http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54750
Bug #: 54750
Summary: conflicting float.h on windows using mingw
Classification: Unclassified
Product: gcc
Version: 4.7.3
Status: UNCONFIRMED
Severity: major
Priority: P3
Component: c
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: p.o...@wanadoo.fr
Since 2010 if I remember correctly (and there was previous discussion about
this on the GCC patch maling-list) there is a conflict with float.h from GCC
and the one from mingw.
Two patches were proposed long time ago to fix this issue. One on GCC tree and
one on MingW tree. At this time the mingw64 project has patched the float.h to
cooperate with GCC's one.
See r2987 from mingw64 subversion tree.
http://mingw-w64.svn.sourceforge.net/viewvc/mingw-w64/trunk/mingw-w64-headers/crt/float.h?r1=2986&r2=2987
So starting with GCC 4.6 it is possible to fix the build issue by just
applying the following simple patch. This is what we do since GCC 4.5
to build the compiler on Windows.
diff --git a/gcc/ginclude/float.h b/gcc/ginclude/float.h
index 2954cc4..d18c390 100644
--- a/gcc/ginclude/float.h
+++ b/gcc/ginclude/float.h
@@ -276,3 +276,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively.
If
#endif /* __STDC_WANT_DEC_FP__ */
#endif /* _FLOAT_H___ */
+
+#ifdef __MINGW32__
+#include_next <float.h>
+#endif
Would it be possible to apply this patch? Or is there still something
missing/wrong for this to happen?
Pascal.
