG_USB4 instead to help people to find the
correct option if they want to enable USB4.
Also do the same for Thunderbolt network driver.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: David S. Miller <davem@davemloft.net>
Link: https://lore.kernel.org/r/20191217123345.31850-6-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
USB4 is the public specification based on Thunderbolt 3 protocol. There
are some differences in register layouts and flows. In addition to PCIe
and DP tunneling, USB4 supports tunneling of USB 3.x. USB4 is also
backward compatible with Thunderbolt 3 (and older generations but the
spec only talks about 3rd generation). USB4 compliant devices can be
identified by checking USB4 version field in router configuration space.
This patch adds initial support for USB4 compliant hosts and devices
which enables following features provided by the existing functionality
in the driver:
- PCIe tunneling
- Display Port tunneling
- Host and device NVM firmware upgrade
- P2P networking
This brings the USB4 support to the same level that we already have for
Thunderbolt 1, 2 and 3 devices.
Note the spec talks about host and device "routers" but in the driver we
still use term "switch" in most places. Both can be used interchangeably.
Co-developed-by: Rajmohan Mani <rajmohan.mani@intel.com>
Signed-off-by: Rajmohan Mani <rajmohan.mani@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-5-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
USB4 1.0 section 6.4.2.7 specifies a new field (PG) in notification
packet that is sent as response of hot plug/unplug events. This field
tells whether the acknowledgment is for plug or unplug event. This needs
to be set accordingly in order the router to send further hot plug
notifications.
To make it simpler we fill the field unconditionally. Legacy devices do
not look at this field so there should be no problems with them.
While there rename tb_cfg_error() to tb_cfg_ack_plug() and update the
log message accordingly. The function is only used to ack plug/unplug
events.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-4-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We are going to re-use tb_drom_read() for USB4 DROM reading as well.
USB4 has separate router operations for this which does not need the
drom_offset. Therefore we move call to tb_eeprom_get_drom_offset() into
tb_eeprom_read_n() where it is needed.
While there change return -ENOSYS to -ENODEV because the former is only
supposed to be used with system calls (invalid syscall nr).
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-3-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We will be needing this when adding initial USB4 support so make it
available to other files in the driver as well. We also rename it to
tb_switch_find_port() to follow conventions used in switch.c.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Link: https://lore.kernel.org/r/20191217123345.31850-2-mika.westerberg@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On zang's Dell XPS 13 9370 after Thunderbolt NVM firmware upgrade the
Thunderbolt controller did not come back as expected. Only after the
system was rebooted it became available again. It is not entirely clear
what happened but I suspect the new NVM firmware image authentication
failed for some reason. Regardless of this the router needs to be power
cycled if NVM authentication fails in order to get it fully functional
again.
This modifies the driver to issue a power cycle in case the NVM
authentication fails immediately when dma_port_flash_update_auth()
returns. We also need to call tb_switch_set_uuid() earlier to be able to
fetch possible NVM authentication failure when DMA port is added.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=205457
Reported-by: zang <dump@tzib.net>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since now we can do pretty much the same thing in the software
connection manager than the firmware would do, there is no point
starting it by default. Instead we can just continue using the software
connection manager.
Make it possible for user to switch between the two by adding a module
pararameter (start_icm) which is by default false. Having this ability
to enable the firmware may be useful at least when debugging possible
issues with the software connection manager implementation.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Titan Ridge supports Display Port 1.4 which adds HBR3 (High Bit Rate)
rates that may be up to 8.1 Gb/s over 4 lanes. This translates to
effective data bandwidth of 25.92 Gb/s (as 8/10 encoding is removed by
the DP adapters when going over Thunderbolt fabric). If another high
rate monitor is connected we may need to reduce the bandwidth it
consumes so that it fits into the total 40 Gb/s available on the
Thunderbolt fabric.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
To perform proper Display Port tunneling for Thunderbolt 3 devices we
need to allocate DP resources for DP IN port before they can be used.
The reason for this is that the user can also connect a monitor directly
to the Type-C ports in which case the Thunderbolt controller acts as
re-driver for Display Port (no tunneling takes place) taking the DP
sinks away from the connection manager. This allocation is done using
special sink allocation registers available through the link controller.
