Corrected access of null pointer in audioproc_f

Corrected access of null pointer in audioproc_f:
The previous CL that added the ability to add
and artificial nearend signal had an issue with
null pointer access.

This is addressed in this CL.

BUG=webrtc:6018
Committed: https://crrev.com/9e1e6c599dc0d911fe1888b8a1788a3e1d73117b
Cr-Commit-Position: refs/heads/master@{#15600}

[peah-webrtc, 2016-12-14 10:19:41]

Description was changed from

==========
Corrected access of null pointer


Corrected access of null pointer

BUG=
==========

to

==========
Corrected access of null pointer in audioproc_f:
The previous CL that added the ability to add
and artificial nearend signal had an issue with
null pointer access.

This is addressed in this CL.

BUG=webrtc:6018
==========

[peah-webrtc, 2016-12-14 10:20:01]

peah@webrtc.org changed reviewers:
+ ivoc@webrtc.org

[peah-webrtc, 2016-12-14 10:20:01]

ivoc@: PTAL

Thanks!

[ivoc, 2016-12-14 10:26:41]

https://codereview.webrtc.org/2573033003/diff/1/webrtc/modules/audio_processi...
File webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc (right):

https://codereview.webrtc.org/2573033003/diff/1/webrtc/modules/audio_processi...
webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc:237:
rtc::CheckedDivExact(sample_rate_hz, kChunksPerSecond), 1));
Maybe I'm missing something, but to me it looks like this statement is exactly
the same (minus formatting) as before the change. How does this prevent access
of a nullptr?

[peah-webrtc, 2016-12-14 10:33:41]

https://codereview.webrtc.org/2573033003/diff/1/webrtc/modules/audio_processi...
File webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc (right):

https://codereview.webrtc.org/2573033003/diff/1/webrtc/modules/audio_processi...
webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc:237:
rtc::CheckedDivExact(sample_rate_hz, kChunksPerSecond), 1));
On 2016/12/14 10:26:41, ivoc wrote:
> Maybe I'm missing something, but to me it looks like this statement is exactly
> the same (minus formatting) as before the change. How does this prevent access
> of a nullptr?

I think what happens is that the pointer artificial_nearend_file is lost  in the
std::move() operation on row 235, which in turn makes the call to
artificial_nearend_file->sample_rate() access an invalid pointer.

But please correct me if I'm wrong in that and that something else was wrong.
The former code fails when I run it, and the new code does not fail.

[ivoc, 2016-12-14 10:37:06]

LGTM

https://codereview.webrtc.org/2573033003/diff/1/webrtc/modules/audio_processi...
File webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc (right):

https://codereview.webrtc.org/2573033003/diff/1/webrtc/modules/audio_processi...
webrtc/modules/audio_processing/test/aec_dump_based_simulator.cc:237:
rtc::CheckedDivExact(sample_rate_hz, kChunksPerSecond), 1));
On 2016/12/14 10:33:40, peah-webrtc wrote:
> On 2016/12/14 10:26:41, ivoc wrote:
> > Maybe I'm missing something, but to me it looks like this statement is
exactly
> > the same (minus formatting) as before the change. How does this prevent
access
> > of a nullptr?
> 
> I think what happens is that the pointer artificial_nearend_file is lost  in
the
> std::move() operation on row 235, which in turn makes the call to
> artificial_nearend_file->sample_rate() access an invalid pointer.
> 
> But please correct me if I'm wrong in that and that something else was wrong.
> The former code fails when I run it, and the new code does not fail.

Ah right, that's a good point. It's easy to forget the effect of std::move (I
guess that's why it was missed in the initial CL).

[peah-webrtc, 2016-12-14 10:57:11]

The CQ bit was checked by peah@webrtc.org

[commit-bot: I haz the power, 2016-12-14 10:57:21]

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

[commit-bot: I haz the power, 2016-12-14 11:12:21]

CQ is committing da patch.

Bot data: {"patchset_id": 1, "attempt_start_ts": 1481713031432030, "parent_rev":
"fdecee31bca990958db8f9327ff42f0c061351a8", "commit_rev":
"22ab230690d8cae64ac1fded32c1024ec8f92d95"}

[commit-bot: I haz the power, 2016-12-14 11:12:23]

Description was changed from

==========
Corrected access of null pointer in audioproc_f:
The previous CL that added the ability to add
and artificial nearend signal had an issue with
null pointer access.

This is addressed in this CL.

BUG=webrtc:6018
==========

to

==========
Corrected access of null pointer in audioproc_f:
The previous CL that added the ability to add
and artificial nearend signal had an issue with
null pointer access.

This is addressed in this CL.

BUG=webrtc:6018

Review-Url: https://codereview.webrtc.org/2573033003
==========

[commit-bot: I haz the power, 2016-12-14 11:12:24]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-12-14 11:12:32]

Description was changed from

==========
Corrected access of null pointer in audioproc_f:
The previous CL that added the ability to add
and artificial nearend signal had an issue with
null pointer access.

This is addressed in this CL.

BUG=webrtc:6018

Review-Url: https://codereview.webrtc.org/2573033003
==========

to

==========
Corrected access of null pointer in audioproc_f:
The previous CL that added the ability to add
and artificial nearend signal had an issue with
null pointer access.

This is addressed in this CL.

BUG=webrtc:6018

Committed: https://crrev.com/9e1e6c599dc0d911fe1888b8a1788a3e1d73117b
Cr-Commit-Position: refs/heads/master@{#15600}
==========

[commit-bot: I haz the power, 2016-12-14 11:12:33]

Patchset 1 (id:??) landed as
https://crrev.com/9e1e6c599dc0d911fe1888b8a1788a3e1d73117b
Cr-Commit-Position: refs/heads/master@{#15600}