On 2022-03-17 9:50 a.m., Christian
König wrote:
Better to leave the decision when to flush the VM changes in the TLB to the VM code. Signed-off-by: Christian König <christian.koenig@xxxxxxx> --- .../gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 21 +++++++------------ drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 4 ++-- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 5 ++--- 6 files changed, 16 insertions(+), 24 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 57b521bb1eec..8b14c55a0793 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1104,7 +1104,7 @@ static int update_gpuvm_pte(struct kgd_mem *mem, return ret; /* Update the page tables */ - ret = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + ret = amdgpu_vm_bo_update(adev, bo_va, false); if (ret) { pr_err("amdgpu_vm_bo_update failed\n"); return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 2d4a89fb264e..62518c7b31c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -806,7 +806,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_vm_bo_update(adev, fpriv->prt_va, false, NULL); + r = amdgpu_vm_bo_update(adev, fpriv->prt_va, false); if (r) return r; @@ -817,7 +817,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (amdgpu_mcbp || amdgpu_sriov_vf(adev)) { bo_va = fpriv->csa_va; BUG_ON(!bo_va); - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) return r; @@ -836,7 +836,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (bo_va == NULL) continue; - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index bab6500728cb..2e16484bf606 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -618,7 +618,7 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, if (operation == AMDGPU_VA_OP_MAP || operation == AMDGPU_VA_OP_REPLACE) { - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) goto error; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index a43302a86254..c29b12fec863 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -808,7 +808,6 @@ static void amdgpu_vm_tlb_seq_cb(struct dma_fence *fence, * @res: ttm_resource to map * @pages_addr: DMA addresses to use for mapping * @fence: optional resulting fence - * @table_freed: return true if page table is freed * * Fill in the page table entries between @start and @last. * @@ -823,8 +822,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, uint64_t flags, uint64_t offset, struct ttm_resource *res, dma_addr_t *pages_addr, - struct dma_fence **fence, - bool *table_freed) + struct dma_fence **fence) { struct amdgpu_vm_update_params params; struct amdgpu_vm_tlb_seq_cb *tlb_cb; @@ -937,9 +935,6 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, tlb_cb = NULL; } - if (table_freed) - *table_freed = *table_freed || params.table_freed; - error_free: kfree(tlb_cb); @@ -999,7 +994,6 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, * @adev: amdgpu_device pointer * @bo_va: requested BO and VM object * @clear: if true clear the entries - * @table_freed: return true if page table is freed * * Fill in the page table entries for @bo_va. * @@ -1007,7 +1001,7 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, * 0 for success, -EINVAL for failure. */ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, - bool clear, bool *table_freed) + bool clear) { struct amdgpu_bo *bo = bo_va->base.bo; struct amdgpu_vm *vm = bo_va->base.vm; @@ -1086,7 +1080,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, resv, mapping->start, mapping->last, update_flags, mapping->offset, mem, - pages_addr, last_update, table_freed); + pages_addr, last_update); if (r) return r; } @@ -1280,7 +1274,7 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, r = amdgpu_vm_bo_update_mapping(adev, adev, vm, false, false, resv, mapping->start, mapping->last, init_pte_value, - 0, NULL, NULL, &f, NULL); + 0, NULL, NULL, &f); amdgpu_vm_free_mapping(adev, vm, mapping, f); if (r) { dma_fence_put(f); @@ -1322,7 +1316,7 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, list_for_each_entry_safe(bo_va, tmp, &vm->moved, base.vm_status) { /* Per VM BOs never need to bo cleared in the page tables */ - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) return r; } @@ -1341,7 +1335,7 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, else clear = true; - r = amdgpu_vm_bo_update(adev, bo_va, clear, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, clear); if (r) return r; @@ -2525,8 +2519,7 @@ bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid, } r = amdgpu_vm_bo_update_mapping(adev, adev, vm, true, false, NULL, addr, - addr, flags, value, NULL, NULL, NULL, - NULL); + addr, flags, value, NULL, NULL, NULL); if (r) goto error_unlock; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 38a1eab1ff74..6b7682fe84f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -410,10 +410,10 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, uint64_t flags, uint64_t offset, struct ttm_resource *res, dma_addr_t *pages_addr, - struct dma_fence **fence, bool *free_table); + struct dma_fence **fence); int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, - bool clear, bool *table_freed); + bool clear); bool amdgpu_vm_evictable(struct amdgpu_bo *bo); void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, struct amdgpu_bo *bo, bool evicted); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 509d915cbe69..1cb0d811dde0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1190,7 +1190,7 @@ svm_range_unmap_from_gpu(struct amdgpu_device *adev, struct amdgpu_vm *vm, return amdgpu_vm_bo_update_mapping(adev, adev, vm, false, true, NULL, start, last, init_pte_value, 0, - NULL, NULL, fence, NULL); + NULL, NULL, fence); } static int @@ -1283,8 +1283,7 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, prange->start + i, pte_flags, last_start - prange->start, NULL, dma_addr, - &vm->last_update, - &table_freed); + &vm->last_update); for (j = last_start - prange->start; j <= i; j++) dma_addr[j] |= last_domain;
You already pointed out this bug, to flush TLB after waiting for vm update done. The below fix can be merged to this patch.
SVM is missing one optimization for mGPUs (Felix pointer out this
before), add fence of mGPUs to a sync, then wait for sync, instead
of waiting for fence of each GPU. This can be added later.
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c@@ -1243,7 +1243,6 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange,
{
struct amdgpu_device *adev = pdd->dev->adev;
struct amdgpu_vm *vm = drm_priv_to_vm(pdd->drm_priv);
- bool table_freed = false;
uint64_t pte_flags;
unsigned long last_start;
int last_domain;
@@ -1305,8 +1304,6 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange,
if (fence)
*fence = dma_fence_get(vm->last_update);
- if (table_freed)
- kfd_flush_tlb(pdd, TLB_FLUSH_LEGACY);
out:
return r;
}
@@ -1362,6 +1359,8 @@ svm_range_map_to_gpus(struct svm_range *prange, unsigned long offset,
break;
}
}
+
+ kfd_flush_tlb(pdd, TLB_FLUSH_LEGACY);
}
return r;
