libgpg-error strerror weirdness [was: Re: libgpg-error: Re: [PATCH] fix logging in t-version.c]

Daniel Kahn Gillmor dkg at fifthhorseman.net
Wed Jun 25 00:39:49 CEST 2014


On Tue 2014-06-24 17:49:57 -0400, Daniel Kahn Gillmor <dkg at fifthhorseman.net> wrote:
> sorry, the preceding patch to fix the logging in t-version.c was for
> libgpg-error, not for any other GnuPG-related development!

When trying to build libgpg-error 1.13 on debian testing/unstable
systems with gcc 4.9, i also notice this weird artifact:

gcc -DHAVE_CONFIG_H -I. -I..   -D_FORTIFY_SOURCE=2  -g -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security -c gen-posix-lock-obj.c
gen-posix-lock-obj.c: In function ‘main’:
gen-posix-lock-obj.c:106:7: warning: format ‘%s’ expects argument of type ‘char *’, but argument 3 has type ‘int’ [-Wformat=]
       fprintf (stderr, PGM ": error writing to stdout: %s\n", strerror (errno));
       ^

i'm used to strerror returning a char*, not an int -- maybe there is
some sort of weird macro wrapper transforming strerror into strerror_r
or something?

Anyway, this code doesn't ship in any user-facing binaries, afaict --
it's just a C code generator itself, so i don't think this is a big
deal, i just wanted to point it out in case anyone knows more about
what's going on here.

regards,

   --dkg
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 948 bytes
Desc: not available
URL: </pipermail/attachments/20140624/d14c2752/attachment.sig>


More information about the Gnupg-devel mailing list