Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Sender: Salvatore Bonaccorso <salvatore.bonaccorso@gmail.com>
Date:   Mon, 12 Oct 2020 17:41:59 +0200
From:   Salvatore Bonaccorso <carnil@debian.org>
To:     "J. Bruce Fields" <bfields@fieldses.org>
Cc:     linux-nfs@vger.kernel.org, Chuck Lever <chuck.lever@oracle.com>
Subject: Re: Kernel panic / list_add corruption when in nfsd4_run_cb_work
Message-ID: <20201012154159.GA49819@eldamar.lan>
References: <20201011075913.GA8065@eldamar.lan>
 <20201012142602.GD26571@fieldses.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20201012142602.GD26571@fieldses.org>
Precedence: bulk

Hi Bruce,

Thanks a lot for your reply, much appreciated.

On Mon, Oct 12, 2020 at 10:26:02AM -0400, J. Bruce Fields wrote:
> On Sun, Oct 11, 2020 at 09:59:13AM +0200, Salvatore Bonaccorso wrote:
> > Hi
> > 
> > On a system running 4.19.146-1 in Debian buster an issue got hit,
> > while the server was under some slight load, but it does not seem
> > easily reproducible, so asking if some more information can be
> > provided to track/narrow this down. On the console the following was
> > caught:
> 
> Worth checking git logs of fs/nfsd/nfs4state.c and
> fs/nfsd/nfs4callback.c.  It might be
> 2bbfed98a4d82ac4e7abfcd4eba40bddfc670b1d "nfsd: Fix races between
> nfsd4_cb_release() and nfsd4_shutdown_callback()" ?

That might be possible. As it was not possible to simply trigger the
issue, do you know if it is possible to simply reproduce the issue
fixed in the above?

2bbfed98a4d8 ("nfsd: Fix races between nfsd4_cb_release() and
nfsd4_shutdown_callback()") would be missing in the v4.19.y stable
series (as it was in 5.5-rc1 but not backported to other stable
versions).

Regards,
Salvatore