Re: Fwd: 10-BETA : some network issues

[Taylor R Campbell <campbell+netbsd-tech-net%mumble.net@localhost>]

> Date: Thu, 29 Dec 2022 10:06:55 +0100
> From: BERTRAND Joël <joel.bertrand%systella.fr@localhost>
> 
> >> 	First constatations: 10-BETA reboots when it tries to configure agr0
> >> (it doesn't panic, only reboots).
> > 
> > 4. Do you have dmesg from the previous boot when this happens, or
> >    serial console?
> 
> 	No, sorry.
> [...]
> > 5. Can you use crash(8) to find the kernel stack trace of the process
> >    that's eating CPU?
> 
> 	I will try, but only when I have to reboot this system (when this
> server reboots, I have to restart all diskless workstations on lan1).

Can you try to reproduce these problems (and the original one) in a VM
with roughly the same configuration?

Having a test platform to run experiments on in a complex
configuration like this will generally make everything less painful --
also for testing configuration changes, not just debugging NetBSD!
Something like this, perhaps:

qemu-img create -f qcow2 hd0.img 8g

qemu-system-x86_64 \
        -machine pc,accel=nvmm \
        -smp 2 \
        -m 1g \
        -drive if=none,file=NetBSD-10.0_BETA-amd64-install.img,id=boot,format=raw \
        -device virtio-blk-pci,drive=boot \
        -drive if=none,file=hd0.qcow2,id=hd0,format=qcow2 \
        -device virtio-blk-pci,drive=hd0 \
        -netdev tap,id=wm0,ifname=tap,script=ifup,downscript=ifdown \
        -device e1000,netdev=wm0 \
        -netdev tap,id=wm1,ifname=tap,script=ifup,downscript=ifdown \
        -device e1000,netdev=wm1 \
        -netdev tap,id=wm2,ifname=tap,script=ifup,downscript=ifdown \
        -device e1000,netdev=wm2 \
        -netdev tap,id=wm3,ifname=tap,script=ifup,downscript=ifdown \
        -device e1000,netdev=wm3 \
        -netdev tap,id=wm4,ifname=tap,script=ifup,downscript=ifdown \
        -device e1000,netdev=wm4

Create scripts `ifup' and `ifdown' that run like `ifup/down tapN' and
do `ifconfig tapN up; brconfig bridge42 add tapN', where bridge42 is a
bridge representing any LANs they need to be on (different scripts
with bridges, if wm0/1/2/3/4 aren't on the same LAN).

After first boot to install NetBSD, remove the boot drive -- the first
`-drive' and `-device' options, with `boot' in them.

You can then put `consdev=com0' in /boot.cfg in the VM's file system,
and run qemu on subsequent boots with `-nographic' to get serial
console.  (To exit qemu when it's running in `-nographic' mode, type
`C-a x'.)