drm/amdgpu: fix clearing mappings for BOs that are always valid in VM
Subsystems (3)
PATCH 1/1 drm/amdgpu: fix clearing mappings for BOs that are always valid in VM
If the BO has been moved the PT should be updated, otherwise the VAs might point to invalid PT.
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.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +12 -0
@@ -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); }
Signature · 1 line
2.41.0
Am 16.06.23 um 08:27 schrieb Samuel Pitoiset:
Quoted 2 lines
If the BO has been moved the PT should be updated, otherwise the VAs might point to invalid PT.
You might want to update this sentence a bit. Something like:
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.
Apart from that really good catch and the patch is Reviewed-by: Christian König <[email protected]>
Regards,
Christian.
Quoted 46 lines
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.
Signed-off-by Samuel Pitoiset <[email protected]>--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +12
-0@@ -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); }
PATCH v2 1/1 drm/amdgpu: fix clearing mappings for BOs that are always valid in VM
In reply to Samuel PitoisetPer 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.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +12 -0
@@ -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); }
Signature · 1 line
2.41.0
Alex Deucher
AppliedApplied. Thanks!
Alex
Quoted 54 lines · Samuel Pitoiset
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.
Signed-off-by Samuel Pitoiset <[email protected]>Reviewed-by Christian König <[email protected]>--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +12
-1@@ -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); } --2.41.0