Android WebView: make use of host multilib support.

Android WebView: make use of host multilib support.

Set the gyp variables that control the use of the Android build system's
host multilib support so that we can:
1) not break host multilib builds
2) correctly specify whether we want our host binaries to be build as 32
   or 64 bit, fixing Mac host builds of V8 for 64-bit targets.

BUG=358141
NOTRY=true
R=mkosiba@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=275110

[Torne, 2014-05-30 11:24:24]

This requires the gyp change https://codereview.chromium.org/301373002

[Ying Wang, 2014-05-30 21:42:29]

https://codereview.chromium.org/305163002/diff/1/Android.mk
File Android.mk (right):

https://codereview.chromium.org/305163002/diff/1/Android.mk#newcode60
Android.mk:60: ifeq ($(2ND_TARGET_IS_64_BIT),true)
This won't happen.

https://codereview.chromium.org/305163002/diff/1/Android.mk#newcode62
Android.mk:62: GYP_HOST_MULTILIB := 64
Even if that happens, shouldn't it be set to 32? because you are building for
the TARGET_2ND_ARCH, which is 32-bit.

[Ying Wang, 2014-05-30 21:49:57]

https://codereview.chromium.org/305163002/diff/1/Android.mk
File Android.mk (right):

https://codereview.chromium.org/305163002/diff/1/Android.mk#newcode62
Android.mk:62: GYP_HOST_MULTILIB := 64
On 2014/05/30 21:42:29, Ying Wang wrote:
> Even if that happens, shouldn't it be set to 32? because you are building for
> the TARGET_2ND_ARCH, which is 32-bit.

Sorry, I mistook the conditional as ifeq ($(2ND_HOST_IS_64_BIT),true)...
This makes sense to me now, though it's unlikely we'll have target 2nd arch as
64-bit, I think.

[Torne, 2014-06-02 10:15:49]

https://codereview.chromium.org/305163002/diff/1/Android.mk
File Android.mk (right):

https://codereview.chromium.org/305163002/diff/1/Android.mk#newcode62
Android.mk:62: GYP_HOST_MULTILIB := 64
On 2014/05/30 21:49:57, Ying Wang wrote:
> On 2014/05/30 21:42:29, Ying Wang wrote:
> > Even if that happens, shouldn't it be set to 32? because you are building
for
> > the TARGET_2ND_ARCH, which is 32-bit.
> 
> Sorry, I mistook the conditional as ifeq ($(2ND_HOST_IS_64_BIT),true)...
> This makes sense to me now, though it's unlikely we'll have target 2nd arch as
> 64-bit, I think.

Even if it's unlikely it doesn't hurt to have the conditional here just to make
it clear, I don't think.

[Ying Wang, 2014-06-02 17:02:53]

https://codereview.chromium.org/305163002/diff/1/Android.mk
File Android.mk (right):

https://codereview.chromium.org/305163002/diff/1/Android.mk#newcode62
Android.mk:62: GYP_HOST_MULTILIB := 64
On 2014/06/02 10:15:49, Torne wrote:
> On 2014/05/30 21:49:57, Ying Wang wrote:
> > On 2014/05/30 21:42:29, Ying Wang wrote:
> > > Even if that happens, shouldn't it be set to 32? because you are building
> for
> > > the TARGET_2ND_ARCH, which is 32-bit.
> > 
> > Sorry, I mistook the conditional as ifeq ($(2ND_HOST_IS_64_BIT),true)...
> > This makes sense to me now, though it's unlikely we'll have target 2nd arch
as
> > 64-bit, I think.
> 
> Even if it's unlikely it doesn't hurt to have the conditional here just to
make
> it clear, I don't think.
Sounds good.
Thanks for the change!

[Torne, 2014-06-05 09:56:53]

Should be safe to land this ahead of the gyp change as it just sets variables
which will not be used by the current makefiles.

[mkosiba (inactive), 2014-06-05 10:11:32]

lgtm

[Torne, 2014-06-05 10:57:23]

The CQ bit was checked by torne@chromium.org

[commit-bot: I haz the power, 2014-06-05 10:59:32]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/torne@chromium.org/305163002/1

[Torne, 2014-06-05 14:50:41]

Committed patchset #1 manually as r275110 (presubmit successful).