Currently libWebCoreDerivedSources.a cannot see the symbol located in WebCore Source files. This patch let the library can see the symbols of WebCore Sources.
Committed r198830: <http://trac.webkit.org/changeset/198830>
(In reply to comment #1) > Committed r198830: <http://trac.webkit.org/changeset/198830> Hmm...I'm not sure if this is correct fix.
(In reply to comment #2) > (In reply to comment #1) > > Committed r198830: <http://trac.webkit.org/changeset/198830> > > Hmm...I'm not sure if this is correct fix. libWebCoreDerivedSources.a size is about 6.5 MB larger than before. - Before: 2223292 3월 30 16:44 libWebCoreDerivedSources.a - After: 8919352 3월 30 16:33 libWebCoreDerivedSources.a
(In reply to comment #3) > (In reply to comment #2) > > (In reply to comment #1) > > > Committed r198830: <http://trac.webkit.org/changeset/198830> > > > > Hmm...I'm not sure if this is correct fix. > > libWebCoreDerivedSources.a size is about 6.5 MB larger than before. > > - Before: 2223292 3월 30 16:44 libWebCoreDerivedSources.a > > - After: 8919352 3월 30 16:33 libWebCoreDerivedSources.a Maybe there will be a better solution. I think this change is just temporary thing to avoid compile failure. I will investigate this issue further.
(In reply to comment #4) > (In reply to comment #3) > > (In reply to comment #2) > > > (In reply to comment #1) > > > > Committed r198830: <http://trac.webkit.org/changeset/198830> > > > > > > Hmm...I'm not sure if this is correct fix. > > > > libWebCoreDerivedSources.a size is about 6.5 MB larger than before. > > > > - Before: 2223292 3월 30 16:44 libWebCoreDerivedSources.a > > > > - After: 8919352 3월 30 16:33 libWebCoreDerivedSources.a > > Maybe there will be a better solution. I think this change is just temporary > thing to avoid compile failure. > I will investigate this issue further. Re-open this bug until fixing this problem correctly.
Created attachment 275368 [details] Patch
Created attachment 275371 [details] Rebase patch
(In reply to comment #7) > Created attachment 275371 [details] > Rebase patch Alex, could you check if there is any issue on win and mac's cmake build when applying this change ?
Comment on attachment 275371 [details] Rebase patch On Mac I see this: ld: unknown option: --start-group
(In reply to comment #9) > Comment on attachment 275371 [details] > Rebase patch > > On Mac I see this: > ld: unknown option: --start-group When I ran Tools/Script/build-webkit --cmake I saw this log message. CMake Warning at Source/cmake/OptionsCommon.cmake:81 (message): GNU gold linker isn't available, using the default system linker. So it seems that mac port uses bsd linker currently. It is needed to guard here with if (USE_LD_GOLD), I think.
Created attachment 275448 [details] Patch
Created attachment 275449 [details] Fix indentation
Comment on attachment 275449 [details] Fix indentation View in context: https://bugs.webkit.org/attachment.cgi?id=275449&action=review > Source/WebKit2/CMakeLists.txt:835 > +if (USE_LD_GOLD) Hm, --start-group and --end-group are supported by ld.bfd; we should use them in that case, too. Maybe if (NOT APPLE) would be a better check here?
Comment on attachment 275449 [details] Fix indentation View in context: https://bugs.webkit.org/attachment.cgi?id=275449&action=review >> Source/WebKit2/CMakeLists.txt:835 >> +if (USE_LD_GOLD) > > Hm, --start-group and --end-group are supported by ld.bfd; we should use them in that case, too. Maybe if (NOT APPLE) would be a better check here? Ok, let me change this to if (NOT APPLE) here.
Created attachment 275520 [details] Patch
Comment on attachment 275520 [details] Patch Clearing flags on attachment: 275520 Committed r199011: <http://trac.webkit.org/changeset/199011>
All reviewed patches have been landed. Closing bug.