From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760147Ab3HOCDj (ORCPT ); Wed, 14 Aug 2013 22:03:39 -0400 Received: from hrndva-omtalb.mail.rr.com ([71.74.56.122]:2950 "EHLO hrndva-omtalb.mail.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758582Ab3HOCDg (ORCPT ); Wed, 14 Aug 2013 22:03:36 -0400 X-Authority-Analysis: v=2.0 cv=aqMw+FlV c=1 sm=0 a=Sro2XwOs0tJUSHxCKfOySw==:17 a=Drc5e87SC40A:10 a=Ciwy3NGCPMMA:10 a=dzgTTVZbPnkA:10 a=5SG0PmZfjMsA:10 a=bbbx4UPp9XUA:10 a=meVymXHHAAAA:8 a=KGjhK52YXX0A:10 a=l2TaRiXcqNYA:10 a=20KFwNOVAAAA:8 a=VwQbUJbxAAAA:8 a=pGLkceISAAAA:8 a=JfrnYn6hAAAA:8 a=Hv4b9A_i4XO6o4Ooc9EA:9 a=jEp0ucaQiEUA:10 a=MSl-tDqOz04A:10 a=3Rfx1nUSh_UA:10 a=jeBq3FmKZ4MA:10 a=Sro2XwOs0tJUSHxCKfOySw==:117 X-Cloudmark-Score: 0 X-Authenticated-User: X-Originating-IP: 67.255.60.225 Message-Id: <20130815020335.292748624@goodmis.org> User-Agent: quilt/0.60-1 Date: Wed, 14 Aug 2013 22:01:07 -0400 From: Steven Rostedt To: linux-kernel@vger.kernel.org Cc: Ingo Molnar , Andrew Morton , Peter Zijlstra , David Ahern , Oleg Nesterov Subject: [PATCH tip/perf/core 3/3] tracing/perf: Avoid perf_trace_buf_*() in perf_trace_##call() when possible References: <20130815020104.813498803@goodmis.org> Content-Disposition: inline; filename=0003-tracing-perf-Avoid-perf_trace_buf_-in-perf_trace_-ca.patch Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Oleg Nesterov perf_trace_buf_prepare() + perf_trace_buf_submit(task => NULL) make no sense if hlist_empty(head). Change perf_trace_##call() to check ->perf_events beforehand and do nothing if it is empty. This removes the overhead for tasks without events associated with them. For example, "perf record -e sched:sched_switch -p1" attaches the counter(s) to the single task, but every task in system will do perf_trace_buf_prepare/submit() just to realize that it was not attached to this event. However, we can only do this if __task == NULL, so we also add the __builtin_constant_p(__task) check. With this patch "perf bench sched pipe" shows approximately 4% improvement when "perf record -p1" runs in parallel, many thanks to Steven for the testing. Link: http://lkml.kernel.org/r/20130806160847.GA2746@redhat.com Tested-by: David Ahern Acked-by: Peter Zijlstra Signed-off-by: Oleg Nesterov Signed-off-by: Steven Rostedt --- include/trace/ftrace.h | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 4163d93..5c7ab17 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -667,6 +667,12 @@ perf_trace_##call(void *__data, proto) \ int rctx; \ \ __data_size = ftrace_get_offsets_##call(&__data_offsets, args); \ + \ + head = this_cpu_ptr(event_call->perf_events); \ + if (__builtin_constant_p(!__task) && !__task && \ + hlist_empty(head)) \ + return; \ + \ __entry_size = ALIGN(__data_size + sizeof(*entry) + sizeof(u32),\ sizeof(u64)); \ __entry_size -= sizeof(u32); \ @@ -681,7 +687,6 @@ perf_trace_##call(void *__data, proto) \ \ { assign; } \ \ - head = this_cpu_ptr(event_call->perf_events); \ perf_trace_buf_submit(entry, __entry_size, rctx, __addr, \ __count, &__regs, head, __task); \ } -- 1.7.10.4