With commit 89773b8559
("drm/amdkfd: Let VRAM allocations go to GTT domain on small APUs")
big and small APU "VRAM" handling in KFD was unified. Since AMD_IS_APU
is set for both big and small APUs, we can simplify the checks in
the code.
v2: clean up a few more places (Lang)
Acked-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 28ebbb4981.
Revert this commit as apparently the LLVM code to take advantage of
this never landed.
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Feifei Xu <feifei.xu@amd.com>
nouveau:
- fix bo metadata uAPI for vm bind
panthor:
- Fixes for panthor's heap logical block.
- Reset on unrecoverable fault
- Fix VM references.
- Reset fix.
xlnx:
- xlnx compile and doc fixes.
amdgpu:
- Handle vbios table integrated info v2.3
amdkfd:
- Handle duplicate BOs in reserve_bo_and_cond_vms
- Handle memory limitations on small APUs
dp/mst:
- MST null deref fix.
bridge:
- Don't let next bridge create connector in adv7511 to make probe work.
-----BEGIN PGP SIGNATURE-----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=pexB
-----END PGP SIGNATURE-----
Merge tag 'drm-next-2024-05-25' of https://gitlab.freedesktop.org/drm/kernel
Pull drm fixes from Dave Airlie:
"Some fixes for the end of the merge window, mostly amdgpu and panthor,
with one nouveau uAPI change that fixes a bad decision we made a few
months back.
nouveau:
- fix bo metadata uAPI for vm bind
panthor:
- Fixes for panthor's heap logical block.
- Reset on unrecoverable fault
- Fix VM references.
- Reset fix.
xlnx:
- xlnx compile and doc fixes.
amdgpu:
- Handle vbios table integrated info v2.3
amdkfd:
- Handle duplicate BOs in reserve_bo_and_cond_vms
- Handle memory limitations on small APUs
dp/mst:
- MST null deref fix.
bridge:
- Don't let next bridge create connector in adv7511 to make probe
work"
* tag 'drm-next-2024-05-25' of https://gitlab.freedesktop.org/drm/kernel:
drm/amdgpu/atomfirmware: add intergrated info v2.3 table
drm/mst: Fix NULL pointer dereference at drm_dp_add_payload_part2
drm/amdkfd: Let VRAM allocations go to GTT domain on small APUs
drm/amdkfd: handle duplicate BOs in reserve_bo_and_cond_vms
drm/bridge: adv7511: Attach next bridge without creating connector
drm/buddy: Fix the warn on's during force merge
drm/nouveau: use tile_mode and pte_kind for VM_BIND bo allocations
drm/panthor: Call panthor_sched_post_reset() even if the reset failed
drm/panthor: Reset the FW VM to NULL on unplug
drm/panthor: Keep a ref to the VM at the panthor_kernel_bo level
drm/panthor: Force an immediate reset on unrecoverable faults
drm/panthor: Document drm_panthor_tiler_heap_destroy::handle validity constraints
drm/panthor: Fix an off-by-one in the heap context retrieval logic
drm/panthor: Relax the constraints on the tiler chunk size
drm/panthor: Make sure the tiler initial/max chunks are consistent
drm/panthor: Fix tiler OOM handling to allow incremental rendering
drm: xlnx: zynqmp_dpsub: Fix compilation error
drm: xlnx: zynqmp_dpsub: Fix few function comments
Small APUs(i.e., consumer, embedded products) usually have a small
carveout device memory which can't satisfy most compute workloads
memory allocation requirements.
We can't even run a Basic MNIST Example with a default 512MB carveout.
https://github.com/pytorch/examples/tree/main/mnist. Error Log:
"torch.cuda.OutOfMemoryError: HIP out of memory. Tried to allocate
84.00 MiB. GPU 0 has a total capacity of 512.00 MiB of which 0 bytes
is free. Of the allocated memory 103.83 MiB is allocated by PyTorch,
and 22.17 MiB is reserved by PyTorch but unallocated"
Though we can change BIOS settings to enlarge carveout size,
which is inflexible and may bring complaint. On the other hand,
the memory resource can't be effectively used between host and device.
The solution is MI300A approach, i.e., let VRAM allocations go to GTT.
