[ofa-general] Re: [PATCH 08 of 11] anon-vma-rwsem

[Christoph Lameter]

On Wed, 7 May 2008, Linus Torvalds wrote:

> On Wed, 7 May 2008, Christoph Lameter wrote:
> > 
> > Multiple vmas may share the same mapping or refer to the same anonymous 
> > vma. The above code will deadlock since we may take some locks multiple 
> > times.
> 
> Ok, so that actually _is_ a problem. It would be easy enough to also add 
> just a flag to the vma (VM_MULTILOCKED), which is still cleaner than doing 
> a vmalloc and a whole sort thing, but if this is really rare, maybe Ben's 
> suggestion of just using stop-machine is actually the right one just 
> because it's _so_ simple.

Set the vma flag when we locked it and then skip when we find it locked 
right? This would be in addition to the global lock?

stop-machine would work for KVM since its a once in a Guest OS time of 
thing. But GRU, KVM and eventually Infiniband need the ability to attach 
in a reasonable timeframe without causing major hiccups for other 
processes.

> (That said, we're not running out of vm flags yet, and if we were, we 
> could just add another word. We're already wasting that space right now on 
> 64-bit by calling it "unsigned long").

We sure have enough flags.