On Thu, Nov 29, 2018 at 02:36:50PM +0200, Amir Goldstein wrote: > > Same again - what's the test plan for these cherry-picked data > > corruption fixes? > > > > Dave, > > Just to check if we are on the same page, if this was the test plan: > https://www.spinics.net/lists/linux-xfs/msg20639.html > > Would "no regressions from baseline" have been sufficient to validate > those specific patches are solid for stable? No, not in this case, because fsx in fstests only does 100k ops at most - it's a smoke test. This isn't sufficient to regression test fixes that, in some cases, took hundreds of millions of fsx ops to expose. Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx
