Re: [libvirt] Implementation deficiency in virInitctlSetRunLevel
On 12/18/2013 02:06 PM, Reco wrote:
Hi.
On Wed, 18 Dec 2013 13:42:25 -0700
Eric Blake <eblake(a)redhat.com> wrote:
> git send-email --in-reply-to='messageid(a)example.com'
>
> for the correct value of messageid. But generally posting a new thread
> is better than threading your v2 to your v1 anyways, as top-level
> threads are easier to spot than a patch nested deep in reply to another
> message.
Thanks. Should I resend the patch the proper way again?
Reco
We also prefer that patches have authorship attributed to a legal name,
rather than a nickname. But rather than resend your patch, I think we
should pay attention to Dan's advice - that any solution we come up with
that doesn't involve forking a child process, then resetting that child
into the correct namespace, is prone to races and therefore does not
solve the problem at hand (namely, that symlinks created in the guest
namespace must NOT be followed in the parent in host namespace).
--
Eric Blake eblake redhat com +1-919-301-3266
Libvirt virtualization library http://libvirt.org
Attachments:
- signature.asc (application/pgp-signature — 621 bytes)