linux-kernel - [PATCH 1/2] Add force-abs-timestamp trace option.

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-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 18 Jun 2020 15:15:40 -0700
From:   Korben Rusek <korben@...gle.com>
To:     Steven Rostedt <rostedt@...dmis.org>,
        Ingo Molnar <mingo@...hat.com>,
        Jonathan Corbet <corbet@....net>
Cc:     Peter Shier <pshier@...gle.com>, linux-doc@...r.kernel.org,
        linux-kernel@...r.kernel.org, jbehmer@...gle.com,
        franksie@...gle.com, Korben Rusek <korben@...gle.com>
Subject: [PATCH 1/2] Add force-abs-timestamp trace option.

Normally a data event includes the delta time between itself and
the previous data event. But if a write interrupts another write
then the event cannot check the timestamp of the previous event
and so it uses a delta of 0.

Adds a new trace option: force-abs-timestamp. When this is on
each event is proceeded by an absolute timestamp event.

Reviewed-by: Peter Shier <pshier@...gle.com>
Signed-off-by: Korben Rusek <korben@...gle.com>
---
 Documentation/trace/ftrace.rst | 5 +++++
 kernel/trace/trace.c           | 4 ++++
 kernel/trace/trace.h           | 1 +
 3 files changed, 10 insertions(+)

diff --git a/Documentation/trace/ftrace.rst b/Documentation/trace/ftrace.rst
index 430a16283103..704264164ffd 100644
--- a/Documentation/trace/ftrace.rst
+++ b/Documentation/trace/ftrace.rst
@@ -1029,6 +1029,7 @@ To see what is available, simply cat the file::
 	irq-info
 	markers
 	noevent-fork
+	noforce-abs-timestamp
 	function-trace
 	nofunction-fork
 	nodisplay-graph
@@ -1212,6 +1213,10 @@ Here are the available options:
 
         This affects PIDs listed in set_event_notrace_pid as well.
 
+  force-abs-timestamp
+	When set, always use absolute timestamps rather than
+	defaulting to deltas.
+
   function-trace
 	The latency tracers will enable function tracing
 	if this option is enabled (default it is). When
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index ec44b0e2a19c..8b5c4e998895 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -4852,6 +4852,12 @@ int set_tracer_flag(struct trace_array *tr, unsigned int mask, int enabled)
 		trace_printk_control(enabled);
 	}
 
+	if (mask == TRACE_ITER_FORCE_ABS_TIMESTAMP) {
+		ring_buffer_set_time_stamp_abs(tr->array_buffer.buffer, enabled);
+#ifdef CONFIG_TRACER_MAX_TRACE
+		ring_buffer_set_time_stamp_abs(tr->max_buffer.buffer, enabled);
+#endif
+	}
 	return 0;
 }
 
diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
index def769df5bf1..cb1ecc540494 100644
--- a/kernel/trace/trace.h
+++ b/kernel/trace/trace.h
@@ -1330,6 +1330,7 @@ extern int trace_get_user(struct trace_parser *parser, const char __user *ubuf,
 		C(MARKERS,		"markers"),		\
 		C(EVENT_FORK,		"event-fork"),		\
 		C(PAUSE_ON_TRACE,	"pause-on-trace"),	\
+		C(FORCE_ABS_TIMESTAMP,	"force-abs-timestamp"),	\
 		FUNCTION_FLAGS					\
 		FGRAPH_FLAGS					\
 		STACK_FLAGS					\
-- 
2.27.0.290.gba653c62da-goog