Remove call to MessageLoop::RunUntilIdle from wait_util.mm.

Remove call to MessageLoop::RunUntilIdle from wait_util.mm.

MessageLoop::RunUntilIdle is deprecated.

BUG=616447
Committed: https://crrev.com/25556a66237890c3e0ae3be457093f72e1c4a387
Cr-Commit-Position: refs/heads/master@{#420416}

[fdoray, 2016-09-20 19:04:56]

fdoray@chromium.org changed reviewers:
+ rohitrao@chromium.org

[fdoray, 2016-09-20 19:04:56]

PTAL

[fdoray, 2016-09-22 17:44:09]

fdoray@chromium.org changed reviewers:
+ sdefresne@chromium.org

[fdoray, 2016-09-22 17:44:10]

rohitrao@: ping

+sdefresne@ in case rohitrao@ isn't available

[rohitrao (ping after 24h), 2016-09-22 17:52:13]

LGTM

Does anyone ever call this function with a message loop that isn't the current
message loop?

[fdoray, 2016-09-22 17:56:53]

On 2016/09/22 17:52:13, rohitrao wrote:
> LGTM
> 
> Does anyone ever call this function with a message loop that isn't the current
> message loop?

I haven't seen any call site that does that in public or internal code.

Anyways, calling RunUntilIdle() on any MessageLoop runs MessageLoop::current(),
not the MessageLoop on which RunUntilIdle() is called. It's counter-intuitive
and that one of the reasons why we are deprecating MessageLoop::RunUntilIdle().

[fdoray, 2016-09-22 18:48:49]

On 2016/09/22 17:56:53, fdoray wrote:
> On 2016/09/22 17:52:13, rohitrao wrote:
> > LGTM
> > 
> > Does anyone ever call this function with a message loop that isn't the
current
> > message loop?
> 
> I haven't seen any call site that does that in public or internal code.
> 
> Anyways, calling RunUntilIdle() on any MessageLoop runs
MessageLoop::current(),
> not the MessageLoop on which RunUntilIdle() is called. It's counter-intuitive
> and that one of the reasons why we are deprecating
MessageLoop::RunUntilIdle().

ios-internal tests passed.

[fdoray, 2016-09-22 18:48:53]

The CQ bit was checked by fdoray@chromium.org

[commit-bot: I haz the power, 2016-09-22 18:49:33]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-22 19:06:58]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-09-22 19:09:25]

Description was changed from

==========
Remove call to MessageLoop::RunUntilIdle from wait_util.mm.

MessageLoop::RunUntilIdle is deprecated.

BUG=616447
==========

to

==========
Remove call to MessageLoop::RunUntilIdle from wait_util.mm.

MessageLoop::RunUntilIdle is deprecated.

BUG=616447

Committed: https://crrev.com/25556a66237890c3e0ae3be457093f72e1c4a387
Cr-Commit-Position: refs/heads/master@{#420416}
==========

[commit-bot: I haz the power, 2016-09-22 19:09:27]

Patchset 3 (id:??) landed as
https://crrev.com/25556a66237890c3e0ae3be457093f72e1c4a387
Cr-Commit-Position: refs/heads/master@{#420416}

[sdefresne, 2016-09-23 09:08:17]

https://codereview.chromium.org/2358633002/diff/40001/base/test/ios/wait_util.mm
File base/test/ios/wait_util.mm (right):

https://codereview.chromium.org/2358633002/diff/40001/base/test/ios/wait_util...
base/test/ios/wait_util.mm:21: bool run_message_loop,
It look like you are passing "true" for this variable in all callsite while
previously the value was null if MessageLoop::current() was returning null
(which is possible).

If this does not break anything, it's good, but there is a semantic change here.

In addition, if we only ever call this method with "true", then maybe we should
remove the extra parameter.

WDYT?

[fdoray, 2016-09-23 12:13:47]

https://codereview.chromium.org/2358633002/diff/40001/base/test/ios/wait_util.mm
File base/test/ios/wait_util.mm (right):

https://codereview.chromium.org/2358633002/diff/40001/base/test/ios/wait_util...
base/test/ios/wait_util.mm:21: bool run_message_loop,
On 2016/09/23 09:08:17, sdefresne wrote:
> It look like you are passing "true" for this variable in all callsite while
> previously the value was null if MessageLoop::current() was returning null
> (which is possible).
> 
> If this does not break anything, it's good, but there is a semantic change
here.
> 
> In addition, if we only ever call this method with "true", then maybe we
should
> remove the extra parameter.
> 
> WDYT?

You're right, this changes the semantic. But since the CL didn't break anything,
I don't think that MessageLoop::current() returned nullptr at any
TimeUntilCondition() call site before.

In ios_internal, there are call sites where nullptr is passed (will be migrated
to false). Therefore, we can't remove the extra parameter.