* [PATCH 0/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-10 12:21 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
This came out of the discussion here [1]. It seems like we can get some
extra big.LITTLE stuff working pretty easily. The test issues mentioned
in the linked thread are actually fairly unrelated and I've fixed them
in a different set on the list.
After adding it in the first commit, the remaining ones tidy up a
related capability that doesn't do anything any more.
I've added a fixes tag for the commit where
PERF_PMU_CAP_EXTENDED_HW_TYPE was originally added because it probably
should have been added to the Arm PMU at the same time. It doesn't apply
cleanly that far back because another capability was added between then,
but the resolution is trivial.
Thanks
James
[1]: https://lore.kernel.org/linux-perf-users/CAP-5=fVkRc9=ySJ=fG-SQ8oAKmE_1mhHHzSASmGHUsda5Qy92A@mail.gmail.com/T/#t
James Clark (4):
arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
arch/x86/events/core.c | 1 -
drivers/perf/arm_pmu.c | 10 ++++++----
include/linux/perf_event.h | 2 +-
3 files changed, 7 insertions(+), 6 deletions(-)
--
2.34.1
^ permalink raw reply [flat|nested] 39+ messages in thread
* [PATCH 0/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-10 12:21 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
This came out of the discussion here [1]. It seems like we can get some
extra big.LITTLE stuff working pretty easily. The test issues mentioned
in the linked thread are actually fairly unrelated and I've fixed them
in a different set on the list.
After adding it in the first commit, the remaining ones tidy up a
related capability that doesn't do anything any more.
I've added a fixes tag for the commit where
PERF_PMU_CAP_EXTENDED_HW_TYPE was originally added because it probably
should have been added to the Arm PMU at the same time. It doesn't apply
cleanly that far back because another capability was added between then,
but the resolution is trivial.
Thanks
James
[1]: https://lore.kernel.org/linux-perf-users/CAP-5=fVkRc9=ySJ=fG-SQ8oAKmE_1mhHHzSASmGHUsda5Qy92A@mail.gmail.com/T/#t
James Clark (4):
arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
arch/x86/events/core.c | 1 -
drivers/perf/arm_pmu.c | 10 ++++++----
include/linux/perf_event.h | 2 +-
3 files changed, 7 insertions(+), 6 deletions(-)
--
2.34.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 12:21 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
This capability gives us the ability to open PERF_TYPE_HARDWARE and
PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
implementation is contained in the Perf core and tool code so no change
to the Arm PMU driver is needed.
The following basic use case now results in Perf opening the event on
all PMUs rather than picking only one in an unpredictable way:
$ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
963279620 armv8_cortex_a57/cycles/ (99.19%)
752745657 armv8_cortex_a53/cycles/ (94.80%)
Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
Suggested-by: Ian Rogers <irogers@google.com>
Signed-off-by: James Clark <james.clark@arm.com>
---
drivers/perf/arm_pmu.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
index 277e29fbd504..d8844a9461a2 100644
--- a/drivers/perf/arm_pmu.c
+++ b/drivers/perf/arm_pmu.c
@@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
* configuration (e.g. big.LITTLE). This is not an uncore PMU,
* and we have taken ctx sharing into account (e.g. with our
* pmu::filter callback and pmu::event_init group validation).
+ *
+ * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
+ * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
+ * specific PMU.
*/
- .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
+ .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
+ PERF_PMU_CAP_EXTENDED_HW_TYPE,
};
pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
--
2.34.1
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-10 12:21 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
This capability gives us the ability to open PERF_TYPE_HARDWARE and
PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
implementation is contained in the Perf core and tool code so no change
to the Arm PMU driver is needed.
The following basic use case now results in Perf opening the event on
all PMUs rather than picking only one in an unpredictable way:
$ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
963279620 armv8_cortex_a57/cycles/ (99.19%)
752745657 armv8_cortex_a53/cycles/ (94.80%)
Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
Suggested-by: Ian Rogers <irogers@google.com>
Signed-off-by: James Clark <james.clark@arm.com>
---
drivers/perf/arm_pmu.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
index 277e29fbd504..d8844a9461a2 100644
--- a/drivers/perf/arm_pmu.c
+++ b/drivers/perf/arm_pmu.c
@@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
* configuration (e.g. big.LITTLE). This is not an uncore PMU,
* and we have taken ctx sharing into account (e.g. with our
* pmu::filter callback and pmu::event_init group validation).
+ *
+ * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
+ * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
+ * specific PMU.
*/
- .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
+ .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
+ PERF_PMU_CAP_EXTENDED_HW_TYPE,
};
pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
--
2.34.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 2/4] perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 12:21 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
Since commit bd2756811766 ("perf: Rewrite core context handling") the
relationship between perf_event_context and PMUs has changed so that
the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
silenced no longer exists.
Remove the capability to avoid confusion that it actually influences
any perf core behavior. This change should be a no-op.
Signed-off-by: James Clark <james.clark@arm.com>
---
arch/x86/events/core.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 9d248703cbdd..2353aaf0b248 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -2168,7 +2168,6 @@ static int __init init_hw_perf_events(void)
hybrid_pmu->pmu = pmu;
hybrid_pmu->pmu.type = -1;
hybrid_pmu->pmu.attr_update = x86_pmu.attr_update;
- hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_HETEROGENEOUS_CPUS;
hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_EXTENDED_HW_TYPE;
err = perf_pmu_register(&hybrid_pmu->pmu, hybrid_pmu->name,
--
2.34.1
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 2/4] perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-10 12:21 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
Since commit bd2756811766 ("perf: Rewrite core context handling") the
relationship between perf_event_context and PMUs has changed so that
the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
silenced no longer exists.
Remove the capability to avoid confusion that it actually influences
any perf core behavior. This change should be a no-op.
Signed-off-by: James Clark <james.clark@arm.com>
---
arch/x86/events/core.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 9d248703cbdd..2353aaf0b248 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -2168,7 +2168,6 @@ static int __init init_hw_perf_events(void)
hybrid_pmu->pmu = pmu;
hybrid_pmu->pmu.type = -1;
hybrid_pmu->pmu.attr_update = x86_pmu.attr_update;
- hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_HETEROGENEOUS_CPUS;
hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_EXTENDED_HW_TYPE;
err = perf_pmu_register(&hybrid_pmu->pmu, hybrid_pmu->name,
--
2.34.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 12:21 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
Since commit bd2756811766 ("perf: Rewrite core context handling") the
relationship between perf_event_context and PMUs has changed so that
the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
silenced no longer exists.
Remove the capability and associated comment to avoid confusion that it
actually influences any perf core behavior. This change should be a
no-op.
