Enable GPU blacklist in tests.

Enable GPU blacklist in tests.

Partial re-land of https://codereview.chromium.org/22198004/) -- Always enable FCM on Windows -- Part 1/3.

On top of fixing some of the issues there; only lands the part that enables the GPU blacklist in tests as the former CL is too hard to land all at once.

Also keeping --skip-gpu-data-loading around for now to be able to deal with failures caused in layout_tests by this on their own later (this needs to land ASAP and layout_tests don't need this ASAP).

Another CL will follow to always enable FCM on non-blacklisted Windows machines.

This part re-enables loading the blacklist in tests (and adds a content_browsertest to make sure that the configuration we expect to be testing is indeed the one we are testing -- this uncertainty is basically the only reason the blacklist was explicitly disabled before).

This CL also cleans up compositor_util.cc which was enforcing the blacklist twice.

The original plan was to do this only for Windows as Mac/Linux was causing trouble, but it turns out to be harder to do it only on Windows; so taking care of http://crbug.com/277242 in this CL too after all... 

BUG=233830, 190942, 277242
TBR=jcivelli, piman 

Originally Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=219132 
Then Reverted: https://src.chromium.org/viewvc/chrome?view=rev&revision=219159

Then re-committed (part 1/3): https://src.chromium.org/viewvc/chrome?view=rev&revision=221114
Reverted in: https://src.chromium.org/viewvc/chrome?view=rev&revision=221145

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=222105
Reverted: https://src.chromium.org/viewvc/chrome?view=rev&revision=222129

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=223397
Reverted: https://src.chromium.org/viewvc/chrome?view=rev&revision=223456

[gab, 2013-08-27 22:24:12]

Trying to re-land https://codereview.chromium.org/22198004/ (TBR jcivelli/piman
as before); zmo, WDYT of this?

@jamesr: Running layout tests, WDYT needs to change here (ref:
https://code.google.com/p/chromium/issues/detail?id=267038#c29).

Thanks!
Gab

https://codereview.chromium.org/23534006/diff/27001/chrome/test/gpu/gpu_featu...
File chrome/test/gpu/gpu_feature_browsertest.cc (right):

https://codereview.chromium.org/23534006/diff/27001/chrome/test/gpu/gpu_featu...
chrome/test/gpu/gpu_feature_browsertest.cc:196: const std::string json_blacklist
=
Moved the test blacklisting directly inside the test (like the other similar
tests are already doing); this should help prevent races with the GPU process
coming back with an update and messing up the test config -- although the real
solution is probably to disable GPU probing for these tests?

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
File gpu/config/gpu_control_list.cc (right):

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
gpu/config/gpu_control_list.cc:1043: LOG(INFO) <<
base::StringPrintf(kDisabledFeaturesMessage, id_);
Used this to debug locally and via try jobs along the way and I think it could
be useful to have such INFO logs always on in browser tests, WDYT?

[jamesr, 2013-08-27 22:31:16]

On 2013/08/27 22:24:12, gab wrote:
> Trying to re-land https://codereview.chromium.org/22198004/ (TBR
jcivelli/piman
> as before); zmo, WDYT of this?
> 
> @jamesr: Running layout tests, WDYT needs to change here (ref:
> https://code.google.com/p/chromium/issues/detail?id=267038#c29).

Somebody has to code up
https://code.google.com/p/chromium/issues/detail?id=267038#c30 before you can
land this.

[Zhenyao Mo, 2013-08-27 22:50:41]

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
File content/public/common/content_switches.cc (right):

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
content/public/common/content_switches.cc:615: const char
kLogGpuBlacklistDecisions[]        = "log-gpu-blacklist-decisions";
kLogGpuControlListDecisions and "log-gpu-control-list-decisions"

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
File content/public/common/content_switches.h (right):

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
content/public/common/content_switches.h:184: CONTENT_EXPORT extern const char
kLogGpuBlacklistDecisions[];
kLogGpuControlListDecisions[]

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
File gpu/config/gpu_control_list.cc (right):

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
gpu/config/gpu_control_list.cc:1043: LOG(INFO) <<
base::StringPrintf(kDisabledFeaturesMessage, id_);
On 2013/08/27 22:24:12, gab wrote:
> Used this to debug locally and via try jobs along the way and I think it could
> be useful to have such INFO logs always on in browser tests, WDYT?

