I/O breakage in today's current

[Kevin Oberman]

After rebuilding the current kernel today (cvsup at 16:20 UTC) on my T30
ThinkPad, I saw to major (though possibly unrelated) problems.

1. My second disk (master on the secondary bus) suddenly probes as
   having a non-ATA66 cable. It is a laptop drive and has not cable. It
   probes fine with the old kernel and runs at UDMA5. Attempts to use
   atacontrol to reset it to UDMA5 received the same message. The drive
   does work and I had not problems reading or writing. It was just
   slow. The master on the primary bus (the system disk) probed correctly.

2. My network card (Intel Pro 100 VE) probes fine and looks fine, but
   never seems to send a packet. I monitored the Ethernet using tcpdump on
   a directly connected system and receives NO packets. dhclient reported
   the error "send_packet: Network is down".

I don't know if these are two different problems that cropped up at the
same time or a common failure. A kernel built from sources dated
yesterday at 16:29 UTC works fine. Our cvs mirror updates every hour, so
there is some fuzz in the times.

Any ideas? (mdodd just fixed one really annoying problem yesterday and
today I get an even more annoying problem. The joy of running
current. :-(

If I don't see any report of what might be wrong, I'll try to do some
more in-depth testing tomorrow. No time today.

Thanks,
-- 
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634