On Thu, May 25, 2023 at 12:45:23AM +0000, Ping-Ke Shih wrote:
>
>
> > -----Original Message-----
> > From: Sascha Hauer <s.hauer@xxxxxxxxxxxxxx>
> > Sent: Wednesday, May 24, 2023 6:40 PM
> > To: linux-wireless <linux-wireless@xxxxxxxxxxxxxxx>
> > Cc: Hans Ulli Kroll <linux@xxxxxxxxxxxxx>; Larry Finger <Larry.Finger@xxxxxxxxxxxx>; Ping-Ke Shih
> > <pkshih@xxxxxxxxxxx>; Tim K <tpkuester@xxxxxxxxx>; Alex G . <mr.nuke.me@xxxxxxxxx>; Nick Morrow
> > <morrownr@xxxxxxxxx>; Viktor Petrenko <g0000ga@xxxxxxxxx>; Andreas Henriksson <andreas@xxxxxxxx>;
> > ValdikSS <iam@xxxxxxxxxxxxxxx>; kernel@xxxxxxxxxxxxxx; petter@xxxxxxxxxx; Sascha Hauer
> > <s.hauer@xxxxxxxxxxxxxx>
> > Subject: [PATCH] wifi: rtw88: usb: silence log flooding error message
> >
> > When receiving more rx packets than the kernel can handle the driver
> > drops the packets and issues an error message.
>
> The workqueue rtw88_usb is using is:
>
> rtwusb->rxwq = create_singlethread_workqueue("rtw88_usb: rx wq");
>
> Have you tried workqueue with flags WQ_UNBOUND and WQ_HIGHPRI? Like,
>
> rtwusb->rxwq = alloc_workqueue("rtw88_usb: rx wq", WQ_UNBOUND | WQ_HIGHPRI, 0);
> or
> rtwusb->rxwq = alloc_ordered_workqueue("rtw88_usb: rx wq", WQ_HIGHPRI);
>
> Then, driver get more time to process RX, so it could ease flooding messages.
No, I haven't tried this. Regardless of that, I think it still makes
sense to rate limit the messages. There will always be a slower system
that can't cope with the number of packets even with a higher priority
workqueue.
Sascha
--
Pengutronix e.K. | |
Steuerwalder Str. 21 | http://www.pengutronix.de/ |
31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
