On 10/08/2015 02:52 PM, Laine Stump wrote:
On 10/07/2015 12:50 PM, Daniel P. Berrange wrote:
> else
> fprintf(stdout, "%s (%s)\n",
> - _(failMessages[level]), hint);
> + _(failMessages[level]), msg);
> + free(msg);
make syntax-check doesn't complain, but should we be consistently
using VIR_FREE()? (actually it looks like there are a lot of these
already existing, although VIR_FREE() has at least one use in
virt-host-validate-common.c dating all the way back to the original
commit)
I spoke too soon - when I fixed the fclose() occurences, *then*
syntax-check complaining about free() (and I also saw that all of the
"already existing" occurrences were actually newly added in this same patch.