cvs commit: src/sys/vm uma_core.c

[Alan L. Cox]

Peter Wemm wrote:
> 
> Alan Cox wrote:
> >   Modified files:
> >     sys/vm               uma_core.c
> >   Log:
> >   Gulp ... call kmem_malloc() without Giant.
> 
> So, let me guess, now would be a great time to update the kernel on my
> desktop, right? :-)
> 

I sincerely hope that it doesn't turn out to be a bad time.  :-)

To everyone: There are two "failure modes" that I would anticipate if
something is wrong.  The first would be an assertion failure that Giant
isn't held somewhere.  This is no big deal.  The second would be
seemingly random kernel crashes, probably in the vm system, due to
corruption of a vm object or vm page queue.  This is worrisome.  It
means that I've missed some access to a vm object or the vm page "cache"
queues that requires synchronization.  Unfortunately, due to the use of
vm objects as a kind of page table page container in the various pmaps,
I haven't been able to add as many assertions as I would like.

Regards,
Alan