On 2012-05-19 Phil Pennock <pdp@???> wrote:
> On 2012-05-19 at 16:26 +0200, Andreas Metzler wrote:
[...]
> If you're going to build with -Werror=format-security then you need to
> #define PRINTF_FUNCTION(A,B) to /**/ in mytypes.h, which will also shut
> up a bunch of other warnings. The PRINTF_FUNCTION() usage has caught a
> number of small issues and been generally useful, but it's not a perfect
> match. If there were a pragma to define a new format and register what
> each escape expects, as a type, and declare that some do not consume
> arguments, we could use that and there would be no mismatches.
[...]
> I think, realistically, people are going to turn on -Wformat=security
> and we need to accept that and remove the safety-checks instead.
> They're useful to the developers, in figuring out where there *might* be
> issues, but there's so many false positive warnings, and this, that it
> is not tenable for a release.
> For now, can you please build without -Wformat=security and see if it
> works?
FWIW I have just uploaded to Debian/experimental to check for
build-errors. In a first try we are building with -Wformat=security
and
--- exim4-4.80~rc2.orig/src/functions.h
+++ exim4-4.80~rc2/src/functions.h
@@ -325,7 +325,7 @@ extern uschar *string_copy_malloc(uschar
extern uschar *string_copylc(uschar *);
extern uschar *string_copynlc(uschar *, int);
extern uschar *string_dequote(uschar **);
-extern BOOL string_format(uschar *, int, const char *, ...) PRINTF_FUNCTION(3,4);
+extern BOOL string_format(uschar *, int, const char *, ...);
as this has worked for me. ;-)
cu andreas
--
`What a good friend you are to him, Dr. Maturin. His other friends are
so grateful to you.'
`I sew his ears on from time to time, sure'
