[MPlayer-cvslog] r26345 - trunk/libao2/ao_sun.c

Jan Knutar jknutar at nic.fi
Sat Feb 7 17:01:48 CET 2009


On Monday 07 April 2008, reimar wrote:
> Author: reimar
> Date: Mon Apr  7 19:39:41 2008
> New Revision: 26345
>
> Log:
> reset() should not senselessly close and reopen
> the device but instead just call flush_audio()

This seems to have broken seeking on Opensolaris, from irc:

<freakabcd> i hope i'm saying it(pts) correctly. This is the output of 
pressing up(w) arrow:
<freakabcd> A:-97325.0 V:  64.5 A-V:-97389.500 ct:  0.000   0/  0 ??% 
??% ??,?% 0 0
<freakabcd> it is audio.
<freakabcd> if i do -ao null, i can seek perfectly.
<freakabcd> it is using -ao sun by default
<freakabcd> audio works good when playing without any seeking. but once 
i seek, everything just hangs and i have to interrupt and kill
<ShadowJK> I guess you could svn log ao_sun.c and svn up -r 26344 
ao_sun.c and go back a few times to check if anything recent-ish broke 
it
<freakabcd> yeah, ao_sun is the problem now. how do i find out why it 
goes -ve? i.e. output shows A:-97378.2 ... or other huge -ve numbers
<freakabcd> +6344 is the one you remember working for ao_sun?
<freakabcd> +'ll certainly try that now
<freakabcd> maybe it wasn;t on mplayer's side and the sun guys 
broke/changed something
<ShadowJK> No it was so long ago I don't even remember
<ShadowJK> I think it was Solaris 9 too
<freakabcd> why did you recommend 26344 then?
<-- speps has quit (Read error: 110 (Connection timed out))
<ShadowJK> Seemed like the first non-cosmetic change in the list
<freakabcd> was that just a random number since it was far back in the 
past? with regards to ao_sun?
<freakabcd> ah, ok. i will get that

<freakabcd> ShadowJK, ok. it was 26345 that screwed it
<freakabcd> 26344 ao_sun.c works perfectly with seeking.
<ShadowJK> freakabcd, I'll reply to that commit msg
<freakabcd> ShadowJK, make sure you mention opensolaris 2008.11 on x86
<freakabcd> cos i don;t want reimar going 'it workes perfectly on 
solaris ...' :)


> Modified:
>    trunk/libao2/ao_sun.c
>
> Modified: trunk/libao2/ao_sun.c
> =====================================================================
>========= --- trunk/libao2/ao_sun.c	(original)
> +++ trunk/libao2/ao_sun.c	Mon Apr  7 19:39:41 2008
> @@ -588,29 +588,7 @@ static void uninit(int immed){
>
>  // stop playing and empty buffers (for seeking/pause)
>  static void reset(void){
> -    audio_info_t info;
> -
> -    uninit(1);
> -    audio_fd=open(audio_dev, O_WRONLY);
> -    if(audio_fd<0){
> -	mp_msg(MSGT_AO, MSGL_FATAL, MSGTR_AO_SUN_CantReopenReset,
> strerror(errno)); -	return;
> -    }
> -
> -    ioctl(audio_fd, AUDIO_DRAIN, 0);
> -
> -    AUDIO_INITINFO(&info);
> -    info.play.encoding = af2sunfmt(ao_data.format);
> -    info.play.precision =
> -	(ao_data.format==AF_FORMAT_S16_NE
> -	 ? AUDIO_PRECISION_16
> -	 : AUDIO_PRECISION_8);
> -    info.play.channels = ao_data.channels;
> -    info.play.sample_rate = ao_data.samplerate;
> -    info.play.samples = 0;
> -    info.play.eof = 0;
> -    info.play.error = 0;
> -    ioctl (audio_fd, AUDIO_SETINFO, &info);
> +    flush_audio(audio_fd);
>      queued_bursts = 0;
>      queued_samples = 0;
>  }
> _______________________________________________
> MPlayer-cvslog mailing list
> MPlayer-cvslog at mplayerhq.hu
> https://lists.mplayerhq.hu/mailman/listinfo/mplayer-cvslog





More information about the MPlayer-cvslog mailing list