Bug#495014: Acknowledgement (linux-source-2.6.26: IDE reset, DMA, 32-bit access killed on hald startup, via driver)

[David Baron <d_baron@012.net.il>]

I replaced the suspect hard disk. Used 2.6.26 (Ingo's RT patch) for a couple 
of days without problems. Then it happened again.

Another problem which could be related to this kernel: Recent nted (compile 
with cairo and pango libraries) do not terminate properly, stuck, unkillable 
in memory (disk-sleep mode?) and this takes other unrelated applications with 
it, in one case qjackctl (qt4) and tripwire stuck with the same status. This 
does not happen with 2.6.23-rt.

Another symptom? is that the disks run very hot.

This needs be checked out thoroughly because of possible disk damage, aside 
from usability issues involved.