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]
Message-Id: <1380113447-17144-15-git-send-email-jolsa@redhat.com>
Date:	Wed, 25 Sep 2013 14:50:40 +0200
From:	Jiri Olsa <jolsa@...hat.com>
To:	linux-kernel@...r.kernel.org
Cc:	Jiri Olsa <jolsa@...hat.com>,
	Arnaldo Carvalho de Melo <acme@...hat.com>,
	Corey Ashford <cjashfor@...ux.vnet.ibm.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	Ingo Molnar <mingo@...e.hu>, Paul Mackerras <paulus@...ba.org>,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>
Subject: [PATCH 14/21] perf tests: Adding event group toggling test

Adding ioctl toggle test to measure instructions
enclosed within following syscall groups:

  geteuid
    openat
     --> measure instructions
    close
  getppid

To show/test we could chain multiple togglers.

Signed-off-by: Jiri Olsa <jolsa@...hat.com>
Cc: Arnaldo Carvalho de Melo <acme@...hat.com>
Cc: Corey Ashford <cjashfor@...ux.vnet.ibm.com>
Cc: Frederic Weisbecker <fweisbec@...il.com>
Cc: Ingo Molnar <mingo@...e.hu>
Cc: Paul Mackerras <paulus@...ba.org>
Cc: Peter Zijlstra <a.p.zijlstra@...llo.nl>
Cc: Arnaldo Carvalho de Melo <acme@...hat.com>
---
 tools/perf/Makefile                   |   1 +
 tools/perf/tests/builtin-test.c       |   4 +
 tools/perf/tests/tests.h              |   1 +
 tools/perf/tests/toggle-event-group.c | 195 ++++++++++++++++++++++++++++++++++
 4 files changed, 201 insertions(+)
 create mode 100644 tools/perf/tests/toggle-event-group.c

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 2072389..90d7127 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -399,6 +399,7 @@ LIB_OBJS += $(OUTPUT)tests/code-reading.o
 LIB_OBJS += $(OUTPUT)tests/sample-parsing.o
 LIB_OBJS += $(OUTPUT)tests/parse-no-sample-id-all.o
 LIB_OBJS += $(OUTPUT)tests/toggle-event-raw.o
+LIB_OBJS += $(OUTPUT)tests/toggle-event-group.o
 ifeq ($(RAW_ARCH),x86_64)
 LIB_OBJS += $(OUTPUT)arch/x86/tests/toggle-event-raw-64.o
 endif
diff --git a/tools/perf/tests/builtin-test.c b/tools/perf/tests/builtin-test.c
index db9d924b..7e96550 100644
--- a/tools/perf/tests/builtin-test.c
+++ b/tools/perf/tests/builtin-test.c
@@ -120,6 +120,10 @@ static struct test {
 		.func = test__toggle_event_raw,
 	},
 	{
+		.desc = "Toggle event group",
+		.func = test__toggle_event_group,
+	},
+	{
 		.func = NULL,
 	},
 };
diff --git a/tools/perf/tests/tests.h b/tools/perf/tests/tests.h
index 4f2a8a1..db692bf 100644
--- a/tools/perf/tests/tests.h
+++ b/tools/perf/tests/tests.h
@@ -41,5 +41,6 @@ int test__sample_parsing(void);
 int test__keep_tracking(void);
 int test__parse_no_sample_id_all(void);
 int test__toggle_event_raw(void);
+int test__toggle_event_group(void);
 
