Diary and notebook of whatever tech problems are irritating me at the moment.

20071210

Keyboard LED flashing panic

I was in the middle of replying to an email from my home-made desktop PC when it suddenly stopped responding to anything. The only activity was the Caps Lock and Scroll Lock LEDs flashing at about a 1 second (1Hz) interval. At first I thought a PS/2 connector had come loose on my D-Link DKVM-4 KVM or on the system but the USB mouse was also not responsive. After some Google searching with my laptop I figured out that the kernel had probably panicked and the LED activity was caused by the panic_blink function which is basically an idiot light. According to what I read the flashing can also involve the Num Lock LED but mine may have been off because the Num Lock was active at the time of the panic. While there is no information conveyed by any of them it does let you know that something serious happened, not just a loose connector or X hanging. There was a kernel patch submitted by Andrew Rodland back in 2002 to blink in Morse code but I don't think it was ever included in the main tree. In my experience, once a Linux system has been stable for a while a panic usually indicates a hardware problem. After a few hours of troubleshooting I determined that my 1GB PC-3200 DIMM had failed, probably because it wasn't seated in the socket properly. I'm surprised it lasted as long as it did.

4 comments:

kwacka said...

Thanks for the pointer - found when 'googling' the same fault, saved me the hours you put in checking.

In my case it appears it failed because of a build-up of dust; memory is placed directly next to the CPU with the fan blowing directly onto the memory.

Swapping modules suggest that one memory stick AND the socket holding it have become faulty.

Hope this infor is of use to someone.

This is an MSI K9VGM motherboard.

jhansonxi said...

I had a socket fail once. It was caused by the module not being seated fully on one end and electrical current flowing through one of the power pins with a bad connection caused it to overheat and peel the contact off the module and burn the matching pins in the socket.

Dirt alone shouldn't cause it to fail but it can cause overheating by thermally insulating the module and preventing proper air flow. Unless the modules are high speed (high heat dissipation) and fairly close together (high heat density) they shouldn't require a lot of cooling and can operate at high temperatures without damage. Check the manufacturer specs to see what their maximum operating temperature is. In your situation the overheating module and its proximity to the CPU could indicate the CPU was also running rather hot but it or the motherboard usually have overheat protection and reduce the clock speed when it occurs.

Often the first thing to fail in these situations is the CPU fan bearing if it is mounted directly on the CPU heat sink. Might want to replace it just to be safe.

Jnanesh said...

I am also facing the same problem...

I am working with PCI wifi cards. Using Madwifi as driver.

I experience the situation of kernel-panic when i try to use adhoc mode.

Does anybody has any idea how to solve it?

jhansonxi said...

I've never encountered that with my Toshiba laptop and its Atheros chip. Try the support forums or mailing lists for your Linux distribution, the MadWifi mailing lists, Linux Wireless site, and LinuxQuestions.org network forum.

About Me

Omnifarious Implementer = I do just about everything. With my usual occupations this means anything an electrical engineer does not feel like doing including PCB design, electronic troubleshooting and repair, part sourcing, inventory control, enclosure machining, label design, PC support, network administration, plant maintenance, janitorial, etc. Non-occupational includes residential plumbing, heating, electrical, farming, automotive and small engine repair. There is plenty more but you get the idea.