mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Ian Rogers <irogers@google.com>
To: Arnaldo Carvalho de Melo <acme@kernel.org>,
	Namhyung Kim <namhyung@kernel.org>,
	 Aaron Tomlin <atomlin@atomlin.com>
Cc: Howard Chu <howardchu95@gmail.com>,
	Jakub Brnak <jbrnak@redhat.com>,
	 Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>, Jiri Olsa <jolsa@kernel.org>,
	 Adrian Hunter <adrian.hunter@intel.com>,
	James Clark <james.clark@linaro.org>,
	 linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org,
	 Ian Rogers <irogers@google.com>
Subject: [PATCH v6 10/26] perf trace: Only take augmented arguments from the BPF output event
Date: Mon, 28 Sep 2026 11:25:49 -0700	[thread overview]
Message-ID: <20260928182605.3649015-11-irogers@google.com> (raw)
In-Reply-To: <20260928182605.3649015-1-irogers@google.com>

syscall__augmented_args() treats any data after the syscall arguments as
augmented arguments. trace__sys_enter() avoids this for
raw_syscalls:sys_enter, but trace__fprintf_sys_enter() also handles
syscalls:sys_enter_* tracepoint events, whose trailing data since v6.19
holds the __data_loc strings of internal fields. Only the BPF output
event carries augmented arguments, so only look for them there.

Assisted-by: Antigravity:gemini-3.1-pro
Signed-off-by: Ian Rogers <irogers@google.com>
---
 tools/perf/builtin-trace.c | 25 ++++++++++---------------
 1 file changed, 10 insertions(+), 15 deletions(-)

diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index 6a54d8bff534..1f89f85ba2e4 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -2947,7 +2947,8 @@ static int trace__fprintf_sample(struct trace *trace, struct perf_sample *sample
 	return printed;
 }
 
-static void *syscall__augmented_args(struct syscall *sc, struct perf_sample *sample, int *augmented_args_size, int raw_augmented_args_size)
+static void *syscall__augmented_args(struct trace *trace, struct syscall *sc,
+				     struct perf_sample *sample, int *augmented_args_size)
 {
 	/*
 	 * For now with BPF raw_augmented we hook into raw_syscalls:sys_enter
@@ -2963,9 +2964,14 @@ static void *syscall__augmented_args(struct syscall *sc, struct perf_sample *sam
 	 * use syscalls:sys_enter_NAME, so that we reduce the kernel/userspace
 	 * traffic to just what is needed for each syscall.
 	 */
-	int args_size = raw_augmented_args_size ?: sc->args_size;
+	int args_size = trace->raw_augmented_syscalls_args_size ?: sc->args_size;
 	static uintptr_t argbuf[1024]; /* assuming single-threaded */
 
+	*augmented_args_size = 0;
+	/* Only the BPF output event carries augmented arguments. */
+	if (sample->evsel != trace->syscalls.events.bpf_output)
+		return NULL;
+
 	*augmented_args_size = sample->raw_size - args_size;
 	/*
 	 * The raw data is padded to a u64 boundary with stale bytes, so less
@@ -3025,18 +3031,7 @@ static int trace__sys_enter(struct trace *trace,
 
 	if (!(trace->duration_filter || trace->summary_only || trace->min_stack))
 		trace__printf_interrupted_entry(trace);
-	/*
-	 * If this is raw_syscalls.sys_enter, then it always comes with the 6 possible
-	 * arguments, even if the syscall being handled, say "openat", uses only 4 arguments
-	 * this breaks syscall__augmented_args() check for augmented args, as we calculate
-	 * syscall->args_size using each syscalls:sys_enter_NAME tracefs format file,
-	 * so when handling, say the openat syscall, we end up getting 6 args for the
-	 * raw_syscalls:sys_enter event, when we expected just 4, we end up mistakenly
-	 * thinking that the extra 2 u64 args are the augmented filename, so just check
-	 * here and avoid using augmented syscalls when the evsel is the raw_syscalls one.
-	 */
-	if (evsel != trace->syscalls.events.sys_enter)
-		augmented_args = syscall__augmented_args(sc, sample, &augmented_args_size, trace->raw_augmented_syscalls_args_size);
+	augmented_args = syscall__augmented_args(trace, sc, sample, &augmented_args_size);
 	ttrace->entry_time = sample->time;
 	ttrace->entry_cpu = sample->cpu;
 	msg = ttrace->entry_str;
@@ -3099,7 +3094,7 @@ static int trace__fprintf_sys_enter(struct trace *trace, struct perf_sample *sam
 		goto out_put;
 
