Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also caught http://crbug.com/634552, disabled TaskSchedulerTaskTrackerTest.SingletonAllowed's
death test on OS_LINUX for now because of it.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459, 634552
Committed: https://crrev.com/5e69cff93950ddd3959f995dcffff2069ff097c0
Cr-Commit-Position: refs/heads/master@{#409980}

[gab, 2016-08-04 15:56:06]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

BUG=553459
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generated an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in this CL.

BUG=553459
==========

[gab, 2016-08-04 15:56:07]

gab@chromium.org changed reviewers:
+ robliao@chromium.org

[gab, 2016-08-04 15:56:35]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generated an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in this CL.

BUG=553459
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in this CL.

BUG=553459
==========

[gab, 2016-08-04 15:57:33]

Rob PTAL at base/task_scheduler/scheduler_lock_impl.cc (don't bother with the
rest of the CL, I'll send to base/ OWNERS once you're okay with
SchedulerLockImpl changes).

Thanks!
Gab

[gab, 2016-08-04 15:57:51]

The CQ bit was checked by gab@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-04 15:58:14]

Dry run: 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, 2016-08-04 16:02:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-04 16:02:29]

Dry run: Try jobs failed on following builders:
  ios-device on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device/builds...)
  ios-simulator on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator/bui...)

[robliao, 2016-08-04 17:25:20]

https://codereview.chromium.org/2213933003/diff/20001/base/task_scheduler/sch...
File base/task_scheduler/scheduler_lock_impl.cc (right):

https://codereview.chromium.org/2213933003/diff/20001/base/task_scheduler/sch...
base/task_scheduler/scheduler_lock_impl.cc:76: // This asserts that lock's
registered predecessor is safe. Since registration
This change should be pulled out of this CL. Nice find! Logic looks good to me.

[gab, 2016-08-04 17:41:25]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in this CL.

BUG=553459
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

BUG=553459
==========

[gab, 2016-08-04 17:43:05]

gab@chromium.org changed reviewers:
+ thestig@chromium.org
- robliao@chromium.org

[gab, 2016-08-04 17:43:05]

Lei PTAL, I added these new death tests for DCHECKs recently and am realizing
that the second parameter (|regex|) was not very useful (not used by most --
resulting in one incorrect check as documented in CL description).

https://codereview.chromium.org/2213933003/diff/20001/base/task_scheduler/sch...
File base/task_scheduler/scheduler_lock_impl.cc (right):

https://codereview.chromium.org/2213933003/diff/20001/base/task_scheduler/sch...
base/task_scheduler/scheduler_lock_impl.cc:76: // This asserts that lock's
registered predecessor is safe. Since registration
On 2016/08/04 17:25:20, robliao wrote:
> This change should be pulled out of this CL. Nice find! Logic looks good to
me.

Done: https://codereview.chromium.org/2218443002/

[gab, 2016-08-04 17:44:14]

On 2016/08/04 17:43:05, gab wrote:
> Lei PTAL, I added these new death tests for DCHECKs recently and am realizing
> that the second parameter (|regex|) was not very useful (not used by most --
> resulting in one incorrect check as documented in CL description).

PS: Don't bother the red bots, patch failure because I of stacked uncommitted
branches, will fix when landing.

You mostly only need to look at gtest_util.h changes, everything else is a
side-effect of that change.

[gab, 2016-08-04 17:47:13]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

BUG=553459
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459
==========

[gab, 2016-08-04 17:48:18]

gab@chromium.org changed reviewers:
+ jam@chromium.org

[gab, 2016-08-04 17:48:19]

+jam for thread_restrictions.cc per base\threading\OWNERS mentioning specific
ownership of those files by you (see CL description for details)

[Lei Zhang, 2016-08-04 19:17:33]

lgtm

[gab, 2016-08-04 21:10:56]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catches an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459
==========

[gab, 2016-08-04 21:17:24]

On 2016/08/04 17:48:19, gab wrote:
> +jam for thread_restrictions.cc per base\threading\OWNERS mentioning specific
> ownership of those files by you (see CL description for details)

jam says he only cares about things that add to
https://cs.chromium.org/chromium/src/base/threading/thread_restrictions.h?rcl...
and as such this lgth, landing.

[gab, 2016-08-04 21:17:29]

The CQ bit was checked by gab@chromium.org

[commit-bot: I haz the power, 2016-08-04 21:18:03]

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

[Lei Zhang, 2016-08-04 21:18:03]

On 2016/08/04 21:17:24, gab wrote:
> jam says he only cares about things that add to
>
https://cs.chromium.org/chromium/src/base/threading/thread_restrictions.h?rcl...
> and as such this lgth, landing.

ack

[commit-bot: I haz the power, 2016-08-04 21:20:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-04 21:20:59]

Try jobs failed on following builders:
  mac_chromium_compile_dbg_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_comp...)

