On 25.07.19 17:36, David Hildenbrand wrote:
> On 25.07.19 17:11, Christian Borntraeger wrote:
>> This adds a check for documented stfle dependencies.
>>
>
> Expected error under TCG:
>
> FAIL: cpumodel: dependency: 37 implies 42
>
> DFP not implemented (yet).
>
> We also don't warn about this in check_consistency(), which is nice for
> TCG ;)
So should I force this to KVM? Or should I try to detect TCG and make this
xfail?
>
>> Signed-off-by: Christian Borntraeger <borntraeger@xxxxxxxxxx>
>> ---
>> s390x/Makefile | 1 +
>> s390x/cpumodel.c | 58 +++++++++++++++++++++++++++++++++++++++++++++
>> s390x/unittests.cfg | 3 +++
>> 3 files changed, 62 insertions(+)
>> create mode 100644 s390x/cpumodel.c
>>
>> diff --git a/s390x/Makefile b/s390x/Makefile
>> index 1f21ddb9c943..574a9a20824d 100644
>> --- a/s390x/Makefile
>> +++ b/s390x/Makefile
>> @@ -11,6 +11,7 @@ tests += $(TEST_DIR)/cmm.elf
>> tests += $(TEST_DIR)/vector.elf
>> tests += $(TEST_DIR)/gs.elf
>> tests += $(TEST_DIR)/iep.elf
>> +tests += $(TEST_DIR)/cpumodel.elf
>> tests_binary = $(patsubst %.elf,%.bin,$(tests))
>>
>> all: directories test_cases test_cases_binary
>> diff --git a/s390x/cpumodel.c b/s390x/cpumodel.c
>> new file mode 100644
>> index 000000000000..8ff61f7f6ec9
>> --- /dev/null
>> +++ b/s390x/cpumodel.c
>> @@ -0,0 +1,58 @@
>> +/*
>> + * Test the known dependencies for facilities
>> + *
>> + * Copyright 2019 IBM Corp.
>> + *
>> + * Authors:
>> + * Christian Borntraeger <borntraeger@xxxxxxxxxx>
>> + *
>> + * This code is free software; you can redistribute it and/or modify it
>> + * under the terms of the GNU Library General Public License version 2.
>> + */
>> +
>> +#include <asm/facility.h>
>> +
>> +static int dep[][2] = {
>> + /* from SA22-7832-11 4-98 facility indications */
>> + { 4, 3},
>> + { 5, 3},
>> + { 5, 4},
>> + { 19, 18},
>> + { 37, 42},
>> + { 43, 42},
>> + { 73, 49},
>> + {134, 129},
>> + {139, 25},
>> + {139, 28},
>> + {146, 76},
>> + /* indirectly documented in description */
>> + { 78, 8}, /* EDAT */
>> + /* new dependencies from gen15 */
>> + { 61, 45},
>> + {148, 129},
>> + {148, 135},
>> + {152, 129},
>> + {152, 134},
>> + {155, 76},
>> + {155, 77},
>> +};
>> +
>> +int main(void)
>> +{
>> + int i;
>> +
>> + report_prefix_push("cpumodel");
>> +
>> + report_prefix_push("dependency");
>> + for (i = 0; i < ARRAY_SIZE(dep); i++) {
>> + if (test_facility(dep[i][0])) {
>> + report("%d implies %d",
>> + !(test_facility(dep[i][0]) && !test_facility(dep[i][1])),
>> + dep[i][0], dep[i][1]);
>> + } else {
>> + report_skip("facility %d not present", dep[i][0]);
>> + }
>> + }
>> + report_prefix_pop();
>
> Are you missing another pop here?
Yes it seems.
>
>> + return report_summary();
>> +}
>> diff --git a/s390x/unittests.cfg b/s390x/unittests.cfg
>> index 546b1f281f8f..db58bad5a038 100644
>> --- a/s390x/unittests.cfg
>> +++ b/s390x/unittests.cfg
>> @@ -61,3 +61,6 @@ file = gs.elf
>>
>> [iep]
>> file = iep.elf
>> +
>> +[cpumodel]
>> +file = cpumodel.elf
>>
>
> Didn't verify the facilities. In general, looks good to me.
>
