Use -fuse-ld=gold when supported to select gold.

Use -fuse-ld=gold when supported to select gold.

Allow usage of gold on gcc 4.8 systems such as debian.

BUG=352046
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=263189

[mithro-old, 2014-04-10 16:16:16]

Hi!

Davve ran into a problem where he was using
GYP_DEFINES="linux_use_gold_binary=0" to disable using third_party/binutils but
then gcc was trying to link using ld.bfd instead of gold.

The correct way to select gold on such a system is this -fuse-ld flag (which is
used already by a bunch of the android targets -- those targets should probably
start setting "linux_use_gold_flags=1" or something now?).

Tim

[Nico, 2014-04-10 21:27:19]

https://codereview.chromium.org/225393007/diff/1/build/common.gypi
File build/common.gypi (right):

https://codereview.chromium.org/225393007/diff/1/build/common.gypi#newcode3841
build/common.gypi:3841: # clang support --
http://clang-developers.42468.n3.nabble.com/Adding-fuse-ld-support-to-clang-t...
That link is just a thread where someone posted a patch for this option, but the
patch didn't go anywhere and isn't committed. Remove this line. (gcc_version
isn't set for clang, so this isn't used, which is fine.)

…but does third_party/gold (referenced a few lines below, I suppose it says
third_party/binutils at trunk) contain a ld.gold? Else gcc won't use the gold in
-B but the system one, rigth?

[mithro-old, 2014-04-11 02:17:56]

https://codereview.chromium.org/225393007/diff/1/build/common.gypi
File build/common.gypi (right):

https://codereview.chromium.org/225393007/diff/1/build/common.gypi#newcode3841
build/common.gypi:3841: # clang support --
http://clang-developers.42468.n3.nabble.com/Adding-fuse-ld-support-to-clang-t...
On 2014/04/10 21:27:19, Nico wrote:
> That link is just a thread where someone posted a patch for this option, but
the
> patch didn't go anywhere and isn't committed. Remove this line. (gcc_version
> isn't set for clang, so this isn't used, which is fine.)

Actually the patch did land and then was reverted. The guy said he would update
the thread when he relands the patch. Changed the comment to be clearer.
---------------
The code was committed and reverted, so if you want to have a go at fixing the
tests, you'd be welcome.  Otherwise, I'll hopefully get back to it in a few
weeks (although it's further down the list than tidying up a load of MIPS back
end fixes for upstreaming).
---------------

> 
> …but does third_party/gold (referenced a few lines below, I suppose it says
> third_party/binutils at trunk) contain a ld.gold? Else gcc won't use the gold
in
> -B but the system one, rigth?

Yes, it does say binutils at trunk. I have rebased this patch and uploaded.

third_party/binutils does contain an ld.gold

With this patch, you can use "linux_use_gold_flags=1" and
"linux_use_gold_binary=0" to use the system gold linker (when using gcc 4.8).

[Nico, 2014-04-11 02:29:21]

cool, lgtm

[mithro-old, 2014-04-11 02:32:48]

The CQ bit was checked by mithro@mithis.com

[commit-bot: I haz the power, 2014-04-11 02:33:37]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mithro@mithis.com/225393007/20001

[commit-bot: I haz the power, 2014-04-11 08:05:16]

Change committed as 263189