Quoting Janosch Frank (2022-07-22 10:30:46)
> On 7/22/22 09:20, Nico Boehr wrote:
> > Under PV, SIGP ecall requires some special handling by the hypervisor
> > when the receiving CPU is in enabled wait. Hence, we should have
> > coverage for the various SIGP call orders when the receiving CPU is in
> > enabled wait.
>
> When the SIGP interpretation facility is in use a SIGP external call to
> a waiting CPU will result in an exit of the calling cpu. For non-pv
> guests it's a code 56 (partial execution) exit otherwise its a code 108
> (secure instruction notification) exit. Those exits are handled
> differently from a normal SIGP instruction intercept that happens
> without interpretation and hence need to be tested.
Changed.
> >
> > The ecall test currently fails under PV due to a KVM bug under
> > investigation.
>
> That shouldn't be true anymore
Yeah, it's not yet in mainline, but is soon gonna be. I will remove this.
> >
> > Signed-off-by: Nico Boehr <nrb@xxxxxxxxxxxxx>
> > ---
> > s390x/smp.c | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++++
> > 1 file changed, 75 insertions(+)
> >
> > diff --git a/s390x/smp.c b/s390x/smp.c
> > index 683b0e618a48..eed7aa3564de 100644
> > --- a/s390x/smp.c
> > +++ b/s390x/smp.c
> > @@ -347,6 +347,80 @@ static void test_calls(void)
> > }
> > }
> >
> > +static void call_in_wait_ext_int_fixup(struct stack_frame_int *stack)
> > +{
> > + /* leave wait after returning */
>
> Clear wait bit so we don't immediately wait again after the fixup
Changed.
>
> > + lowcore.ext_old_psw.mask &= ~PSW_MASK_WAIT;
> > +
> > + stack->crs[0] &= ~current_sigp_call_case->cr0_bit;
>
> You need a mask but have a bit, no?
>
> ~BIT(current_sigp_call_case->cr0_bit)
Oopsie, thanks, good find.
This reminds me the ctl_clear_bit() I added in call_in_wait_received() is completely useless, since I handle it here. So, I will remove it there as well.
[...]
> > +static void test_calls_in_wait(void)
> > +{
> > + int i;
> > + struct psw psw;
> > +
> > + report_prefix_push("psw wait");
> > + for (i = 0; i < ARRAY_SIZE(cases_sigp_call); i++) {
> > + current_sigp_call_case = &cases_sigp_call[i];
> > +
> > + report_prefix_push(current_sigp_call_case->name);
> > + if (!current_sigp_call_case->supports_pv && uv_os_is_guest()) {
> > + report_skip("Not supported under PV");
> > + report_prefix_pop();
> > + continue;
> > + }
> > +
> /* Let the secondary CPU setup the external mask and the external
> interrupt cleanup function */
Changed.
