Use FakeServer-based invalidations for Sync tests (try #2)

Use FakeServer-based invalidations for Sync tests

Note: this is a second go-round for this CL. It was originally committed
as r267422 but was reverted due to a compile error in some dead code.
The dead code was deleted in https://codereview.chromium.org/264773007/
 (r267686). See https://codereview.chromium.org/234113002/ for the
original CL.

This CL creates a new InvalidationService implementation,
FakeServerInvalidationService, to remove FakeServer-based tests'
dependency on the Python XMPP server.

Another major change is that FakeServer's HandleCommand is now executed
on the UI thread and locking has been removed from FakeServer itself.

BUG=323265
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=268311

[rlarocque, 2014-05-02 20:51:21]

https://codereview.chromium.org/267723012/diff/1/sync/test/fake_server/fake_s...
File sync/test/fake_server/fake_server_http_post_provider.cc (right):

https://codereview.chromium.org/267723012/diff/1/sync/test/fake_server/fake_s...
sync/test/fake_server/fake_server_http_post_provider.cc:87: const int
kTimeoutSecs = 5;
Is a timeout really necessary?

If there's a bug that causes us to hang, then the test framework should kill the
test.  I don't think it's this part of the code's responsibility to work around
serious logic errors.

[pval...(no longer on Chromium), 2014-05-05 18:33:36]

https://codereview.chromium.org/267723012/diff/1/sync/test/fake_server/fake_s...
File sync/test/fake_server/fake_server_http_post_provider.cc (right):

https://codereview.chromium.org/267723012/diff/1/sync/test/fake_server/fake_s...
sync/test/fake_server/fake_server_http_post_provider.cc:87: const int
kTimeoutSecs = 5;
On 2014/05/02 20:51:21, rlarocque wrote:
> Is a timeout really necessary?
> 
> If there's a bug that causes us to hang, then the test framework should kill
the
> test.  I don't think it's this part of the code's responsibility to work
around
> serious logic errors.

I agree; I can't justify the added complexity.

FYI I also found a bug here (checking *error_code before it was set) that it now
resolved.

https://codereview.chromium.org/267723012/diff/20001/chrome/chrome_tests.gypi
File chrome/chrome_tests.gypi (right):

https://codereview.chromium.org/267723012/diff/20001/chrome/chrome_tests.gypi...
chrome/chrome_tests.gypi:1004:
'browser/extensions/api/browser/browser_apitest.cc',
I must have synced to trunk on Friday?

[rlarocque, 2014-05-05 18:38:36]

lgtm

[pval...(no longer on Chromium), 2014-05-05 19:54:24]

The CQ bit was checked by pvalenzuela@chromium.org

[commit-bot: I haz the power, 2014-05-05 19:55:57]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/pvalenzuela@chromium.org/267723012/20001

[commit-bot: I haz the power, 2014-05-05 23:23:49]

Change committed as 268311