Skip to content

Commit 2d300eb

Browse files
ruijingdongalexdeucher
authored andcommitted
drm/amdgpu: fix strsep() corrupting lockup_timeout on multi-GPU (v3)
amdgpu_device_get_job_timeout_settings() passes a pointer directly to the global amdgpu_lockup_timeout[] buffer into strsep(). strsep() destructively replaces delimiter characters with '\0' in-place. On multi-GPU systems, this function is called once per device. When a multi-value setting like "0,0,0,-1" is used, the first GPU's call transforms the global buffer into "0\00\00\0-1". The second GPU then sees only "0" (terminated at the first '\0'), parses a single value, hits the single-value fallthrough (index == 1), and applies timeout=0 to all rings — causing immediate false job timeouts. Fix this by copying into a stack-local array before calling strsep(), so the global module parameter buffer remains intact across calls. The buffer is AMDGPU_MAX_TIMEOUT_PARAM_LENGTH (256) bytes, which is safe for the stack. v2: wrap commit message to 72 columns, add Assisted-by tag. v3: use stack array with strscpy() instead of kstrdup()/kfree() to avoid unnecessary heap allocation (Christian). This patch was developed with assistance from Claude (claude-opus-4-6). Assisted-by: Claude:claude-opus-4-6 Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Ruijing Dong <ruijing.dong@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> (cherry picked from commit 94d79f5) Cc: stable@vger.kernel.org
1 parent aed3d04 commit 2d300eb

1 file changed

Lines changed: 11 additions & 2 deletions

File tree

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

Lines changed: 11 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -4207,7 +4207,8 @@ static void amdgpu_device_xgmi_reset_func(struct work_struct *__work)
42074207

42084208
static int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
42094209
{
4210-
char *input = amdgpu_lockup_timeout;
4210+
char buf[AMDGPU_MAX_TIMEOUT_PARAM_LENGTH];
4211+
char *input = buf;
42114212
char *timeout_setting = NULL;
42124213
int index = 0;
42134214
long timeout;
@@ -4217,9 +4218,17 @@ static int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
42174218
adev->gfx_timeout = adev->compute_timeout = adev->sdma_timeout =
42184219
adev->video_timeout = msecs_to_jiffies(2000);
42194220

4220-
if (!strnlen(input, AMDGPU_MAX_TIMEOUT_PARAM_LENGTH))
4221+
if (!strnlen(amdgpu_lockup_timeout, AMDGPU_MAX_TIMEOUT_PARAM_LENGTH))
42214222
return 0;
42224223

4224+
/*
4225+
* strsep() destructively modifies its input by replacing delimiters
4226+
* with '\0'. Use a stack copy so the global module parameter buffer
4227+
* remains intact for multi-GPU systems where this function is called
4228+
* once per device.
4229+
*/
4230+
strscpy(buf, amdgpu_lockup_timeout, sizeof(buf));
4231+
42234232
while ((timeout_setting = strsep(&input, ",")) &&
42244233
strnlen(timeout_setting, AMDGPU_MAX_TIMEOUT_PARAM_LENGTH)) {
42254234
ret = kstrtol(timeout_setting, 0, &timeout);

0 commit comments

Comments
 (0)