Replace RefCountImpl with rtc::RefCountedObject.

Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org

Committed: https://crrev.com/1d1944187fe1304b5933bc8f8a8a36a54da3e705
Cr-Commit-Position: refs/heads/master@{#12075}

[pbos-webrtc, 2015-11-27 13:07:03]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2015-11-27 13:07:11]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/1

[commit-bot: I haz the power, 2015-11-27 13:09:12]

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

[commit-bot: I haz the power, 2015-11-27 13:09:12]

Dry run: Try jobs failed on following builders:
  ios32_sim_dbg on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios32_sim_dbg/builds/3608)
  mac_x64_gn_dbg on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/mac_x64_gn_dbg/builds/5569)

[pbos-webrtc, 2015-11-27 13:11:02]

include header typo

[pbos-webrtc, 2015-11-27 13:11:20]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2015-11-27 13:11:27]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/20001

[commit-bot: I haz the power, 2015-11-27 13:39:03]

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

[commit-bot: I haz the power, 2015-11-27 13:39:04]

Dry run: Try jobs failed on following builders:
  linux_asan on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/linux_asan/builds/10925)

[pbos-webrtc, 2015-11-27 17:57:40]

Patchset #3 (id:40001) has been deleted

[pbos-webrtc, 2015-11-27 17:57:48]

scoped_refptr everywhere

[pbos-webrtc, 2015-11-27 17:59:04]

Description was changed from

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

BUG=
R=tommi@webrtc.org
==========

to

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org
==========

[pbos-webrtc, 2015-11-27 17:59:51]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2015-11-27 18:00:06]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/60001

[commit-bot: I haz the power, 2015-11-27 18:03:00]

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

[commit-bot: I haz the power, 2015-11-27 18:03:00]

Dry run: Try jobs failed on following builders:
  ios64_sim_dbg on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios64_sim_dbg/builds/3611)
  ios_arm64_dbg on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios_arm64_dbg/builds/5924)
  ios_arm64_rel on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios_arm64_rel/builds/5860)
  ios_dbg on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios_dbg/builds/11079)
  ios_rel on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios_rel/builds/9733)
  win_x64_rel on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/win_x64_rel/builds/10970)

[pbos-webrtc, 2015-11-27 18:04:24]

fix compile errors on win/ios

[pbos-webrtc, 2015-11-27 18:11:55]

On 2015/11/27 18:04:24, pbos-webrtc wrote:
> fix compile errors on win/ios

PTAL, this reduces code duplication but also introduces scoped_refptrs where
they should be (in a couple of places).

Related: I think scoped_refptr: operator T*() const { return ptr_; } is
dangerous (since you can do T* foo = scoped_refptr<T>(new T)) and have UAFs
easily.

This happens now if you do VideoCodingModule* foo =
VideoCodingModule::Create(...); since we've then gotten rid of all the
references to foo.

[pbos-webrtc, 2015-11-27 18:12:12]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2015-11-27 18:12:26]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/80001

[pbos-webrtc, 2015-11-27 18:46:42]

remove vcm reference in Destroy

[pbos-webrtc, 2016-01-07 16:19:50]

rebase

[pbos-webrtc, 2016-01-07 16:21:37]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-07 16:21:46]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/120001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/120001

[pbos-webrtc, 2016-01-07 16:21:48]

PTAL

[tommi, 2016-01-07 16:41:40]

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/audio_dev...
File webrtc/modules/audio_device/test/audio_device_test_api.cc (right):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/audio_dev...
webrtc/modules/audio_device/test/audio_device_test_api.cc:49:
AudioEventObserverAPI(rtc::scoped_refptr<AudioDeviceModule> audioDevice)
const&  otherwise we're adding calls to AddRef/Release

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
File webrtc/modules/video_capture/ios/video_capture_ios.mm (left):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
webrtc/modules/video_capture/ios/video_capture_ios.mm:54: return NULL;
uhm... I guess we've just been leaking then :-/

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
File webrtc/modules/video_capture/video_capture_impl.cc (right):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
webrtc/modules/video_capture/video_capture_impl.cc:32: externalCapture =
implementation.get();
hmm... should we file a bug for this?

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/voice_engine/shar...
File webrtc/voice_engine/shared_data.h (right):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/voice_engine/shar...
webrtc/voice_engine/shared_data.h:42: void
set_audio_device(rtc::scoped_refptr<AudioDeviceModule> audio_device);
const &

[commit-bot: I haz the power, 2016-01-07 17:02:33]

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

[commit-bot: I haz the power, 2016-01-07 17:02:33]

Dry run: Try jobs failed on following builders:
  win_baremetal on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/win_baremetal/builds/8227)

[pbos-webrtc, 2016-01-07 17:08:37]

feedback

[pbos-webrtc, 2016-01-07 17:23:01]

scoped_refptr in AudioTransportAPITest

[pbos-webrtc, 2016-01-07 17:23:16]

PTAL

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/audio_dev...
File webrtc/modules/audio_device/test/audio_device_test_api.cc (right):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/audio_dev...
webrtc/modules/audio_device/test/audio_device_test_api.cc:49:
AudioEventObserverAPI(rtc::scoped_refptr<AudioDeviceModule> audioDevice)
On 2016/01/07 16:41:40, tommi-webrtc wrote:
> const&  otherwise we're adding calls to AddRef/Release

Done.

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
File webrtc/modules/video_capture/ios/video_capture_ios.mm (left):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
webrtc/modules/video_capture/ios/video_capture_ios.mm:54: return NULL;
On 2016/01/07 16:41:40, tommi-webrtc wrote:
> uhm... I guess we've just been leaking then :-/

Acknowledged.

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
File webrtc/modules/video_capture/video_capture_impl.cc (right):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/modules/video_cap...
webrtc/modules/video_capture/video_capture_impl.cc:32: externalCapture =
implementation.get();
On 2016/01/07 16:41:40, tommi-webrtc wrote:
> hmm... should we file a bug for this?

Maybe, not sure if anyone wants to own it though, so it might just be a
tumbleweed bug. :\

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/voice_engine/shar...
File webrtc/voice_engine/shared_data.h (right):

https://codereview.webrtc.org/1477013005/diff/120001/webrtc/voice_engine/shar...
webrtc/voice_engine/shared_data.h:42: void
set_audio_device(rtc::scoped_refptr<AudioDeviceModule> audio_device);
On 2016/01/07 16:41:40, tommi-webrtc wrote:
> const &

Done.

[pbos-webrtc, 2016-01-07 17:23:23]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2016-01-07 17:23:30]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/160001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/160001

[commit-bot: I haz the power, 2016-01-07 20:29:07]

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

[commit-bot: I haz the power, 2016-01-07 20:29:08]

Dry run: Try jobs failed on following builders:
  linux_baremetal on tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/linux_baremetal/builds/...)

[pbos-webrtc, 2016-03-16 11:14:33]

Tommi do you think we wanna land something like this? This probably fails rebase
by a ton.

[tommi (sloooow) - chröme, 2016-03-16 16:00:37]

tommi@chromium.org changed reviewers:
+ tommi@chromium.org

[tommi (sloooow) - chröme, 2016-03-16 16:00:38]

yes - have you given rebase a try?

[pbos-webrtc, 2016-03-16 20:34:34]

rebase

[pbos-webrtc, 2016-03-16 20:37:49]

rebase seemed doable, that's surprising

PTAL

[pbos-webrtc, 2016-03-16 20:37:59]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2016-03-16 20:38:07]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/180001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/180001

[commit-bot: I haz the power, 2016-03-16 22:37:58]

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

[commit-bot: I haz the power, 2016-03-16 22:37:59]

Dry run: This issue passed the CQ dry run.

[tommi, 2016-03-17 09:15:12]

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/modules/audio_dev...
File webrtc/modules/audio_device/test/audio_device_test_api.cc (right):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/modules/audio_dev...
webrtc/modules/audio_device/test/audio_device_test_api.cc:164:
audio_device_.release();
ehm, yeah you'd think that .release() calls Release(), right? nope! :)
you either want this instead:
audio_device_ = nullptr;

or, if you want to maintain the previous check (which might be a better idea),
this:

EXPECT_EQ(0, audio_device_.release()->Release());

The explicit call to AddRef() that was here, will now be done implicitly during
assignment to audio_device_.

same below.

