Earnie Boyd wrote:
> This will become part of the updated msysDTK and is currently
> located in the Snapshot->MSYS release.
Just curious: did you run this through its test suite? How well
did it perform?
It might also be worth considering an upgrade of m4, to m4-1.4.5;
(bison is dependent on m4, and m4-1.4.5 is reported to include a
number of significant bug fixes, over the m4-1.4 we currently
provide). I recently built this as a native app. OOTB, it built
without a murmur, but then promptly failed every one of its test
cases. However, with the attached minor patch, it passes all but
three of the 104 tests in the suite.
I don't really have time to pursue the three failures, at present;
I think I should submit a bug report to bug-m4@....
Regards,
Keith.
--- src/m4.c 2006/07/17 15:23:02 1.1
+++ src/m4.c 2006/07/18 11:38:57
@@ -279,6 +279,11 @@
program_name = argv[0];
+ /* Prefer binary mode output for stdout/stderr
+ * on platforms which differentiate it from text. */
+ SET_BINARY (STDOUT_FILENO);
+ SET_BINARY (STDERR_FILENO);
+
include_init ();
debug_init ();
#ifdef USE_STACKOVF
--- lib/binary-io.h 2006/07/17 15:23:39 1.1
+++ lib/binary-io.h 2006/07/17 15:24:58
@@ -37,12 +37,12 @@
# undef O_TEXT
#endif
#if O_BINARY
-# if !(defined __EMX__ || defined __DJGPP__ || defined __CYGWIN__)
+# if !(defined __EMX__ || defined __DJGPP__ || defined __CYGWIN__ ||
defined __MINGW32__)
# define setmode _setmode
# undef fileno
# define fileno _fileno
# endif
-# if defined __DJGPP__ || defined __CYGWIN__
+# if defined __DJGPP__ || defined __CYGWIN__ || defined __MINGW32__
# include <io.h> /* declares setmode() */
# endif
# ifdef __DJGPP__
|