Re: [PATCH 3/3] hugetlbfs: replace hugetlb_vma_lock with invalidate_lock

[Rik van Riel]

On Mon, 2023-09-25 at 13:06 -0700, Mike Kravetz wrote:
> On 09/25/23 15:22, Rik van Riel wrote:
> > On Mon, 2023-09-25 at 10:04 +0800, kernel test robot wrote:
> > > Hi,
> > > 
> > > kernel test robot noticed the following build errors:
> > > 
> > > [auto build test ERROR on akpm-mm/mm-everything]
> > > [also build test ERROR on linus/master v6.6-rc3 next-20230921]
> > > [If your patch is applied to the wrong git tree, kindly drop us a
> > > note.
> > > And when submitting patch, we suggest to use '--base' as
> > > documented
> > > in
> > > https://git-scm.com/docs/git-format-patch#_base_tree_information]
> > 
> > OK, so I have a fix for patch 3/3 that gets rid of the
> > compile error, but the libhugetlbfs test cases show that
> > patch 3/3 opens up a condition where resv_hugepages 
> > underflows.
> > 
> > I have not figured out the cause of that yet, but
> > patches 1 & 2 seem to survive all tests fine.
> 
> In addition, I suspect patch 3 is going to cause a performance
> regression.
> It is taking me a little while to resurrect the test environment used
> when
> the hugetlb vma lock was introduced.  My plan is to exercise the
> series in
> that environment.
> 
I am planning to send a v3 of the series soon, once I have
confirmed that the bugs in patch 3 have all been fixed.

I have no strong opinion on whether or not patch 3 gets
merged at all. Patches 1 & 2 fix the actual bug that I am
trying to fix, and I am perfectly fine if patch 3 ends up
getting dropped in the end.

It seemed worth trying to get that cleanup though ;)

> I should be able to review patches 1 & 2 later (my) today.

Thank you!

-- 
All Rights Reversed.