Signed-off-by: James Clark <james.clark@arm.com>
---
drivers/perf/arm_pmu.c | 7 ++-----
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
index d8844a9461a2..297906df6628 100644
--- a/drivers/perf/arm_pmu.c
+++ b/drivers/perf/arm_pmu.c
@@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
.attr_groups = pmu->attr_groups,
/*
* This is a CPU PMU potentially in a heterogeneous
- * configuration (e.g. big.LITTLE). This is not an uncore PMU,
- * and we have taken ctx sharing into account (e.g. with our
- * pmu::filter callback and pmu::event_init group validation).
- *
+ * configuration (e.g. big.LITTLE) so
* PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
* PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
* specific PMU.
*/
- .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
+ .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
PERF_PMU_CAP_EXTENDED_HW_TYPE,
};
--
2.34.1
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-10 12:21 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
Since commit bd2756811766 ("perf: Rewrite core context handling") the
relationship between perf_event_context and PMUs has changed so that
the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
silenced no longer exists.
Remove the capability and associated comment to avoid confusion that it
actually influences any perf core behavior. This change should be a
no-op.
Signed-off-by: James Clark <james.clark@arm.com>
---
drivers/perf/arm_pmu.c | 7 ++-----
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
index d8844a9461a2..297906df6628 100644
--- a/drivers/perf/arm_pmu.c
+++ b/drivers/perf/arm_pmu.c
@@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
.attr_groups = pmu->attr_groups,
/*
* This is a CPU PMU potentially in a heterogeneous
- * configuration (e.g. big.LITTLE). This is not an uncore PMU,
- * and we have taken ctx sharing into account (e.g. with our
- * pmu::filter callback and pmu::event_init group validation).
- *
+ * configuration (e.g. big.LITTLE) so
* PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
* PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
* specific PMU.
*/
- .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
+ .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
PERF_PMU_CAP_EXTENDED_HW_TYPE,
};
--
2.34.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 12:21 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
Since commit bd2756811766 ("perf: Rewrite core context handling") the
relationship between perf_event_context and PMUs has changed so that
the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
silenced no longer exists.
Remove the capability to avoid confusion that it actually influences
any perf core behavior. This change should be a no-op.
Signed-off-by: James Clark <james.clark@arm.com>
---
include/linux/perf_event.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index d5628a7b5eaa..3f4d941fd6c5 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -288,7 +288,7 @@ struct perf_event_pmu_context;
#define PERF_PMU_CAP_EXTENDED_REGS 0x0008
#define PERF_PMU_CAP_EXCLUSIVE 0x0010
#define PERF_PMU_CAP_ITRACE 0x0020
-#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
+/* Unused 0x0040 */
#define PERF_PMU_CAP_NO_EXCLUDE 0x0080
#define PERF_PMU_CAP_AUX_OUTPUT 0x0100
#define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
--
2.34.1
^ permalink raw reply related [flat|nested] 39+ messages in thread
* [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-10 12:21 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-10 12:21 UTC (permalink / raw)
To: linux-perf-users, irogers
Cc: James Clark, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
Since commit bd2756811766 ("perf: Rewrite core context handling") the
relationship between perf_event_context and PMUs has changed so that
the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
silenced no longer exists.
Remove the capability to avoid confusion that it actually influences
any perf core behavior. This change should be a no-op.
Signed-off-by: James Clark <james.clark@arm.com>
---
include/linux/perf_event.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index d5628a7b5eaa..3f4d941fd6c5 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -288,7 +288,7 @@ struct perf_event_pmu_context;
#define PERF_PMU_CAP_EXTENDED_REGS 0x0008
#define PERF_PMU_CAP_EXCLUSIVE 0x0010
#define PERF_PMU_CAP_ITRACE 0x0020
-#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
+/* Unused 0x0040 */
#define PERF_PMU_CAP_NO_EXCLUDE 0x0080
#define PERF_PMU_CAP_AUX_OUTPUT 0x0100
#define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
--
2.34.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 16:04 ` Ian Rogers
-1 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:04 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel, Yabin Cui
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> implementation is contained in the Perf core and tool code so no change
> to the Arm PMU driver is needed.
>
> The following basic use case now results in Perf opening the event on
> all PMUs rather than picking only one in an unpredictable way:
>
> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>
> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>
> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>
> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> Suggested-by: Ian Rogers <irogers@google.com>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> drivers/perf/arm_pmu.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index 277e29fbd504..d8844a9461a2 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> * and we have taken ctx sharing into account (e.g. with our
> * pmu::filter callback and pmu::event_init group validation).
> + *
> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> + * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
> };
>
> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
> --
> 2.34.1
>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-10 16:04 ` Ian Rogers
0 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:04 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel, Yabin Cui
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> implementation is contained in the Perf core and tool code so no change
> to the Arm PMU driver is needed.
>
> The following basic use case now results in Perf opening the event on
> all PMUs rather than picking only one in an unpredictable way:
>
> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>
> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>
> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>
> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> Suggested-by: Ian Rogers <irogers@google.com>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> drivers/perf/arm_pmu.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index 277e29fbd504..d8844a9461a2 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> * and we have taken ctx sharing into account (e.g. with our
> * pmu::filter callback and pmu::event_init group validation).
> + *
> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> + * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
> };
>
> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
> --
> 2.34.1
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 2/4] perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 16:07 ` Ian Rogers
-1 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:07 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability to avoid confusion that it actually influences
> any perf core behavior. This change should be a no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> arch/x86/events/core.c | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
> index 9d248703cbdd..2353aaf0b248 100644
> --- a/arch/x86/events/core.c
> +++ b/arch/x86/events/core.c
> @@ -2168,7 +2168,6 @@ static int __init init_hw_perf_events(void)
> hybrid_pmu->pmu = pmu;
> hybrid_pmu->pmu.type = -1;
> hybrid_pmu->pmu.attr_update = x86_pmu.attr_update;
> - hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_HETEROGENEOUS_CPUS;
> hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_EXTENDED_HW_TYPE;
>
> err = perf_pmu_register(&hybrid_pmu->pmu, hybrid_pmu->name,
> --
> 2.34.1
>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 2/4] perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-10 16:07 ` Ian Rogers
0 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:07 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability to avoid confusion that it actually influences
> any perf core behavior. This change should be a no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> arch/x86/events/core.c | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
> index 9d248703cbdd..2353aaf0b248 100644
> --- a/arch/x86/events/core.c
> +++ b/arch/x86/events/core.c
> @@ -2168,7 +2168,6 @@ static int __init init_hw_perf_events(void)
> hybrid_pmu->pmu = pmu;
> hybrid_pmu->pmu.type = -1;
> hybrid_pmu->pmu.attr_update = x86_pmu.attr_update;
> - hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_HETEROGENEOUS_CPUS;
> hybrid_pmu->pmu.capabilities |= PERF_PMU_CAP_EXTENDED_HW_TYPE;
>
> err = perf_pmu_register(&hybrid_pmu->pmu, hybrid_pmu->name,
> --
> 2.34.1
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 16:08 ` Ian Rogers
-1 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:08 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability and associated comment to avoid confusion that it
> actually influences any perf core behavior. This change should be a
> no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> drivers/perf/arm_pmu.c | 7 ++-----
> 1 file changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index d8844a9461a2..297906df6628 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
> .attr_groups = pmu->attr_groups,
> /*
> * This is a CPU PMU potentially in a heterogeneous
> - * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> - * and we have taken ctx sharing into account (e.g. with our
> - * pmu::filter callback and pmu::event_init group validation).
