Re: [PATCH] xfs: always free inline data before resetting inode fork during ifree

[Michal Hocko]

On Wed 28-03-18 19:33:06, Sasha Levin wrote:
> On Wed, Mar 28, 2018 at 03:21:48PM +0200, Michal Hocko wrote:
> >On Tue 27-03-18 19:54:35, Luis R. Rodriguez wrote:
> >> On Tue, Mar 27, 2018 at 09:06:37AM +0200, Michal Hocko wrote:
> >> > So by no means the MM backports were reviewed by me. And considering how hard
> >> > it is to get any review for MM patches in general I strongly suspect that
> >> > others didn't review either.
> >> >
> >> > In general I am quite skeptical about the automagic backports
> >> > selections, to be honest. MM patches should be reasonably good at
> >> > selecting stable backports and adding more patches on top just risks
> >> > regressions.
> >>
> >> BTW other than suggesting we needing *actual review* of the MM patches, are
> >> there known unit tests which could be run as well? Thinking long term.
> >
> >There are some in selftests but most fixes are quite hard to get a
> >specialized testcase for. Rememeber the MM is a pile of heuristics to
> >handle large scale of workloads.
> 
> Would running mmtests for each patch help here at all?

mmtests are more for performance than regression/correctness testing
AFAIR.

-- 
Michal Hocko
SUSE Labs