Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 5 May 2023 22:29:40 +0800
From:   Zorro Lang <zlang@redhat.com>
To:     Anna Schumaker <anna@kernel.org>
Cc:     linux-nfs@vger.kernel.org, fstests@vger.kernel.org
Subject: Re: [PATCH] nfs/002: Add a test for xattr ctime updates
Message-ID: <20230505142940.46iulwinf27o4wil@zlang-mailbox>
References: <20230504204847.405037-1-anna@kernel.org>
 <20230505044458.wrvhvamrkcuqin3z@zlang-mailbox>
 <CAFX2Jfk15M9_097BjTPicZ-G0p+TYVqjGmzvZX8r-D7mAOX21Q@mail.gmail.com>
 <20230505140519.xtloxhejjw7hoc24@zlang-mailbox>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20230505140519.xtloxhejjw7hoc24@zlang-mailbox>
Precedence: bulk

On Fri, May 05, 2023 at 10:05:19PM +0800, Zorro Lang wrote:
> On Fri, May 05, 2023 at 09:11:52AM -0400, Anna Schumaker wrote:
> > Hi Zorro,
> > 
> > On Fri, May 5, 2023 at 12:45 AM Zorro Lang <zlang@redhat.com> wrote:
> > >
> > > On Thu, May 04, 2023 at 04:48:47PM -0400, Anna Schumaker wrote:
> > > > From: Anna Schumaker <Anna.Schumaker@Netapp.com>
> > > >
> > > > The NFS client wasn't updating ctime after a setxattr request. This is a
> > > > test written while fixing the bug.
> > > >
> > > > Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
> > > > ---
> > > >  tests/nfs/002     | 39 +++++++++++++++++++++++++++++++++++++++
> > > >  tests/nfs/002.out |  2 ++
> > > >  2 files changed, 41 insertions(+)
> > > >  create mode 100755 tests/nfs/002
> > > >  create mode 100644 tests/nfs/002.out
> > > >
> > > > diff --git a/tests/nfs/002 b/tests/nfs/002
> > > > new file mode 100755
> > > > index 000000000000..5bfedef6c57d
> > > > --- /dev/null
> > > > +++ b/tests/nfs/002
> > > > @@ -0,0 +1,39 @@
> > > > +#! /bin/bash
> > > > +# SPDX-License-Identifier: GPL-2.0
> > > > +# Copyright (c) 2023 Netapp Inc., All Rights Reserved.
> > > > +#
> > > > +# FS QA Test 002
> > > > +#
> > > > +# Test a bug whene the NFS client wasn't sending a post-op GETATTR to the
                      ^^ ?

> > > > +# server after setting an xattr, resulting in `stat` reporting a stale ctime.

If there's a known bug fix (git commit) in linux, you can use
_fixed_by_kernel_commit(), refer to other cases which use this function.

> > > > +#
> > > > +. ./common/preamble
> > > > +_begin_fstest auto quick attr
> > > > +
> > > > +# Import common functions
> > > > +. ./common/filter

I think the common/filter isn't needed.

> > > > +. ./common/attr
> > > > +
> > > > +# real QA test starts here
> > > > +_supported_fs nfs
> > >
> > > Great, a new nfs test case!
> > >
> > > > +_require_test_nfs_version 4.2
> > >
> > > But I'm wondering if this case can be a generic test case, due to the operations
> > > in this case are common (need xattr and ctime support), don't depend on
> > > any nfs specific features/operations.
> > 
> > This probably could be a generic test case.
> 
> Great :)
> 
> > 
> > >
> > > Not sure why nfs4.2 is necessary, can it be removed or replaced ?
> > 
> > That's because xattrs were added to the NFS protocol in NFS v4.2, so I
> > filtered out the other versions since they're not going to run anyway.
> > I think xattr support is already checked to properly skip this on
> > other versions, however, so changing this to a generic test shouldn't
> > create a new failure on earlier NFS versions.
> 
> That makes sense.
> 
> > 
> > Should I send a v2 with those changes? And should I find an open test
> 
> Sure, please send v2 to change this case to be a generic test case. Then we
> can check if more other filesystems has this issue :)
> 
> > number, or choose something like "generic/999"?
> 
> That depends on you. Due to there's only one test case in this patch, so you
> just need to base on latest for-next branch, then choose a number which has
> been taken, I'll deal with that if there's conflict when I merge.
> 
> > Anna
> > >
> > > Thanks,
> > > Zorro
> > >

_require_test

> > > > +_require_attrs
> > > > +
> > > > +touch $TEST_DIR/testfile

We can't be sure there's not a file (or even a directory or others) named
"testfile" in TEST_DIR when fstests is running. So better to remove it at
first. E.g

  rm -rf $TEST_DIR/testfile

> > > > +
> > > > +before_ctime=$(stat -c %z $TEST_DIR/testfile)
> > > > +$SETFATTR_PROG -n user.foobar -v 123 $TEST_DIR/testfile
> > > > +after_ctime=$(stat -c %z $TEST_DIR/testfile)

Might "%Z" be better to be compared? (optional)

Thanks,
Zorro

> > > > +
> > > > +test "$before_ctime" != "$after_ctime" || echo "Expected ctime to change."
> > > > +
> > > > +
> > > > +before_ctime=$after_ctime
> > > > +$SETFATTR_PROG -x user.foobar $TEST_DIR/testfile
> > > > +after_ctime=$(stat -c %z $TEST_DIR/testfile)
> > > > +
> > > > +test "$before_ctime" != "$after_ctime" || echo "Expected ctime to change."
> > > > +
> > > > +echo "Silence is golden"
> > > > +status=0
> > > > +exit
> > > > diff --git a/tests/nfs/002.out b/tests/nfs/002.out
> > > > new file mode 100644
> > > > index 000000000000..61705c7cc203
> > > > --- /dev/null
> > > > +++ b/tests/nfs/002.out
> > > > @@ -0,0 +1,2 @@
> > > > +QA output created by 002
> > > > +Silence is golden
> > > > --
> > > > 2.40.1
> > > >
> > >
> >