leandrof/linux
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

drm/amdgpu: Print the IH client ID name when vm fault happens

Browse Source 
This gives more information and improves productivity.

Signed-off-by: Yong Zhao <Yong.Zhao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Yong Zhao
2020-08-21 14:01:18 +08:00
committed by Alex Deucher
parent 0ef6845c8c
commit be14729a33
4 changed files with 46 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

34
drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
View File

@@ -65,6 +65,40 @@
#define AMDGPU_WAIT_IDLE_TIMEOUT 200 #define AMDGPU_WAIT_IDLE_TIMEOUT 200
const char *soc15_ih_clientid_name[] = {
"IH",
"SDMA2 or ACP",
"ATHUB",
"BIF",
"SDMA3 or DCE",
"SDMA4 or ISP",
"VMC1 or PCIE0",
"RLC",
"SDMA0",
"SDMA1",
"SE0SH",
"SE1SH",
"SE2SH",
"SE3SH",
"VCN1 or UVD1",
"THM",
"VCN or UVD",
"SDMA5 or VCE0",
"VMC",
"SDMA6 or XDMA",
"GRBM_CP",
"ATS",
"ROM_SMUIO",
"DF",
"SDMA7 or VCE1",
"PWR",
"UTCL2",
"EA",
"UTCL2LOG",
"MP0",
"MP1"
};
/** /**
* amdgpu_hotplug_work_func - work handler for display hotplug event * amdgpu_hotplug_work_func - work handler for display hotplug event
* *

5
drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
View File

@@ -152,8 +152,9 @@ static int gmc_v10_0_process_interrupt(struct amdgpu_device *adev,
entry->src_id, entry->ring_id, entry->vmid, entry->src_id, entry->ring_id, entry->vmid,
entry->pasid, task_info.process_name, task_info.tgid, entry->pasid, task_info.process_name, task_info.tgid,
task_info.task_name, task_info.pid); task_info.task_name, task_info.pid);
dev_err(adev->dev, " in page starting at address 0x%012llx from client %d\n", dev_err(adev->dev, " in page starting at address 0x%016llx from client 0x%x (%s)\n",
addr, entry->client_id); addr, entry->client_id,
soc15_ih_clientid_name[entry->client_id]);
if (!amdgpu_sriov_vf(adev)) if (!amdgpu_sriov_vf(adev))
hub->vmhub_funcs->print_l2_protection_fault_status(adev, hub->vmhub_funcs->print_l2_protection_fault_status(adev,

5
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
View File

@@ -521,8 +521,9 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev,
entry->src_id, entry->ring_id, entry->vmid, entry->src_id, entry->ring_id, entry->vmid,
entry->pasid, task_info.process_name, task_info.tgid, entry->pasid, task_info.process_name, task_info.tgid,
task_info.task_name, task_info.pid); task_info.task_name, task_info.pid);
dev_err(adev->dev, " in page starting at address 0x%012llx from client %d\n", dev_err(adev->dev, " in page starting at address 0x%016llx from IH client 0x%x (%s)\n",
addr, entry->client_id); addr, entry->client_id,
soc15_ih_clientid_name[entry->client_id]);
if (amdgpu_sriov_vf(adev)) if (amdgpu_sriov_vf(adev))
return 0; return 0;

6
drivers/gpu/drm/amd/include/soc15_ih_clientid.h
View File

@@ -25,7 +25,9 @@
#define __SOC15_IH_CLIENTID_H__ #define __SOC15_IH_CLIENTID_H__
/* /*
* vega10+ IH clients * Vega10+ IH clients
* Whenever this structure is updated, which should not happen, make sure
* soc15_ih_clientid_name in the below is also updated accordingly.
*/ */
enum soc15_ih_clientid { enum soc15_ih_clientid {
SOC15_IH_CLIENTID_IH = 0x00, SOC15_IH_CLIENTID_IH = 0x00,
@@ -74,6 +76,8 @@ enum soc15_ih_clientid {
SOC15_IH_CLIENTID_VMC1 = SOC15_IH_CLIENTID_PCIE0, SOC15_IH_CLIENTID_VMC1 = SOC15_IH_CLIENTID_PCIE0,
}; };
extern const char *soc15_ih_clientid_name[];
#endif #endif