kernel/linux.git/include/uapi/linux/kfd_ioctl.h, branch v6.12.80

drm/amdkfd: clamp queue size to minimum

2025-04-20T08:15:24+00:00

[ Upstream commit e90711946b53590371ecce32e8fcc381a99d6333 ] If queue size is less than minimum, clamp it to minimum to prevent underflow when writing queue mqd. Signed-off-by: David Yat Sin Reviewed-by: Jay Cornwall Reviewed-by: Harish Kasiviswanathan Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Document and define SVM events message macro

2024-09-06T21:55:05+00:00

Document how to use SMI system management interface to enable and receive SVM events. Document SVM event triggers. Define SVM events message string format macro that could be used by user mode for sscanf to parse the event. Add it to uAPI header file to make it obvious that is changing uAPI in future. No functional changes. Signed-off-by: Philip Yang Reviewed-by: James Zhu Signed-off-by: Alex Deucher

drm/amdkfd: allow users to target recommended SDMA engines

2024-07-25T21:43:41+00:00

Certain GPUs have better copy performance over xGMI on specific SDMA engines depending on the source and destination GPU. Allow users to create SDMA queues on these recommended engines. Close to 2x overall performance has been observed with this optimization. Signed-off-by: Jonathan Kim Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher

drm/amdkfd: enable single alu ops for gfx12

2024-05-02T20:18:13+00:00

GFX12 debugging requires setting up precise ALU operation for catching ALU exceptions. Signed-off-by: Jonathan Kim Tested-by: Lancelot Six Reviewed-by: Eric Huang Signed-off-by: Alex Deucher

drm/amdkfd: Bump kfd version for contiguous VRAM allocation

2024-04-30T14:00:43+00:00

Bump the kfd ioctl minor version to delcare the contiguous VRAM allocation flag support. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher

drm/amdgpu: Support contiguous VRAM allocation

2024-04-30T13:44:46+00:00

RDMA device with limited scatter-gather ability requires contiguous VRAM buffer allocation for RDMA peer direct support. Add a new KFD alloc memory flag and store as bo alloc flag AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS. When pin this bo to export for RDMA peerdirect access, this will set TTM_PL_FLAG_CONTIFUOUS flag, and ask VRAM buddy allocator to get contiguous VRAM. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher

drm/amdkfd: range check cp bad op exception interrupts

2024-03-27T12:53:02+00:00

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 Tested-by: Jesse Zhang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher

drm/amdkfd: Bump KFD ioctl version

2024-01-15T23:35:35+00:00

This is not strictly a change in the IOCTL API. This version bump is meant to indicate to user mode the presence of a number of changes and fixes that enable the management of VA mappings in compute VMs using the GEM_VA ioctl for DMABufs exported from KFD. Signed-off-by: Felix Kuehling Reviewed-by: Xiaogang Chen Signed-off-by: Alex Deucher

drm/amdgpu: Add EXT_COHERENT memory allocation flags

2023-09-20T20:24:06+00:00

These flags (for GEM and SVM allocations) allocate memory that allows for system-scope atomic semantics. On GFX943 these flags cause caches to be avoided on non-local memory. On all other ASICs they are identical in functionality to the equivalent COHERENT flags. Corresponding Thunk patch is at https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/pull/88 Reviewed-by: David Yat Sin Signed-off-by: David Francis Signed-off-by: Alex Deucher

drm/amdkfd: bump kfd ioctl minor version for event age availability

2023-06-15T15:37:55+00:00

Bump the minor version to declare event age tracking feature is now available. In kernel amdgpu driver, kfd_wait_on_events is used to support user space signal event wait function. For multiple threads waiting on same event scenery, race condition could occur since some threads after checking signal condition, before calling kfd_wait_on_events, the event interrupt could be fired and wake up other thread which are sleeping on this event. Then those threads could fall into sleep without waking up again. Adding event age tracking in both kernel and user mode, will help avoiding this race condition. Proposed ROCT-Thunk-Interface: https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/efdbf6cfbc026bd68ac3c35d00dacf84370eb81e https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/1820ae0a2db85b6f584611dc0cde1a00e7c22915 Proposed ROCR-Runtime: https://github.com/RadeonOpenCompute/ROCR-Runtime/compare/master...zhums:ROCR-Runtime:new_event_wait_review https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/e1f5bdb88eb882ac798aeca2c00ea3fbb2dba459 https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/7d26afd14107b5c2a754c1a3f415d89f3aabb503 Signed-off-by: James Zhu Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher