[Android] Add remaining build/common.gypi changes from downstream.

[Android] Add remaining build/common.gypi changes from downstream and strip NOTIMPLEMENTED()s from official builds.

Includes:
1) Allowing overridding of android_ndk_sysroot (addition of "%" in
variable name).
2) Disable NOTIMPLEMENTED output for official builds (see base/logging.h)
3) Add clang build configurations
4) Remove unnecessary cflags from regular build.
5) Include crtbegin_so.o and crtend_so.so for _type=="loadable_module"
which is needed for ppapi_tests.

BUG=137569
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=147986

[Yaron, 2012-07-19 23:24:51]

The remaining diff in this file is related to order profiling which is in
http://codereview.chromium.org/10697079/

[John Grabowski, 2012-07-19 23:33:00]

LGTM, but please wait for Mento to review line 973 (NOTIMPLEMENTED_POLICY)

http://codereview.chromium.org/10808042/diff/1/build/common.gypi
File build/common.gypi (right):

http://codereview.chromium.org/10808042/diff/1/build/common.gypi#newcode973
build/common.gypi:973: 'NOTIMPLEMENTED_POLICY=0'
Not sure what to think of this.  I'll add an extra reviewer to confirm this is a
good idea.

http://codereview.chromium.org/10808042/diff/1/build/common.gypi#newcode2234
build/common.gypi:2234: # TODO(beverloo): WebKit warns about using the "nullptr"
bug#?

[Yaron, 2012-07-20 00:16:40]

http://codereview.chromium.org/10808042/diff/1/build/common.gypi
File build/common.gypi (right):

http://codereview.chromium.org/10808042/diff/1/build/common.gypi#newcode973
build/common.gypi:973: 'NOTIMPLEMENTED_POLICY=0'
On 2012/07/19 23:33:00, John Grabowski wrote:
> Not sure what to think of this.  I'll add an extra reviewer to confirm this is
a
> good idea.

It was originally added as part of b/5364886 with the goal to reduce logging for
released builds. I'm not sure it's really valuable to include this.. isn't it
more useful for dev builds? I'm fine with removing it if it's preferred.

http://codereview.chromium.org/10808042/diff/1/build/common.gypi#newcode2234
build/common.gypi:2234: # TODO(beverloo): WebKit warns about using the "nullptr"
On 2012/07/19 23:33:00, John Grabowski wrote:
> bug#?

filed crbug.com/138166

[John Grabowski, 2012-07-20 00:17:48]

http://codereview.chromium.org/10808042/diff/1/build/common.gypi
File build/common.gypi (right):

http://codereview.chromium.org/10808042/diff/1/build/common.gypi#newcode973
build/common.gypi:973: 'NOTIMPLEMENTED_POLICY=0'
On 2012/07/20 00:16:40, Yaron wrote:
> On 2012/07/19 23:33:00, John Grabowski wrote:
> > Not sure what to think of this.  I'll add an extra reviewer to confirm this
is
> a
> > good idea.
> 
> It was originally added as part of b/5364886 with the goal to reduce logging
for
> released builds. I'm not sure it's really valuable to include this.. isn't it
> more useful for dev builds? I'm fine with removing it if it's preferred.

I'd guess it's fine but there may be specific policies about this.  mento would
know.

[Mark Mentovai, 2012-07-20 02:11:46]

http://codereview.chromium.org/10808042/diff/8001/build/common.gypi
File build/common.gypi (right):

http://codereview.chromium.org/10808042/diff/8001/build/common.gypi#newcode973
build/common.gypi:973: 'NOTIMPLEMENTED_POLICY=0'
Can you make this change in logging.h or wherever NOTIMPLEMENTED happens? This
isn’t really the sort of thing that needs to be defined globally like this.

[Yaron, 2012-07-20 18:09:06]

moved NOTIMPLEMENTED_POLICY to base/logging.h

[Yaron, 2012-07-23 20:55:26]

mark: ptal at base/logging.h

[Mark Mentovai, 2012-07-23 21:07:11]

LGTM but please revise the checkin comment to say something about logging.h or
break the logging.h modification into a separate change.

[Mark Mentovai, 2012-07-23 21:08:49]

Oh, you did update it, but the logging.h thing is buried, while you gave
common.gypi top billing.

“now in base/logging.h” is meaningless in the context of commit history, because
nobody’s going to look at this code review to recognize that “now in” means that
it was one in common.gypi but we’ve chosen to take a different approach. The CL
message should be clear and stand on its own (maybe in the context of other
commit messages), and not make ambiguous references to the review history
without actually saying what it means.

[Yaron, 2012-07-23 21:36:23]

Thanks for review. Added mention of NOTIMPLEMENTED change to headline of review

[Mark Mentovai, 2012-07-23 21:37:38]

LGTM

[commit-bot: I haz the power, 2012-07-23 21:38:26]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/yfriedman@chromium.org/10808042/5003

[commit-bot: I haz the power, 2012-07-23 23:22:31]

Change committed as 147986

[Ami GONE FROM CHROMIUM, 2012-07-24 18:06:28]

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi
File build/common.gypi (left):

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi#o...
build/common.gypi:2193: '-mfloat-abi=<(arm_float_abi)',
Why did you remove this?

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi
File build/common.gypi (right):

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi#n...
build/common.gypi:2199: 'cflags': [ '-mfloat-abi=soft', ],
This is broken for cros/arm builds (which don't use clang AFAIK, but still).

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi#n...
build/common.gypi:2211: 'cflags': [ '-mfloat-abi=softfp', ],
This is just busted.  It hardcodes softfp for all builds, which breaks ffmpeg on
cros/arm, for example.

[Yaron, 2012-07-24 18:16:45]

Reverted this and http://codereview.chromium.org/10810071/
I'll roll again with these issues sorted out

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi
File build/common.gypi (left):

https://chromiumcodereview.appspot.com/10808042/diff/5003/build/common.gypi#o...
build/common.gypi:2193: '-mfloat-abi=<(arm_float_abi)',
On 2012/07/24 18:06:28, Ami Fischman wrote:
> Why did you remove this? 

I was merging in changes from downstream. The condition was pushed into the
clang sections below but the changes we had were specific to Android.
Unfortunately I didn't give this enough scrutiny pushing it up. Sorry for the
breakage.