6.6-stable review patch. If anyone has any objections, please let me know. ------------------ From: Priya Hosur [ Upstream commit 94e25cb6ab7f4f025bcdcd8ea79fda30f12843a4 ] MES (Micro Engine Scheduler) does not perform heavy-weight TLB invalidation after unmapping queues, unlike HWS which does this automatically. This causes a race condition where in-flight DMA descriptors can access memory that has been unmapped, leading to page faults and GPU queue hangs during SVM page migration. The issue manifests as KFDSVMRangeTest.MultiThreadMigrationTest failures on gfx1151 (Strix Point) with XNACK mode 1 enabled - the GPU compute queue hangs with packets submitted but never consumed. Add kfd_flush_tlb() calls after MES queue removal in two locations: - evict_process_queues_cpsch(): after all queues removed during eviction - suspend_queues(): after debug/criu queue suspension (with mem_fence barrier) This ensures all in-flight memory accesses from unmapped queues are flushed before memory is freed or migrated. Signed-off-by: Priya Hosur Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher (cherry picked from commit f5c4f88e0f9c45a8fb9dfac0c1df726c95e41b77) Cc: stable@vger.kernel.org Signed-off-by: Sasha Levin --- .../gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 6b274e4f0af35..7e4bba9d07f8b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -1151,6 +1151,14 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm, } } } + /* + * Heavy-weight TLB flush after MES removes queues to ensure + * in-flight memory accesses complete before memory is freed/migrated. + * HWS does this automatically, MES does not. + */ + if (dqm->dev->kfd->shared_resources.enable_mes) + kfd_flush_tlb(pdd, TLB_FLUSH_HEAVYWEIGHT); + pdd->last_evict_timestamp = get_jiffies_64(); if (!dqm->dev->kfd->shared_resources.enable_mes) retval = execute_queues_cpsch(dqm, @@ -2992,8 +3000,12 @@ int suspend_queues(struct kfd_process *p, if (!per_device_suspended) { dqm_unlock(dqm); mutex_unlock(&p->event_mutex); - if (total_suspended) + if (total_suspended) { amdgpu_amdkfd_debug_mem_fence(dqm->dev->adev); + /* Heavy-weight TLB flush after MES suspends queues */ + if (pdd->drm_priv) + kfd_flush_tlb(pdd, TLB_FLUSH_HEAVYWEIGHT); + } continue; } -- 2.53.0