We can pair DP IN to DP OUT only if
* DP IN has sink allocated via link controller
* DP OUT port receives hotplug event
For DP IN adapters (only for the host router) we first query whether
there is DP resource available (it may be the previous instance of the
driver for example already allocated it) and if it is we add it to the
list. We then update the list when after each plug/unplug event to a DP
IN/OUT adapter. Each time the list is updated we try to find additional
DP IN <-> DP OUT pairs for tunnel establishment. This strategy also
makes it possible to establish another tunnel in case there are 3
monitors connected and one gets unplugged releasing the DP IN adapter
for the new tunnel.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Titan Ridge needs an additional connection manager handshake in order to
do proper Display Port tunneling so implement it here.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
In order to keep PCIe hierarchies consistent across hotplugs, add
hard-coded PCIe downstream port to Thunderbolt port for Alpine Ridge and
Titan Ridge as well.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
For a casual reader tb_switch_is_cr() does not tell much so instead
spell out the full controller name in the function name. For example
tb_switch_is_cr() becomes tb_switch_is_cactus_ridge() which is easier
to understand.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
We currently read how sibling lane adapter ports relate each other from
DROM (Device ROM). If the two lane adapter ports go through the same
physical connector these lanes can then be bonded together. However,
some cases DROM does not provide this information or it is missing
completely (host routers typically do not have DROM). In this case we
have hard-coded the relationship.
Expand this to work with both legacy devices where lane adapter ports 1
and 2, and 3 and 4 are always linked together, and with USB4 devices
where lane adapter 1 is always following lane adapter 0 or is disabled
completely (see USB4 section 5.2.1 for more information).
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Lane bonding allows aggregating two 10/20 Gb/s (depending on the
generation) lanes into a single 20/40 Gb/s bonded link. This allows
sharing the full bandwidth more efficiently. In order to establish lane
bonding we need to check that lane bonding is possible through link
controller and that both ends of the link actually supports 2x widths.
This also means that all the paths should be established through the
primary port so update tb_path_alloc() to handle this as well.
Lane bonding is supported starting from Falcon Ridge (2nd generation)
controllers.
We also expose the current speed and number of lanes under each device
except the host router following similar attribute naming than USB bus.
Expose speed and number of lanes for both directions to allow possibility
of asymmetric link in the future.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Currently add_switch() takes a huge amount of parameters that makes it
hard to maintain. Instead of passing all those parameters we can split
the function into two parts (alloc and add) and fill the additional
switch fields directly in the functions calling those.
While there remove redundant error logging in case kmemdup() fails.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
There are quite many places in the driver where we iterate over each
port in the switch. To make it bit more convenient, add a macro that can
be used to iterate over each port and convert existing call sites to use it.
This is based on code by Lukas Wunner.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Now that USB4 spec has names for these DP adapter registers we can use
them instead. This makes it easier to match certain register to the spec.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Now that USB4 spec has names for these PCIe adapter registers we can use
them instead. This makes it easier to match certain register to the spec.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Now that USB4 spec has names for these basic registers we can use them
instead. This makes it easier to match certain register to the spec.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
If we fail to add a switch for some reason log an error instead of
keeping silent. This is useful for debugging.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
We currently differentiate between SW CM (Software Connection Manager,
sometimes also called External Connection Manager) and ICM (Firmware
based Connection Manager, Internal Connection Manager) by looking
directly at the sw->config.enabled field which may be rather hard to
understand for the casual reader. For this reason introduce a wrapper
function with documentation that should make the intention more clear.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The Thunderbolt standard went through several major iterations, here
called generation. USB4, which will be based on Thunderbolt, will be
generation 4. Let userspace know the generation of the controller in
the devices in order to distinguish between Thunderbolt and USB4, so
it can be shown in various user interfaces.
Signed-off-by: Christian Kellner <christian@kellner.me>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The read is not needed as we overwrite the returned value in the next
line anyway so drop it.
