Remove --enable-ipv6 and --disable-ipv6 command line flags.

Remove --enable-ipv6 and --disable-ipv6 command line flags.

These were introduced for debugging and corner-case purposes. However, these were not permanently supported configurations - policy support was never added, which is a key indicator that these are options which will be supported going into the future.

--enable-ipv6 is hardly ever used. Internet searches for "--enable-ipv6 chrome" do not turn up many uses.

--disable-ipv6 is used more regularly though.

One place where --disable-ipv6 is used is in the SafeBrowsing tests. This was introduced in http://src.chromium.org/viewvc/chrome?revision=66196 to help accelerate test cycle times. The main issue was windows was resolving localhost to ::1 which made tests flaky.

The other general case where it has been used is when Chrome detects IPv6 availability but accesses are slow. Since this flag was added, Chrome added support for Happy Eyeballs, as well as improved the IPv6 probing for DNS, so the main motivations should be limited at this point.

BUG=344685
Committed: https://crrev.com/a391f4a06d66975a2a4da722ccd26935f39bd697
Cr-Commit-Position: refs/heads/master@{#332245}

[cbentzel, 2015-06-01 16:40:30]

cbentzel@chromium.org changed reviewers:
+ mmenke@chromium.org, shess@chromium.org

[cbentzel, 2015-06-01 16:40:31]

mmenke: io_thread
shess: safebrowsing_test

My primary concern with this CL revolves around removing disable_ipv6 for the
SafeBrowsing tests and whether it hurts performance on the windows bots.

[mmenke, 2015-06-01 17:05:54]

io_thread (And chrome_switches) LGTM

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc
File chrome/browser/io_thread.cc (left):

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc...
chrome/browser/io_thread.cc:206:
global_host_resolver->SetDefaultAddressFamily(net::ADDRESS_FAMILY_IPV4);
Hrm...net-internals is now the only user of SetDefaultAddressFamily, to force
IPv6 to be enabled.

That does allow one to force IPv6 to be enabled, to know if an issue is due to
broken probes or broken IPv6, but think it may make sense to just get rid of
SetDefaultAddressFamily entirely, at some point.

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc
File chrome/browser/io_thread.cc (right):

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc...
chrome/browser/io_thread.cc:200: 
nit: Remove extra blank line.

[Scott Hess - ex-Googler, 2015-06-01 17:07:27]

LGTM.

I have no idea why this was landed, but it really sounds like something that
should be in the trybot environment not the test setup.  Does it make the
trybots slower?

[cbentzel, 2015-06-01 17:14:04]

On 2015/06/01 17:07:27, Scott Hess wrote:
> LGTM.
> 
> I have no idea why this was landed, but it really sounds like something that
> should be in the trybot environment not the test setup.  Does it make the
> trybots slower?

I agree that this should have been trybot environment. I don't know yet if it
makes things slower. I'm going to measure to see if there are performance
regressions.

[cbentzel, 2015-06-01 17:16:32]

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc
File chrome/browser/io_thread.cc (left):

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc...
chrome/browser/io_thread.cc:206:
global_host_resolver->SetDefaultAddressFamily(net::ADDRESS_FAMILY_IPV4);
On 2015/06/01 17:05:53, mmenke wrote:
> Hrm...net-internals is now the only user of SetDefaultAddressFamily, to force
> IPv6 to be enabled.
> 
> That does allow one to force IPv6 to be enabled, to know if an issue is due to
> broken probes or broken IPv6, but think it may make sense to just get rid of
> SetDefaultAddressFamily entirely, at some point.

I have a separate CL to remove that from net-internals. After that's done I also
plan to remove SetDefaultAddressFamily.

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc
File chrome/browser/io_thread.cc (right):

https://codereview.chromium.org/1149303005/diff/1/chrome/browser/io_thread.cc...
chrome/browser/io_thread.cc:200: 
On 2015/06/01 17:05:53, mmenke wrote:
> nit: Remove extra blank line.

Done.

[cbentzel, 2015-06-01 18:13:26]

The CQ bit was checked by cbentzel@chromium.org

[cbentzel, 2015-06-01 18:13:27]

The patchset sent to the CQ was uploaded after l-g-t-m from mmenke@chromium.org,
shess@chromium.org
Link to the patchset: https://codereview.chromium.org/1149303005/#ps20001
(title: "Remove extra newline")

[commit-bot: I haz the power, 2015-06-01 18:15:30]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1149303005/20001

[commit-bot: I haz the power, 2015-06-01 19:42:48]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-06-01 19:44:41]

Patchset 2 (id:??) landed as
https://crrev.com/a391f4a06d66975a2a4da722ccd26935f39bd697
Cr-Commit-Position: refs/heads/master@{#332245}