When the uvc request will get parsed by uvc_video_decode_start it will leave the function with -EAGAIN to be restarted on the next frame. While the first wrong parse the statistics will already be updated with uvc_video_stats_decode. One value e.g. is the error_count, which therefor will be incremented twice in case the fid has changed on the way. This patch fixes the unnecessary extra parsing by moving the decode functions to the end of decode_start, when it is save to really parse the data. Signed-off-by: Michael Grzeschik <m.grzeschik@xxxxxxxxxxxxxx> --- Changes in v3: - Moved the clock_decode and stat_decode to end of function as suggested by Ricardo Ribalda - Link to v2: https://lore.kernel.org/r/20240221-uvc-host-video-decode-start-v2-1-88c6e17e487a@xxxxxxxxxxxxxx Changes in v2: - Moved the EAGAIN bailout after the sequence handling as mentioned by Ricardo Ribalda - Link to v1: https://lore.kernel.org/r/20240221-uvc-host-video-decode-start-v1-1-228995925c70@xxxxxxxxxxxxxx --- drivers/media/usb/uvc/uvc_video.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/media/usb/uvc/uvc_video.c b/drivers/media/usb/uvc/uvc_video.c index 7cbf4692bd875..7471bff0ca894 100644 --- a/drivers/media/usb/uvc/uvc_video.c +++ b/drivers/media/usb/uvc/uvc_video.c @@ -1078,9 +1078,6 @@ static int uvc_video_decode_start(struct uvc_streaming *stream, uvc_video_stats_update(stream); } - uvc_video_clock_decode(stream, buf, data, len); - uvc_video_stats_decode(stream, data, len); - /* * Store the payload FID bit and return immediately when the buffer is * NULL. @@ -1147,6 +1144,9 @@ static int uvc_video_decode_start(struct uvc_streaming *stream, return -EAGAIN; } + uvc_video_clock_decode(stream, buf, data, len); + uvc_video_stats_decode(stream, data, len); + stream->last_fid = fid; return data[0]; --- base-commit: d99e42ce6b8341d3f09e22c6706461ec900fe172 change-id: 20240221-uvc-host-video-decode-start-af53df5924cd Best regards, -- Michael Grzeschik <m.grzeschik@xxxxxxxxxxxxxx>