lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230623082031.1402-1-ravi.bangoria@amd.com>
Date:   Fri, 23 Jun 2023 13:50:31 +0530
From:   Ravi Bangoria <ravi.bangoria@....com>
To:     <acme@...nel.org>
CC:     <ravi.bangoria@....com>, <jolsa@...nel.org>, <namhyung@...nel.org>,
        <irogers@...gle.com>, <peterz@...radead.org>,
        <adrian.hunter@...el.com>, <kan.liang@...el.com>,
        <eranian@...gle.com>, <ak@...ux.intel.com>,
        <linux-perf-users@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <sandipan.das@....com>, <ananth.narayan@....com>,
        <santosh.shukla@....com>
Subject: [RFC] perf record: Use PERF_RECORD_LOST for synthesizing samples from read_format->lost

Currently perf synthesizes PERF_RECORD_LOST_SAMPLES samples for values
returned by read_format->lost. IIUC, PERF_RECORD_LOST_SAMPLES is used
only when hw provides corrupted samples and thus kernel has to drop
those. OTOH, PERF_RECORD_LOST is used when kernel has valid samples but
it fails to push them to ring-buffer because ring-buffer is already full.

So I feel PERF_RECORD_LOST is more appropriate for synthesizing samples
from read_format->lost.

Signed-off-by: Ravi Bangoria <ravi.bangoria@....com>
---
Notes:
 - Posting this as RFC to get feedback. I haven't tested it well.
 - There is one more minor issue: Aggregated Stats shows count of LOST/
   LOST_SAMPLES records instead of actual number of lost events/samples:
     ```
     $ sudo ./perf report -D | tail -20
     Warning:
     Processed 78923 events and lost 1153 chunks!
     Warning:
     Processed 47427 samples and lost 28.30%!
     ...
     Aggregated stats:
               LOST events:       1153  ( 1.5%)
       LOST_SAMPLES events:        132  ( 0.2%)
     ```
   Not sure if this is intentional.

 tools/perf/builtin-record.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index efa03e4ac2c9..5baed42437b2 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -1858,7 +1858,7 @@ record__switch_output(struct record *rec, bool at_exit)
 }
 
 static void __record__save_lost_samples(struct record *rec, struct evsel *evsel,
-					struct perf_record_lost_samples *lost,
+					struct perf_record_lost *lost,
 					int cpu_idx, int thread_idx, u64 lost_count,
 					u16 misc_flag)
 {
@@ -1871,6 +1871,7 @@ static void __record__save_lost_samples(struct record *rec, struct evsel *evsel,
 		sid = xyarray__entry(evsel->core.sample_id, cpu_idx, thread_idx);
 		sample.id = sid->id;
 	}
+	lost->id = sid->id;
 
 	id_hdr_size = perf_event__synthesize_id_sample((void *)(lost + 1),
 						       evsel->core.attr.sample_type, &sample);
@@ -1882,7 +1883,7 @@ static void __record__save_lost_samples(struct record *rec, struct evsel *evsel,
 static void record__read_lost_samples(struct record *rec)
 {
 	struct perf_session *session = rec->session;
-	struct perf_record_lost_samples *lost;
+	struct perf_record_lost *lost;
 	struct evsel *evsel;
 
 	/* there was an error during record__open */
@@ -1895,7 +1896,7 @@ static void record__read_lost_samples(struct record *rec)
 		return;
 	}
 
-	lost->header.type = PERF_RECORD_LOST_SAMPLES;
+	lost->header.type = PERF_RECORD_LOST;
 
 	evlist__for_each_entry(session->evlist, evsel) {
 		struct xyarray *xy = evsel->core.sample_id;
-- 
2.41.0

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