On 2011-09-26 09:58, Andreas Färber wrote: > Hello Jan, > > Am 20.09.2011 um 18:53 schrieb Jan Kiszka: > >> Using the error management of QemuMutex allows to simplify the code. >> >> CC: malc <av1...@comtv.ru> >> CC: Andreas Färber <andreas.faer...@web.de> >> Signed-off-by: Jan Kiszka <jan.kis...@siemens.com> >> --- >> audio/coreaudio.c | 56 +++++++--------------------------------------------- >> 1 files changed, 8 insertions(+), 48 deletions(-) > > You missed one coreaudio_unlock() occurrence in line 187. Other than that > looks okay, with some more softfloat workarounds it compiles and doesn't > noticeably regress on Darwin/i386 v10.6.8. > > diff --git a/audio/coreaudio.c b/audio/coreaudio.c > index c34a593..1cb3fce 100644 > --- a/audio/coreaudio.c > +++ b/audio/coreaudio.c > @@ -184,7 +184,7 @@ static int coreaudio_run_out (HWVoiceOut *hw, int live) > core->live = live - decr; > hw->rpos = core->rpos; > > - coreaudio_unlock (core, "coreaudio_run_out"); > + qemu_mutex_unlock(&core->mutex); > return decr; > } > > > Do you have a particular test case? Any parameters needed? I've never used > audio in my guests. ;)
Nope, I hoped you had. :) But I received indications that the implicit logging removal is not welcome. So I will likely refrain from touching the audio subsystem in the first step. Thanks, Jan -- Siemens AG, Corporate Technology, CT T DE IT 1 Corporate Competence Center Embedded Linux
