Re: [PATCH v3 5/9] KVM: s390: selftest: memop: Move testlist into main

[Janis Schoetterl-Glausch]

On Tue, 2022-11-22 at 08:52 +0100, Thomas Huth wrote:
> On 17/11/2022 23.17, Janis Schoetterl-Glausch wrote:
> > This allows checking if the necessary requirements for a test case are
> > met via an arbitrary expression. In particular, it is easy to check if
> > certain bits are set in the memop extension capability.
> > 
> > Signed-off-by: Janis Schoetterl-Glausch <scgl@xxxxxxxxxxxxx>
> > ---
> >   tools/testing/selftests/kvm/s390x/memop.c | 132 +++++++++++-----------
> >   1 file changed, 66 insertions(+), 66 deletions(-)
> > 
> > diff --git a/tools/testing/selftests/kvm/s390x/memop.c b/tools/testing/selftests/kvm/s390x/memop.c
> > index 286185a59238..10f34c629cac 100644
> > --- a/tools/testing/selftests/kvm/s390x/memop.c
> > +++ b/tools/testing/selftests/kvm/s390x/memop.c
> > @@ -690,87 +690,87 @@ static void test_errors(void)
> >   	kvm_vm_free(t.kvm_vm);
> >   }
> >   
[...]
> > 
> > +	} testlist[] = {
> > +		{
> > +			.name = "simple copy",
> > +			.test = test_copy,
> > +			.requirements_met = true,
> > +		},
> > +		{
> > +			.name = "generic error checks",
> > +			.test = test_errors,
> > +			.requirements_met = true,
> > +		},
> > +		{
> > +			.name = "copy with storage keys",
> > +			.test = test_copy_key,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "copy with key storage protection override",
> > +			.test = test_copy_key_storage_prot_override,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "copy with key fetch protection",
> > +			.test = test_copy_key_fetch_prot,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "copy with key fetch protection override",
> > +			.test = test_copy_key_fetch_prot_override,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "error checks with key",
> > +			.test = test_errors_key,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "termination",
> > +			.test = test_termination,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "error checks with key storage protection override",
> > +			.test = test_errors_key_storage_prot_override,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "error checks without key fetch prot override",
> > +			.test = test_errors_key_fetch_prot_override_not_enabled,
> > +			.requirements_met = extension_cap > 0,
> > +		},
> > +		{
> > +			.name = "error checks with key fetch prot override",
> > +			.test = test_errors_key_fetch_prot_override_enabled,
> > +			.requirements_met = extension_cap > 0,
> 
> I wonder whether it would rather make sense to check for "extension_cap & 1" 
> instead of "extension_cap > 0" ?

The cap should always have been a bitmap, but unfortunately I didn't initially
define it as one, the storage key extension must be supported if the cap > 0.
So the test reflects that and may catch an error in the future.
> 
> Anyway:
> Reviewed-by: Thomas Huth <thuth@xxxxxxxxxx>
> 
Thanks!