On all the platforms I tested, PulseAudio is frequently awaken and doesn't sleep for the duration specified for the poll timeout. Sometimes wake-ups occur within milliseconds of the poll call for no good reason; this seems to be related to ALSA issues (see my posts on the ALSA mailing list on null poll events). This patch enables a better log of requested sleep times v. actual sleep times. Enable DEBUG_TIMING to see actual messages. Please let me know if you see odd behaviors like the one below - Pierre E: rtpoll.c: rtpoll_run E: rtpoll.c: poll timeout: 188 ms E: rtpoll.c: Process time 0 ms; sleep time 48 ms E: rtpoll.c: rtpoll_run E: rtpoll.c: rtpoll finish E: rtpoll.c: rtpoll_run E: rtpoll.c: poll timeout: 139 ms E: rtpoll.c: Process time 0 ms; sleep time 49 ms E: rtpoll.c: rtpoll_run E: rtpoll.c: rtpoll finish E: rtpoll.c: rtpoll_run E: rtpoll.c: poll timeout: 189 ms E: rtpoll.c: Process time 0 ms; sleep time 0 ms E: rtpoll.c: rtpoll_run E: rtpoll.c: rtpoll finish E: rtpoll.c: rtpoll_run E: rtpoll.c: poll timeout: 189 ms E: rtpoll.c: Process time 0 ms; sleep time 49 ms -------------- next part -------------- A non-text attachment was scrubbed... Name: 0001-rtpoll-better-support-for-DEBUG_TIMING-logs.patch Type: text/x-patch Size: 2093 bytes Desc: not available URL: <http://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20091223/3cde1b88/attachment.bin>
