diff mbox series

[V18,3/9] drivers: perf: arm_pmu: Add infrastructure for branch stack sampling

Message ID 20240613061731.3109448-4-anshuman.khandual@arm.com (mailing list archive)
State New, archived
Headers show
Series arm64/perf: Enable branch stack sampling | expand

Commit Message

Anshuman Khandual June 13, 2024, 6:17 a.m. UTC
In order to support the Branch Record Buffer Extension (BRBE), we need to
extend the arm_pmu framework with some basic infrastructure for branch
stack sampling which arm_pmu drivers can opt-in to using. Subsequent
patches will use this to add support for BRBE in the PMUv3 driver.

With BRBE, the hardware records branches into a hardware FIFO, which will
be sampled by software when perf events overflow. A task may be context-
switched an arbitrary number of times between overflows, and to avoid
losing samples we need to save the current records when a task is context-
switched out. To do these we'll need to use the pmu::sched_task() callback,
and we'll also need to allocate some per-task storage space via event flag
PERF_ATTACH_TASK_DATA.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
----
Changes in V18:

- Scan valid branch stack events in armpmu_start() to create merged filter
- Updated the commit message

 drivers/perf/arm_pmu.c       | 42 +++++++++++++++++++++++++++++++++---
 include/linux/perf/arm_pmu.h | 32 ++++++++++++++++++++++++++-
 2 files changed, 70 insertions(+), 4 deletions(-)

Comments

Mark Rutland June 14, 2024, 3:01 p.m. UTC | #1
On Thu, Jun 13, 2024 at 11:47:25AM +0530, Anshuman Khandual wrote:
> @@ -289,6 +289,23 @@ static void armpmu_start(struct perf_event *event, int flags)
>  {
>  	struct arm_pmu *armpmu = to_arm_pmu(event->pmu);
>  	struct hw_perf_event *hwc = &event->hw;
> +	struct pmu_hw_events *cpuc = this_cpu_ptr(armpmu->hw_events);
> +	int idx;
> +
> +	/*
> +	 * Merge all branch filter requests from different perf
> +	 * events being added into this PMU. This includes both
> +	 * privilege and branch type filters.
> +	 */
> +	if (armpmu->has_branch_stack) {
> +		cpuc->branch_sample_type = 0;
> +		for (idx = 0; idx < ARMPMU_MAX_HWEVENTS; idx++) {
> +			struct perf_event *event_idx = cpuc->events[idx];
> +
> +			if (event_idx && has_branch_stack(event_idx))
> +				cpuc->branch_sample_type |= event_idx->attr.branch_sample_type;
> +		}
> +	}

When we spoke about this, I meant that we should do this under armpmu::start(),
or a callee or caller thereof once we know all the events are configured, just
before we actually enable the PMU.

For example, this could live in armv8pmu_branch_enable(), which'd allow
all the actual logic to be added in the BRBE enablement patch.

Doing this in armpmu_start() doesn't work as well because it won't handle
events being removed.

[...]

> diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
> index b3b34f6670cf..9eda16dd684e 100644
> --- a/include/linux/perf/arm_pmu.h
> +++ b/include/linux/perf/arm_pmu.h
> @@ -46,6 +46,18 @@ static_assert((PERF_EVENT_FLAG_ARCH & ARMPMU_EVT_63BIT) == ARMPMU_EVT_63BIT);
>  	},								\
>  }
>  
> +/*
> + * Maximum branch record entries which could be processed
> + * for core perf branch stack sampling support, regardless
> + * of the hardware support available on a given ARM PMU.
> + */
> +#define MAX_BRANCH_RECORDS 64
> +
> +struct branch_records {
> +	struct perf_branch_stack	branch_stack;
> +	struct perf_branch_entry	branch_entries[MAX_BRANCH_RECORDS];
> +};
> +
>  /* The events for a given PMU register set. */
>  struct pmu_hw_events {
>  	/*
> @@ -66,6 +78,17 @@ struct pmu_hw_events {
>  	struct arm_pmu		*percpu_pmu;
>  
>  	int irq;
> +
> +	struct branch_records	*branches;
> +
> +	/* Active context for task events */
> +	void			*branch_context;

Using 'void *' here makes this harder to reason about and hides type
safety issues.

Give this a real type. IIUC it should be 'perf_event_context *'.

