Commit 6c1c07b3 authored by Kan Liang's avatar Kan Liang Committed by Ingo Molnar

perf/x86/intel: Avoid unnecessary PEBS_ENABLE MSR access in PMI

The perf PMI handler, intel_pmu_handle_irq(), currently does
unnecessary MSR accesses for PEBS_ENABLE MSR in
__intel_pmu_enable/disable_all() when PEBS is enabled.

When entering the handler, global ctrl is explicitly disabled. All
counters do not count anymore. It doesn't matter if PEBS is enabled
or not in a PMI handler.
Furthermore, for most cases, the cpuc->pebs_enabled is not changed in
PMI. The PEBS status doesn't change. The PEBS_ENABLE MSR doesn't need to
be changed either when exiting the handler.

PMI throttle may change the PEBS status during PMI handler. The
x86_pmu_stop() ends up in intel_pmu_pebs_disable() which can update
cpuc->pebs_enabled. But the MSR_IA32_PEBS_ENABLE is not updated
at the same time. Because the cpuc->enabled has been forced to 0.
The patch explicitly update the MSR_IA32_PEBS_ENABLE for this case.

Use ftrace to measure the duration of intel_pmu_handle_irq() on BDX.
   #perf record -e cycles:P -- ./tchain_edit

The average duration of intel_pmu_handle_irq():

  Without the patch       1.144 us
  With the patch          1.025 us
Signed-off-by: default avatarKan Liang <kan.liang@linux.intel.com>
Signed-off-by: default avatarPeter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200121181338.3234-1-kan.liang@linux.intel.com
parent f861854e
...@@ -1945,6 +1945,14 @@ static __initconst const u64 knl_hw_cache_extra_regs ...@@ -1945,6 +1945,14 @@ static __initconst const u64 knl_hw_cache_extra_regs
* intel_bts events don't coexist with intel PMU's BTS events because of * intel_bts events don't coexist with intel PMU's BTS events because of
* x86_add_exclusive(x86_lbr_exclusive_lbr); there's no need to keep them * x86_add_exclusive(x86_lbr_exclusive_lbr); there's no need to keep them
* disabled around intel PMU's event batching etc, only inside the PMI handler. * disabled around intel PMU's event batching etc, only inside the PMI handler.
*
* Avoid PEBS_ENABLE MSR access in PMIs.
* The GLOBAL_CTRL has been disabled. All the counters do not count anymore.
* It doesn't matter if the PEBS is enabled or not.
* Usually, the PEBS status are not changed in PMIs. It's unnecessary to
* access PEBS_ENABLE MSR in disable_all()/enable_all().
* 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 void __intel_pmu_disable_all(void)
{ {
...@@ -1954,13 +1962,12 @@ static void __intel_pmu_disable_all(void) ...@@ -1954,13 +1962,12 @@ static void __intel_pmu_disable_all(void)
if (test_bit(INTEL_PMC_IDX_FIXED_BTS, cpuc->active_mask)) if (test_bit(INTEL_PMC_IDX_FIXED_BTS, cpuc->active_mask))
intel_pmu_disable_bts(); intel_pmu_disable_bts();
intel_pmu_pebs_disable_all();
} }
static void intel_pmu_disable_all(void) static void intel_pmu_disable_all(void)
{ {
__intel_pmu_disable_all(); __intel_pmu_disable_all();
intel_pmu_pebs_disable_all();
intel_pmu_lbr_disable_all(); intel_pmu_lbr_disable_all();
} }
...@@ -1968,7 +1975,6 @@ static void __intel_pmu_enable_all(int added, bool pmi) ...@@ -1968,7 +1975,6 @@ static void __intel_pmu_enable_all(int added, bool pmi)
{ {
struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
intel_pmu_pebs_enable_all();
intel_pmu_lbr_enable_all(pmi); intel_pmu_lbr_enable_all(pmi);
wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL,
x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_guest_mask); x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_guest_mask);
...@@ -1986,6 +1992,7 @@ static void __intel_pmu_enable_all(int added, bool pmi) ...@@ -1986,6 +1992,7 @@ static void __intel_pmu_enable_all(int added, bool pmi)
static void intel_pmu_enable_all(int added) static void intel_pmu_enable_all(int added)
{ {
intel_pmu_pebs_enable_all();
__intel_pmu_enable_all(added, false); __intel_pmu_enable_all(added, false);
} }
...@@ -2374,9 +2381,21 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status) ...@@ -2374,9 +2381,21 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status)
* PEBS overflow sets bit 62 in the global status register * PEBS overflow sets bit 62 in the global status register
*/ */
if (__test_and_clear_bit(62, (unsigned long *)&status)) { if (__test_and_clear_bit(62, (unsigned long *)&status)) {
u64 pebs_enabled = cpuc->pebs_enabled;
handled++; handled++;
x86_pmu.drain_pebs(regs); x86_pmu.drain_pebs(regs);
status &= x86_pmu.intel_ctrl | GLOBAL_STATUS_TRACE_TOPAPMI; status &= x86_pmu.intel_ctrl | GLOBAL_STATUS_TRACE_TOPAPMI;
/*
* PMI throttle may be triggered, which stops the PEBS event.
* Although cpuc->pebs_enabled is updated accordingly, the
* MSR_IA32_PEBS_ENABLE is not updated. Because the
* cpuc->enabled has been forced to 0 in PMI.
* Update the MSR if pebs_enabled is changed.
*/
if (pebs_enabled != cpuc->pebs_enabled)
wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled);
} }
/* /*
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment