On Wed, Jul 22, 2009 at 04:19:48PM +0900, Nguyen Anh Quynh wrote:
On Wed, Jul 22, 2009 at 4:02 PM, Daniel Veillard<veillard@redhat.com> wrote: [...]
Then having no error handling here would make sense, but currently if you pass a wrong argument you would just silently assume flags == VIR_MEMORY_PHYSICAL, so the patch as-is is not correct.
Actually I dont think it is necessary to do sanity check inside the driver, as it is already checked outside, inside virDomainMemoryPeek() (I will fix the patch, as agreed above).
So do you really want to check it again here?
yes because a driver may support only a subset of the flags allowed by the API, it's a bit redundant but it's the status quo, and I think it makes sense. Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/