I have to disappoint all who complained here. Looks like it's not a bug, but a feature. My disk was dying for a long time, and new kernel just made it clear to me. After replacement, I have no problems with kernel 2.6.32-32-generic. Probably new drivers allow for less rereads and have stricter DMA wait limits.
I have to disappoint all who complained here. Looks like it's not a bug, but a feature. My disk was dying for a long time, and new kernel just made it clear to me. After replacement, I have no problems with kernel 2.6.32-32-generic. Probably new drivers allow for less rereads and have stricter DMA wait limits.