Then device and host can flexibly and effectively share memory resource.
v2: Report local_mem_size_private as 0. (Felix)
Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
new drivers:
- panthor: ARM Mali/Immortalis CSF-based GPU driver
core:
- add a CONFIG_DRM_WERROR option
- make more headers self-contained
- grab resv lock in pin/unpin
- fix vmap resv locking
- EDID/eDP panel matching
- Kconfig cleanups
- DT sound bindings
- Add SIZE_HINTS property for cursor planes
- Add struct drm_edid_product_id and helpers.
- Use drm device based logging in more drm functions.
- drop seq_file.h from a bunch of places
- use drm_edid driver conversions
dp:
- DP Tunnel documentation
- MST read sideband cap
- Adaptive sync SDP prep work
ttm:
- improve placement for TTM BOs in idle/busy handling
panic:
- Fixes for drm-panic, and option to test it.
- Add drm panic to simpledrm, mgag200, imx, ast
bridge:
- improve init ordering
- adv7511: allow GPIO pin sharing
- tc358775: add tc358675 support
panel:
- AUO B120XAN01.0
- Samsung s6e3fa7
- BOE NT116WHM-N44
- CMN N116BCA-EA1,
- CrystalClear CMT430B19N00
- Startek KD050HDFIA020-C020A
- powertip PH128800T006-ZHC01
- Innolux G121X1-L03
- LG sw43408
- Khadas TS050 V2
- EDO RM69380 OLED
- CSOT MNB601LS1-1
amdgpu:
- HDCP/ODM/RAS fixes
- Devcoredump improvements
- Expose VCN activity via sysfs
- SMY 13.0.x updates
- Enable fast updates on DCN 3.1.4
- Add dclk and vclk reporting on additional devices
- Add ACA RAS infrastructure
- Implement TLB flush fence
- EEPROM handling fixes
- SMUIO 14.0.2 support
- SMU 14.0.1 Updates
- SMU 14.0.2 support
- Sync page table freeing with TLB flushes
- DML2 refactor
- DC debug improvements
- DCN 3.5.x Updates
- GPU reset fixes
- HDP fix for second GFX pipe on GC 10.x
- Enable secondary GFX pipe on GC 10.3
- Refactor and clean up BACO/BOCO/BAMACO handling
- Remove invalid TTM resource start check
- UAF fix in VA IOCTL
- GPUVM page fault redirection to secondary IH rings for IH 6.x
- Initial support for mapping kernel queues via MES
- Fix VRAM memory accounting
amdkfd:
- MQD handling cleanup
- Preemption handling fixes for XCDs
- TLB flush fix for GC 9.4.2
- Properly clean up workqueue during module unload
- Fix memory leak process create failure
- Range check CP bad op exception targets to avoid reporting invalid exceptions to userspace
- Fix eviction fence handling
- Fix leak in GPU memory allocation failure case
- DMABuf import handling fix
- Enable SQ watchpoint for gfx10
i915:
- Adding new DG2 PCI ID
- add context hints for GT frequency
- enable only one CCS for compute workloads
- new workarounds
- Fix UAF on destroy against retire race and remove two earlier partial fixes
- Limit the reserved VM space to only the platforms that need it
- Fix gt reset with GuC submission is disable
- Add and use gt_to_guc() wrapper
i915/xe display:
- Lunar Lake display enabling, including cdclk and other refactors
- BIOS/VBT/opregion related refactor
- Digital port related refactor/clean-up
- Fix 2s boot time regression on DP panel replay init
- Remove duplication on audio enable/disable on SDVO and g4x+ DP
- Disable AuxCCS framebuffers if built for Xe
- Make crtc disable more atomic
- Increase DP idle pattern wait timeout to 2ms
- Start using container_of_const() for some extra const safety
- Fix Jasper Lake boot freeze
- Enable MST mode for 128b/132b single-stream sideband
- Enable Adaptive Sync SDP Support for DP
- Fix MTL supported DP rates - removal of UHBR13.5
- PLL refactoring
- Limit eDP MSO pipe only for display version 20
- More display refactor towards independence from i915 dev_priv
- Convert i915/xe fbdev to DRM client
- More initial work to make display code more independent from i915
xe:
- improved error capture
- clean up some uAPI leftovers
- devcoredump update
- Add BMG mocs table
- Handle GSCCS ER interrupt
- Implement xe2- and GuC workarounds
- struct xe_device cleanup
- Hwmon updates
- Add LRC parsing for more GPU instruction
- Increase VM_BIND number of per-ioctl Ops
- drm/xe: Add XE_BO_GGTT_INVALIDATE flag
- Initial development for SR-IOV support
- Add new PCI IDs to DG2 platform
- Move userptr over to start using hmm_range_fault
msm:
- Switched to generating register header files during build process
instead of shipping pre-generated headers
- Merged DPU and MDP4 format databases.
