Re: kvmtool tree (Was: Re: [patch] config: fix make kvmconfig)

[Paolo Bonzini <pbonzini@redhat.com>]

Il 12/02/2013 10:52, Ingo Molnar ha scritto:
> Check the list I gave (unmodified):
> 
> "- Pekka listed new virtio drivers that were done via tools/kvm. 

vhost-scsi got in first in tools/kvm, but out-of-tree patches had
existed for QEMU for more than a year.  It was developed with QEMU.

>  - Pekka listed ARM KVM support which was written/prototyped
>    using tools/kvm.

Again, I think both QEMU and tools/kvm were used, with QEMU probably
coming first.  However, I'm not 100% sure.

>  - There's over a dozen bugfixes in your kernel which were found
>    via syscall fuzzing built into tools/kvm. (I can dig them all
>    out if you want.)

This sounds like something that runs in a guest, not in the host.  As
such it need not be tied to tools/kvm.  If there is a host component
(host C code), I'd be interested.  I know that Fujitsu guys have patches
for panic notification from the guest kernel, for example.

But if there is a host component to it, then please go back to the
original purpose of tools/kvm.  Drop all the QCOW, VNC, SeaBIOS
nonsense.  Drop the userspace networking stack (tap is enough).  Make
the tool into a library that supports nothing but virtio-{blk,net,9p}
and VFIO.  Build _many_ narrow-focused tools around that library.

>> "sparse" is useful for kernel development. "git" is useful for 
>> kernel development. "xterm" is useful for kernel development.
> 
> That argument is silly beyond belief. Read this list:
> 
>   - tools/kvm
>   - sparse
>   - git
>   - xterm
>   - perf
> 
> Which two tools in this list:
> 
>  - Use and focus on Linux specific system calls to provide Linux
>    specific functionality?
> 
>  - Are never - and will conceivably never - run on any kernel
>    which is not extremely Linux compatible?

QEMU runs KVM on Solaris, FWIW.

Paolo