This patch adds support to program trap handler settings
when loading driver with software scheduler (sched_policy=2).
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Suggested-by: Jay Cornwall <Jay.Cornwall@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For xnack on, if range ACCESS or ACCESS_IN_PLACE (AIP) by single GPU, or
range is ACCESS_IN_PLACE by mGPUs and all mGPUs connection on XGMI same
hive, the best prefetch location is prefetch_loc GPU. Otherwise, the best
prefetch location is always CPU because GPU does not have coherent
mapping VRAM of other GPUs even with large-BAR PCIe connection.
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>
Currently the SVM get_attr call allows querying, which flags are set
in the entire address range. Add the opposite query, which flags are
clear in the entire address range. Both queries can be combined in a
single get_attr call, which allows answering questions such as, "is
this address range coherent, non-coherent, or a mix of both"?
Proposed userspace for UAPI:
https://github.com/RadeonOpenCompute/ROCR-Runtime/tree/memory_model_queries
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Philip Yand <philip.yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add u32 gfx_target_version field to kfd_node_properties and
kfd_device_info. Populate <asic>_device_info structs accordingly and
expose to sysfs.
This allows eliminating device-ID-based lookup tables in user mode for
future ASICs.
Signed-off-by: Graham Sider <Graham.Sider@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It is to workaround HW bug on other Asics and based on
reverting two commits back:
drm/amdkfd: Add heavy-weight TLB flush after unmapping
drm/amdkfd: Add memory sync before TLB flush on unmap
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 4bba567c8c.
Revert reason: The issue has been resolved.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 7ed9876c97.
Revert reason: The issue has been resolved.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 430f8e6edb.
Revert reason: Issue has been resolved.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 4bba567c8c.
Revert reason: The issue has been resolved.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 7ed9876c97.
Revert reason: The issue has been resolved.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 430f8e6edb.
Revert reason: Issue has been resolved.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add KFD support for cyan_skillfish.
v2: whitespace fixes (Alex)
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
start_cpsch and stop_cpsch can be called during kfd device
initialization or during gpu reset/recovery. So they can
run concurrently. Currently in start_cpsch and stop_cpsch,
pm_init and pm_uninit is not protected by the dpm lock.
Imagine such a case that user use packet manager's function
to submit a pm4 packet to hang hws (ie through command
cat /sys/class/kfd/kfd/topology/nodes/1/gpu_id | sudo tee
/sys/kernel/debug/kfd/hang_hws), while kfd device is under
device reset/recovery so packet manager can be not initialized.
There will be unpredictable protection fault in such case.
This patch moves pm_init/uninit inside the dpm lock and check
packet manager is initialized before using packet manager
function.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Acked-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This variable will be used to determine whether packet
manager is initialized or not, in a future patch.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Acked-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Renaming packets to packet_mgr to reflect the real meaning
of this variable.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Acked-by: Christian Konig <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Similar to xGMI reporting the min/max bandwidth between direct peers, PCIe
will report the min/max bandwidth to the KFD.
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>
Report the min/max bandwidth in megabytes to the kfd for direct
xgmi connections only. Indirect peers will report 0 since
indirect route is unknown.
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>
This reverts commit 1098d658be.
Reason for revert: it causes regressions on several Asics.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 31f3324378.
Reason for revert: it causes regressions on several Asics.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 3be4dca197.
Reason for revert: it causes regressions on several Asics.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
prange is NULL if vm fault retry on invalid address, for this case, can
not use prange to get pdd, use adev to get gpuidx and then get pdd
instead, then increase pdd vm fault counter.
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 reverts commit 1098d658be.
Reason for revert: it causes regressions on several Asics.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 31f3324378.
Reason for revert: it causes regressions on several Asics.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reverts commit 3be4dca197.
Reason for revert: it causes regressions on several Asics.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
prange is NULL if vm fault retry on invalid address, for this case, can
not use prange to get pdd, use adev to get gpuidx and then get pdd
instead, then increase pdd vm fault counter.
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>
[Why]
svm ranges can have mixed pages from device or system memory.
A good example is, after a prange has been allocated in VRAM and a
copy-on-write is triggered by a fork. This invalidates some pages
inside the prange. Endding up in mixed pages.
