Return fixed hardware buffer size for odd sample rates.

Return fixed hardware buffer size for odd sample rates.

Windows will return a sample rate of zero if audio can't
be output, in this case we shouldn't return a buffer size
of zero.

Some hardware also might return a hardware sample rate < 100,
in this case we should choose a non-zero fixed buffer size to
prevent crashes.

In either case, AudioOutputResampler will later check the hardware
config for validity and fallback to the high latency audio path in
cases where a bogus sample rate exists.

BUG=152073
TEST=compiles

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=158651

[Ami GONE FROM CHROMIUM, 2012-09-25 18:25:46]

LGTM % <IDontKnowWhatImDoing/>

https://codereview.chromium.org/10993013/diff/3001/media/audio/audio_util.cc
File media/audio/audio_util.cc (right):

https://codereview.chromium.org/10993013/diff/3001/media/audio/audio_util.cc#...
media/audio/audio_util.cc:351: return kFallbackBufferSize;
any reason you're going to 2048 here from the 960 it was before the previous
change?

https://codereview.chromium.org/10993013/diff/3001/media/audio/audio_util.cc#...
media/audio/audio_util.cc:353: return 2048;
constify

[henrika (OOO until Aug 14), 2012-09-25 18:28:26]

LGTM. Perhaps add a comment that we can see 0 as sample rate using e.g. Remote
Desktop.

[DaleCurtis, 2012-09-25 18:39:30]

https://codereview.chromium.org/10993013/diff/3001/media/audio/audio_util.cc
File media/audio/audio_util.cc (right):

https://codereview.chromium.org/10993013/diff/3001/media/audio/audio_util.cc#...
media/audio/audio_util.cc:351: return kFallbackBufferSize;
On 2012/09/25 18:25:46, Ami Fischman wrote:
> any reason you're going to 2048 here from the 960 it was before the previous
> change?

Most cases are now handled by the mixing_sample_rate / 100, in cases where we've
got a crazy sample rate < 8kHz, I think a larger buffer size is more prudent.

Although it doesn't really matter since the fallback code is going to say hell
no when it tries to verify the configuration and choose a high latency
appropriate buffer size via another algorithm.

https://codereview.chromium.org/10993013/diff/3001/media/audio/audio_util.cc#...
media/audio/audio_util.cc:353: return 2048;
On 2012/09/25 18:25:46, Ami Fischman wrote:
> constify

I intentionally skipped this because it's not in the #OS_WIN section but rather
the default buffer size for Linux, i.e. not a fallback buffer size :)

[commit-bot: I haz the power, 2012-09-25 19:40:59]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dalecurtis@chromium.org/10993013/9001

[commit-bot: I haz the power, 2012-09-25 19:48:05]

Sorry for I got bad news for ya.
Compile failed with a clobber build.
Your code is likely broken or HEAD is junk. Please ensure your
code is not broken then alert the build sheriffs.
Look at the try server FAQ for more details.

[DaleCurtis, 2012-09-25 21:00:45]

On 2012/09/25 19:48:05, I haz the power (commit-bot) wrote:
> Sorry for I got bad news for ya.
> Compile failed with a clobber build.
> Your code is likely broken or HEAD is junk. Please ensure your
> code is not broken then alert the build sheriffs.
> Look at the try server FAQ for more details.

Landed manually since it passed Windows.