Re: [PATCH v5 3/6] perf test: Add a test for the new Arm CPU ID comparison behavior

[John Garry <john.g.garry@oracle.com>]

On 16/08/2023 10:14, James Clark wrote:
> 
> 
> On 15/08/2023 10:47, John Garry wrote:
>> On 11/08/2023 15:39, James Clark wrote:
>>> Now that variant and revision fields are taken into account the behavior
>>> is slightly more complicated so add a test to ensure that this behaves
>>> as expected.
>>>
>>> Signed-off-by: James Clark <james.clark@arm.com>
>>> ---
>>>    tools/perf/arch/arm64/include/arch-tests.h |  3 ++
>>>    tools/perf/arch/arm64/tests/Build          |  1 +
>>>    tools/perf/arch/arm64/tests/arch-tests.c   |  4 +++
>>>    tools/perf/arch/arm64/tests/cpuid-match.c  | 38 ++++++++++++++++++++++
>>>    4 files changed, 46 insertions(+)
>>>    create mode 100644 tools/perf/arch/arm64/tests/cpuid-match.c
>>>
>>> diff --git a/tools/perf/arch/arm64/include/arch-tests.h
>>> b/tools/perf/arch/arm64/include/arch-tests.h
>>> index 452b3d904521..474d7cf5afbd 100644
>>> --- a/tools/perf/arch/arm64/include/arch-tests.h
>>> +++ b/tools/perf/arch/arm64/include/arch-tests.h
>>> @@ -2,6 +2,9 @@
>>>    #ifndef ARCH_TESTS_H
>>>    #define ARCH_TESTS_H
>>>    +struct test_suite;
>>> +
>>> +int test__cpuid_match(struct test_suite *test, int subtest);
>>>    extern struct test_suite *arch_tests[];
>>>      #endif
>>> diff --git a/tools/perf/arch/arm64/tests/Build
>>> b/tools/perf/arch/arm64/tests/Build
>>> index a61c06bdb757..e337c09e7f56 100644
>>> --- a/tools/perf/arch/arm64/tests/Build
>>> +++ b/tools/perf/arch/arm64/tests/Build
>>> @@ -2,3 +2,4 @@ perf-y += regs_load.o
>>>    perf-$(CONFIG_DWARF_UNWIND) += dwarf-unwind.o
>>>      perf-y += arch-tests.o
>>> +perf-y += cpuid-match.o
>>> diff --git a/tools/perf/arch/arm64/tests/arch-tests.c
>>> b/tools/perf/arch/arm64/tests/arch-tests.c
>>> index ad16b4f8f63e..74932e72c727 100644
>>> --- a/tools/perf/arch/arm64/tests/arch-tests.c
>>> +++ b/tools/perf/arch/arm64/tests/arch-tests.c
>>> @@ -3,9 +3,13 @@
>>>    #include "tests/tests.h"
>>>    #include "arch-tests.h"
>>>    +
>>> +DEFINE_SUITE("arm64 CPUID matching", cpuid_match);
>>> +
>>>    struct test_suite *arch_tests[] = {
>>>    #ifdef HAVE_DWARF_UNWIND_SUPPORT
>>>        &suite__dwarf_unwind,
>>>    #endif
>>> +    &suite__cpuid_match,
>>>        NULL,
>>>    };
>>> diff --git a/tools/perf/arch/arm64/tests/cpuid-match.c
>>> b/tools/perf/arch/arm64/tests/cpuid-match.c
>>> new file mode 100644
>>> index 000000000000..af0871b54ae7
>>> --- /dev/null
>>> +++ b/tools/perf/arch/arm64/tests/cpuid-match.c
>>> @@ -0,0 +1,38 @@
>>> +// SPDX-License-Identifier: GPL-2.0
>>> +#include <linux/compiler.h>
>>> +
>>> +#include "arch-tests.h"
>>> +#include "tests/tests.h"
>>> +#include "util/header.h"
>>> +
>>> +int test__cpuid_match(struct test_suite *test __maybe_unused,
>>> +                 int subtest __maybe_unused)
>>> +{
>>> +    /* midr with no leading zeros matches */
>>> +    if (strcmp_cpuid_str("0x410fd0c0", "0x00000000410fd0c0"))
>>> +        return -1;
>>> +    /* Upper case matches */
>>> +    if (strcmp_cpuid_str("0x410fd0c0", "0x00000000410FD0C0"))
>>> +        return -1;
>>> +    /* r0p0 = r0p0 matches */
>>> +    if (strcmp_cpuid_str("0x00000000410fd480", "0x00000000410fd480"))
>>> +        return -1;
>>> +    /* r0p1 > r0p0 matches */
>>> +    if (strcmp_cpuid_str("0x00000000410fd480", "0x00000000410fd481"))
>>> +        return -1;
>>> +    /* r1p0 > r0p0 matches*/
>>> +    if (strcmp_cpuid_str("0x00000000410fd480", "0x00000000411fd480"))
>>> +        return -1;
>>> +    /* r0p0 < r0p1 doesn't match */
>>> +    if (!strcmp_cpuid_str("0x00000000410fd481", "0x00000000410fd480"))
>>> +        return -1;
>>> +    /* r0p0 < r1p0 doesn't match */
>>> +    if (!strcmp_cpuid_str("0x00000000411fd480", "0x00000000410fd480"))
>>> +        return -1;
>>> +    /* Different CPU doesn't match */
>>> +    if (!strcmp_cpuid_str("0x00000000410fd4c0", "0x00000000430f0af0"))
>>> +        return -1;
>>> +
>>> +    return 0;
>>> +}
>>> +
>>
>> Would it be possible to put this in core test code, since x86 also
>> supports strcmp_cpuid_str()?
>>
> 
> That's how I started, but Ian suggested to move it to an arch specific
> folder because that's what it was testing.
> 

Yeah, I see that comment now.

> We could still add test__cpuid_match() in the x86 folder rather than
> adding it with #ifdefs

I was thinking to make cpuid_match_array[] exposed by the arch code and 
have a "weak", i.e. version for other archs.

, but I don't think it needs to be done here
> because I haven't touched the x86 code.

For the moment, I don't feel too strongly about this and it can be done 
as a follow-up

Reviewed-by: John Garry <john.g.garry@oracle.com>

> 
>> Maybe we would have an structure per arch of cpuids and expected
>> results, like
>>
>> struct cpuid_match {
>>      char *cpuid1;
>>      char *cpuid1;
>>      int expected_result;
>> };
>>
>>
>> #ifdef ARM64
>>   cpuid_match_array[] = {
>>      {"0x410fd0c0", "0x00000000410FD0C0", -1},
>>      {"0x00000000410fd480", "0x00000000410fd480", -1},
>>      ...
>>      {} /* sentinel */
>>
>> };
>> #else if defined(X86)
>>   cpuid_match_array[] = {
>>      {....}
>>      ...
>>      {} /* sentinel */
>>
>> };
>> #else
>> /* no support */
>> #endif
>>
>> Thanks,
>> John