> - *
> + * configuration (e.g. big.LITTLE) so
> * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
> PERF_PMU_CAP_EXTENDED_HW_TYPE,
> };
>
> --
> 2.34.1
>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-10 16:08 ` Ian Rogers
0 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:08 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability and associated comment to avoid confusion that it
> actually influences any perf core behavior. This change should be a
> no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> drivers/perf/arm_pmu.c | 7 ++-----
> 1 file changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index d8844a9461a2..297906df6628 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
> .attr_groups = pmu->attr_groups,
> /*
> * This is a CPU PMU potentially in a heterogeneous
> - * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> - * and we have taken ctx sharing into account (e.g. with our
> - * pmu::filter callback and pmu::event_init group validation).
> - *
> + * configuration (e.g. big.LITTLE) so
> * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
> PERF_PMU_CAP_EXTENDED_HW_TYPE,
> };
>
> --
> 2.34.1
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-10 16:09 ` Ian Rogers
-1 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:09 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability to avoid confusion that it actually influences
> any perf core behavior. This change should be a no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> include/linux/perf_event.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
> index d5628a7b5eaa..3f4d941fd6c5 100644
> --- a/include/linux/perf_event.h
> +++ b/include/linux/perf_event.h
> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
> #define PERF_PMU_CAP_ITRACE 0x0020
> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
> +/* Unused 0x0040 */
> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
> --
> 2.34.1
>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-10 16:09 ` Ian Rogers
0 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-10 16:09 UTC (permalink / raw)
To: James Clark
Cc: linux-perf-users, Peter Zijlstra, Ingo Molnar,
Arnaldo Carvalho de Melo, Mark Rutland, Alexander Shishkin,
Jiri Olsa, Namhyung Kim, Adrian Hunter, Thomas Gleixner,
Borislav Petkov, Dave Hansen, x86, H. Peter Anvin, Will Deacon,
Kan Liang, linux-kernel, linux-arm-kernel
On Mon, Jul 10, 2023 at 5:22 AM James Clark <james.clark@arm.com> wrote:
>
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability to avoid confusion that it actually influences
> any perf core behavior. This change should be a no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Thanks,
Ian
> ---
> include/linux/perf_event.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
> index d5628a7b5eaa..3f4d941fd6c5 100644
> --- a/include/linux/perf_event.h
> +++ b/include/linux/perf_event.h
> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
> #define PERF_PMU_CAP_ITRACE 0x0020
> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
> +/* Unused 0x0040 */
> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
> --
> 2.34.1
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-10 12:21 ` James Clark
@ 2023-07-11 12:01 ` Anshuman Khandual
-1 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-11 12:01 UTC (permalink / raw)
To: James Clark, linux-perf-users, irogers
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel
On 7/10/23 17:51, James Clark wrote:
> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> implementation is contained in the Perf core and tool code so no change
> to the Arm PMU driver is needed.
>
> The following basic use case now results in Perf opening the event on
> all PMUs rather than picking only one in an unpredictable way:
>
> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>
> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>
> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>
> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> Suggested-by: Ian Rogers <irogers@google.com>
> Signed-off-by: James Clark <james.clark@arm.com>
> ---
> drivers/perf/arm_pmu.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index 277e29fbd504..d8844a9461a2 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> * and we have taken ctx sharing into account (e.g. with our
> * pmu::filter callback and pmu::event_init group validation).
> + *
> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
s/legacy/generic ? These hardware events are still around.
> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> + * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
> };
>
> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-11 12:01 ` Anshuman Khandual
0 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-11 12:01 UTC (permalink / raw)
To: James Clark, linux-perf-users, irogers
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel
On 7/10/23 17:51, James Clark wrote:
> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> implementation is contained in the Perf core and tool code so no change
> to the Arm PMU driver is needed.
>
> The following basic use case now results in Perf opening the event on
> all PMUs rather than picking only one in an unpredictable way:
>
> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>
> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>
> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>
> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> Suggested-by: Ian Rogers <irogers@google.com>
> Signed-off-by: James Clark <james.clark@arm.com>
> ---
> drivers/perf/arm_pmu.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index 277e29fbd504..d8844a9461a2 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> * and we have taken ctx sharing into account (e.g. with our
> * pmu::filter callback and pmu::event_init group validation).
> + *
> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
s/legacy/generic ? These hardware events are still around.
> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> + * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
> };
>
> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-11 12:08 ` Anshuman Khandual
-1 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-11 12:08 UTC (permalink / raw)
To: James Clark, linux-perf-users, irogers
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel
On 7/10/23 17:51, James Clark wrote:
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability and associated comment to avoid confusion that it
> actually influences any perf core behavior. This change should be a
> no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
> ---
> drivers/perf/arm_pmu.c | 7 ++-----
> 1 file changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index d8844a9461a2..297906df6628 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
> .attr_groups = pmu->attr_groups,
> /*
> * This is a CPU PMU potentially in a heterogeneous
> - * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> - * and we have taken ctx sharing into account (e.g. with our
> - * pmu::filter callback and pmu::event_init group validation).
> - *
> + * configuration (e.g. big.LITTLE) so
> * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
> PERF_PMU_CAP_EXTENDED_HW_TYPE,
Small nit, the second line could be folded back into the first one.
> };
>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-11 12:08 ` Anshuman Khandual
0 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-11 12:08 UTC (permalink / raw)
To: James Clark, linux-perf-users, irogers
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel
On 7/10/23 17:51, James Clark wrote:
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability and associated comment to avoid confusion that it
> actually influences any perf core behavior. This change should be a
> no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
> ---
> drivers/perf/arm_pmu.c | 7 ++-----
> 1 file changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> index d8844a9461a2..297906df6628 100644
> --- a/drivers/perf/arm_pmu.c
> +++ b/drivers/perf/arm_pmu.c
> @@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
> .attr_groups = pmu->attr_groups,
> /*
> * This is a CPU PMU potentially in a heterogeneous
> - * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> - * and we have taken ctx sharing into account (e.g. with our
> - * pmu::filter callback and pmu::event_init group validation).
> - *
> + * configuration (e.g. big.LITTLE) so
> * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> * specific PMU.
> */
> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> + .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
> PERF_PMU_CAP_EXTENDED_HW_TYPE,
Small nit, the second line could be folded back into the first one.
