On 3/31/20 7:27 PM, David Hildenbrand wrote:
> On 31.03.20 11:28, Janosch Frank wrote:
>> On 3/31/20 11:07 AM, David Hildenbrand wrote:
>>> On 24.03.20 09:12, Janosch Frank wrote:
>>>> Local interrupts (external and emergency call) should be cleared after
>>>> any cpu reset.
>>>>
>>>> Signed-off-by: Janosch Frank <frankja@xxxxxxxxxxxxx>
>>>> ---
>>>> s390x/smp.c | 22 ++++++++++++++++++++++
>>>> 1 file changed, 22 insertions(+)
>>>>
>>>> diff --git a/s390x/smp.c b/s390x/smp.c
>>>> index 8a6cd1d8b17d76c6..a8e3dd7aac0c788c 100644
>>>> --- a/s390x/smp.c
>>>> +++ b/s390x/smp.c
>>>> @@ -243,6 +243,20 @@ static void test_reset_initial(void)
>>>> report_prefix_pop();
>>>> }
>>>>
>>>> +static void test_local_ints(void)
>>>> +{
>>>> + unsigned long mask;
>>>> +
>>>> + expect_ext_int();
>>>> + /* Open masks for ecall and emcall */
>>>> + ctl_set_bit(0, 13);
>>>> + ctl_set_bit(0, 14);
>>>> + mask = extract_psw_mask();
>>>> + mask |= PSW_MASK_EXT;
>>>> + load_psw_mask(mask);
>>>> + set_flag(1);
>>>> +}
>>>> +
>>>> static void test_reset(void)
>>>> {
>>>> struct psw psw;
>>>> @@ -251,10 +265,18 @@ static void test_reset(void)
>>>> psw.addr = (unsigned long)test_func;
>>>>
>>>> report_prefix_push("cpu reset");
>>>> + sigp(1, SIGP_EMERGENCY_SIGNAL, 0, NULL);
>>>> + sigp(1, SIGP_EXTERNAL_CALL, 0, NULL);
>>>> smp_cpu_start(1, psw);
>>>>
>>>> sigp_retry(1, SIGP_CPU_RESET, 0, NULL);
>>>> report(smp_cpu_stopped(1), "cpu stopped");
>>>> +
>>>> + set_flag(0);
>>>> + psw.addr = (unsigned long)test_local_ints;
>>>> + smp_cpu_start(1, psw);
>>>> + wait_for_flag();
>>>> + report(true, "local interrupts cleared");
>>>
>>>
>>> How can you be sure they were actually cleared/delivered?
>>>
>> Because cpu 1 would get a ext int it didn't expect and would do a
>> report_abort() as pecified in lib/s390x/interrupt.c
>
> But what if it *didn't* get the interrupts delivered.
Well, the target is still looping until the reset initial test is
executed and from personal experience I can say that this test bites
rather fast.
We could execute an instruction with a mandatory exit which will prompt
KVM to inject any remaining IRQs before setting the flag to 1.
Unfortunately we do not have a polling instruction for non-IO interrupts
at the moment to verify this test.
>
> Then cpu 1 will simply (test_local_ints()) unlock interrupts, load the
> psw mask, set the flag and be done with it. What am I missing? How can
> you be sure the interrupts on cpu 1 were actually delivered?
>
Attachment:
signature.asc
Description: OpenPGP digital signature
