Merged upstream to topic/asoc branch

Merged upstream to topic/asoc branch

[Takashi Iwai]

Hi,

FYI, I merged the current Linus tree back to topic/asoc branch on
sound git tree for clean-up works including the latest commit in 
sound/soc/fsl.

If you have any problems with it, please let me know.


thanks,

Takashi

Re: Merged upstream to topic/asoc branch

[Takashi Iwai]

At Mon, 01 Dec 2008 18:00:04 +0100,
I wrote:
> 
> Hi,
> 
> FYI, I merged the current Linus tree back to topic/asoc branch on
> sound git tree for clean-up works including the latest commit in 
> sound/soc/fsl.

Oops, there was the latest update of sound/sparc on Linus tree, so
I re-merged again...


Takashi

Re: Merged upstream to topic/asoc branch

[Mark Brown]

On Mon, Dec 01, 2008 at 06:00:04PM +0100, Takashi Iwai wrote:

> FYI, I merged the current Linus tree back to topic/asoc branch on
> sound git tree for clean-up works including the latest commit in 
> sound/soc/fsl.

> If you have any problems with it, please let me know.

Hrm.  Which compiler are you using?  The WM8903 warning patch you
introduced is not needed on any of my systems (primarily ARM and
x86-64).  It smells like a problem with the BUG() definition on your
platform...

Re: Merged upstream to topic/asoc branch

[Takashi Iwai]

At Mon, 1 Dec 2008 17:06:34 +0000,
Mark Brown wrote:
> 
> On Mon, Dec 01, 2008 at 06:00:04PM +0100, Takashi Iwai wrote:
> 
> > FYI, I merged the current Linus tree back to topic/asoc branch on
> > sound git tree for clean-up works including the latest commit in 
> > sound/soc/fsl.
> 
> > If you have any problems with it, please let me know.
> 
> Hrm.  Which compiler are you using?

gcc-4.1 for ARM, if I remember correctly...

>  The WM8903 warning patch you
> introduced is not needed on any of my systems (primarily ARM and
> x86-64).  It smells like a problem with the BUG() definition on your
> platform...

Possibly.  I'm not pretty sure on which systems, since I run test-builds
regularly for ARM, PPC, x86 and x86-64 with cross compilers.


thanks,

Takashi

Re: Merged upstream to topic/asoc branch

[Mark Brown]

On Mon, Dec 01, 2008 at 06:14:12PM +0100, Takashi Iwai wrote:
> Mark Brown wrote:

> > Hrm.  Which compiler are you using?

> gcc-4.1 for ARM, if I remember correctly...

I'm at 4.2.4 for ARM.

> >  The WM8903 warning patch you
> > introduced is not needed on any of my systems (primarily ARM and
> > x86-64).  It smells like a problem with the BUG() definition on your
> > platform...

> Possibly.  I'm not pretty sure on which systems, since I run test-builds
> regularly for ARM, PPC, x86 and x86-64 with cross compilers.

Could you investigate, please?  I'd rather not add annotations like that
since they mask genuine compiler warnings and are liable to crop up
elsewhere.  That particular function isn't likely to change but I'm
worried in case this sort of thing starts cropping up elsewhere.

For that particular function I'd rather add return statements after the
BUG()s...

As far as I'm aware the current ASoC code is warning free.

Re: Merged upstream to topic/asoc branch

[Takashi Iwai]

At Mon, 1 Dec 2008 17:24:55 +0000,
Mark Brown wrote:
> 
> On Mon, Dec 01, 2008 at 06:14:12PM +0100, Takashi Iwai wrote:
> > Mark Brown wrote:
> 
> > > Hrm.  Which compiler are you using?
> 
> > gcc-4.1 for ARM, if I remember correctly...
> 
> I'm at 4.2.4 for ARM.
> 
> > >  The WM8903 warning patch you
> > > introduced is not needed on any of my systems (primarily ARM and
> > > x86-64).  It smells like a problem with the BUG() definition on your
> > > platform...
> 
> > Possibly.  I'm not pretty sure on which systems, since I run test-builds
> > regularly for ARM, PPC, x86 and x86-64 with cross compilers.
> 
> Could you investigate, please?  I'd rather not add annotations like that
> since they mask genuine compiler warnings and are liable to crop up
> elsewhere.  That particular function isn't likely to change but I'm
> worried in case this sort of thing starts cropping up elsewhere.
> 
> For that particular function I'd rather add return statements after the
> BUG()s...

