Re: NetBSD current and virtualbox .. not a good place right now

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

2009/3/5 Adam Hoka <adam.hoka%gmail.com@localhost>:
> George Michaelson wrote:
>
>>
>> OSX VirtualBox 2.1.4, NetBSD current as of a few days back. Not
>> playing nice.

Same here...

>>
>> Tried ACPI disable/enable
>>
>> Tried VT-X enable/disable
>>
>> Tried boot with/without ACPI/SMP at boot prompt.
>>
>> range of runtime shell/other coredumps, also kernel panics, also
>> crashes VirtualBox.

I haven't VirtualBox crash on me because of NetBSD (or anything else
for that matter) so far.

>>
>> Seems a shame that a freely available virtualizer like this can't be
>> used to run NetBSD guests.

Precisely.

>>
>> (OpenBSD, FreeBSD also have issues. I'm not trying to apportion blame)
>>
>> -G
>
> I have generally found Virtualbox to be incorrectly owkring with anything,
> but the few supported OSes. On Windows I usually use VirtualPC which is also 
> free.

That is quite a lame excuse I am afraid. I have found NetBSD to be
actually the worst as far as VirtualBox is concerned. I am presently
running FreeBSD 7.1, OpenBSD (current as from maybe six weeks, X
doesn't work) , Haiku (the BeOS replacement), various Windows systems,
some beta (W7), some historical (i.e. Server 3.51, WFWg etc.). All
this with reasonable success. [W7 build under VBox hosted on
OpenSolaris runs better than native Vista on the same hardware...].

The best I've had with NetBSD was 5.99.1 under VBox 2.1.2 a few weeks
ago - I managed to get to the stage to format the slice and got the
panic - but at least it booted the install kernel at the time.

I am not that bothered about VirtualBox, actually. I have been running
NetBSD under VMWare server 1,2 and ESX3i for some time with good
results; the last -current which worked for me was again 5.99.1 -
since that time I consistently get panic loading the installation
kernel. I thought I might have some problem with my local src tree, so
downloaded a few .iso files from releng, as recent as week ago - with
exactly the same result.

>
>

Chavdar Ivanov

-- 
----