 	args = perf_evsel__sc_tp_ptr(args, sample);
-	augmented_args = syscall__augmented_args(sc, sample, &augmented_args_size, trace->raw_augmented_syscalls_args_size);
+	augmented_args = syscall__augmented_args(trace, sc, sample, &augmented_args_size);
 	printed += syscall__scnprintf_args(sc, msg, sizeof(msg), args, augmented_args, augmented_args_size, trace, thread);
 	fprintf(trace->output, "%.*s", (int)printed, msg);
 	err = 0;
-- 
2.56.0.rc1.315.gc6ed9934b7-goog


  parent reply	other threads:[~2026-09-28 18:26 UTC|newest]

Thread overview: 126+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-17  6:42 [PATCH v1 00/13] perf trace: Fix BPF filtering and make tracing tests non-exclusive Ian Rogers
2026-09-17  6:42 ` [PATCH v1 01/13] perf trace: Start BPF summary before starting workload Ian Rogers
2026-09-17  6:42 ` [PATCH v1 02/13] perf trace: Skip internal tracepoint fields in formatting and beauty map Ian Rogers
2026-09-17  6:42 ` [PATCH v1 03/13] perf trace: Do not set unaugmented BPF program on sys_exit map Ian Rogers
2026-09-17  6:42 ` [PATCH v1 04/13] perf trace: Filter events in BPF and avoid tracepoint vetoes Ian Rogers
2026-09-17  6:42 ` [PATCH v1 05/13] perf trace: Handle fork and exit directly in BPF filter maps Ian Rogers
2026-09-17  6:42 ` [PATCH v1 06/13] perf test test_task_analyzer: Isolate in temporary directory and make non-exclusive Ian Rogers
2026-09-17  6:42 ` [PATCH v1 07/13] perf test common: Do not globally disable tracing events in clear_all_probes Ian Rogers
2026-09-17  6:42 ` [PATCH v1 08/13] perf test probe_vfs_getname: Scope probe name to PID and make non-exclusive Ian Rogers
2026-09-17  6:42 ` [PATCH v1 09/13] perf test record+probe_libc_inet_pton: Scope event to PID, add retries, " Ian Rogers
2026-09-17  6:42 ` [PATCH v1 10/13] perf test trace_summary: Improve error diagnostics Ian Rogers
2026-09-17  6:42 ` [PATCH v1 11/13] perf test trace_btf_general: Drop --max-events=1 and make non-exclusive Ian Rogers
2026-09-17  6:42 ` [PATCH v1 12/13] perf test trace_summary: Make non-exclusive Ian Rogers
2026-09-17  6:42 ` [PATCH v1 13/13] perf test uprobe_from_different_cu: Scope probe name to PID Ian Rogers
2026-09-17 16:38 ` [PATCH v2 00/14] perf trace: Fix BPF filtering and make tracing tests non-exclusive Ian Rogers
2026-09-17 16:38   ` [PATCH v2 01/14] perf trace: Include the headers declaring pid_t and strcmp Ian Rogers
2026-09-17 16:38   ` [PATCH v2 02/14] perf trace: Start BPF summary before starting workload Ian Rogers
2026-09-17 16:38   ` [PATCH v2 03/14] perf trace: Skip internal tracepoint fields in formatting and beauty map Ian Rogers
2026-09-17 16:38   ` [PATCH v2 04/14] perf trace: Do not set unaugmented BPF program on sys_exit map Ian Rogers
2026-09-17 16:38   ` [PATCH v2 05/14] perf trace: Filter events in BPF and avoid tracepoint vetoes Ian Rogers
2026-09-17 16:38   ` [PATCH v2 06/14] perf trace: Handle fork and exit directly in BPF filter maps Ian Rogers
2026-09-17 16:38   ` [PATCH v2 07/14] perf test test_task_analyzer: Isolate in temporary directory and make non-exclusive Ian Rogers
2026-09-17 16:38   ` [PATCH v2 08/14] perf test common: Do not globally disable tracing events in clear_all_probes Ian Rogers
2026-09-17 16:38   ` [PATCH v2 09/14] perf test probe_vfs_getname: Scope probe name to PID and make non-exclusive Ian Rogers
2026-09-17 16:38   ` [PATCH v2 10/14] perf test record+probe_libc_inet_pton: Scope event to PID, add retries, " Ian Rogers
2026-09-17 16:38   ` [PATCH v2 11/14] perf test trace_summary: Improve error diagnostics Ian Rogers
2026-09-17 16:39   ` [PATCH v2 12/14] perf test trace_btf_general: Drop --max-events=1 and make non-exclusive Ian Rogers
2026-09-17 16:39   ` [PATCH v2 13/14] perf test trace_summary: Make non-exclusive Ian Rogers
