[parisc-linux-cvs] Diff for recent submission

[Randolph Chung]

In reference to a message from John Marvin, dated Oct 24:
> Here is the diff for a bug fix I just checked in. t0 was holding the
> address of pa_dbit_lock in the 32 bit version of the PA2.0 dirty bit
> handler. Some code was inserted later that wound up trashing t0,
> causing us to take an unaligned fault with the Q bit off. I changed
> all instances of the inserted code to use t1, rather than t0, although
> only one instance was causing the problem.

awesome! jsm, can you please check this into the 2.5 tree too?

randolph