Matthew Brost [Tue, 23 Jul 2024 15:10:45 +0000 (08:10 -0700)]
drm/xe: Store process name and pid in xe file
An xe file can outlive the associated process as the GPU cleanup is just
triggered upon file close (process kill) and completes sometime later.
If the file close triggers error conditions (GPU hangs) the process
cannot be safely referenced to retrieve the name and pid for debug
information. Store the process name and pid directly in the xe file to
be safe.
v2:
- Access file->pid via rcu_access_pointer (Matthew Auld)
Fixes: b10d0c5e9df7 ("drm/xe: Add process name to devcoredump") Fixes: f6ca930d974e ("drm/xe: Add process name and PID to job timedout message") Signed-off-by: Matthew Brost <matthew.brost@intel.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240723151045.1725417-1-matthew.brost@intel.com
Matthew Brost [Tue, 23 Jul 2024 01:17:02 +0000 (18:17 -0700)]
drm/xe: Return -ENOBUFS if a kmalloc fails which is tied to an array of binds
The size of an array of binds is directly tied to several kmalloc in the
KMD, thus making these kmalloc more likely to fail. Return -ENOBUFS in
the case of these failures.
The expected UMD behavior upon returning -ENOBUFS is to split an array
of binds into a series of single binds.
Matthew Brost [Tue, 23 Jul 2024 01:02:30 +0000 (18:02 -0700)]
drm/xe: Fix xe_pt_abort_unbind
When restoring the children PT entries on a bind failure the incorrect
loop index has used resulting in PT entries being leaked. This is shown
by running xe_vm.bind-array-conflict-error-inject on a VRAM device going
into a suspend state after the test completes.
Lucas De Marchi [Fri, 19 Jul 2024 19:15:34 +0000 (12:15 -0700)]
drm/xe: Fix warning on unreachable statement
eu_type_to_str() relies on -Wswitch to warn (and -Werror) to make sure
it handles all enum values. However it's perfectly legal to pass an int
to that function so in the end that function may happen to return
nothing. There's too much implicit knowledge about the initialization
of eu_type for a compiler to notice eu_type is never assigned to
anything other than those values.
Trying to reproduce this issue, none of gcc-9, gcc-10 and gcc-13
triggered for me, but this was reported in a different system with
gcc-10:
drivers/gpu/drm/xe/xe.o: warning: objtool: xe_gt_topology_dump() falls through to next function xe_gt_topology_init()
Also it was reported these warnings when building with clang:
drivers/gpu/drm/xe/xe.o: warning: objtool: xe_gt_topology_dump+0x77: sibling call from callable instruction with modified stack frame
drivers/gpu/drm/xe/xe.o: warning: objtool: xe_gt_topology_dump() falls through to next function xe_dss_mask_group_ffs()
drivers/gpu/drm/xe/xe.o: warning: objtool: xe_gt_topology_dump+0x77: can't find jump dest instruction at .text.xe_gt_topology_dump+0xc0
Since that value is not really possible in real world, just take the
simple approach and return NULL.
Fixes: 7108b4a589cd ("drm/xe/uapi: Expose SIMD16 EU mask in topology query") Reviewed-by: Nathan Chancellor <nathan@kernel.org> Tested-by: Nathan Chancellor <nathan@kernel.org> Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240719191534.3845469-1-lucas.demarchi@intel.com Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Michal Wajdeczko [Mon, 15 Jul 2024 18:05:38 +0000 (20:05 +0200)]
drm/xe: Add NEEDS_2M BO flag
In addition of NEEDS_64K BO flag, add similar one to force 2 MiB
alignment of the buffer objects. Explicitly use this flag during
VF LMEM provisioning as LMTT uses 2 MiB pages and one day we may
drop requirement of allocating pinned objects as contiguous.
Michal Wajdeczko [Sat, 20 Jul 2024 14:25:23 +0000 (16:25 +0200)]
drm/xe/tests: Add helpers for use in live tests
Instead of iterating over available Xe devices within a testcase,
without being able to distinguish potential failures from different
devices on system with many Xe devices, introduce helpers that will
allow to treat each Xe device as a parameter for the testcase like:
Michal Wajdeczko [Sat, 20 Jul 2024 14:25:22 +0000 (16:25 +0200)]
drm/xe: Introduce const cast helper
Typically we want to preserve pointer constness when converting
from one xe pointer to another, but in some rare cases, like kunit
parameter conversions, we might want to discard this constness.
Add a helper that we will use to clearly indicate our intention.
Matthew Brost [Fri, 19 Jul 2024 17:29:05 +0000 (10:29 -0700)]
drm/xe: Build PM into GuC CT layer
Take PM ref when any G2H are outstanding, drop when none are
outstanding.
