Re: virtio panic during reboot (NetBSD 9)

[Chavdar Ivanov <ci4ic4%gmail.com@localhost>]

On Tue, 11 Aug 2020 at 14:59, Thomas Klausner <tk%giga.or.at@localhost> wrote:
>
> Hi!
>
> With a NetBSD 9 release in a virtual server (not sure what technology,
> dmesg available on request), I saw a panic during reboot:

That would be interesting - to see where vioscsi actually attaches.
I've always wondered if it would be difficult to modify the vioscsi
driver to attach under VirtualBox - as it is recognized at the moment
as

Qumranet product 1048 (SCSI mass storage, revision 0x01) at pci0 dev
15 function 0 not configured

I appreciate that VirtualBox documentation states 'Currently the
virtio-scsi controller is experimental'.

vioif attaches just fine and as far as I can see, works the same.

>
> netbsd: [ 425.9816243] uvm_fault(0xffffceeb5b6b2a20, 0x0, 1) -> e
> netbsd: [ 425.9816243] fatal page fault in supervisor mode
> netbsd: [ 425.9816243] trap type 6 code 0 rip 0xffffffff80240fb0 cs 0x8 rflags 0x10246 cr2 0x41 ilevel 0 rsp 0xffffda80b0
> 2fed70
> netbsd: [ 425.9816243] curlwp 0xffffceeb58a62080 pid 72.1 lowest kstack 0xffffda80b02fc2c0
> netbsd: [ 425.9816243] Skipping crash dump on recursive panic
> netbsd: [ 425.9816243] panic: trap
> netbsd: [ 425.9893443] cpu0: Begin traceback...
> netbsd: [ 425.9893443] vpanic() at netbsd:vpanic+0x160
> netbsd: [ 425.9893443] snprintf() at netbsd:snprintf
> netbsd: [ 425.9893443] startlwp() at netbsd:startlwp
> netbsd: [ 425.9893443] alltraps() at netbsd:alltraps+0xbb
> netbsd: [ 425.9992299] virtio_pci_free_interrupts() at netbsd:virtio_pci_free_interrupts+0x3c
> netbsd: [ 426.0220926] virtio_child_detach() at netbsd:virtio_child_detach+0x61
> netbsd: [ 426.0220926] vioscsi_detach() at netbsd:vioscsi_detach+0x9d
> netbsd: [ 426.0292585] config_detach() at netbsd:config_detach+0x85
> netbsd: [ 426.0292585] config_detach_all() at netbsd:config_detach_all+0x9f
> netbsd: [ 426.0292585] cpu_reboot() at netbsd:cpu_reboot+0x137
> netbsd: [ 426.0292585] sys_reboot() at netbsd:sys_reboot+0x75
> netbsd: [ 426.0292585] syscall() at netbsd:syscall+0x157
> netbsd: [ 426.0392749] --- syscall (number 208) ---
> netbsd: [ 426.0392749] 7af270242f5a:
> netbsd: [ 426.0392749] cpu0: End traceback...
> netbsd: [ 426.0392749] rebooting...
>
> Is this a known problem or should I file a bug report?
>  Thomas

Chavdar



-- 
----