Move more common settings to webcore_prerequisites

Move more common settings to webcore_prerequisites

dependent on http://codereview.chromium.org/3278002

is https://bugs.webkit.org/show_bug.cgi?id=44812

[M-A Ruel, 2010-08-28 02:39:56]

Moving more configurations to webcore_prerequisites. Now I'll be able to go into
more fun stuff.

Try jobs:
http://build.chromium.org/buildbot/try-server/builders/win/builds/47217
http://build.chromium.org/buildbot/try-server/builders/win/builds/47216

http://build.chromium.org/buildbot/try-server/builders/mac/builds/44409
http://build.chromium.org/buildbot/try-server/builders/mac/builds/44408

http://build.chromium.org/buildbot/try-server/builders/linux/builds/44525
http://build.chromium.org/buildbot/try-server/builders/linux/builds/44524

[Mark Mentovai, 2010-08-28 03:24:02]

LGTM

[Victor Wang, 2010-08-28 03:38:02]

LGTM

[M-A Ruel, 2010-08-28 03:47:42]

In fact this doesn't work. For an unknown reason the dependency tree is wrong.

In webcore_remaining, in condition OS==mac, dependencies:
webkit_system_interface is there. The problem is that the dependency is not
creating a build link. So webkit_system_interface is built after
webcore_remaining. See the compile failure on mac as evidence. Does any of your
has an idea on how to fix this?

[M-A Ruel, 2010-08-28 04:19:41]

Fixed the issue by putting hard_dependency:1 in webkit_system_interface. I'm not
sure about it but it works. The documentation about hard_dependency isn't clear
either.

Sent new try jobs to confirm:
http://build.chromium.org/buildbot/try-server/builders/mac/builds/44415
http://build.chromium.org/buildbot/try-server/builders/mac/builds/44416

[Victor Wang, 2010-08-28 04:23:14]

LGTM

On 2010/08/28 04:19:41, Marc-Antoine Ruel wrote:
> Fixed the issue by putting hard_dependency:1 in webkit_system_interface. I'm
not
> sure about it but it works. The documentation about hard_dependency isn't
clear
> either.
> 
> Sent new try jobs to confirm:
> http://build.chromium.org/buildbot/try-server/builders/mac/builds/44415
> http://build.chromium.org/buildbot/try-server/builders/mac/builds/44416

[M-A Ruel, 2010-08-29 14:57:07]

Actually the last change didn't work either.

Mark, it seems that defining includes_dir in direct_dependent_settings works for
C++ but not for Objective-C. I can try to write a regression test for that if
you want but you can just try the patchset were I don't duplicate includes_dir,
you'll see it .mm file fails to compile in webcore_remaining.