Use LogControlListMatch() instead and also the message, switch name, as it is a
more generic class than GpuBlacklist.

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
gpu/config/gpu_control_list.cc:1320: if (blacklist_logging_enabled_)
same, control_list_logging_enabled_ instead.

[gab, 2013-08-28 20:20:38]

@zmo, PTAL

@jamesr, how involved do you think this change is? Any idea who could do it? I
guess I could take a look, but I don't have any experience with the
content_shell...

Thanks,
Gab

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
File content/public/common/content_switches.cc (right):

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
content/public/common/content_switches.cc:615: const char
kLogGpuBlacklistDecisions[]        = "log-gpu-blacklist-decisions";
On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> kLogGpuControlListDecisions and "log-gpu-control-list-decisions"

Done.

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
File content/public/common/content_switches.h (right):

https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
content/public/common/content_switches.h:184: CONTENT_EXPORT extern const char
kLogGpuBlacklistDecisions[];
On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> kLogGpuControlListDecisions[]

Done.

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
File gpu/config/gpu_control_list.cc (right):

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
gpu/config/gpu_control_list.cc:1043: LOG(INFO) <<
base::StringPrintf(kDisabledFeaturesMessage, id_);
On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> On 2013/08/27 22:24:12, gab wrote:
> > Used this to debug locally and via try jobs along the way and I think it
could
> > be useful to have such INFO logs always on in browser tests, WDYT?
> 
> Use LogControlListMatch() instead and also the message, switch name, as it is
a
> more generic class than GpuBlacklist.

Done.

https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
gpu/config/gpu_control_list.cc:1320: if (blacklist_logging_enabled_)
On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> same, control_list_logging_enabled_ instead.

Done.

[Zhenyao Mo, 2013-08-28 21:12:48]

On 2013/08/28 20:20:38, gab wrote:
> @zmo, PTAL
> 
> @jamesr, how involved do you think this change is? Any idea who could do it? I
> guess I could take a look, but I don't have any experience with the
> content_shell...
> 
> Thanks,
> Gab
> 
>
https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
> File content/public/common/content_switches.cc (right):
> 
>
https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
> content/public/common/content_switches.cc:615: const char
> kLogGpuBlacklistDecisions[]        = "log-gpu-blacklist-decisions";
> On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> > kLogGpuControlListDecisions and "log-gpu-control-list-decisions"
> 
> Done.
> 
>
https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
> File content/public/common/content_switches.h (right):
> 
>
https://codereview.chromium.org/23534006/diff/27001/content/public/common/con...
> content/public/common/content_switches.h:184: CONTENT_EXPORT extern const char
> kLogGpuBlacklistDecisions[];
> On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> > kLogGpuControlListDecisions[]
> 
> Done.
> 
>
https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
> File gpu/config/gpu_control_list.cc (right):
> 
>
https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
> gpu/config/gpu_control_list.cc:1043: LOG(INFO) <<
> base::StringPrintf(kDisabledFeaturesMessage, id_);
> On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> > On 2013/08/27 22:24:12, gab wrote:
> > > Used this to debug locally and via try jobs along the way and I think it
> could
> > > be useful to have such INFO logs always on in browser tests, WDYT?
> > 
> > Use LogControlListMatch() instead and also the message, switch name, as it
is
> a
> > more generic class than GpuBlacklist.
> 
> Done.
> 
>
https://codereview.chromium.org/23534006/diff/27001/gpu/config/gpu_control_li...
> gpu/config/gpu_control_list.cc:1320: if (blacklist_logging_enabled_)
> On 2013/08/27 22:50:41, Zhenyao Mo wrote:
> > same, control_list_logging_enabled_ instead.
> 
> Done.

LGTM

[gab, 2013-09-03 22:04:46]

Hello all,

I neutered this CL to only take care of enabling the blacklist on the bots (and
the refactoring around that).

