Age | Commit message (Collapse) | Author | Files | Lines |
|
Switch to IP version checking instead of asic_type on various KFD
version checks.
Signed-off-by: Graham Sider <Graham.Sider@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Pull more drm updates from Dave Airlie:
"I missed a drm-misc-next pull for the main pull last week. It wasn't
that major and isn't the bulk of this at all. This has a bunch of
fixes all over, a lot for amdgpu and i915.
bridge:
- HPD improvments for lt9611uxc
- eDP aux-bus support for ps8640
- LVDS data-mapping selection support
ttm:
- remove huge page functionality (needs reworking)
- fix a race condition during BO eviction
panels:
- add some new panels
fbdev:
- fix double-free
- remove unused scrolling acceleration
- CONFIG_FB dep improvements
locking:
- improve contended locking logging
- naming collision fix
dma-buf:
- add dma_resv_for_each_fence iterator
- fix fence refcounting bug
- name locking fixesA
prime:
- fix object references during mmap
nouveau:
- various code style changes
- refcount fix
- device removal fixes
- protect client list with a mutex
- fix CE0 address calculation
i915:
- DP rates related fixes
- Revert disabling dual eDP that was causing state readout problems
- put the cdclk vtables in const data
- Fix DVO port type for older platforms
- Fix blankscreen by turning DP++ TMDS output buffers on encoder->shutdown
- CCS FBs related fixes
- Fix recursive lock in GuC submission
- Revert guc_id from i915_request tracepoint
- Build fix around dmabuf
amdgpu:
- GPU reset fix
- Aldebaran fix
- Yellow Carp fixes
- DCN2.1 DMCUB fix
- IOMMU regression fix for Picasso
- DSC display fixes
- BPC display calculation fixes
- Other misc display fixes
- Don't allow partial copy from user for DC debugfs
- SRIOV fixes
- GFX9 CSB pin count fix
- Various IP version check fixes
- DP 2.0 fixes
- Limit DCN1 MPO fix to DCN1
amdkfd:
- SVM fixes
- Fix gfx version for renoir
- Reset fixes
udl:
- timeout fix
imx:
- circular locking fix
virtio:
- NULL ptr deref fix"
* tag 'drm-next-2021-11-12' of git://anongit.freedesktop.org/drm/drm: (126 commits)
drm/ttm: Double check mem_type of BO while eviction
drm/amdgpu: add missed support for UVD IP_VERSION(3, 0, 64)
drm/amdgpu: drop jpeg IP initialization in SRIOV case
drm/amd/display: reject both non-zero src_x and src_y only for DCN1x
drm/amd/display: Add callbacks for DMUB HPD IRQ notifications
drm/amd/display: Don't lock connection_mutex for DMUB HPD
drm/amd/display: Add comment where CONFIG_DRM_AMD_DC_DCN macro ends
drm/amdkfd: Fix retry fault drain race conditions
drm/amdkfd: lower the VAs base offset to 8KB
drm/amd/display: fix exit from amdgpu_dm_atomic_check() abruptly
drm/amd/amdgpu: fix the kfd pre_reset sequence in sriov
drm/amdgpu: fix uvd crash on Polaris12 during driver unloading
drm/i915/adlp/fb: Prevent the mapping of redundant trailing padding NULL pages
drm/i915/fb: Fix rounding error in subsampled plane size calculation
drm/i915/hdmi: Turn DP++ TMDS output buffers back on in encoder->shutdown()
drm/locking: fix __stack_depot_* name conflict
drm/virtio: Fix NULL dereference error in virtio_gpu_poll
drm/amdgpu: fix SI handling in amdgpu_device_asic_has_dc_support()
drm/amdgpu: Fix dangling kfd_bo pointer for shared BOs
drm/amd/amdkfd: Don't sent command to HWS on kfd reset
...
|
|
MIGRATE_PFN_LOCKED is used to indicate to migrate_vma_prepare() that a
source page was already locked during migrate_vma_collect(). If it
wasn't then the a second attempt is made to lock the page. However if
the first attempt failed it's unlikely a second attempt will succeed,
and the retry adds complexity. So clean this up by removing the retry
and MIGRATE_PFN_LOCKED flag.
Destination pages are also meant to have the MIGRATE_PFN_LOCKED flag
set, but nothing actually checks that.
