Re: status update

I have one more bit of updated status on the impl. The problem stopping 
glibc from building during the intermediate phase for s390 and s390x has 
been identified. This is a known problem for both platforms where 
glibc-2.3.2 will not build successfully when using gcc-3.3 due to some 
header file changes. The workaround is a simple one line patch to a single 
file in the glibc source. After making this patch I was able to get the 
impl to complete the intermediate phase and start phase3. Unfortunately, I 
made a small mistake in the glibc.xml for phase3 and had to restart the 
impl. I'll find out later today if it dies at the same point as ia32 and 
ia64. The fix for s390 and s390x will be checked into cvs in a little 
while.

Regards,

Marvin Heffler
Linux Standard Base
IBM Linux Technology Center
11400 Burnet Road, Zip 905-7A017
Austin, TX 78758
(512) 838-0953    T/L 678-0953




"Wichmann, Mats D" <[email protected]>
Sent by: [email protected]
02/21/2004 09:49 AM

 
        To:     <[email protected]>
        cc: 
        Subject:        [lsb-impl] status update




One more update to the lsbsi story.
ia64 now builds all the way through
on one machine, then fails at the very end
(just like ia32).

I also kicked off a build on a
Red Hat EL3 ppc64 system, it failed
very early on, trying to build gcc.

The ia64 problem:  the static copy of bash
built in the bootstrap is not reliable,
so I inserted a build of a dynamic bash
as soon as possible in the intermediate -
right after glibc. This was enough to
make things work on one ia64, but still
too late on another (on the other, the
shell died trying to unpack the bash
tarball).  I think this is toolchain
dependent and thus hard to fix, but not
sure and out of time at the moment.
(This is clearly not the "right" fix
but let things move forward).

rpm 4.1 is in but fails at the very end
for platforms which get that far, in the
step where an empty database is made in
the lsbsi. rpm-final is built in the intermediate
chroot, which has a dynamic libbz2; it then
does not run in the final, which does not
have a dynamic libbz2.  Two ways to fix,
neither terribly complex: phase2 should install
only a static, not a dynamic libbz2; or phase3
should build and install a dynamic libbz2.
Since the LSB does not specify libbz2,
perhaps the former fix is the appropriate one?




_______________________________________________
lsb-impl mailing list
[email protected]
http://freestandards.org/mailman/listinfo/lsb-impl

_______________________________________________
lsb-impl mailing list
[email protected]
http://freestandards.org/mailman/listinfo/lsb-impl