From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: Thomas Richter <tmricht@linux.ibm.com>
Cc: linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org,
svens@linux.ibm.com, gor@linux.ibm.com, sumanthk@linux.ibm.com,
hca@linux.ibm.com
Subject: Re: [PATCH] perf stat: Suppress warning when using cpum_cf events on s390
Date: Fri, 24 Mar 2023 17:18:15 -0300 [thread overview]
Message-ID: <ZB4Fh71hRVPXW+Ne@kernel.org> (raw)
In-Reply-To: <20230316074946.41110-1-tmricht@linux.ibm.com>
Em Thu, Mar 16, 2023 at 08:49:46AM +0100, Thomas Richter escreveu:
> Running command perf stat -vv -e cpu_cycles -C0 -- true
> displays this warning:
>
> Attempting to add event pmu 'cpum_cf' with 'cpu_cycles,'
> that may result in non-fatal errors
>
> Make the PMU cpum_cf selectable and avoid this warning.
> While at it also fix this warning for PMUs pai_crypto and pai_ext.
>
> Output before:
Thanks, applied.
- Arnaldo
> # ./perf stat -vv -e cpu_cycles -C0 -- true
> Using CPUID IBM,3931,704,A01,3.7,002f
> Attempting to add event pmu 'cpum_cf' with 'cpu_cycles,'
> that may result in non-fatal errors
> After aliases, add event pmu 'cpum_cf' with 'event,'
> that may result in non-fatal errors
> cpu_cycles -> cpum_cf/event=0/
> Control descriptor is not initialized
> ------------------------------------------------------------
> perf_event_attr:
> type 10
> size 128
> config 0x1001
> sample_type IDENTIFIER
> read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING
> disabled 1
> inherit 1
> exclude_guest 1
> ------------------------------------------------------------
> sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8 = 3
> cpu_cycles: 0: 290434 2479172 2479172:
> cpu_cycles: 290434 2479172 2479172
>
> Performance counter stats for 'CPU(s) 0':
>
> 290,434 cpu_cycles
>
> 0.002465617 seconds time elapsed
> #
>
> Now the warning "Attempting to add event pmu 'cpum_cf' ..."
> does not show up anymore.
>
> Output after:
> # ./perf stat -vv -e cpu_cycles -C0 -- true
> Using CPUID IBM,3931,704,A01,3.7,002f
> After aliases, add event pmu 'cpum_cf' with 'event,'
> that may result in non-fatal errors
> cpu_cycles -> cpum_cf/event=0/
> Control descriptor is not initialized
> ....
>
> Performance counter stats for 'CPU(s) 0':
>
> 357,023 cpu_cycles
>
> 0.002454995 seconds time elapsed
> #
>
> Signed-off-by: Thomas Richter <tmricht@linux.ibm.com>
> ---
> tools/perf/arch/s390/util/Build | 1 +
> tools/perf/arch/s390/util/pmu.c | 23 +++++++++++++++++++++++
> 2 files changed, 24 insertions(+)
> create mode 100644 tools/perf/arch/s390/util/pmu.c
>
> diff --git a/tools/perf/arch/s390/util/Build b/tools/perf/arch/s390/util/Build
> index db6884086997..fa66f15a14ec 100644
> --- a/tools/perf/arch/s390/util/Build
> +++ b/tools/perf/arch/s390/util/Build
> @@ -6,5 +6,6 @@ perf-$(CONFIG_DWARF) += dwarf-regs.o
> perf-$(CONFIG_LIBDW_DWARF_UNWIND) += unwind-libdw.o
>
> perf-y += machine.o
> +perf-y += pmu.o
>
> perf-$(CONFIG_AUXTRACE) += auxtrace.o
> diff --git a/tools/perf/arch/s390/util/pmu.c b/tools/perf/arch/s390/util/pmu.c
> new file mode 100644
> index 000000000000..11f03f32e3fd
> --- /dev/null
> +++ b/tools/perf/arch/s390/util/pmu.c
> @@ -0,0 +1,23 @@
> +// SPDX-License-Identifier: GPL-2.0
> +
> +/*
> + * Copyright IBM Corp. 2023
> + * Author(s): Thomas Richter <tmricht@linux.ibm.com>
> + */
> +
> +#include <string.h>
> +
> +#include "../../../util/pmu.h"
> +
> +#define S390_PMUPAI_CRYPTO "pai_crypto"
> +#define S390_PMUPAI_EXT "pai_ext"
> +#define S390_PMUCPUM_CF "cpum_cf"
> +
> +struct perf_event_attr *perf_pmu__get_default_config(struct perf_pmu *pmu)
> +{
> + if (!strcmp(pmu->name, S390_PMUPAI_CRYPTO) ||
> + !strcmp(pmu->name, S390_PMUPAI_EXT) ||
> + !strcmp(pmu->name, S390_PMUCPUM_CF))
> + pmu->selectable = true;
> + return NULL;
> +}
> --
> 2.39.1
>
--
- Arnaldo
prev parent reply other threads:[~2023-03-24 20:18 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-16 7:49 [PATCH] perf stat: Suppress warning when using cpum_cf events on s390 Thomas Richter
2023-03-24 20:18 ` Arnaldo Carvalho de Melo [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZB4Fh71hRVPXW+Ne@kernel.org \
--to=acme@kernel.org \
--cc=gor@linux.ibm.com \
--cc=hca@linux.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=sumanthk@linux.ibm.com \
--cc=svens@linux.ibm.com \
--cc=tmricht@linux.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).