- DP:
- Stop using compat string to distinguish DP and eDP cases
- Added support for X Elite platform (X1E80100)
- Reworked DP aux/audio support
- Added SM6350 DP to the bindings
- GPU:
- a7xx perfcntr reg fixes
- MAINTAINERS updates
- a750 devcoredump support
radeon:
- Silence UBSAN warnings related to flexible arrays
nouveau:
- move some uAPI objects to uapi headers
omapdrm:
- console fix
ast:
- add i2c polling
qaic:
- add debugfs entries
exynos:
- fix platform_driver .owner
- drop cleanup code
mediatek:
- Use devm_platform_get_and_ioremap_resource() in mtk_hdmi_ddc_probe()
- Add GAMMA 12-bit LUT support for MT8188
- Rename mtk_drm_* to mtk_*
- Drop driver owner initialization
- Correct calculation formula of PHY Timing
-----BEGIN PGP SIGNATURE-----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=ocVM
-----END PGP SIGNATURE-----
Merge tag 'drm-next-2024-05-15' of https://gitlab.freedesktop.org/drm/kernel
Pull drm updates from Dave Airlie:
"This is the main pull request for the drm subsystems for 6.10.
In drivers the main thing is a new driver for ARM Mali firmware based
GPUs, otherwise there are a lot of changes to amdgpu/xe/i915/msm and
scattered changes to everything else.
In the core a bunch of headers and Kconfig was refactored, along with
the addition of a new panic handler which is meant to provide a user
friendly message when a panic happens and graphical display is
enabled.
New drivers:
- panthor: ARM Mali/Immortalis CSF-based GPU driver
Core:
- add a CONFIG_DRM_WERROR option
- make more headers self-contained
- grab resv lock in pin/unpin
- fix vmap resv locking
- EDID/eDP panel matching
- Kconfig cleanups
- DT sound bindings
- Add SIZE_HINTS property for cursor planes
- Add struct drm_edid_product_id and helpers.
- Use drm device based logging in more drm functions.
- drop seq_file.h from a bunch of places
- use drm_edid driver conversions
dp:
- DP Tunnel documentation
- MST read sideband cap
- Adaptive sync SDP prep work
ttm:
- improve placement for TTM BOs in idle/busy handling
panic:
- Fixes for drm-panic, and option to test it.
- Add drm panic to simpledrm, mgag200, imx, ast
bridge:
- improve init ordering
- adv7511: allow GPIO pin sharing
- tc358775: add tc358675 support
panel:
- AUO B120XAN01.0
- Samsung s6e3fa7
- BOE NT116WHM-N44
- CMN N116BCA-EA1,
- CrystalClear CMT430B19N00
- Startek KD050HDFIA020-C020A
- powertip PH128800T006-ZHC01
- Innolux G121X1-L03
- LG sw43408
- Khadas TS050 V2
- EDO RM69380 OLED
- CSOT MNB601LS1-1
amdgpu:
- HDCP/ODM/RAS fixes
- Devcoredump improvements
- Expose VCN activity via sysfs
- SMY 13.0.x updates
- Enable fast updates on DCN 3.1.4
- Add dclk and vclk reporting on additional devices
- Add ACA RAS infrastructure
- Implement TLB flush fence
- EEPROM handling fixes
- SMUIO 14.0.2 support
- SMU 14.0.1 Updates
- SMU 14.0.2 support
- Sync page table freeing with TLB flushes
- DML2 refactor
- DC debug improvements
- DCN 3.5.x Updates
- GPU reset fixes
- HDP fix for second GFX pipe on GC 10.x
- Enable secondary GFX pipe on GC 10.3
- Refactor and clean up BACO/BOCO/BAMACO handling
- Remove invalid TTM resource start check
- UAF fix in VA IOCTL
- GPUVM page fault redirection to secondary IH rings for IH 6.x
- Initial support for mapping kernel queues via MES
- Fix VRAM memory accounting
amdkfd:
- MQD handling cleanup
- Preemption handling fixes for XCDs
- TLB flush fix for GC 9.4.2
- Properly clean up workqueue during module unload
- Fix memory leak process create failure
- Range check CP bad op exception targets to avoid reporting invalid exceptions to userspace
- Fix eviction fence handling
- Fix leak in GPU memory allocation failure case
- DMABuf import handling fix
- Enable SQ watchpoint for gfx10
