On Thu, 16 Jan 2020 13:52:21 +0100
Thomas Huth <th...@redhat.com> wrote:
> On 16/01/2020 13.37, Cornelia Huck wrote:
> > If the kernel irqchip has been disabled, we don't want the
> > {add,release}_adapter_routes routines to call any kvm_irqchip_*
> > interfaces, as they may rely on an irqchip actually having been
> > created. Just take a quick exit in that case instead.
> >
> > Fixes: d426d9fba8ea ("s390x/virtio-ccw: wire up irq routing and irqfds")
> > Signed-off-by: Cornelia Huck <coh...@redhat.com>
> > ---
> >
> > Without this patch, QEMU with kernel_irqchip=off will crash in
> > kvm_irqchip_release_virq(), so alternatively, we could add a check
> > there. kvm_irqchip_add_adapter_route() is actually fine.
> >
> > ---
> > hw/intc/s390_flic_kvm.c | 8 ++++++++
> > 1 file changed, 8 insertions(+)
> >
> > diff --git a/hw/intc/s390_flic_kvm.c b/hw/intc/s390_flic_kvm.c
> > index dddd33ea61c8..44b7960ebcc8 100644
> > --- a/hw/intc/s390_flic_kvm.c
> > +++ b/hw/intc/s390_flic_kvm.c
> > @@ -331,6 +331,10 @@ static int kvm_s390_add_adapter_routes(S390FLICState
> > *fs,
> > int ret, i;
> > uint64_t ind_offset = routes->adapter.ind_offset;
> >
> > + if (!kvm_gsi_routing_enabled()) {
> > + return -ENOSYS;
> > + }
>
> As you wrote, this check is not really necessary since it is already
> done in kvm_irqchip_add_adapter_route() ...
I do think it is cleaner, though.
>
> > for (i = 0; i < routes->num_routes; i++) {
> > ret = kvm_irqchip_add_adapter_route(kvm_state, &routes->adapter);
> > if (ret < 0) {
>
> ... so I wonder if it would be simply best to set
>
> routes->gsi[i] = -1;
>
> before the "goto" instead to make sure that
> kvm_s390_release_adapter_routes() does not try to clean it up? That
> would also fix a potential crash in case kvm_irqchip_add_adapter_route()
> returned an error code in case of a different problem, I think.
I think we should pre-initialize gsi[] to -1 instead, just to be on the
safe side.
