Roll clang 247874:253678

Roll clang 247874:253678

Also revert https://codereview.chromium.org/1375193005 for bootstrap builds,
else compiler-rt won't build in that config.  See comment 45 on the bug.

BUG=538614
R=thakis@chromium.org

Committed: https://chromium.googlesource.com/chromium/src/+/6a0a6e2c285cf215a15755ab9d32b386947a77b1

[hans, 2015-11-20 18:30:37]

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

[hans, 2015-11-20 18:30:38]

thestig: It seems linux_chromium_trusty32_rel doesn't exist anymore. What trybot
should we use to verify rolls for 32-bit Linux?

[hans, 2015-11-20 18:31:07]

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

[hans, 2015-11-20 18:31:07]

thestig -> cc
thakis -> review

[Nico, 2015-11-20 18:34:19]

lgtm

[Nico, 2015-11-20 21:44:43]

The CQ bit was checked by thakis@chromium.org

[commit-bot: I haz the power, 2015-11-20 21:45:33]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1461003007/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1461003007/1

[Lei Zhang, 2015-11-20 22:07:47]

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

[Lei Zhang, 2015-11-20 22:07:48]

+dpranke, I don't know what happened to the 32-bit try bots.

[commit-bot: I haz the power, 2015-11-20 23:06:16]

Patchset 1 (id:??) landed as
https://crrev.com/6a0a6e2c285cf215a15755ab9d32b386947a77b1
Cr-Commit-Position: refs/heads/master@{#360938}

[hans, 2015-11-20 23:06:46]

Committed patchset #1 (id:1) manually as
6a0a6e2c285cf215a15755ab9d32b386947a77b1 (presubmit successful).

[Dirk Pranke, 2015-11-20 23:12:48]

I'm not sure either, it looks like bot flakiness (builds timing out).

On Fri, Nov 20, 2015 at 2:07 PM, <thestig@chromium.org> wrote:

> +dpranke, I don't know what happened to the 32-bit try bots.
>
> https://codereview.chromium.org/1461003007/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[Dirk Pranke, 2015-11-20 23:37:08]

Oh, sorry, I misunderstood the question. We got rid of the trusty32 bot.

linux_chromium_compile_dbg_32_ng tests the linux32 build.

[hans, 2015-11-20 23:39:27]

On 2015/11/20 23:37:08, Dirk Pranke wrote:
> linux_chromium_compile_dbg_32_ng tests the linux32 build.

Do we have something that runs tests as well?

[Dirk Pranke, 2015-11-20 23:48:38]

On 2015/11/20 23:37:08, Dirk Pranke wrote:
> Oh, sorry, I misunderstood the question. We got rid of the trusty32 bot.
> 
> linux_chromium_compile_dbg_32_ng tests the linux32 build.

I'll take a stab at updating the list on the wiki page
(//docs/updating_clang.md).

[hans, 2015-11-21 00:00:40]

On 2015/11/20 23:48:38, Dirk Pranke wrote:
> On 2015/11/20 23:37:08, Dirk Pranke wrote:
> > Oh, sorry, I misunderstood the question. We got rid of the trusty32 bot.
> > 
> > linux_chromium_compile_dbg_32_ng tests the linux32 build.
> 
> I'll take a stab at updating the list on the wiki page
> (//docs/updating_clang.md).

Thanks!

[hans, 2015-11-21 06:06:17]

A revert of this CL (patchset #1 id:1) has been created in
https://codereview.chromium.org/1460503011/ by hans@chromium.org.

The reason for reverting is: MSan was broken by the roll..

[Michael Achenbach, 2015-11-21 08:28:12]

machenbach@chromium.org changed reviewers:
+ machenbach@chromium.org

[Michael Achenbach, 2015-11-21 08:28:13]

FYI: This also broke the cfi builder on the v8 side:
https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20-%20cfi/builds...

[Michael Achenbach, 2015-11-23 07:54:06]

On 2015/11/21 08:28:13, Michael Achenbach wrote:
> FYI: This also broke the cfi builder on the v8 side:
>
https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20-%20cfi/builds...

On top of a possible clang/gold plugin mismatch, v8 also seems to have an
infrastructure problem with the gold-plugin download script or the incremental
build. After the clang roll revert, our bot didn't stabilize by itself. But
after a nuke of the build directory it did...

[Nico, 2015-11-23 17:23:34]

Does v8 have something like
https://code.google.com/p/chromium/codesearch#chromium/src/build/common.gypi&...

      ['clang==1 or host_clang==1', {
        # This is here so that all files get recompiled after a clang roll and
        # when turning clang on or off.
        # (defines are passed via the command line, and build systems rebuild
        # things when their commandline changes). Nothing should ever read this
        # define.
        'defines': ['CR_CLANG_REVISION=<!(python
<(DEPTH)/tools/clang/scripts/update.py --print-revision)'],
      }],

somewhere? If not, it needs that.

[Michael Achenbach, 2015-11-23 18:49:56]

On 2015/11/23 17:23:34, Nico wrote:
> Does v8 have something like
>
https://code.google.com/p/chromium/codesearch#chromium/src/build/common.gypi&...
> 
>       ['clang==1 or host_clang==1', {
>         # This is here so that all files get recompiled after a clang roll and
>         # when turning clang on or off.
>         # (defines are passed via the command line, and build systems rebuild
>         # things when their commandline changes). Nothing should ever read
this
>         # define.
>         'defines': ['CR_CLANG_REVISION=<!(python
> <(DEPTH)/tools/clang/scripts/update.py --print-revision)'],
>       }],
> 
> somewhere? If not, it needs that.

We have that. It recompiled everything, but started complaining about missing
the gold plugin:
https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20-%20cfi/builds...

But before taking any other action, I need to try locally if it builds in this
version on a clean checkout. Maybe the gold plugin update script has a bug on
clang version changes.