i915:
- Adding new DG2 PCI ID
- add context hints for GT frequency
- enable only one CCS for compute workloads
- new workarounds
- Fix UAF on destroy against retire race and remove two earlier partial fixes
- Limit the reserved VM space to only the platforms that need it
- Fix gt reset with GuC submission is disable
- Add and use gt_to_guc() wrapper
i915/xe display:
- Lunar Lake display enabling, including cdclk and other refactors
- BIOS/VBT/opregion related refactor
- Digital port related refactor/clean-up
- Fix 2s boot time regression on DP panel replay init
- Remove duplication on audio enable/disable on SDVO and g4x+ DP
- Disable AuxCCS framebuffers if built for Xe
- Make crtc disable more atomic
- Increase DP idle pattern wait timeout to 2ms
- Start using container_of_const() for some extra const safety
- Fix Jasper Lake boot freeze
- Enable MST mode for 128b/132b single-stream sideband
- Enable Adaptive Sync SDP Support for DP
- Fix MTL supported DP rates - removal of UHBR13.5
- PLL refactoring
- Limit eDP MSO pipe only for display version 20
- More display refactor towards independence from i915 dev_priv
- Convert i915/xe fbdev to DRM client
- More initial work to make display code more independent from i915
xe:
- improved error capture
- clean up some uAPI leftovers
- devcoredump update
- Add BMG mocs table
- Handle GSCCS ER interrupt
- Implement xe2- and GuC workarounds
- struct xe_device cleanup
- Hwmon updates
- Add LRC parsing for more GPU instruction
- Increase VM_BIND number of per-ioctl Ops
- drm/xe: Add XE_BO_GGTT_INVALIDATE flag
- Initial development for SR-IOV support
- Add new PCI IDs to DG2 platform
- Move userptr over to start using hmm_range_fault
msm:
- Switched to generating register header files during build process
instead of shipping pre-generated headers
- Merged DPU and MDP4 format databases.
- DP:
- Stop using compat string to distinguish DP and eDP cases
- Added support for X Elite platform (X1E80100)
- Reworked DP aux/audio support
- Added SM6350 DP to the bindings
- GPU:
- a7xx perfcntr reg fixes
- MAINTAINERS updates
- a750 devcoredump support
radeon:
- Silence UBSAN warnings related to flexible arrays
nouveau:
- move some uAPI objects to uapi headers
omapdrm:
- console fix
ast:
- add i2c polling
qaic:
- add debugfs entries
exynos:
- fix platform_driver .owner
- drop cleanup code
mediatek:
- Use devm_platform_get_and_ioremap_resource() in mtk_hdmi_ddc_probe()
- Add GAMMA 12-bit LUT support for MT8188
- Rename mtk_drm_* to mtk_*
- Drop driver owner initialization
- Correct calculation formula of PHY Timing"
* tag 'drm-next-2024-05-15' of https://gitlab.freedesktop.org/drm/kernel: (1477 commits)
drm/xe/ads: Use flexible-array
drm/xe: Use ordered WQ for G2H handler
drm/msm/gen_header: allow skipping the validation
drm/msm/a6xx: Cleanup indexed regs const'ness
drm/msm: Add devcoredump support for a750
drm/msm: Adjust a7xx GBIF debugbus dumping
drm/msm: Update a6xx registers XML
drm/msm: Fix imported a750 snapshot header for upstream
drm/msm: Import a750 snapshot registers from kgsl
MAINTAINERS: Add Konrad Dybcio as a reviewer for the Adreno driver
MAINTAINERS: Add a separate entry for Qualcomm Adreno GPU drivers
drm/msm/a6xx: Avoid a nullptr dereference when speedbin setting fails
drm/msm/adreno: fix CP cycles stat retrieval on a7xx
drm/msm/a7xx: allow writing to CP_BV counter selection registers
drm: zynqmp_dpsub: Always register bridge
Revert "drm/bridge: ti-sn65dsi83: Fix enable error path"
drm/fb_dma: Add checks in drm_fb_dma_get_scanout_buffer()
drm/fbdev-generic: Do not set physical framebuffer address
drm/panthor: Fix the FW reset logic
drm/panthor: Make sure we handle 'unknown group state' case properly
...