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer.cc
File webrtc/test/vcm_capturer.cc (left):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer...
webrtc/test/vcm_capturer.cc:90: // TODO(pbos): How do I destroy the
VideoCaptureModule? This still leaves
is this still true? (if so, please leave the comment in)

[pbos-webrtc, 2016-03-17 15:04:00]

PTAL

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/modules/audio_dev...
File webrtc/modules/audio_device/test/audio_device_test_api.cc (right):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/modules/audio_dev...
webrtc/modules/audio_device/test/audio_device_test_api.cc:164:
audio_device_.release();
On 2016/03/17 09:15:12, tommi-webrtc wrote:
> ehm, yeah you'd think that .release() calls Release(), right? nope! :)
> you either want this instead:
> audio_device_ = nullptr;
> 
> or, if you want to maintain the previous check (which might be a better idea),
> this:
> 
> EXPECT_EQ(0, audio_device_.release()->Release());
> 
> The explicit call to AddRef() that was here, will now be done implicitly
during
> assignment to audio_device_.
> 
> same below.

done (as = nullptr)

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer.cc
File webrtc/test/vcm_capturer.cc (left):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer...
webrtc/test/vcm_capturer.cc:90: // TODO(pbos): How do I destroy the
VideoCaptureModule? This still leaves
On 2016/03/17 09:15:12, tommi-webrtc wrote:
> is this still true? (if so, please leave the comment in)

Not true anymore, since the creation does AddRef() now the assignment Release()s
properly.

[pbos-webrtc, 2016-03-17 15:04:17]

remove bogus release() calls

[pbos-webrtc, 2016-03-17 15:05:35]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2016-03-17 15:05:43]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/200001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/200001

[tommi, 2016-03-17 15:10:57]

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer.cc
File webrtc/test/vcm_capturer.cc (left):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer...
webrtc/test/vcm_capturer.cc:90: // TODO(pbos): How do I destroy the
VideoCaptureModule? This still leaves
On 2016/03/17 15:04:00, pbos-webrtc wrote:
> On 2016/03/17 09:15:12, tommi-webrtc wrote:
> > is this still true? (if so, please leave the comment in)
> 
> Not true anymore, since the creation does AddRef() now the assignment
Release()s
> properly.

I'm not following... the previous code already did vcm_->Release(), what's
different?

https://codereview.webrtc.org/1477013005/diff/200001/webrtc/modules/audio_dev...
File webrtc/modules/audio_device/test/audio_device_test_api.cc (right):

https://codereview.webrtc.org/1477013005/diff/200001/webrtc/modules/audio_dev...
webrtc/modules/audio_device/test/audio_device_test_api.cc:164: audio_device_ =
nullptr;
nit: since you went with this approach, you might as well remove this line since
the next line is assignment anyway.

Btw, did you test that the behavior is still that Release() returns 0?

[commit-bot: I haz the power, 2016-03-17 17:06:11]

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

[commit-bot: I haz the power, 2016-03-17 17:06:12]

Dry run: Try jobs failed on following builders:
  android_arm64_rel on tryserver.webrtc (JOB_TIMED_OUT, no build URL)
  win_baremetal on tryserver.webrtc (JOB_TIMED_OUT, no build URL)

[pbos-webrtc, 2016-03-21 10:04:11]

add back Release()s in test

[pbos-webrtc, 2016-03-21 10:04:31]

PTAL again, went with Release() to check it.

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer.cc
File webrtc/test/vcm_capturer.cc (left):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer...
webrtc/test/vcm_capturer.cc:90: // TODO(pbos): How do I destroy the
VideoCaptureModule? This still leaves
On 2016/03/17 15:10:56, tommi-webrtc wrote:
> On 2016/03/17 15:04:00, pbos-webrtc wrote:
> > On 2016/03/17 09:15:12, tommi-webrtc wrote:
> > > is this still true? (if so, please leave the comment in)
> > 
> > Not true anymore, since the creation does AddRef() now the assignment
> Release()s
> > properly.
> 
> I'm not following... the previous code already did vcm_->Release(), what's
> different?

That it did not do AddRef() which is now done implicitly through the
scoped_refptr.

https://codereview.webrtc.org/1477013005/diff/200001/webrtc/modules/audio_dev...
File webrtc/modules/audio_device/test/audio_device_test_api.cc (right):

https://codereview.webrtc.org/1477013005/diff/200001/webrtc/modules/audio_dev...
webrtc/modules/audio_device/test/audio_device_test_api.cc:164: audio_device_ =
nullptr;
On 2016/03/17 15:10:56, tommi-webrtc wrote:
> nit: since you went with this approach, you might as well remove this line
since
> the next line is assignment anyway.
> 
> Btw, did you test that the behavior is still that Release() returns 0?

Added back Release() and == 0 checks.