2026-09-17 16:39   ` [PATCH v2 14/14] perf test uprobe_from_different_cu: Scope probe name to PID Ian Rogers
2026-09-18 14:06   ` [PATCH v3 00/16] perf trace: Fix BPF filtering and make tracing tests non-exclusive Ian Rogers
2026-09-18 14:06     ` [PATCH v3 01/16] perf trace: Include the headers declaring pid_t, strcmp and assert Ian Rogers
2026-09-18 14:06     ` [PATCH v3 02/16] perf trace: Free the whole evsel_trace in evsel__put_and_free_priv Ian Rogers
2026-09-18 14:06     ` [PATCH v3 03/16] perf trace: Start BPF summary before starting workload Ian Rogers
2026-09-18 14:06     ` [PATCH v3 04/16] perf trace: Skip internal tracepoint fields in formatting and beauty map Ian Rogers
2026-09-18 14:06     ` [PATCH v3 05/16] perf trace: Do not set unaugmented BPF program on sys_exit map Ian Rogers
2026-09-18 14:06     ` [PATCH v3 06/16] perf trace: Filter events in BPF and avoid tracepoint vetoes Ian Rogers
2026-09-18 14:06     ` [PATCH v3 07/16] perf trace: Handle fork and exit directly in BPF filter maps Ian Rogers
2026-09-18 14:06     ` [PATCH v3 08/16] perf trace: Enumerate the target again once BPF is attached Ian Rogers
2026-09-18 14:06     ` [PATCH v3 09/16] perf test test_task_analyzer: Isolate in temporary directory and make non-exclusive Ian Rogers
2026-09-18 14:06     ` [PATCH v3 10/16] perf test common: Only disable probes in clear_all_probes Ian Rogers
2026-09-18 14:06     ` [PATCH v3 11/16] perf test probe_vfs_getname: Scope probe name to PID and make non-exclusive Ian Rogers
2026-09-18 14:06     ` [PATCH v3 12/16] perf test record+probe_libc_inet_pton: Scope event to PID, add retries, " Ian Rogers
2026-09-18 14:06     ` [PATCH v3 13/16] perf test trace_summary: Improve error diagnostics Ian Rogers
2026-09-18 14:06     ` [PATCH v3 14/16] perf test trace_btf_general: Drop --max-events=1 and make non-exclusive Ian Rogers
2026-09-18 14:06     ` [PATCH v3 15/16] perf test trace_summary: Make non-exclusive Ian Rogers
2026-09-18 14:06     ` [PATCH v3 16/16] perf test uprobe_from_different_cu: Scope probe name to PID Ian Rogers
2026-09-18 21:19     ` [PATCH v4 00/18] perf trace: Fix BPF filtering and make tracing tests non-exclusive Ian Rogers
2026-09-18 21:19       ` [PATCH v4 01/18] perf trace: Include the headers declaring pid_t, strcmp and assert Ian Rogers
2026-09-18 21:19       ` [PATCH v4 02/18] perf trace: Free the whole evsel_trace in evsel__put_and_free_priv Ian Rogers
2026-09-18 21:19       ` [PATCH v4 03/18] perf evsel: Report an allocation failure as ENOMEM when setting filters Ian Rogers
2026-09-18 21:19       ` [PATCH v4 04/18] perf trace: Start BPF summary before starting workload Ian Rogers
2026-09-18 21:19       ` [PATCH v4 05/18] perf trace: Skip internal tracepoint fields in formatting and beauty map Ian Rogers
2026-09-18 21:19       ` [PATCH v4 06/18] perf trace: Do not set unaugmented BPF program on sys_exit map Ian Rogers
2026-09-18 21:19       ` [PATCH v4 07/18] perf trace: Filter events in BPF and avoid tracepoint vetoes Ian Rogers
2026-09-18 21:19       ` [PATCH v4 08/18] perf trace: Handle fork and exit directly in BPF filter maps Ian Rogers
2026-09-18 21:19       ` [PATCH v4 09/18] perf trace: Enumerate the target again once BPF is attached Ian Rogers
2026-09-18 21:19       ` [PATCH v4 10/18] perf trace: Drop targets that died before they were filtered Ian Rogers
2026-09-18 21:19       ` [PATCH v4 11/18] perf test test_task_analyzer: Isolate in temporary directory and make non-exclusive Ian Rogers
2026-09-18 21:19       ` [PATCH v4 12/18] perf test common: Only disable probes in clear_all_probes Ian Rogers
2026-09-18 21:19       ` [PATCH v4 13/18] perf test probe_vfs_getname: Scope probe name to PID and make non-exclusive Ian Rogers
