Commit graph

126,670 commits

Author SHA1 Message Date
Michal Wajdeczko
95336cfd5b drm/xe: Add memirq report page address helpers
Both xe_memirq_{source,status}_ptr functions are now strictly for
obtaining an address of the memory based interrupt report pages
used by the HW engines. When initializing the GuC that does not
require special per instance page preparation, we don't need to
abuse these public functions and pass a NULL instead of valid hwe
pointer. Also, without further fixes, this actually may lead to
NPD crash once the hw_reports_to_instance_zero() will be true.

Add internal helpers that will provide report page addresses based
solely on the instance number, which will be always 0 for both GuCs.

Fixes: ef6103d20f ("drm/xe: memirq infra changes for MSI-X")
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Piotr Piórkowski <piotr.piorkowski@intel.com>
Cc: Ilia Levi <ilia.levi@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240925084445.1495-1-michal.wajdeczko@intel.com
2024-10-06 14:24:28 +02:00
Linus Torvalds
fe6fceceae drm fixes for 6.12-rc2
atomic:
 - Use correct type when reading damage rectangles
 
 display:
 - Fix kernel docs
 
 dp-mst:
 - Fix DSC decompression detection
 
 hdmi:
 - Fix infoframe size
 
 sched:
 - Update maintainers
 - Fix race condition whne queueing up jobs
 - Fix locking in drm_sched_entity_modify_sched()
 - Fix pointer deref if entity queue changes
 
 sysfb:
 - Disable sysfb if framebuffer parent device is unknown
 
 amdgpu:
 - DML2 fix
 - DSC fix
 - Dispclk fix
 - eDP HDR fix
 - IPS fix
 - TBT fix
 
 i915:
 - One fix for bitwise and logical "and" mixup in PM code
 
 xe:
 - Restore pci state on resume
 - Fix locking on submission, queue and vm
 - Fix UAF on queue destruction
 - Fix resource release on freq init error path
 - Use rw_semaphore to reduce contention on ASID->VM lookup
 - Fix steering for media on Xe2_HPM
 - Tuning updates to Xe2
 - Resume TDR after GT reset to prevent jobs running forever
 - Move id allocation to avoid userspace using a guessed number
   to trigger UAF
 - Fix OA stream close preventing pbatch buffers to complete
 - Fix NPD when migrating memory on LNL
 - Fix memory leak when aborting binds
 
 panthor:
 - Fix locking
 - Set FOP_UNSIGNED_OFFSET in fops instance
 - Acquire lock in panthor_vm_prepare_map_op_ctx()
 - Avoid uninitialized variable in tick_ctx_cleanup()
 - Do not block scheduler queue if work is pending
 - Do not add write fences to the shared BOs
 
 vbox:
 - Fix VLA handling
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCAAdFiEEEKbZHaGwW9KfbeusDHTzWXnEhr4FAmb/YwgACgkQDHTzWXnE
 hr6VbxAAoq9FYTAdRPWzfG1HYpG96UyTh+IT6lz1bk/Hblxhi7oRdfmRy/bVPQYh
 vj+Q2xrnyS6JYhyfeDT2nU75tD3gvR1V/qSamxXS7c1nrqcb431DaMzuSQ5ST6MZ
 jrmob2TxlbXDDw70dxtiGCmSu0a9QInbelEamJQySKOdun0Il5C0LRZIBMpicDGc
 7Y3eSpCIwgTSU6bnApGyOchppvzptiqBWGmhoIuACMOgXI8eaLUPqbROKEHlPe5g
 JIG603rRK7cf+on/KEwvgrd2ZO59fJZvmwFrM5yY5bOsDCwTIJ6mHhOmutUNQvmd
 G5n6ZFnVxlBRSVWCAqPRBgA405s/0wi2IQprilaPCu2qAXToBXAUpIHuuat5I/8b
 BVwurVRAGV/GSeg7E51H3o8cu/fcQr4aGNW4Ul6fS1G123ZuUISpcUp9IEnqG7nB
 5PSnHapadb5Pu+7kwhbWUD4kONp16oEacZPhymlN+74Q6X3v8UVlK/YSyD6wq4fj
 2s4TBWUmXmNxztNEkgJhyJORYQhZeBaD0PtPq8kSzMUCFj3Q7Wf0bAODhbnmdCw1
 iPgxRd9+38IpfW621AROJUoTcyCaLtlSUvHWgFfska5CYnbtbKuNPBW6baQxeEHe
 Rhns01dZhNTIbKNEw37cfOf2DqcjpmRm4cVJj4xjZawxWYlNldk=
 =wcmz
 -----END PGP SIGNATURE-----

