24539 2009-03-11 19:32:17 -0700 Unresolved CSSMutableStyleDeclaration 2009-07-13 11:35:52 -0700 1 1 1 Unclassified WebKit WebCore Misc. 528+ (Nightly build) S60 Hardware S60 3rd edition RESOLVED FIXED P2 Normal --- 27065 1 norbert.leser webkit-unassigned darin hausmann laszlo.gombos mrowe oldest_to_newest 113299 0 norbert.leser 2009-03-11 19:32:17 -0700 Symbian linker cannot resolve CSSMutableStyleDeclaration class without explicit include of CSSMutableStyleDeclaration.h Proposed solution is to explicitly include the header file. See attached patch. 113300 1 28515 norbert.leser 2009-03-11 19:33:04 -0700 Created attachment 28515 Proposed fix for bug 24539 113306 2 mrowe 2009-03-11 19:39:03 -0700 Why is the forward-declaration of CSSMutableStyleDeclaration not sufficient? 115355 3 28515 eric 2009-03-26 11:05:39 -0700 Comment on attachment 28515 Proposed fix for bug 24539 This looks fine. Because CSSMutableStyleDeclaration is used in a RefPtr<CSSMutableStyleDeclaration> other ports must require this header too. They're likely only not failing due to RemoveCSSPropertyCommand.halways being included after CSSMutableStyleDeclaration in the same file. 115378 4 darin 2009-03-26 12:43:00 -0700 (In reply to comment #3) > This looks fine. Because CSSMutableStyleDeclaration is used in a > RefPtr<CSSMutableStyleDeclaration> other ports must require this header too. > They're likely only not failing due to RemoveCSSPropertyCommand.halways being > included after CSSMutableStyleDeclaration in the same file. Eric, I think that's wrong. Normally a forward declaration is sufficient even if the file does define a RefPtr, as long as the header doesn't contain any code that actually needs to manipulate the RefPtr. Are you sure this is a good change? 116954 5 28515 levin 2009-04-08 16:33:54 -0700 Comment on attachment 28515 Proposed fix for bug 24539 I'm clearing the r+ flag to remove this from the commit queue due to Drain's comment. It is waiting for a comment from Eric before it can be committed and he could add the r+ when he adds a comment. 117000 6 eric 2009-04-09 01:46:15 -0700 (In reply to comment #4) > Eric, I think that's wrong. Normally a forward declaration is sufficient even > if the file does define a RefPtr, as long as the header doesn't contain any > code that actually needs to manipulate the RefPtr. Are you sure this is a good > change? Again, this is one of those c++ questions where I'm gonna trust your (lengthy) c++ experience over my own... but in this case, I think the instantiation of PassRefPtr<CSSMutableStyleDeclaration> via: 6 static PassRefPtr<RemoveCSSPropertyCommand> create(Document* document, PassRefPtr<CSSMutableStyleDeclaration> style, CSSPropertyID property) 37 { 38 return adoptRef(new RemoveCSSPropertyCommand(document, style, property)); 39 } is going to require actually including CSSMutableStyleDeclaration.h so that PassRefPtr can see that CSSMutableStyleDeclaration actually has a ref() method, no? I'm still confused by why the Symbian build would fail here and no other port would, but I don't see including this header as harmful, if PassRefPtr<CSSMutableStyleDeclaration> instantiation does really require the full class definition. 117037 7 darin 2009-04-09 08:38:36 -0700 (In reply to comment #6) Yes, the create function does require the CSSMutableStyleDeclaration class definition. So either that function needs to be moved to the .cpp file and made non-inline, or the include needs to be added. 117533 8 norbert.leser 2009-04-14 16:01:49 -0700 (In reply to comment #7) > (In reply to comment #6) > > Yes, the create function does require the CSSMutableStyleDeclaration class > definition. So either that function needs to be moved to the .cpp file and made > non-inline, or the include needs to be added. > I will leave it up to you guys to decide. Either of the proposed solutions seems to work in our (symbian linker) case. 120744 9 darin 2009-05-11 16:54:50 -0700 http://trac.webkit.org/changeset/43523 122132 10 eric 2009-05-20 05:37:49 -0700 I missed a couple files when landing: Committing to http://svn.webkit.org/repository/webkit/trunk ... A LayoutTests/platform/mac/fast/forms/search-styled-expected.checksum A LayoutTests/platform/mac/fast/forms/search-styled-expected.png Committed r43904 122133 11 eric 2009-05-20 05:38:41 -0700 That comment was meant for https://bugs.webkit.org/show_bug.cgi?id=25742. 28515 2009-03-11 19:33:04 -0700 2009-05-11 16:54:32 -0700 Proposed fix for bug 24539 patch_fix11.txt text/plain 989 norbert.leser SW5kZXg6IFdlYkNvcmUvQ2hhbmdlTG9nCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFdlYkNvcmUvQ2hhbmdlTG9n CShyZXZpc2lvbiA0MTYwNCkKKysrIFdlYkNvcmUvQ2hhbmdlTG9nCSh3b3JraW5nIGNvcHkpCkBA IC0xLDMgKzEsMTQgQEAKKzIwMDktMDMtMTEgIE5va2lhIFVzZXIgIDxub3JiZXJ0Lmxlc2VyQG5v a2lhLmNvbT4KKworICAgICAgICBSZXZpZXdlZCBieSBOT0JPRFkgKE9PUFMhKS4KKworICAgICAg ICBXQVJOSU5HOiBOTyBURVNUIENBU0VTIEFEREVEIE9SIENIQU5HRUQKKworICAgICAgICBTeW1i aWFuIGxpbmtlciBjYW5ub3QgcmVzb2x2ZSBDU1NNdXRhYmxlU3R5bGVEZWNsYXJhdGlvbiBjbGFz cworICAgICAgICB3aXRob3V0IGV4cGxpY2l0IGluY2x1ZGUgb2YgQ1NTTXV0YWJsZVN0eWxlRGVj bGFyYXRpb24uaAorCisgICAgICAgICogZWRpdGluZy9SZW1vdmVDU1NQcm9wZXJ0eUNvbW1hbmQu aDoKKwogMjAwOS0wMy0xMSAgTWFyayBSb3dlICA8bXJvd2VAYXBwbGUuY29tPgogCiAgICAgICAg IFJ1YmJlci1zdGFtcGVkIGJ5IERhcmluIEFkbGVyLgpJbmRleDogV2ViQ29yZS9lZGl0aW5nL1Jl bW92ZUNTU1Byb3BlcnR5Q29tbWFuZC5oCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFdlYkNvcmUvZWRpdGluZy9S ZW1vdmVDU1NQcm9wZXJ0eUNvbW1hbmQuaAkocmV2aXNpb24gNDE2MDIpCisrKyBXZWJDb3JlL2Vk aXRpbmcvUmVtb3ZlQ1NTUHJvcGVydHlDb21tYW5kLmgJKHdvcmtpbmcgY29weSkKQEAgLTI4LDYg KzI4LDcgQEAKIAogI2luY2x1ZGUgIkVkaXRDb21tYW5kLmgiCiAjaW5jbHVkZSAiQ1NTUHJvcGVy dHlOYW1lcy5oIgorI2luY2x1ZGUgIkNTU011dGFibGVTdHlsZURlY2xhcmF0aW9uLmgiCiAKIG5h bWVzcGFjZSBXZWJDb3JlIHsKIAo=