Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <c8306569f880021bae853caf9f347e7c806ffaaa.camel@linux.ibm.com>
Subject: Re: [syzbot] possible deadlock in ovl_maybe_copy_up
From:   Mimi Zohar <zohar@linux.ibm.com>
To:     Amir Goldstein <amir73il@gmail.com>
Cc:     Hillf Danton <hdanton@sina.com>,
        syzbot <syzbot+c18f2f6a7b08c51e3025@syzkaller.appspotmail.com>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        syzkaller-bugs <syzkaller-bugs@googlegroups.com>
Date:   Tue, 22 Jun 2021 07:40:53 -0400
In-Reply-To: <CAOQ4uxjYJTG3p2ALXAWnAY+6Kmoi_L0=Z42okT6R+ovzQ7dQoQ@mail.gmail.com>
References: <000000000000c5b77105b4c3546e@google.com>
         <000000000000df47be05bf165394@google.com>
         <20210618040135.950-1-hdanton@sina.com>
         <23ba225593be391c384109af527bd0f3cb122a0d.camel@linux.ibm.com>
         <CAOQ4uxjYJTG3p2ALXAWnAY+6Kmoi_L0=Z42okT6R+ovzQ7dQoQ@mail.gmail.com>
Content-Type: text/plain; charset="ISO-8859-15"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

> > Should annotating the iint be limited to files on overlay filesystems?
> >
> 
> Not to overlay files specifically but to files on stacked fs,
> i.e. (inode->i_sb->s_stack_depth > 0)
> Assuming that this patch is tested(?), how come it did not hit the
> WARN_ON_ONCE(depth < 0... above?

Thanks, Amir!

As per the overlayfs comment, the depth can never be 0.  It sounds like
in this case we only want to annotate the iint mutex for regular files,
if the stacking depth is greater than 0, but less than the max depth.

(I'm still trying to reproduce the lockdep.)

Mimi