Re: [PATCH] make voyager work again after the cpumask_t changes

[William Lee Irwin III <wli@holomorphy.com>]

James Bottomley <James.Bottomley@SteelEye.com> wrote:
>> Most is just simple fixes; however, the needless change from atomic to
>> non-atomic operations in smp_invalidate_interrupt() caused me a lot of
>> pain to track down since it introduced some very subtle bugs.

On Thu, Aug 28, 2003 at 12:10:16PM -0700, Andrew Morton wrote:
> Yes, the generic code was like that too.  It was causing lockups.  Sorry, I
> did not realise that voyager had a private invalidatation implementation.
> Officially smp_invalidate_needed should be a cpumask_t and
> smp_invalidate_interrupt() should be using cpu_isset() rather than
> open-coded bitops.  For all those 64-way voyagers out there ;)
> (Actually it is legitimate: you may want to run a NR_CPUS=48 kernel on a
> 2-way voyager just for testing purposes).  I'll drop your patch in as-is,
> and maybe Bill can take a look at cpumaskifying it sometime?

I'm not convinced it's worth it; AIUI there are architectural limits to
Voyager that prevent it from ever supporting > 32x in hardware, though
it could be worth doing so in tandem with an across-the-board all-
subarch extension of generic i386 support.


-- wli