On 6/10/20 3:51 AM, Erik Skultety wrote:
On Wed, Jun 10, 2020 at 12:16:50AM -0400, Laine Stump wrote:
> This was mostly boilerplate conversion, but in one case I needed to
> define several differently named char* to take the place of a single
> char *tmp that was re-used multiple times, and in another place there
> was a single char* that was used at the toplevel of the function, and
> then later used repeatedly inside a for loop, so I defined a new
> separate char* inside the loop.
>
> Signed-off-by: Laine Stump <laine(a)redhat.com>
> ---
>
> This should be applied on top of Dan's IPv6 NAT patch series (it was
> reviewing that series that showed me this file hadn't yet been
> converted).
...
> @@ -689,14 +678,12 @@ virNetworkDHCPDefParseXML(const char *networkName,
>
> if (server &&
> virSocketAddrParse(&inaddr, server, AF_UNSPEC) < 0) {
> - VIR_FREE(file);
> - VIR_FREE(server);
> goto cleanup;
> }
>
> def->bootfile = file;
> + file = NULL;
g_steal_pointer would do as well
Reviewed-by: Erik Skultety <eskultet(a)redhat.com>
Well, Duh! Where was my brain while I was doing that mindless conversion??
This made me realized there's actually several places with the x = y; y
= NULL; pattern associated with no-autofree'd pointers. If it's okay
with you, I'll squash the following into the patch before I push it (or,
if you'd prefer I can push it separately):
(Thanks for the review, BTW)