 #endif /* TESTS_H */
diff --git a/tools/perf/tests/toggle-event-group.c b/tools/perf/tests/toggle-event-group.c
new file mode 100644
index 0000000..e781b30
--- /dev/null
+++ b/tools/perf/tests/toggle-event-group.c
@@ -0,0 +1,195 @@
+#include <sys/types.h>
+#include <unistd.h>
+#include <traceevent/event-parse.h>
+#include "thread_map.h"
+#include "evsel.h"
+#include "debug.h"
+#include "tests.h"
+
+/*
+ * We want to toggle instructions on/off only after chained
+ * execution of defined tracepoints, like:
+ *
+ * getuid();
+ *    openat();
+ *       instructions to count
+ *    close();
+ * getppid();
+ *
+ * This test creates following events:
+ *
+ * 1) tracepoint sys_enter_getuid
+ * 2) tracepoint sys_enter_getppid
+ * 3) tracepoint sys_enter_openat
+ * 4) tracepoint sys_enter_close
+ * 5) HW event instruction
+ *
+ *
+ * Events 3) and 4) are created as a group with 3) as the leader.
+ * Events 3) and 4) toggle ON and OFF respectively event 5).
+ * Events 1) and 2) toggle ON and OFF respectively event 3).
+ *
+ * This means:
+ *   - when the workload executes getuid(), the group (events 3
+ *     and 4) is toggled ON.
+ *   - when the workload executes close, the group (events 3
+ *     and 4) is toggled OFF.
+ *   - when the workload executes started events 3) and 4) they
+ *     toggle ON and OFF respectively instructions event.
+ *
+ */
+
+extern int test__toggle_event_raw_arch(void);
+
+static int get_tp_id(const char *name)
+{
+	struct event_format *tp_format = event_format__new("syscalls", name);
+	u64 id = 0;
+
+	if (tp_format) {
+		id = tp_format->id;
+		pevent_free_format(tp_format);
+	}
+
+	return id;
+}
+
+#ifndef __x86_64__
+int test__toggle_event_group(void)
+{
+	pr_err("The toggle event test not implemented for arch.\n");
+	return 0;
+}
+#else
+
+static int test(void)
+{
+	int instr;
+
+	getuid();
+	instr = test__toggle_event_raw_arch();
+	getppid();
+	return instr;
+}
+
+#ifndef PERF_EVENT_IOC_SET_TOGGLE
+#define PERF_EVENT_IOC_SET_TOGGLE 1074275336
+#endif
+
+static int toggle_event(int fd_event, int flag, int fd_toggled)
+{
+	u64 args[2] = { fd_toggled, flag };
+	return ioctl(fd_event, PERF_EVENT_IOC_SET_TOGGLE, args);
+}
+
+int test__toggle_event_group(void)
+{
+	struct perf_event_attr attr_group_on = {
+		.type   = PERF_TYPE_TRACEPOINT,
+		.config = get_tp_id("sys_enter_getuid"),
+		.sample_period = 1,
+	};
+	struct perf_event_attr attr_group_off = {
+		.type   = PERF_TYPE_TRACEPOINT,
+		.config = get_tp_id("sys_enter_getppid"),
+		.sample_period = 1,
+	};
+	struct perf_event_attr attr_on = {
+		.type   = PERF_TYPE_TRACEPOINT,
+		.config = get_tp_id("sys_enter_openat"),
+		.sample_period = 1,
+		.paused		= 1,
+	};
+	struct perf_event_attr attr_off = {
+		.type   = PERF_TYPE_TRACEPOINT,
+		.config = get_tp_id("sys_enter_close"),
+		.sample_period = 1,
+		.paused		= 1,
+	};
+	struct perf_event_attr attr_instr = {
+		.type		= PERF_TYPE_HARDWARE,
+		.config		= PERF_COUNT_HW_INSTRUCTIONS,
+		.paused		= 1,
+		.exclude_kernel = 1,
+		.exclude_hv	= 1,
+	};
+	int fd_group_on, fd_group_off, fd_on, fd_off, fd_instr;
+	__u64 value, instr;
+
+	fd_instr = sys_perf_event_open(&attr_instr, 0, -1, -1, 0);
+	if (fd_instr < 0) {
+		pr_err("failed to open instruction event, errno %d\n", errno);
+		return -1;
+	}
+
+	fd_on = sys_perf_event_open(&attr_on, 0, -1, -1, 0);
+	if (fd_on < 0) {
+		pr_err("failed to open 'on' event, errno %d\n", errno);
+		return -1;
+	}
+
+	fd_off = sys_perf_event_open(&attr_off, 0, -1, fd_on, 0);
+	if (fd_off < 0) {
+		pr_err("failed to open 'off' event, errno %d\n", errno);
+		return -1;
+	}
+
+	if (toggle_event(fd_on, PERF_FLAG_TOGGLE_ON, fd_instr)) {
+		pr_err("failed to set toggle 'on', errno %d\n", errno);
+		return -1;
+	}
+
+	if (toggle_event(fd_off, PERF_FLAG_TOGGLE_OFF, fd_instr)) {
+		pr_err("failed to set toggle 'off', errno %d\n", errno);
+		return -1;
+	}
+
+	fd_group_on = sys_perf_event_open(&attr_group_on, 0, -1,
+					  fd_on, PERF_FLAG_TOGGLE_ON);
+	if (fd_group_on < 0) {
+		pr_err("failed to open 'group_on' event, errno %d\n", errno);
+		return -1;
+	}
+
+	fd_group_off = sys_perf_event_open(&attr_group_off, 0, -1,
+					   fd_on, PERF_FLAG_TOGGLE_OFF);
+	if (fd_group_off < 0) {
+		pr_err("failed to open 'group_off' event, errno %d\n", errno);
+		return -1;
+	}
+
+#define READ(i, exp)								\
+do {										\
+	if (sizeof(value) != read(fd_instr, &value, sizeof(value))) {		\
+		pr_err("failed to read instruction event, errno %d\n", errno);	\
+		return -1;							\
+	}									\
+	pr_debug("%d got count %llu vs %lu\n", i, value, (unsigned long) exp);	\
+} while (0)
+
+
+	READ(1, 0);
+
+	test__toggle_event_raw_arch();
+
+	READ(2, 0);
+
+	instr = test();
+
+	READ(3, instr);
+
+	test__toggle_event_raw_arch();
+
+	READ(4, instr);
+
+	close(fd_on);
+	close(fd_off);
+	close(fd_group_on);
+	close(fd_group_off);
+
+	READ(5, instr);
+
+	close(fd_instr);
+	return instr != value;
+}
+#endif /* __x86_64__ */
-- 
1.7.11.7

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