github-mirror/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-05-23 22:52:19 +02:00
Code Issues Actions Packages Projects Releases Wiki Activity Graph

drm/amd/ras: Reduce stack usage in amdgpu_virt_ras_get_cper_records()

Browse Source 
amdgpu_virt_ras_get_cper_records() was using a large stack array
of ras_log_info pointers. This contributed to the frame size
warning on this function.

Replace the fixed-size stack array:

    struct ras_log_info *trace[MAX_RECORD_PER_BATCH];

with a heap-allocated array using kcalloc().

We free the trace buffer together with out_buf on all exit paths.
If allocation of trace or out_buf fails, we return a generic RAS
error code.

This reduces stack usage and keeps the runtime behaviour
unchanged.

Fixes:
stack frame size: 1112 bytes (limit: 1024)

Cc: Tao Zhou <tao.zhou1@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Srinivasan Shanmugam 2025-12-05 17:45:10 +05:30 committed by Alex Deucher
parent 5b57c3c3f2
commit 6ef93f6253
1 changed files with 14 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_virt_ras_cmd.c
View File

@ -183,7 +183,7 @@ static int amdgpu_virt_ras_get_cper_records(struct ras_core_context *ras_core,
(struct ras_cmd_cper_record_rsp *)cmd->output_buff_raw;
struct ras_log_batch_overview *overview = &virt_ras->batch_mgr.batch_overview;
struct ras_cmd_batch_trace_record_rsp *rsp_cache = &virt_ras->batch_mgr.batch_trace;
struct ras_log_info *trace[MAX_RECORD_PER_BATCH] = {0};
struct ras_log_info **trace;
uint32_t offset = 0, real_data_len = 0;
uint64_t batch_id;
uint8_t *out_buf;
@ -195,18 +195,25 @@ static int amdgpu_virt_ras_get_cper_records(struct ras_core_context *ras_core,
if (!req->buf_size || !req->buf_ptr || !req->cper_num)
return RAS_CMD__ERROR_INVALID_INPUT_DATA;
out_buf = kzalloc(req->buf_size, GFP_KERNEL);
if (!out_buf)
trace = kcalloc(MAX_RECORD_PER_BATCH, sizeof(*trace), GFP_KERNEL);
if (!trace)
return RAS_CMD__ERROR_GENERIC;
out_buf = kzalloc(req->buf_size, GFP_KERNEL);
if (!out_buf) {
kfree(trace);
return RAS_CMD__ERROR_GENERIC;
}
memset(out_buf, 0, req->buf_size);
for (i = 0; i < req->cper_num; i++) {
batch_id = req->cper_start_id + i;
if (batch_id >= overview->last_batch_id)
break;
count = amdgpu_virt_ras_get_batch_records(ras_core, batch_id, trace,
ARRAY_SIZE(trace), rsp_cache);
count = amdgpu_virt_ras_get_batch_records(ras_core, batch_id,
trace, MAX_RECORD_PER_BATCH,
rsp_cache);
if (count > 0) {
ret = ras_cper_generate_cper(ras_core, trace, count,
&out_buf[offset], req->buf_size - offset, &real_data_len);
@ -220,6 +227,7 @@ static int amdgpu_virt_ras_get_cper_records(struct ras_core_context *ras_core,
if ((ret && (ret != -ENOMEM)) ||
copy_to_user(u64_to_user_ptr(req->buf_ptr), out_buf, offset)) {
kfree(out_buf);
kfree(trace);
return RAS_CMD__ERROR_GENERIC;
}
@ -231,6 +239,7 @@ static int amdgpu_virt_ras_get_cper_records(struct ras_core_context *ras_core,
cmd->output_size = sizeof(struct ras_cmd_cper_record_rsp);
kfree(out_buf);
kfree(trace);
return RAS_CMD__SUCCESS;
}