> };
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-10 12:21 ` James Clark
@ 2023-07-11 12:10 ` Anshuman Khandual
-1 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-11 12:10 UTC (permalink / raw)
To: James Clark, linux-perf-users, irogers
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel
On 7/10/23 17:51, James Clark wrote:
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability to avoid confusion that it actually influences
> any perf core behavior. This change should be a no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
> ---
> include/linux/perf_event.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
> index d5628a7b5eaa..3f4d941fd6c5 100644
> --- a/include/linux/perf_event.h
> +++ b/include/linux/perf_event.h
> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
> #define PERF_PMU_CAP_ITRACE 0x0020
> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
> +/* Unused 0x0040 */
Small nit, "Unused" marking might not be required here.
> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-11 12:10 ` Anshuman Khandual
0 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-11 12:10 UTC (permalink / raw)
To: James Clark, linux-perf-users, irogers
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel
On 7/10/23 17:51, James Clark wrote:
> Since commit bd2756811766 ("perf: Rewrite core context handling") the
> relationship between perf_event_context and PMUs has changed so that
> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
> silenced no longer exists.
>
> Remove the capability to avoid confusion that it actually influences
> any perf core behavior. This change should be a no-op.
>
> Signed-off-by: James Clark <james.clark@arm.com>
> ---
> include/linux/perf_event.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
> index d5628a7b5eaa..3f4d941fd6c5 100644
> --- a/include/linux/perf_event.h
> +++ b/include/linux/perf_event.h
> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
> #define PERF_PMU_CAP_ITRACE 0x0020
> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
> +/* Unused 0x0040 */
Small nit, "Unused" marking might not be required here.
> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-11 12:08 ` Anshuman Khandual
@ 2023-07-11 14:04 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-11 14:04 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 11/07/2023 13:08, Anshuman Khandual wrote:
>
>
> On 7/10/23 17:51, James Clark wrote:
>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>> relationship between perf_event_context and PMUs has changed so that
>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>> silenced no longer exists.
>>
>> Remove the capability and associated comment to avoid confusion that it
>> actually influences any perf core behavior. This change should be a
>> no-op.
>>
>> Signed-off-by: James Clark <james.clark@arm.com>
>> ---
>> drivers/perf/arm_pmu.c | 7 ++-----
>> 1 file changed, 2 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
>> index d8844a9461a2..297906df6628 100644
>> --- a/drivers/perf/arm_pmu.c
>> +++ b/drivers/perf/arm_pmu.c
>> @@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
>> .attr_groups = pmu->attr_groups,
>> /*
>> * This is a CPU PMU potentially in a heterogeneous
>> - * configuration (e.g. big.LITTLE). This is not an uncore PMU,
>> - * and we have taken ctx sharing into account (e.g. with our
>> - * pmu::filter callback and pmu::event_init group validation).
>> - *
>> + * configuration (e.g. big.LITTLE) so
>> * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
>> * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
>> * specific PMU.
>> */
>> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
>> + .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
>> PERF_PMU_CAP_EXTENDED_HW_TYPE,
>
> Small nit, the second line could be folded back into the first one.
.clang-format has a column limit of 80 which is why it gets folded. The
kernel coding style guide also still says 80 is a strongly preferred
limit so I'll probably leave it.
>
>> };
>>
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 3/4] arm_pmu: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-11 14:04 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-11 14:04 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 11/07/2023 13:08, Anshuman Khandual wrote:
>
>
> On 7/10/23 17:51, James Clark wrote:
>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>> relationship between perf_event_context and PMUs has changed so that
>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>> silenced no longer exists.
>>
>> Remove the capability and associated comment to avoid confusion that it
>> actually influences any perf core behavior. This change should be a
>> no-op.
>>
>> Signed-off-by: James Clark <james.clark@arm.com>
>> ---
>> drivers/perf/arm_pmu.c | 7 ++-----
>> 1 file changed, 2 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
>> index d8844a9461a2..297906df6628 100644
>> --- a/drivers/perf/arm_pmu.c
>> +++ b/drivers/perf/arm_pmu.c
>> @@ -872,15 +872,12 @@ struct arm_pmu *armpmu_alloc(void)
>> .attr_groups = pmu->attr_groups,
>> /*
>> * This is a CPU PMU potentially in a heterogeneous
>> - * configuration (e.g. big.LITTLE). This is not an uncore PMU,
>> - * and we have taken ctx sharing into account (e.g. with our
>> - * pmu::filter callback and pmu::event_init group validation).
>> - *
>> + * configuration (e.g. big.LITTLE) so
>> * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
>> * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
>> * specific PMU.
>> */
>> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
>> + .capabilities = PERF_PMU_CAP_EXTENDED_REGS |
>> PERF_PMU_CAP_EXTENDED_HW_TYPE,
>
> Small nit, the second line could be folded back into the first one.
.clang-format has a column limit of 80 which is why it gets folded. The
kernel coding style guide also still says 80 is a strongly preferred
limit so I'll probably leave it.
>
>> };
>>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-11 12:01 ` Anshuman Khandual
@ 2023-07-11 14:12 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-11 14:12 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 11/07/2023 13:01, Anshuman Khandual wrote:
>
>
> On 7/10/23 17:51, James Clark wrote:
>> This capability gives us the ability to open PERF_TYPE_HARDWARE and
>> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
>> implementation is contained in the Perf core and tool code so no change
>> to the Arm PMU driver is needed.
>>
>> The following basic use case now results in Perf opening the event on
>> all PMUs rather than picking only one in an unpredictable way:
>>
>> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>>
>> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>>
>> 963279620 armv8_cortex_a57/cycles/ (99.19%)
>> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>>
>> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
>> Suggested-by: Ian Rogers <irogers@google.com>
>> Signed-off-by: James Clark <james.clark@arm.com>
>> ---
>> drivers/perf/arm_pmu.c | 7 ++++++-
>> 1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
>> index 277e29fbd504..d8844a9461a2 100644
>> --- a/drivers/perf/arm_pmu.c
>> +++ b/drivers/perf/arm_pmu.c
>> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
>> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
>> * and we have taken ctx sharing into account (e.g. with our
>> * pmu::filter callback and pmu::event_init group validation).
>> + *
>> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
>
> s/legacy/generic ? These hardware events are still around.
True, I thought I saw it mentioned that way somewhere, but I can
probably just remove it altogether. PERF_TYPE_HARDWARE and
PERF_TYPE_HW_CACHE is enough.
>
>> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
>> + * specific PMU.
>> */
>> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
>> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
>> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
>> };
>>
>> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-11 14:12 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-11 14:12 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 11/07/2023 13:01, Anshuman Khandual wrote:
>
>
> On 7/10/23 17:51, James Clark wrote:
>> This capability gives us the ability to open PERF_TYPE_HARDWARE and
>> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
>> implementation is contained in the Perf core and tool code so no change
>> to the Arm PMU driver is needed.