To safely ensure we have PM ref when in the GuC CT layer, a PM ref needs
to be held when scheduler messages are pending too.
v2:
- Add outer PM protections to xe_file_close (CI)
v3:
- Only take PM ref 0->1 and drop on 1->0 (Matthew Auld)
v4:
- Add assert to G2H increment function
v5:
- Rebase
v6:
- Declare xe as local variable in xe_file_close (CI)
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs") Cc: Matthew Auld <matthew.auld@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Nirmoy Das <nirmoy.das@intel.com> Signed-off-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Nirmoy Das <nirmoy.das@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240719172905.1527927-5-matthew.brost@intel.com
Matthew Brost [Fri, 19 Jul 2024 17:29:03 +0000 (10:29 -0700)]
drm/xe: Drop xe_gt_tlb_invalidation_wait
Having two methods to wait on GT TLB invalidations is not ideal. Remove
xe_gt_tlb_invalidation_wait and only use GT TLB invalidation fences.
In addition to two methods being less than ideal, once GT TLB
invalidations are coalesced the seqno cannot be assigned during
xe_gt_tlb_invalidation_ggtt/range. Thus xe_gt_tlb_invalidation_wait
would not have a seqno to wait one. A fence however can be armed and
later signaled.
v3:
- Add explaination about coalescing to commit message
v4:
- Don't put dma fence if defined on stack (CI)
v5:
- Initialize ret to zero (CI)
v6:
- Use invalidation_fence_signal helper in tlb timeout (Matthew Auld)
Michal Wajdeczko [Thu, 18 Jul 2024 20:31:55 +0000 (22:31 +0200)]
drm/xe/vf: Fix register value lookup
We should use the number of actual entries stored in the runtime
register buffer, not the maximum number of entries that this buffer
can hold, otherwise bsearch() may fail and we may miss the data and
wrongly report unexpected access to some registers.
Fixes: 4edadc41a3a4 ("drm/xe/vf: Use register values obtained from the PF") Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Piotr Piórkowski <piotr.piorkowski@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240718203155.486-1-michal.wajdeczko@intel.com
drm/xe: Fix use after free when client stats are captured
xe_file_close triggers an asynchronous queue cleanup and then frees up
the xef object. Since queue cleanup flushes all pending jobs and the KMD
stores client usage stats into the xef object after jobs are flushed, we
see a use-after-free for the xef object. Resolve this by taking a
reference to xef from xe_exec_queue.
While at it, revert an earlier change that contained a partial work
around for this issue.
v2:
- Take a ref to xef even for the VM bind queue (Matt)
- Squash patches relevant to that fix and work around (Lucas)
v3: Fix typo (Lucas)
Fixes: ce62827bc294 ("drm/xe: Do not access xe file when updating exec queue run_ticks") Fixes: 6109f24f87d7 ("drm/xe: Add helper to accumulate exec queue runtime") Closes: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1908 Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com> Reviewed-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240718210548.3580382-5-umesh.nerlige.ramappa@intel.com Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Lucas De Marchi [Wed, 10 Jul 2024 22:02:27 +0000 (15:02 -0700)]
drm/xe/uapi: Expose SIMD16 EU mask in topology query
PVC, Xe2 and later platforms have 16-wide EUs. We were implicitly
reporting for PVC the number of 16-wide EUs without giving userspace any
hint that they were different than for other platforms. Xe2 and later
also have 16-wide, but in those cases the reported number would
correspond to the 8-wide count.
To avoid confusion and make sure the right number is used by userspace
depending on the platform, add a new item to the topology query and drop
the one that is not available. The new mask reported for both PVC and
Xe2 should now match the numbers reported via hwconfig.
v2: Use a different topo item with EU type in its name to report the
new mask instead of adding the type itself as the item (Matt Roper)
Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: José Roberto de Souza <jose.souza@intel.com> Acked-by: Mateusz Jablonski <mateusz.jablonski@intel.com> Acked-by: Wenbin Lu <wenbin.lu@intel.com> Acked-by: Effie Yu <effie.yu@intel.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240710220446.2169797-1-lucas.demarchi@intel.com Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Uma Shankar [Wed, 17 Jul 2024 08:22:52 +0000 (13:52 +0530)]
drm/xe/fbdev: Limit the usage of stolen for LNL+
As per recommendation in the workarounds:
WA_22019338487
There is an issue with accessing Stolen memory pages due a
hardware limitation. Limit the usage of stolen memory for
fbdev for LNL+. Don't use BIOS FB from stolen on LNL+ and
assign the same from system memory.
v2: Corrected the WA Number, limited WA to LNL and
Adopted XE_WA framework as suggested by Lucas and Matt.
v3: Introduced the waxxx_display to implement display side
of WA changes on Lunarlake. Used xe_root_mmio_gt and
avoid the for loop (Suggested by Lucas)
In xe2+ dgfx, we don't need to handle the copying of ccs
metadata during migration. This test validates the ccs data post
clear and copy during evict/restore operation. Thus, we can skip
this test on xe2+ dgfx.
drm/xe/migrate: Add kunit to test migration functionality for BMG
This part of kunit verifies that
- main data is decompressed and ccs data is clear post bo eviction.
