Hugetlb HPTE hashing weirdness

[Benjamin Herrenschmidt]

On Mon, 2005-09-26 at 16:31 -0500, Adam Litke wrote:
> On Tue, 2005-09-27 at 07:18 +1000, Benjamin Herrenschmidt wrote:
> > > That patch (as included in 2.6.14-rc2-git5) makes the problem disappear
> > > for the prefault case but it is still broken for demand faulting.  Seems
> > > some HPTEs are still not flushed.  I'll try on some LPAR systems and see
> > > if I can reproduce there.
> > 
> > I haven't tried demand faulting, do you have a test case ?
> 
> I am working on a kernel patch (with David Gibson) to support it.  I've
> attached the patch below.  The test case is to run the same thing
> (./mmap-gettest 1 10) repeatedly with the kernel patch applied.

Ok, I'll have a look, thanks

Ben.