Build and package LLD unconditionally with the bundled Clang

Build and package LLD unconditionally with the bundled Clang

LLD is an experimental linker that links chrome faster.
However, it's not in the chromium-bundled toolchain on non-windows
platform, and hard to try out.
This CL adds LLD to the bundled toolchain package, so that developers
can use it just by setting `use_lld = true` in args.gn.

[tzik, 2016-12-12 06:40:31]

Description was changed from

==========
tol# Enter a description of the change.
Build and package LLD unconditionally with the bundled Clang
==========

to

==========
Build and package LLD unconditionally with the bundled Clang
==========

[tzik, 2016-12-12 06:40:38]

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

[commit-bot: I haz the power, 2016-12-12 06:41:03]

Dry run: 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, 2016-12-12 08:21:33]

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

[commit-bot: I haz the power, 2016-12-12 08:21:33]

Dry run: This issue passed the CQ dry run.

[tzik, 2016-12-12 08:55:28]

Description was changed from

==========
Build and package LLD unconditionally with the bundled Clang
==========

to

==========
Build and package LLD unconditionally with the bundled Clang

LLD is an experimental linker that links chrome faster.
However, it's not in the chromium-bundled toolchain on non-windows
platform, and hard to try out.
This CL adds LLD to the bundled toolchain package, so that developers
can use it just by setting `use_lld = true` in args.gn.
==========

[tzik, 2016-12-12 08:55:28]

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

[tzik, 2016-12-12 08:55:39]

PTAL

[Reid Kleckner, 2016-12-12 16:29:54]

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

[Reid Kleckner, 2016-12-12 16:29:54]

This will grow the clang package by the same amount as libLLVMgold.so would have
because LLVM is statically linked into LLD for LTO purposes. We split that out
into a separate package over size concerns. Maybe we don't care anymore.

[Nico, 2016-12-12 18:25:17]

My bar is that everything in the clang bundle should be used by ~everyone. I
think we do want to ship lld and use that by default, but someone needs to own
switching the default (this includes measuring build times in various scenarios
with old and new linker, being on the hook for failing tests, etc). If you're
willing to work on that, then I'm generally fine with this change. Else, I'd
want to wait with this until someone else (including me, eventually) wants to do
that.

(ps: this patch adds lld on mac too I think; on mac we very likely don't want to
use lld for quite a while, if ever, depending on what apple does with ld64.)

[Nico, 2016-12-13 01:17:31]

(see also https://crbug.com/607968)

On Mon, Dec 12, 2016 at 1:25 PM, <thakis@chromium.org> wrote:

> My bar is that everything in the clang bundle should be used by ~everyone.
> I
> think we do want to ship lld and use that by default, but someone needs to
> own
> switching the default (this includes measuring build times in various
> scenarios
> with old and new linker, being on the hook for failing tests, etc). If
> you're
> willing to work on that, then I'm generally fine with this change. Else,
> I'd
> want to wait with this until someone else (including me, eventually) wants
> to do
> that.
>
> (ps: this patch adds lld on mac too I think; on mac we very likely don't
> want to
> use lld for quite a while, if ever, depending on what apple does with
> ld64.)
>
> https://codereview.chromium.org/2560423004/
>

-- 
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.