kernel/linux.git/drivers/gpu/drm/amd/amdgpu, branch linux-6.9.y

drm/amdgpu: Fix signedness bug in sdma_v4_0_process_trap_irq()

2024-07-27T09:36:15+00:00

commit 6769a23697f17f9bf9365ca8ed62fe37e361a05a upstream. The "instance" variable needs to be signed for the error handling to work. Fixes: 8b2faf1a4f3b ("drm/amdgpu: add error handle to avoid out-of-bounds") Reviewed-by: Bob Zhou Signed-off-by: Dan Carpenter Signed-off-by: Alex Deucher Cc: Siddh Raman Pant Signed-off-by: Greg Kroah-Hartman

drm/amdgpu: init TA fw for psp v14

2024-07-25T07:53:30+00:00

[ Upstream commit ed5a4484f074aa2bfb1dad99ff3628ea8da4acdc ] Add support to init TA firmware for psp v14. Signed-off-by: Likun Gao Acked-by: Alex Deucher Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: Indicate CU havest info to CP

2024-07-25T07:53:30+00:00

[ Upstream commit 49c9ffabde555c841392858d8b9e6cf58998a50c ] To achieve full occupancy CP hardware needs to know if CUs in SE are symmetrically or asymmetrically harvested v2: Reset is_symmetric_cus for each loop Signed-off-by: Harish Kasiviswanathan Acked-by: Alex Deucher Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: correct hbm field in boot status

2024-07-11T10:51:21+00:00

[ Upstream commit ec58991054e899c9d86f7e3c8a96cb602d4b5938 ] hbm filed takes bit 13 and bit 14 in boot status. Signed-off-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Let VRAM allocations go to GTT domain on small APUs

2024-07-11T10:51:21+00:00

[ Upstream commit eb853413d02c8d9b27942429b261a9eef228f005 ] 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 Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: fix the warning about the expression (int)size - len

2024-07-11T10:51:00+00:00

[ Upstream commit ea686fef5489ef7a2450a9fdbcc732b837fb46a8 ] Converting size from size_t to int may overflow. v2: keep reverse xmas tree order (Christian) Signed-off-by: Jesse Zhang Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: fix uninitialized scalar variable warning

2024-07-11T10:51:00+00:00

[ Upstream commit 9a5f15d2a29d06ce5bd50919da7221cda92afb69 ] Clear warning that uses uninitialized value fw_size. Signed-off-by: Tim Huang Reviewed-by: Christian König Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: fix double free err_addr pointer warnings

2024-07-11T10:50:59+00:00

[ Upstream commit 506c245f3f1cd989cb89811a7f06e04ff8813a0d ] In amdgpu_umc_bad_page_polling_timeout, the amdgpu_umc_handle_bad_pages will be run many times so that double free err_addr in some special case. So set the err_addr to NULL to avoid the warnings. Signed-off-by: Bob Zhou Acked-by: Alex Deucher Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: Initialize timestamp for some legacy SOCs

2024-07-11T10:50:59+00:00

[ Upstream commit 2e55bcf3d742a4946d862b86e39e75a95cc6f1c0 ] Initialize the interrupt timestamp for some legacy SOCs to fix the coverity issue "Uninitialized scalar variable" Signed-off-by: Ma Jun Suggested-by: Christian König Reviewed-by: Christian König Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: Using uninitialized value *size when calling amdgpu_vce_cs_reloc

2024-07-11T10:50:58+00:00

[ Upstream commit 88a9a467c548d0b3c7761b4fd54a68e70f9c0944 ] Initialize the size before calling amdgpu_vce_cs_reloc, such as case 0x03000001. V2: To really improve the handling we would actually need to have a separate value of 0xffffffff.(Christian) Signed-off-by: Jesse Zhang Suggested-by: Christian König Reviewed-by: Christian König Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin