Split LoadAndRunNativeApplication() ...

Split LoadAndRunNativeApplication() ...

(... to separate LoadNativeApplication() and RunNativeApplication()).

* This is to better support sandboxing in the near future. (Note: All
  the thunks are set in RunNativeApplication(). This should be fine with
  respect to sandboxing. More questionable is the calling of
  InitGoRuntime() (though this is supposedly temporary). We'll worry
  about that if it ever becomes a problem.)
* Also rename the files from dynamic_service_runner.* to
  native_application_support.*.
* Also make a separate native_application_support target. (The inclusion
  of dynamic_service_runner.* in the in_process_native_runner target was
  never really right anyway.
* Remove the (trivial -- just for an enum declaration) dependency of
  native_application_support.* on application_manager. This is so the
  future mojo_shell_child binary won't need to depend on half the world,
  and thus link in dynamic libraries it may not need.

R=jamesr@chromium.org

Committed: https://chromium.googlesource.com/external/mojo/+/e126c86bc85bc9a2b8a0951a1573c45a03e12c18

[jamesr, 2015-03-30 22:15:22]

Looks like android is Teh Sad

https://codereview.chromium.org/1046013002/diff/1/shell/native_application_su...
File shell/native_application_support.h (right):

https://codereview.chromium.org/1046013002/diff/1/shell/native_application_su...
shell/native_application_support.h:31: bool should_delete_app_path);
could you leave this an enum? a callsite passing 'true' is hard to interpret

[viettrungluu, 2015-03-30 22:53:08]

Thanks, PTAL.

On 2015/03/30 22:15:22, jamesr wrote:
> Looks like android is Teh Sad

Fixed, maybe.

> 
>
https://codereview.chromium.org/1046013002/diff/1/shell/native_application_su...
> File shell/native_application_support.h (right):
> 
>
https://codereview.chromium.org/1046013002/diff/1/shell/native_application_su...
> shell/native_application_support.h:31: bool should_delete_app_path);
> could you leave this an enum? a callsite passing 'true' is hard to interpret

Done.

[jamesr, 2015-03-30 23:53:54]

lgtm

[viettrungluu, 2015-03-31 02:20:50]

Committed patchset #5 (id:80001) manually as
e126c86bc85bc9a2b8a0951a1573c45a03e12c18 (presubmit successful).