Server push cancellation: fix host resolver set up in URLRequestQuicTest

Server push cancellation: fix host resolver set up in URLRequestQuicTest
so that customized host resolver rules can be apply when QuicStreamFactory
resolves host. Also change to force quic on test.example.com:443 so that
https requests will always use QUIC, no TCP racing.

BUG=232040
Review-Url: https://codereview.chromium.org/2693083002
Cr-Commit-Position: refs/heads/master@{#450508}
Committed: https://chromium.googlesource.com/chromium/src/+/6f20e71dec40fa752fb16ade304cda16d537bbdd

[Zhongyi Shi, 2017-02-14 05:09:04]

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

[commit-bot: I haz the power, 2017-02-14 05:09:36]

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

[Zhongyi Shi, 2017-02-14 05:10:05]

zhongyi@chromium.org changed reviewers:
+ rch@chromium.org

[Zhongyi Shi, 2017-02-14 05:10:06]

This CL will fix the host resolver setup in tests, previous tests weren't using
the host resolver. PTAL!

[Ryan Hamilton, 2017-02-14 05:22:04]

Thanks for doing this!

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
File net/url_request/url_request_quic_unittest.cc (right):

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:65:
HostPortPair::FromString("test.example.com:443"));
nit: HostPortPair(kTestServerHost, 443));

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:68:
context_->set_host_resolver(host_resolver_.get());
Did the old code not actually work?

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:117: // Use a mapped host resolver
so that request for test.example.com (port 80)
Is port 80 really correct in this comment?

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:119: std::string map_rule = "MAP
test.example.com:443 test.example.com:" +
Is 443 required here?

[commit-bot: I haz the power, 2017-02-14 05:48:47]

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

[commit-bot: I haz the power, 2017-02-14 05:48:47]

Dry run: This issue passed the CQ dry run.

[Zhongyi Shi, 2017-02-14 05:52:57]

Thanks for help review, comments addressed, ptal :)

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
File net/url_request/url_request_quic_unittest.cc (right):

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:65:
HostPortPair::FromString("test.example.com:443"));
On 2017/02/14 05:22:04, Ryan Hamilton wrote:
> nit: HostPortPair(kTestServerHost, 443));

Done.

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:68:
context_->set_host_resolver(host_resolver_.get());
On 2017/02/14 05:22:03, Ryan Hamilton wrote:
> Did the old code not actually work?

The old code is not using the host resolver rules we set here, the host_resolver
in HttpNetworkSession::Params is overwritten in TestURLRequestContext::Init().

https://cs.chromium.org/chromium/src/net/url_request/url_request_test_util.cc...

The old tests happen to work because no host_resolve rules needed to connect to
the local server as the request url is like:
https://test.example.com:6121/hello.txt.

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:117: // Use a mapped host resolver
so that request for test.example.com (port 80)
On 2017/02/14 05:22:04, Ryan Hamilton wrote:
> Is port 80 really correct in this comment?

Done.

https://codereview.chromium.org/2693083002/diff/1/net/url_request/url_request...
net/url_request/url_request_quic_unittest.cc:119: std::string map_rule = "MAP
test.example.com:443 test.example.com:" +
On 2017/02/14 05:22:04, Ryan Hamilton wrote:
> Is 443 required here?

Aha, it's not required here. I was following the
https://www.chromium.org/quic/playing-with-quic though. Happy to revert this.

[Ryan Hamilton, 2017-02-14 22:19:58]

The CQ bit was checked by rch@chromium.org

[Ryan Hamilton, 2017-02-14 22:19:59]

lgtm

[commit-bot: I haz the power, 2017-02-14 22:20:41]

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, 2017-02-14 23:20:43]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1487110798934550,
"parent_rev": "a7f5aca8dad4b6166f6515b927b3a0e9fe6f832f", "commit_rev":
"6f20e71dec40fa752fb16ade304cda16d537bbdd"}

[commit-bot: I haz the power, 2017-02-14 23:21:18]

Description was changed from

==========
Server push cancellation: fix host resolver set up in URLRequestQuicTest
so that customized host resolver rules can be apply when QuicStreamFactory
resolves host. Also change to force quic on test.example.com:443 so that
https requests will always use QUIC, no TCP racing.

BUG=232040
==========

to

==========
Server push cancellation: fix host resolver set up in URLRequestQuicTest
so that customized host resolver rules can be apply when QuicStreamFactory
resolves host. Also change to force quic on test.example.com:443 so that
https requests will always use QUIC, no TCP racing.

BUG=232040

Review-Url: https://codereview.chromium.org/2693083002
Cr-Commit-Position: refs/heads/master@{#450508}
Committed:
https://chromium.googlesource.com/chromium/src/+/6f20e71dec40fa752fb16ade304c...
==========

[commit-bot: I haz the power, 2017-02-14 23:21:20]

Committed patchset #2 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/6f20e71dec40fa752fb16ade304c...