linux: don't build gold as part of install-build-deps.sh

linux: don't build gold as part of install-build-deps.sh

Now that it's in the tree we don't need to force everyone to build
it.  (Especially since the code to build it didn't put it in the
path anyway.)

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=121005

[Evan Martin, 2012-02-03 19:53:09]

(won't land this until my change to use gold sticks)

If you think this code is useful enough to be saved, I'd be ok with making it a
separate script.  But my opinion is that anyone who wants to use their own gold
is better off just downloading it and setting it up themselves, much in the way
we support other tools like clang or valgrind.

[Nico, 2012-02-03 19:55:33]

I'll leave this one to Lei. Maybe want to sort out the 32bit story first.

> If you think this code is useful enough to be saved, I'd be ok with making it
a
> separate script.  But my opinion is that anyone who wants to use their own
gold
> is better off just downloading it and setting it up themselves, much in the
way
> we support other tools like clang or valgrind.

(for clang, people can run tools/clang/scripts/update.sh --force-local-build if
they want</offtopic>)

[Evan Martin, 2012-02-03 20:02:57]

On 2012/02/03 19:55:33, Nico wrote:
> I'll leave this one to Lei. Maybe want to sort out the 32bit story first.
> 
> > If you think this code is useful enough to be saved, I'd be ok with making
it
> a
> > separate script.  But my opinion is that anyone who wants to use their own
> gold
> > is better off just downloading it and setting it up themselves, much in the
> way
> > we support other tools like clang or valgrind.
> 
> (for clang, people can run tools/clang/scripts/update.sh --force-local-build
if
> they want</offtopic>)

Hm, that's interesting.  Perhaps I should move this code to third_party/gold to
make it easy for someone to update the binary found there?

[Evan Martin, 2012-02-03 20:05:47]

+tony too since he asked about it

[Lei Zhang, 2012-02-03 20:39:48]

On 2012/02/03 20:02:57, Evan Martin wrote:
> On 2012/02/03 19:55:33, Nico wrote:
> > I'll leave this one to Lei. Maybe want to sort out the 32bit story first.
> > 
> > > If you think this code is useful enough to be saved, I'd be ok with making
> it
> > a
> > > separate script.  But my opinion is that anyone who wants to use their own
> > gold
> > > is better off just downloading it and setting it up themselves, much in
the
> > way
> > > we support other tools like clang or valgrind.
> > 
> > (for clang, people can run tools/clang/scripts/update.sh --force-local-build
> if
> > they want</offtopic>)
> 
> Hm, that's interesting.  Perhaps I should move this code to third_party/gold
to
> make it easy for someone to update the binary found there?

I am in favor of this, so anyone with the same toolchain/platform can reproduce
the binary output we've checked in.

[Lei Zhang, 2012-02-03 22:29:04]

On 2012/02/03 19:55:33, Nico wrote:
> I'll leave this one to Lei. Maybe want to sort out the 32bit story first.

I personally don't care about the 32-bit story. It's 2012, DDR3 memory is
currently 5 USD / GB. Developers should really have more than 4 GB of RAM and be
running 64-bit build machines. </rant>

[Evan Martin, 2012-02-07 22:44:01]

PTAR(econsideration)

Now that I've updated the docs in third_party/gold, I think it's ok to delete
this.

[Lei Zhang, 2012-02-07 23:02:02]

On 2012/02/07 22:44:01, Evan Martin wrote:
> PTAR(econsideration)
> 
> Now that I've updated the docs in third_party/gold, I think it's ok to delete
> this.

LGTM

I'll add a few more works to third_party/gold/README.chromium

[Lei Zhang, 2012-02-07 23:02:13]

> I'll add a few more works to third_party/gold/README.chromium

s/works/words/

[Evan Martin, 2012-02-07 23:02:23]

make sure to "svn up" first

On Tue, Feb 7, 2012 at 3:02 PM,  <thestig@chromium.org> wrote:
> On 2012/02/07 22:44:01, Evan Martin wrote:
>>
>> PTAR(econsideration)
>
>
>> Now that I've updated the docs in third_party/gold, I think it's ok to
>> delete
>> this.
>
>
> LGTM
>
> I'll add a few more works to third_party/gold/README.chromium
>
> http://codereview.chromium.org/9317085/