Eric Blake wrote:
> On 05/23/2015 02:45 PM, Martin Kletzander wrote:
> > On Sat, May 23, 2015 at 08:05:23PM +0300, Roman Bogorodskiy wrote:
> >> Build with gcc 4.8 fails with:
>
> Arguably a bug in gcc; but since we can work around it without too much
> pain, we should.
>
> >>
> >> bhyve/bhyve_monitor.c: In function 'bhyveMonitorIO':
>
> >> bhyveMonitorIO(int watch, int kq, int events ATTRIBUTE_UNUSED, void
> >> *opaque)
> >> {
> >> - const struct timespec zerowait = {};
> >> + const struct timespec zerowait = { 0, 0 };
>
> Would also be sufficient to do 'zerowait = { 0 };' - any C compiler that
> warns about an initializer of { 0 } is broken, because that is THE
> idiomatic way to zero-initialize anything (scalar or structure)
> according to C99.
>
> >
> > You "need" to set at least minimum one field, all others will be set
> > to 0. But this is of course very right thing to do.
> >
> > ACK, structures shouldn't be initialized this way.
>
> Go ahead and push as you have it, though, with two members, since we
> know struct timespec has (at least) two members.Pushed, thanks! Roman Bogorodskiy
pgpPA5xOzLb6o.pgp
Description: PGP signature
-- libvir-list mailing list [email protected] https://www.redhat.com/mailman/listinfo/libvir-list
