linux-kernel - [PATCH 4/4] perf/x86/intel: Throttle PEBS events only from pmi

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1482931866-6018-5-git-send-email-jolsa@kernel.org>
Date:   Wed, 28 Dec 2016 14:31:06 +0100
From:   Jiri Olsa <jolsa@...nel.org>
To:     Peter Zijlstra <a.p.zijlstra@...llo.nl>
Cc:     lkml <linux-kernel@...r.kernel.org>,
        Ingo Molnar <mingo@...nel.org>,
        Andi Kleen <andi@...stfloor.org>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Arnaldo Carvalho de Melo <acme@...nel.org>,
        Vince Weaver <vince@...ter.net>
Subject: [PATCH 4/4] perf/x86/intel: Throttle PEBS events only from pmi

This patch fixes following WARNING:

  WARNING: CPU: 15 PID: 15768 at arch/x86/events/core.c:1256 x86_pmu_start+0x1b3/0x1c0
  ...
  Call Trace:
   <IRQ>
   dump_stack+0x86/0xc3
   __warn+0xcb/0xf0
   warn_slowpath_null+0x1d/0x20
   x86_pmu_start+0x1b3/0x1c0
   perf_event_task_tick+0x342/0x3f0
   scheduler_tick+0x75/0xd0
   update_process_times+0x47/0x60
   tick_sched_handle.isra.19+0x25/0x60
   tick_sched_timer+0x3d/0x70
   __hrtimer_run_queues+0xfb/0x510
   hrtimer_interrupt+0x9d/0x1a0
   local_apic_timer_interrupt+0x38/0x60
   smp_trace_apic_timer_interrupt+0x56/0x25a
   trace_apic_timer_interrupt+0x9d/0xb0
   ...

which happens AFAICS under following conditions:
(we have PEBS events configured)

  - x86_pmu_enable reconfigures counters and calls:
       - x86_pmu_stop on PEBS event
       - x86_pmu_stop drains the PEBS buffer, crosses
         the throttle limit and sets:
           event->hw.interrupts = MAX_INTERRUPTS
       - following x86_pmu_start call starts the event
  - perf_event_task_tick is triggered
    - perf_adjust_freq_unthr_context sees event with
      MAX_INTERRUPTS set and calls x86_pmu_start on already
      started event, which triggers the warning

My first attempt to fix this was to unthrottle the event
before starting it in x86_pmu_enable. But I think that
omitting the throttling completely when we are not in the
PMI is better.

Signed-off-by: Jiri Olsa <jolsa@...nel.org>
---
 arch/x86/events/intel/core.c |  2 +-
 arch/x86/events/intel/ds.c   | 17 +++++++++--------
 arch/x86/events/perf_event.h |  2 +-
 3 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index 86138267b68a..16332867d435 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -2137,7 +2137,7 @@ static int intel_pmu_handle_irq(struct pt_regs *regs)
 	 */
 	if (__test_and_clear_bit(62, (unsigned long *)&status)) {
 		handled++;
-		x86_pmu.drain_pebs(regs);
+		x86_pmu.drain_pebs(regs, 1);
 		status &= x86_pmu.intel_ctrl | GLOBAL_STATUS_TRACE_TOPAPMI;
 	}
 
diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index 9dfeeeca0ea8..75ba3cb06012 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -603,7 +603,7 @@ static inline void intel_pmu_drain_pebs_buffer(void)
 {
 	struct pt_regs regs;
 
-	x86_pmu.drain_pebs(&regs);
+	x86_pmu.drain_pebs(&regs, 0);
 }
 
 void intel_pmu_pebs_sched_task(struct perf_event_context *ctx, bool sched_in)
@@ -1233,7 +1233,7 @@ get_next_pebs_record_by_bit(void *base, void *top, int bit)
 static void __intel_pmu_pebs_event(struct perf_event *event,
 				   struct pt_regs *iregs,
 				   void *base, void *top,
-				   int bit, int count)
+				   int bit, int count, int pmi)
 {
 	struct perf_sample_data data;
 	struct pt_regs regs;
@@ -1257,14 +1257,14 @@ static void __intel_pmu_pebs_event(struct perf_event *event,
 	 * All but the last records are processed.
 	 * The last one is left to be able to call the overflow handler.
 	 */
-	if (perf_event_overflow(event, &data, &regs)) {
+	if (perf_event_overflow_throttle(event, pmi, &data, &regs)) {
 		x86_pmu_stop(event, 0);
 		return;
 	}
 
 }
 
-static void intel_pmu_drain_pebs_core(struct pt_regs *iregs)
+static void intel_pmu_drain_pebs_core(struct pt_regs *iregs, int pmi)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct debug_store *ds = cpuc->ds;
@@ -1295,10 +1295,10 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs)
 	if (n <= 0)
 		return;
 
-	__intel_pmu_pebs_event(event, iregs, at, top, 0, n);
+	__intel_pmu_pebs_event(event, iregs, at, top, 0, n, pmi);
 }
 
-static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs)
+static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs, int pmi)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct debug_store *ds = cpuc->ds;
@@ -1392,13 +1392,14 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs)
 		if (error[bit]) {
 			perf_log_lost_samples(event, error[bit]);
 
-			if (perf_event_account_interrupt(event))
+			if (pmi && perf_event_account_interrupt(event))
 				x86_pmu_stop(event, 0);
 		}
 
 		if (counts[bit]) {
 			__intel_pmu_pebs_event(event, iregs, base,
-					       top, bit, counts[bit]);
+					       top, bit, counts[bit],
+					       pmi);
 		}
 	}
 }
diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
index bcbb1d2ae10b..cdcfdd06b2d2 100644
--- a/arch/x86/events/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -590,7 +590,7 @@ struct x86_pmu {
 			pebs_prec_dist	:1;
 	int		pebs_record_size;
 	int		pebs_buffer_size;
-	void		(*drain_pebs)(struct pt_regs *regs);
+	void		(*drain_pebs)(struct pt_regs *regs, int pmi);
 	struct event_constraint *pebs_constraints;
 	void		(*pebs_aliases)(struct perf_event *event);
 	int 		max_pebs_events;
-- 
2.7.4