vkernel & GSoC, some questions

Jordan Gordeev jgordeev at dir.bg
Tue Mar 18 19:29:11 UTC 2008


Matthew Dillon wrote:

>:Matt,
>:
>:We use VMWare Server at work. It does not have the same nice image management interface and/or video capture as commercial counterparts. However, it is is free and testing on it helps us out big time. We never concluded whether it maked sense to pay for VMWare licenses, instead of using free shell scripts legally available for free.
>:
>:I have used UML for development in the past. I even used bochs once to debug a boot loader. All nice tools. Beats real hardware for me.
>:
>:Xen and KVM are significantly slower than commercial products due to hardware switching. There is a GPLed product that works about as fast as VMWare's BT - VirtualBox by innotek. Sun recently scooped them up.
>
>    You've tried them all pretty much.  VirtualBox is the only one that's
>    actually open-source (well, bochs is too but bochs is a full simulator
>    and unusably slow).  VB is very nicely documented too, though I get a 
>    headache trying to read all that C++.   I know several people using
>    VB on linux host systems.
>
>    The reason I ask is simply because I am trying to point out *the* major
>    issue for open-source systems when it comes to hardware virtualization,
>    that being the availability of an open-source solution.  I have never
>    trusted VMWare to maintain compatibility with any longevity.  VB has
>    a better chance and even though it is hard to gauge what Sun will do
>    with it, I'd rather it be Sun then MS.
>
>:Don't you use something like VMWare for development and debugging?
>
>    We use vkernel's for development and debugging.  Pretty much everything
>    except hardware device driver development can be done using a vkernel
>    and being able to gdb the kernel binary (gdb, not kgdb) on a whim is
>    rather nice.  Kinda hard to beat a 5-second reboot, it makes the
>    engineering test cycle almost as short as hitting ^C and re-running a
>    program.  I don't even bother to do a clean halt of the vkernel most of
>    time, I just pop it into it's DDB with ctl-\ and then kill it with ^C,
>    then restart.
>
>    One interesting side-effect of having a vkernel so easily accessible
>    is that it opens up kernel development to normal programmers.  More
>    DragonFly developers have been dipping their fingers into the kernel
>    code in the last 6 months then in all the time before then.  That alone
>    justifies the time spent doing it.  Except for hardware device driver
>    development, the agonizing engineering cycle for kernel development
>    is completely gone now.
>
>						-Matt
>
>:In production, we don't use any of these products - too slow and too much RAM would be required.
>:
>:Sincerely,
>:
>:Igor Shmukler, http://www.elusiva.com
>
>
>  
>
I have thought of the vkernel primarily as an aid to kernel development 
(where performance is not a prime concern), not as a virtualisation 
solution that will compete with Xen and VMWare. It's difficult to 
compete with thousands of men-hours paid by corporate funding.

So far nobody has expressed interest in vkernels as a tool for kernel 
development. And I got the general impression that I've proposed 
something stupid and useless.

I can see the value in having a virtualisation solution in the base 
system that allows one to easily try out kernel development. I know it 
would have helped _me_ get into the forbidden land.

I'm augmenting my project idea by dropping the AMD64 version, and adding 
"performance work to reduce overhead for disk I/O and networking". I'll 
also make sure that no kernel function renaming will be needed, through 
some linking magic.


More information about the freebsd-hackers mailing list