On Wed, 2007-05-23 at 17:54 +0800, zamri wrote: > On 5/21/07, Aaron Stone <[EMAIL PROTECTED]> wrote: > > On Mon, 2007-05-21 at 16:58 +0200, Paul J Stevens wrote: > > > zamri wrote: > > > > On 5/16/07, zamri <[EMAIL PROTECTED]> wrote: > > > >> I got this "warnings" and similar in various parts of compilation: > > > >> > > > >> maintenance.c: In function 'do_purge_deleted': > > > >> maintenance.c:296: warning: format '%llu' expects type 'long long > > > >> unsigned int', but argument 2 has type 'guint64' > > > > > > Aaron, this smells like the solaris fix we did recently in mailbox_status. > > > > > > Can we safely use %llu to printf u64_t, or should we change them all to > > > %u? > > > > > > > I'm not sure. I manually checked the type sizes and they are OK on > > Linux / amd64, despite the warnings. We might need some deeper > > autoconfoolery to pick the right type. As guint64 is a Glib provided > > typedef, let's ask over there if people are seeing problems with > > actually using this type in printf statements. > > > > Aaron
> Any news about this? I won't upgrade yet since I am paranoid with all > the problems I read in this ML about 2.2.5. The warnings appear to be harmless. We will try to find a better type that gcc on x86-64 doesn't get angry about, but I have not seen any functionality problems on my x86-64 box. Aaron _______________________________________________ DBmail mailing list [email protected] https://mailman.fastxs.nl/mailman/listinfo/dbmail