>>
>> The following basic use case now results in Perf opening the event on
>> all PMUs rather than picking only one in an unpredictable way:
>>
>> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>>
>> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>>
>> 963279620 armv8_cortex_a57/cycles/ (99.19%)
>> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>>
>> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
>> Suggested-by: Ian Rogers <irogers@google.com>
>> Signed-off-by: James Clark <james.clark@arm.com>
>> ---
>> drivers/perf/arm_pmu.c | 7 ++++++-
>> 1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
>> index 277e29fbd504..d8844a9461a2 100644
>> --- a/drivers/perf/arm_pmu.c
>> +++ b/drivers/perf/arm_pmu.c
>> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
>> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
>> * and we have taken ctx sharing into account (e.g. with our
>> * pmu::filter callback and pmu::event_init group validation).
>> + *
>> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
>
> s/legacy/generic ? These hardware events are still around.
True, I thought I saw it mentioned that way somewhere, but I can
probably just remove it altogether. PERF_TYPE_HARDWARE and
PERF_TYPE_HW_CACHE is enough.
>
>> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
>> + * specific PMU.
>> */
>> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
>> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
>> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
>> };
>>
>> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-11 12:10 ` Anshuman Khandual
@ 2023-07-11 14:15 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-11 14:15 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 11/07/2023 13:10, Anshuman Khandual wrote:
>
>
> On 7/10/23 17:51, James Clark wrote:
>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>> relationship between perf_event_context and PMUs has changed so that
>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>> silenced no longer exists.
>>
>> Remove the capability to avoid confusion that it actually influences
>> any perf core behavior. This change should be a no-op.
>>
>> Signed-off-by: James Clark <james.clark@arm.com>
>> ---
>> include/linux/perf_event.h | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
>> index d5628a7b5eaa..3f4d941fd6c5 100644
>> --- a/include/linux/perf_event.h
>> +++ b/include/linux/perf_event.h
>> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
>> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
>> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
>> #define PERF_PMU_CAP_ITRACE 0x0020
>> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
>> +/* Unused 0x0040 */
>
> Small nit, "Unused" marking might not be required here.
>
But then it would be very easy to miss that there is a free bit if I
don't leave the comment. Is it really better without it?
I could shift all the following ones down by one bit, but it would be a
lot of work to make sure that nobody has hard coded some check for one
of the bits instead of using the define somewhere.
>> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
>> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
>> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-11 14:15 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-11 14:15 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 11/07/2023 13:10, Anshuman Khandual wrote:
>
>
> On 7/10/23 17:51, James Clark wrote:
>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>> relationship between perf_event_context and PMUs has changed so that
>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>> silenced no longer exists.
>>
>> Remove the capability to avoid confusion that it actually influences
>> any perf core behavior. This change should be a no-op.
>>
>> Signed-off-by: James Clark <james.clark@arm.com>
>> ---
>> include/linux/perf_event.h | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
>> index d5628a7b5eaa..3f4d941fd6c5 100644
>> --- a/include/linux/perf_event.h
>> +++ b/include/linux/perf_event.h
>> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
>> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
>> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
>> #define PERF_PMU_CAP_ITRACE 0x0020
>> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
>> +/* Unused 0x0040 */
>
> Small nit, "Unused" marking might not be required here.
>
But then it would be very easy to miss that there is a free bit if I
don't leave the comment. Is it really better without it?
I could shift all the following ones down by one bit, but it would be a
lot of work to make sure that nobody has hard coded some check for one
of the bits instead of using the define somewhere.
>> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
>> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
>> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-11 14:15 ` James Clark
@ 2023-07-13 7:30 ` Anshuman Khandual
-1 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-13 7:30 UTC (permalink / raw)
To: James Clark
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 7/11/23 19:45, James Clark wrote:
>
>
> On 11/07/2023 13:10, Anshuman Khandual wrote:
>>
>>
>> On 7/10/23 17:51, James Clark wrote:
>>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>>> relationship between perf_event_context and PMUs has changed so that
>>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>>> silenced no longer exists.
>>>
>>> Remove the capability to avoid confusion that it actually influences
>>> any perf core behavior. This change should be a no-op.
>>>
>>> Signed-off-by: James Clark <james.clark@arm.com>
>>> ---
>>> include/linux/perf_event.h | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
>>> index d5628a7b5eaa..3f4d941fd6c5 100644
>>> --- a/include/linux/perf_event.h
>>> +++ b/include/linux/perf_event.h
>>> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
>>> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
>>> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
>>> #define PERF_PMU_CAP_ITRACE 0x0020
>>> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
>>> +/* Unused 0x0040 */
>>
>> Small nit, "Unused" marking might not be required here.
>>
>
> But then it would be very easy to miss that there is a free bit if I
> don't leave the comment. Is it really better without it?
>
> I could shift all the following ones down by one bit, but it would be a
Sounds as a better option IMHO.
> lot of work to make sure that nobody has hard coded some check for one
> of the bits instead of using the define somewhere.
These are not user visible ABI and hence defined in include/linux/perf_event.h
to be used by drivers registering a PMU for capability enumeration. I am just
wondering why they might have been hard coded any where ?
>
>>> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
>>> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
>>> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-13 7:30 ` Anshuman Khandual
0 siblings, 0 replies; 39+ messages in thread
From: Anshuman Khandual @ 2023-07-13 7:30 UTC (permalink / raw)
To: James Clark
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 7/11/23 19:45, James Clark wrote:
>
>
> On 11/07/2023 13:10, Anshuman Khandual wrote:
>>
>>
>> On 7/10/23 17:51, James Clark wrote:
>>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>>> relationship between perf_event_context and PMUs has changed so that
>>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>>> silenced no longer exists.
>>>
>>> Remove the capability to avoid confusion that it actually influences
>>> any perf core behavior. This change should be a no-op.
>>>
>>> Signed-off-by: James Clark <james.clark@arm.com>
>>> ---
>>> include/linux/perf_event.h | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
>>> index d5628a7b5eaa..3f4d941fd6c5 100644
>>> --- a/include/linux/perf_event.h
>>> +++ b/include/linux/perf_event.h
>>> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
>>> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
>>> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
>>> #define PERF_PMU_CAP_ITRACE 0x0020
>>> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
>>> +/* Unused 0x0040 */
>>
>> Small nit, "Unused" marking might not be required here.
>>
>
> But then it would be very easy to miss that there is a free bit if I
> don't leave the comment. Is it really better without it?
>
> I could shift all the following ones down by one bit, but it would be a
Sounds as a better option IMHO.
> lot of work to make sure that nobody has hard coded some check for one
> of the bits instead of using the define somewhere.
These are not user visible ABI and hence defined in include/linux/perf_event.h
to be used by drivers registering a PMU for capability enumeration. I am just
wondering why they might have been hard coded any where ?
