diff options
| author | Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com> | 2024-07-18 14:05:48 -0700 |
|---|---|---|
| committer | Lucas De Marchi <lucas.demarchi@intel.com> | 2024-07-19 00:32:37 -0700 |
| commit | 2149ded63079449b8dddf9da38392632f155e6b5 (patch) | |
| tree | 3e0c3ea5150be185844336bc1ca6339004a7108a /drivers/gpu/drm/xe/xe_exec_queue.c | |
| parent | a2387e69493df3de706f14e4573ee123d23d5d34 (diff) | |
| download | linux-2149ded63079449b8dddf9da38392632f155e6b5.tar.gz linux-2149ded63079449b8dddf9da38392632f155e6b5.tar.bz2 linux-2149ded63079449b8dddf9da38392632f155e6b5.zip | |
drm/xe: Fix use after free when client stats are captured
xe_file_close triggers an asynchronous queue cleanup and then frees up
the xef object. Since queue cleanup flushes all pending jobs and the KMD
stores client usage stats into the xef object after jobs are flushed, we
see a use-after-free for the xef object. Resolve this by taking a
reference to xef from xe_exec_queue.
While at it, revert an earlier change that contained a partial work
around for this issue.
v2:
- Take a ref to xef even for the VM bind queue (Matt)
- Squash patches relevant to that fix and work around (Lucas)
v3: Fix typo (Lucas)
Fixes: ce62827bc294 ("drm/xe: Do not access xe file when updating exec queue run_ticks")
Fixes: 6109f24f87d7 ("drm/xe: Add helper to accumulate exec queue runtime")
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/issues/1908
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240718210548.3580382-5-umesh.nerlige.ramappa@intel.com
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Diffstat (limited to 'drivers/gpu/drm/xe/xe_exec_queue.c')
| -rw-r--r-- | drivers/gpu/drm/xe/xe_exec_queue.c | 10 |
1 files changed, 9 insertions, 1 deletions
diff --git a/drivers/gpu/drm/xe/xe_exec_queue.c b/drivers/gpu/drm/xe/xe_exec_queue.c index 3336a01a1006..69867a7b7c77 100644 --- a/drivers/gpu/drm/xe/xe_exec_queue.c +++ b/drivers/gpu/drm/xe/xe_exec_queue.c @@ -37,6 +37,10 @@ static void __xe_exec_queue_free(struct xe_exec_queue *q) { if (q->vm) xe_vm_put(q->vm); + + if (q->xef) + xe_file_put(q->xef); + kfree(q); } @@ -649,6 +653,7 @@ int xe_exec_queue_create_ioctl(struct drm_device *dev, void *data, goto kill_exec_queue; args->exec_queue_id = id; + q->xef = xe_file_get(xef); return 0; @@ -762,6 +767,7 @@ bool xe_exec_queue_is_idle(struct xe_exec_queue *q) */ void xe_exec_queue_update_run_ticks(struct xe_exec_queue *q) { + struct xe_file *xef; struct xe_lrc *lrc; u32 old_ts, new_ts; @@ -773,6 +779,8 @@ void xe_exec_queue_update_run_ticks(struct xe_exec_queue *q) if (!q->vm || !q->vm->xef) return; + xef = q->vm->xef; + /* * Only sample the first LRC. For parallel submission, all of them are * scheduled together and we compensate that below by multiplying by @@ -783,7 +791,7 @@ void xe_exec_queue_update_run_ticks(struct xe_exec_queue *q) */ lrc = q->lrc[0]; new_ts = xe_lrc_update_timestamp(lrc, &old_ts); - q->run_ticks += (new_ts - old_ts) * q->width; + xef->run_ticks[q->class] += (new_ts - old_ts) * q->width; } void xe_exec_queue_kill(struct xe_exec_queue *q) |