Merge tag 'drm-fixes-2024-10-04' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Weekly fixes, xe and amdgpu lead the way, with panthor, and few core
  components getting various fixes. Nothing seems too out of the
  ordinary.

  atomic:
   - Use correct type when reading damage rectangles

  display:
   - Fix kernel docs

  dp-mst:
   - Fix DSC decompression detection

  hdmi:
   - Fix infoframe size

  sched:
   - Update maintainers
   - Fix race condition whne queueing up jobs
   - Fix locking in drm_sched_entity_modify_sched()
   - Fix pointer deref if entity queue changes

  sysfb:
   - Disable sysfb if framebuffer parent device is unknown

  amdgpu:
   - DML2 fix
   - DSC fix
   - Dispclk fix
   - eDP HDR fix
   - IPS fix
   - TBT fix

  i915:
   - One fix for bitwise and logical "and" mixup in PM code

  xe:
   - Restore pci state on resume
   - Fix locking on submission, queue and vm
   - Fix UAF on queue destruction
   - Fix resource release on freq init error path
   - Use rw_semaphore to reduce contention on ASID->VM lookup
   - Fix steering for media on Xe2_HPM
   - Tuning updates to Xe2
   - Resume TDR after GT reset to prevent jobs running forever
   - Move id allocation to avoid userspace using a guessed number to
     trigger UAF
   - Fix OA stream close preventing pbatch buffers to complete
   - Fix NPD when migrating memory on LNL
   - Fix memory leak when aborting binds

  panthor:
   - Fix locking
   - Set FOP_UNSIGNED_OFFSET in fops instance
   - Acquire lock in panthor_vm_prepare_map_op_ctx()
   - Avoid uninitialized variable in tick_ctx_cleanup()
   - Do not block scheduler queue if work is pending
   - Do not add write fences to the shared BOs

  vbox:
   - Fix VLA handling"

* tag 'drm-fixes-2024-10-04' of https://gitlab.freedesktop.org/drm/kernel: (41 commits)
  drm/xe: Fix memory leak when aborting binds
  drm/xe: Prevent null pointer access in xe_migrate_copy
  drm/xe/oa: Don't reset OAC_CONTEXT_ENABLE on OA stream close
  drm/xe/queue: move xa_alloc to prevent UAF
  drm/xe/vm: move xa_alloc to prevent UAF
  drm/xe: Clean up VM / exec queue file lock usage.
  drm/xe: Resume TDR after GT reset
  drm/xe/xe2: Add performance tuning for L3 cache flushing
  drm/xe/xe2: Extend performance tuning to media GT
  drm/xe/mcr: Use Xe2_LPM steering tables for Xe2_HPM
  drm/xe: Use helper for ASID -> VM in GPU faults and access counters
  drm/xe: Convert to USM lock to rwsem
  drm/xe: use devm_add_action_or_reset() helper
  drm/xe: fix UAF around queue destruction
  drm/xe/guc_submit: add missing locking in wedged_fini
  drm/xe: Restore pci state upon resume
  drm/amd/display: Fix system hang while resume with TBT monitor
  drm/amd/display: Enable idle workqueue for more IPS modes
  drm/amd/display: Add HDR workaround for specific eDP
  drm/amd/display: avoid set dispclk to 0
  ...
2024-10-04 11:25:14 -07:00
Matt Roper
93d9381342 drm/xe: Make wedged_mode debugfs writable
The intent of this debugfs entry is to allow modification of wedging
behavior, either from IGT tests or during manual debug; it should be
marked as writable to properly reflect this.  In practice this hasn't
caused a problem because we always access wedged_mode as root, which
ignores file permissions, but it's still misleading to have the entry
incorrectly marked as RO.

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Fixes: 6b8ef44cc0 ("drm/xe: Introduce the wedged_mode debugfs")
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241002230620.1249258-2-matthew.d.roper@intel.com
2024-10-04 08:10:35 -07:00
Thomas Hellström
b88132ceb3 Merge drm/drm-next into drm-xe-next
Backmerging to resolve a conflict with core locally.

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
2024-10-04 11:46:30 +02:00
Jani Nikula
8231ac7e72 drm/i915: use NULL for zero wakeref_t instead of plain integer 0
As of commit 2edc6a75f2 ("drm/i915: switch intel_wakeref_t underlying
type to struct ref_tracker *") we gained quite a few sparse warnings
about "Using plain integer as NULL pointer" for using 0 to initialize
wakeref_t. Switch to NULL everywhere.

