Use --icf=all instead of --icf=safe

Use --icf=all instead of --icf=safe

With --icf=safe, functions whose address is taken aren't folded.
Folding could change the results of function pointer comparisons.

However, on x86_64 with pie, --icf=safe disables almost all folding,
for technical reasons. Since using --icf=all will save 7 MB on x86_64
and all tests seem to pass with it, switch to --icf=all. If we end up
having a function where we absolutely need equality, we can pass
--keep-unique,<symbol-name> for that symbol.

And if this turns out to be some maintenance burden, we can switch
back to --icf=safe.

Change the flag on all platforms where we use gold, not just x86_64, for
consistency so that if we end up having problems with this, we notice
it faster.

BUG=76280, 492177
Committed: https://crrev.com/45ab39af3cc293e150158bf06515a47319db9f79
Cr-Commit-Position: refs/heads/master@{#331474}

[Nico, 2015-05-26 20:40:55]

thakis@chromium.org changed reviewers:
+ thestig@chromium.org

[Lei Zhang, 2015-05-26 21:26:34]

Do you have an unstripped chrome binary build with --ifc=all handy? I'm
wondering if Breakpad's dump_syms tool will handle it.

[Nico, 2015-05-26 21:27:49]

On 2015/05/26 21:26:34, Lei Zhang wrote:
> Do you have an unstripped chrome binary build with --ifc=all handy? I'm
> wondering if Breakpad's dump_syms tool will handle it.

I don't. But it should be roughly the same as --icf=safe on arm or i386 (except
for functions which get their address taken, which now get merged too), and it
works there too, right?

[Lei Zhang, 2015-05-26 21:40:25]

On 2015/05/26 21:27:49, Nico wrote:
> On 2015/05/26 21:26:34, Lei Zhang wrote:
> > Do you have an unstripped chrome binary build with --ifc=all handy? I'm
> > wondering if Breakpad's dump_syms tool will handle it.
> 
> I don't. But it should be roughly the same as --icf=safe on arm or i386
(except
> for functions which get their address taken, which now get merged too), and it
> works there too, right?

It should be roughly the same. Well, let's see what the official bots say. lgtm.

[Nico, 2015-05-26 23:12:57]

The CQ bit was checked by thakis@chromium.org

[commit-bot: I haz the power, 2015-05-26 23:13:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1153143004/20001

[commit-bot: I haz the power, 2015-05-26 23:19:57]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-05-26 23:20:45]

Patchset 2 (id:??) landed as
https://crrev.com/45ab39af3cc293e150158bf06515a47319db9f79
Cr-Commit-Position: refs/heads/master@{#331474}

[Nico, 2015-05-28 02:24:21]

A revert of this CL (patchset #2 id:20001) has been created in
https://codereview.chromium.org/1153153003/ by thakis@chromium.org.

The reason for reverting is: Speculative, to see if it's related to
http://crbug.com/492809.