On Tue, Sep 23, 2003, Dennis McRitchie wrote:
> [...]
> Looking earlier in the bootstrap output, I see that, while it appears that
> rpm was built without any errors, whenever it is executed (as below) it
> seems to get killed. This includes when it is being used to create the
> source and binary RPMs. So I suspect this is closer to the root of the
> problem. But I have no idea why it is getting killed. This did not show up
> in either of the other two successful build outputs.
>
> vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv
> ...
> + export HOME
> + echo ++ initializing RPM database
> ++ initializing RPM database
> + /tmp/rpm --initdb
> Killed
> + echo ++ re-iterating over RPM specification procedures
> ++ re-iterating over RPM specification procedures
> OPENPKG_BOOT=1
> + export OPENPKG_BOOT
> + /tmp/rpm -bb openpkg.spec
> Killed
> ...
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> Has anyone seen this before or have any idea what might be causing this?
> [...]
Looks like the "Killed" there actually is some sort "segmentation fault"
of "bus error" signal which is delivered. Hmm... the question is why
this happens on this box although the stuff works fine on other Solaris
9 boxes (included our sun4u boxes here) and what you can do to fix
or workaround it? Is this box on the latest Solaris 9 patch level?
What compiler have you used to build the bootstrap? If it is a Sun
compiler, can you retry with GCC (by installing OpenPKG 1.3 binary
packages temporarily and adding it to $PATH) or vice versa?
Ralf S. Engelschall
[EMAIL PROTECTED]
www.engelschall.com
______________________________________________________________________
The OpenPKG Project www.openpkg.org
User Communication List [EMAIL PROTECTED]
