[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [libvirt] [PATCH 5/6] Taint domains configured with cpu mode=host-passthrough



On 01/06/2012 08:04 AM, Jiri Denemark wrote:
> ---
>  src/conf/domain_conf.c |    3 ++-
>  src/conf/domain_conf.h |    1 +
>  src/qemu/qemu_domain.c |    3 +++
>  3 files changed, 6 insertions(+), 1 deletions(-)

Aargh - 'git grep taint docs/*.in' turned up nothing, so we don't have
any documentation of tainting.  That would be a nice thing to add before
this patch, along with mentioning each category of taint (including this
new category).

But I agree that using host-passthrough is a reason for taint - if
current qemu does not fully model the host, then upgrading to a newer
version of qemu that adds new modeling features will pass through new
capabilities to the guest, and that is a guest-visible ABI change (in
particular, it might cause a Windows reactivation).  Since we try to
promise stability, but the passthrough puts us at the mercy of what qemu
upgrades implement, declaring taint seems like an appropriate reaction
to warn the user of the potential for problems.

Maybe you should include something like the above paragraph as
justification in your commit message.

Your code additions look sane, but I'd feel more comfortable with docs
as well.

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org

Attachment: signature.asc
Description: OpenPGP digital signature


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]