Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753039AbZF3NKc (ORCPT ); Tue, 30 Jun 2009 09:10:32 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751529AbZF3NKZ (ORCPT ); Tue, 30 Jun 2009 09:10:25 -0400 Received: from hera.kernel.org ([140.211.167.34]:51402 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751379AbZF3NKY (ORCPT ); Tue, 30 Jun 2009 09:10:24 -0400 Subject: Re: [PATCH -tip] perf stat: define CHECK_ATTRS for easy attrs checking From: Jaswinder Singh Rajput To: Ingo Molnar Cc: Thomas Gleixner , Peter Zijlstra , LKML In-Reply-To: <20090630100835.GE6942@elte.hu> References: <1246274348.3018.3.camel@hpdv5.satnam> <20090630100835.GE6942@elte.hu> Content-Type: text/plain Date: Tue, 30 Jun 2009 18:39:48 +0530 Message-Id: <1246367388.3026.2.camel@hpdv5.satnam> Mime-Version: 1.0 X-Mailer: Evolution 2.24.5 (2.24.5-1.fc10) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3631 Lines: 102 On Tue, 2009-06-30 at 12:08 +0200, Ingo Molnar wrote: > * Jaswinder Singh Rajput wrote: > > > CHECK_ATTRS is useful : > > 1. for multiple attrs checking > > 2. avoid repetition of PERF_TYPE_ and PERF_COUNT_ and save space > > 3. avoids line breakage > > > > Signed-off-by: Jaswinder Singh Rajput > > --- > > tools/perf/builtin-stat.c | 18 ++++++++---------- > > 1 files changed, 8 insertions(+), 10 deletions(-) > > > > diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c > > index 3e5ea4e..6231b39 100644 > > --- a/tools/perf/builtin-stat.c > > +++ b/tools/perf/builtin-stat.c > > @@ -96,6 +96,10 @@ static u64 walltime_nsecs_noise; > > static u64 runtime_cycles_avg; > > static u64 runtime_cycles_noise; > > > > +#define CHECK_ATTRS(t, c, counter) \ > > + (attrs[counter].type == PERF_TYPE_##t && \ > > + attrs[counter].config == PERF_COUNT_##c) > > looks good except that the name should be something like > MATCH_EVENT() not CHECK_ATTRS(). > [PATCH -tip] perf stat: define MATCH_EVENT for easy attrs checking MATCH_EVENT is useful : 1. for multiple attrs checking 2. avoid repetition of PERF_TYPE_ and PERF_COUNT_ and save space 3. avoids line breakage Signed-off-by: Jaswinder Singh Rajput --- tools/perf/builtin-stat.c | 18 ++++++++---------- 1 files changed, 8 insertions(+), 10 deletions(-) diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 3e5ea4e..00feb8d 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -96,6 +96,10 @@ static u64 walltime_nsecs_noise; static u64 runtime_cycles_avg; static u64 runtime_cycles_noise; +#define MATCH_EVENT(t, c, counter) \ + (attrs[counter].type == PERF_TYPE_##t && \ + attrs[counter].config == PERF_COUNT_##c) + #define ERR_PERF_OPEN \ "Error: counter %d, sys_perf_counter_open() syscall returned with %d (%s)\n" @@ -191,11 +195,9 @@ static void read_counter(int counter) /* * Save the full runtime - to allow normalization during printout: */ - if (attrs[counter].type == PERF_TYPE_SOFTWARE && - attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) + if (MATCH_EVENT(SOFTWARE, SW_TASK_CLOCK, counter)) runtime_nsecs[run_idx] = count[0]; - if (attrs[counter].type == PERF_TYPE_HARDWARE && - attrs[counter].config == PERF_COUNT_HW_CPU_CYCLES) + if (MATCH_EVENT(HARDWARE, HW_CPU_CYCLES, counter)) runtime_cycles[run_idx] = count[0]; } @@ -254,9 +256,7 @@ static void nsec_printout(int counter, u64 *count, u64 *noise) fprintf(stderr, " %14.6f %-24s", msecs, event_name(counter)); - if (attrs[counter].type == PERF_TYPE_SOFTWARE && - attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) { - + if (MATCH_EVENT(SOFTWARE, SW_TASK_CLOCK, counter)) { if (walltime_nsecs_avg) fprintf(stderr, " # %10.3f CPUs ", (double)count[0] / (double)walltime_nsecs_avg); @@ -269,9 +269,7 @@ static void abs_printout(int counter, u64 *count, u64 *noise) fprintf(stderr, " %14Ld %-24s", count[0], event_name(counter)); if (runtime_cycles_avg && - attrs[counter].type == PERF_TYPE_HARDWARE && - attrs[counter].config == PERF_COUNT_HW_INSTRUCTIONS) { - + MATCH_EVENT(HARDWARE, HW_INSTRUCTIONS, counter)) { fprintf(stderr, " # %10.3f IPC ", (double)count[0] / (double)runtime_cycles_avg); } else { -- 1.6.0.6 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/