Using fake clock for TURN port tests and un-disabling some tests.

Using fake clock for TURN port tests and un-disabling some tests.

The fake clock has a few advantages:
1. It lets use verify that operations take the expected number of
   round trips.
2. It makes the tests faster by letting us remove the equivalent
   of "Sleep(500)" all over the tests.
3. It makes the tests less flaky, because sometimes sleeping for
   500ms or waiting for 1s is not enough.

R=honghaiz@webrtc.org, pthatcher@webrtc.org

Committed: https://crrev.com/716d07a2414389638eba6ed03b704bf752a269ef
Cr-Commit-Position: refs/heads/master@{#13304}

[Taylor Brandstetter, 2016-06-24 16:33:40]

While I was updating the TURN test yesterday, I noticed some ugly things here
(such as rtc::ProcessMessages(500)), so I decided to clean it up. Take a look
when you have time.

https://codereview.webrtc.org/2097793003/diff/1/webrtc/base/gunit.h
File webrtc/base/gunit.h (right):

https://codereview.webrtc.org/2097793003/diff/1/webrtc/base/gunit.h#newcode26
webrtc/base/gunit.h:26: !(ex) && rtc::SystemTimeMillis() < start + (timeout);) {
\
I'm adding parentheses around "timeout" because this macro was breaking if I
tried to pass anything funky as the timeout argument. Such as a ternary
expression; it compiled fine but resulted in an endless loop due to operator
precedence and such. "time < start + condition ? timeout_a : timeout_b" equates
to "(bool)((time < start + condition) ? timeout_a : timeout_b)"

https://codereview.webrtc.org/2097793003/diff/1/webrtc/p2p/base/turnport_unit...
File webrtc/p2p/base/turnport_unittest.cc (right):

https://codereview.webrtc.org/2097793003/diff/1/webrtc/p2p/base/turnport_unit...
webrtc/p2p/base/turnport_unittest.cc:70: static constexpr unsigned int
kSimulatedRtt = 50;
This might be overkill, but I just thought "if we're using the fake clock to
make the test less flaky, we might as well do this too since we can". It would
catch an issue such as if, say, we accidentally started making the port send an
extra Allocate request that isn't necessary before it signals that it's ready.

https://codereview.webrtc.org/2097793003/diff/20001/webrtc/base/virtualsocket...
File webrtc/base/virtualsocketserver.cc (right):

https://codereview.webrtc.org/2097793003/diff/20001/webrtc/base/virtualsocket...
webrtc/base/virtualsocketserver.cc:776: uint32_t delay =
GetRandomTransitDelay();
This is needed for the "TCP release allocation" test. When the TURN port is
destroyed it sends a packet to release the allocation, but if we don't delay the
"MSG_ID_DISCONNECT", the receiving socket is already destroyed by the time the
packet arrives, in essence.

[Taylor Brandstetter, 2016-06-24 16:34:32]

deadbeef@webrtc.org changed reviewers:
+ honghaiz@webrtc.org, pthatcher@webrtc.org

[Taylor Brandstetter, 2016-06-24 16:34:32]

Forgot to add reviewers. Repeat of previous comment:

While I was updating the TURN test yesterday, I noticed some ugly things here
(such as rtc::ProcessMessages(500)), so I decided to clean it up. Take a look
when you have time.

[honghaiz3, 2016-06-27 18:21:36]

lgtm

[pthatcher1, 2016-06-27 19:42:02]

lgtm

Nice work.

[Taylor Brandstetter, 2016-06-27 20:13:49]

The CQ bit was checked by deadbeef@webrtc.org

[commit-bot: I haz the power, 2016-06-27 20:13:57]

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, 2016-06-27 20:25:06]

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

[commit-bot: I haz the power, 2016-06-27 20:25:07]

Try jobs failed on following builders:
  android_arm64_rel on master.tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/android_arm64_rel/build...)

[commit-bot: I haz the power, 2016-06-27 21:07:55]

Description was changed from

==========
Using fake clock for TURN port tests and un-disabling some tests.

The fake clock has a few advantages:
1. It lets use verify that operations take the expected number of
   round trips.
2. It makes the tests faster by letting us remove the equivalent
   of "Sleep(500)" all over the tests.
3. It makes the tests less flaky, because sometimes sleeping for
   500ms or waiting for 1s is not enough.
==========

to

==========
Using fake clock for TURN port tests and un-disabling some tests.

The fake clock has a few advantages:
1. It lets use verify that operations take the expected number of
   round trips.
2. It makes the tests faster by letting us remove the equivalent
   of "Sleep(500)" all over the tests.
3. It makes the tests less flaky, because sometimes sleeping for
   500ms or waiting for 1s is not enough.

R=honghaiz@webrtc.org, pthatcher@webrtc.org

Committed: https://crrev.com/716d07a2414389638eba6ed03b704bf752a269ef
Cr-Commit-Position: refs/heads/master@{#13304}
==========

[Taylor Brandstetter, 2016-06-27 21:07:57]

Description was changed from

==========
Using fake clock for TURN port tests and un-disabling some tests.

The fake clock has a few advantages:
1. It lets use verify that operations take the expected number of
   round trips.
2. It makes the tests faster by letting us remove the equivalent
   of "Sleep(500)" all over the tests.
3. It makes the tests less flaky, because sometimes sleeping for
   500ms or waiting for 1s is not enough.

R=honghaiz@webrtc.org, pthatcher@webrtc.org

Committed: https://crrev.com/716d07a2414389638eba6ed03b704bf752a269ef
Cr-Commit-Position: refs/heads/master@{#13304}
==========

to

==========
Using fake clock for TURN port tests and un-disabling some tests.

The fake clock has a few advantages:
1. It lets use verify that operations take the expected number of
   round trips.
2. It makes the tests faster by letting us remove the equivalent
   of "Sleep(500)" all over the tests.
3. It makes the tests less flaky, because sometimes sleeping for
   500ms or waiting for 1s is not enough.

R=honghaiz@webrtc.org, pthatcher@webrtc.org

Committed:
https://chromium.googlesource.com/external/webrtc/+/716d07a2414389638eba6ed03...
==========

[commit-bot: I haz the power, 2016-06-27 21:07:57]

Patchset 2 (id:??) landed as
https://crrev.com/716d07a2414389638eba6ed03b704bf752a269ef
Cr-Commit-Position: refs/heads/master@{#13304}

[Taylor Brandstetter, 2016-06-27 21:07:59]

Committed patchset #2 (id:20001) manually as
716d07a2414389638eba6ed03b704bf752a269ef (presubmit successful).