"Daniel P. Berrange" <berrange(a)redhat.com> wrote:
On Mon, Dec 10, 2007 at 10:13:03PM +0100, Jim Meyering wrote:
> Until now, there has been no coverage of libvirtd's config-processing code.
> In adding this, (and esp. in comparing actual/expected diagnostics),
> I noticed and fixed a trivial inconsistency: for a parameter expecting
> a string value, an invalid values, you'd get a different diagnostic for
> the four config parameters that are parsed via remoteConfigGetAuth than
> for all of the other config parameters (which go through checkType).
> This fixes that, too:
>
> Test libvirtd's config-processing code.
> And remove a minor diagnostic inconsistency.
> * tests/daemon-conf: New test.
> * tests/Makefile.am (TESTS_ENVIRONMENT): Prepend qemud/ to PATH,
> so we can invoke libvirtd without an absolute name.
> (test_scripts): Add daemon-conf.
> * qemud/qemud.c (remoteConfigGetAuth): Use checkType,
> rather than open-coding it with a different diagnostic.
ACK
Thanks.
Committed.