Bug or what?

[P-O Blomstrand]

Found what seems like a bug in kernel 2.6.9-1.ydl.8g5. If not, excuse
the ignorance of a newbie. This is the story:

Installed YDL 4.0 on a G5, 1,6 GHz with 1280 Mb memory. No problems
installing. Video, screen, network etc was identified properly and
worked out of the box. Used yum to update to the 2.6.9 kernel. When I
tried to shutdown for the day hell broke lose. Half way through the
shutdown process, just before "Sending all processes the TERM signal",
the process stops with unreadable messages quickly scrolling down the
screen. The only way to break out was to use the power button. Booted
again and studied logs. In /var/log/messages I find tons of this:

Jan  7 09:18:39 localhost kernel: Badness in pci_find_subsys at
drivers/pci/search.c:167
Jan  7 09:18:39 localhost kernel: Call trace:
Jan  7 09:18:39 localhost kernel:  [c00075e0] check_bug_trap+0xc0/0x128
Jan  7 09:18:39 localhost kernel:  [c0007bdc]
ProgramCheckException+0x1c0/0x2a0
Jan  7 09:18:39 localhost kernel:  [c0007038]
ret_from_except_full+0x0/0x4c
Jan  7 09:18:39 localhost kernel:  [c015ba24] pci_find_subsys+0x1c/0xe0
Jan  7 09:18:39 localhost kernel:  [c000f5dc]
pci_device_from_OF_node+0xb4/0x13c
Jan  7 09:18:39 localhost kernel:  [c04630cc] 0xc04630cc
Jan  7 09:18:39 localhost kernel:  [c04631f4] 0xc04631f4
Jan  7 09:18:39 localhost kernel:  [f1184bb0] gem_put_cell+0x6c/0x7c
[sungem]
Jan  7 09:18:39 localhost kernel:  [f118731c] gem_link_timer+0x108/0x43c
[sungem]
Jan  7 09:18:39 localhost kernel:  [c0029e68]
run_timer_softirq+0xe0/0x1ac
Jan  7 09:18:39 localhost kernel:  [c0025ff8] __do_softirq+0x78/0xf8
Jan  7 09:18:39 localhost kernel:  [c00260d0] do_softirq+0x58/0x5c
Jan  7 09:18:40 localhost kernel:  [c00094f8]
timer_interrupt+0x204/0x24c
Jan  7 09:18:40 localhost kernel:  [c0007084] ret_from_except+0x0/0x14

Understand enough of this to realize that it's not possible to kill the
network. The same messages scroll by if I tail /var/log/messages and use
the Network GUI to deactivate eth0. Activated again everything is back
to normal after a few seconds.
Switched back to the original 2.6.8 kernel and the problem was gone.

What is this? Bug or what? Anyone with insight?

/PO