linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Adrian Hunter <adrian.hunter@intel.com>
To: tanze <tanze@kylinos.cn>, <james.clark@linaro.org>,
	<leo.yan@linux.dev>, <irogers@google.com>,
	<john.g.garry@oracle.com>
Cc: <linux-arm-kernel@lists.infradead.org>,
	<linux-perf-users@vger.kernel.org>,
	<linux-kernel@vger.kernel.org>, <graham.woodward@arm.com>,
	<mike.leach@linaro.org>
Subject: Re: [PATCH v2] perf arm_spe: Add a macro definition to handle offset value
Date: Fri, 24 Oct 2025 09:27:25 +0300	[thread overview]
Message-ID: <22b7b11e-b1fe-4d9d-8749-afe57ea1c35c@intel.com> (raw)
In-Reply-To: <20251017021540.45930-1-tanze@kylinos.cn>

On 17/10/2025 05:15, tanze wrote:
> Add a macro definition SPE_SYNTH_ID_OFFSET to handle the offset value
> and improve readability.
> 
> Signed-off-by: tanze <tanze@kylinos.cn>
> ---
> Hi, Leo Yan
> 
> Thank you for your guidance and suggestions. I have made revisions according to your requirements. 
> Do you have any further suggestions?
> ---
> Changes in v2:
> - Migrate the macro definitions to the synthetic-events.h file
> - Add modifications to other source files that use the offset value
> ---
>  tools/perf/util/arm-spe.c          | 2 +-
>  tools/perf/util/cs-etm.c           | 2 +-
>  tools/perf/util/intel-bts.c        | 2 +-
>  tools/perf/util/intel-pt.c         | 2 +-
>  tools/perf/util/powerpc-vpadtl.c   | 3 ++-
>  tools/perf/util/synthetic-events.h | 2 ++
>  6 files changed, 8 insertions(+), 5 deletions(-)
> 
> diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c
> index 71be979f5077..b082cb1666a6 100644
> --- a/tools/perf/util/arm-spe.c
> +++ b/tools/perf/util/arm-spe.c
> @@ -1732,7 +1732,7 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session)
>  	attr.sample_period = spe->synth_opts.period;
>  
>  	/* create new id val to be a fixed offset from evsel id */
> -	id = evsel->core.id[0] + 1000000000;
> +	id = evsel->core.id[0] + PERF_SYNTH_EVENT_ID_OFFSET;
>  
>  	if (!id)
>  		id = 1;

Functionally this is to find a free range of IDs for synthesized events,
but also the range can only have one user e.g. in this case auxtrace

So add to auxtrace.c

#define AUXTRACE_SYNTH_EVENT_ID_OFFSET	1000000000ULL

/*
 * Event IDs are allocated sequentially, so a big offset from any
 * existing ID will reach a unused range.
 */
u64 auxtrace_synth_id_range_start(struct evsel *evsel)
{
	u64 id = evsel->core.id[0] + AUXTRACE_SYNTH_EVENT_ID_OFFSET;

  	if (!id)
  		id = 1;

	return id;
}

And then use that:

- 	/* create new id val to be a fixed offset from evsel id */
-	id = evsel->core.id[0] + 1000000000;
+	id = auxtrace_synth_id_range_start(evsel);
-
-	if (!id)
-		id = 1;

