Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com>
Date:   Wed, 27 Nov 2019 21:01:30 -0300
User-Agent: K-9 Mail for Android
In-Reply-To: <20191127232657.GL84886@tassilo.jf.intel.com>
References: <20191121001522.180827-1-andi@firstfloor.org> <20191127151657.GE22719@kernel.org> <20191127154305.GJ22719@kernel.org> <20191127232657.GL84886@tassilo.jf.intel.com>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: Optimize perf stat for large number of events/cpus
To:     Andi Kleen <ak@linux.intel.com>,
        Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com>
CC:     Andi Kleen <andi@firstfloor.org>, jolsa@kernel.org,
        linux-kernel@vger.kernel.org
Message-ID: <033E8FCB-2430-48E0-8631-0F0236D5E904@kernel.org>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On November 27, 2019 8:26:57 PM GMT-03:00, Andi Kleen <ak@linux=2Eintel=2Ec=
om> wrote:
>On Wed, Nov 27, 2019 at 12:43:05PM -0300, Arnaldo Carvalho de Melo
>wrote:
>> So, have you tried running 'perf test' after each cset is applied and
>> built?
>
>I ran it at the end, but there are quite a few fails out of the box,
>so I missed that one thanks=2E
>
>This patch fixes it=2E Let me know if I should submit it in a more
>formal way=2E
>
>---
>
>Fix event times test case
>
>Reported-by: Arnaldo
>Signed-off-by: Andi Kleen <ak@linux=2Eintel=2Ecom>
>
>diff --git a/tools/perf/lib/evsel=2Ec b/tools/perf/lib/evsel=2Ec
>index 4c6485fc31b9=2E=2E4dc06289f4c7 100644
>--- a/tools/perf/lib/evsel=2Ec
>+++ b/tools/perf/lib/evsel=2Ec
>@@ -224,7 +224,7 @@ int perf_evsel__enable(struct perf_evsel *evsel)
> 	int i;
> 	int err =3D 0;
>=20
>-	for (i =3D 0; i < evsel->cpus->nr && !err; i++)
>+	for (i =3D 0; i < xyarray__max_x(evsel->fd) && !err; i++)
> 		err =3D perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, i);
> 	return err;
> }
>@@ -239,7 +239,7 @@ int perf_evsel__disable(struct perf_evsel *evsel)
> 	int i;
> 	int err =3D 0;
>=20
>-	for (i =3D 0; i < evsel->cpus->nr && !err; i++)
>+	for (i =3D 0; i < xyarray__max_x(evsel->fd) && !err; i++)
> 		err =3D perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, i);
> 	return err;
> }
>diff --git a/tools/perf/util/evsel=2Ec b/tools/perf/util/evsel=2Ec
>index 59b9b4f3fe34=2E=2E0844e3e29fb0 100644
>--- a/tools/perf/util/evsel=2Ec
>+++ b/tools/perf/util/evsel=2Ec
>@@ -1853,6 +1853,10 @@ int perf_evsel__open_per_cpu(struct evsel
>*evsel,
> 			     struct perf_cpu_map *cpus,
> 			     int cpu)
> {
>+	if (cpu =3D=3D -1)
>+		return evsel__open_cpu(evsel, cpus, NULL, 0,
>+					cpus ? cpus->nr : 1);
>+
> 	return evsel__open_cpu(evsel, cpus, NULL, cpu, cpu + 1);
> }
>=20

Just save me some time by saying to which cset in v8 I should squash this =
into, so that we keep the whole shebang bisectable,

Thanks,

- Arnaldo