Re: MySQL at Linux/PPC
On Thu, Oct 14, 1999 at 10:27:35PM +0300, Michael Widenius wrote:
> >>>>> "Kostas" == Kostas Gewrgiou <firstname.lastname@example.org> writes:
> Kostas> After looking at Bug#: 5732 in RedHat's bugzilla,
> Kostas> (http://developer.redhat.com/bugzilla/show_bug.cgi?id=5732)
> Kostas> i found out why enabling LFS makes a bad mysql binary, quoting from the
> Kostas> description of the patch bug:
> Kostas> The configure scripts of patch-2.5 detect the LFS support in
> Kostas> glibc-2.1, but due to the lack of a definition of
> Kostas> _GNU_SOURCE, no prototypes for fseeko are declared. That way
> Kostas> the compiler thinks fseeko is taking a 32bit argument for
> Kostas> the offset, when in reality it takes a 64bit offset.
> Kostas> After compiling MySQL with -D_GNU_SOURCE (with LFS enabled) the problems
> Kostas> went away, i am not sure if defining _LARGEFILE_SOURCE is supposed to give
> Kostas> the fseeko declarations but if it is then its a bug in the glibc headers.
> Kostas> Kostas Gewrgiou.
> MySQL 3.22.27 and MySQL 3.23.4 and above we now detect the glibc
> version and will not use LFS if glibc < 2.1
> We have successfully compiled MySQL on RedHat 6.0 without having to
> patch anything, so it looks like this is now fixed.
> If you still get the problem on your machine with the above MySQL
> version, please mail me and we will look at it once more!
I haven't tried 3.22.7, but 3.23.4 builds and run just
fine with LinuxPPC R5/Q3 on my PowerBook and on a G4/400.
The LFS problem doesn't occur like it did with earlier versions.