- main data is raw copied and ccs data is clear post bo restore.
drm/xe/xe_migrate: Handle migration logic for xe2+ dgfx
During eviction (vram->sysmem), we use compressed -> uncompressed mapping.
During restore (sysmem->vram), we need to use mapping from
uncompressed -> uncompressed.
Handle logic for selecting the compressed identity map for eviction,
and selecting uncompressed map for restore operations.
v2: Move check of xe_migrate_ccs_emit() before calling
xe_migrate_ccs_copy(). (Nirmoy)
drm/xe/xe2: Introduce identity map for compressed pat for vram
Xe2+ has unified compression (exactly one compression mode/format),
where compression is now controlled via PAT at PTE level.
This simplifies KMD operations, as it can now decompress freely
without concern for the buffer's original compression format—unlike DG2,
which had multiple compression formats and thus required copying the
raw CCS state during VRAM eviction. In addition mixed VRAM and system
memory buffers were not supported with compression enabled.
On Xe2 dGPU compression is still only supported with VRAM, however we
can now support compression with VRAM and system memory buffers,
with GPU access being seamless underneath. So long as when doing
VRAM -> system memory the KMD uses compressed -> uncompressed,
to decompress it. This also allows CPU access to such buffers,
assuming that userspace first decompress the corresponding
pages being accessed.
If the pages are already in system memory then KMD would have already
decompressed them. When restoring such buffers with sysmem -> VRAM
the KMD can't easily know which pages were originally compressed,
so we always use uncompressed -> uncompressed here.
With this it also means we can drop all the raw CCS handling on such
platforms (including needing to allocate extra CCS storage).
In order to support this we now need to have two different identity
mappings for compressed and uncompressed VRAM.
In this patch, we set up the additional identity map for the VRAM with
compressed pat_index. We then select the appropriate mapping during
migration/clear. During eviction (vram->sysmem), we use the mapping
from compressed -> uncompressed. During restore (sysmem->vram), we need
the mapping from uncompressed -> uncompressed.
Therefore, we need to have two different mappings for compressed and
uncompressed vram. We set up an additional identity map for the vram
with compressed pat_index.
We then select the appropriate mapping during migration/clear.
v2: Formatting nits, Updated code to match recent changes in
xe_migrate_prepare_vm(). (Matt)
v3: Move identity map loop to a helper function. (Matt Brost)
v4: Split helper function in different patch, and
add asserts and nits. (Matt Brost)
v5: Convert the 2 bool arguments of pte_update_size to flags
argument (Matt Brost)
drm/xe/migrate: Add kunit to test clear functionality
This test verifies if the main and ccs data are cleared during bo creation.
The motivation to use Kunit instead of IGT is that, although we can verify
whether the data is zero following bo creation,
we cannot confirm whether the zero value after bo creation is the result of
our clear function or simply because the initial data present was zero.
v2: Updated the mutex_lock and unlock logic,
Changed out_unlock to out_put. (Matt)
drm/xe/migrate: Handle clear ccs logic for xe2 dgfx
For Xe2 dGPU, we clear the bo by modifying the VRAM using an
uncompressed pat index which then indirectly updates the
compression status as uncompressed i.e zeroed CCS.
So xe_migrate_clear() should be updated for BMG to not
emit CCS surf copy commands.
v2: Moved xe_device_needs_ccs_emit() to xe_migrate.c and changed
name to xe_migrate_needs_ccs_emit() since its very specific to
migration.(Matt)
Matthew Brost [Tue, 16 Jul 2024 06:39:01 +0000 (23:39 -0700)]
drm/xe: Wedge the entire device
Wedge the entire device, not just GT which may have triggered the wedge.
To implement this, cleanup the layering so xe_device_declare_wedged()
calls into the lower layers (GT) to ensure entire device is wedged.
While we are here, also signal any pending GT TLB invalidations upon
wedging device.
Lastly, short circuit reset wait if device is wedged.
v2:
- Short circuit reset wait if device is wedged (Local testing)
Michal Wajdeczko [Sat, 13 Jul 2024 14:26:43 +0000 (16:26 +0200)]
drm/xe/vf: Track writes to inaccessible registers from VF
Only limited set of registers is accessible for the VF driver and
the hardware will silently drop writes to inaccessible registers.
