On 2/27/19 10:32 AM, John Ferlan wrote:
>
> Unfortunately, libvirt versions between 1.2.12 and 5.0.0 will
> silently ignore the new flag, rather than diagnosing that they
> don't support it; but at least silent lack of snapshots from
> an older server is not a security hole.
>
> Signed-off-by: Eric Blake <eblake(a)redhat.com>
> ---
> include/libvirt/libvirt-domain.h | 1 +
> src/conf/domain_conf.c | 13 ++++++++-----
> src/libvirt-domain.c | 5 +++++
> 3 files changed, 14 insertions(+), 5 deletions(-)
>
[...]
> diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c
> index 072b92b717..2691698bd5 100644
> --- a/src/libvirt-domain.c
> +++ b/src/libvirt-domain.c
> @@ -2570,6 +2570,11 @@ virDomainGetControlInfo(virDomainPtr domain,
> * XML might not validate against the schema, so it is mainly for
> * internal use.
> *
> + * If @flags contains VIR_DOMAIN_XML_SNAPSHOTS, the XML will include
Should we even try to say that "and supported by the target libvirt
system with the appropriate version of the software installed" ;-)... I
know implied somewhat - but perhaps notable in this (and future) cases
because of the issue mentioned in the commit message that outward facing
docs consumers may never read.
Maybe, since this is indeed enough of a break from the usual norms of
rejecting unknown flags (at least for a couple of years) to be worth it.
The upcoming VIR_DOMAIN_XML_CHECKPOINTS will have the same wording, of
course.
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization:
qemu.org |
libvirt.org