Enable Windows Flash hw video decode trial.

Add Windows Flash hw video decode trial.

This is to support the Windows rollout of the feature.

The distribution on different channels will be controlled from the
server. Initially we will target 100% on canary/dev, 50% on beta and
starting slowly (say from 1%) on stable but this will be gradual and
is subject to approval.

Also remove low latency audio trail which has expired a long time ago.
(It also is hard wired to enabled in all Flash binaries.)

BUG=chromium:335836
TEST=Verified Flash flags are properly passed on Link.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=246446

[ilja, 2014-01-18 02:01:18]

I still need to setup the config for the backend in p4 but I am wondering if
this is all I need to change in the browser?

[cpu_(ooo_6.6-7.5), 2014-01-19 20:56:38]

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
File chrome/browser/chrome_browser_field_trials_desktop.cc (right):

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
chrome/browser/chrome_browser_field_trials_desktop.cc:71: #if defined(OS_WIN)
do we need #if defined (OS_WIN) ?

Would the non-trial enabled flash players on mac/linux just ignore the flag so
when they don't ignore we don't have to change this code?

[DaleCurtis, 2014-01-21 18:54:41]

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
File chrome/browser/chrome_browser_field_trials_desktop.cc (right):

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
chrome/browser/chrome_browser_field_trials_desktop.cc:74:
content::kFlashHwVideoDecodeFieldTrialName, 1000, "Standard",
Why 1000? 1 and 50 would work below, no?

[ilja, 2014-01-21 21:15:28]

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
File chrome/browser/chrome_browser_field_trials_desktop.cc (right):

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
chrome/browser/chrome_browser_field_trials_desktop.cc:71: #if defined(OS_WIN)
On 2014/01/19 20:56:39, cpu wrote:
> do we need #if defined (OS_WIN) ?
> 
> Would the non-trial enabled flash players on mac/linux just ignore the flag so
> when they don't ignore we don't have to change this code?

No, the mac/linux players don't ignore the Flag. If set they will try
and ask the browser for hardware decode, get an error, handle the error
and fall back for the duration of the video, just to try again with the
next. I think we should focus this effort on Windows. (And on CrOS we
would pass the argument twice, which seems wrong considering that
we have length limitation.)

https://codereview.chromium.org/141753006/diff/110001/chrome/browser/chrome_b...
chrome/browser/chrome_browser_field_trials_desktop.cc:74:
content::kFlashHwVideoDecodeFieldTrialName, 1000, "Standard",
On 2014/01/21 18:54:41, DaleCurtis wrote:
> Why 1000? 1 and 50 would work below, no?

If we stick with the current numbers then I will change it back to 1 and 50.

[DaleCurtis, 2014-01-21 21:17:13]

lgtm % switching the numbers then.

[ilja, 2014-01-22 02:16:28]

I spoke with Adobe and Anthony and adjusted a few numbers. I think it is better
to have it fixed enabled on any builds developers might use
(dev/canary/unbranded). Also I think we should get better user feedback by using
ONE_TIME_RANDOMIZED.

Piman: OWNERs review please.

[piman, 2014-01-22 02:21:46]

lgtm

[commit-bot: I haz the power, 2014-01-22 02:48:48]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ihf@chromium.org/141753006/230001

[commit-bot: I haz the power, 2014-01-22 03:10:22]

Retried try job too often on chromium_presubmit for step(s) presubmit
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=chromium_p...

[ilja, 2014-01-22 03:49:31]

Alexei: could you also please do an OWNERs review?

[ilja, 2014-01-22 06:36:23]

Sorry for the rebase. While testing with the finch server I realized that Flash
arguments should be passed comma delimited.

Dale, please take a second look.

Alexei: I have tried testing this (on ChromeOS if that matters - but with the
#defines adjusted) with --fake-variations-channel=beta and have not been able to
force the beta or stable channel. See also citique/59987046.

[Alexei Svitkine (slow), 2014-01-22 16:27:24]

I can help you figure out why you're having trouble activating this trial
manually through chat today. Ping me when my calendar's free.

https://codereview.chromium.org/141753006/diff/470001/chrome/browser/chrome_b...
File chrome/browser/chrome_browser_field_trials_desktop.cc (right):

https://codereview.chromium.org/141753006/diff/470001/chrome/browser/chrome_b...
chrome/browser/chrome_browser_field_trials_desktop.cc:76: 2014, 6, 10,
base::FieldTrial::ONE_TIME_RANDOMIZED, NULL));
It's better to initialize field trials entirely from the server, unless there's
a really good reason against that (which I don't think there is in this case).

If the trial is set up from the server, then this block of code is not needed at
all and can be removed.

https://codereview.chromium.org/141753006/diff/470001/content/browser/ppapi_p...
File content/browser/ppapi_plugin_process_host.cc (right):

https://codereview.chromium.org/141753006/diff/470001/content/browser/ppapi_p...
content/browser/ppapi_plugin_process_host.cc:315: if (existing_args.size())
Nit: !existing_args.empty()

[cpu_(ooo_6.6-7.5), 2014-01-22 20:27:20]

lgtm

[Alexei Svitkine (slow), 2014-01-22 20:42:19]

lgtm % nit

https://codereview.chromium.org/141753006/diff/470001/content/browser/ppapi_p...
File content/browser/ppapi_plugin_process_host.cc (right):

https://codereview.chromium.org/141753006/diff/470001/content/browser/ppapi_p...
content/browser/ppapi_plugin_process_host.cc:315: if (existing_args.size())
On 2014/01/22 16:27:24, Alexei Svitkine wrote:
> Nit: !existing_args.empty()

[ilja, 2014-01-22 20:50:03]

We decided to remove the hard wired numbers and control everything from he
server.

[commit-bot: I haz the power, 2014-01-22 20:52:06]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ihf@chromium.org/141753006/630001

[commit-bot: I haz the power, 2014-01-22 22:31:49]

Retried try job too often on linux_chromeos for step(s) interactive_ui_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_chro...

[commit-bot: I haz the power, 2014-01-22 22:48:07]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ihf@chromium.org/141753006/630001

[commit-bot: I haz the power, 2014-01-22 23:35:40]

Change committed as 246446