I will keep FCM on Windows for another CL (it's currently blocked on at least
crbug.com/281726 anyways).

I will also keep --skip-gpu-data-loading around for layout_tests only for now
and deal with those failures after FCM goes in.

This was getting too hard to land and understand failures with all of this in a
single CL.

Updated title/description to reflect this.

Cheers!
Gab

[commit-bot: I haz the power, 2013-09-03 22:05:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gab@chromium.org/23534006/143001

[commit-bot: I haz the power, 2013-09-04 02:01:58]

Change committed as 221114

[gab, 2013-09-04 18:54:26]

Fixing GpuFeatureTest.AcceleratedCompositingBlocked; see
https://codereview.chromium.org/23775004/ for details on revert.

https://codereview.chromium.org/23534006/diff/122001/content/browser/gpu/gpu_...
File content/browser/gpu/gpu_data_manager_impl_private.h (right):

https://codereview.chromium.org/23534006/diff/122001/content/browser/gpu/gpu_...
content/browser/gpu/gpu_data_manager_impl_private.h:250: bool finalized_;
Adding this state to make sure the initialization resulting of GPU process
reporting back doesn't end up overwriting test blacklists initialized via
InitializeForTesting...

[gab, 2013-09-09 15:45:44]

FCM landed in http://crrev.com/221643; trying this again...

[commit-bot: I haz the power, 2013-09-09 15:47:50]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gab@chromium.org/23534006/143002

[commit-bot: I haz the power, 2013-09-09 21:52:20]

Change committed as 222105

[please use gerrit instead, 2013-09-09 23:23:05]

On 2013/09/09 21:52:20, I haz the power (commit-bot) wrote:
> Change committed as 222105

FYI, AcceleratedCompositingBlockedTest.AcceleratedCompositingBlocked failed on
mac-asan builder:

AcceleratedCompositingBlockedTest.AcceleratedCompositingBlocked: 
[7719:3847:0909/155847:1828989999085:INFO:gpu_control_list.cc(1043)] Control
list match for rule #48.
[7719:3847:0909/155847:1829419815783:INFO:gpu_control_list.cc(1043)] Control
list match for rule #48.
[7719:3847:0909/155847:1829591550844:INFO:gpu_control_list.cc(1043)] Control
list match for rule #.
../../chrome/test/gpu/gpu_feature_browsertest.cc:143: Failure
Value of: 0U
Actual: 0
Expected: event_count
Which is: 1

If this failure persists, then I will be reverting r222105.

[please use gerrit instead, 2013-09-09 23:36:24]

On 2013/09/09 23:23:05, Rouslan Solomakhin wrote:
> On 2013/09/09 21:52:20, I haz the power (commit-bot) wrote:
> > Change committed as 222105
> 
> FYI, AcceleratedCompositingBlockedTest.AcceleratedCompositingBlocked failed on
> mac-asan builder:
> 
> AcceleratedCompositingBlockedTest.AcceleratedCompositingBlocked: 
> [7719:3847:0909/155847:1828989999085:INFO:gpu_control_list.cc(1043)] Control
> list match for rule #48.
> [7719:3847:0909/155847:1829419815783:INFO:gpu_control_list.cc(1043)] Control
> list match for rule #48.
> [7719:3847:0909/155847:1829591550844:INFO:gpu_control_list.cc(1043)] Control
> list match for rule #.
> ../../chrome/test/gpu/gpu_feature_browsertest.cc:143: Failure
> Value of: 0U
> Actual: 0
> Expected: event_count
> Which is: 1
> 
> If this failure persists, then I will be reverting r222105.

Failed twice in a row. Reverting r222105.

