Make combined native/manual JNI registration work for clang.

Make combined native/manual JNI registration work for clang.

Since clang has trouble with aliases for static functions, go back to
using a trivial wrapper function instead of an alias. Since there is
only one caller of the wrapped, static function, and its address is no
longer taken to generate the manual JNI registration table (the
wrapper's address is already being taken instead), there's no actual
benefit to using an alias anyway, since the compiler will simply inline
the static function into the wrapper.

BUG=442327, 460857
TBR=cjhopman@chromium.org

Committed: https://crrev.com/cab875835ac9e39183ba0ac4fe30e96d1bdfec45
Cr-Commit-Position: refs/heads/master@{#317787}

[Torne, 2015-02-23 18:51:35]

torne@chromium.org changed reviewers:
+ rmcilroy@chromium.org

[Torne, 2015-02-23 19:24:34]

This turns out to be super trivial, I have no idea why I didn't just do this in
the first place :)

[rmcilroy, 2015-02-23 22:46:40]

lgtm with a couple of nits. Thanks!

https://codereview.chromium.org/953523003/diff/1/build/jar_file_jni_generator...
File build/jar_file_jni_generator.gypi (right):

https://codereview.chromium.org/953523003/diff/1/build/jar_file_jni_generator...
build/jar_file_jni_generator.gypi:70: 'native_exports%': '--native_exports',
can we get rid of this block?

https://codereview.chromium.org/953523003/diff/1/build/jni_generator.gypi
File build/jni_generator.gypi (right):

https://codereview.chromium.org/953523003/diff/1/build/jni_generator.gypi#new...
build/jni_generator.gypi:89: 'native_exports%': '--native_exports',
ditto

[pasko, 2015-02-24 10:35:20]

fly by lgtm

that extra aliasing seems to have broken our -section-ordering-file logic (at
least it is in the range of commits). I'm not sure why, seems like the
instrumentation records alias calls properly, but -section-ordering-file does
not bother ordering them properly (in the same section as the function being
aliased?)

See:
http://crbug.com/460906
https://codereview.chromium.org/947073003

[Torne, 2015-02-24 11:09:07]

On 2015/02/24 10:35:20, pasko wrote:
> fly by lgtm
> 
> that extra aliasing seems to have broken our -section-ordering-file logic (at
> least it is in the range of commits). I'm not sure why, seems like the
> instrumentation records alias calls properly, but -section-ordering-file does
> not bother ordering them properly (in the same section as the function being
> aliased?)
> 
> See:
> http://crbug.com/460906
> https://codereview.chromium.org/947073003

Yeah, sorry for the various exciting problems all this alias magic caused, it
really wasn't necessary (as seen in this CL) and I'm not sure why I didn't think
of the simpler way first :)

[Torne, 2015-02-24 11:09:11]

The CQ bit was checked by torne@chromium.org

[commit-bot: I haz the power, 2015-02-24 11:09:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/953523003/1

[Torne, 2015-02-24 11:16:55]

On 2015/02/23 22:46:40, rmcilroy wrote:
> lgtm with a couple of nits. Thanks!
> 
>
https://codereview.chromium.org/953523003/diff/1/build/jar_file_jni_generator...
> File build/jar_file_jni_generator.gypi (right):
> 
>
https://codereview.chromium.org/953523003/diff/1/build/jar_file_jni_generator...
> build/jar_file_jni_generator.gypi:70: 'native_exports%': '--native_exports',
> can we get rid of this block?
> 
> https://codereview.chromium.org/953523003/diff/1/build/jni_generator.gypi
> File build/jni_generator.gypi (right):
> 
>
https://codereview.chromium.org/953523003/diff/1/build/jni_generator.gypi#new...
> build/jni_generator.gypi:89: 'native_exports%': '--native_exports',
> ditto

I don't want to change the behaviour of the downstream webview build, especially
when we're so close to killing it, so I am going to leave these in for now until
the android_webview_build stuff is deleted entirely.

[Torne, 2015-02-24 11:18:34]

torne@chromium.org changed reviewers:
+ cjhopman@chromium.org

[Torne, 2015-02-24 11:18:35]

TBRing cjhopman@ for build config stuff as I forgot to ask him to review and
this fixes asan bots downstream. The change to the build config is trivial (just
removing clang=0 conditonal).

[Torne, 2015-02-24 11:18:55]

The CQ bit was unchecked by torne@chromium.org

[Torne, 2015-02-24 11:19:10]

The CQ bit was checked by torne@chromium.org

[commit-bot: I haz the power, 2015-02-24 11:19:16]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/953523003/1

[commit-bot: I haz the power, 2015-02-24 12:55:11]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-02-24 12:55:44]

Patchset 1 (id:??) landed as
https://crrev.com/cab875835ac9e39183ba0ac4fe30e96d1bdfec45
Cr-Commit-Position: refs/heads/master@{#317787}

[cjhopman, 2015-02-24 21:03:39]

lgtm