From: Namhyung Kim <namhyung@kernel.org>
To: Tengda Wu <wutengda@huaweicloud.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>,
james.clark@linaro.org, xueshuai@linux.alibaba.com,
Peter Zijlstra <peterz@infradead.org>,
leo.yan@linux.dev, Li Huafei <lihuafei1@huawei.com>,
Ian Rogers <irogers@google.com>,
Kim Phillips <kim.phillips@arm.com>,
Mark Rutland <mark.rutland@arm.com>,
Ingo Molnar <mingo@redhat.com>, Bill Wendling <morbo@google.com>,
Nick Desaulniers <nick.desaulniers+lkml@gmail.com>,
Alexander Shishkin <alexander.shishkin@linux.intel.com>,
Adrian Hunter <adrian.hunter@intel.com>,
Zecheng Li <zli94@ncsu.edu>,
linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org,
llvm@lists.linux.dev
Subject: Re: [PATCH] perf annotate-data: Show event name in the data-type title
Date: Fri, 30 Oct 2026 23:09:11 -0700 [thread overview]
Message-ID: <auWGB-ONmDMktvRB@z2> (raw)
In-Reply-To: <20261029071510.460873-1-wutengda@huaweicloud.com>
On Thu, Oct 29, 2026 at 07:15:10AM +0000, Tengda Wu wrote:
> The same data type can be annotated for multiple sampling events, in
> which case the data-type header is printed once per event. But the
> header only shows the type name and the sample count, so several
> identical-looking headers are emitted and the user cannot tell which
> event each of them corresponds to. E.g.
>
> $ perf annotate --data-type=page --stdio --itrace=f
> Annotate type: 'struct page' in [kernel.kallsyms] (59208 samples):
> ============================================================================
> Percent offset size field
> 100.00 0 0x40 struct page {
> 9.95 0 0x8 long unsigned int flags;
>
> Annotate type: 'struct page' in [kernel.kallsyms] (59480 samples):
> ============================================================================
> Percent offset size field
> 100.00 0 0x40 struct page {
> 10.15 0 0x8 long unsigned int flags;
>
> Add the event name to the data-type title in both the TUI browser and
> the stdio output, so each header can be attributed to its event. For
> group events, use evsel__group_desc() to print the group description
> instead of the leader's name, so the whole group is identified.
>
> While at it, size the underline in the stdio output dynamically based
> on the length of the header line instead of using a fixed-width string,
> so it keeps matching the header as the text grows.
>
> Result:
>
> $ perf annotate --data-type=page --stdio --itrace=f
> Annotate type: 'struct page' in [kernel.kallsyms] for l1d-miss (59208 samples):
> ===============================================================================
> Percent offset size field
> 100.00 0 0x40 struct page {
> 9.95 0 0x8 long unsigned int flags;
>
> Annotate type: 'struct page' in [kernel.kallsyms] for l1d-access (59480 samples):
> =================================================================================
> Percent offset size field
> 100.00 0 0x40 struct page {
> 10.15 0 0x8 long unsigned int flags;
>
> Signed-off-by: Tengda Wu <wutengda@huaweicloud.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Thanks,
Namhyung
> ---
> tools/perf/ui/browsers/annotate-data.c | 14 +++++++++++---
> tools/perf/util/annotate-data.c | 15 ++++++++++++---
> 2 files changed, 23 insertions(+), 6 deletions(-)
>
> diff --git a/tools/perf/ui/browsers/annotate-data.c b/tools/perf/ui/browsers/annotate-data.c
> index a15608f6ee53..0f778788df4a 100644
> --- a/tools/perf/ui/browsers/annotate-data.c
> +++ b/tools/perf/ui/browsers/annotate-data.c
> @@ -340,11 +340,19 @@ static int browser__show(struct ui_browser *uib)
> struct hist_entry *he = uib->priv;
> struct annotated_data_type *adt = he->mem_type;
> struct annotated_data_browser *browser = get_browser(uib);
> + struct evsel *evsel = hists_to_evsel(he->hists);
> const char *help = "Press 'h' for help on key bindings";
> - char title[256];
> + const char *evsel_name = evsel__name(evsel);
> + char buf[512];
> + char title[sizeof(buf) + 256];
> +
> + if (evsel__is_group_event(evsel)) {
> + evsel__group_desc(evsel, buf, sizeof(buf));
> + evsel_name = buf;
> + }
>
> - snprintf(title, sizeof(title), "Annotate type: '%s' (%d samples)",
> - adt->self.type_name, he->stat.nr_events);
> + snprintf(title, sizeof(title), "Annotate type: '%s' for %s (%d samples)",
> + adt->self.type_name, evsel_name, he->stat.nr_events);
>
> if (ui_browser__show(uib, title, help) < 0)
> return -1;
> diff --git a/tools/perf/util/annotate-data.c b/tools/perf/util/annotate-data.c
> index b3de1ca688a2..25e3169ea8ad 100644
> --- a/tools/perf/util/annotate-data.c
> +++ b/tools/perf/util/annotate-data.c
> @@ -2149,16 +2149,23 @@ static void print_annotated_data_header(struct hist_entry *he, struct evsel *evs
> int nr_samples = he->stat.nr_events;
> int width = 7;
> const char *val_hdr = "Percent";
> + const char *evsel_name = evsel__name(evsel);
> + char buf[512];
> + int len;
>
> if (evsel__is_group_event(evsel)) {
> struct hist_entry *pair;
>
> list_for_each_entry(pair, &he->pairs.head, pairs.node)
> nr_samples += pair->stat.nr_events;
> +
> + evsel__group_desc(evsel, buf, sizeof(buf));
> + evsel_name = buf;
> }
>
> - printf("Annotate type: '%s' in %s (%d samples):\n",
> - he->mem_type->self.type_name, dso__name(dso), nr_samples);
> + len = printf("Annotate type: '%s' in %s for %s (%d samples):\n",
> + he->mem_type->self.type_name, dso__name(dso), evsel_name,
> + nr_samples);
>
> if (evsel__is_group_event(evsel)) {
> struct evsel *pos;
> @@ -2183,7 +2190,9 @@ static void print_annotated_data_header(struct hist_entry *he, struct evsel *evs
> val_hdr = "Samples";
> }
>
> - printf("============================================================================\n");
> + while (len-- > 1)
> + putchar('=');
> + putchar('\n');
> printf("%*s %10s %10s %s\n", (width + 1) * nr_members, val_hdr,
> "offset", "size", "field");
> }
>
> base-commit: 151a134a4e6ec66f69db8750c66006f5207f7a9d
> --
> 2.34.1
>
prev parent reply other threads:[~2026-10-31 6:09 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-29 7:15 Tengda Wu
2026-10-31 6:09 ` Namhyung Kim [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=auWGB-ONmDMktvRB@z2 \
--to=namhyung@kernel.org \
--cc=acme@kernel.org \
--cc=adrian.hunter@intel.com \
--cc=alexander.shishkin@linux.intel.com \
--cc=irogers@google.com \
--cc=james.clark@linaro.org \
--cc=kim.phillips@arm.com \
--cc=leo.yan@linux.dev \
--cc=lihuafei1@huawei.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=llvm@lists.linux.dev \
--cc=mark.rutland@arm.com \
--cc=mingo@redhat.com \
--cc=morbo@google.com \
--cc=nick.desaulniers+lkml@gmail.com \
--cc=peterz@infradead.org \
--cc=wutengda@huaweicloud.com \
--cc=xueshuai@linux.alibaba.com \
--cc=zli94@ncsu.edu \
/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
all inboxes | Powered by JetHome®