> diff --git a/tools/perf/util/cs-etm.c b/tools/perf/util/cs-etm.c
> index 30f4bb3e7fa3..46902f7b69f1 100644
> --- a/tools/perf/util/cs-etm.c
> +++ b/tools/perf/util/cs-etm.c
> @@ -1726,7 +1726,7 @@ static int cs_etm__synth_events(struct cs_etm_auxtrace *etm,
>  	attr.read_format = evsel->core.attr.read_format;
>  
>  	/* create new id val to be a fixed offset from evsel id */
> -	id = evsel->core.id[0] + 1000000000;
> +	id = evsel->core.id[0] + PERF_SYNTH_EVENT_ID_OFFSET;
>  
>  	if (!id)
>  		id = 1;
> diff --git a/tools/perf/util/intel-bts.c b/tools/perf/util/intel-bts.c
> index 3625c6224750..98b928eca724 100644
> --- a/tools/perf/util/intel-bts.c
> +++ b/tools/perf/util/intel-bts.c
> @@ -777,7 +777,7 @@ static int intel_bts_synth_events(struct intel_bts *bts,
>  	attr.sample_id_all = evsel->core.attr.sample_id_all;
>  	attr.read_format = evsel->core.attr.read_format;
>  
> -	id = evsel->core.id[0] + 1000000000;
> +	id = evsel->core.id[0] + PERF_SYNTH_EVENT_ID_OFFSET;
>  	if (!id)
>  		id = 1;
>  
> diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
> index 9b1011fe4826..4fb9600a7369 100644
> --- a/tools/perf/util/intel-pt.c
> +++ b/tools/perf/util/intel-pt.c
> @@ -3987,7 +3987,7 @@ static int intel_pt_synth_events(struct intel_pt *pt,
>  	attr.sample_id_all = evsel->core.attr.sample_id_all;
>  	attr.read_format = evsel->core.attr.read_format;
>  
> -	id = evsel->core.id[0] + 1000000000;
> +	id = evsel->core.id[0] + PERF_SYNTH_EVENT_ID_OFFSET;
>  	if (!id)
>  		id = 1;
>  
> diff --git a/tools/perf/util/powerpc-vpadtl.c b/tools/perf/util/powerpc-vpadtl.c
> index 39a3fb3f1330..5ce5d2bf4c6c 100644
> --- a/tools/perf/util/powerpc-vpadtl.c
> +++ b/tools/perf/util/powerpc-vpadtl.c
> @@ -15,6 +15,7 @@
>  #include "powerpc-vpadtl.h"
>  #include "sample.h"
>  #include "tool.h"
> +#include "util/synthetic-events.h"
>  
>  /*
>   * Structure to save the auxtrace queue
> @@ -656,7 +657,7 @@ powerpc_vpadtl_synth_events(struct powerpc_vpadtl *vpa, struct perf_session *ses
>  	attr.config = PERF_SYNTH_POWERPC_VPA_DTL;
>  
>  	/* create new id val to be a fixed offset from evsel id */
> -	id = evsel->core.id[0] + 1000000000;
> +	id = evsel->core.id[0] + PERF_SYNTH_EVENT_ID_OFFSET;
>  	if (!id)
>  		id = 1;
>  
> diff --git a/tools/perf/util/synthetic-events.h b/tools/perf/util/synthetic-events.h
> index ee29615d68e5..c8a1fe7a3fa4 100644
> --- a/tools/perf/util/synthetic-events.h
> +++ b/tools/perf/util/synthetic-events.h
> @@ -39,6 +39,8 @@ enum perf_record_synth {
>  };
>  #define PERF_SYNTH_ALL  (PERF_SYNTH_MAX - 1)
>  
> +#define PERF_SYNTH_EVENT_ID_OFFSET	(1000000000ULL)
> +
>  int parse_synth_opt(char *str);
>  
>  typedef int (*perf_event__handler_t)(const struct perf_tool *tool, union perf_event *event,


  parent reply	other threads:[~2025-10-24  6:27 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-16  8:30 [PATCH v1] perf arm_spe: Add a macro definition to handle offset value tanze
2025-10-16  9:09 ` Leo Yan
     [not found] ` <1760620322251982.15.seg@mailgw.kylinos.cn>
2025-10-17  1:07   ` tanze
2025-10-17  2:15 ` [PATCH v2] " tanze
2025-10-17 10:09   ` Leo Yan
2025-10-19  3:05   ` Namhyung Kim
2025-10-24  6:27   ` Adrian Hunter [this message]
2025-10-24  8:10     ` Leo Yan

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=22b7b11e-b1fe-4d9d-8749-afe57ea1c35c@intel.com \
    --to=adrian.hunter@intel.com \
    --cc=graham.woodward@arm.com \
    --cc=irogers@google.com \
    --cc=james.clark@linaro.org \
    --cc=john.g.garry@oracle.com \
    --cc=leo.yan@linux.dev \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mike.leach@linaro.org \
    --cc=tanze@kylinos.cn \
    /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).