Eric Blake wrote:
> > For me, with -Wall, plain 0 as a statement provokes a warning
> > "statement with no effect", but writing (void) 0 suppresses that
> > warning.
>
> Which implies the need for:
>
> # define SET_BINARY(fd) ((void) 0)
Yes, I agree. And when the result type is 'void', it should be 'void' in all
cases. A Google codesearch reveals that no one is using the SET_BINARY macro
for its value, so this is safe. Applied as follows:
2009-12-24 Bruno Haible <br...@clisp.org>
binary-io: Avoid gcc warning due to SET_BINARY.
* lib/binary-io.h (SET_BINARY): Cast the result to void.
Reported by Jim Meyering <j...@meyering.net>. Suggestion by Eric Blake.
--- lib/binary-io.h.orig Thu Dec 24 22:13:34 2009
+++ lib/binary-io.h Thu Dec 24 22:11:50 2009
@@ -36,6 +36,9 @@
# undef O_BINARY
# undef O_TEXT
#endif
+
+/* SET_BINARY (fd);
+ changes the file descriptor fd to perform binary I/O. */
#if O_BINARY
# if defined __EMX__ || defined __DJGPP__ || defined __CYGWIN__
# include <io.h> /* declares setmode() */
@@ -49,15 +52,15 @@
/* Avoid putting stdin/stdout in binary mode if it is connected to
the console, because that would make it impossible for the user
to interrupt the program through Ctrl-C or Ctrl-Break. */
-# define SET_BINARY(fd) (!isatty (fd) ? (setmode (fd, O_BINARY), 0) : 0)
+# define SET_BINARY(fd) ((void) (!isatty (fd) ? (setmode (fd, O_BINARY), 0) :
0))
# else
-# define SET_BINARY(fd) setmode (fd, O_BINARY)
+# define SET_BINARY(fd) ((void) setmode (fd, O_BINARY))
# endif
#else
/* On reasonable systems, binary I/O is the default. */
# undef O_BINARY
# define O_BINARY 0
-# define SET_BINARY(fd) /* nothing */
+# define SET_BINARY(fd) /* do nothing */ ((void) 0)
#endif
#endif /* _BINARY_H */