[gab, 2013-09-16 17:55:30]

On 2013/09/09 23:36:24, Rouslan Solomakhin wrote:
> On 2013/09/09 23:23:05, Rouslan Solomakhin wrote:
> > On 2013/09/09 21:52:20, I haz the power (commit-bot) wrote:
> > > Change committed as 222105
> > 
> > FYI, AcceleratedCompositingBlockedTest.AcceleratedCompositingBlocked failed
on
> > mac-asan builder:
> > 
> > AcceleratedCompositingBlockedTest.AcceleratedCompositingBlocked: 
> > [7719:3847:0909/155847:1828989999085:INFO:gpu_control_list.cc(1043)] Control
> > list match for rule #48.
> > [7719:3847:0909/155847:1829419815783:INFO:gpu_control_list.cc(1043)] Control
> > list match for rule #48.
> > [7719:3847:0909/155847:1829591550844:INFO:gpu_control_list.cc(1043)] Control
> > list match for rule #.
> > ../../chrome/test/gpu/gpu_feature_browsertest.cc:143: Failure
> > Value of: 0U
> > Actual: 0
> > Expected: event_count
> > Which is: 1
> > 
> > If this failure persists, then I will be reverting r222105.
> 
> Failed twice in a row. Reverting r222105.

Aura landed, me/cpu think that maybe this bug is fixed, trying again...

[commit-bot: I haz the power, 2013-09-16 17:56:44]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gab@chromium.org/23534006/2001

[commit-bot: I haz the power, 2013-09-16 20:23:50]

Change committed as 223397

[gab, 2013-09-17 15:49:37]

On 2013/09/16 20:23:50, I haz the power (commit-bot) wrote:
> Change committed as 223397

Reverted in 223456:

VISTA browser_tests:

WebViewTest.Shim_TestAutosizeAfterNavigation: 
[4360:2360:0916/144749:2708687:INFO:gpu_control_list.cc(1043)] Control list
match for rule #64.
[4360:2360:0916/144749:2708687:INFO:gpu_control_list.cc(1043)] Control list
match for rule #65.
[4360:2360:0916/144749:2708718:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4360:2360:0916/144749:2708734:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4296:4712:0916/144749:ERROR:base_feature_provider.cc(122)] manifestTypes:
Allowing web_page contexts requires supplying a value for matches.
[4360:2360:0916/144749:2709000:ERROR:base_feature_provider.cc(122)]
manifestTypes: Allowing web_page contexts requires supplying a value for
matches.
[4360:2360:0916/144749:2709015:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4360:2360:0916/144749:2709015:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4360:2360:0916/144754:2713750:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
Killed (timed out).

WebViewTest.Shim_TestAutosizeWithPartialAttributes: 
[4080:3604:0916/144719:2678578:INFO:gpu_control_list.cc(1043)] Control list
match for rule #64.
[4080:3604:0916/144719:2678578:INFO:gpu_control_list.cc(1043)] Control list
match for rule #65.
[4080:3604:0916/144719:2678609:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4080:3604:0916/144719:2678609:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[956:3276:0916/144719:ERROR:base_feature_provider.cc(122)] manifestTypes:
Allowing web_page contexts requires supplying a value for matches.
[4080:3604:0916/144719:2678843:ERROR:base_feature_provider.cc(122)]
manifestTypes: Allowing web_page contexts requires supplying a value for
matches.
[4080:3604:0916/144719:2678859:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4080:3604:0916/144719:2678875:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4080:3604:0916/144719:2678921:INFO:CONSOLE(168)]
"testAutosizeWithPartialAttributes", source:
chrome-extension://pkempkigbpaghgdcbdpelombjbioainh/main.js (168)
[4080:3604:0916/144724:2683625:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
Killed (timed out).

