summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--arch/x86/events/core.c10
-rw-r--r--arch/x86/events/intel/core.c63
-rw-r--r--arch/x86/events/perf_event.h3
-rw-r--r--arch/x86/include/asm/msr-index.h2
-rw-r--r--kernel/events/core.c30
5 files changed, 93 insertions, 15 deletions
diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 580b60f5ac83..e6f5e4b163ac 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -1750,6 +1750,8 @@ ssize_t x86_event_sysfs_show(char *page, u64 config, u64 event)
return ret;
}
+static struct attribute_group x86_pmu_attr_group;
+
static int __init init_hw_perf_events(void)
{
struct x86_pmu_quirk *quirk;
@@ -1813,6 +1815,14 @@ static int __init init_hw_perf_events(void)
x86_pmu_events_group.attrs = tmp;
}
+ if (x86_pmu.attrs) {
+ struct attribute **tmp;
+
+ tmp = merge_attr(x86_pmu_attr_group.attrs, x86_pmu.attrs);
+ if (!WARN_ON(!tmp))
+ x86_pmu_attr_group.attrs = tmp;
+ }
+
pr_info("... version: %d\n", x86_pmu.version);
pr_info("... bit width: %d\n", x86_pmu.cntval_bits);
pr_info("... generic registers: %d\n", x86_pmu.num_counters);
diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index a6d91d4e37a1..da9047eec7ba 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -3160,6 +3160,19 @@ err:
return -ENOMEM;
}
+static void flip_smm_bit(void *data)
+{
+ unsigned long set = *(unsigned long *)data;
+
+ if (set > 0) {
+ msr_set_bit(MSR_IA32_DEBUGCTLMSR,
+ DEBUGCTLMSR_FREEZE_IN_SMM_BIT);
+ } else {
+ msr_clear_bit(MSR_IA32_DEBUGCTLMSR,
+ DEBUGCTLMSR_FREEZE_IN_SMM_BIT);
+ }
+}
+
static void intel_pmu_cpu_starting(int cpu)
{
struct cpu_hw_events *cpuc = &per_cpu(cpu_hw_events, cpu);
@@ -3174,6 +3187,8 @@ static void intel_pmu_cpu_starting(int cpu)
cpuc->lbr_sel = NULL;
+ flip_smm_bit(&x86_pmu.attr_freeze_on_smi);
+
if (!cpuc->shared_regs)
return;
@@ -3595,6 +3610,52 @@ static struct attribute *hsw_events_attrs[] = {
NULL
};
+static ssize_t freeze_on_smi_show(struct device *cdev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ return sprintf(buf, "%lu\n", x86_pmu.attr_freeze_on_smi);
+}
+
+static DEFINE_MUTEX(freeze_on_smi_mutex);
+
+static ssize_t freeze_on_smi_store(struct device *cdev,
+ struct device_attribute *attr,
+ const char *buf, size_t count)
+{
+ unsigned long val;
+ ssize_t ret;
+
+ ret = kstrtoul(buf, 0, &val);
+ if (ret)
+ return ret;
+
+ if (val > 1)
+ return -EINVAL;
+
+ mutex_lock(&freeze_on_smi_mutex);
+
+ if (x86_pmu.attr_freeze_on_smi == val)
+ goto done;
+
+ x86_pmu.attr_freeze_on_smi = val;
+
+ get_online_cpus();
+ on_each_cpu(flip_smm_bit, &val, 1);
+ put_online_cpus();
+done:
+ mutex_unlock(&freeze_on_smi_mutex);
+
+ return count;
+}
+
+static DEVICE_ATTR_RW(freeze_on_smi);
+
+static struct attribute *intel_pmu_attrs[] = {
+ &dev_attr_freeze_on_smi.attr,
+ NULL,
+};
+
__init int intel_pmu_init(void)
{
union cpuid10_edx edx;
@@ -3641,6 +3702,8 @@ __init int intel_pmu_init(void)
x86_pmu.max_pebs_events = min_t(unsigned, MAX_PEBS_EVENTS, x86_pmu.num_counters);
+
+ x86_pmu.attrs = intel_pmu_attrs;
/*
* Quirk: v2 perfmon does not report fixed-purpose events, so
* assume at least 3 events, when not running in a hypervisor:
diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
index be3d36254040..53728eea1bed 100644
--- a/arch/x86/events/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -562,6 +562,9 @@ struct x86_pmu {
ssize_t (*events_sysfs_show)(char *page, u64 config);
struct attribute **cpu_events;
+ unsigned long attr_freeze_on_smi;
+ struct attribute **attrs;
+
/*
* CPU Hotplug hooks
*/
diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h
index 673f9ac50f6d..18b162322eff 100644
--- a/arch/x86/include/asm/msr-index.h
+++ b/arch/x86/include/asm/msr-index.h
@@ -137,6 +137,8 @@
#define DEBUGCTLMSR_BTS_OFF_OS (1UL << 9)
#define DEBUGCTLMSR_BTS_OFF_USR (1UL << 10)
#define DEBUGCTLMSR_FREEZE_LBRS_ON_PMI (1UL << 11)
+#define DEBUGCTLMSR_FREEZE_IN_SMM_BIT 14
+#define DEBUGCTLMSR_FREEZE_IN_SMM (1UL << DEBUGCTLMSR_FREEZE_IN_SMM_BIT)
#define MSR_PEBS_FRONTEND 0x000003f7
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 6c4e523dc1e2..bc63f8db1b0d 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -925,11 +925,6 @@ static inline int is_cgroup_event(struct perf_event *event)
return 0;
}
-static inline u64 perf_cgroup_event_cgrp_time(struct perf_event *event)
-{
- return 0;
-}
-
static inline void update_cgrp_time_from_event(struct perf_event *event)
{
}
@@ -5729,9 +5724,6 @@ static void perf_output_read_one(struct perf_output_handle *handle,
__output_copy(handle, values, n * sizeof(u64));
}
-/*
- * XXX PERF_FORMAT_GROUP vs inherited events seems difficult.
- */
static void perf_output_read_group(struct perf_output_handle *handle,
struct perf_event *event,
u64 enabled, u64 running)
@@ -5776,6 +5768,13 @@ static void perf_output_read_group(struct perf_output_handle *handle,
#define PERF_FORMAT_TOTAL_TIMES (PERF_FORMAT_TOTAL_TIME_ENABLED|\
PERF_FORMAT_TOTAL_TIME_RUNNING)
+/*
+ * XXX PERF_SAMPLE_READ vs inherited events seems difficult.
+ *
+ * The problem is that its both hard and excessively expensive to iterate the
+ * child list, not to mention that its impossible to IPI the children running
+ * on another CPU, from interrupt/NMI context.
+ */
static void perf_output_read(struct perf_output_handle *handle,
struct perf_event *event)
{
@@ -9193,7 +9192,7 @@ static int perf_try_init_event(struct pmu *pmu, struct perf_event *event)
static struct pmu *perf_init_event(struct perf_event *event)
{
- struct pmu *pmu = NULL;
+ struct pmu *pmu;
int idx;
int ret;
@@ -9462,9 +9461,10 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu,
local64_set(&hwc->period_left, hwc->sample_period);
/*
- * we currently do not support PERF_FORMAT_GROUP on inherited events
+ * We currently do not support PERF_SAMPLE_READ on inherited events.
+ * See perf_output_read().
*/
- if (attr->inherit && (attr->read_format & PERF_FORMAT_GROUP))
+ if (attr->inherit && (attr->sample_type & PERF_SAMPLE_READ))
goto err_ns;
if (!has_branch_stack(event))
@@ -9477,9 +9477,7 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu,
}
pmu = perf_init_event(event);
- if (!pmu)
- goto err_ns;
- else if (IS_ERR(pmu)) {
+ if (IS_ERR(pmu)) {
err = PTR_ERR(pmu);
goto err_ns;
}
@@ -9492,8 +9490,10 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu,
event->addr_filters_offs = kcalloc(pmu->nr_addr_filters,
sizeof(unsigned long),
GFP_KERNEL);
- if (!event->addr_filters_offs)
+ if (!event->addr_filters_offs) {
+ err = -ENOMEM;
goto err_per_task;
+ }
/* force hw sync on the address filters */
event->addr_filters_gen = 1;
generated by cgit v1.2.3 (git 2.46.0) at 2025-07-18 11:01:00 +0000