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241002181655.582597-1-jani.nikula@intel.com
2024-10-04 10:29:24 +03:00
Dave Airlie
5b272bf7dc Driver Changes:
- Restore pci state on resume (Rodrigo Vivi)
 - Fix locking on submission, queue and vm (Matthew Auld, Matthew Brost)
 - Fix UAF on queue destruction (Matthew Auld)
 - Fix resource release on freq init error path (He Lugang)
 - Use rw_semaphore to reduce contention on ASID->VM lookup (Matthew Brost)
 - Fix steering for media on Xe2_HPM (Gustavo Sousa)
 - Tuning updates to Xe2 (Gustavo Sousa)
 - Resume TDR after GT reset to prevent jobs running forever (Matthew Brost)
 - Move id allocation to avoid userspace using a guessed number
   to trigger UAF (Matthew Auld, Matthew Brost)
 - Fix OA stream close preventing pbatch buffers to complete (José)
 - Fix NPD when migrating memory on LNL (Zhanjun Dong)
 - Fix memory leak when aborting binds (Matthew Brost)
 -----BEGIN PGP SIGNATURE-----
 
 iQJNBAABCAA3FiEE6rM8lpABPHM5FqyDm6KlpjDL6lMFAmb+pSoZHGx1Y2FzLmRl
 bWFyY2hpQGludGVsLmNvbQAKCRCboqWmMMvqU4N9EACWXIAOU0/U/CME5pNL8zmf
 P6ISnP7XpvTKEnm0eKkj2UNqHfI3oU1Nr+jFVZwCvu7zCZQHaKc7bW1HfHqSvZkq
 ZFNmhN6JeDPcA+bKolZKWbQJiz+XfXRWQZ7eGu0rnvjBsXWHQFWo+8w0SpuSUfYU
 EDES6hGDfLjwe8I1uahJqyseQQxV66orUSpooSS+assCPnH+dexuH459IEY5Uooy
 GRj9Mo76j9aV5eLXK5jr0fD1Vj+OY1TI+7UNSaG1x3ZPembpfrXeNqToU/IdYQ2R
 CzdHbvSDNu8VX+v48QViBM1QsnW1oyGQfN5+3G2eOC+SPvGmqi/Mr+0ytMc27lVd
 +wfwUrBCY1ksZpUl2dt7Ct4mxMnUuZfQY8CAFtHhzkkVb5zKxoxTdXdQS4EATzYA
 ayXNcJPrj+E9Z73MGx5mB9cYSW9F58iuM4Yr9hzP2VoiGQ84PoK022sBA7DMAR0u
 NNTqDCr/ntQkC6heLorPKxfojeTQnl7u3oMzbai0XtadNp77sH4Uoi/zOKrzDeSp
 wz8n7qdNoqDYB2TTgAhei8+UN/04e2KhD5coAzBxJGMsrFxhiYnRq849ozLhxegy
 glA6QmScoHZywqy8VSaDi06pZ0SwoSCtjMnj+q6Wu4raAwuvh4HPsdI2p/5lBKn3
 hDHKS6bpcuMyl1H0ekfmag==
 =FjY1
 -----END PGP SIGNATURE-----

Merge tag 'drm-xe-fixes-2024-10-03' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-fixes

Driver Changes:
- Restore pci state on resume (Rodrigo Vivi)
- Fix locking on submission, queue and vm (Matthew Auld, Matthew Brost)
- Fix UAF on queue destruction (Matthew Auld)
- Fix resource release on freq init error path (He Lugang)
- Use rw_semaphore to reduce contention on ASID->VM lookup (Matthew Brost)
- Fix steering for media on Xe2_HPM (Gustavo Sousa)
- Tuning updates to Xe2 (Gustavo Sousa)
- Resume TDR after GT reset to prevent jobs running forever (Matthew Brost)
- Move id allocation to avoid userspace using a guessed number
  to trigger UAF (Matthew Auld, Matthew Brost)
- Fix OA stream close preventing pbatch buffers to complete (José)
- Fix NPD when migrating memory on LNL (Zhanjun Dong)
- Fix memory leak when aborting binds (Matthew Brost)

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/2fiv63yanlal5mpw3mxtotte6yvkvtex74c7mkjxca4bazlyja@o4iejcfragxy
2024-10-04 11:00:39 +10:00
Colin Ian King
301d194d01 drm/nouveau/gsp: remove extraneous ; after mutex
The mutex field has two following semicolons, replace this with just
one semicolon.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240917120856.1877733-1-colin.i.king@gmail.com
2024-10-03 23:49:40 +02:00
Ville Syrjälä
45c548642b drm/i915/dsb: Use DSB for plane/color management updates
Push regular plane/color management updates to the DSB,
if other constraints allow it.

