'Twas brillig, and xing wang at 25/07/11 02:27 did gyre and gimble: > Hi Pierre, Colin, > > Any comments? No specific comments from me as I'm not super clued up on that area. Pierre any specific comments at your end? Cheers. Col > 2011/7/22 xing wang <wangxingchao2011 at gmail.com> >> >> >> 2011/7/23 pl bossart <bossart.nospam at gmail.com> >>> >>>> When meet Overrun or Underrun, PA will increase watermark. Due to some >>>> hardware limitation, the hw_ptr will not be updated in PA's request latency. >>>> So there maybe some cases "faked overrun/underrun". Take capture for >>>> example, when Pulseaudio was waken up to read new capture data from driver, >>>> and find the "available" area is much more than the record space. Although >>>> it's fast enough to read out all data in short time, PA regard that as >>>> "overrun" and increase watermark/latency step by step, that's really >>>> useless. >>>> So attached patch used to avoid such case and only increase >>>> watermark/latency in real "overrun" case. >>> >>> Humm, it's not clear what problem you are trying to solve and on what >>> platform. Looks like a very old MRST firmware/driver? >>> If your driver doesn't update the hw_ptr, then how can you know if >>> this is a real or a fake overrun? >> >> In fact it's a solution for such firmware hasing low hw_ptr updating rate. >> Pulseaudio use hw_unused field to track the unused buffer, the real overrun >> means the "avail" space is geting close to hwbuf_size. >> --xingchao >>> >>> -Pierre >>> _______________________________________________ >>> pulseaudio-discuss mailing list >>> pulseaudio-discuss at lists.freedesktop.org >>> http://lists.freedesktop.org/mailman/listinfo/pulseaudio-discuss >> -- Colin Guthrie cguthrie(at)mandriva.org http://colin.guthr.ie/ Day Job: Tribalogic Limited [http://www.tribalogic.net/] Open Source: Mageia Contributor [http://www.mageia.org/] PulseAudio Hacker [http://www.pulseaudio.org/] Trac Hacker [http://trac.edgewall.org/]
