On Tue, Sep 04, 2018 at 05:18:48PM +0200, Andrew Jones wrote:
> On Tue, Sep 04, 2018 at 03:22:36PM +0200, Igor Mammedov wrote:
> > commit
> > (5cdc9b76e3 vl.c: Remove dead assignment)
> > removed sockets calculation when 'sockets' weren't provided on CLI
> > since there wasn't any users for it back then. Exiting checks
> > are neither reachable
> > } else if (sockets * cores * threads < cpus) {
> > or nor triggable
> > if (sockets * cores * threads > max_cpus)
> > so we weren't noticing wrong topology since then, since users
> > recalculate sockets adhoc on their own.
> >
> > However with deprecation check it becomes noticable, for example
> > -smp 2
> > will start printing warning:
> > "warning: Invalid CPU topology deprecated: sockets (1) * cores (1) *
> > threads (1) != maxcpus (2)"
> > calculating sockets if they weren't specified.
> >
> > Fix it by returning back sockets calculation if it's omited on CLI.
> >
> > Signed-off-by: Igor Mammedov <imamm...@redhat.com>
> > ---
> > vl.c | 5 ++++-
> > 1 file changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/vl.c b/vl.c
> > index 7fd700e..333d638 100644
> > --- a/vl.c
> > +++ b/vl.c
> > @@ -1210,11 +1210,14 @@ static void smp_parse(QemuOpts *opts)
> >
> > /* compute missing values, prefer sockets over cores over threads
> > */
> > if (cpus == 0 || sockets == 0) {
> > - sockets = sockets > 0 ? sockets : 1;
> > cores = cores > 0 ? cores : 1;
> > threads = threads > 0 ? threads : 1;
> > if (cpus == 0) {
> > + sockets = sockets > 0 ? sockets : 1;
> > cpus = cores * threads * sockets;
> > + } else {
> > + max_cpus = qemu_opt_get_number(opts, "maxcpus", cpus);
> > + sockets = !sockets ? max_cpus / (cores * threads) :
> > sockets;
Why the !sockets check here? If we have reached this statement
we know that (cpus == 0 || sockets == 0) is true and
(cpus == 0) is false, so sockets is guaranteed to be 0.
Sorry for not spotting this earlier.
> > }
> > } else if (cores == 0) {
> > threads = threads > 0 ? threads : 1;
> > --
> > 2.7.4
> >
> >
>
> Reviewed-by: Andrew Jones <drjo...@redhat.com>
--
Eduardo
--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
