[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210311161030.5ed11805@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
Date: Thu, 11 Mar 2021 16:10:30 -0800
From: Jakub Kicinski <kuba@...nel.org>
To: Xie He <xie.he.0141@...il.com>
Cc: Martin Schiller <ms@....tdt.de>,
"David S. Miller" <davem@...emloft.net>,
Linux X25 <linux-x25@...r.kernel.org>,
Linux Kernel Network Developers <netdev@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH net] net: lapbether: Prevent racing when checking
whether the netif is running
On Thu, 11 Mar 2021 15:13:01 -0800 Xie He wrote:
> On Thu, Mar 11, 2021 at 2:52 PM Jakub Kicinski <kuba@...nel.org> wrote:
> >
> > Normally driver's ndo_stop() calls netif_tx_disable() which takes TX
> > locks, so unless your driver is lockless (LLTX) there should be no xmit
> > calls after that point.
>
> Do you mean I should call "netif_tx_disable" inside my "ndo_stop"
> function as a fix for the racing between "ndo_stop" and
> "ndo_start_xmit"?
>
> I can't call "netif_tx_disable" inside my "ndo_stop" function because
> "netif_tx_disable" will call "netif_tx_stop_queue", which causes
> another racing problem. Please see my recent commit f7d9d4854519
> ("net: lapbether: Remove netif_start_queue / netif_stop_queue")
And the "noqueue" queue is there because it's on top of hdlc_fr.c
somehow or some out of tree driver? Or do you install it manually?
Powered by blists - more mailing lists