Re: wm(4) interrupt timeouts under Xen (as a dom0)

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Thu, Oct 08, 2020 at 12:28:01AM -0500, Dustin Marquess wrote:
> Just when I thought things were finally working...
> 
> A XEN3_DOM0 kernel built today from -current:
> 
> NetBSD apu2.fdf.net 9.99.73 NetBSD 9.99.73 (XEN3_DOM0) #0: Thu Oct  8
> 00:02:41 CDT 2020
> 
> Running under Xen 4.13 (xenkernel413) on a PC Engines apu4d4, the
> onboard Intel NIC is really flakey:
> 
> [   125.039792] wm0: device timeout (lost interrupt)
> [   173.015794] wm0: device timeout (lost interrupt)
> [   192.006321] wm0: device timeout (lost interrupt)
> [   202.001142] wm0: device timeout (lost interrupt)
> [   208.997794] wm0: device timeout (lost interrupt)
> [   241.981164] wm0: device timeout (lost interrupt)
> [   257.973293] wm0: device timeout (lost interrupt)
> [   331.966900] wm0: device timeout (lost interrupt)
> [   347.966998] wm0: device timeout (lost interrupt)
> 
> Booting into a normal GENERIC kernel built from the exact same source
> (without Xen), it works fine.
> 
> I'm not using any special Xen command line options.  Comparing 'xl
> dmesg' to a working L*x Xen 4.12 box, I don't see any obvious
> differences between the two.

You should compare the dmesg between GENERIC and XEN3_dOM0 kernels

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--