>
>>> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
>>> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
>>> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
2023-07-13 7:30 ` Anshuman Khandual
@ 2023-07-13 9:15 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-13 9:15 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 13/07/2023 08:30, Anshuman Khandual wrote:
>
>
> On 7/11/23 19:45, James Clark wrote:
>>
>>
>> On 11/07/2023 13:10, Anshuman Khandual wrote:
>>>
>>>
>>> On 7/10/23 17:51, James Clark wrote:
>>>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>>>> relationship between perf_event_context and PMUs has changed so that
>>>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>>>> silenced no longer exists.
>>>>
>>>> Remove the capability to avoid confusion that it actually influences
>>>> any perf core behavior. This change should be a no-op.
>>>>
>>>> Signed-off-by: James Clark <james.clark@arm.com>
>>>> ---
>>>> include/linux/perf_event.h | 2 +-
>>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
>>>> index d5628a7b5eaa..3f4d941fd6c5 100644
>>>> --- a/include/linux/perf_event.h
>>>> +++ b/include/linux/perf_event.h
>>>> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
>>>> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
>>>> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
>>>> #define PERF_PMU_CAP_ITRACE 0x0020
>>>> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
>>>> +/* Unused 0x0040 */
>>>
>>> Small nit, "Unused" marking might not be required here.
>>>
>>
>> But then it would be very easy to miss that there is a free bit if I
>> don't leave the comment. Is it really better without it?
>>
>> I could shift all the following ones down by one bit, but it would be a
>
> Sounds as a better option IMHO.
>
>> lot of work to make sure that nobody has hard coded some check for one
>> of the bits instead of using the define somewhere.
>
> These are not user visible ABI and hence defined in include/linux/perf_event.h
> to be used by drivers registering a PMU for capability enumeration. I am just
> wondering why they might have been hard coded any where ?
>
You never know, I've seen worse! I suppose we'll find out if I change it
and anything breaks. It will probably be fine though, I'll make the
change on the next version.
>>
>>>> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
>>>> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
>>>> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 4/4] perf: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability
@ 2023-07-13 9:15 ` James Clark
0 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-13 9:15 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo,
Mark Rutland, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Will Deacon, Kan Liang, linux-kernel,
linux-arm-kernel, linux-perf-users, irogers
On 13/07/2023 08:30, Anshuman Khandual wrote:
>
>
> On 7/11/23 19:45, James Clark wrote:
>>
>>
>> On 11/07/2023 13:10, Anshuman Khandual wrote:
>>>
>>>
>>> On 7/10/23 17:51, James Clark wrote:
>>>> Since commit bd2756811766 ("perf: Rewrite core context handling") the
>>>> relationship between perf_event_context and PMUs has changed so that
>>>> the error scenario that PERF_PMU_CAP_HETEROGENEOUS_CPUS originally
>>>> silenced no longer exists.
>>>>
>>>> Remove the capability to avoid confusion that it actually influences
>>>> any perf core behavior. This change should be a no-op.
>>>>
>>>> Signed-off-by: James Clark <james.clark@arm.com>
>>>> ---
>>>> include/linux/perf_event.h | 2 +-
>>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
>>>> index d5628a7b5eaa..3f4d941fd6c5 100644
>>>> --- a/include/linux/perf_event.h
>>>> +++ b/include/linux/perf_event.h
>>>> @@ -288,7 +288,7 @@ struct perf_event_pmu_context;
>>>> #define PERF_PMU_CAP_EXTENDED_REGS 0x0008
>>>> #define PERF_PMU_CAP_EXCLUSIVE 0x0010
>>>> #define PERF_PMU_CAP_ITRACE 0x0020
>>>> -#define PERF_PMU_CAP_HETEROGENEOUS_CPUS 0x0040
>>>> +/* Unused 0x0040 */
>>>
>>> Small nit, "Unused" marking might not be required here.
>>>
>>
>> But then it would be very easy to miss that there is a free bit if I
>> don't leave the comment. Is it really better without it?
>>
>> I could shift all the following ones down by one bit, but it would be a
>
> Sounds as a better option IMHO.
>
>> lot of work to make sure that nobody has hard coded some check for one
>> of the bits instead of using the define somewhere.
>
> These are not user visible ABI and hence defined in include/linux/perf_event.h
> to be used by drivers registering a PMU for capability enumeration. I am just
> wondering why they might have been hard coded any where ?
>
You never know, I've seen worse! I suppose we'll find out if I change it
and anything breaks. It will probably be fine though, I'll make the
change on the next version.
>>
>>>> #define PERF_PMU_CAP_NO_EXCLUDE 0x0080
>>>> #define PERF_PMU_CAP_AUX_OUTPUT 0x0100
>>>> #define PERF_PMU_CAP_EXTENDED_HW_TYPE 0x0200
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-11 14:12 ` James Clark
@ 2023-07-20 17:12 ` Ian Rogers
-1 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-20 17:12 UTC (permalink / raw)
To: James Clark, linux-kernel, Will Deacon, Mark Rutland, linux-arm-kernel
Cc: Anshuman Khandual, Peter Zijlstra, Ingo Molnar,
Alexander Shishkin, Jiri Olsa, Namhyung Kim, Adrian Hunter,
Thomas Gleixner, Borislav Petkov, Dave Hansen, x86,
H. Peter Anvin, Kan Liang, linux-perf-users,
Arnaldo Carvalho de Melo
On Tue, Jul 11, 2023 at 7:12 AM James Clark <james.clark@arm.com> wrote:
>
>
>
> On 11/07/2023 13:01, Anshuman Khandual wrote:
> >
> >
> > On 7/10/23 17:51, James Clark wrote:
> >> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> >> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> >> implementation is contained in the Perf core and tool code so no change
> >> to the Arm PMU driver is needed.
> >>
> >> The following basic use case now results in Perf opening the event on
> >> all PMUs rather than picking only one in an unpredictable way:
> >>
> >> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
> >>
> >> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
> >>
> >> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> >> 752745657 armv8_cortex_a53/cycles/ (94.80%)
> >>
> >> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> >> Suggested-by: Ian Rogers <irogers@google.com>
> >> Signed-off-by: James Clark <james.clark@arm.com>
Hi ARM Linux and ARM Linux PMU people,
Could this patch be picked up for Linux 6.5? I don't see it in the
tree and it seems a shame to have to wait for it. The other patches do
cleanup and so waiting for 6.6 seems okay.
Thanks,
Ian
> >> ---
> >> drivers/perf/arm_pmu.c | 7 ++++++-
> >> 1 file changed, 6 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> >> index 277e29fbd504..d8844a9461a2 100644
> >> --- a/drivers/perf/arm_pmu.c
> >> +++ b/drivers/perf/arm_pmu.c
> >> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
> >> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> >> * and we have taken ctx sharing into account (e.g. with our
> >> * pmu::filter callback and pmu::event_init group validation).
