Messages in this thread | | | Date | Thu, 5 Nov 2015 03:43:07 +0100 | From | Andi Kleen <> | Subject | Re: [PATCH 3/9] perf, tools, stat: Abstract stat metrics printing |
| |
On Wed, Nov 04, 2015 at 09:42:48AM +0100, Jiri Olsa wrote: > On Mon, Nov 02, 2015 at 05:50:22PM -0800, Andi Kleen wrote: > > SNIP > > > evsel->attr.type == PERF_TYPE_HW_CACHE && > > evsel->attr.config == ( PERF_COUNT_HW_CACHE_L1D | > > ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | > > - ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) && > > - runtime_l1_dcache_stats[ctx][cpu].n != 0) { > > - print_l1_dcache_misses(out, cpu, evsel, avg); > > + ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { > > + if (runtime_l1_dcache_stats[ctx][cpu].n != 0) > > + print_l1_dcache_misses(cpu, evsel, avg, out); > > + else > > + print_metric(ctxp, NULL, NULL, "of all L1-dcache hits", 0); > > hum, what's the reason for all those else cases > ending up printing nothing? > > we have one metric per line anyway right?
There can be stuff after the metric, need to print spaces then to get the correct indentation.
Also it's needed for the metrics only patch (posted separately) so that it can catch all column headers, even if some are missing initially.
-Andi
-- ak@linux.intel.com -- Speaking for myself only.
| |