WTF shouldn't use a script build phase to detect the presence of headers when the compiler can do it for us

• Component: Web Template Framework
• Status: RESOLVED FIXED

[Mark Rowe (bdash)]

WTF has a build phase to detect the presence of certain headers at build time. This is unnecessary now that the Mac port always builds with Clang, which has the __has_include directive that can detect headers for us.

[Mark Rowe (bdash)]

Created attachment 186216 [details]
Patch v1

[WebKit Review Bot]

Attachment 186216 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/JavaScriptCore/ChangeLog', u'Source/JavaScriptCore/DerivedSources.make', u'Source/WTF/ChangeLog', u'Source/WTF/WTF.xcodeproj/project.pbxproj', u'Source/WTF/wtf/FastMalloc.cpp']" exit_code: 1
Source/WTF/wtf/FastMalloc.cpp:451:  Missing spaces around /  [whitespace/operators] [3]
Total errors found: 1 in 3 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Mark Rowe (bdash)]

Comment on attachment 186216 [details]
Patch v1

Clearing flags since this appears to not build with GCC.

[EFL EWS Bot]

Comment on attachment 186216 [details]
Patch v1

Attachment 186216 [details] did not pass efl-ews (efl):
Output: http://queues.webkit.org/results/16343541

[Mark Rowe (bdash)]

Created attachment 186220 [details]
Patch v2

Split the detection of __has_include out from the use of it so that compilers that don't support it don't choke.

[WebKit Review Bot]

Attachment 186220 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/JavaScriptCore/ChangeLog', u'Source/JavaScriptCore/DerivedSources.make', u'Source/WTF/ChangeLog', u'Source/WTF/WTF.xcodeproj/project.pbxproj', u'Source/WTF/wtf/FastMalloc.cpp']" exit_code: 1
Source/WTF/wtf/FastMalloc.cpp:452:  Missing spaces around /  [whitespace/operators] [3]
Total errors found: 1 in 3 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Alexey Proskuryakov]

Comment on attachment 186220 [details]
Patch v2

View in context: https://bugs.webkit.org/attachment.cgi?id=186220&action=review

> Source/WTF/wtf/FastMalloc.cpp:451
> +#ifdef __has_include

Would it make sense to have a WTF macro for this (like HAVE(HAS_INCLUDE))?

[WebKit Review Bot]

Comment on attachment 186220 [details]
Patch v2

Clearing flags on attachment: 186220

Committed r141700: <http://trac.webkit.org/changeset/141700>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Mark Rowe (bdash)]

(In reply to comment #7)
> (From update of attachment 186220 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=186220&action=review
> 
> > Source/WTF/wtf/FastMalloc.cpp:451
> > +#ifdef __has_include
> 
> Would it make sense to have a WTF macro for this (like HAVE(HAS_INCLUDE))?

I considered doing something like COMPILER_SUPPORTS(HEADER_DETECTION). In order to make it worth it you'd want to wrap the actual use of __has_include in a macro like HAS_INCLUDE  to support compilers with different names for the functionality. Sadly Clang currently seems to crash if you try to use __has_include in the body of a preprocessor macro.