aura: Explicitly disable GTK

aura: Explicitly disable GTK.

Explicitly disable GTK by setting TOOLKIT_USES_GTK. This is the first in a series of changes required to have chrome on aura build and link without GTK. This CL makes base/ build without GTK. Subsequent CLs will do this for other components.

BUG=97131
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=102005

[sadrul, 2011-09-19 16:49:17]

If this looks acceptable, I will fix up the build failure in linux_touch

[oshima, 2011-09-19 17:08:07]

Please update the bug description to make it easy to understand what this CL is
about.

The direction looks good, so please go ahead.

http://codereview.chromium.org/7904034/diff/1/base/base.gyp
File base/base.gyp (right):

http://codereview.chromium.org/7904034/diff/1/base/base.gyp#newcode260
base/base.gyp:260: ['toolkit_uses_glib==1', {
"toolkit" uses glib may be misleading because it doesn't. how about just
"use_glib"?

http://codereview.chromium.org/7904034/diff/1/base/message_loop.cc
File base/message_loop.cc (right):

http://codereview.chromium.org/7904034/diff/1/base/message_loop.cc#newcode32
base/message_loop.cc:32: #if defined(OS_POSIX) && !defined(OS_MACOSX) &&
!defined(OS_ANDROID) && !defined(USE_AURA)
Can this simply be TOOLKIT_USES_GTK  now?

http://codereview.chromium.org/7904034/diff/1/build/build_config.h
File build/build_config.h (right):

http://codereview.chromium.org/7904034/diff/1/build/build_config.h#newcode48
build/build_config.h:48: #if defined(TOOLKIT_GTK) || (defined(TOOLKIT_VIEWS) &&
!defined(OS_WIN)) && !defined(USE_AURA)
(defined(TOOLKIT_VIEWS) && !defined(OS_WIN) && !defined(USE_URA))

is more correct

I thought emmanuel was going to work on this change, so please coordinate with
him (if you haven't already so)

[sadrul, 2011-09-19 18:44:48]

+saintlou, mostly for FYI

http://codereview.chromium.org/7904034/diff/1/base/base.gyp
File base/base.gyp (right):

http://codereview.chromium.org/7904034/diff/1/base/base.gyp#newcode260
base/base.gyp:260: ['toolkit_uses_glib==1', {
On 2011/09/19 17:08:08, oshima wrote:
> "toolkit" uses glib may be misleading because it doesn't. how about just
> "use_glib"?

Good point. Done.

http://codereview.chromium.org/7904034/diff/1/base/message_loop.cc
File base/message_loop.cc (right):

http://codereview.chromium.org/7904034/diff/1/base/message_loop.cc#newcode32
base/message_loop.cc:32: #if defined(OS_POSIX) && !defined(OS_MACOSX) &&
!defined(OS_ANDROID) && !defined(USE_AURA)
On 2011/09/19 17:08:08, oshima wrote:
> Can this simply be TOOLKIT_USES_GTK  now?

Done.

http://codereview.chromium.org/7904034/diff/5002/build/build_config.h
File build/build_config.h (right):

http://codereview.chromium.org/7904034/diff/5002/build/build_config.h#newcode48
build/build_config.h:48: #if defined(TOOLKIT_GTK) || (defined(TOOLKIT_VIEWS) &&
!defined(OS_WIN)) && !defined(USE_AURA)
We can leave this particular change out to allow chrome to build with use_aura=1
for now if so desired.

[Emmanuel Saint-loubert-Bié, 2011-09-19 19:23:14]

It is great. Thanks Sadrul.

[Nico, 2011-09-19 22:08:49]

You might want to cc a few of the linux folks. Since you also need base
approval, maybe just add evan as a reviewer.

[sadrul, 2011-09-19 22:16:04]

Thanks thakis!

Evan: could you please take a look at this? The idea is to split the dependency
into glib dependency and GTK+ dependency, so that the latter can be gradually
removed from components. A couple of follow up CLs:
http://codereview.chromium.org/7963001/ http://codereview.chromium.org/7944011/

I am currently working on a CL to build 'ui' with aura and without GTK. If you
have suggestions to approach this better/differently, please let me know.

[oshima, 2011-09-20 05:10:17]

http://codereview.chromium.org/7904034/diff/9003/base/base.gypi
File base/base.gypi (right):

http://codereview.chromium.org/7904034/diff/9003/base/base.gypi#newcode397
base/base.gypi:397: ],
nit: single line may be easer to compare with below.

http://codereview.chromium.org/7904034/diff/9003/base/message_pump_x.cc
File base/message_pump_x.cc (right):

http://codereview.chromium.org/7904034/diff/9003/base/message_pump_x.cc#newco...
base/message_pump_x.cc:104: GPollFD* x_poll = new GPollFD();
Don't you need the have to skip g_source_destroy etc in destructor?

[sadrul, 2011-09-20 18:24:27]

http://codereview.chromium.org/7904034/diff/9003/base/base.gypi
File base/base.gypi (right):

http://codereview.chromium.org/7904034/diff/9003/base/base.gypi#newcode397
base/base.gypi:397: ],
On 2011/09/20 05:10:17, oshima wrote:
> nit: single line may be easer to compare with below.

Done.

http://codereview.chromium.org/7904034/diff/9003/base/message_pump_x.cc
File base/message_pump_x.cc (right):

http://codereview.chromium.org/7904034/diff/9003/base/message_pump_x.cc#newco...
base/message_pump_x.cc:104: GPollFD* x_poll = new GPollFD();
On 2011/09/20 05:10:17, oshima wrote:
> Don't you need the have to skip g_source_destroy etc in destructor?

This was actually wrong. The event-source for X was not being initialized for
non-GTK use at all! Oops! Fixed that.

[oshima, 2011-09-20 18:30:59]

LGTM

[Evan Martin, 2011-09-20 19:00:55]

lgtm