lkml.org 
[lkml]   [2015]   [Mar]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH 3/4] perf, tools, stat: Fix IPC and other formulas with -A
Date
From: Andi Kleen <ak@linux.intel.com>

perf stat didn't compute the IPC and other formulas for individual
CPUs with -A. Fix this for the easy -A case. As before,
--per-core and --per-socket do not handle it, they
simply print nothing.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
tools/perf/builtin-stat.c | 36 +++++++++++++++++++-----------------
1 file changed, 19 insertions(+), 17 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index b5ba033..05c7a00 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -367,41 +367,42 @@ static struct perf_evsel *nth_evsel(int n)
* more semantic information such as miss/hit ratios,
* instruction rates, etc:
*/
-static void update_shadow_stats(struct perf_evsel *counter, u64 *count)
+static void update_shadow_stats(struct perf_evsel *counter, u64 *count,
+ int cpu)
{
int ctx = evsel_context(counter);

if (perf_evsel__match(counter, SOFTWARE, SW_TASK_CLOCK))
- update_stats(&runtime_nsecs_stats[0], count[0]);
+ update_stats(&runtime_nsecs_stats[cpu], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
- update_stats(&runtime_cycles_stats[ctx][0], count[0]);
+ update_stats(&runtime_cycles_stats[ctx][cpu], count[0]);
else if (transaction_run &&
perf_evsel__cmp(counter, nth_evsel(T_CYCLES_IN_TX)))
- update_stats(&runtime_cycles_in_tx_stats[ctx][0], count[0]);
+ update_stats(&runtime_cycles_in_tx_stats[ctx][cpu], count[0]);
else if (transaction_run &&
perf_evsel__cmp(counter, nth_evsel(T_TRANSACTION_START)))
- update_stats(&runtime_transaction_stats[ctx][0], count[0]);
+ update_stats(&runtime_transaction_stats[ctx][cpu], count[0]);
else if (transaction_run &&
perf_evsel__cmp(counter, nth_evsel(T_ELISION_START)))
- update_stats(&runtime_elision_stats[ctx][0], count[0]);
+ update_stats(&runtime_elision_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_FRONTEND))
- update_stats(&runtime_stalled_cycles_front_stats[ctx][0], count[0]);
+ update_stats(&runtime_stalled_cycles_front_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_BACKEND))
- update_stats(&runtime_stalled_cycles_back_stats[ctx][0], count[0]);
+ update_stats(&runtime_stalled_cycles_back_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
- update_stats(&runtime_branches_stats[ctx][0], count[0]);
+ update_stats(&runtime_branches_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_CACHE_REFERENCES))
- update_stats(&runtime_cacherefs_stats[ctx][0], count[0]);
+ update_stats(&runtime_cacherefs_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1D))
- update_stats(&runtime_l1_dcache_stats[ctx][0], count[0]);
+ update_stats(&runtime_l1_dcache_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1I))
- update_stats(&runtime_l1_icache_stats[ctx][0], count[0]);
+ update_stats(&runtime_l1_icache_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_LL))
- update_stats(&runtime_ll_cache_stats[ctx][0], count[0]);
+ update_stats(&runtime_ll_cache_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_DTLB))
- update_stats(&runtime_dtlb_cache_stats[ctx][0], count[0]);
+ update_stats(&runtime_dtlb_cache_stats[ctx][cpu], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_ITLB))
- update_stats(&runtime_itlb_cache_stats[ctx][0], count[0]);
+ update_stats(&runtime_itlb_cache_stats[ctx][cpu], count[0]);
}

static void zero_per_pkg(struct perf_evsel *counter)
@@ -463,7 +464,8 @@ static int read_cb(struct perf_evsel *evsel, int cpu, int thread __maybe_unused,
perf_evsel__compute_deltas(evsel, cpu, count);
perf_counts_values__scale(count, scale, NULL);
evsel->counts->cpu[cpu] = *count;
- update_shadow_stats(evsel, count->values);
+ if (aggr_mode == AGGR_NONE)
+ update_shadow_stats(evsel, count->values, cpu);
break;
case AGGR_GLOBAL:
aggr->val += count->val;
@@ -511,7 +513,7 @@ static int read_counter_aggr(struct perf_evsel *counter)
/*
* Save the full runtime - to allow normalization during printout:
*/
- update_shadow_stats(counter, count);
+ update_shadow_stats(counter, count, 0);

return 0;
}
--
1.9.3


\
 
 \ /
  Last update: 2015-03-09 01:01    [W:0.120 / U:4.476 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site