Hi Philipp, 2014-10-21 18:21 GMT+02:00 Philipp Zabel <p.zabel@xxxxxxxxxxxxxx>: > Hi Jean-Michel, > > Am Dienstag, den 21.10.2014, 17:39 +0200 schrieb Jean-Michel Hautbois: > [...] >> And the output is now : >> v4l2-ctl -d1 --stream-out-mmap --stream-mmap --stream-to x.raw >> [ 6208.240919] coda 2040000.vpu: Not output type >> [ 6208.245316] coda 2040000.vpu: streamon_out (N), streamon_cap (Y) >> [ 6208.251353] coda 2040000.vpu: fill bitstream >> [ 6208.255653] coda 2040000.vpu: fill bitstream payload : 0 >> VIDIOC_STREAMON: failed: Invalid argument >> >> Any idea ? >> JM > > $ trace-cmd record -e v4l2* v4l2-ctl -d13 --stream-out-mmap --stream-mmap --stream-to x.raw > [...] > $ trace-cmd report -R | grep bytesused > [...] > v4l2-ctl-308 [003] 1030.861067: v4l2_qbuf: minor=44 index=0 type=1 bytesused=0 flags=16387 field=0 timestamp=0 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.861292: v4l2_qbuf: minor=44 index=1 type=1 bytesused=0 flags=16387 field=0 timestamp=0 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.861471: v4l2_qbuf: minor=44 index=2 type=1 bytesused=0 flags=16387 field=0 timestamp=0 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.861638: v4l2_qbuf: minor=44 index=3 type=1 bytesused=0 flags=16387 field=0 timestamp=0 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.862301: v4l2_qbuf: minor=44 index=0 type=2 bytesused=3133440 flags=16387 field=1 timestamp=1030852944000 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.862490: v4l2_qbuf: minor=44 index=1 type=2 bytesused=3133440 flags=16387 field=1 timestamp=1030853139000 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.862672: v4l2_qbuf: minor=44 index=2 type=2 bytesused=3133440 flags=16387 field=1 timestamp=1030853322000 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > v4l2-ctl-308 [003] 1030.862841: v4l2_qbuf: minor=44 index=3 type=2 bytesused=3133440 flags=16387 field=1 timestamp=1030853491000 timecode_type=0 timecode_flags=0 timecode_frames=0 timecode_seconds=0 timecode_minutes=0 timecode_hours=0 timecode_userbits0=0 timecode_userbits1=0 timecode_userbits2=0 timecode_userbits3=0 sequence=0 > > The decoder is fed ~ 3 MiB input buffers, which it tries (and fails) to > copy into the 1 MiB bitstream ringbuffer (currently hard-coded via the > badly named CODA_MAX_FRAME_SIZE constant), so the bitstream payload in > the ringbuffer is 0 during start_streaming. Mmmh, nice, didn't thought to get perf out of there :). Well, I understand it can't feed the ringbuffer, but is there a way to make the encoder work ? I could of course modify CODA_MAX_FRAME_SIZE but this is clearly not the good thing to do... Thanks, JM -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
