apm appears to lockup/freeze PC
 

I have recently installed slackware 10.2 on an old Compaq Presario 5170. This PC has a Pentium II 350MHz with 320 meg memory.

I did a full install of slackware from 2 CDs and installed the default 2.4.31 kernel.

I uncommented "/sbin/modprobe apm" in /etc/rc.d/rc.modules to allow the PC to be automatically powered off after issuing a "shutdown -h now" command.

Having the apm module loaded appears to be causing a lockup problem to the point where the PC does not respond to any keyboard commands and I can not connect to the PC by ssh. After doing a hard shutdown and then commenting out "/sbin/modprobe apm" in rc.modules, and then rebooting, the lockup problem goes away. The lockup appears more quickly while in a X session (XFCE).

The apm module does power off the PC. I have successfully used live distros like DSL and KNOPPIX 3.4 which use apm without this problem.

Any thoughts on how to correct this problem?

An update; my problem is something other than apm. Without the apm module loaded, the PC locked up again, but this time it had a Kernel Panic after some type of I/O errors with a reiserfs partition.

Booted up with KNOPPIX and ran "reiserfsck --check" on the partitions. They checked out fine after replaying some journal entries on the root partition.

Any thoughts on what to check to determine the source of the problem?

I don't know if this thread will help:

http://www.linuxquestions.org/questi...d.php?t=446141

That person also had problems with keyboard lockups with a Presario.

Woodsman, Thanks for the reply. The Presario 5170 is a desktop PC and had been very stable. My problems usually occur after being in a X session for a period of time.

I also have slackware 10.2 installed on an old Presario 1210 laptop (Pentium 150 Mhz) and it has not given me any problems.

Open your box, then clean and reseat your memory sticks. Sometimes the least amount of oxidation, not detectable by the human eye, will cause memory stick problems. A light rubbing with a pencil eraser is sufficient to clean contacts.

You might want to check your BIOS settings for your RAM and bump the bus speed to one level slower. I had a box that would work fine for days and then suddenly would burp similar to what you describe. I changed the RAM speed to one level slower and never had a related problem thereafter. I noticed no performance hit.

Also, reseat all ribbon and device connectors.

Usually or always? After trying the above you then might want to focus on your /etc/X11/xorg.conf and video driver. A lot of people have problems with certain video drivers. Browse the forum for more info. You could try using the generic VESA driver for several days to see what happens.

A challenge with troubleshooting is isolating the problem before trying to derive a solution. Consider temporarily disabling all possible services: networking, APM, CUPS, etc. Run with the bare bones possible. The less overhead you have running the easier to narrow the possible problems.

Hopefully some other people will pipe in to help.

Woodsman, I appreciate your thorough reply with some very good suggestions.

After hearing an odd sound, like a click, coming from the PC, I suspected my problem was a bad hard drive.

So using "partimage", "qtparted", and "k3b" while running KNOPPIX on the PC, I was able to successfully make image copies of the old hard drive's partitions and then burn them to CDs. From there, I partitioned the replacement drive to match the old drive's partitions. And then using "partimage", restored the image copies to the replacement drive. Booted from a floppy disk to do the final step of installing lilo using "liloconfig".

So far, "knock on wood", replacing the hard drive appears to have solve my problem. The old drive, is a Western Digital 40GB that I had used in other PCs and is over 4 years old. The replacement drive is a newer Maxtor 80GB. I was pleasantly surprised when the BIOS in this 8 year old PC recognized the full 80 GBs.

As it has been said, the common wisdom for hard drives is not if they are going to fail but instead when they are going to fail. I feel fortunate that I was able to image copy the old drive before it completely failed.

If replacing the hard drive does not solve the problem and I will follow your suggestions.