drm/amdkfd: add RAS ECC event support (v3)

RAS ECC event will combine with GPU reset event, due to
ECC interrupts are caused by uncorrectable error that triggers
GPU reset.

v2: Fix misleading-indentation warning
v3: fix build with CONFIG_HSA_AMD disabled

Signed-off-by: Eric Huang <JinhuiEric.Huang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Eric Huang
2019-01-11 14:38:51 -05:00
committed by Alex Deucher
parent 0dee45a25a
commit 9b54d20176
9 changed files with 51 additions and 2 deletions

View File

@@ -211,6 +211,11 @@ struct kfd_ioctl_dbg_wave_control_args {
#define KFD_HW_EXCEPTION_GPU_HANG 0
#define KFD_HW_EXCEPTION_ECC 1
/* For kfd_hsa_memory_exception_data.ErrorType */
#define KFD_MEM_ERR_NO_RAS 0
#define KFD_MEM_ERR_SRAM_ECC 1
#define KFD_MEM_ERR_POISON_CONSUMED 2
#define KFD_MEM_ERR_GPU_HANG 3
struct kfd_ioctl_create_event_args {
__u64 event_page_offset; /* from KFD */
@@ -250,7 +255,12 @@ struct kfd_hsa_memory_exception_data {
struct kfd_memory_exception_failure failure;
__u64 va;
__u32 gpu_id;
__u32 pad;
__u32 ErrorType; /* 0 = no RAS error,
* 1 = ECC_SRAM,
* 2 = Link_SYNFLOOD (poison),
* 3 = GPU hang (not attributable to a specific cause),
* other values reserved
*/
};
/* hw exception data */