On Thu, Feb 25, 2016 at 11:47:03AM -0500, John Ferlan wrote:
On 02/25/2016 09:25 AM, Daniel P. Berrange wrote:
int virPolkitCheckAuth(const char *actionid, pid_t pid, unsigned long long startTime, @@ -31,4 +33,7 @@ int virPolkitCheckAuth(const char *actionid, const char **details, bool allowInteraction);
+void virPolkitAgentDestroy(virCommandPtr cmd); +virCommandPtr virPolkitAgentCreate(void);
Rather than exposing use of virCommand in the API, I'd suggest you create a
typedef struct virPolkitAgent virPolkitAgent; typedef virPolkitAgent *virPolkitAgentPtr;
okidoke... Funny I had done it this way at some point, but when virCommandPtr was the only thing in the structure, I just opted to use virCommandPtr directly. Anyway, the following is now defined:
It is just nice from the POV of isolating callers from impl detail. It means we can switch to use direct DBus calls in the future without changing any callers.
Forgot to note in my response to Martin that virpolkittest.c doesn't need a change here since virpolkit.h now includes vircommand.h. Same of course for virsh.h.
Should I post a v4 of patches 2 & 3?
Yeah, it'd be good to send v4 Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|