To improve our VF driver lets intercept all such writes to warn
about such unexpected writes on debug builds or optionally allow
to provide some substitution (as a potential future extension).
Wa_15015404425 asks us to perform four "dummy" writes to a
non-existent register offset before every real register read.
Although the specific offset of the writes doesn't directly
matter, the workaround suggests offset 0x130030 as a good target
so that these writes will be easy to recognize and filter out in
debugging traces.
V5(MattR):
- Avoid negating an equality comparison
V4(MattR):
- Use writel and remove xe_reg usage
V3(MattR):
- Define dummy reg local to function
- Avoid tracing dummy writes
- Update commit message
V2:
- Add WA to 8/16/32bit reads also - MattR
- Corrected dummy reg address - MattR
- Use for loop to avoid mental pause - JaniN
Michal Wajdeczko [Thu, 11 Jul 2024 19:23:19 +0000 (21:23 +0200)]
drm/xe/pf: Limit fair VF LMEM provisioning
Due to the current design of the BO and VRAM manager, any object
with XE_BO_FLAG_PINNED flag, which the PF driver uses during VF
LMEM provisionining, is created with the TTM_PL_FLAG_CONTIGUOUS
flag, which may cause VRAM fragmentation that prevents subsequent
allocations of larger objects, like fair VF LMEM provisioning.
To avoid such failures, round down fair VF LMEM provisioning size
to next power of two size, to compensate what xe_ttm_vram_mgr is
doing to achieve contiguous allocations.
Fixes: ac6598aed1b3 ("drm/xe/pf: Add support to configure SR-IOV VFs") Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Reviewed-by: Piotr Piórkowski <piotr.piorkowski@intel.com> Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240711192320.1198-2-michal.wajdeczko@intel.com Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Michal Wajdeczko [Mon, 8 Jul 2024 11:12:10 +0000 (13:12 +0200)]
drm/xe/kunit: Simplify xe_mocs live tests code layout
The test case logic is implemented by the functions compiled as
part of the core Xe driver module and then exported to build and
register the test suite in the live test module.
But we don't need to export individual test case functions, we may
just export the entire test suite. And we don't need to register
this test suite in a separate file, it can be done in the main
file of the live test module.
Michal Wajdeczko [Mon, 8 Jul 2024 11:12:09 +0000 (13:12 +0200)]
drm/xe/kunit: Simplify xe_migrate live tests code layout
The test case logic is implemented by the functions compiled as
part of the core Xe driver module and then exported to build and
register the test suite in the live test module.
But we don't need to export individual test case functions, we may
just export the entire test suite. And we don't need to register
this test suite in a separate file, it can be done in the main
file of the live test module.
Michal Wajdeczko [Mon, 8 Jul 2024 11:12:08 +0000 (13:12 +0200)]
drm/xe/kunit: Simplify xe_dma_buf live tests code layout
The test case logic is implemented by the functions compiled as
part of the core Xe driver module and then exported to build and
register the test suite in the live test module.
But we don't need to export individual test case functions, we may
just export the entire test suite. And we don't need to register
this test suite in a separate file, it can be done in the main
file of the live test module.
Michal Wajdeczko [Mon, 8 Jul 2024 11:12:07 +0000 (13:12 +0200)]
drm/xe/kunit: Simplify xe_bo live tests code layout
The test case logic is implemented by the functions compiled as
part of the core Xe driver module and then exported to build and
register the test suite in the live test module.
But we don't need to export individual test case functions, we may
just export the entire test suite. And we don't need to register
this test suite in a separate file, it can be done in the main
file of the live test module.
Lucas De Marchi [Mon, 8 Jul 2024 21:29:06 +0000 (14:29 -0700)]
drm/xe: Generate oob before compiling anything
Instead of keep adding more dependencies as WAs are needed in different
places of the driver, just add a rule with all the objects so the code
generation happens before anything else.
While at it, group lines related to wa_oob in the Makefile.
Matthew Brost [Mon, 8 Jul 2024 21:10:08 +0000 (14:10 -0700)]
drm/xe: Drop trace_xe_hw_fence_free
fence->ctx may be stale memory when trace_xe_hw_fence_free is called
resuling UAF bug when deriving the device name. This tracepoint is not
all that useful, so just drop it.
Fixes: 501c4255c409 ("drm/xe/trace: Print device_id in xe_trace events") Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Gustavo Sousa <gustavo.sousa@intel.com> Cc: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240708211008.956384-1-matthew.brost@intel.com
Thomas Hellström [Fri, 5 Jul 2024 13:28:28 +0000 (15:28 +0200)]
drm/xe: Use write-back caching mode for system memory on DGFX
The caching mode for buffer objects with VRAM as a possible
placement was forced to write-combined, regardless of placement.