[pbos-webrtc, 2016-03-21 10:42:33]

Description was changed from

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org
==========

to

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229, webrtc:5660
R=tommi@webrtc.org
==========

[pbos-webrtc, 2016-03-21 10:42:49]

The CQ bit was checked by pbos@webrtc.org to run a CQ dry run

[commit-bot: I haz the power, 2016-03-21 10:42:59]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1477013005/220001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1477013005/220001

[tommi, 2016-03-21 11:23:21]

lgtm

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer.cc
File webrtc/test/vcm_capturer.cc (left):

https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer...
webrtc/test/vcm_capturer.cc:90: // TODO(pbos): How do I destroy the
VideoCaptureModule? This still leaves
On 2016/03/21 10:04:31, pbos-webrtc wrote:
> On 2016/03/17 15:10:56, tommi-webrtc wrote:
> > On 2016/03/17 15:04:00, pbos-webrtc wrote:
> > > On 2016/03/17 09:15:12, tommi-webrtc wrote:
> > > > is this still true? (if so, please leave the comment in)
> > > 
> > > Not true anymore, since the creation does AddRef() now the assignment
> > Release()s
> > > properly.
> > 
> > I'm not following... the previous code already did vcm_->Release(), what's
> > different?
> 
> That it did not do AddRef() which is now done implicitly through the
> scoped_refptr.

Oh I see!  Yikes, that could also be an invite to a use-after-free issue (e.g.
if someone else AddRef+Releases the object). Do you know if there are more cases
of this (specifically in non-test code)?

[pbos-webrtc, 2016-03-21 11:55:32]

On 2016/03/21 11:23:21, tommi-webrtc wrote:
> lgtm
> 
>
https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer.cc
> File webrtc/test/vcm_capturer.cc (left):
> 
>
https://codereview.webrtc.org/1477013005/diff/180001/webrtc/test/vcm_capturer...
> webrtc/test/vcm_capturer.cc:90: // TODO(pbos): How do I destroy the
> VideoCaptureModule? This still leaves
> On 2016/03/21 10:04:31, pbos-webrtc wrote:
> > On 2016/03/17 15:10:56, tommi-webrtc wrote:
> > > On 2016/03/17 15:04:00, pbos-webrtc wrote:
> > > > On 2016/03/17 09:15:12, tommi-webrtc wrote:
> > > > > is this still true? (if so, please leave the comment in)
> > > > 
> > > > Not true anymore, since the creation does AddRef() now the assignment
> > > Release()s
> > > > properly.
> > > 
> > > I'm not following... the previous code already did vcm_->Release(), what's
> > > different?
> > 
> > That it did not do AddRef() which is now done implicitly through the
> > scoped_refptr.
> 
> Oh I see!  Yikes, that could also be an invite to a use-after-free issue (e.g.
> if someone else AddRef+Releases the object). Do you know if there are more
cases
> of this (specifically in non-test code)?

Don't know any more really-incorrect usages than this. This barely requires
Chromium code updates (just updating an include, is what it looks like so far),
so that code is "probably safe". Capture code might've been incorrectly used in
Firefox/by other third-party developers though. But it's definitely a very
inviting foot-shooting gun.

[pbos-webrtc, 2016-03-21 12:08:39]

Description was changed from

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229, webrtc:5660
R=tommi@webrtc.org
==========

to

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org
==========

[commit-bot: I haz the power, 2016-03-21 12:53:12]

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

[commit-bot: I haz the power, 2016-03-21 12:53:14]

Dry run: This issue passed the CQ dry run.

[pbos-webrtc, 2016-03-21 15:44:47]

Description was changed from

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org
==========

to

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org

Committed:
https://chromium.googlesource.com/external/webrtc/+/1d1944187fe1304b5933bc8f8...
==========

[commit-bot: I haz the power, 2016-03-21 15:44:50]

Description was changed from

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org

Committed:
https://chromium.googlesource.com/external/webrtc/+/1d1944187fe1304b5933bc8f8...
==========

to

==========
Replace RefCountImpl with rtc::RefCountedObject.

Removes code duplication and use of the dangerous public destructor in
RefCountImpl.

Also making wider use of scoped_refptr and fixing various leaks in the
process.

BUG=webrtc:5229
R=tommi@webrtc.org

Committed: https://crrev.com/1d1944187fe1304b5933bc8f8a8a36a54da3e705
Cr-Commit-Position: refs/heads/master@{#12075}
==========

[pbos-webrtc, 2016-03-21 15:44:50]

Committed patchset #11 (id:220001) manually as
1d1944187fe1304b5933bc8f8a8a36a54da3e705 (presubmit successful).

[commit-bot: I haz the power, 2016-03-21 15:44:52]

Patchset 11 (id:??) landed as
https://crrev.com/1d1944187fe1304b5933bc8f8a8a36a54da3e705
Cr-Commit-Position: refs/heads/master@{#12075}