Milter-greylist

http://ftp.espci.fr/pub/milter-greylist/milter-greylist-3.0rc2.tgz
MD5 (milter-greylist-3.0rc2.tgz) = 445604a69f90647ad4438bbc9c1d362e 

Changes since rc1:
        Honour LINE_MAX for syslog (Attila Bruncsak)
        fix MX sync again on Tru64 (Attila Bruncsak)

-- 
Emmanuel Dreyfus
http://hcpnet.free.fr/pubz
manu@...

> 
> http://ftp.espci.fr/pub/milter-greylist/milter-greylist-3.0rc2.tgz
> MD5 (milter-greylist-3.0rc2.tgz) = 445604a69f90647ad4438bbc9c1d362e 
> 
> Changes since rc1:
>         Honour LINE_MAX for syslog (Attila Bruncsak)
>         fix MX sync again on Tru64 (Attila Bruncsak)
> 

Hello,

Emmanuel, why have you changed again the line 
   syslog(level, "%s", messagebuffer);
to
  syslog(level, messagebug);
?

The man syslog(3) says:

>  void syslog(
>          int priority,
>          const char *format,...);

So syslog(3) requires a format string.
The messagebug/messagebuffer already contains a formatted string as a result of the previous vsnprintf.

If there would be an e-mail address like rr%ss@... to be printed literally, the syslog would look for an additional non-existent argument incorrectly.

The "%s" is a must there.

On the other hand there is still a bit of confusion on the name of the buffer, so it does not even compile.

Bests,
Attila

<attila.bruncsak@...> wrote:

> Emmanuel, why have you changed again the line 
>    syslog(level, "%s", messagebuffer);
> to
>   syslog(level, messagebug);
> ?

Obviously I overlooked it. The real issue is that I cannot simply check
that it builds, since undefining HAVE_VSYSLOG will cause a redefinition
of vsyslog at mine. Let's hope the next shot will be okay

Here is the rc3:
http://ftp.espci.fr/pub/milter-greylist/milter-greylist-3.0rc3.tgz
MD5 (milter-greylist-3.0rc3.tgz) = 59881b4ae5eeca3d451013518b9a0ba1

-- 
Emmanuel Dreyfus
http://hcpnet.free.fr/pubz
manu@...

> > Emmanuel, why have you changed again the line 
> >    syslog(level, "%s", messagebuffer);
> > to
> >   syslog(level, messagebug);
> > ?
> 
> Obviously I overlooked it. The real issue is that I cannot 
> simply check
> that it builds, since undefining HAVE_VSYSLOG will cause a 
> redefinition
> of vsyslog at mine. Let's hope the next shot will be okay
> 
> Here is the rc3:
> http://ftp.espci.fr/pub/milter-greylist/milter-greylist-3.0rc3.tgz
> MD5 (milter-greylist-3.0rc3.tgz) = 59881b4ae5eeca3d451013518b9a0ba1
> 

Hello,

This compiles cleanly and runs fine for me.

Bests,
Attila