[gab, 2016-08-04 21:31:52]

The CQ bit was checked by gab@chromium.org

[gab, 2016-08-04 21:31:52]

The patchset sent to the CQ was uploaded after l-g-t-m from thestig@chromium.org
Link to the patchset: https://codereview.chromium.org/2213933003/#ps60001
(title: "merge up to r409903")

[commit-bot: I haz the power, 2016-08-04 21:32:27]

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, 2016-08-04 21:54:52]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-04 21:54:54]

Try jobs failed on following builders:
  linux_chromium_chromeos_ozone_rel_ng on master.tryserver.chromium.linux
(JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[gab, 2016-08-05 00:11:30]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also caught http://crbug.com/634552, disabled
TaskSchedulerTaskTrackerTest.SingletonAllowed's
death test on OS_LINUX for now because of it.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459, 634552
==========

[gab, 2016-08-05 00:12:46]

Patchset #5 (id:80001) has been deleted

[gab, 2016-08-05 00:13:23]

The CQ bit was checked by gab@chromium.org

[gab, 2016-08-05 00:13:24]

The patchset sent to the CQ was uploaded after l-g-t-m from thestig@chromium.org
Link to the patchset: https://codereview.chromium.org/2213933003/#ps100001
(title: "Disabling a death test which crashed instead of DCHECKing :
http://crbug.com/634552")

[commit-bot: I haz the power, 2016-08-05 00:13:57]

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

[gab, 2016-08-05 00:14:41]

FYI, had to disable TaskSchedulerTaskTrackerTest.SingletonAllowed's death test
on OS_LINUX as well because this CL caught http://crbug.com/634552 (so weird!).

This test is under base/task_scheduler which I own, so I'll CQ without
re-review, thanks!

[commit-bot: I haz the power, 2016-08-05 01:28:34]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-05 01:28:36]

Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[gab, 2016-08-05 01:29:01]

The CQ bit was checked by gab@chromium.org

[commit-bot: I haz the power, 2016-08-05 01:29:16]

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, 2016-08-05 02:29:48]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-05 02:29:50]

Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[gab, 2016-08-05 02:55:51]

The CQ bit was checked by gab@chromium.org

[commit-bot: I haz the power, 2016-08-05 02:56:08]

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, 2016-08-05 03:26:36]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also caught http://crbug.com/634552, disabled
TaskSchedulerTaskTrackerTest.SingletonAllowed's
death test on OS_LINUX for now because of it.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459, 634552
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also caught http://crbug.com/634552, disabled
TaskSchedulerTaskTrackerTest.SingletonAllowed's
death test on OS_LINUX for now because of it.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459, 634552
==========

[commit-bot: I haz the power, 2016-08-05 03:26:37]

Committed patchset #5 (id:100001)

[commit-bot: I haz the power, 2016-08-05 03:30:20]

Description was changed from

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on 
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also caught http://crbug.com/634552, disabled
TaskSchedulerTaskTrackerTest.SingletonAllowed's
death test on OS_LINUX for now because of it.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459, 634552
==========

to

==========
Change EXPECT/ASSERT_DCHECK_DEATH macro to not expose the |regex| parameter.

Most users of the macro were ignoring the regex, now it makes sure a DCHECK
caused the death (message contains "Check failed") without adding burden on
callers to encode the exact error message.

This catched an error in SchedulerLock which was generating an exception in
std::unordered_map::at() instead of hitting the DCHECK it was intending to
hit. Fixed in precursor CL @ https://codereview.chromium.org/2218443002/.

Also caught http://crbug.com/634552, disabled
TaskSchedulerTaskTrackerTest.SingletonAllowed's
death test on OS_LINUX for now because of it.

Also updated thread_restrictions.cc to use NOTREACHED() instead of LOG(FATAL)
in order for it to generate a DCHECK failure (it's only defined when
DCHECK_IS_ON()
so it makes sense to align it with that).

BUG=553459, 634552

Committed: https://crrev.com/5e69cff93950ddd3959f995dcffff2069ff097c0
Cr-Commit-Position: refs/heads/master@{#409980}
==========

[commit-bot: I haz the power, 2016-08-05 03:30:21]

Patchset 5 (id:??) landed as
https://crrev.com/5e69cff93950ddd3959f995dcffff2069ff097c0
Cr-Commit-Position: refs/heads/master@{#409980}