On 2019-12-05 8:39 a.m., Christian König wrote:
This allows to invalidate VM entries without taking the reservation lock.
Signed-off-by: Christian König
---
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 39 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 4 +++
2 files changed, 36 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 0d700e8154c4..839d6df394fc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -656,7 +656,7 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
void *param)
{
struct amdgpu_vm_bo_base *bo_base, *tmp;
- int r = 0;
+ int r;
vm->bulk_moveable &= list_empty(>evicted);
@@ -665,7 +665,7 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
r = validate(param, bo);
if (r)
- break;
+ return r;
if (bo->tbo.type != ttm_bo_type_kernel) {
amdgpu_vm_bo_moved(bo_base);
@@ -678,7 +678,11 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
}
}
- return r;
+ mutex_lock(>eviction_lock);
+ vm->evicting = false;
+ mutex_unlock(>eviction_lock);
+
+ return 0;
}
/**
@@ -1555,15 +1559,25 @@ static int amdgpu_vm_bo_update_mapping(struct
amdgpu_device *adev,
if (!(flags & AMDGPU_PTE_VALID))
owner = AMDGPU_FENCE_OWNER_KFD;
+ mutex_lock(>eviction_lock);
+ if (vm->evicting) {
+ r = EINPROGRESS;
Not sure about this error code. As far as I can find, this is normally
used on non-blocking sockets. I found this explanation:
https://stackoverflow.com/questions/8277970/what-are-possible-reason-for-socket-error-einprogress-in-solaris
Quote: "so there is another error for non-blocking connect: EINPROGRESS,
which tells you that the operation is in progress and you should check
its status later."
This call is neither non-blocking nor is the requested page table update
in progress when this error is returned. So I'd think a better error to
return here would be EBUSY.
Other than that, this patch is
Reviewed-by: Felix Kuehling
+ goto error_unlock;
+ }
+
r = vm->update_funcs->prepare(, owner, exclusive);
if (r)
- return r;
+ goto error_unlock;
r = amdgpu_vm_update_ptes(, start, last + 1, addr, flags);
if (r)
- return r;
+ goto error_unlock;
- return vm->update_funcs->commit(, fence);
+ r = vm->update_funcs->commit(, fence);
+
+error_unlock:
+ mutex_unlock(>eviction_lock);
+ return r;
}
/**
@@ -2522,11 +2536,19 @@ bool amdgpu_vm_evictable(struct amdgpu_bo *bo)
if (!dma_resv_test_signaled_rcu(bo->tbo.base.resv, true))
return false;
+ /* Try to block ongoing updates */
+ if (!mutex_trylock(_base->vm->eviction_lock))
+ return false;
+
/* Don't evict VM page tables while they are updated */
if (!dma_fence_is_signaled(bo_base->vm->last_direct) ||
- !dma_fence_is_signaled(bo_base->vm->last_delayed))
+ !dma_fence_is_signaled(bo_base->vm->last_delayed)) {
+ mutex_unlock(_base->vm->eviction_lock);
return false;
+ }
+ bo_base->vm->evicting = true;
+ mutex_unlock(_base->vm->eviction_lock);
return true;
}
@@ -2773,6 +2795,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
vm->last_direct = dma_fence_get_stub();
vm->last_delayed = dma_fence_get_stub();
+ mutex_init(>eviction_lock);
+ vm->evicting = false;
+
amdgpu_vm_bo_param(adev, vm, adev->vm_manager.root_level, false, );
if (vm_context == AMDGPU_VM_CONTEXT_COMPUTE)
bp.flags &= ~AMDGPU_GEM_CREATE_SHADOW;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index d93ea9ad879e..d5613d184e99 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -242,6 +242,10 @@ struct amdgpu_vm {
/* tree of virtual addresses mapped */
struct rb_root_cached va;
+ /* Lock to prevent eviction while we are updating page tables */
+ struct mutexeviction_lock;
+ boolevicting;
+
/* BOs who needs a validation */
struct list_headevicted;
___
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