We don't get the right offset in that case. The GPU has
an unused 4K area of the register BAR space into which you can
remap registers. We remap the HDP flush registers into this
space to allow userspace (CPU or GPU) to flush the HDP when it
updates VRAM. However, on systems with >4K pages, we end up
exposing PAGE_SIZE of MMIO space.
Fixes: d8e408a827 ("drm/amdkfd: Expose HDP registers to user space")
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
This reverts commit c37ce764cd.
RCCL library is currently not treating spatial partitions differently,
hence this change is causing issues. Revert temporarily till RCCL
implementation is ready for spatial partitions.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Jonathan Kim <jonathan.kim@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
There is a race condition when re-creating a kfd_process for a process.
This has been observed when a process under the debugger executes
exec(3). In this scenario:
- The process executes exec.
- This will eventually release the process's mm, which will cause the
kfd_process object associated with the process to be freed
(kfd_process_free_notifier decrements the reference count to the
kfd_process to 0). This causes kfd_process_ref_release to enqueue
kfd_process_wq_release to the kfd_process_wq.
- The debugger receives the PTRACE_EVENT_EXEC notification, and tries to
re-enable AMDGPU traps (KFD_IOC_DBG_TRAP_ENABLE).
- When handling this request, KFD tries to re-create a kfd_process.
This eventually calls kfd_create_process and kobject_init_and_add.
At this point the call to kobject_init_and_add can fail because the
old kfd_process.kobj has not been freed yet by kfd_process_wq_release.
This patch proposes to avoid this race by making sure to drain
kfd_process_wq before creating a new kfd_process object. This way, we
know that any cleanup task is done executing when we reach
kobject_init_and_add.
Signed-off-by: Lancelot SIX <lancelot.six@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Queue buffer, though it is in system memory, has to be created using the
correct amdgpu device. Enforce this as the BO needs to mapped to the
GART for MES Hardware scheduler to access it.
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Replace tmz flag into buffer flag to make it easier to understand
and extend
Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Frank Min <Frank.Min@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Do VRAM accounting when doing migrations to vram to make sure
there is enough available VRAM and migrating to VRAM doesn't evict
other possible non-unified memory BOs. If migrating to VRAM fails,
driver can fall back to using system memory seamlessly.
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Handle the case that the restore worker was already scheduled by another
eviction while the restore was in progress.
Fixes: 9a1c1339ab ("drm/amdkfd: Run restore_workers on freezable WQs")
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Tested-by: Yunxiang Li <Yunxiang.Li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Do VRAM accounting when doing migrations to vram to make sure
there is enough available VRAM and migrating to VRAM doesn't evict
other possible non-unified memory BOs. If migrating to VRAM fails,
driver can fall back to using system memory seamlessly.
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Handle the case that the restore worker was already scheduled by another
eviction while the restore was in progress.
Fixes: 9a1c1339ab ("drm/amdkfd: Run restore_workers on freezable WQs")
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Tested-by: Yunxiang Li <Yunxiang.Li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Handle case that dma_fence_get_rcu_safe returns NULL.
If restore work is already scheduled, only update its timer. The same
work item cannot be queued twice, so undo the extra queue eviction.
Fixes: 9a1c1339ab ("drm/amdkfd: Run restore_workers on freezable WQs")
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Tested-by: Gang BA <Gang.Ba@amd.com>
Reviewed-by: Gang BA <Gang.Ba@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
It's not really an error since the devices don't support
the necessary hardware functionality.
