Reproduction steps: 1. Login to Gmail 2. Click on any message. Expected result: The message is open on a pane Actual result: Nothing happens. I don't see any error messages in the console either. It appears as if click events are never fired?
<rdar://problem/14564599>
It worked fine when I was landing http://trac.webkit.org/changeset/153381; after all, to get the crash to repro you had to click on messages.
I guess the regression range is somewhat misleading since WebKit nightly used to crash prior to http://trac.webkit.org/changeset/153381
<rdar://problem/14564630>
<rdar://problem/14564632>
(In reply to comment #2) > It worked fine when I was landing http://trac.webkit.org/changeset/153381; after all, to get the crash to repro you had to click on messages. Do you remember the revision you were able to open a message?
This is probably caused by http://trac.webkit.org/changeset/153380.
I've confirmed that the regression is caused by http://trac.webkit.org/changeset/153380 by manually bisecting the issue.
kling points out that the issue doesn't reproduce on this local machine. We're suspecting it's yet another instance of an old version of clang not being able to compile FINAL correctly.
It appears that this bug reproduces whenever we compile WebKit with clang 4 but not when we compile it with clang 5.
Created attachment 207569 [details] Disables final on clang 4.0
Created attachment 207570 [details] Fixed a typo
Comment on attachment 207570 [details] Fixed a typo View in context: https://bugs.webkit.org/attachment.cgi?id=207570&action=review > Source/WTF/wtf/Compiler.h:44 > +#define GCC_VERSION_AT_LEAST(major, minor) (__clang_major__ >= major && __clang_major__ >= minor) Ugh... clearly this needs to be CLANG_VERSION_AT_LEAST instead. > Source/WTF/wtf/Compiler.h:65 > +#define WTF_COMPILER_SUPPORTS_CXX_FINAL_CONTROL WTF_COMPILER_SUPPORTS_CXX_OVERRIDE_CONTROL && GCC_VERSION_AT_LEAST(4, 2) Ditto here.
Created attachment 207571 [details] Patch for landing
Comment on attachment 207571 [details] Patch for landing I'm gonna wait for EWS bots to process this patch.
Comment on attachment 207571 [details] Patch for landing View in context: https://bugs.webkit.org/attachment.cgi?id=207571&action=review > Source/WTF/wtf/Compiler.h:44 > +#define CLANG_VERSION_AT_LEAST(major, minor) (defined(__clang_major__) && __clang_major__ >= major && __clang_major__ >= minor) Typo at the end, __clang_major__ >= minor Should be __clang_minor__ >= minor
Created attachment 207574 [details] Make the fix Apple specific since __clang_major__ and __clang_minor__ are vendor dependent
Attachment 207574 [details] did not pass style-queue: Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WTF/ChangeLog', u'Source/WTF/wtf/Compiler.h']" exit_code: 1 Source/WTF/wtf/Compiler.h:63: One space before end of line comments [whitespace/comments] [5] Source/WTF/wtf/Compiler.h:63: Should have a space between // and comment [whitespace/comments] [4] Total errors found: 2 in 2 files If any of these errors are false positives, please file a bug against check-webkit-style.
Created attachment 207575 [details] Use ifdef instead of if defined
Attachment 207575 [details] did not pass style-queue: Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WTF/ChangeLog', u'Source/WTF/wtf/Compiler.h']" exit_code: 1 Source/WTF/wtf/Compiler.h:63: One space before end of line comments [whitespace/comments] [5] Source/WTF/wtf/Compiler.h:63: Should have a space between // and comment [whitespace/comments] [4] Total errors found: 2 in 2 files If any of these errors are false positives, please file a bug against check-webkit-style.
Comment on attachment 207575 [details] Use ifdef instead of if defined Looks like a reasonable solution. I'll land this patch right away to get nightlies back in shape on ML.
Comment on attachment 207575 [details] Use ifdef instead of if defined Clearing flags on attachment: 207575 Committed r153400: <http://trac.webkit.org/changeset/153400>
All reviewed patches have been landed. Closing bug.
(In reply to comment #21) > (From update of attachment 207575 [details]) > Looks like a reasonable solution. I'll land this patch right away to get nightlies back in shape on ML. Thanks!
*** Bug 119176 has been marked as a duplicate of this bug. ***
*** Bug 119193 has been marked as a duplicate of this bug. ***
*** Bug 119202 has been marked as a duplicate of this bug. ***