Re: [PATCH] avfilter/dnn_backend_torch: fix missing ret in async error path
"Guo, Yejun via ffmpeg-devel" <[email protected]> Fri, 3 Jul 2026 13:26:02 +0000
| Newsgroups | gmane.comp.video.ffmpeg.devel |
|---|---|
| Message-ID | <PH7PR11MB595720B15E6C4C049D6722EDF1F42@PH7PR11MB5957.namprd11.prod.outlook.com> |
-----Original Message----- From: tknaveen via ffmpeg-devel <[email protected]> Sent: Thursday, April 30, 2026 7:14 PM To: [email protected] Cc: tknaveen <[email protected]> Subject: [FFmpeg-devel] [PATCH] avfilter/dnn_backend_torch: fix missing ret in async error path ff_dnn_start_inference_async() return value was not stored in ret, causing execute_model_th() to return success on async startup failure. This left vf_dnn_processing stuck in its flush loop waiting on a task that was never started. Fixes regression in 346fe01607ba. --- libavfilter/dnn/dnn_backend_torch.cpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/libavfilter/dnn/dnn_backend_torch.cpp b/libavfilter/dnn/dnn_backend_torch.cpp index 6cd93fd136..bba166ce5f 100644 --- a/libavfilter/dnn/dnn_backend_torch.cpp +++ b/libavfilter/dnn/dnn_backend_torch.cpp @@ -342,9 +342,9 @@ static int execute_model_th(THRequestItem *request, Queue *lltask_queue) } if (task->async) { - if (ff_dnn_start_inference_async(th_model->ctx, &request->exec_module) != 0) { + ret = ff_dnn_start_inference_async(th_model->ctx, &request->exec_module); + if (ret != 0) goto err; - } Nice catch, will rebase locally and push soon. Thanks. _______________________________________________ ffmpeg-devel mailing list -- [email protected] To unsubscribe send an email to [email protected]