Linking to WebKit2 should be enough, like Network and Plugin processes.
Created attachment 236430 [details]
Gyuyoung, could you please confirm this patch doesn't break the EFL build with shared WebCore?
I don't remember why it was required long time ago.
But I can confirm that EFL port looks fine now with this patch whatever SHARED_CORE is ON/OFF.
(In reply to comment #1)
> Created an attachment (id=236430) [details]
> Gyuyoung, could you please confirm this patch doesn't break the EFL build with shared WebCore?
When I test this patch locally, there is no build break. Besides I let efl-ews to build with SHARED_CORE=ON yesterday.
Additionally I'm taking a look how to run layout test on EFL port regardless of SHARED_CORE.
(In reply to comment #2)
> I don't remember why it was required long time ago.
IIRC, initial EFL WK1 DumpRenderTree was implemented only with SHARED_CORE=ON because SAHRED_CORE reduces linking time on debug build. It looks nobody checked EFL layout test failed on release build at that time. IMO, that's why EFL port should enable SHARED_CORE to run layout test. However, it looks EFL WK2 doesn't need to enable it if we set something environment variables. Let me post when I solve this issue.
Committed r172516: <http://trac.webkit.org/changeset/172516>