The first part of the sequence will go as follows:
- CPU will kick off DSB0 immediately
- DSB0 writes double bufferd non-arming registers
- DSB0 evades the vblank
- DSB0 writes double buffered arming registers

If no color management updates is needed we follow that up with:
- DSB0 waits for the undelayed vblank
- DSB0 waits for the delayed vblank (usec wait)
- DSB0 emits an interrupt which will cause the CPU to complete the commit

If color management update is needed:
- DSB0 will start DSB1 with wait for undelayed vblank
- DSB0 will in parallel perform the force DEwake tricks
- DSB1 writes single buffered LUT registers
- DSB1 waits for the delayed vblank (usec wait)
- DSB1 emits an interrupt which will cause the CPU to complete the commit

With this sequence we don't need to increase the vblank delay
to make room for register programming during vblank, which is
a good thing for high refresh rate display. But I'll need to
still think of some way to eliminate VRR commit completion
related races under this scheme.

Stuff that isn't ready for DSB yet:
- modesets (potentially we could do
  at least the plane enabling via DSB)
- fastsets
- VRR
- PSR
- scalers
- async flips

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-14-ville.syrjala@linux.intel.com
2024-10-04 00:33:44 +03:00
Ville Syrjälä
a6d4d9776e drm/i915: Plumb 'dsb' all way to the color commit hooks
Pass the 'dsb' all the way down to the color commit hooks so that
we'll be able to update the double buffered color management registers
(eg. CSC) via the DSB.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-13-ville.syrjala@linux.intel.com
2024-10-04 00:33:39 +03:00
Ville Syrjälä
01389846f7 drm/i915: Plumb 'dsb' all way to the plane hooks
We need to be able to do both MMIO and DSB based pipe/plane
programming. To that end plumb the 'dsb' all way from the top
into the plane commit hooks.

The compiler appears smart enough to combine the branches from
all the back-to-back register writes into a single branch.
So the generated asm ends up looking more or less like this:
plane_hook()
{
	if (dsb) {
		intel_dsb_reg_write();
		intel_dsb_reg_write();
		...
	} else {
		intel_de_write_fw();
		intel_de_write_fw();
		...
	}
}
which seems like a reasonably efficient way to do this.

An alternative I was also considering is some kind of closure
(register write function + display vs. dsb pointer passed to it).
That does result is smaller code as there are no branches anymore,
but having each register access go via function pointer sounds
less efficient.

Not that I actually measured the overhead of either approach yet.
Also the reg_rw tracepoint seems to be making a huge mess of the
generated code for the mmio path. And additionally there's some
kind of IS_GSI_REG() hack in __raw_uncore_read() which ends up
generating a pointless branch for every mmio register access.
So looks like there might be quite a bit of room for improvement
in the mmio path still.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-12-ville.syrjala@linux.intel.com
2024-10-04 00:33:28 +03:00
Ville Syrjälä
dd6ec895ba drm/i915: Extract intel_crtc_prepare_vblank_event()
Extract the code for staging the vblank event for the
flip done interrupt handler. We'll reuse this for DSB
stuff later.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-11-ville.syrjala@linux.intel.com
2024-10-04 00:33:25 +03:00
Ville Syrjälä
08222ca194 drm/i915/dsb: Introduce intel_dsb_wait_vblank_delay()
Add intel_dsb_wait_vblank_delay() which instructs the DSB
to wait for duration between the undelayed and delayed vblanks.
We'll need this as the DSB can only directly wait for the
undelayed vblank, but we'll need to wait until the delayed
vblank has elapsed as well.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-10-ville.syrjala@linux.intel.com
2024-10-04 00:33:21 +03:00
Ville Syrjälä
d4f9a053bf drm/i915: Introduce intel_scanlines_to_usecs()
Introduce intel_scanlines_to_usecs() as a counterpart to
intel_usecs_to_scanlines().

We'll have some use for this in DSB code as we want to do
relative scanline waits to evade the delayed vblank, but
unfortunately DSB can't do relative scanline waits (only
absolute). So we'll instead convert the relative scanline
count to usec and do a relative usec wait instead (which the
DSB knows how to do).

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-9-ville.syrjala@linux.intel.com
2024-10-04 00:33:18 +03:00
Ville Syrjälä
d6dfbc6f81 drm/i915/dsb: Introduce intel_dsb_wait_vblanks()
Add a function to emit a DSB wait for vblank instruction. This
just waits until the specified number of vblanks.