Link: https://lkml.kernel.org/r/20211025041608.289017-1-apopple@nvidia.com
Signed-off-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Ralph Campbell <rcampbell@nvidia.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
[Why]:
When we call hmm_range_fault to map memory after a migration, we don't
expect memory to be migrated again as a result of hmm_range_fault. The
driver ensures that all memory is in GPU-accessible locations so that
no migration should be needed. However, there is one corner case where
hmm_range_fault can unexpectedly cause a migration from DEVICE_PRIVATE
back to system memory due to a write-fault when a system memory page in
the same range was mapped read-only (e.g. COW). Ranges with individual
pages in different locations are usually the result of failed page
migrations (e.g. page lock contention). The unexpected migration back
to system memory causes a deadlock from recursive locking in our
driver.
[How]:
Creating a task reference new member under svm_range_list struct.
Setting this with "current" reference, right before the hmm_range_fault
is called. This member is checked against "current" reference at
svm_migrate_to_ram callback function. If equal, the migration will be
ignored.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
If some pages fail to migrate to system memory, don't update
prange->actual_loc = 0. This prevents endless CPU page faults after
partial migration failures due to contested page locks.
Migration to RAM must be complete during migrations from VRAM to VRAM and
during evictions. Implement retry and fail if the migration to RAM fails.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Not all migrate.cpages returned from migrate_vma_setup can be migrated,
for example non anonymous page, or out of device memory. So after
migrate_vma_pages returns, add debug message to count pages are
successfully migrated which has MIGRATE_PFN_VALID and
MIGRATE_PFN_MIGRATE flag set.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
cpages is only updated by migrate_vma_setup. So capture its value at
that point to clarify the significance of the number. The next patch
will add counting of actually migrated pages after migrate_vma_pages for
debug purposes.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
migrate_vma_setup may return cpages 0, means 0 page can be migrated,
treat this as error case to skip the rest of vma migration steps.
Change svm_migrate_vma_to_vram and svm_migrate_vma_to_ram to return the
number of pages migrated successfully or error code. The caller add up
all the successful migration pages and update prange->actual_loc only if
the total migrated pages is not 0.
This also removes the warning message "VRAM BO missing during
validation" if migration cpages is 0.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
No function change, use pr_debug_ratelimited to avoid per page debug
message overflowing dmesg buf and console log.
use dev_err to show error message from unexpected situation, to provide
clue to help debug without enabling dynamic debug log. Define dev_fmt to
output function name in error message.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Use resource_size function on resource object
instead of explicit computation.
Clean up coccicheck warning:
./drivers/gpu/drm/amd/amdkfd/kfd_migrate.c:905:10-13: ERROR: Missing
resource_size with res
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Reviewed-by: Amos Kong <kongjianjun@gmail.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Device manager releases device-specific resources when a driver
disconnects from a device, devm_memunmap_pages and
devm_release_mem_region calls in svm_migrate_fini are redundant.
It causes below warning trace after patch "drm/amdgpu: Split
amdgpu_device_fini into early and late", so remove function
svm_migrate_fini.
BUG: https://gitlab.freedesktop.org/drm/amd/-/issues/1718
WARNING: CPU: 1 PID: 3646 at drivers/base/devres.c:795
devm_release_action+0x51/0x60
Call Trace:
? memunmap_pages+0x360/0x360
svm_migrate_fini+0x2d/0x60 [amdgpu]
kgd2kfd_device_exit+0x23/0xa0 [amdgpu]
amdgpu_amdkfd_device_fini_sw+0x1d/0x30 [amdgpu]
amdgpu_device_fini_sw+0x45/0x290 [amdgpu]
amdgpu_driver_release_kms+0x12/0x30 [amdgpu]
drm_dev_release+0x20/0x40 [drm]
release_nodes+0x196/0x1e0
device_release_driver_internal+0x104/0x1d0
driver_detach+0x47/0x90
bus_remove_driver+0x7a/0xd0
pci_unregister_driver+0x3d/0x90
amdgpu_exit+0x11/0x20 [amdgpu]
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
If svm migration init failed to create pgmap for device memory, set
pgmap type to 0 to disable device SVM support capability.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Each zone-device page holds a reference to the SVM BO that manages its
backing storage. This is necessary to correctly hold on to the BO in
case zone_device pages are shared with a child-process.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
This is for debug purposes only.
It conditionally generates partial migrations to test mixed
CPU/GPU memory domain pages in a prange easily.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Migration skipped for pages that are already in VRAM
domain. These could be the result of previous partial
migrations to SYS RAM, and prefetch back to VRAM.
Ex. Coherent pages in VRAM that were not written/invalidated after
a copy-on-write.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Invalid pages can be the result of pages that have been migrated
already due to copy-on-write procedure or pages that were never
migrated to VRAM in first place. This is not an issue anymore,
as pranges now support mixed memory domains (CPU/GPU).
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
svm_range_prefault is called right before migrations to VRAM,
to make sure pages are resident in system memory before the migration.
