From: Yong Zhao <yong.zhao@xxxxxxx> On Raven multiple PPRs can be queued up by the hardware. When the first of those requests is handled by the IOMMU driver, the memory access succeeds. After that the application may be done with the memory and unmap it. At that point the page table entries are invalidated, but there are still outstanding duplicate PPRs for those addresses. When the IOMMU driver processes those duplicate requests, it finds invalid page table entries and triggers an invalid PPR fault. As a workaround, don't signal invalid PPR faults on Raven to avoid segfaulting applications that haven't done anything wrong. As a side effect, real GPU memory access faults may go unnoticed by the application. Signed-off-by: Yong Zhao <yong.zhao at amd.com> Reviewed-by: Felix Kuehling <Felix.Kuehling at amd.com> Signed-off-by: Felix Kuehling <Felix.Kuehling at amd.com> --- drivers/gpu/drm/amd/amdkfd/kfd_events.c | 21 ++++++++++++++++----- 1 file changed, 16 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 820133c..4dcacce 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -932,13 +932,24 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, unsigned int pasid, up_read(&mm->mmap_sem); mmput(mm); - mutex_lock(&p->event_mutex); + pr_debug("notpresent %d, noexecute %d, readonly %d\n", + memory_exception_data.failure.NotPresent, + memory_exception_data.failure.NoExecute, + memory_exception_data.failure.ReadOnly); - /* Lookup events by type and signal them */ - lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, - &memory_exception_data); + /* Workaround on Raven to not kill the process when memory is freed + * before IOMMU is able to finish processing all the excessive PPRs + */ + if (dev->device_info->asic_family != CHIP_RAVEN) { + mutex_lock(&p->event_mutex); + + /* Lookup events by type and signal them */ + lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, + &memory_exception_data); + + mutex_unlock(&p->event_mutex); + } - mutex_unlock(&p->event_mutex); kfd_unref_process(p); } #endif /* KFD_SUPPORT_IOMMU_V2 */ -- 2.7.4