WebViewTest.Shim_TestResizeWebviewResizesContent: 
[4452:3772:0916/144648:2648468:INFO:gpu_control_list.cc(1043)] Control list
match for rule #64.
[4452:3772:0916/144648:2648468:INFO:gpu_control_list.cc(1043)] Control list
match for rule #65.
[4452:3772:0916/144648:2648500:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4452:3772:0916/144648:2648500:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[5048:4448:0916/144649:ERROR:base_feature_provider.cc(122)] manifestTypes:
Allowing web_page contexts requires supplying a value for matches.
[4452:3772:0916/144649:2648703:ERROR:base_feature_provider.cc(122)]
manifestTypes: Allowing web_page contexts requires supplying a value for
matches.
[4452:3772:0916/144649:2648718:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4452:3772:0916/144649:2648718:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4188:4400:0916/144649:ERROR:base_feature_provider.cc(122)] manifestTypes:
Allowing web_page contexts requires supplying a value for matches.
[4452:3772:0916/144649:2648906:INFO:CONSOLE(982)] "Script has been injected into
webview.", source: chrome-extension://pkempkigbpaghgdcbdpelombjbioainh/main.js
(982)
[4452:3772:0916/144649:2648937:INFO:CONSOLE(992)] "A communication channel has
been established with webview.", source:
chrome-extension://pkempkigbpaghgdcbdpelombjbioainh/main.js (992)
[4452:3772:0916/144649:2648937:INFO:CONSOLE(993)] "Resizing <webview> width from
300px to 400px.", source:
chrome-extension://pkempkigbpaghgdcbdpelombjbioainh/main.js (993)
[4452:3772:0916/144653:2653515:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
Killed (timed out).




XP browser_tests:

WebViewTest.Shim_TestResizeWebviewResizesContent: 
[4032:1596:0916/153058:5037625:INFO:gpu_control_list.cc(1043)] Control list
match for rule #41.
[4032:1596:0916/153058:5037625:INFO:gpu_control_list.cc(1043)] Control list
match for rule #64.
[4032:1596:0916/153058:5037625:INFO:gpu_control_list.cc(1043)] Control list
match for rule #67.
[4032:1596:0916/153058:5037625:INFO:gpu_control_list.cc(1043)] Control list
match for rule #78.
[4032:1596:0916/153101:5040781:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4032:1596:0916/153101:5040781:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4032:1596:0916/153107:5045937:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4032:1596:0916/153109:5047953:ERROR:base_feature_provider.cc(122)]
manifestTypes: Allowing web_page contexts requires supplying a value for
matches.
[4032:1596:0916/153109:5048093:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[4032:1596:0916/153109:5048125:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[4032:1596:0916/153117:5056078:INFO:CONSOLE(982)] "Script has been injected into
webview.", source: chrome-extension://cihojnfhelemnmfajmkmgpbbiceaachl/main.js
(982)
[4032:1596:0916/153117:5056421:INFO:CONSOLE(992)] "A communication channel has
been established with webview.", source:
chrome-extension://cihojnfhelemnmfajmkmgpbbiceaachl/main.js (992)
[4032:1596:0916/153117:5056421:INFO:CONSOLE(993)] "Resizing <webview> width from
300px to 400px.", source:
chrome-extension://cihojnfhelemnmfajmkmgpbbiceaachl/main.js (993)
Killed (timed out).

WebViewTest.Shim_TestAutosizeWithPartialAttributes: 
[3220:4072:0916/162017:3964000:INFO:gpu_control_list.cc(1043)] Control list
match for rule #41.
[3220:4072:0916/162017:3964000:INFO:gpu_control_list.cc(1043)] Control list
match for rule #64.
[3220:4072:0916/162017:3964000:INFO:gpu_control_list.cc(1043)] Control list
match for rule #67.
[3220:4072:0916/162017:3964000:INFO:gpu_control_list.cc(1043)] Control list
match for rule #78.
[3220:4072:0916/162019:3966031:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[3220:4072:0916/162019:3966031:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[3220:4072:0916/162024:3971078:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[3220:4072:0916/162025:3971562:ERROR:base_feature_provider.cc(122)]
manifestTypes: Allowing web_page contexts requires supplying a value for
matches.
[3220:4072:0916/162025:3971609:ERROR:chrome_views_delegate.cc(158)] NOT
IMPLEMENTED
[3220:4072:0916/162025:3971625:ERROR:desktop_root_window_host_win.cc(671)] NOT
IMPLEMENTED
[3220:4072:0916/162026:3972140:INFO:CONSOLE(168)]
"testAutosizeWithPartialAttributes", source:
chrome-extension://ehnjcbganihefdmlcmaggflijkfeookm/main.js (168)
Killed (timed out).

[gab, 2013-09-23 20:58:50]

Work for this will be continued in https://codereview.chromium.org/23703017/;
this CL already gave enough :)!

Cheers,
Gab