[alsa-devel] [PATCH] alsa-lib: Add htimestamp operation in plugin file

[Takashi Iwai]

On Tue, 29 Nov 2016 16:15:57 +0100,
Takashi Sakamoto wrote:
> 
> Hi,
> 
> On Nov 29 2016 04:18, Takashi Iwai wrote:
> > On Fri, 25 Nov 2016 12:36:37 +0100,
> > sutar.mounesh at gmail.com wrote:
> >>
> >> From: Andreas Pape <apape at de.adit-jv.com>
> >>
> >> PCM operation htimestamp is not implemented in plugin file.
> >> Calling snd_pcm_htimestamp() on a plugin file crashes. This scenario
> >> is considered now.
> >>
> >> Signed-off-by: Andreas Pape <apape at de.adit-jv.com>
> >> Signed-off-by: Joshua Frkuska <joshua_frkuska at mentor.com>
> >
> > Applied, thanks.
> >
> >
> > Takashi
> >
> >> ---
> >>
> >> --- a/src/pcm/pcm_file.c	2013-07-08 14:31:36.000000000 +0200
> >> +++ b/src/pcm/pcm_file.c	2015-05-04 16:26:10.413615403 +0200
> >> @@ -698,6 +698,7 @@
> >>  	.readi = snd_pcm_file_readi,
> >>  	.readn = snd_pcm_file_readn,
> >>  	.avail_update = snd_pcm_generic_avail_update,
> >> +	.htimestamp = snd_pcm_generic_htimestamp,
> >>  	.mmap_commit = snd_pcm_file_mmap_commit,
> >>  	.poll_descriptors_count = snd_pcm_generic_poll_descriptors_count,
> >>  	.poll_descriptors = snd_pcm_generic_poll_descriptors,
> 
> I oppose this application, because designated initialization is
> already applied to the .htimestamp member.
> 
> ...
>         .poll_descriptors = snd_pcm_generic_poll_descriptors,
>         .poll_revents = snd_pcm_generic_poll_revents,
>         .htimestamp = snd_pcm_generic_htimestamp,
> };
> 
> Please see below commit:
> pcm:file: add the missing htimestamp callback
> http://git.alsa-project.org/?p=alsa-lib.git;a=commit;h=68ae0c72a53704d416fe77d4c612d1f88d791e02
> 
> This mistake causes below warning with -Woverride-init option.
> pcm_file.c:714:16: warning: initialized field overwritten [-Woverride-init]
>   .htimestamp = snd_pcm_generic_real_htimestamp,
>                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> I guess mentor/ADIT developers works for former snapshot, then missed
> to rebase to current mainline of alsa-lib. I suggest them to di
> re-evaluation with current mainline without this patch.

I overlooked it, too.  Now the commit got reverted.

Thanks!


Takashi