Re: [PATCH] 2.1.88 Hanging Processes (Uninterruptible Sleep)

Gerard Roudier (groudier@club-internet.fr)
Wed, 4 Mar 1998 00:04:01 +0100 (MET)

I forgot to mention SMP in my reply, but Linus's explanation was for
SMP and so I just assumed we are only speaking about SMP.
I agree than UP should not be affected by the problem (assuming, perhaps
wrongly (?), that interrupts are delivered after a whole instruction
having been executed in program order and that the CPU has a consistent
view of the memory it intended to modify regardless its buffering
features under UP).

Gerard.

On Tue, 3 Mar 1998, MOLNAR Ingo wrote:

>
> > I would think that the problem is not specific to PPro/PII and that the
> > good explanation is likely a wrong memory ordering assumption in the
> > code rather that intruction reordering. However, I agree that instruction
> > re-ordering increases the probability of memory read/write being reordered.
>
> note that those assumptions are perfectly valid on UP. It's just on SMP
> that we run IRQ handlers and the kernel in parallel on two or more CPUs,
> so such issues start popping up. (and we run kernel code and kernel code
> in parallel for more and more subsystems as well)
>
> > So, my thought is that the initial construct that did not force any memory
> > ordering was only safe for 386 processors.
>
> it was/is safe on any UP system, and it's also safe on 2.0 SMP.
>
> -- mingo

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu