Kudzu segfault

[Tom Chiverton]

This is a Fedora Core 5 box, upgraded recently from 7.3 Legacy. Yum is
using only extras, updates and core - all non-testing - and reports it
is up to date.

Kudzu crashes during boot up and when run by root once the system is up
(the crash doesn't appear to have effected the boot process).

Moving /etc/sysconfig/hwconf out of the way doesn't appear to help.

Here is a snipit from an strace (strace -f kudzu >/tmp/a 2>&1):
close(3)                                = 0
open("/dev/xvc0", O_RDONLY)             = -1 ENOENT (No such file or
directory)
readlink("/proc/self/fd/0", "/dev/pts/0", 4096) = 10
readlink("/proc/self/fd/1", "/tmp/a", 4096) = 6
readlink("/proc/self/fd/2", "/tmp/a", 4096) = 6
open("/dev/console", O_RDWR)            = 3
ioctl(3, TIOCLINUX, 0xbfa42fd3)         = 0
vm86old(0)                              = -1 EFAULT (Bad address)
access("/proc/xen", X_OK)               = -1 ENOENT (No such file or
directory)
open("/dev/zero", O_RDWR)               = 6
mmap2(0x10000, 524288, PROT_READ|PROT_WRITE|PROT_EXEC,
MAP_SHARED|MAP_FIXED, 6, 0) = 0x10000
close(6)                                = 0
open("/dev/mem", O_RDWR)                = 6
mmap2(NULL, 1282, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_SHARED|MAP_FIXED,
6, 0) = 0
mmap2(0xa0000, 393216, PROT_READ|PROT_WRITE, MAP_SHARED|MAP_FIXED, 6,
0xa0) = 0xa0000
close(6)                                = 0
iopl(0x3)                               = 0
ioperm(0, 0x400, 0x1)                   = 0
rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0
vm86old(0x80898ac)                      = -1 ENOSYS (Function not
implemented)

The last two lines (rt_sig and vm86old) repeat for several hundred lines
until:

rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0
vm86old(0x80898ac)                      = -1 ENOSYS (Function not
implemented)
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++
Process 3020 detached

Any ideas ?

-- 

Tom C