Rather than back 075 and 112 out I reverted xfstest tests/generic/075 and 112 to their state from a few weeks ago by backing out the patch I mentioned below Kicked off http://smb3-test-rhel-75.southcentralus.cloudapp.azure.com/#/builders/4/builds/110 with that change to the buildbot On Mon, Mar 4, 2019 at 11:00 PM Steve French <smfrench@xxxxxxxxx> wrote: > > I tried it to Samba a few minutes ago and it worked fine with current for-next > and also:q it looks important (data integrity etc.). Test description > > # Test that mmap read doesn't see non-zero data past EOF on truncate down. > # > # This is inspired by an XFS bug that truncate down fails to zero page cache > # beyond new EOF and causes stale data written to disk unexpectedly and a > # subsequent mmap reads and sees non-zeros post EOF. > > I have two test targets, both Samba localhost. One succeeds for all 8 of > the ones that we were worried about: > > ./check -cifs generic/013 generic/014 generic/024 generic/030 > generic/069 generic/075 generic/112 generic/125 generic/346 > generic/469 > > The other succeeds on all but 075 and 112 which worked on the 'old' > xfstests from a month ago, but fail on the one with the newer > xfstests. > > So I think we are ok with 469 ... but we do have to figure out what to > do with what seems to be either a regression in 075 and 112 xfstests > (a bug in the tests) or something that the updated tests are now > seeing as a cifs bug. My theory is that it is due to this xfstest > commit: > > root@smf-Thinkpad-P51:~/xfstests-dev# git log tests/generic/075 > commit ec295d73ac19a42d1f022cb074d0bd506252cb3b > Author: Christoph Hellwig <hch@xxxxxx> > Date: Fri Feb 15 13:41:40 2019 +0100 > > generic/075,112: detect preallocation support for fsx tests > > Currently generic/075 and generic/112 have two extra fsx passes each > that exercise fsx with preallocation, which are only enabled for > XFS. > > These tests can also be run with other file systems, given that the > XFS prealloc ioctls are implemented in generic code since the > addition of the fallocate system call. This also means a version of > XFS that does not support preallocation (e.g. because it always > writes out of place) can skip the prealloc tests while still > completing the normal fsx tests just fine. > > Signed-off-by: Christoph Hellwig <hch@xxxxxx> > Reviewed-by: Eryu Guan <guaneryu@xxxxxxxxx> > Signed-off-by: Eryu Guan <guaneryu@xxxxxxxxx> > > On Mon, Mar 4, 2019 at 10:10 PM ronnie sahlberg > <ronniesahlberg@xxxxxxxxx> wrote: > > > > generic/469 does it work for you? > > It fails in the tests. > > > > I have tried it locally and it always fails here. With current > > for-next as well as for-next as of a month ago. > > Same for xfs-tests, with current as well as months old master. > > > > Lets remove it. > > > > On Sun, Mar 3, 2019 at 3:05 PM Steve French <smfrench@xxxxxxxxx> wrote: > > > > > > Added four xfs subtests to the cifs-testing buildbot (and 3 of these > > > were also missing from the Azure buildbot so added them to Azure as > > > well) > > > + [ "generic/464", "smb3"], > > > + [ "generic/469", "smb3"], > > > + [ "generic/524", "smb3"], > > > + [ "generic/528", "smb3samba], > > > > > > Testing the three new tests for Azure in > > > http://smb3-test-rhel-75.southcentralus.cloudapp.azure.com/#/builders/4/builds/105 > > > > > > > > > -- > > > Thanks, > > > > > > Steve > > > > -- > Thanks, > > Steve -- Thanks, Steve
