The patch below does not apply to the 4.14-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable@xxxxxxxxxxxxxxx>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-4.14.y git checkout FETCH_HEAD git cherry-pick -x ea2c3c08554601b051d91403a241266e1cf490a5 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable@xxxxxxxxxxxxxxx>' --in-reply-to '2023072132-dupe-darkened-6015@gregkh' --subject-prefix 'PATCH 4.14.y' HEAD^.. Possible dependencies: ea2c3c085546 ("drm/amdgpu: fix clearing mappings for BOs that are always valid in VM") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From ea2c3c08554601b051d91403a241266e1cf490a5 Mon Sep 17 00:00:00 2001 From: Samuel Pitoiset <samuel.pitoiset@xxxxxxxxx> Date: Fri, 16 Jun 2023 15:14:07 +0200 Subject: [PATCH] drm/amdgpu: fix clearing mappings for BOs that are always valid in VM MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Per VM BOs must be marked as moved or otherwise their ranges are not updated on use which might be necessary when the replace operation splits mappings. This fixes random GPU hangs when replacing sparse mappings from the userspace, while OP_MAP/OP_UNMAP works fine because always valid BOs are correctly handled there. Cc: stable@xxxxxxxxxxxxxxx Signed-off-by: Samuel Pitoiset <samuel.pitoiset@xxxxxxxxx> Reviewed-by: Christian König <christian.koenig@xxxxxxx> Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 143d11afe0e5..eff73c428b12 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -1771,18 +1771,30 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, /* Insert partial mapping before the range */ if (!list_empty(&before->list)) { + struct amdgpu_bo *bo = before->bo_va->base.bo; + amdgpu_vm_it_insert(before, &vm->va); if (before->flags & AMDGPU_PTE_PRT) amdgpu_vm_prt_get(adev); + + if (bo && bo->tbo.base.resv == vm->root.bo->tbo.base.resv && + !before->bo_va->base.moved) + amdgpu_vm_bo_moved(&before->bo_va->base); } else { kfree(before); } /* Insert partial mapping after the range */ if (!list_empty(&after->list)) { + struct amdgpu_bo *bo = after->bo_va->base.bo; + amdgpu_vm_it_insert(after, &vm->va); if (after->flags & AMDGPU_PTE_PRT) amdgpu_vm_prt_get(adev); + + if (bo && bo->tbo.base.resv == vm->root.bo->tbo.base.resv && + !after->bo_va->base.moved) + amdgpu_vm_bo_moved(&after->bo_va->base); } else { kfree(after); }