Fixes: 3cdb9446a1 ("thunderbolt: Add support for Intel Ice Lake")
Reported-by: Nicholas Johnson <nicholas.johnson-opensource@outlook.com.au>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
When lockdep is enabled, plugging Thunderbolt dock on Dominik's laptop
triggers following splat:
======================================================
WARNING: possible circular locking dependency detected
5.3.0-rc6+ #1 Tainted: G T
------------------------------------------------------
pool-/usr/lib/b/1258 is trying to acquire lock:
000000005ab0ad43 (pci_rescan_remove_lock){+.+.}, at: authorized_store+0xe8/0x210
but task is already holding lock:
00000000bfb796b5 (&tb->lock){+.+.}, at: authorized_store+0x7c/0x210
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&tb->lock){+.+.}:
__mutex_lock+0xac/0x9a0
tb_domain_add+0x2d/0x130
nhi_probe+0x1dd/0x330
pci_device_probe+0xd2/0x150
really_probe+0xee/0x280
driver_probe_device+0x50/0xc0
bus_for_each_drv+0x84/0xd0
__device_attach+0xe4/0x150
pci_bus_add_device+0x4e/0x70
pci_bus_add_devices+0x2e/0x66
pci_bus_add_devices+0x59/0x66
pci_bus_add_devices+0x59/0x66
enable_slot+0x344/0x450
acpiphp_check_bridge.part.0+0x119/0x150
acpiphp_hotplug_notify+0xaa/0x140
acpi_device_hotplug+0xa2/0x3f0
acpi_hotplug_work_fn+0x1a/0x30
process_one_work+0x234/0x580
worker_thread+0x50/0x3b0
kthread+0x10a/0x140
ret_from_fork+0x3a/0x50
-> #0 (pci_rescan_remove_lock){+.+.}:
__lock_acquire+0xe54/0x1ac0
lock_acquire+0xb8/0x1b0
__mutex_lock+0xac/0x9a0
authorized_store+0xe8/0x210
kernfs_fop_write+0x125/0x1b0
vfs_write+0xc2/0x1d0
ksys_write+0x6c/0xf0
do_syscall_64+0x50/0x180
entry_SYSCALL_64_after_hwframe+0x49/0xbe
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&tb->lock);
lock(pci_rescan_remove_lock);
lock(&tb->lock);
lock(pci_rescan_remove_lock);
*** DEADLOCK ***
5 locks held by pool-/usr/lib/b/1258:
#0: 000000003df1a1ad (&f->f_pos_lock){+.+.}, at: __fdget_pos+0x4d/0x60
#1: 0000000095a40b02 (sb_writers#6){.+.+}, at: vfs_write+0x185/0x1d0
#2: 0000000017a7d714 (&of->mutex){+.+.}, at: kernfs_fop_write+0xf2/0x1b0
#3: 000000004f262981 (kn->count#208){.+.+}, at: kernfs_fop_write+0xfa/0x1b0
#4: 00000000bfb796b5 (&tb->lock){+.+.}, at: authorized_store+0x7c/0x210
stack backtrace:
CPU: 0 PID: 1258 Comm: pool-/usr/lib/b Tainted: G T 5.3.0-rc6+ #1
On an system using ACPI hotplug the host router gets hotplugged first and then
the firmware starts sending notifications about connected devices so the above
scenario should not happen in reality. However, after taking a second
look at commit a03e828915 ("thunderbolt: Serialize PCIe tunnel
creation with PCI rescan") that introduced the locking, I don't think it
is actually correct. It may have cured the symptom but probably the real
root cause was somewhere closer to PCI stack and possibly is already
fixed with recent kernels. I also tried to reproduce the original issue
with the commit reverted but could not.
So to keep lockdep happy and the code bit less complex drop calls to
pci_lock_rescan_remove()/pci_unlock_rescan_remove() in
tb_switch_set_authorized() effectively reverting a03e828915.
Link: https://lkml.org/lkml/2019/8/30/513
Fixes: a03e828915 ("thunderbolt: Serialize PCIe tunnel creation with PCI rescan")
Reported-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
When we discover existing DP tunnels the code checks whether DP IN
adapter port is enabled by calling tb_dp_port_is_enabled() before it
continues the discovery process. On Light Ridge (gen 1) controller
reading only the first dword of the DP IN config space causes subsequent
access to the same DP IN port path config space to fail or return
invalid data as can be seen in the below splat:
thunderbolt 0000:07:00.0: CFG_ERROR(0:d): Invalid config space or offset
Call Trace:
tb_cfg_read+0xb9/0xd0
__tb_path_deactivate_hop+0x98/0x210
tb_path_activate+0x228/0x7d0
tb_tunnel_restart+0x95/0x200
tb_handle_hotplug+0x30e/0x630
process_one_work+0x1b4/0x340
worker_thread+0x44/0x3d0
kthread+0xeb/0x120
? process_one_work+0x340/0x340
? kthread_park+0xa0/0xa0
ret_from_fork+0x1f/0x30
If both DP In adapter config dwords are read in one go the issue does
not reproduce. This is likely firmware bug but we can work it around by
always reading the two dwords in one go. There should be no harm for
other controllers either so can do it unconditionally.
Link: https://lkml.org/lkml/2019/8/28/160
Reported-by: Brad Campbell <lists2009@fnarfbargle.com>
Tested-by: Brad Campbell <lists2009@fnarfbargle.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The Thunderbolt controller is integrated into the Ice Lake CPU itself
and requires special flows to power it on and off using force power bit
in NHI VSEC registers. Runtime PM (RTD3) and Sx flows also differ from
the discrete solutions. Now the firmware notifies the driver whether
RTD3 entry or exit are possible. The driver is responsible of sending
Go2Sx command through link controller mailbox when system enters Sx
states (suspend-to-mem/disk). Rest of the ICM firwmare flows follow
Titan Ridge.
Signed-off-by: Raanan Avargil <raanan.avargil@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Tested-by: Mario Limonciello <mario.limonciello@dell.com>
Ice Lake Thunderbolt controller NVM firmware is part of the BIOS image
which means it is not writable through the DMA port anymore. However, we
can still read it so we can keep nvm_version and active parts of NVM.
This way users still can find out the active NVM version and other
potentially useful information directly from Linux.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Tested-by: Mario Limonciello <mario.limonciello@dell.com>
Thunderbolt host routers may not always contain DROM that includes
device identification information. This is mostly needed for Ice Lake
systems but some Falcon Ridge controllers on PCs also do not have DROM.
In that case hide the identification attributes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Tested-by: Mario Limonciello <mario.limonciello@dell.com>
There are two ways to mark a port as unimplemented. Typical way is to
return port type as TB_TYPE_INACTIVE when its config space is read.
Alternatively if the port is not physically present (such as ports 10
and 11 in ICL) reading from port config space returns
TB_CFG_ERROR_INVALID_CONFIG_SPACE instead. Currently the driver bails
out from adding the switch if it receives any error during port
inititialization which is wrong.
Handle this properly and just leave the port as TB_TYPE_INACTIVE before
continuing to the next port.
This also allows us to get rid of special casing for Light Ridge port 5
in eeprom.c.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Tested-by: Mario Limonciello <mario.limonciello@dell.com>
The register access should be using 32-bit reads/writes according to the
datasheet. With the previous generation hardware 16-bit writes have been
working but starting with ICL this is not the case anymore so fix
producer/consumer register update to use correct width register address.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Tested-by: Mario Limonciello <mario.limonciello@dell.com>
This is depends on the controller and on the platform/CPU we are
running. Move it to struct icm so we can set it per controller.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Tested-by: Mario Limonciello <mario.limonciello@dell.com>
PCIe tunnel path indices got mixed up when we added support for tunnels
between switches that are not adjacent. This did not affect the
functionality as it is just an index but fix it now nevertheless to make
the code easier to understand.
Reported-by: Rajmohan Mani <rajmohan.mani@intel.com>
Fixes: 8c7acaaf02 ("thunderbolt: Extend tunnel creation to more than 2 adjacent switches")
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Yehezkel Bernat <YehezkelShB@gmail.com>
%*pEp (without "h" or "o") is a no-op. This string could contain
arbitrary (non-NULL) characters, so we do want escaping. Use %*pE like
every other caller.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Use device_property_count_uXX() directly, that makes code neater.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Yehezkel Bernat <YehezkelShB@gmail.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Rajmohan Mani