> +
> +	/* Active events requesting branch records */
> +	unsigned int		branch_users;
> +
> +	/* Active branch sample type filters */
> +	unsigned long		branch_sample_type;
>  };
>  
>  enum armpmu_attr_groups {
> @@ -96,8 +119,15 @@ struct arm_pmu {
>  	void		(*stop)(struct arm_pmu *);
>  	void		(*reset)(void *);
>  	int		(*map_event)(struct perf_event *event);
> +	void		(*sched_task)(struct perf_event_pmu_context *pmu_ctx, bool sched_in);
> +	bool		(*branch_stack_init)(struct perf_event *event);
> +	void		(*branch_stack_add)(struct perf_event *event, struct pmu_hw_events *cpuc);
> +	void		(*branch_stack_del)(struct perf_event *event, struct pmu_hw_events *cpuc);
> +	void		(*branch_stack_reset)(void);

The reset callback isn't used in this series; s

Subsequent patches call armv8pmu_branch_stack_reset() directly from
PMUv3 and the BRBE driver, and arm_pmu::branch_stack_reset() is never
used, so we can delete it.

Mark.
Anshuman Khandual June 17, 2024, 4:37 a.m. UTC | #2
On 6/14/24 20:31, Mark Rutland wrote:
> On Thu, Jun 13, 2024 at 11:47:25AM +0530, Anshuman Khandual wrote:
>> @@ -289,6 +289,23 @@ static void armpmu_start(struct perf_event *event, int flags)
>>  {
>>  	struct arm_pmu *armpmu = to_arm_pmu(event->pmu);
>>  	struct hw_perf_event *hwc = &event->hw;
>> +	struct pmu_hw_events *cpuc = this_cpu_ptr(armpmu->hw_events);
>> +	int idx;
>> +
>> +	/*
>> +	 * Merge all branch filter requests from different perf
>> +	 * events being added into this PMU. This includes both
>> +	 * privilege and branch type filters.
>> +	 */
>> +	if (armpmu->has_branch_stack) {
>> +		cpuc->branch_sample_type = 0;
>> +		for (idx = 0; idx < ARMPMU_MAX_HWEVENTS; idx++) {
>> +			struct perf_event *event_idx = cpuc->events[idx];
>> +
>> +			if (event_idx && has_branch_stack(event_idx))
>> +				cpuc->branch_sample_type |= event_idx->attr.branch_sample_type;
>> +		}
>> +	}
> 
> When we spoke about this, I meant that we should do this under armpmu::start(),
> or a callee or caller thereof once we know all the events are configured, just
> before we actually enable the PMU.
> 
> For example, this could live in armv8pmu_branch_enable(), which'd allow
> all the actual logic to be added in the BRBE enablement patch.
> 
> Doing this in armpmu_start() doesn't work as well because it won't handle
> events being removed.

Sure, will move this filter aggregation inside armv8pmu_branch_enable() instead
which is being added via the BRBE driver.

diff --git a/drivers/perf/arm_brbe.c b/drivers/perf/arm_brbe.c
index d795e8fd646f..9cf824bdc8b7 100644
--- a/drivers/perf/arm_brbe.c
+++ b/drivers/perf/arm_brbe.c
@@ -856,6 +856,22 @@ void armv8pmu_branch_enable(struct arm_pmu *arm_pmu)
 {
        struct pmu_hw_events *cpuc = this_cpu_ptr(arm_pmu->hw_events);
        u64 brbfcr, brbcr;
+       int idx;
+
+       /*
+        * Merge all branch filter requests from different perf
+        * events being added into this PMU. This includes both
+        * privilege and branch type filters.
+        */
+       if (arm_pmu->has_branch_stack) {
+               cpuc->branch_sample_type = 0;
+               for (idx = 0; idx < ARMPMU_MAX_HWEVENTS; idx++) {
+                       struct perf_event *event_idx = cpuc->events[idx];
+
+                       if (event_idx && has_branch_stack(event_idx))
+                               cpuc->branch_sample_type |= event_idx->attr.branch_sample_type;
+               }
+       }
 
        if (!(cpuc->branch_sample_type && cpuc->branch_users))
                return;