However, write-combined system memory is expensive to allocate and
even though it is pooled, the pool is expensive to shrink, since
it involves global CPU TLB flushes.
Moreover write-combined system memory from TTM is only reliably
available on x86 and DGFX doesn't have an x86 restriction.
So regardless of the cpu caching mode selected for a bo,
internally use write-back caching mode for system memory on DGFX.
Coherency is maintained, but user-space clients may perceive a
difference in cpu access speeds.
v2:
- Update RB- and Ack tags.
- Rephrase wording in xe_drm.h (Matt Roper)
v3:
- Really rephrase wording.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Fixes: 622f709ca629 ("drm/xe/uapi: Add support for CPU caching mode") Cc: Pallavi Mishra <pallavi.mishra@intel.com> Cc: Matthew Auld <matthew.auld@intel.com> Cc: dri-devel@lists.freedesktop.org Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: Effie Yu <effie.yu@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Jose Souza <jose.souza@intel.com> Cc: Michal Mrozek <michal.mrozek@intel.com> Cc: <stable@vger.kernel.org> # v6.8+ Acked-by: Matthew Auld <matthew.auld@intel.com> Acked-by: José Roberto de Souza <jose.souza@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Fixes: 622f709ca629 ("drm/xe/uapi: Add support for CPU caching mode") Acked-by: Michal Mrozek <michal.mrozek@intel.com> Acked-by: Effie Yu <effie.yu@intel.com> #On chat Link: https://patchwork.freedesktop.org/patch/msgid/20240705132828.27714-1-thomas.hellstrom@linux.intel.com
Matthew Auld [Wed, 3 Jul 2024 12:43:39 +0000 (13:43 +0100)]
drm/i915: disable fbc due to Wa_16023588340
On BMG-G21 we need to disable fbc due to complications around the WA.
v2:
- Try to handle with i915_drv.h and compat layer. (Rodrigo)
v3:
- For simplicity retreat back to the original design for now.
- Drop the extra \ from the Makefile (Jani)
Signed-off-by: Matthew Auld <matthew.auld@intel.com> Cc: Jonathan Cavitt <jonathan.cavitt@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Vinod Govindapillai <vinod.govindapillai@intel.com> Cc: Jani Nikula <jani.nikula@intel.com> Cc: intel-gfx@lists.freedesktop.org Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240703124338.208220-4-matthew.auld@intel.com
Matthew Auld [Wed, 3 Jul 2024 12:43:38 +0000 (13:43 +0100)]
drm/xe/bmg: implement Wa_16023588340
This involves enabling l2 caching of host side memory access to VRAM
through the CPU BAR. The main fallout here is with display since VRAM
writes from CPU can now be cached in GPU l2, and display is never
coherent with caches, so needs various manual flushing. In the case of
fbc we disable it due to complications in getting this to work
correctly (in a later patch).
Signed-off-by: Matthew Auld <matthew.auld@intel.com> Cc: Jonathan Cavitt <jonathan.cavitt@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240703124338.208220-3-matthew.auld@intel.com
Michal Wajdeczko [Tue, 2 Jul 2024 18:37:04 +0000 (20:37 +0200)]
drm/xe: Use VF_CAP_REG for device wmb
To force a write barrier on the device memory, we write to the
SOFTWARE_FLAGS_SPR33 register, but this particular register was
selected because it was one of the writable and unused register.
Since a write barrier should also work if we use the read-only
register, switch to VF_CAP_REG register that is also marked as
accessible for VFs.
While at it, add simple kernel-doc for xe_device_wmb() function.
Matthew Brost [Thu, 4 Jul 2024 04:16:52 +0000 (21:16 -0700)]
drm/xe: Add VM bind IOCTL error injection
Add VM bind IOCTL error injection which steals MSB of the bind flags
field which if set injects errors at various points in the VM bind
IOCTL. Intended to validate error paths. Enabled by CONFIG_DRM_XE_DEBUG.
Matthew Brost [Thu, 4 Jul 2024 04:16:49 +0000 (21:16 -0700)]
drm/xe: Convert multiple bind ops into single job
This aligns with the uAPI of an array of binds or single bind that
results in multiple GPUVA ops to be considered a single atomic
operations.
The design is roughly:
- xe_vma_ops is a list of xe_vma_op (GPUVA op)
- each xe_vma_op resolves to 0-3 PT ops
- xe_vma_ops creates a single job
- if at any point during binding a failure occurs, xe_vma_ops contains
the information necessary unwind the PT and VMA (GPUVA) state
v2:
- add missing dma-resv slot reservation (CI, testing)
v4:
- Fix TLB invalidation (Paulo)
- Add missing xe_sched_job_last_fence_add/test_dep check (Inspection)
v5:
- Invert i, j usage (Matthew Auld)
- Add helper to test and add job dep (Matthew Auld)
- Return on anything but -ETIME for cpu bind (Matthew Auld)
- Return -ENOBUFS if suballoc of BB fails due to size (Matthew Auld)
- s/do/Do (Matthew Auld)
- Add missing comma (Matthew Auld)
- Do not assign return value to xe_range_fence_insert (Matthew Auld)
v6:
- s/0x1ff/MAX_PTE_PER_SDI (Matthew Auld, CI)
- Check to large of SA in Xe to avoid triggering WARN (Matthew Auld)
- Fix checkpatch issues
v7:
- Rebase
- Support more than 510 PTEs updates in a bind job (Paulo, mesa testing)
v8:
- Rebase
Matthew Brost [Thu, 4 Jul 2024 04:16:47 +0000 (21:16 -0700)]
drm/xe: Add xe_vm_pgtable_update_op to xe_vma_ops
Each xe_vma_op resolves to 0-3 pt_ops. Add storage for the pt_ops to
xe_vma_ops which is dynamically allocated based the number and types of
xe_vma_op in the xe_vma_ops list. Allocation only implemented in this
patch.
This will help with converting xe_vma_ops (multiple xe_vma_op) in a
atomic update unit.
Ashutosh Dixit [Wed, 3 Jul 2024 16:48:01 +0000 (09:48 -0700)]
drm/xe/uapi: Rename xe perf layer as xe observation layer
In Xe, the perf layer allows capture of HW counter streams. These HW
counters are generally performance related but don't have to be necessarily
so. Also, the name "perf" is a carryover from i915 and is not preferred.
Here we propose the name "observation" for this common layer which allows
capture of different types of these counter streams.
v2: Rename observability layer to observation layer (Lucas/Rodrigo)
v3: Rename sysctl file to "observation_paranoid" (Jose)
Fixes: 52c2e956dceb ("drm/xe/perf/uapi: "Perf" layer to support multiple perf counter stream types") Fixes: fe8929bdf835 ("drm/xe/perf/uapi: Add perf_stream_paranoid sysctl") Acked-by: Lucas De Marchi <lucas.demarchi@intel.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com> Acked-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240703164801.2561423-1-ashutosh.dixit@intel.com
Matthew Brost [Wed, 26 Jun 2024 00:41:37 +0000 (17:41 -0700)]
drm/xe: Add timeout to preempt fences
To adhere to dma fencing rules that fences must signal within a
reasonable amount of time, add a 5 second timeout to preempt fences. If
this timeout occurs, kill the associated VM as this fatal to the VM.
v2:
- Add comment for smp_wmb (Checkpatch)
- Fix kernel doc typo (Inspection)
- Add comment for killed check (Niranjana)
v3:
- Drop smp_wmb (Matthew Auld)
- Don't take vm->lock in preempt fence worker (Matthew Auld)
- Drop RB given changes to patch
v4:
- Add WRITE/READ_ONCE (Niranjana)
- Don't export xe_vm_kill (Niranjana)
Cc: Matthew Auld <matthew.auld@intel.com> Cc: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com> Signed-off-by: Matthew Brost <matthew.brost@intel.com> Tested-by: Stuart Summers <stuart.summers@intel.com> Reviewed-by: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240626004137.4060806-1-matthew.brost@intel.com
Michal Wajdeczko [Mon, 1 Jul 2024 19:30:30 +0000 (21:30 +0200)]
drm/xe/guc: Demote GuC IDs usage message to debug
Printing message at INFO level about available GuC IDs is not that
important, DEBUG level is enough. It will also match message about
available doorbells:
Extend this WA to BMG GT as well. In this case media GT is
not affected. The cap frequencies and max allowed ggtt writes
are different as well. On BMG, we need to do a flush after 1100
GGTT writes, and we need to limit the GT frequency request
to 2133 Mhz during driver load and leave it at that value after
driver unloads.
Lucas De Marchi [Fri, 28 Jun 2024 16:17:26 +0000 (09:17 -0700)]
drm/xe/rtp: Fix out-of-bounds array access
Increment the counter before checking for number of rules, otherwise
when there's no XE_RTP_MATCH_OR an out-of-bounds access is done, as
reported by kasan:
BUG: KASAN: global-out-of-bounds in rule_matches+0xb6d/0x11c0 [xe]
Read of size 1 at addr ffffffffa0a50b70 by task systemd-udevd/243
Michal Wajdeczko [Mon, 1 Jul 2024 10:27:37 +0000 (12:27 +0200)]
drm/xe/pf: Restart VFs provisioning after GT reset
Any prior configurations pushed to the GuC are lost when the GT
is reset. Push again all non-empty VF configurations to the GuC
as part of the GuC reset procedure.
