2023-06-13 10:16:30

by Ravi Bangoria

[permalink] [raw]
Subject: [PATCH 3/4] perf mem amd: Fix perf_pmus__num_mem_pmus()

AMD cpus does not contain hybrid cores. Also, perf mem/c2c on AMD
internally uses IBS OP PMU, not the core PMU.

Signed-off-by: Ravi Bangoria <[email protected]>
---
tools/perf/arch/x86/util/pmu.c | 15 +++++++++++++++
tools/perf/util/pmus.c | 2 +-
2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/tools/perf/arch/x86/util/pmu.c b/tools/perf/arch/x86/util/pmu.c
index 3c0de3370d7e..8a20d28d9927 100644
--- a/tools/perf/arch/x86/util/pmu.c
+++ b/tools/perf/arch/x86/util/pmu.c
@@ -14,6 +14,8 @@
#include "../../../util/intel-bts.h"
#include "../../../util/pmu.h"
#include "../../../util/fncache.h"
+#include "../../../util/pmus.h"
+#include "env.h"

struct pmu_alias {
char *name;
@@ -168,3 +170,16 @@ char *pmu_find_alias_name(const char *name)

return __pmu_find_alias_name(name);
}
+
+int perf_pmus__num_mem_pmus(void)
+{
+ /*
+ * AMD does not have hybrid cores and also uses IBS OP
+ * pmu for perf mem/c2c.
+ */
+ if (x86__is_amd_cpu())
+ return 1;
+
+ /* Intel uses core pmus for perf mem/c2c */
+ return perf_pmus__num_core_pmus();
+}
diff --git a/tools/perf/util/pmus.c b/tools/perf/util/pmus.c
index e505d2fef828..0ed943932945 100644
--- a/tools/perf/util/pmus.c
+++ b/tools/perf/util/pmus.c
@@ -240,7 +240,7 @@ const struct perf_pmu *perf_pmus__pmu_for_pmu_filter(const char *str)
return NULL;
}

-int perf_pmus__num_mem_pmus(void)
+int __weak perf_pmus__num_mem_pmus(void)
{
/* All core PMUs are for mem events. */
return perf_pmus__num_core_pmus();
--
2.40.1



2023-06-13 15:36:19

by Ian Rogers

[permalink] [raw]
Subject: Re: [PATCH 3/4] perf mem amd: Fix perf_pmus__num_mem_pmus()

On Tue, Jun 13, 2023 at 2:56 AM Ravi Bangoria <[email protected]> wrote:
>
> AMD cpus does not contain hybrid cores. Also, perf mem/c2c on AMD
> internally uses IBS OP PMU, not the core PMU.
>
> Signed-off-by: Ravi Bangoria <[email protected]>
> ---
> tools/perf/arch/x86/util/pmu.c | 15 +++++++++++++++
> tools/perf/util/pmus.c | 2 +-
> 2 files changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/tools/perf/arch/x86/util/pmu.c b/tools/perf/arch/x86/util/pmu.c
> index 3c0de3370d7e..8a20d28d9927 100644
> --- a/tools/perf/arch/x86/util/pmu.c
> +++ b/tools/perf/arch/x86/util/pmu.c
> @@ -14,6 +14,8 @@
> #include "../../../util/intel-bts.h"
> #include "../../../util/pmu.h"
> #include "../../../util/fncache.h"
> +#include "../../../util/pmus.h"
> +#include "env.h"
>
> struct pmu_alias {
> char *name;
> @@ -168,3 +170,16 @@ char *pmu_find_alias_name(const char *name)
>
> return __pmu_find_alias_name(name);
> }
> +
> +int perf_pmus__num_mem_pmus(void)
> +{
> + /*
> + * AMD does not have hybrid cores and also uses IBS OP
> + * pmu for perf mem/c2c.
> + */
> + if (x86__is_amd_cpu())
> + return 1;

The code and comment seem out of sync here. For the hybrid part
perf_pmus__num_core_pmus() will yield 1 if there is no hybrid, so we
can just use perf_pmus__num_core_pmus(). For the IBS OP part, does
that mean that AMD should have 2 mem pmus? Or is IBS OP a core PMU?
Can we add this as an example in the core/other documentation in patch
1, as you've done for ARM, for clarity.

Thanks,
Ian

> +
> + /* Intel uses core pmus for perf mem/c2c */
> + return perf_pmus__num_core_pmus();
> +}
> diff --git a/tools/perf/util/pmus.c b/tools/perf/util/pmus.c
> index e505d2fef828..0ed943932945 100644
> --- a/tools/perf/util/pmus.c
> +++ b/tools/perf/util/pmus.c
> @@ -240,7 +240,7 @@ const struct perf_pmu *perf_pmus__pmu_for_pmu_filter(const char *str)
> return NULL;
> }
>
> -int perf_pmus__num_mem_pmus(void)
> +int __weak perf_pmus__num_mem_pmus(void)
> {
> /* All core PMUs are for mem events. */
> return perf_pmus__num_core_pmus();
> --
> 2.40.1
>

2023-06-14 03:59:34

by Ravi Bangoria

[permalink] [raw]
Subject: Re: [PATCH 3/4] perf mem amd: Fix perf_pmus__num_mem_pmus()

>> +int perf_pmus__num_mem_pmus(void)
>> +{
>> + /*
>> + * AMD does not have hybrid cores and also uses IBS OP
>> + * pmu for perf mem/c2c.
>> + */
>> + if (x86__is_amd_cpu())
>> + return 1;
>
> The code and comment seem out of sync here. For the hybrid part
> perf_pmus__num_core_pmus() will yield 1 if there is no hybrid, so we
> can just use perf_pmus__num_core_pmus(). For the IBS OP part, does
> that mean that AMD should have 2 mem pmus? Or is IBS OP a core PMU?

Sure. Let me remove hybrid part from the comment.

There are two IBS pmus: ibs_fetch// and ibs_op//. Both of them are
independent of the core pmu (cpu//). An instance of all 3 pmus is
present in each hw SMT thread. And, perf mem/c2c internally uses
ibs_op// pmu on AMD. See tools/perf/arch/x86/util/mem-events.c

static struct perf_mem_event perf_mem_events_amd[PERF_MEM_EVENTS__MAX] = {
E(NULL, NULL, NULL),
E(NULL, NULL, NULL),
E("mem-ldst", "ibs_op//", "ibs_op"),
};

Hope this clarifies.

> Can we add this as an example in the core/other documentation in patch
> 1, as you've done for ARM, for clarity.

Sure. Let me also add IBS example in the patch #1 comment.

Thanks,
Ravi