[How]
By classifying each page inside a prange, based on its type. Device or
system memory, during dma mapping call. If page corresponds
to VRAM domain, a flag is set to its dma_addr entry for each GPU.
Then, at the GPU page table mapping. All group of contiguous pages within
the same type are mapped with their proper pte flags.
v2:
Instead of using ttm_res to calculate vram pfns in the svm_range. It is now
done by setting the vram real physical address into drm_addr array.
This makes more flexible VRAM management, plus removes the need to have
a BO reference in the svm_range.
v3:
Remove mapping member from svm_range
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>
Now that prange could have mixed domains (VRAM or SYSRAM),
actual_loc nor svm_bo can not be used to check its current
domain and eventually get its pfns to map them in GPU.
Instead, pfns from both domains, are now obtained from
hmm_range_fault through amdgpu_hmm_range_get_pages
call. This is done everytime a GPU map occur.
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>
Get the proper owner reference for amdgpu_hmm_range_get_pages function.
This is useful for partial migrations. To avoid migrating back to
system memory, VRAM pages, that are accessible by all devices in the
same memory domain.
Ex. multiple devices in the same hive.
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>
The parameter is used in the dev_private_owner to decide if device
pages in the range require to be migrated back to system memory, based
if they are or not in the same memory domain.
In this case, this reference could come from the same memory domain
with devices connected to the same hive.
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>
During GPU page table invalidation with xnack off, new ranges
split may occur concurrently in the same prange. Creating a new
child per split. Each child should also increment its
invalid counter, to assure GPU page table updates in these
ranges.
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>
This is part of SVM profiling API, export sysfs counters for
per-process, per-GPU vm retry fault, pages migrated in and out of GPU vram.
counters will not be updated in parallel in GPU retry fault handler and
migration to vram/ram path, use READ_ONCE to avoid compiler
optimization.
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>
3 cases of kobj leak, which causes memory leak:
kobj_type must have release() method to free memory from release
callback. Don't need NULL default_attrs to init kobj.
sysfs files created under kobj_status should be removed with kobj_status
as parent kobject.
Remove queue sysfs files when releasing queue from process MMU notifier
release callback.
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 functionality change. Modify kfd_sysfs_create_file to use kobject as
parameter, so it becomes common helper function to remove duplicate code
and will simplify new kfd sysfs file create in future.
Move pr_warn to helper function if sysfs file create failed. Set helper
function as void return because caller doesn't use the helper function
return value.
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>
To avoid any list corruption.
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix non-coherent bit of iolink properties flag
which always is 0.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Calling free_mqd inside of destroy_queue_nocpsch_locked can cause a
circular lock. destroy_queue_nocpsch_locked is called under a DQM lock,
which is taken in MMU notifiers, potentially in FS reclaim context.
Taking another lock, which is BO reservation lock from free_mqd, while
causing an FS reclaim inside the DQM lock creates a problematic circular
lock dependency. Therefore move free_mqd out of
destroy_queue_nocpsch_locked and call it after unlocking DQM.
Signed-off-by: Amber Lin <Amber.Lin@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Page table entries are now in embedded in VM BO, so
we do not need struct amdgpu_vm_pt. This patch replaces
struct amdgpu_vm_pt with struct amdgpu_vm_bo_base.
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When some GPUs don't support SVM, don't disabe it for the entire process.
That would be inconsistent with the information the process got from the
topology, which indicates SVM support per GPU.
Instead disable SVM support only for the unsupported GPUs. This is done
by checking any per-device attributes against the bitmap of supported
GPUs. Also use the supported GPU bitmap to initialize access bitmaps for
new SVM address ranges.
Don't handle recoverable page faults from unsupported GPUs. (I don't
think there will be unsupported GPUs that can generate recoverable page
faults. But better safe than sorry.)
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>
get_wave_state acquires the mmap_lock on copy_to_user but so do
mmu_notifiers. mmu_notifiers allows dqm locking so do get_wave_state
outside the dqm_lock to prevent circular locking.
v2: squash in unused variable removal.
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>
CoherentHostAccess flag support has moved from HSA_MEMORYPROPERTY
to HSA_CAPABILITY struct. Proper changes have made also at the thunk
to support this change.
CoherentHostAccess: whether or not device memory can be coherently
accessed by the host CPU.
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>
It is to fix a failure for SDMA updating PTEs.
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