This will also help restore early manual provisioning, when the
PF was in the meantime suspended and then resumed.
Michal Wajdeczko [Mon, 1 Jul 2024 10:27:36 +0000 (12:27 +0200)]
drm/xe/pf: Skip fair VFs provisioning if already provisioned
Our debugfs allows to view and change VFs' provisioning configs.
If we attempt to experiment with VFs provisioning before enabling
them, this early config will affect fair provisioning calculations,
and will also be overwritten, which is undesirable behavior.
To improve this, check if the VFs configs are empty (unprovisioned)
before starting the fair provisioning procedure.
Nirmoy Das [Fri, 28 Jun 2024 08:58:45 +0000 (10:58 +0200)]
drm/xe/guc: Configure TLB timeout based on CT buffer size
GuC TLB invalidation depends on GuC to process the request from the CT
queue and then the real time to invalidate TLB. Add a function to return
overestimated possible time a TLB inval H2G might take which can be used
as timeout value for TLB invalidation wait time.
v4: Make sure CTB is in 4K blocks(Michal) and other doc fixes
v3: Pass CT to xe_guc_ct_queue_proc_time_jiffies() (Michal)
Add tlb_timeout_jiffies() that replaces TLB_TIMEOUT(Michal)
v2: Address reviews from Michal.
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/1622 Cc: Matthew Brost <matthew.brost@intel.com> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Suggested-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Acked-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240628085845.2369-1-nirmoy.das@intel.com Signed-off-by: Nirmoy Das <nirmoy.das@intel.com>
Matt Roper [Wed, 26 Jun 2024 21:05:37 +0000 (14:05 -0700)]
drm/xe/mcr: Avoid clobbering DSS steering
A couple copy/paste mistakes in the code that selects steering targets
for OADDRM and INSTANCE0 unintentionally clobbered the steering target
for DSS ranges in some cases.
The OADDRM/INSTANCE0 values were also not assigned as intended, although
that mistake wound up being harmless since the desired values for those
specific ranges were '0' which the kzalloc of the GT structure should
have already taken care of implicitly.
Matt Roper [Thu, 27 Jun 2024 20:37:43 +0000 (13:37 -0700)]
drm/xe/mocs: Clarify difference between hw and sw sizes
It's not very obvious what the difference is between the 'size' and
'n_entries' fields of the MOCS structure. Rename both fields slightly
and add some comments explaining that one is the documentation-defined
table size, while the other is the number of entries that can be
programmed into the hardware (and the documented table size can
potentially be smaller than the number of hardware entries).
Matt Roper [Thu, 27 Jun 2024 20:37:42 +0000 (13:37 -0700)]
drm/xe/mocs: Update MOCS assertions and remove redundant checks
Rely more heavily on assertions to describe the MOCS programming
invariants. CI checks these assertions and will ensure no violations
sneak in due to programmer error, so we can remove some of the redundant
WARN and silent return checks from non-debug builds.
Also tweak/augment some of the existing assertions: there's no reason
we'd ever want a platform not to have a MOCS 'ops' structure hooked up
so ensure info->ops is non-NULL. Likewise, we should never have a case
where the bspec-defined MOCS setting table is larger than the number of
MOCS registers exposed by the hardware, so add an extra assert on those
sizes as well.
drm/xe: Get hwe domain specific FW to read RING_TIMESTAMP
Per client engine utilization uses RING_TIMESTAMP to return
drm-total-cycles to the user. Current code uses XE_FW_GT to read this
register on the first available engine in a GT. When testing on DG2, it
is observed that this value is 0 when running test on some engines. To
resolve that, get the hwe domain specific FW for reading the engine
timestamp.
v2:
- update commit message
- use domain specific FW (Matt)
v3:
- Drop check for hwe in the helper (Matt, Michal)
Nirmoy Das [Tue, 25 Jun 2024 09:42:28 +0000 (11:42 +0200)]
drm/xe/client: Check return value of xe_force_wake_get
xe_force_wake_get() can return error so check it's return value
before reading gpu_timestamp value.
v2: set HWE to NULL instead of setting timestamp to 0(Lucas)
Add a warn on for xe_force_wake_put(Himal)
Fixes: 188ced1e0ff8 ("drm/xe/client: Print runtime to fdinfo") Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Nirmoy Das <nirmoy.das@intel.com> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240625094228.5327-1-nirmoy.das@intel.com Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Matthew Auld [Thu, 20 Jun 2024 10:20:26 +0000 (11:20 +0100)]
drm/xe: fix error handling in xe_migrate_update_pgtables
Don't call drm_suballoc_free with sa_bo pointing to PTR_ERR.
