Subject: [ecasound] now that's reliable!
From: Kai Vehmanen (k_AT_eca.cx)
Date: Tue Oct 09 2001 - 18:21:51 EEST
Ok, now that the -z:nointbuf bug is behind us, we can get back to
improving the new buffering system. Here's a good example of its
current state:
Ecasound command-line (runs as root):
$ ecasound_debug -i smile.mp3 -o alsa -ea:100 -c
Top shows:
--cut--
6:11pm up 1 day, 9:55, 9 users, load average: 4.11, 2.35, 1.56
97 processes: 88 sleeping, 8 running, 1 zombie, 0 stopped
CPU states: 49.5% user, 50.4% system, 0.0% nice, 0.0% idle
Mem: 384848K av, 366420K used, 18428K free, 0K shrd, 4356K buff
Swap: 136544K av, 6812K used, 129732K free 157536K cached
PID USER PRI NI SIZE RSS SHARE FLAGS STAT %CPU %MEM COMMAND
18533 root 20 0 8032 8024 4660 100040 R 99.9 2.0 ecasound_debug
1401 kaiv 19 0 36584 34M 3908 0 R 24.9 9.2 netscape-naviga
18532 root 12 0 8032 8024 4660 140 S 10.4 2.0 ecasound_debug
19035 kaiv 18 0 3276 3276 1152 0 R 6.3 0.8 cc1
--cut--
Because of the '-ea:100', ecasound chooses the 'rtlowlatency' mode, which
means that minimal buffering is used. But even with small buffers,
runtime performance is rock solid. And as you can see, machine is
_heavily_ loaded (mozilla, compiling alsa-driver, compiling ecasound and
ecasound itself). User-interface responds slowly, but that's ok as long as
audio is ok.
PS I'm using a non-patched 2.4.10 smp-kernel and ALSA-CVS ~0.9beta8.
PPS When compiled with (--enable-debug, produces ecasound_debug),
ecasound runs _extremely_ slow (all debug-checks on, no optimizations,
nothing inlined).
-- http://www.eca.cx Audio software for Linux!-- To unsubscribe send message 'unsubscribe' in the body of the message to <ecasound-list-request_AT_wakkanet.fi>.
This archive was generated by hypermail 2b28 : Tue Oct 09 2001 - 18:19:07 EEST