With partial migrations, this reference is used by hmm range get pages
to avoid migrating pages that are already in the same VRAM domain.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
GPUs in the same XGMI hive have direct access to all
members'VRAM. When mapping memory to a GPU, we don't need
hmm_range_fault to fault device-private pages in the same
hive back to the host. Identifying the page owner as the hive,
rather than the individual GPU, accomplishes this.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Add helper function to get process device data structure from adev to
update counters.
Update vm faults, page_in, page_out counters will no be executed in
parallel, use WRITE_ONCE to avoid any form of compiler optimizations.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Access to the mm_node is now forbidden. So instead of hand wiring that
use the cursor functionality.
v2: fix handling as pointed out by Philip.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by and Tested-by: Philip Yang <philip.yang@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210602100914.46246-5-christian.koenig@amd.com
|
|
If migration copy failed because process is killed, or out of VRAM or
system memory, pass error code back to caller to handle error
gracefully.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
This function is only used in this source file.
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Use devm_memunmap_pages instead of memunmap_pages to release pgmap
and remove pgmap from device action, to avoid double free pgmap when
unloading driver module.
Release device memory region if failed to create device memory pages
structure.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
The variable r is being initialized with a value that is never read
and it is being updated later with a new value. The initialization is
redundant and can be removed.
Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
If prefetch range to gpu with acutal location is another gpu, or GPU
retry fault restore pages to migrate the range with acutal location is
gpu, then migrate from one gpu to another gpu.
Use system memory as bridge because sdma engine may not able to access
another gpu vram, use sdma of source gpu to migrate to system memory,
then use sdma of destination gpu to migrate from system memory to gpu.
Print out gpuid or gpuidx in debug messages.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
With xnack on, GPU vm fault handler decide the best restore location,
then migrate range to the best restore location and update GPU mapping
to recover the GPU vm fault.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
GPU page tables are invalidated by unmapping prange directly at
the mmu notifier, when page fault retry is enabled through
amdgpu_noretry global parameter. The restore page table is
performed at the page fault handler.
If xnack is on, we update GPU mappings after migration to avoid
unnecessary GPUVM faults.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
If CPU page fault happens, HMM pgmap_ops callback migrate_to_ram start
migrate memory from vram to ram in steps:
1. migrate_vma_pages get vram pages, and notify HMM to invalidate the
pages, HMM interval notifier callback evict process queues
2. Allocate system memory pages
3. Use svm copy memory to migrate data from vram to ram
4. migrate_vma_pages copy pages structure from vram pages to ram pages
5. Return VM_FAULT_SIGBUS if migration failed, to notify application
6. migrate_vma_finalize put vram pages, page_free callback free vram
pages and vram nodes
7. Restore work wait for migration is finished, then update GPU page
table mapping to system memory, and resume process queues
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Register svm range with same address and size but perferred_location
is changed from CPU to GPU or from GPU to CPU, trigger migration the svm
range from ram to vram or from vram to ram.
If svm range prefetch location is GPU with flags
KFD_IOCTL_SVM_FLAG_HOST_ACCESS, validate the svm range on ram first,
then migrate it from ram to vram.
After migrating to vram is done, CPU access will have cpu page fault,
page fault handler migrate it back to ram and resume cpu access.
Migration steps:
1. migrate_vma_pages get svm range ram pages, notify the
interval is invalidated and unmap from CPU page table, HMM interval
notifier callback evict process queues
2. Allocate new pages in vram using TTM
3. Use svm copy memory to sdma copy data from ram to vram
4. migrate_vma_pages copy ram pages structure to vram pages structure
5. migrate_vma_finalize put ram pages to free ram pages and memory
6. Restore work wait for migration is finished, then update GPUs page
table mapping to new vram pages, resume process queues
If migrate_vma_setup failed to collect all ram pages of range, retry 3
times until success to start migration.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Use sdma linear copy to migrate data between ram and vram. The sdma
linear copy command uses kernel buffer function queue to access system
memory through gart table.
Use reserved gart table window 0 to map system page address, and vram
page address is direct mapping. Use the same kernel buffer function to
fill in gart table mapping, so this is serialized with memory copy by
sdma job submit. We only need wait for the last memory copy sdma fence
for larger buffer migration.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
HMM migration alloc sizeof(struct page) on system memory for each VRAM
page, it is 1GB system memory reserved for 64GB VRAM. To avoid
application OOM, increase system memory used size based on VRAM size of
all GPUs, then application alloc memory will fail if system memory usage
reach the limit.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Register vram memory as MEMORY_DEVICE_PRIVATE type resource, to
allocate vram backing pages for page migration.
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|