Hugetlb HPTE hashing weirdness
benh at kernel.crashing.org
Tue Sep 27 08:23:04 EST 2005
On Mon, 2005-09-26 at 17:21 -0500, Adam Litke wrote:
> On Fri, 2005-09-23 at 15:15 +1000, Benjamin Herrenschmidt wrote:
> > > Attached is my somewhat hacky hugetlb test suite. Build it with 'make
> > > all'. I've been reproducing with:
> > > for i in `seq 1 10`; do ./mmap-gettest 1 10; done
> > > After about the third or fourth iteration, you should start to see test
> > > failures.
> > What machine do you use ?
> > I can reproduce the problem on a g5 here but strangely not on any LPAR
> > machine I've tested so far (all POWER5)...
> OK, another data point. I've tested on a Power5 (p570) and I don't see
> the bug. So it must be Power4-specific.
Either that, or non-LPAR specific. The previous problem, for which I
pushed a patch, was specific to non-LPAR hash management code.
More information about the Linuxppc64-dev