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3331
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
-----BEGIN PGP SIGNATURE-----
iQFSBAABCAA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAmYlapoeHHRvcnZhbGRz
QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiGJLgH/RFrsXefpxAWACJv
wRmIx/YFjI3Z4nypBZPXysX/JOuf/cf4bR22jMHC1vhQDtxVgB7eUpEYx0C0Y5RD
ll7cunAZCpP+PB725/bOWbaF94eigl8xC4RrKPon4tG1CS9NAvCJFDyNgWpqh48s
sSKKae68a16zMtIX5Za8nvmsBQOeD67ZFBgI4m7wIDDhTXY3XlG84r8Rz/ZwcdE/
rExFkDMUqjpoJVAMBtETQNk4pvKZVMrW2B0f/xZSR+IKXw5l2FO0raXkIG/6TERs
CxJvCuUjKIosqFLIK2O/cU9G+5V2axZ/WD8YRwDr2vlnnOPht+dFQzyPcuSqezbG
ShIhQo8=
=GI5F
-----END PGP SIGNATURE-----
Backmerge tag 'v6.9-rc5' into drm-next
Linux 6.9-rc5
I've had a persistent msm failure on clang, and the fix is in fixes
so just pull it back to fix that.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Handle case that dma_fence_get_rcu_safe returns NULL.
If restore work is already scheduled, only update its timer. The same
work item cannot be queued twice, so undo the extra queue eviction.
Fixes: 9a1c1339ab ("drm/amdkfd: Run restore_workers on freezable WQs")
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Tested-by: Gang BA <Gang.Ba@amd.com>
Reviewed-by: Gang BA <Gang.Ba@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
mode-2 reset is the only reliable method that can get
GC/SDMA back when poison is consumed. mmhub requires
mode-1 reset.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix memory leak due to a leaked mmget reference on an error handling
code path that is triggered when attempting to create KFD processes
while a GPU reset is in progress.
Fixes: 0ab2d7532b ("drm/amdkfd: prepare per-process debug enable and disable")
CC: Xiaogang Chen <xiaogang.chen@amd.com>
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Tested-by: Harish Kasiviswanthan <Harish.Kasiviswanthan@amd.com>
Reviewed-by: Mukul Joshi <mukul.joshi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Fix memory leak due to a leaked mmget reference on an error handling
code path that is triggered when attempting to create KFD processes
while a GPU reset is in progress.
Fixes: 0ab2d7532b ("drm/amdkfd: prepare per-process debug enable and disable")
CC: Xiaogang Chen <xiaogang.chen@amd.com>
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Tested-by: Harish Kasiviswanthan <Harish.Kasiviswanthan@amd.com>
Reviewed-by: Mukul Joshi <mukul.joshi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If there are more than one device doing reset in parallel, the first
device will call kfd_suspend_all_processes() to evict all processes
on all devices, this call takes time to finish. other device will
start reset and recover without waiting. if the process has not been
evicted before doing recover, it will be restored, then caused page
fault.
Signed-off-by: Zhigang Luo <Zhigang.Luo@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Currently, with F32 HWS GPU reset is only when unmap queue fails.
However, if compute queue doesn't repond to preemption request in time
unmap will return without any error. In this case, only preemption error
is logged and Reset is not triggered. Call GPU reset in this case also.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Reviewed-by: Mukul Joshi <mukul.joshi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
If there are more than one device doing reset in parallel, the first
device will call kfd_suspend_all_processes() to evict all processes
on all devices, this call takes time to finish. other device will
start reset and recover without waiting. if the process has not been
evicted before doing recover, it will be restored, then caused page
fault.
Signed-off-by: Zhigang Luo <Zhigang.Luo@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Due to a CP interrupt bug, bad packet garbage exception codes are raised.
Do a range check so that the debugger and runtime do not receive garbage
codes.
Update the user api to guard exception code type checking as well.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Tested-by: Jesse Zhang <jesse.zhang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
TLB flush after unmap accidentially was removed on
gfx9.4.2. It is to add it back.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Check cgroup permissions when returning DMA-buf info and
based on cgroup info return the GPU id of the GPU that have
access to the BO.
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Currently, with F32 HWS GPU reset is only when unmap queue fails.
However, if compute queue doesn't repond to preemption request in time
unmap will return without any error. In this case, only preemption error
is logged and Reset is not triggered. Call GPU reset in this case also.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Reviewed-by: Mukul Joshi <mukul.joshi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Destroy the high priority workqueue that handles interrupts
during KFD node cleanup.
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Due to a CP interrupt bug, bad packet garbage exception codes are raised.
Do a range check so that the debugger and runtime do not receive garbage
codes.
Update the user api to guard exception code type checking as well.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Tested-by: Jesse Zhang <jesse.zhang@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
TLB flush after unmap accidentially was removed on
gfx9.4.2. It is to add it back.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Check cgroup permissions when returning DMA-buf info and
based on cgroup info return the GPU id of the GPU that have
access to the BO.
