MIDDLE_ENDIAN and BIG_ENDIAN are CPU attributes and not platform. They should be used with CPU() macro, and not with PLATFORM().
Created attachment 223320 [details] patch
Comment on attachment 223320 [details] patch r=me Is there any way we can prevent this? From briefly searching, GCC has option -Wundef to warn about undefined identifiers. Would enabling it lead to many false positives?
Comment on attachment 223320 [details] patch Clearing flags on attachment: 223320 Committed r170081: <http://trac.webkit.org/changeset/170081>
All reviewed patches have been landed. Closing bug.
(In reply to comment #2) > (From update of attachment 223320 [details]) > r=me > > Is there any way we can prevent this? From briefly searching, GCC has option -Wundef to warn about undefined identifiers. Would enabling it lead to many false positives? -Wundef is used by EFL port, but it doesn't help similar issues, because PLATFORM macro has the following definition: #define PLATFORM(WTF_FEATURE) (defined WTF_PLATFORM_##WTF_FEATURE && WTF_PLATFORM_##WTF_FEATURE)