Note that this triggers on the transcoder's undelayed vblank,
as opposed to the pipe's delayed vblank.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-8-ville.syrjala@linux.intel.com
2024-10-04 00:33:13 +03:00
Ville Syrjälä
de968532fd drm/i915/dsb: Introduce intel_dsb_wait_usec()
Add a function to emit the DSB "wait usecs" instruction.
This is just a usleep() for the DSB.

As a lower bound it seems pretty accurate, but the upper bound
seemed oddly relaxed (ie. sometimes I've seen waits that are
quite a bit longer than specified, not sure why).

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-7-ville.syrjala@linux.intel.com
2024-10-04 00:33:10 +03:00
Ville Syrjälä
63b41d207d drm/i915/dsb: Introduce intel_dsb_vblank_evade()
Add a helper for performing vblank evasion on the DSB. DSB based
plane updates will need this to guarantee all the double buffered
arming registers will get programmed atomically within the same
frame.

With VRR we more or less have two vblanks to worry about:
- vmax vblank start in case no push was sent
- vmin vblank start in case a push was already sent during
  the vertical active. Only a concern for mailbox updates,
  which I suppose could happen if the legacy cursor updates
  take the non-fastpath without setting
  state->legacy_cursor_update to false.
Since we don't know which case is relevant we'll just evade
both.

We must also make sure to evade both the delayed vblank
(for pipe/plane registers) and the undelayed vblank
(for transcoder registers and chained DSBs w/
DSB_WAIT_FOR_VBLANK).

TODO: come up with a sensible usec number for the evasion...

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-6-ville.syrjala@linux.intel.com
2024-10-04 00:33:06 +03:00
Ville Syrjälä
9e9953715e drm/i915/dsb: Enable programmable DSB interrupt
The DSB can signal a programmable interrupt in response to
a specific DSB command getting executed. Hook that up.

For now we'll just use this to signal the completion of the
commit via a vblank event. If, in the future, we'll need to
do other things in response to DSB interrupts we may need to
come up with some kind of fancier DSB interrupt framework where
the caller can specify a custom handler...

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-5-ville.syrjala@linux.intel.com
2024-10-04 00:33:02 +03:00
Ville Syrjälä
b0413571bc drm/i915/dsb: Generate the DSB buffer in commit_tail()
Once we start using DSB for plane updates we'll need to defer
generating the DSB buffer until the clear color has been
read out. So we need to move at some of the DSB stuff into
commit_tail(). That is perhaps a better place for it anyway
as the ioctl thread can move on immediately without spending
time building the DSB commands.

We always have the MMIO fallback (in case the DSB buffer
allocation fails), so there's no real reason to keep any
of this in the synchronous part of the ioctl.

Because the DSB LUT programming doesn't depend on the plane
clear color we can still do that part before waiting for
fences/etc. which should help paralleize things a bit more.
The DSB plane programming will need to happen after those
however as that depends on the clear color.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-4-ville.syrjala@linux.intel.com
2024-10-04 00:32:58 +03:00
Ville Syrjälä
b7e247b3c9 drm/i915: Prepare clear color before wait_for_dependencies()
Read out the clear color as soon as fences and the transient
data flush have finished. There is no need to wait for
all the display specific operations that might still be
going on. This could parallelize things a bit more effectively.

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-3-ville.syrjala@linux.intel.com
2024-10-04 00:32:55 +03:00
Ville Syrjälä
0e94059113 drm/i915/dsb: Avoid reads of the DSB buffer for indexed register writes
Reading from the DSB command buffer might be somewhat expensive on
discrete GPUs because the buffer resides in GPU local memory. Avoid
such reads in the indexed register write handling by tracking the
previous instruction in intel_dsb.

TODO: actually measure this

Reviewed-by: Animesh Manna <animesh.manna@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930170415.23841-2-ville.syrjala@linux.intel.com
2024-10-04 00:32:50 +03:00
Pin-yen Lin
574c558ddb drm/bridge: it6505: Drop EDID cache on bridge power off
The bridge might miss the display change events when it's powered off.
This happens when a user changes the external monitor when the system
is suspended and the embedded controller doesn't not wake AP up.

It's also observed that one DP-to-HDMI bridge doesn't work correctly
when there is no EDID read after it is powered on.

Drop the cache to force an EDID read after system resume to fix this.