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Each RAS block has different requirement for gpu reset in poison
consumption handling.
Add support for mmhub RAS poison consumption handling.
v2: remove the mmhub poison support for kfd int v10.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Support the query for both gfxhub and mmhub, also replace
xcc_id with hub_inst.
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch adds the following functionality:
- Check the queue preemption status on all XCDs in a partition
for GFX 9.4.3.
- Update the queue preemption debug message to print the queue
doorbell id for which preemption failed.
- Change the signature of check preemption failed function to
return a bool instead of uint32_t and pass the MQD manager
as an argument.
Suggested-by: Jay Cornwall <jay.cornwall@amd.com>
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rename read_doorbell_id function to a more meaningful name,
implying what it is used for. No functional change.
Suggested-by: Jay Cornwall <jay.cornwall@amd.com>
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Replace it with related interface in gfxhub functions.
v2: replace node id with xcc id.
get node id for query_utcl2_poison_status
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since commit 43a7206b09 ("driver core: class: make class_register() take
a const *"), the driver core allows for struct class to be in read-only
memory, so move the kfd_class structure to be declared at build time
placing it into read-only memory, instead of having to be dynamically
allocated at boot time.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net>
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Similarly to gfx9, gfx10.1 drops vector stores when an xnack error is
raised. To work around this issue, use scalar stores instead of vector
stores when trapsts.xnack_error == 1.
Signed-off-by: Laurent Morichetti <laurent.morichetti@amd.com>
Reviewed-by: Jay Cornwall <jay.cornwall@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In a future commit, the cwsr trap handler code size for gfx10.1 will
increase to slightly above the one page mark. Since the TMA does not
need to be page aligned, and only 2 pointers are stored in it, push
the TMA offset by 2 KiB and keep the TBA+TMA reserved memory size
to two pages.
Signed-off-by: Laurent Morichetti <laurent.morichetti@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch changes the handling and lifecycle of vm->task_info object.
The major changes are:
- vm->task_info is a dynamically allocated ptr now, and its uasge is
reference counted.
- introducing two new helper funcs for task_info lifecycle management
- amdgpu_vm_get_task_info: reference counts up task_info before
returning this info
- amdgpu_vm_put_task_info: reference counts down task_info
- last put to task_info() frees task_info from the vm.
This patch also does logistical changes required for existing usage
of vm->task_info.
V2: Do not block all the prints when task_info not found (Felix)
V3: Fixed review comments from Felix
- Fix wrong indentation
- No debug message for -ENOMEM
- Add NULL check for task_info
- Do not duplicate the debug messages (ti vs no ti)
- Get first reference of task_info in vm_init(), put last
in vm_fini()
V4: Fixed review comments from Felix
- fix double reference increment in create_task_info
- change amdgpu_vm_get_task_info_pasid
- additional changes in amdgpu_gem.c while porting
Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Shashank Sharma <shashank.sharma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The adev can be found from bo by amdgpu_ttm_adev(bo->tbo.bdev),
and adev is also not used in the function
amdgpu_amdkfd_map_gtt_bo_to_gart().
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
On devices which have multi-partition nodes, keep partition id in
location_id[31:28].
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Jonathan Kim <jonathan.kim@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If fatal error is detected, packet submission won't go through. Return
error in such cases. Also, avoid waiting for fence when fatal error is
detected.
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Prevent dropping the KFD process reference at the end of a debug
IOCTL call where the acquired process value is an error.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The TBA and TMA, along with an unused IB allocation, reside at low
addresses in the VM address space. A stray VM fault which hits these
pages must be serviced by making their page table entries invalid.
The scheduler depends upon these pages being resident and fails,
preventing a debugger from inspecting the failure state.
By relocating these pages above 47 bits in the VM address space they
can only be reached when bits [63:48] are set to 1. This makes it much
less likely for a misbehaving program to generate accesses to them.
The current placement at VA (PAGE_SIZE*2) is readily hit by a NULL
access with a small offset.
v2:
- Move it to the reserved space to avoid concflicts with Mesa
- Add macros to make reserved space management easier
v3:
- Move VM max PFN calculation into AMDGPU_VA_RESERVED macros
Cc: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Cc: Christian Koenig <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jay Cornwall <jay.cornwall@amd.com>
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