2026-09-18 21:19       ` [PATCH v4 14/18] perf test record+probe_libc_inet_pton: Scope event to PID, add retries, " Ian Rogers
2026-09-18 21:19       ` [PATCH v4 15/18] perf test trace_summary: Improve error diagnostics Ian Rogers
2026-09-18 21:19       ` [PATCH v4 16/18] perf test trace_btf_general: Drop --max-events=1 and make non-exclusive Ian Rogers
2026-09-18 21:19       ` [PATCH v4 17/18] perf test trace_summary: Make non-exclusive Ian Rogers
2026-09-18 21:19       ` [PATCH v4 18/18] perf test uprobe_from_different_cu: Scope probe name to PID Ian Rogers
2026-09-22 13:11       ` [PATCH v4 00/18] perf trace: Fix BPF filtering and make tracing tests non-exclusive Arnaldo Carvalho de Melo
2026-09-22 22:43         ` Ian Rogers
2026-09-23  7:13       ` [PATCH v5 00/23] " Ian Rogers
2026-09-23  7:13         ` [PATCH v5 01/23] perf trace: Set the augmented arg header in the augmenters that omit it Ian Rogers
2026-09-24  5:39           ` Namhyung Kim
2026-09-23  7:13         ` [PATCH v5 02/23] perf trace: Include the augmented arg header in nanosleep's payload length Ian Rogers
2026-09-23  7:13         ` [PATCH v5 03/23] perf trace: Include the headers declaring pid_t, strcmp and assert Ian Rogers
2026-09-24  5:44           ` Namhyung Kim
2026-09-23  7:13         ` [PATCH v5 04/23] perf trace: Free the whole evsel_trace in evsel__put_and_free_priv Ian Rogers
2026-09-23  7:13         ` [PATCH v5 05/23] perf evsel: Report an allocation failure as ENOMEM when setting filters Ian Rogers
2026-09-23  7:13         ` [PATCH v5 06/23] perf trace: Start BPF summary before starting workload Ian Rogers
2026-09-23  7:13         ` [PATCH v5 07/23] perf trace: Skip internal tracepoint fields in formatting and beauty map Ian Rogers
2026-09-24  6:44           ` Namhyung Kim
2026-09-23  7:13         ` [PATCH v5 08/23] perf trace: Bounds check augmented arguments before reading them Ian Rogers
2026-09-24  6:58           ` Namhyung Kim
2026-09-23  7:13         ` [PATCH v5 09/23] perf trace: Bound the fixed size augmented argument beautifiers Ian Rogers
2026-09-23  7:13         ` [PATCH v5 10/23] perf trace: Do not read sample padding as an augmented argument Ian Rogers
2026-09-23  7:13         ` [PATCH v5 11/23] perf trace: Do not set unaugmented BPF program on sys_exit map Ian Rogers
2026-09-24 22:49           ` Namhyung Kim
2026-09-23  7:13         ` [PATCH v5 12/23] perf trace: Filter events in BPF and avoid tracepoint vetoes Ian Rogers
2026-09-25  2:59           ` Namhyung Kim
2026-09-25  4:10             ` Ian Rogers
2026-09-23  7:13         ` [PATCH v5 13/23] perf trace: Handle fork and exit directly in BPF filter maps Ian Rogers
2026-09-23  7:13         ` [PATCH v5 14/23] perf trace: Enumerate the target again once BPF is attached Ian Rogers
2026-09-23  7:13         ` [PATCH v5 15/23] perf trace: Drop targets that died before they were filtered Ian Rogers
2026-09-23  7:13         ` [PATCH v5 16/23] perf test test_task_analyzer: Isolate in temporary directory and make non-exclusive Ian Rogers
2026-09-23  7:13         ` [PATCH v5 17/23] perf test common: Only disable probes in clear_all_probes Ian Rogers
2026-09-23  7:13         ` [PATCH v5 18/23] perf test probe_vfs_getname: Scope probe name to PID and make non-exclusive Ian Rogers
2026-09-23  7:13         ` [PATCH v5 19/23] perf test record+probe_libc_inet_pton: Scope event to PID, add retries, " Ian Rogers
2026-09-23  7:14         ` [PATCH v5 20/23] perf test trace_summary: Improve error diagnostics Ian Rogers
2026-09-23  7:14         ` [PATCH v5 21/23] perf test trace_btf_general: Drop --max-events=1 and make non-exclusive Ian Rogers
