From: Robert Richter robert.richter@linaro.org
Expose persistent events in the system to userland using sysfs. Perf tools are able to read existing pmu events from sysfs. Now we use a persistent pmu as an event container containing all registered persistent events of the system. This patch adds dynamically registration of persistent events to sysfs. E.g. something like this:
/sys/bus/event_source/devices/persistent/events/mce_record:persistent,config=106 /sys/bus/event_source/devices/persistent/format/persistent:attr5:23
Perf tools need to support the attr<num> syntax that is added in a separate patch set. With it we are able to run perf tool commands to read persistent events, e.g.:
# perf record -e persistent/mce_record/ sleep 10 # perf top -e persistent/mce_record/
[ Document attr<index> syntax in sysfs ABI ] Reported-by: Jiri Olsa jolsa@redhat.com Signed-off-by: Robert Richter robert.richter@linaro.org Signed-off-by: Robert Richter rric@kernel.org --- .../testing/sysfs-bus-event_source-devices-format | 43 ++++++++++++----- kernel/events/persistent.c | 55 +++++++++++++++++++++- 2 files changed, 86 insertions(+), 12 deletions(-)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-format b/Documentation/ABI/testing/sysfs-bus-event_source-devices-format index 77f47ff..47b7353 100644 --- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-format +++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-format @@ -1,13 +1,14 @@ -Where: /sys/bus/event_source/devices/<dev>/format +Where: /sys/bus/event_source/devices/<pmu>/format/<name> Date: January 2012 -Kernel Version: 3.3 +Kernel Version: 3.3 + 3.12 (added attr<index>:<bits>) Contact: Jiri Olsa jolsa@redhat.com -Description: - Attribute group to describe the magic bits that go into - perf_event_attr::config[012] for a particular pmu. - Each attribute of this group defines the 'hardware' bitmask - we want to export, so that userspace can deal with sane - name/value pairs. + +Description: Define formats for bit ranges in perf_event_attr + + Attribute group to describe the magic bits that go + into struct perf_event_attr for a particular pmu. Bit + range may be any bit mask of an u64 (bits 0 to 63).
Userspace must be prepared for the possibility that attributes define overlapping bit ranges. For example: @@ -15,6 +16,26 @@ Description: attr2 = 'config:0-7' attr3 = 'config:12-35'
- Example: 'config1:1,6-10,44' - Defines contents of attribute that occupies bits 1,6-10,44 of - perf_event_attr::config1. + Syntax Description + + config[012]*:<bits> Each attribute of this group + defines the 'hardware' bitmask + we want to export, so that + userspace can deal with sane + name/value pairs. + + attr<index>:<bits> Set any field of the event + attribute. The index is a + decimal number that specifies + the u64 value to be set within + struct perf_event_attr. + + Examples: + + 'config1:1,6-10,44' Defines contents of attribute + that occupies bits 1,6-10,44 + of perf_event_attr::config1. + + 'attr5:23' Define the persistent event + flag (bit 23 of the attribute + flags) diff --git a/kernel/events/persistent.c b/kernel/events/persistent.c index 96201c1..8be7c05 100644 --- a/kernel/events/persistent.c +++ b/kernel/events/persistent.c @@ -17,8 +17,10 @@ struct pers_event_desc { struct pers_event { char *name; struct perf_event_attr attr; + struct perf_pmu_events_attr sysfs; };
+static struct pmu persistent_pmu; static DEFINE_PER_CPU(struct list_head, pers_events); static DEFINE_PER_CPU(struct mutex, pers_events_lock);
@@ -137,6 +139,8 @@ unwind: return PTR_ERR(event); }
+static int pers_event_sysfs_register(struct pers_event *event); + int perf_add_persistent_event_by_id(char* name, int id) { struct pers_event *event; @@ -150,6 +154,8 @@ int perf_add_persistent_event_by_id(char* name, int id) if (!event->name) goto fail;
+ event->sysfs.id = id; + attr = &event->attr; attr->sample_period = 1; attr->wakeup_events = 1; @@ -163,6 +169,8 @@ int perf_add_persistent_event_by_id(char* name, int id) if (ret) goto fail;
+ pers_event_sysfs_register(event); + return 0; fail: kfree(event->name); @@ -207,12 +215,57 @@ static struct attribute_group persistent_format_group = { .attrs = persistent_format_attrs, };
+#define MAX_EVENTS 16 + +static struct attribute *persistent_events_attr[MAX_EVENTS + 1] = { }; + +static struct attribute_group persistent_events_group = { + .name = "events", + .attrs = persistent_events_attr, +}; + static const struct attribute_group *persistent_attr_groups[] = { &persistent_format_group, + NULL, /* placeholder: &persistent_events_group */ NULL, }; +#define EVENTS_GROUP (persistent_attr_groups[1])
-static struct pmu persistent_pmu; +static ssize_t pers_event_sysfs_show(struct device *dev, + struct device_attribute *__attr, char *page) +{ + struct perf_pmu_events_attr *attr = + container_of(__attr, struct perf_pmu_events_attr, attr); + return sprintf(page, "persistent,config=%lld", + (unsigned long long)attr->id); +} + +static int pers_event_sysfs_register(struct pers_event *event) +{ + struct device_attribute *attr = &event->sysfs.attr; + int idx; + + *attr = (struct device_attribute)__ATTR(, 0444, pers_event_sysfs_show, + NULL); + attr->attr.name = event->name; + + /* add sysfs attr to events: */ + for (idx = 0; idx < MAX_EVENTS; idx++) { + if (!cmpxchg(persistent_events_attr + idx, NULL, &attr->attr)) + break; + } + + if (idx >= MAX_EVENTS) + return -ENOSPC; + if (!idx) + EVENTS_GROUP = &persistent_events_group; + if (!persistent_pmu.dev) + return 0; /* sysfs not yet initialized */ + if (idx) + return sysfs_update_group(&persistent_pmu.dev->kobj, + EVENTS_GROUP); + return sysfs_create_group(&persistent_pmu.dev->kobj, EVENTS_GROUP); +}
static int persistent_pmu_init(struct perf_event *event) {