> >> + *
> >> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> >
> > s/legacy/generic ? These hardware events are still around.
>
> True, I thought I saw it mentioned that way somewhere, but I can
> probably just remove it altogether. PERF_TYPE_HARDWARE and
> PERF_TYPE_HW_CACHE is enough.
>
>
> >
> >> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> >> + * specific PMU.
> >> */
> >> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
> >> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> >> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
> >> };
> >>
> >> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-20 17:12 ` Ian Rogers
0 siblings, 0 replies; 39+ messages in thread
From: Ian Rogers @ 2023-07-20 17:12 UTC (permalink / raw)
To: James Clark, linux-kernel, Will Deacon, Mark Rutland, linux-arm-kernel
Cc: Anshuman Khandual, Peter Zijlstra, Ingo Molnar,
Alexander Shishkin, Jiri Olsa, Namhyung Kim, Adrian Hunter,
Thomas Gleixner, Borislav Petkov, Dave Hansen, x86,
H. Peter Anvin, Kan Liang, linux-perf-users,
Arnaldo Carvalho de Melo
On Tue, Jul 11, 2023 at 7:12 AM James Clark <james.clark@arm.com> wrote:
>
>
>
> On 11/07/2023 13:01, Anshuman Khandual wrote:
> >
> >
> > On 7/10/23 17:51, James Clark wrote:
> >> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> >> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> >> implementation is contained in the Perf core and tool code so no change
> >> to the Arm PMU driver is needed.
> >>
> >> The following basic use case now results in Perf opening the event on
> >> all PMUs rather than picking only one in an unpredictable way:
> >>
> >> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
> >>
> >> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
> >>
> >> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> >> 752745657 armv8_cortex_a53/cycles/ (94.80%)
> >>
> >> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> >> Suggested-by: Ian Rogers <irogers@google.com>
> >> Signed-off-by: James Clark <james.clark@arm.com>
Hi ARM Linux and ARM Linux PMU people,
Could this patch be picked up for Linux 6.5? I don't see it in the
tree and it seems a shame to have to wait for it. The other patches do
cleanup and so waiting for 6.6 seems okay.
Thanks,
Ian
> >> ---
> >> drivers/perf/arm_pmu.c | 7 ++++++-
> >> 1 file changed, 6 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
> >> index 277e29fbd504..d8844a9461a2 100644
> >> --- a/drivers/perf/arm_pmu.c
> >> +++ b/drivers/perf/arm_pmu.c
> >> @@ -875,8 +875,13 @@ struct arm_pmu *armpmu_alloc(void)
> >> * configuration (e.g. big.LITTLE). This is not an uncore PMU,
> >> * and we have taken ctx sharing into account (e.g. with our
> >> * pmu::filter callback and pmu::event_init group validation).
> >> + *
> >> + * PERF_PMU_CAP_EXTENDED_HW_TYPE is required to open the legacy
> >
> > s/legacy/generic ? These hardware events are still around.
>
> True, I thought I saw it mentioned that way somewhere, but I can
> probably just remove it altogether. PERF_TYPE_HARDWARE and
> PERF_TYPE_HW_CACHE is enough.
>
>
> >
> >> + * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE events on a
> >> + * specific PMU.
> >> */
> >> - .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS,
> >> + .capabilities = PERF_PMU_CAP_HETEROGENEOUS_CPUS | PERF_PMU_CAP_EXTENDED_REGS |
> >> + PERF_PMU_CAP_EXTENDED_HW_TYPE,
> >> };
> >>
> >> pmu->attr_groups[ARMPMU_ATTR_GROUP_COMMON] =
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-20 17:12 ` Ian Rogers
@ 2023-07-21 10:21 ` Will Deacon
-1 siblings, 0 replies; 39+ messages in thread
From: Will Deacon @ 2023-07-21 10:21 UTC (permalink / raw)
To: Ian Rogers
Cc: James Clark, linux-kernel, Mark Rutland, linux-arm-kernel,
Anshuman Khandual, Peter Zijlstra, Ingo Molnar,
Alexander Shishkin, Jiri Olsa, Namhyung Kim, Adrian Hunter,
Thomas Gleixner, Borislav Petkov, Dave Hansen, x86,
H. Peter Anvin, Kan Liang, linux-perf-users,
Arnaldo Carvalho de Melo
On Thu, Jul 20, 2023 at 10:12:21AM -0700, Ian Rogers wrote:
> On Tue, Jul 11, 2023 at 7:12 AM James Clark <james.clark@arm.com> wrote:
> >
> >
> >
> > On 11/07/2023 13:01, Anshuman Khandual wrote:
> > >
> > >
> > > On 7/10/23 17:51, James Clark wrote:
> > >> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> > >> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> > >> implementation is contained in the Perf core and tool code so no change
> > >> to the Arm PMU driver is needed.
> > >>
> > >> The following basic use case now results in Perf opening the event on
> > >> all PMUs rather than picking only one in an unpredictable way:
> > >>
> > >> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
> > >>
> > >> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
> > >>
> > >> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> > >> 752745657 armv8_cortex_a53/cycles/ (94.80%)
> > >>
> > >> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> > >> Suggested-by: Ian Rogers <irogers@google.com>
> > >> Signed-off-by: James Clark <james.clark@arm.com>
>
> Hi ARM Linux and ARM Linux PMU people,
>
> Could this patch be picked up for Linux 6.5? I don't see it in the
> tree and it seems a shame to have to wait for it. The other patches do
> cleanup and so waiting for 6.6 seems okay.
I'm only taking fixes for 6.5 and I don't think this qualifies.
If it was an oversight introduced during the recent merge window, then
I'd be happier fixing it up, but 55bcf6ef314a was merged ages ago (v5.12?),
so I think we can wait.
I'll be queuing perf changes for 6.6 next week, so I'll look at this
then.
Cheers,
Will
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
@ 2023-07-21 10:21 ` Will Deacon
0 siblings, 0 replies; 39+ messages in thread
From: Will Deacon @ 2023-07-21 10:21 UTC (permalink / raw)
To: Ian Rogers
Cc: James Clark, linux-kernel, Mark Rutland, linux-arm-kernel,
Anshuman Khandual, Peter Zijlstra, Ingo Molnar,
Alexander Shishkin, Jiri Olsa, Namhyung Kim, Adrian Hunter,
Thomas Gleixner, Borislav Petkov, Dave Hansen, x86,
H. Peter Anvin, Kan Liang, linux-perf-users,
Arnaldo Carvalho de Melo
On Thu, Jul 20, 2023 at 10:12:21AM -0700, Ian Rogers wrote:
> On Tue, Jul 11, 2023 at 7:12 AM James Clark <james.clark@arm.com> wrote:
> >
> >
> >
> > On 11/07/2023 13:01, Anshuman Khandual wrote:
> > >
> > >
> > > On 7/10/23 17:51, James Clark wrote:
> > >> This capability gives us the ability to open PERF_TYPE_HARDWARE and
> > >> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
> > >> implementation is contained in the Perf core and tool code so no change
> > >> to the Arm PMU driver is needed.
