From 8725aecac331954e0827d6bed7be02eb7b8f1e9e Mon Sep 17 00:00:00 2001 From: Yong Zhao Date: Fri, 13 Jul 2018 16:17:46 -0400 Subject: [PATCH] drm/amdkfd: Workaround to accommodate Raven too many PPR issue 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 Reviewed-by: Felix Kuehling Signed-off-by: Felix Kuehling Acked-by: Alex Deucher Signed-off-by: Oded Gabbay --- 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 820133cdef83..4dcacce2db86 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 */ -- 2.45.2