6.12-stable review patch. If anyone has any objections, please let me know. ------------------ From: Olivier Kaloudoff [ Upstream commit 3254ed0b15f2fe656c9cb36ccb59b92422440bf8 ] amdgpu_amdkfd_restore_userptr_worker() and svm_range_restore_work() reschedule themselves when validation fails with evicted objects pending. After a surprise removal the restore paths fail with -ENODEV permanently, so rescheduling just spins the worker uselessly until cancel_delayed_work_sync() in fini finally stops it - which can take a while, since process teardown itself is delayed by the dead device. Treat -ENODEV as terminal in both workers. Cc: Xiaogang Chen Signed-off-by: Olivier Kaloudoff Signed-off-by: Felix Kuehling Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher (cherry picked from commit 266fab83139600e58db7e900e930c722e619b78c) Cc: stable@vger.kernel.org Signed-off-by: Sasha Levin --- .../gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 18 +++++++++++++----- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 8 +++++--- 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 6dabdf3fd30dc..bc284939cf66d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -2796,6 +2796,7 @@ static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work) struct task_struct *usertask; struct mm_struct *mm; uint32_t evicted_bos; + int ret; mutex_lock(&process_info->notifier_lock); evicted_bos = process_info->evicted_bos; @@ -2815,14 +2816,16 @@ static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work) mutex_lock(&process_info->lock); - if (update_invalid_user_pages(process_info, mm)) + ret = update_invalid_user_pages(process_info, mm); + if (ret) goto unlock_out; /* userptr_inval_list can be empty if all evicted userptr BOs * have been freed. In that case there is nothing to validate * and we can just restart the queues. */ if (!list_empty(&process_info->userptr_inval_list)) { - if (validate_invalid_user_pages(process_info)) + ret = validate_invalid_user_pages(process_info); + if (ret) goto unlock_out; } /* Final check for concurrent evicton and atomic update. If @@ -2831,11 +2834,14 @@ static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work) * reference counting inside KFD will handle this case. */ mutex_lock(&process_info->notifier_lock); - if (process_info->evicted_bos != evicted_bos) + if (process_info->evicted_bos != evicted_bos) { + ret = -EAGAIN; goto unlock_notifier_out; + } if (confirm_valid_user_pages_locked(process_info)) { WARN(1, "User pages unexpectedly invalid"); + ret = -EINVAL; goto unlock_notifier_out; } @@ -2853,8 +2859,10 @@ static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work) unlock_out: mutex_unlock(&process_info->lock); - /* If validation failed, reschedule another attempt */ - if (evicted_bos) { + /* If validation failed, reschedule another attempt, unless the + * device is gone for good + */ + if (evicted_bos && ret != -ENODEV) { queue_delayed_work(system_freezable_wq, &process_info->restore_userptr_work, msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS)); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 0b86da7d7c502..27085694697a2 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1844,7 +1844,7 @@ static void svm_range_restore_work(struct work_struct *work) struct mm_struct *mm; int evicted_ranges; int invalid; - int r; + int r = 0; svms = container_of(dwork, struct svm_range_list, restore_work); evicted_ranges = atomic_read(&svms->evicted_ranges); @@ -1918,8 +1918,10 @@ static void svm_range_restore_work(struct work_struct *work) mmap_write_unlock(mm); mutex_unlock(&process_info->lock); - /* If validation failed, reschedule another attempt */ - if (evicted_ranges) { + /* If validation failed, reschedule another attempt, unless the + * device is gone for good + */ + if (evicted_ranges && r != -ENODEV) { pr_debug("reschedule to restore svm range\n"); queue_delayed_work(system_freezable_wq, &svms->restore_work, msecs_to_jiffies(AMDGPU_SVM_RANGE_RESTORE_DELAY_MS)); -- 2.53.0