On Tue, Jan 7, 2020 at 9:45 AM Xiao Yang <yangx.jy@xxxxxxxxxxxxxx> wrote: > > On 2020/1/7 14:47, Amir Goldstein wrote: > > No it is not wrong, but this is different than skipping keep_size. > > My point is that when fsx is run with some parameters and > > produces a replay-ops log. If that replay-ops log is replayed with > > exact same parameters it should produce the exact same sequence. > > > > In the case of punch/zero/collapse, if those are skipped when recording > > the ops log, then ops log includes the keyword "skip ..." explicitly, but > > but for skipped keep_size, keep_size will simply not be in the log. > Hi Amir, > > If keep_size is not supported, ops log can also includes the keyword > "skip ..." by Eryu's patch: > https://lore.kernel.org/fstests/20191022123115.12250-1-eguan@xxxxxxxxxxxxxxxxx/ > > Do you want to add require support for keep_size/punch/zero/collapse and > accept Eryu's patch as well? > I am perfectly fine with Eryu's patch and I think it is correct to merge it, but it is completely independent to fixing the 4 tests. Eryu's patch is changing the way the ops log is recorded (for the better), but it doesn't change (and I think it is wrong to change) the way that ops log is replayed. The hardcoded ops logs in those 4 tests are not going to change because you modify fsx. The hardcoded ops logs in those 4 tests intentionally include the keep_size directive, so those tests should not be run on NFS. Thanks, Amir.
