Linux NFS development
 help / color / mirror / Atom feed
From: "J. Bruce Fields" <bfields@fieldses.org>
To: Chuck Lever <chuck.lever@oracle.com>
Cc: linux-nfs@vger.kernel.org
Subject: Re: [PATCH v2] Test renaming onto an open file
Date: Sat, 28 May 2022 16:19:24 -0400	[thread overview]
Message-ID: <20220528201924.GE9929@fieldses.org> (raw)
In-Reply-To: <165368065607.3280217.2106378952611449081.stgit@morisot.1015granger.net>

Applied, thanks.--b.

On Fri, May 27, 2022 at 03:44:55PM -0400, Chuck Lever wrote:
> There has been a recent report of RENAME onto an open file resulting
> in a bogus ESTALE when that open (and now deleted) file is closed.
> Build in some testing with and without SEQUENCE to ensure a server
> is handling this case as expected.
> 
> Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
> 
> Changes since v1:
> - None. This is a just a repost.
> 
> ---
>  nfs4.0/servertests/st_rename.py   |   16 ++++++++++++++++
>  nfs4.1/server41tests/st_rename.py |   19 ++++++++++++++++++-
>  2 files changed, 34 insertions(+), 1 deletion(-)
> 
> diff --git a/nfs4.0/servertests/st_rename.py b/nfs4.0/servertests/st_rename.py
> index afe9a7fe0a58..3302608719e3 100644
> --- a/nfs4.0/servertests/st_rename.py
> +++ b/nfs4.0/servertests/st_rename.py
> @@ -524,6 +524,22 @@ def testLinkRename(t, env):
>          t.fail("RENAME of file into its hard link should do nothing, "
>                 "but cinfo was changed")
>  
> +def testStaleRename(t, env):
> +    """RENAME file over an open file should allow CLOSE
> +
> +    FLAGS: rename all
> +    CODE: RNM21
> +    """
> +    c = env.c1
> +    c.init_connection()
> +    basedir = c.homedir + [t.word()]
> +    c.maketree([t.word(), b'file'])
> +    fh, stateid = c.create_confirm(t.word(), path=basedir + [b'file2'])
> +    res = c.rename_obj(basedir + [b'file'], basedir + [b'file2'])
> +    check(res)
> +    res = c.close_file(t.word(), fh, stateid)
> +    check(res, msg="CLOSE after RENAME deletes target returns ESTALE")
> +
>  ###########################################
>  
>      def testNamingPolicy(t, env):
> diff --git a/nfs4.1/server41tests/st_rename.py b/nfs4.1/server41tests/st_rename.py
> index 066df749379f..26ed4d275067 100644
> --- a/nfs4.1/server41tests/st_rename.py
> +++ b/nfs4.1/server41tests/st_rename.py
> @@ -1,5 +1,5 @@
>  from xdrdef.nfs4_const import *
> -from .environment import check, fail, maketree, rename_obj, get_invalid_utf8strings, create_obj, create_confirm, link, use_obj, create_file
> +from .environment import check, fail, maketree, rename_obj, get_invalid_utf8strings, create_obj, create_confirm, link, use_obj, create_file, close_file
>  import nfs_ops
>  op = nfs_ops.NFS4ops()
>  from xdrdef.nfs4_type import *
> @@ -529,3 +529,20 @@ def testLinkRename(t, env):
>      if scinfo.before != scinfo.after or tcinfo.before != tcinfo.after:
>          t.fail("RENAME of file into its hard link should do nothing, "
>                 "but cinfo was changed")
> +
> +def testStaleRename(t, env):
> +    """RENAME file over an open file should allow CLOSE
> +
> +    FLAGS: rename all
> +    CODE: RNM21
> +    """
> +    name = env.testname(t)
> +    owner = b"owner_%s" % name
> +    sess = env.c1.new_client_session(name)
> +    maketree(sess, [name, b'file'])
> +    basedir = env.c1.homedir + [name]
> +    fh, stateid = create_confirm(sess, owner, path=basedir + [b'file2'])
> +    res = rename_obj(sess, basedir + [b'file'], basedir + [b'file2'])
> +    check(res)
> +    res = close_file(sess, fh, stateid)
> +    check(res, msg="CLOSE after RENAME deletes target returns ESTALE")
> 

      reply	other threads:[~2022-05-28 20:19 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-27 19:44 [PATCH v2] Test renaming onto an open file Chuck Lever
2022-05-28 20:19 ` J. Bruce Fields [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220528201924.GE9929@fieldses.org \
    --to=bfields@fieldses.org \
    --cc=chuck.lever@oracle.com \
    --cc=linux-nfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox