Comment 97 for bug 204996

Thanks for that link, brainiac8008. Those are exactly the symptoms I am experiencing (and I think this report is about), and someone had previously thought it could be an I/O problem because the kernel was not writing logs even with Alt-SysRQ stuff. (Although it was perfectly content blinking the caps lock light).

I am concerned that there seem to be a lot of completely unrelated bug reports mingled with this one.
This particular report regards a lockup that has not been attributed to any particular user-space software, and is not a typical freeze, or xorg's doing. The kernel itself is dieing (at random) with a kernel panic, and not recording information about it. A kernel panic can often be recognized by flashing caps lock and scroll lock lights.

Having said that, we seem to be also looking at a glaring problem with a very easily upset kernel here. I have had a problem with hard drive becoming suddenly unusable before, both on this laptop with a SATA drive and on a desktop with an IDE drive. Could this be that same issue evolved to epic proportions because of the kernel now falling over on its side as a result?

In fact, I am seeing a lot of old bugs I have experienced myself mentioned here, suddenly causing system-wide lockups instead of just Samba no longer working or wireless (ndiswrapper?) disappearing.

I guess that gives us an important question of filing: Is this bug report about the kernel being really unstable, or a particular i/o bug? I am thinking the latter, since it would seem that is the one we can blame for logs not being written. With that in mind, perhaps the general instability stuff that is not an i/o problem could be pulled to another bug report? (Or have I gone mad?)

Regarding i/o bug: On my laptop, I have turned down swappiness and have a lot of real memory. The swap file is rarely, if ever, written to. Could it be that each system's reliance on the swap file is affecting the experienced symptoms?