References: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/2120 Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs") Signed-off-by: Matthew Auld <matthew.auld@intel.com> Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Cc: <stable@vger.kernel.org> # v6.8+ Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240620102025.127699-2-matthew.auld@intel.com
Ashutosh Dixit [Wed, 26 Jun 2024 18:18:17 +0000 (11:18 -0700)]
drm/xe/oa/uapi: Allow preemption to be disabled on the stream exec queue
Mesa VK_KHR_performance_query use case requires preemption and timeslicing
to be disabled for the stream exec queue. Implement this functionality
here.
v2: Minor change to debug print to print both ret values (Umesh)
Ashutosh Dixit [Wed, 26 Jun 2024 18:18:16 +0000 (11:18 -0700)]
drm/xe/oa: Allow stream enable/disable functions to return error
Stream enable/disable functions previously had void return because failure
during function execution was not possible. This will change when we
introduce functionality to disable preemption on the stream exec
queue. Therefore, in preparation for this functionality, prepare this code
to be able to handle error returns.
Ilia Levi [Thu, 6 Jun 2024 12:47:05 +0000 (15:47 +0300)]
drm/xe/irq: remove xe_irq_shutdown
The cleanup is done by devres in irq_uninstall.
Commit bbc9651fe9f4 ("drm/xe/irq: move irq_uninstall over to devm")
resolved the ordering issue where irq_uninstall (registered with drmm)
was called after pci_free_irq_vectors (registered with devm upon calling
pci_alloc_irq_vectors). This happened because drmm action list is
registered with devm very early in the init flow - before
pci_alloc_irq_vectors.
Now that irq_uninstall is registered with devm, it will be called before
pci_free_irq_vectors and we can remove xe_irq_shutdown.
Michal Wajdeczko [Tue, 25 Jun 2024 19:45:46 +0000 (21:45 +0200)]
drm/xe/pf: Trigger explicit FLR while disabling VFs
We attempt to unprovision all VFs GuC when disabling them, but
GuC may reject such request if the target VF was previously active
but VF driver didn't unload with explicit VF reset H2G action or
the VMM has not started the VF FLR.
To avoid mismatches between configs maintained the PF and GuC,
trigger an explicit FLR sequences just before releasing resources.
Michal Wajdeczko [Tue, 25 Jun 2024 14:12:57 +0000 (16:12 +0200)]
drm/xe/guc: Add more GuC error codes to ABI
There are many more error codes used that the GuC firmware can
return in the RESPONSE_FAILURE message. Add to the ABI header
those which are more likely to be seen by the PF or VF drivers.
Michal Wajdeczko [Tue, 25 Jun 2024 14:12:56 +0000 (16:12 +0200)]
drm/xe/guc: Demote the H2G retry log message to debug
The G2H RETRY message sent by the GuC does not necessary indicate
any serious problem and can be a part of the normal communication
flow. Switch the log level from warning to more appropriate debug.
This will also let the CI ignore these logs which were seen in few
SR-IOV scenarios.
While at it, use hex to print the reason and add missing \n.
Michal Wajdeczko [Sat, 22 Jun 2024 09:42:53 +0000 (11:42 +0200)]
drm/xe/vf: Skip attempt to start GuC PC if VF
We have already marked the GuC PC feature as not applicable for
VF devices, but we missed the fact that there may be still some
privileged activities performed by this component, who does much
more than its name suggests.
Explicitly skip xe_guc_pc_start() if running as a VF driver and
use a GT oriented message to report any error.
Vinay Belgaumkar [Thu, 20 Jun 2024 22:49:27 +0000 (15:49 -0700)]
drm/xe/lnl: Apply Wa_22019338487
This WA requires us to limit media GT frequency requests to a certain
cap value during driver load. Freq limits are restored after load
completes, so perf will not be affected during normal operations.
During normal driver operation, this WA requires dummy writes to media
offset 0x380D8C after every ~63 GGTT writes. This will ensure completion
of the LMEM writes originating from Gunit.
During driver unload(before FLR), the WA requires that we set requested
frequency to the cap value again.
v3: Do not use WA number in function name. Call WA wrapper from xe_device.
Rename some variables, check for locks in the correct function (Rodrigo).
Ensure reset path is also covered for this WA.
v4: Fix BAT failure
v5: Add a function pointer for ggtt_ops (Michal W)
v6: Fix name collision and use static function (Rodrigo)
projects / users / jedix / linux-maple.git / log