Yes, this would work, too.

> As far as I'm aware the current ASoC code is warning free.

OK, I'll test through all builds again.


Takashi

Re: Merged upstream to topic/asoc branch

[Takashi Iwai]

At Mon, 01 Dec 2008 18:32:27 +0100,
I wrote:
> 
> At Mon, 1 Dec 2008 17:24:55 +0000,
> Mark Brown wrote:
> > 
> > On Mon, Dec 01, 2008 at 06:14:12PM +0100, Takashi Iwai wrote:
> > > Mark Brown wrote:
> > 
> > > > Hrm.  Which compiler are you using?
> > 
> > > gcc-4.1 for ARM, if I remember correctly...
> > 
> > I'm at 4.2.4 for ARM.
> > 
> > > >  The WM8903 warning patch you
> > > > introduced is not needed on any of my systems (primarily ARM and
> > > > x86-64).  It smells like a problem with the BUG() definition on your
> > > > platform...
> > 
> > > Possibly.  I'm not pretty sure on which systems, since I run test-builds
> > > regularly for ARM, PPC, x86 and x86-64 with cross compilers.
> > 
> > Could you investigate, please?  I'd rather not add annotations like that
> > since they mask genuine compiler warnings and are liable to crop up
> > elsewhere.  That particular function isn't likely to change but I'm
> > worried in case this sort of thing starts cropping up elsewhere.
> > 
> > For that particular function I'd rather add return statements after the
> > BUG()s...
> 
> Yes, this would work, too.
> 
> > As far as I'm aware the current ASoC code is warning free.
> 
> OK, I'll test through all builds again.

Oh, BTW, the following fix is also still pending on my local queue.


===

From: Takashi Iwai <tiwai@suse.de>
Subject: ALSA: ASoC: Fix compile warnings on corgi.c

  sound/soc/pxa/corgi.c: In function 'corgi_shutdown':
  sound/soc/pxa/corgi.c:114: warning: unused variable 'codec'
  sound/soc/pxa/corgi.c: At top level:
  sound/soc/pxa/corgi.c:175: warning: initialization from incompatible pointer type

Signed-off-by: Takashi Iwai <tiwai@suse.de>

---
diff --git a/sound/soc/pxa/corgi.c b/sound/soc/pxa/corgi.c
index 647f056..e56bf4b 100644
--- a/sound/soc/pxa/corgi.c
+++ b/sound/soc/pxa/corgi.c
@@ -108,15 +108,11 @@ static int corgi_startup(struct snd_pcm_substream *substream)
 }
 
 /* we need to unmute the HP at shutdown as the mute burns power on corgi */
-static int corgi_shutdown(struct snd_pcm_substream *substream)
+static void corgi_shutdown(struct snd_pcm_substream *substream)
 {
-	struct snd_soc_pcm_runtime *rtd = substream->private_data;
-	struct snd_soc_codec *codec = rtd->socdev->codec;
-
 	/* set = unmute headphone */
 	gpio_set_value(CORGI_GPIO_MUTE_L, 1);
 	gpio_set_value(CORGI_GPIO_MUTE_R, 1);
-	return 0;
 }
 
 static int corgi_hw_params(struct snd_pcm_substream *substream,

Re: Merged upstream to topic/asoc branch

[Mark Brown]

On Mon, Dec 01, 2008 at 06:37:17PM +0100, Takashi Iwai wrote:

> Oh, BTW, the following fix is also still pending on my local queue.

That fix looks good, I don't regularly track the Zaurii since keeping
track of the build failures in mainline was more trouble than it was
worth last time I tried.

Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

Re: Merged upstream to topic/asoc branch

[Takashi Iwai]

At Mon, 1 Dec 2008 17:42:52 +0000,
Mark Brown wrote:
> 
> On Mon, Dec 01, 2008 at 06:37:17PM +0100, Takashi Iwai wrote:
> 
> > Oh, BTW, the following fix is also still pending on my local queue.
> 
> That fix looks good, I don't regularly track the Zaurii since keeping
> track of the build failures in mainline was more trouble than it was
> worth last time I tried.
> 
> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

OK.  I also applied one more fix below.


Takashi

===
>From 682d5874f3d654b5d13d9b8dd56b9e05cfadd01b Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Mon, 1 Dec 2008 20:03:54 +0100
Subject: [PATCH] ALSA: ASoC: Fix old-style trigger callback in s3c2443-ac97.c

Fix the old-style trigger callback in s3c2443-ac97.c:
    sound/soc/s3c24xx/s3c2443-ac97.c:378: warning: initialization from incompatible pointer type

Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/soc/s3c24xx/s3c2443-ac97.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/sound/soc/s3c24xx/s3c2443-ac97.c b/sound/soc/s3c24xx/s3c2443-ac97.c
index 41bde6a..f0bc9b7 100644
--- a/sound/soc/s3c24xx/s3c2443-ac97.c
+++ b/sound/soc/s3c24xx/s3c2443-ac97.c
@@ -285,7 +285,8 @@ static int s3c2443_ac97_hw_params(struct snd_pcm_substream *substream,
 	return 0;
 }
 
-static int s3c2443_ac97_trigger(struct snd_pcm_substream *substream, int cmd)
+static int s3c2443_ac97_trigger(struct snd_pcm_substream *substream, int cmd,
+				struct snd_soc_dai *dai)
 {
 	u32 ac_glbctrl;
 
-- 
1.6.0.4

Re: Merged upstream to topic/asoc branch

[Takashi Iwai]

At Mon, 01 Dec 2008 18:32:27 +0100,
I wrote:
> 
> At Mon, 1 Dec 2008 17:24:55 +0000,
> Mark Brown wrote:
> > 
> > On Mon, Dec 01, 2008 at 06:14:12PM +0100, Takashi Iwai wrote:
> > > Mark Brown wrote:
> > 
> > > > Hrm.  Which compiler are you using?
> > 
> > > gcc-4.1 for ARM, if I remember correctly...
> > 
> > I'm at 4.2.4 for ARM.
> > 
> > > >  The WM8903 warning patch you
> > > > introduced is not needed on any of my systems (primarily ARM and
> > > > x86-64).  It smells like a problem with the BUG() definition on your
> > > > platform...
> > 
> > > Possibly.  I'm not pretty sure on which systems, since I run test-builds
> > > regularly for ARM, PPC, x86 and x86-64 with cross compilers.
> > 
> > Could you investigate, please?  I'd rather not add annotations like that
> > since they mask genuine compiler warnings and are liable to crop up
> > elsewhere.  That particular function isn't likely to change but I'm
> > worried in case this sort of thing starts cropping up elsewhere.
> > 
> > For that particular function I'd rather add return statements after the
> > BUG()s...
> 
> Yes, this would work, too.
> 
> > As far as I'm aware the current ASoC code is warning free.
> 
> OK, I'll test through all builds again.

Through all builds, I couldn't reproduce the warnings again.
So I reverted the commit.  Sorry for noises.


Takashi

Re: Warnings in wm8903.c (Re: Merged upstream to topic/asoc branch)

[Mark Brown]

On Wed, Dec 10, 2008 at 08:26:56AM +0100, Takashi Iwai wrote:

> Hmm, I got the warnings again on a x86-64 with gcc-4.3.2 randconfig
> test build again.

> The config is attached below.

OK, I've managed to reproduce this with that .config.  I'll push a patch
later today which adds return statements to silence the warning, meaning
that any genuine issues won't be suppressed.