Re: smp irq affinity and new irq stuff

[Andrea Arcangeli]

On Sun, 27 Feb 2000, Andrea Arcangeli wrote:

>[..] and so it's also the only
>one who currently needs lowlevel controller locking [..]

Thinko: sys_sable and sys_rawhide after the per-desc-lock thing will need
a lowlevel controller lock too (as sys_dp264) that I have not implemented
in my patch yet. That's an incremental work. Adding a spinlock around the
I/O controller accesses in such two files will make them SMP safe too.
I'll do that tomorrow...

Andrea