Received-SPF: pass (google.com: domain of linux-kernel+bounces-164827-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
From: Hillf Danton <hdanton@sina.com>
To: Mike Christie <michael.christie@oracle.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>,
	Edward Adam Davis <eadavis@qq.com>,
	syzbot+98edc2df894917b3431f@syzkaller.appspotmail.com,
	jasowang@redhat.com,
	kvm@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	netdev@vger.kernel.org,
	syzkaller-bugs@googlegroups.com,
	virtualization@lists.linux.dev
Subject: Re: [PATCH next] vhost_task: after freeing vhost_task it should not be accessed in vhost_task_fn
Date: Wed,  1 May 2024 08:15:44 +0800
Message-Id: <20240501001544.1606-1-hdanton@sina.com>
In-Reply-To: <b959b82a-510f-45c0-9e06-acf526c2f4a1@oracle.com>
References: 
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

On Tue, Apr 30, 2024 at 11:23:04AM -0500, Mike Christie wrote:
> On 4/30/24 8:05 AM, Edward Adam Davis wrote:
> >  static int vhost_task_fn(void *data)
> >  {
> >  	struct vhost_task *vtsk = data;
> > @@ -51,7 +51,7 @@ static int vhost_task_fn(void *data)
> >  			schedule();
> >  	}
> >  
> > -	mutex_lock(&vtsk->exit_mutex);
> > +	mutex_lock(&exit_mutex);
> >  	/*
> >  	 * If a vhost_task_stop and SIGKILL race, we can ignore the SIGKILL.
> >  	 * When the vhost layer has called vhost_task_stop it's already stopped
> > @@ -62,7 +62,7 @@ static int vhost_task_fn(void *data)
> >  		vtsk->handle_sigkill(vtsk->data);
> >  	}
> >  	complete(&vtsk->exited);
> > -	mutex_unlock(&vtsk->exit_mutex);
> > +	mutex_unlock(&exit_mutex);
> >  
> 
> Edward, thanks for the patch. I think though I just needed to swap the
> order of the calls above.
> 
> Instead of:
> 
> complete(&vtsk->exited);
> mutex_unlock(&vtsk->exit_mutex);
> 
> it should have been:
> 
> mutex_unlock(&vtsk->exit_mutex);
> complete(&vtsk->exited);

JFYI Edward did it [1]

[1] https://lore.kernel.org/lkml/tencent_546DA49414E876EEBECF2C78D26D242EE50A@qq.com/
> 
> If my analysis is correct, then Michael do you want me to resubmit a
> patch on top of your vhost branch or resubmit the entire patchset?