bad block wackiness

[irony]

I recently upgraded the kernel on one of my boxes from 2.6.1 to 2.6.7, and when I tried the new kernel, it stopped while reiserfsck-ing the root partition. It found a bad block that couldn't be remapped. I ordered a new hard drive, since it's not worth messing about with.

When I tried to boot with the older kernel, it didn't find the bad block and booted fine.

Why would one kernel have a bad block and the older kernel not? The bad block is in the root partition, not the boot partition, which is the only thing that changed from kernel to kernel.
_________________
"and if rain brings winds of change, let it rain on us forever..."