Fixes: 11feaef69d ("drm/bridge: it6505: Add caching for EDID")
Signed-off-by: Pin-yen Lin <treapking@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240926092931.3870342-3-treapking@chromium.org
2024-10-03 13:24:33 -07:00
Pin-yen Lin
00ae002116 drm/bridge: anx7625: Drop EDID cache on bridge power off
The bridge might miss the display change events when it's powered off.
This happens when a user changes the external monitor when the system
is suspended and the embedded controller doesn't not wake AP up.

It's also observed that one DP-to-HDMI bridge doesn't work correctly
when there is no EDID read after it is powered on.

Drop the cache to force an EDID read after system resume to fix this.

Fixes: 8bdfc5dae4 ("drm/bridge: anx7625: Add anx7625 MIPI DSI/DPI to DP")
Signed-off-by: Pin-yen Lin <treapking@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240926092931.3870342-2-treapking@chromium.org
2024-10-03 13:23:51 -07:00
Antonino Maniscalco
20f3327575 drm/msm/a6xx: Enable preemption for tested a7xx targets
Initialize with 4 rings to enable preemption.

Add the "preemption_enabled" module parameter to override this.

Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618029/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
7788d320ba drm/msm/a6xx: Add a flag to allow preemption to submitqueue_create
Some userspace changes are necessary so add a flag for userspace to
advertise support for preemption when creating the submitqueue.

When this flag is not set preemption will not be allowed in the middle
of the submitted IBs therefore mantaining compatibility with older
userspace.

The flag is rejected if preemption is not supported on the target, this
allows userspace to know whether preemption is supported.

Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618028/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
35d36dc169 drm/msm/a6xx: Add traces for preemption
Add trace points corresponding to preemption being triggered and being
completed for latency measurement purposes.

Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618026/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
50117cad0c drm/msm/a6xx: Use posamble to reset counters on preemption
Use the postamble to reset perf counters when switching between rings,
except when sysprof is enabled, analogously to how they are reset
between submissions when switching pagetables.

Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618024/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
3044f928cc drm/msm/a6xx: Sync relevant adreno_pm4.xml changes
In mesa CP_SET_CTXSWITCH_IB is renamed to CP_SET_AMBLE and some other
names are changed to match KGSL. Import those changes.

The changes have not been merged yet in mesa but are necessary for this
series.

Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618023/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
e7ae83da4a drm/msm/a6xx: Implement preemption for a7xx targets
This patch implements preemption feature for A6xx targets, this allows
the GPU to switch to a higher priority ringbuffer if one is ready. A6XX
hardware as such supports multiple levels of preemption granularities,
ranging from coarse grained(ringbuffer level) to a more fine grained
such as draw-call level or a bin boundary level preemption. This patch
enables the basic preemption level, with more fine grained preemption
support to follow.

Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Sharat Masetty <smasetty@codeaurora.org>
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618021/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
91389b4e32 drm/msm/a6xx: Add a pwrup_list field to a6xx_info
Add a field to contain the pwup_reglist needed for preemption.

Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618018/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:21:52 -07:00
Antonino Maniscalco
c7546e2c3c drm/msm: Add CONTEXT_SWITCH_CNTL bitfields
Add missing bitfields to CONTEXT_SWITCH_CNTL in a6xx.xml.

Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618016/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:18:35 -07:00
Antonino Maniscalco
b9365f4114 drm/msm: Add a preempt_record_size field
Adds a field to `adreno_info` to store the GPU specific preempt record
size.

Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618015/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:18:35 -07:00
Antonino Maniscalco
3241504ea2 drm/msm/a6xx: Track current_ctx_seqno per ring
With preemption it is not enough to track the current_ctx_seqno globally
as execution might switch between rings.

This is especially problematic when current_ctx_seqno is used to
determine whether a page table switch is necessary as it might lead to
security bugs.

Track current context per ring.

Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618012/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:18:34 -07:00
Antonino Maniscalco
76a28f4c0c drm/msm: Fix bv_fence being used as bv_rptr
The bv_fence field of rbmemptrs was being used incorrectly as the BV
rptr shadow pointer in some places.

Add a bv_rptr field and change the code to use that instead.

Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Tested-by: Rob Clark <robdclark@gmail.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8650-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8550-QRD
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on SM8450-HDK
Signed-off-by: Antonino Maniscalco <antomani103@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/618010/
Signed-off-by: Rob Clark <robdclark@chromium.org>
2024-10-03 13:18:34 -07:00
Vinay Belgaumkar
491418a258
drm/xe: Restore GT freq on GSC load error
As part of a Wa_22019338487, ensure that GT freq is restored
even when GSC reload is not successful.

