Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identity during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
Committed: https://crrev.com/dd34c46303e555ce506cf683eedf3bc95033548d
Cr-Commit-Position: refs/heads/master@{#406995}

[Sam McNally, 2016-07-20 02:59:49]

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

[commit-bot: I haz the power, 2016-07-20 03:00:31]

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

[Sam McNally, 2016-07-20 03:02:11]

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

[commit-bot: I haz the power, 2016-07-20 03:02:46]

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, 2016-07-20 03:08:57]

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

[commit-bot: I haz the power, 2016-07-20 03:08:57]

Dry run: Try jobs failed on following builders:
  win_chromium_compile_dbg_ng on master.tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_comp...)
  win_chromium_x64_rel_ng on master.tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_x64_...)

[Sam McNally, 2016-07-20 03:14:15]

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

[commit-bot: I haz the power, 2016-07-20 03:14:53]

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, 2016-07-20 04:01:13]

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

[commit-bot: I haz the power, 2016-07-20 04:01:14]

Dry run: This issue passed the CQ dry run.

[Sam McNally, 2016-07-20 04:01:42]

Patchset #1 (id:1) has been deleted

[Sam McNally, 2016-07-20 04:01:51]

Patchset #1 (id:20001) has been deleted

[Sam McNally, 2016-07-20 06:31:41]

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

[commit-bot: I haz the power, 2016-07-20 06:31:57]

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

[Sam McNally, 2016-07-20 06:44:50]

Patchset #1 (id:40001) has been deleted

[Sam McNally, 2016-07-20 06:46:26]

Description was changed from

==========
Use a thread checker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identify during shutdown, causing fun
errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

to

==========
Use a thread checker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identify during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

[Sam McNally, 2016-07-20 06:47:43]

Description was changed from

==========
Use a thread checker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identify during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

to

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identify during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

[Sam McNally, 2016-07-20 06:48:05]

sammc@chromium.org changed reviewers:
+ timvolodine@chromium.org

[commit-bot: I haz the power, 2016-07-20 07:18:55]

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

[commit-bot: I haz the power, 2016-07-20 07:18:56]

Dry run: This issue passed the CQ dry run.

[timvolodine, 2016-07-20 18:41:56]

yeah, weird error message...
in description should "identify" -> "identity"?

https://codereview.chromium.org/2160913005/diff/60001/content/browser/device_...
File content/browser/device_sensors/device_sensor_host.h (right):

https://codereview.chromium.org/2160913005/diff/60001/content/browser/device_...
content/browser/device_sensors/device_sensor_host.h:25: // All methods below to
be called on the IO thread.
hmm I understand this is now actually enforced in render_process_host_impl.cc by
adding on io_task_runner. Can we at least add a corresponding dcheck somewhere
in create/construction of this host?

[Sam McNally, 2016-07-20 23:40:04]

Description was changed from

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identify during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

to

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identity during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

[Sam McNally, 2016-07-20 23:45:25]

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

[commit-bot: I haz the power, 2016-07-20 23:45:44]

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

[Sam McNally, 2016-07-21 00:09:29]

On 2016/07/20 18:41:56, timvolodine wrote:
> yeah, weird error message...
> in description should "identify" -> "identity"?
> 
Yes, done.

https://codereview.chromium.org/2160913005/diff/60001/content/browser/device_...
File content/browser/device_sensors/device_sensor_host.h (right):

https://codereview.chromium.org/2160913005/diff/60001/content/browser/device_...
content/browser/device_sensors/device_sensor_host.h:25: // All methods below to
be called on the IO thread.
On 2016/07/20 18:41:56, timvolodine wrote:
> hmm I understand this is now actually enforced in render_process_host_impl.cc
by
> adding on io_task_runner. Can we at least add a corresponding dcheck somewhere
> in create/construction of this host?

Done.

[commit-bot: I haz the power, 2016-07-21 00:59:06]

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

[commit-bot: I haz the power, 2016-07-21 00:59:07]

Dry run: This issue passed the CQ dry run.

[timvolodine, 2016-07-21 15:58:51]

On 2016/07/21 00:09:29, Sam McNally wrote:
> On 2016/07/20 18:41:56, timvolodine wrote:
> > yeah, weird error message...
> > in description should "identify" -> "identity"?
> > 
> Yes, done.
> 
>
https://codereview.chromium.org/2160913005/diff/60001/content/browser/device_...
> File content/browser/device_sensors/device_sensor_host.h (right):
> 
>
https://codereview.chromium.org/2160913005/diff/60001/content/browser/device_...
> content/browser/device_sensors/device_sensor_host.h:25: // All methods below
to
> be called on the IO thread.
> On 2016/07/20 18:41:56, timvolodine wrote:
> > hmm I understand this is now actually enforced in
render_process_host_impl.cc
> by
> > adding on io_task_runner. Can we at least add a corresponding dcheck
somewhere
> > in create/construction of this host?
> 
> Done.

lgtm, thanks!

[Sam McNally, 2016-07-21 23:41:25]

The CQ bit was checked by sammc@chromium.org

[commit-bot: I haz the power, 2016-07-21 23:41:56]

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, 2016-07-21 23:47:04]

Description was changed from

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identity during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

to

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identity during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

[commit-bot: I haz the power, 2016-07-21 23:47:05]

Committed patchset #2 (id:80001)

[commit-bot: I haz the power, 2016-07-21 23:50:43]

Description was changed from

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identity during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988
==========

to

==========
Use a ThreadChecker in DeviceSensorHost instead of DCHECK_CURRENTLY_ON.

BrowserThreads forget their identity during shutdown, causing
nonsensical errors like
"Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::IO).
Must be called on Chrome_IOThread; actually called on Chrome_IOThread."
if DCHECK_CURRENTLY_ON is used in something that is called from a
MessageLoop shutdown observer.

This CL fixes this problem in DeviceSensorHost by using a ThreadChecker
instead.

BUG=625988

Committed: https://crrev.com/dd34c46303e555ce506cf683eedf3bc95033548d
Cr-Commit-Position: refs/heads/master@{#406995}
==========

[commit-bot: I haz the power, 2016-07-21 23:50:44]

Patchset 2 (id:??) landed as
https://crrev.com/dd34c46303e555ce506cf683eedf3bc95033548d
Cr-Commit-Position: refs/heads/master@{#406995}