From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932737Ab2BOCeM (ORCPT ); Tue, 14 Feb 2012 21:34:12 -0500 Received: from e35.co.us.ibm.com ([32.97.110.153]:58733 "EHLO e35.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932129Ab2BOCeJ (ORCPT ); Tue, 14 Feb 2012 21:34:09 -0500 Date: Tue, 14 Feb 2012 18:37:05 -0800 From: Sukadev Bhattiprolu To: Benjamin Herrenschmidt Cc: Stephane Eranian , Anton Blanchard , linux-kernel@vger.kernel.org, peterz@infradead.org, mingo@elte.hu, gleb@redhat.com, wcohen@redhat.com, vince@deater.net, asharma@fb.com, andi@firstfloor.org, paulus@samba.org, emunson@mgebm.net, imunsie@au1.ibm.com, Matt Helsley Subject: Re: [PATCH] perf_events: fix broken intr throttling (v3) Message-ID: <20120215023705.GA13042@us.ibm.com> References: <20120126160319.GA5655@quad> <20120208223250.282861cb@kryten> <20120210031930.GA28023@us.ibm.com> <1328854519.3019.3.camel@pasglop> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1328854519.3019.3.camel@pasglop> X-Operating-System: Linux 2.0.32 on an i486 User-Agent: Mutt/1.5.20 (2009-06-14) X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12021502-6148-0000-0000-00000372F80A Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Benjamin Herrenschmidt [benh@kernel.crashing.org] wrote: | On Thu, 2012-02-09 at 19:19 -0800, Sukadev Bhattiprolu wrote: | > Stephane Eranian [eranian@google.com] wrote: | > | On Wed, Feb 8, 2012 at 12:32 PM, Anton Blanchard wrote: | > | > | > | > Hi, | > | > | > | > On Thu, 26 Jan 2012 17:03:19 +0100 | The powerpc .git tree is not useful at this point, I haven't opened | powerpc -next yet so it's just all old stale stuff. Please work with | upstream always unless you have a very good reason to use my tree. It's | really only meant as a conduit between me and Linus or for -next. Ok. I tried this on current mainline (commit 13d261932) which now includes both of Stephane's patches and repeated the nooploop test. The problem still exists - the number of events reported on Power is far fewer than without the patches. Aggregated stats: TOTAL events: 78 MMAP events: 47 COMM events: 2 EXIT events: 2 SAMPLE events: 27 cycles stats: TOTAL events: 21 MMAP events: 4 COMM events: 1 EXIT events: 2 SAMPLE events: 14 cycles stats: TOTAL events: 13 SAMPLE events: 13 Sukadev