netdev - Re: netconsole deadlock with virtnet

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20201124082035.3e658fa4@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
Date:   Tue, 24 Nov 2020 08:20:35 -0800
From:   Jakub Kicinski <kuba@...nel.org>
To:     Jason Wang <jasowang@...hat.com>
Cc:     Steven Rostedt <rostedt@...dmis.org>,
        Leon Romanovsky <leon@...nel.org>,
        Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
        "Michael S. Tsirkin" <mst@...hat.com>,
        Petr Mladek <pmladek@...e.com>,
        John Ogness <john.ogness@...utronix.de>,
        virtualization@...ts.linux-foundation.org,
        Amit Shah <amit@...nel.org>, Itay Aveksis <itayav@...dia.com>,
        Ran Rozenstein <ranro@...dia.com>,
        netdev <netdev@...r.kernel.org>
Subject: Re: netconsole deadlock with virtnet

On Tue, 24 Nov 2020 11:22:03 +0800 Jason Wang wrote:
> >> Perhaps you need the trylock in virtnet_poll_tx()?  
> > That could work. Best if we used normal lock if !!budget, and trylock
> > when budget is 0. But maybe that's too hairy.  
> 
> If we use trylock, we probably lose(or delay) tx notification that may 
> have side effects to the stack.

That's why I said only trylock with budget == 0. Only netpoll calls with
budget == 0, AFAIK.

> > I'm assuming all this trickiness comes from virtqueue_get_buf() needing
> > locking vs the TX path? It's pretty unusual for the completion path to
> > need locking vs xmit path.  
> 
> Two reasons for doing this:
> 
> 1) For some historical reason, we try to free transmitted tx packets in 
> xmit (see free_old_xmit_skbs() in start_xmit()), we can probably remove 
> this if we remove the non tx interrupt mode.
> 2) virtio core requires virtqueue_get_buf() to be synchronized with 
> virtqueue_add(), we probably can solve this but it requires some non 
> trivial refactoring in the virtio core
> 
> Btw, have a quick search, there are several other drivers that uses tx 
> lock in the tx NAPI.

Unless they do:

	netdev->priv_flags |= IFF_DISABLE_NETPOLL;

they are all broken.