Fixes: 3b1592fb78 ("drm/xe/lnl: Apply Wa_22019338487")

Signed-off-by: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240925204918.1989574-1-vinay.belgaumkar@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2024-10-03 14:12:04 -04:00
Benjamin Szőke
231bb9b4c4 drm/nouveau/i2c: rename aux.c and aux.h to auxch.c and auxch.h
The goal is to clean-up Linux repository from AUX file names, because
the use of such file names is prohibited on other operating systems
such as Windows, so the Linux repository cannot be cloned and
edited on them.

Signed-off-by: Benjamin Szőke <egyszeregy@freemail.hu>
Reviewed-by: Ben Skeggs <bskeggs@nvidia.com>
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240603091558.35672-1-egyszeregy@freemail.hu
2024-10-03 20:05:16 +02:00
Uros Bizjak
b23eff812a drm/lib: Include <linux/prandom.h> instead of <linux/random.h>
Substitute the inclusion of <linux/random.h> header with
<linux/prandom.h> to allow the removal of legacy inclusion
of <linux/prandom.h> from <linux/random.h>.

Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: David Airlie <airlied@gmail.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2024-10-03 18:19:38 +02:00
Uros Bizjak
75caf39655 drm/i915/selftests: Include <linux/prandom.h> instead of <linux/random.h>
Substitute the inclusion of <linux/random.h> header with
<linux/prandom.h> to allow the removal of legacy inclusion
of <linux/prandom.h> from <linux/random.h>.

Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Tvrtko Ursulin <tursulin@ursulin.net>
Cc: David Airlie <airlied@gmail.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
2024-10-03 18:19:03 +02:00
Francois Dugast
91b2c42c21
drm/xe: Use fault injection infrastructure to find issues at probe time
The kernel fault injection infrastructure is used to test proper error
handling during probe. The return code of the functions using
ALLOW_ERROR_INJECTION() can be conditionnally modified at runtime by
tuning some debugfs entries. This requires CONFIG_FUNCTION_ERROR_INJECTION
(among others).

One way to use fault injection at probe time by making each of those
functions fail one at a time is:

    FAILTYPE=fail_function
    DEVICE="0000:00:08.0" # depends on the system
    ERRNO=-12 # -ENOMEM, can depend on the function

    echo N > /sys/kernel/debug/$FAILTYPE/task-filter
    echo 100 > /sys/kernel/debug/$FAILTYPE/probability
    echo 0 > /sys/kernel/debug/$FAILTYPE/interval
    echo -1 > /sys/kernel/debug/$FAILTYPE/times
    echo 0 > /sys/kernel/debug/$FAILTYPE/space
    echo 1 > /sys/kernel/debug/$FAILTYPE/verbose

    modprobe xe
    echo $DEVICE > /sys/bus/pci/drivers/xe/unbind

    grep -oP "^.* \[xe\]" /sys/kernel/debug/$FAILTYPE/injectable | \
    cut -d ' ' -f 1 | while read -r FUNCTION ; do
        echo "Injecting fault in $FUNCTION"
        echo "" > /sys/kernel/debug/$FAILTYPE/inject
        echo $FUNCTION > /sys/kernel/debug/$FAILTYPE/inject
        printf %#x $ERRNO > /sys/kernel/debug/$FAILTYPE/$FUNCTION/retval
        echo $DEVICE > /sys/bus/pci/drivers/xe/bind
    done

    rmmod xe

It will also be integrated into IGT for systematic execution by CI.

v2: Wrappers are not needed in the cases covered by this patch, so
    remove them and use ALLOW_ERROR_INJECTION() directly.

v3: Document the use of fault injection at probe time in xe_pci_probe
    and refer to it where ALLOW_ERROR_INJECTION() is used.

Signed-off-by: Francois Dugast <francois.dugast@intel.com>
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240927151207.399354-1-francois.dugast@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
2024-10-03 08:58:26 -04:00
Jani Nikula
de0cbc7418 drm/i915/irq: remove GEN8_IRQ_RESET_NDX() and GEN8_IRQ_INIT_NDX() macros
Define register offset triplets for all registers used with
GEN8_IRQ_RESET_NDX() and GEN8_IRQ_INIT_NDX() macros, and call the
underlying gen3_irq_reset() and gen3_irq_init() functions
directly. Remove the macros, along with the macro name concatenation
hackery.

Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241002102645.136155-3-jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 15:42:41 +03:00
Jani Nikula
7a26b3f1f6 drm/i915/irq: remove GEN3_IRQ_RESET() and GEN3_IRQ_INIT() macros
Define register offset triplets for all registers used with
GEN3_IRQ_RESET() and GEN3_IRQ_INIT() macros, and call the underlying
gen3_irq_reset() and gen3_irq_init() functions directly. Remove the
macros, along with the macro name concatenation hackery.

Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241002102645.136155-2-jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 15:42:41 +03:00
Jani Nikula
9b63562694 drm/i915/irq: add struct i915_irq_regs triplet
Add struct i915_irq_regs to hold IMR/IER/IIR register offsets to pass to
gen3_irq_reset() and gen3_irq_init(). This helps in grouping the
registers and further cleanup.

Note: gen3_irq_reset() and gen3_irq_init() really did have the
IMR/IER/IIR parameters in different order.

Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241002102645.136155-1-jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 15:42:41 +03:00
Ville Syrjälä
cade191506 drm/i915/dp: Extract intel_edp_set_sink_rates()
Declutter intel_edp_init_dpcd() a bit by extracting the sink
rates probing into its own function.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240918190441.29071-3-ville.syrjala@linux.intel.com
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
2024-10-03 14:54:04 +03:00
Ville Syrjälä
f31b2cfe62 drm/i915/dp: Make intel_dp_get_colorimetry_status() static
intel_dp_get_colorimetry_status() is not used outside of
intel_dp.c. Make it static.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240918190441.29071-2-ville.syrjala@linux.intel.com
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
2024-10-03 14:53:42 +03:00
Ville Syrjälä
e2f5812ebf drm/i915: Apply the i915gm/i945gm irq C-state w/a to CRC interrupts
Turns out CRC interrupts also fail to wake up i915gm/i945gm from
C2+. I suppose this is a generic problem, but for most other
interrupts the system will be busy enough already prior to
the irq being issued. But CRC interrupts are like vblank interrupts
and only fire once per frame, so plenty of time to fall asleep
in between them.

Apply the same core clock gating trick to CRC interrupts
that we use for vblank interrupts.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001195803.3371-5-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 14:50:25 +03:00
Ville Syrjälä
f45cc1d373 drm/i915: Extract i915gm_irq_cstate_wa_{disable,enable}()
Extract the i915gm/i945gm vblank irq C-state workaround to
separate functions. We'll need to reuse these in order to
guarantee timely CRC interrupt delivery as well.

The irq.vblank_enabled count is currently protected by the
drm vblank locks, so let's assert that the innermost of those
is held, in anticipation of other callers.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001195803.3371-4-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 14:40:30 +03:00
Ville Syrjälä
8923422ba4 drm/i915/irq: Pair up the vblank enable/disable functions
The current way of organizing all .vblank_enable() functions
before all .vblabk_disable() functions is infuriating. It's
really hard to compare the enable() vs. disable() for the
same platform to make sure they properly mirror each other.

Reorganize the functions so that the enable+disable for
the same platoform are next to each.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001195803.3371-3-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 14:40:20 +03:00
Ville Syrjälä
9075efdd96 drm/i915/irq: Nuke stale comments
Remove the tall tales about getting passed pipe indices into
the .vblank_{enable,disable}() hooks. This hasn't been true since
commit 08fa8fd0fa ("drm/i915: Switch to per-crtc vblank vfuncs").

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001195803.3371-2-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-03 14:40:10 +03:00
Matthew Auld
11bfc4a2cf drm/xe/ct: drop irq usage of xa_erase()
Unclear why disabling interrupts is needed here. Nothing seems to be
touching fence_lookup and its corresponding lock from an irq so there
should be no risk of deadlock.

Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Badal Nilawar <badal.nilawar@intel.com>
Reviewed-by: Badal Nilawar <badal.nilawar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001084346.98516-8-matthew.auld@intel.com
2024-10-03 08:34:21 +01:00
Matthew Auld
f040327238 drm/xe/guc_submit: fix xa_store() error checking
Looks like we are meant to use xa_err() to extract the error encoded in
the ptr.

Fixes: dd08ebf6c3 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Badal Nilawar <badal.nilawar@intel.com>
Cc: <stable@vger.kernel.org> # v6.8+
Reviewed-by: Badal Nilawar <badal.nilawar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001084346.98516-7-matthew.auld@intel.com
2024-10-03 08:34:20 +01:00
Matthew Auld
1aa4b78647 drm/xe/ct: fix xa_store() error checking
Looks like we are meant to use xa_err() to extract the error encoded in
the ptr.

Fixes: dd08ebf6c3 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Badal Nilawar <badal.nilawar@intel.com>
Cc: <stable@vger.kernel.org> # v6.8+
Reviewed-by: Badal Nilawar <badal.nilawar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001084346.98516-6-matthew.auld@intel.com
2024-10-03 08:34:19 +01:00