Re: [BUG] 2.6.25-rc4 hang/softlockups after freeing hugepages

[Lee Schermerhorn <Lee.Schermerhorn@hp.com>]

On Thu, 2008-03-06 at 09:53 -0800, Nishanth Aravamudan wrote:
> On 06.03.2008 [12:23:03 -0500], Lee Schermerhorn wrote:
> > Test platform:  HP Proliant DL585 server - 4 socket, dual core AMD with
> > 32GB memory.
> > 
> > I first saw this on 25-rc2-mm1 with Mel's zonelist patches, while
> > investigating the interaction of hugepages and cpusets.  Thinking that
> > it might be caused by the zonelist patches, I went back to 25-rc2-mm1
> > w/o the patches and saw the same thing.  It sometimes takes a while for
> > the softlockups to start appearing, and I wanted to find a fairly
> > minimal duplicator.  Meanwhile 25-rc3 and rc4 have come out, so I tried
> > the latest upstream kernel and see the same thing.
> 
> So, does 2.6.25-rc2 show the problem? Or was it something introduced in
> that -mm which has since gone upstream?
> 

I don't recall that I went back that far.  I'll try Ingo's patch [later,
after an obligatory meeting...] and let you know.

<snip>
> > I took a look at the recent hugetlb patches from Adam and Nish, but none
> > seemed to address this symptom.  I don't think I'm dealing with surplus
> > pages here.
> 
> If /proc/sys/vm/nr_overcommit_hugepages = 0, then no, you're not.

I didn't set that, so it should have been zero.

Lee