> > >>
> > >> The following basic use case now results in Perf opening the event on
> > >> all PMUs rather than picking only one in an unpredictable way:
> > >>
> > >> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
> > >>
> > >> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
> > >>
> > >> 963279620 armv8_cortex_a57/cycles/ (99.19%)
> > >> 752745657 armv8_cortex_a53/cycles/ (94.80%)
> > >>
> > >> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
> > >> Suggested-by: Ian Rogers <irogers@google.com>
> > >> Signed-off-by: James Clark <james.clark@arm.com>
>
> Hi ARM Linux and ARM Linux PMU people,
>
> Could this patch be picked up for Linux 6.5? I don't see it in the
> tree and it seems a shame to have to wait for it. The other patches do
> cleanup and so waiting for 6.6 seems okay.
I'm only taking fixes for 6.5 and I don't think this qualifies.
If it was an oversight introduced during the recent merge window, then
I'd be happier fixing it up, but 55bcf6ef314a was merged ages ago (v5.12?),
so I think we can wait.
I'll be queuing perf changes for 6.6 next week, so I'll look at this
then.
Cheers,
Will
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 39+ messages in thread
* Re: [PATCH 1/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability
2023-07-21 10:21 ` Will Deacon
(?)
@ 2023-07-24 13:46 ` James Clark
-1 siblings, 0 replies; 39+ messages in thread
From: James Clark @ 2023-07-24 13:46 UTC (permalink / raw)
To: Will Deacon, Ian Rogers, Anshuman Khandual
Cc: linux-kernel, Mark Rutland, linux-arm-kernel, Peter Zijlstra,
Ingo Molnar, Alexander Shishkin, Jiri Olsa, Namhyung Kim,
Adrian Hunter, Thomas Gleixner, Borislav Petkov, Dave Hansen,
x86, H. Peter Anvin, Kan Liang, linux-perf-users,
Arnaldo Carvalho de Melo
On 21/07/2023 11:21, Will Deacon wrote:
> On Thu, Jul 20, 2023 at 10:12:21AM -0700, Ian Rogers wrote:
>> On Tue, Jul 11, 2023 at 7:12 AM James Clark <james.clark@arm.com> wrote:
>>>
>>>
>>>
>>> On 11/07/2023 13:01, Anshuman Khandual wrote:
>>>>
>>>>
>>>> On 7/10/23 17:51, James Clark wrote:
>>>>> This capability gives us the ability to open PERF_TYPE_HARDWARE and
>>>>> PERF_TYPE_HW_CACHE events on a specific PMU for free. All the
>>>>> implementation is contained in the Perf core and tool code so no change
>>>>> to the Arm PMU driver is needed.
>>>>>
>>>>> The following basic use case now results in Perf opening the event on
>>>>> all PMUs rather than picking only one in an unpredictable way:
>>>>>
>>>>> $ perf stat -e cycles -- taskset --cpu-list 0,1 stress -c 2
>>>>>
>>>>> Performance counter stats for 'taskset --cpu-list 0,1 stress -c 2':
>>>>>
>>>>> 963279620 armv8_cortex_a57/cycles/ (99.19%)
>>>>> 752745657 armv8_cortex_a53/cycles/ (94.80%)
>>>>>
>>>>> Fixes: 55bcf6ef314a ("perf: Extend PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE")
>>>>> Suggested-by: Ian Rogers <irogers@google.com>
>>>>> Signed-off-by: James Clark <james.clark@arm.com>
>>
>> Hi ARM Linux and ARM Linux PMU people,
>>
>> Could this patch be picked up for Linux 6.5? I don't see it in the
>> tree and it seems a shame to have to wait for it. The other patches do
>> cleanup and so waiting for 6.6 seems okay.
>
> I'm only taking fixes for 6.5 and I don't think this qualifies.
>
> If it was an oversight introduced during the recent merge window, then
> I'd be happier fixing it up, but 55bcf6ef314a was merged ages ago (v5.12?),
> so I think we can wait.
>
> I'll be queuing perf changes for 6.6 next week, so I'll look at this
> then.
>
> Cheers,
>
> Will
Hi Will,
Thanks for looking at this. I've sent a v2 with Anshuman's fixes.
James
^ permalink raw reply [flat|nested] 39+ messages in thread
end of thread, other threads:[~2023-07-24 13:50 UTC | newest]
Thread overview: 39+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-07-10 12:21 [PATCH 0/4] arm_pmu: Add PERF_PMU_CAP_EXTENDED_HW_TYPE capability James Clark
2023-07-10 12:21 ` James Clark
2023-07-10 12:21 ` [PATCH 1/4] " James Clark
2023-07-10 12:21 ` James Clark
2023-07-10 16:04 ` Ian Rogers
2023-07-10 16:04 ` Ian Rogers
2023-07-11 12:01 ` Anshuman Khandual
2023-07-11 12:01 ` Anshuman Khandual
2023-07-11 14:12 ` James Clark
2023-07-11 14:12 ` James Clark
2023-07-20 17:12 ` Ian Rogers
2023-07-20 17:12 ` Ian Rogers
2023-07-21 10:21 ` Will Deacon
2023-07-21 10:21 ` Will Deacon
2023-07-24 13:46 ` James Clark
2023-07-10 12:21 ` [PATCH 2/4] perf/x86: Remove unused PERF_PMU_CAP_HETEROGENEOUS_CPUS capability James Clark
2023-07-10 12:21 ` James Clark
2023-07-10 16:07 ` Ian Rogers
2023-07-10 16:07 ` Ian Rogers
2023-07-10 12:21 ` [PATCH 3/4] arm_pmu: " James Clark
2023-07-10 12:21 ` James Clark
2023-07-10 16:08 ` Ian Rogers
2023-07-10 16:08 ` Ian Rogers
2023-07-11 12:08 ` Anshuman Khandual
2023-07-11 12:08 ` Anshuman Khandual
2023-07-11 14:04 ` James Clark
2023-07-11 14:04 ` James Clark
2023-07-10 12:21 ` [PATCH 4/4] perf: " James Clark
2023-07-10 12:21 ` James Clark
2023-07-10 16:09 ` Ian Rogers
2023-07-10 16:09 ` Ian Rogers
2023-07-11 12:10 ` Anshuman Khandual
2023-07-11 12:10 ` Anshuman Khandual
2023-07-11 14:15 ` James Clark
2023-07-11 14:15 ` James Clark
2023-07-13 7:30 ` Anshuman Khandual
2023-07-13 7:30 ` Anshuman Khandual
2023-07-13 9:15 ` James Clark
2023-07-13 9:15 ` James Clark
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.