> 
> [...]
> 
>> diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
>> index b3b34f6670cf..9eda16dd684e 100644
>> --- a/include/linux/perf/arm_pmu.h
>> +++ b/include/linux/perf/arm_pmu.h
>> @@ -46,6 +46,18 @@ static_assert((PERF_EVENT_FLAG_ARCH & ARMPMU_EVT_63BIT) == ARMPMU_EVT_63BIT);
>>  	},								\
>>  }
>>  
>> +/*
>> + * Maximum branch record entries which could be processed
>> + * for core perf branch stack sampling support, regardless
>> + * of the hardware support available on a given ARM PMU.
>> + */
>> +#define MAX_BRANCH_RECORDS 64
>> +
>> +struct branch_records {
>> +	struct perf_branch_stack	branch_stack;
>> +	struct perf_branch_entry	branch_entries[MAX_BRANCH_RECORDS];
>> +};
>> +
>>  /* The events for a given PMU register set. */
>>  struct pmu_hw_events {
>>  	/*
>> @@ -66,6 +78,17 @@ struct pmu_hw_events {
>>  	struct arm_pmu		*percpu_pmu;
>>  
>>  	int irq;
>> +
>> +	struct branch_records	*branches;
>> +
>> +	/* Active context for task events */
>> +	void			*branch_context;
> 
> Using 'void *' here makes this harder to reason about and hides type
> safety issues.
> 
> Give this a real type. IIUC it should be 'perf_event_context *'.

Sure, will change the type.

> 
>> +
>> +	/* Active events requesting branch records */
>> +	unsigned int		branch_users;
>> +
>> +	/* Active branch sample type filters */
>> +	unsigned long		branch_sample_type;
>>  };
>>  
>>  enum armpmu_attr_groups {
>> @@ -96,8 +119,15 @@ struct arm_pmu {
>>  	void		(*stop)(struct arm_pmu *);
>>  	void		(*reset)(void *);
>>  	int		(*map_event)(struct perf_event *event);
>> +	void		(*sched_task)(struct perf_event_pmu_context *pmu_ctx, bool sched_in);
>> +	bool		(*branch_stack_init)(struct perf_event *event);
>> +	void		(*branch_stack_add)(struct perf_event *event, struct pmu_hw_events *cpuc);
>> +	void		(*branch_stack_del)(struct perf_event *event, struct pmu_hw_events *cpuc);
>> +	void		(*branch_stack_reset)(void);
> 
> The reset callback isn't used in this series; s
> 
> Subsequent patches call armv8pmu_branch_stack_reset() directly from
> PMUv3 and the BRBE driver, and arm_pmu::branch_stack_reset() is never
> used, so we can delete it.

Sure, will drop branch_stack_reset() callback.
diff mbox series

Patch

diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c
index 8458fe2cebb4..219c1e276327 100644
--- a/drivers/perf/arm_pmu.c
+++ b/drivers/perf/arm_pmu.c
@@ -289,6 +289,23 @@  static void armpmu_start(struct perf_event *event, int flags)
 {
 	struct arm_pmu *armpmu = to_arm_pmu(event->pmu);
 	struct hw_perf_event *hwc = &event->hw;
+	struct pmu_hw_events *cpuc = this_cpu_ptr(armpmu->hw_events);
+	int idx;
+
+	/*
+	 * Merge all branch filter requests from different perf
+	 * events being added into this PMU. This includes both
+	 * privilege and branch type filters.
+	 */
+	if (armpmu->has_branch_stack) {
+		cpuc->branch_sample_type = 0;
+		for (idx = 0; idx < ARMPMU_MAX_HWEVENTS; idx++) {
+			struct perf_event *event_idx = cpuc->events[idx];
+
+			if (event_idx && has_branch_stack(event_idx))
+				cpuc->branch_sample_type |= event_idx->attr.branch_sample_type;
+		}
+	}
 
 	/*
 	 * ARM pmu always has to reprogram the period, so ignore
@@ -317,6 +334,9 @@  armpmu_del(struct perf_event *event, int flags)
 	struct hw_perf_event *hwc = &event->hw;
 	int idx = hwc->idx;
 
+	if (has_branch_stack(event))
+		armpmu->branch_stack_del(event, hw_events);
+
 	armpmu_stop(event, PERF_EF_UPDATE);
 	hw_events->events[idx] = NULL;
 	armpmu->clear_event_idx(hw_events, event);
@@ -342,6 +362,9 @@  armpmu_add(struct perf_event *event, int flags)
 	if (idx < 0)
 		return idx;
 
+	if (has_branch_stack(event))
+		armpmu->branch_stack_add(event, hw_events);
+
 	/*
 	 * If there is an event in the counter we are going to use then make
 	 * sure it is disabled.
@@ -511,13 +534,25 @@  static int armpmu_event_init(struct perf_event *event)
 		!cpumask_test_cpu(event->cpu, &armpmu->supported_cpus))
 		return -ENOENT;
 
-	/* does not support taken branch sampling */
-	if (has_branch_stack(event))
-		return -EOPNOTSUPP;
+	if (has_branch_stack(event)) {
+		if (!armpmu->has_branch_stack)
+			return -EOPNOTSUPP;
+
+		if (!armpmu->branch_stack_init(event))
+			return -EOPNOTSUPP;
+	}
 
 	return __hw_perf_event_init(event);
 }
 
+static void armpmu_sched_task(struct perf_event_pmu_context *pmu_ctx, bool sched_in)
+{
+	struct arm_pmu *armpmu = to_arm_pmu(pmu_ctx->pmu);
+
+	if (armpmu->sched_task)
+		armpmu->sched_task(pmu_ctx, sched_in);
+}
+
 static void armpmu_enable(struct pmu *pmu)
 {
 	struct arm_pmu *armpmu = to_arm_pmu(pmu);
@@ -864,6 +899,7 @@  struct arm_pmu *armpmu_alloc(void)
 	}
 
 	pmu->pmu = (struct pmu) {
+		.sched_task	= armpmu_sched_task,
 		.pmu_enable	= armpmu_enable,
 		.pmu_disable	= armpmu_disable,
 		.event_init	= armpmu_event_init,
diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
index b3b34f6670cf..9eda16dd684e 100644
--- a/include/linux/perf/arm_pmu.h
+++ b/include/linux/perf/arm_pmu.h
@@ -46,6 +46,18 @@  static_assert((PERF_EVENT_FLAG_ARCH & ARMPMU_EVT_63BIT) == ARMPMU_EVT_63BIT);
 	},								\
 }
 
+/*
+ * Maximum branch record entries which could be processed
+ * for core perf branch stack sampling support, regardless
+ * of the hardware support available on a given ARM PMU.
+ */
+#define MAX_BRANCH_RECORDS 64
+
+struct branch_records {
+	struct perf_branch_stack	branch_stack;
+	struct perf_branch_entry	branch_entries[MAX_BRANCH_RECORDS];
+};
+
 /* The events for a given PMU register set. */
 struct pmu_hw_events {
 	/*
@@ -66,6 +78,17 @@  struct pmu_hw_events {
 	struct arm_pmu		*percpu_pmu;
 
 	int irq;
+
+	struct branch_records	*branches;
+
+	/* Active context for task events */
+	void			*branch_context;
+
+	/* Active events requesting branch records */
+	unsigned int		branch_users;
+
+	/* Active branch sample type filters */
+	unsigned long		branch_sample_type;
 };
 
 enum armpmu_attr_groups {
@@ -96,8 +119,15 @@  struct arm_pmu {
 	void		(*stop)(struct arm_pmu *);
 	void		(*reset)(void *);
 	int		(*map_event)(struct perf_event *event);
+	void		(*sched_task)(struct perf_event_pmu_context *pmu_ctx, bool sched_in);
+	bool		(*branch_stack_init)(struct perf_event *event);
+	void		(*branch_stack_add)(struct perf_event *event, struct pmu_hw_events *cpuc);
+	void		(*branch_stack_del)(struct perf_event *event, struct pmu_hw_events *cpuc);
+	void		(*branch_stack_reset)(void);
 	int		num_events;
-	bool		secure_access; /* 32-bit ARM only */
+	unsigned int	secure_access	: 1, /* 32-bit ARM only */
+			has_branch_stack: 1, /* 64-bit ARM only */
+			reserved	: 30;
 #define ARMV8_PMUV3_MAX_COMMON_EVENTS		0x40
 	DECLARE_BITMAP(pmceid_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
 #define ARMV8_PMUV3_EXT_COMMON_EVENT_BASE	0x4000