linux/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c, branch v5.17

drm/amdgpu: fix suspend/resume hang regression

2022-03-02T23:36:43Z

Regression has been reported that suspend/resume may hang with the previous vm ready check commit. So bring back the evicted list check as a temp fix. Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1922 Fixes: c1a66c3bc425 ("drm/amdgpu: check vm ready by amdgpu_vm->evicting flag") Reviewed-by: Christian König Signed-off-by: Qiang Yu Signed-off-by: Alex Deucher

drm/amdgpu: check vm ready by amdgpu_vm->evicting flag

2022-02-23T21:31:06Z

Workstation application ANSA/META v21.1.4 get this error dmesg when running CI test suite provided by ANSA/META: [drm:amdgpu_gem_va_ioctl [amdgpu]] *ERROR* Couldn't update BO_VA (-16) This is caused by: 1. create a 256MB buffer in invisible VRAM 2. CPU map the buffer and access it causes vm_fault and try to move it to visible VRAM 3. force visible VRAM space and traverse all VRAM bos to check if evicting this bo is valuable 4. when checking a VM bo (in invisible VRAM), amdgpu_vm_evictable() will set amdgpu_vm->evicting, but latter due to not in visible VRAM, won't really evict it so not add it to amdgpu_vm->evicted 5. before next CS to clear the amdgpu_vm->evicting, user VM ops ioctl will pass amdgpu_vm_ready() (check amdgpu_vm->evicted) but fail in amdgpu_vm_bo_update_mapping() (check amdgpu_vm->evicting) and get this error log This error won't affect functionality as next CS will finish the waiting VM ops. But we'd better clear the error log by checking the amdgpu_vm->evicting flag in amdgpu_vm_ready() to stop calling amdgpu_vm_bo_update_mapping() later. Another reason is amdgpu_vm->evicted list holds all BOs (both user buffer and page table), but only page table BOs' eviction prevent VM ops. amdgpu_vm->evicting flag is set only for page table BOs, so we should use evicting flag instead of evicted list in amdgpu_vm_ready(). The side effect of this change is: previously blocked VM op (user buffer in "evicted" list but no page table in it) gets done immediately. v2: update commit comments. Acked-by: Paul Menzel Reviewed-by: Christian König Signed-off-by: Qiang Yu Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org

Merge tag 'amd-drm-next-5.17-2021-12-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

2021-12-23T01:55:28Z

amdgpu: - Add some display debugfs entries - RAS fixes - SR-IOV fixes - W=1 fixes - Documentation fixes - IH timestamp fix - Misc power fixes - IP discovery fixes - Large driver documentation updates - Multi-GPU memory use reductions - Misc display fixes and cleanups - Add new SMU debug option amdkfd: - SVM fixes radeon: - Fix typo in comment From: Alex Deucher Link: https://patchwork.freedesktop.org/patch/msgid/20211216202731.5900-1-alexander.deucher@amd.com

amdgpu: fix some comment typos

2021-12-14T21:10:58Z

Signed-off-by: Yann Dirson Signed-off-by: Alex Deucher

Merge drm/drm-next into drm-misc-next

2021-10-25T13:27:56Z

drm-misc-next hasn't been updated in a while and I need a post -rc2 state to merge some vc4 patches. Signed-off-by: Maxime Ripard

drm/amdgpu: use new iterator in amdgpu_vm_prt_fini

2021-10-20T12:06:35Z

No need to actually allocate an array of fences here. Signed-off-by: Christian König Reviewed-by: Daniel Vetter Link: https://patchwork.freedesktop.org/patch/msgid/20211005113742.1101-14-christian.koenig@amd.com

drm/amdgpu: use adev_to_drm for consistency when accessing drm_device

2021-10-08T17:22:13Z

adev_to_drm is used everywhere, so improve recent changes when accessing drm_device pointer from amdgpu_device. Signed-off-by: Guchun Chen Acked-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: Put drm_dev_enter/exit outside hot codepath

2021-09-23T19:17:29Z

We hit soft hang while doing memory pressure test on one numa system. After a qucik look, this is because kfd invalid/valid userptr memory frequently with process_info lock hold. Looks like update page table mapping use too much cpu time. perf top says below, 75.81% [kernel] [k] __srcu_read_unlock 6.19% [amdgpu] [k] amdgpu_gmc_set_pte_pde 3.56% [kernel] [k] __srcu_read_lock 2.20% [amdgpu] [k] amdgpu_vm_cpu_update 2.20% [kernel] [k] __sg_page_iter_dma_next 2.15% [drm] [k] drm_dev_enter 1.70% [drm] [k] drm_prime_sg_to_dma_addr_array 1.18% [kernel] [k] __sg_alloc_table_from_pages 1.09% [drm] [k] drm_dev_exit So move drm_dev_enter/exit outside gmc code, instead let caller do it. They are gart_unbind, gart_map, vm_clear_bo, vm_update_pdes and gmc_init_pdb0. vm_bo_update_mapping already calls it. Signed-off-by: xinhui pan Reviewed-and-tested-by: Andrey Grodzovsky Signed-off-by: Alex Deucher

drm/amdgpu: rename amdgpu_bo_get_preferred_pin_domain

2021-08-25T22:15:17Z

amdgpu_bo_get_preferred_pin_domain is used for page tables creation, which is not involved with page pinning. And it is used in more cases than display scanout, modify its documentation as well. Signed-off-by: Yifan Zhang Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdkfd: check access permisson to restore retry fault

2021-08-24T19:36:50Z

Check range access permission to restore GPU retry fault, if GPU retry fault on address which belongs to VMA, and VMA has no read or write permission requested by GPU, failed to restore the address. The vm fault event will pass back to user space. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher