Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Song Liu <songliubraving@fb.com>
To:     <bpf@vger.kernel.org>, <linux-kernel@vger.kernel.org>
CC:     <acme@kernel.org>, <peterz@infradead.org>, <mingo@redhat.com>,
        <kjain@linux.ibm.com>, <kernel-team@fb.com>,
        Song Liu <songliubraving@fb.com>
Subject: [PATCH v4 bpf-next 1/3] perf: enable branch record for software events
Date:   Tue, 31 Aug 2021 17:35:15 -0700
Message-ID: <20210901003517.3953145-2-songliubraving@fb.com>
In-Reply-To: <20210901003517.3953145-1-songliubraving@fb.com>
References: <20210901003517.3953145-1-songliubraving@fb.com>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain
Precedence: bulk

The typical way to access branch record (e.g. Intel LBR) is via hardware
perf_event. For CPUs with FREEZE_LBRS_ON_PMI support, PMI could capture
reliable LBR. On the other hand, LBR could also be useful in non-PMI
scenario. For example, in kretprobe or bpf fexit program, LBR could
provide a lot of information on what happened with the function. Add API
to use branch record for software use.

Note that, when the software event triggers, it is necessary to stop the
branch record hardware asap. Therefore, static_call is used to remove som=
e
branch instructions in this process.

Signed-off-by: Song Liu <songliubraving@fb.com>
---
 arch/x86/events/intel/core.c | 26 +++++++++++++++++++++++---
 arch/x86/events/intel/ds.c   |  8 --------
 arch/x86/events/perf_event.h | 10 ++++++++--
 include/linux/perf_event.h   | 26 ++++++++++++++++++++++++++
 kernel/events/core.c         |  2 ++
 5 files changed, 59 insertions(+), 13 deletions(-)

diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index ac6fd2dabf6a2..fe9bec93eb53b 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -2143,7 +2143,7 @@ static __initconst const u64 knl_hw_cache_extra_reg=
s
  * However, there are some cases which may change PEBS status, e.g. PMI
  * throttle. The PEBS_ENABLE should be updated where the status changes.
  */
-static void __intel_pmu_disable_all(void)
+static __always_inline void __intel_pmu_disable_all(void)
 {
 	struct cpu_hw_events *cpuc =3D this_cpu_ptr(&cpu_hw_events);
=20
@@ -2153,7 +2153,7 @@ static void __intel_pmu_disable_all(void)
 		intel_pmu_disable_bts();
 }
=20
-static void intel_pmu_disable_all(void)
+static __always_inline void intel_pmu_disable_all(void)
 {
 	__intel_pmu_disable_all();
 	intel_pmu_pebs_disable_all();
@@ -2186,6 +2186,20 @@ static void intel_pmu_enable_all(int added)
 	__intel_pmu_enable_all(added, false);
 }
=20
+static int
+intel_pmu_snapshot_branch_stack(struct perf_branch_snapshot *br_snapshot=
)
+{
+	struct cpu_hw_events *cpuc =3D this_cpu_ptr(&cpu_hw_events);
+
+	intel_pmu_disable_all();
+	intel_pmu_lbr_read();
+	memcpy(br_snapshot->entries, cpuc->lbr_entries,
+	       sizeof(struct perf_branch_entry) * x86_pmu.lbr_nr);
+	br_snapshot->nr =3D x86_pmu.lbr_nr;
+	intel_pmu_enable_all(0);
+	return 0;
+}
+
 /*
  * Workaround for:
  *   Intel Errata AAK100 (model 26)
@@ -6283,9 +6297,15 @@ __init int intel_pmu_init(void)
 			x86_pmu.lbr_nr =3D 0;
 	}
=20
-	if (x86_pmu.lbr_nr)
+	if (x86_pmu.lbr_nr) {
 		pr_cont("%d-deep LBR, ", x86_pmu.lbr_nr);
=20
+		/* only support branch_stack snapshot for perfmon >=3D v2 */
+		if (x86_pmu.disable_all =3D=3D intel_pmu_disable_all)
+			static_call_update(perf_snapshot_branch_stack,
+					   intel_pmu_snapshot_branch_stack);
+	}
+
 	intel_pmu_check_extra_regs(x86_pmu.extra_regs);
=20
 	/* Support full width counters using alternative MSR range */
diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index 8647713276a73..8a832986578a9 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -1296,14 +1296,6 @@ void intel_pmu_pebs_enable_all(void)
 		wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
 }
=20
-void intel_pmu_pebs_disable_all(void)
-{
-	struct cpu_hw_events *cpuc =3D this_cpu_ptr(&cpu_hw_events);
-
-	if (cpuc->pebs_enabled)
-		wrmsrl(MSR_IA32_PEBS_ENABLE, 0);
-}
-
 static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs)
 {
 	struct cpu_hw_events *cpuc =3D this_cpu_ptr(&cpu_hw_events);
diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
index e3ac05c97b5e5..171abbb359fe5 100644
--- a/arch/x86/events/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -1240,6 +1240,14 @@ static inline bool intel_pmu_has_bts(struct perf_e=
vent *event)
 	return intel_pmu_has_bts_period(event, hwc->sample_period);
 }
=20
+static __always_inline void intel_pmu_pebs_disable_all(void)
+{
+	struct cpu_hw_events *cpuc =3D this_cpu_ptr(&cpu_hw_events);
+
+	if (cpuc->pebs_enabled)
+		wrmsrl(MSR_IA32_PEBS_ENABLE, 0);
+}
+
 int intel_pmu_save_and_restart(struct perf_event *event);
=20
 struct event_constraint *
@@ -1314,8 +1322,6 @@ void intel_pmu_pebs_disable(struct perf_event *even=
t);
=20
 void intel_pmu_pebs_enable_all(void);
=20
-void intel_pmu_pebs_disable_all(void);
-
 void intel_pmu_pebs_sched_task(struct perf_event_context *ctx, bool sche=
d_in);
=20
 void intel_pmu_auto_reload_read(struct perf_event *event);
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index fe156a8170aa3..a368dfd754608 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -57,6 +57,7 @@ struct perf_guest_info_callbacks {
 #include <linux/cgroup.h>
 #include <linux/refcount.h>
 #include <linux/security.h>
+#include <linux/static_call.h>
 #include <asm/local.h>
=20
 struct perf_callchain_entry {
@@ -1612,4 +1613,29 @@ extern void __weak arch_perf_update_userpage(struc=
t perf_event *event,
 extern __weak u64 arch_perf_get_page_size(struct mm_struct *mm, unsigned=
 long addr);
 #endif
=20
+/*
+ * Snapshot branch stack on software events.
+ *
+ * Branch stack can be very useful in understanding software events. For
+ * example, when a long function, e.g. sys_perf_event_open, returns an
+ * errno, it is not obvious why the function failed. Branch stack could
+ * provide very helpful information in this type of scenarios.
+ *
+ * On software event, it is necessary to stop the hardware branch record=
er
+ * fast. Otherwise, the hardware register/buffer will be flushed with
+ * entries af the triggering event. Therefore, static call is used to
+ * stop the hardware recorder.
+ */
+enum {
+	PERF_MAX_BRANCH_SNAPSHOT =3D 32,
+};
+
+struct perf_branch_snapshot {
+	unsigned int nr;
+	struct perf_branch_entry entries[PERF_MAX_BRANCH_SNAPSHOT];
+};
+
+typedef int (perf_snapshot_branch_stack_t)(struct perf_branch_snapshot *=
);
+DECLARE_STATIC_CALL(perf_snapshot_branch_stack, perf_snapshot_branch_sta=
ck_t);
+
 #endif /* _LINUX_PERF_EVENT_H */
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 011cc5069b7ba..d32a3cf37eb90 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -13437,3 +13437,5 @@ struct cgroup_subsys perf_event_cgrp_subsys =3D {
 	.threaded	=3D true,
 };
 #endif /* CONFIG_CGROUP_PERF */
+
+DEFINE_STATIC_CALL_RET0(perf_snapshot_branch_stack, perf_snapshot_branch=
_stack_t);
--=20
2.30.2