Date:   Tue, 4 Jan 2022 11:56:34 -0500
From:   Bruce Fields <bfields@fieldses.org>
To:     Chuck Lever III <chuck.lever@oracle.com>
Cc:     Linux NFS Mailing List <linux-nfs@vger.kernel.org>
Subject: Re: NFSv4 OPEN returns a zero cinfo.after on tmpfs
Message-ID: <20220104165634.GE7815@fieldses.org>
References: <49640909-A7F0-4004-AF55-859621B26D38@oracle.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <49640909-A7F0-4004-AF55-859621B26D38@oracle.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Precedence: bulk

On Fri, Dec 24, 2021 at 04:46:26AM +0000, Chuck Lever III wrote:
> During some testing I noticed that OPEN frequently returns a
> zero in the cinfo.after field on my test share, which is tmpfs.
> Does not seem to be an issue for xfs.

Thanks for catching this.

> An easy way to address this would be to revert 428a23d2bf0c. But I
> wonder if there are any particular regression tests in the pynfs
> suite that could detect this kind of misbehavior, in case someone
> would like to try to re-implement the optimization in 428a23d2bf0c.

From a quick grep the only tests I see checking cinfo are in
nfs4.0/servertests/st_rename.py

--b.