Re: [PATCH v6 2/3] arm64: perf: Add support for event counting threshold

From: Anshuman Khandual
Date: Mon Nov 27 2023 - 00:34:09 EST




On 11/24/23 15:58, James Clark wrote:
> FEAT_PMUv3_TH (Armv8.8) permits a PMU counter to increment only on
> events whose count meets a specified threshold condition. For example if
> PMEVTYPERn.TC (Threshold Control) is set to 0b101 (Greater than or
> equal, count), and the threshold is set to 2, then the PMU counter will
> now only increment by 1 when an event would have previously incremented
> the PMU counter by 2 or more on a single processor cycle.
>
> Three new Perf event config fields, 'threshold', 'threshold_compare' and
> 'threshold_count' have been added to control the feature.
> threshold_compare maps to the upper two bits of PMEVTYPERn.TC and
> threshold_count maps to the first bit of TC. These separate attributes
> have been picked rather than enumerating all the possible combinations
> of the TC field as in the Arm ARM. The attributes would be used on a
> Perf command line like this:
>
> $ perf stat -e stall_slot/threshold=2,threshold_compare=2/
>
> A new capability for reading out the maximum supported threshold value
> has also been added:
>
> $ cat /sys/bus/event_source/devices/armv8_pmuv3/caps/threshold_max
>
> 0x000000ff
>
> If a threshold higher than threshold_max is provided, then no error is
> generated but the threshold is clamped to the max value. If
> FEAT_PMUv3_TH isn't implemented or a 32 bit kernel is running, then
> threshold_max reads zero, and neither the 'threshold' nor
> 'threshold_control' parameters will be used.
>
> The threshold is per PMU counter, and there are potentially different
> threshold_max values per PMU type on heterogeneous systems.
>
> Bits higher than 32 now need to be written into PMEVTYPER, so
> armv8pmu_write_evtype() has to be updated to take an unsigned long value
> rather than u32 which gives the correct behavior on both aarch32 and 64.
>
> Reviewed-by: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
> Signed-off-by: James Clark <james.clark@xxxxxxx>
> ---
> drivers/perf/arm_pmuv3.c | 84 +++++++++++++++++++++++++++++++++-
> include/linux/perf/arm_pmuv3.h | 1 +
> 2 files changed, 84 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
> index 1d40d794f5e4..eb1ef84e1dbb 100644
> --- a/drivers/perf/arm_pmuv3.c
> +++ b/drivers/perf/arm_pmuv3.c
> @@ -15,6 +15,7 @@
> #include <clocksource/arm_arch_timer.h>
>
> #include <linux/acpi.h>
> +#include <linux/bitfield.h>
> #include <linux/clocksource.h>
> #include <linux/of.h>
> #include <linux/perf/arm_pmu.h>
> @@ -294,9 +295,20 @@ static const struct attribute_group armv8_pmuv3_events_attr_group = {
> .is_visible = armv8pmu_event_attr_is_visible,
> };
>
> +#define THRESHOLD_LOW 2
> +#define THRESHOLD_HIGH 13
> +#define THRESHOLD_CNT 14
> +#define THRESHOLD_CMP_LO 15
> +#define THRESHOLD_CMP_HI 16
> +
> PMU_FORMAT_ATTR(event, "config:0-15");
> PMU_FORMAT_ATTR(long, "config1:0");
> PMU_FORMAT_ATTR(rdpmc, "config1:1");
> +PMU_FORMAT_ATTR(threshold, "config1:" __stringify(THRESHOLD_LOW) "-"
> + __stringify(THRESHOLD_HIGH));
> +PMU_FORMAT_ATTR(threshold_compare, "config1:" __stringify(THRESHOLD_CMP_LO) "-"
> + __stringify(THRESHOLD_CMP_HI));
> +PMU_FORMAT_ATTR(threshold_count, "config1:" __stringify(THRESHOLD_CNT));

Small nit - could this be formatted better ? Is not that the column could go
upto 100 without setting off checkpatch.pl warning these days ?

>
> static int sysctl_perf_user_access __read_mostly;
>
> @@ -310,10 +322,33 @@ static inline bool armv8pmu_event_want_user_access(struct perf_event *event)
> return event->attr.config1 & 0x2;
> }
>
> +static inline u32 armv8pmu_event_threshold(struct perf_event_attr *attr)
> +{
> + return FIELD_GET(GENMASK(THRESHOLD_HIGH, THRESHOLD_LOW), attr->config1);
> +}
> +
> +static inline u8 armv8pmu_event_threshold_control(struct perf_event_attr *attr)
> +{
> + u8 th_compare = FIELD_GET(GENMASK(THRESHOLD_CMP_HI, THRESHOLD_CMP_LO),
> + attr->config1);

Ditto

> + u8 th_count = FIELD_GET(BIT(THRESHOLD_CNT), attr->config1);
> +
> + /*
> + * The count bit is always the bottom bit of the full control field, and
> + * the comparison is the upper two bits, but it's not explicitly
> + * labelled in the Arm ARM. For the Perf interface we split it into two
> + * fields, so reconstruct it here.
> + */
> + return (th_compare << 1) | th_count;
> +}
> +
> static struct attribute *armv8_pmuv3_format_attrs[] = {
> &format_attr_event.attr,
> &format_attr_long.attr,
> &format_attr_rdpmc.attr,
> + &format_attr_threshold.attr,
> + &format_attr_threshold_compare.attr,
> + &format_attr_threshold_count.attr,
> NULL,
> };
>
> @@ -365,10 +400,38 @@ static ssize_t bus_width_show(struct device *dev, struct device_attribute *attr,
>
> static DEVICE_ATTR_RO(bus_width);
>
> +static u32 threshold_max(struct arm_pmu *cpu_pmu)
> +{
> + /*
> + * PMMIR.THWIDTH is readable and non-zero on aarch32, but it would be
> + * impossible to write the threshold in the upper 32 bits of PMEVTYPER.
> + */
> + if (IS_ENABLED(CONFIG_ARM))
> + return 0;
> +
> + /*
> + * The largest value that can be written to PMEVTYPER<n>_EL0.TH is
> + * (2 ^ PMMIR.THWIDTH) - 1.
> + */
> + return (1 << FIELD_GET(ARMV8_PMU_THWIDTH, cpu_pmu->reg_pmmir)) - 1;
> +}
> +
> +static ssize_t threshold_max_show(struct device *dev,
> + struct device_attribute *attr, char *page)
> +{
> + struct pmu *pmu = dev_get_drvdata(dev);
> + struct arm_pmu *cpu_pmu = container_of(pmu, struct arm_pmu, pmu);
> +
> + return sysfs_emit(page, "0x%08x\n", threshold_max(cpu_pmu));
> +}
> +
> +static DEVICE_ATTR_RO(threshold_max);
> +
> static struct attribute *armv8_pmuv3_caps_attrs[] = {
> &dev_attr_slots.attr,
> &dev_attr_bus_slots.attr,
> &dev_attr_bus_width.attr,
> + &dev_attr_threshold_max.attr,
> NULL,
> };
>
> @@ -552,7 +615,7 @@ static void armv8pmu_write_counter(struct perf_event *event, u64 value)
> armv8pmu_write_hw_counter(event, value);
> }
>
> -static inline void armv8pmu_write_evtype(int idx, u32 val)
> +static inline void armv8pmu_write_evtype(int idx, unsigned long val)
> {
> u32 counter = ARMV8_IDX_TO_COUNTER(idx);
> unsigned long mask = ARMV8_PMU_EVTYPE_EVENT |
> @@ -921,6 +984,10 @@ static int armv8pmu_set_event_filter(struct hw_perf_event *event,
> struct perf_event_attr *attr)
> {
> unsigned long config_base = 0;
> + struct perf_event *perf_event = container_of(attr, struct perf_event,
> + attr);

Ditto

> + struct arm_pmu *cpu_pmu = to_arm_pmu(perf_event->pmu);
> + u32 th, th_max;
>
> if (attr->exclude_idle)
> return -EPERM;
> @@ -952,6 +1019,21 @@ static int armv8pmu_set_event_filter(struct hw_perf_event *event,
> if (attr->exclude_user)
> config_base |= ARMV8_PMU_EXCLUDE_EL0;
>
> + /*
> + * Insert event counting threshold (FEAT_PMUv3_TH) values. If
> + * FEAT_PMUv3_TH isn't implemented, then THWIDTH (threshold_max) will be
> + * 0 and no values will be written.
> + */
> + th_max = threshold_max(cpu_pmu);
> + if (IS_ENABLED(CONFIG_ARM64) && th_max) {
> + th = min(armv8pmu_event_threshold(attr), th_max);
> + if (th) {
> + config_base |= FIELD_PREP(ARMV8_PMU_EVTYPE_TH, th);
> + config_base |= FIELD_PREP(ARMV8_PMU_EVTYPE_TC,
> + armv8pmu_event_threshold_control(attr));

Ditto. As mentioned earlier this could have been avoided using a local variable.

> + }
> + }
> +
> /*
> * Install the filter into config_base as this is used to
> * construct the event type.
> diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
> index ddd1fec86739..ccbc0f9a74d8 100644
> --- a/include/linux/perf/arm_pmuv3.h
> +++ b/include/linux/perf/arm_pmuv3.h
> @@ -258,6 +258,7 @@
> #define ARMV8_PMU_BUS_SLOTS_MASK 0xff
> #define ARMV8_PMU_BUS_WIDTH_SHIFT 16
> #define ARMV8_PMU_BUS_WIDTH_MASK 0xf
> +#define ARMV8_PMU_THWIDTH GENMASK(23, 20)
>
> /*
> * This code is really good

Otherwise LGTM

Reviewed-by: Anshuman Khandual <anshuman.khandual@xxxxxxx>