2026-09-23  7:14         ` [PATCH v5 22/23] perf test trace_summary: Make non-exclusive Ian Rogers
2026-09-23  7:14         ` [PATCH v5 23/23] perf test uprobe_from_different_cu: Scope probe name to PID Ian Rogers
2026-09-24  7:05         ` [PATCH v5 00/23] perf trace: Fix BPF filtering and make tracing tests non-exclusive Namhyung Kim
2026-09-28 18:25         ` [PATCH v6 00/26] " Ian Rogers
2026-09-28 18:25           ` [PATCH v6 01/26] perf trace: Set the augmented arg header in the augmenters that omit it Ian Rogers
2026-09-28 18:25           ` [PATCH v6 02/26] perf trace: Include the augmented arg header in nanosleep's payload length Ian Rogers
2026-09-28 18:25           ` [PATCH v6 03/26] perf trace: Don't read sample padding as an augmented argument Ian Rogers
2026-09-28 18:25           ` [PATCH v6 04/26] perf trace: Bounds check augmented arguments before reading them Ian Rogers
2026-09-28 18:25           ` [PATCH v6 05/26] perf trace: Bound the fixed size augmented argument beautifiers Ian Rogers
2026-09-28 18:25           ` [PATCH v6 06/26] perf trace: Copy sockaddr arguments by their length Ian Rogers
2026-09-28 18:25           ` [PATCH v6 07/26] perf trace: Start BPF summary before starting workload Ian Rogers
2026-09-28 18:25           ` [PATCH v6 08/26] perf trace: Stop at internal fields when walking syscall arguments Ian Rogers
2026-09-28 18:25           ` [PATCH v6 09/26] perf trace: Don't allocate syscall arg formats for internal fields Ian Rogers
2026-09-28 18:25           ` Ian Rogers [this message]
2026-09-28 18:25           ` [PATCH v6 11/26] perf trace: Split unaugmented sys_exit program Ian Rogers
2026-09-28 18:25           ` [PATCH v6 12/26] perf trace: Use the CPU map index for the BPF output event's fds Ian Rogers
2026-09-28 18:25           ` [PATCH v6 13/26] perf trace: Destroy the BPF skeleton if it fails to load Ian Rogers
2026-09-28 18:25           ` [PATCH v6 14/26] perf thread_map: Add thread_map__tgid() Ian Rogers
2026-09-28 18:25           ` [PATCH v6 15/26] perf trace: Filter the target's tasks in BPF Ian Rogers
2026-09-28 18:25           ` [PATCH v6 16/26] perf trace: Open the BPF output event on every CPU for task targets Ian Rogers
2026-09-28 18:25           ` [PATCH v6 17/26] perf trace: Do not return 0 from syscall tracepoint BPF Ian Rogers
2026-09-28 18:25           ` [PATCH v6 18/26] perf trace: Remove unused code Ian Rogers
2026-09-28 18:25           ` [PATCH v6 19/26] perf trace: Add an option to disable syscall augmentation Ian Rogers
2026-09-28 18:25           ` [PATCH v6 20/26] perf test common: Only disable probes in clear_all_probes Ian Rogers
2026-09-28 18:26           ` [PATCH v6 21/26] perf test probe_vfs_getname: Scope probe name to PID and make non-exclusive Ian Rogers
2026-09-28 18:26           ` [PATCH v6 22/26] perf test record+probe_libc_inet_pton: Scope event " Ian Rogers
2026-09-28 18:26           ` [PATCH v6 23/26] perf test trace_summary: Improve error diagnostics Ian Rogers
2026-09-28 18:26           ` [PATCH v6 24/26] perf test trace_btf_general: Drop --max-events=1 and make non-exclusive Ian Rogers
2026-09-28 18:26           ` [PATCH v6 25/26] perf test: Remove exclusive tag from perf trace tests Ian Rogers
2026-09-28 18:26           ` [PATCH v6 26/26] perf test uprobe_from_different_cu: Scope probe name to PID Ian Rogers

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=20260928182605.3649015-11-irogers@google.com \
    --to=irogers@google.com \
    --cc=acme@kernel.org \
    --cc=adrian.hunter@intel.com \
    --cc=atomlin@atomlin.com \
    --cc=howardchu95@gmail.com \
    --cc=james.clark@linaro.org \
    --cc=jbrnak@redhat.com \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    /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®