Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1468857 > unrolled thread

[GIT PULL 00/35] perf/core improvements and fixes

Back to article view | Back to linux.kernel

Contents

  [GIT PULL 00/35] perf/core improvements and fixes Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-23 23:10 +0200
    [PATCH 10/35] perf kvm: Use NSEC_PER_USEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-23 23:50 +0200
    [PATCH 11/35] perf bench futex: Use NSEC_PER_USEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-23 23:50 +0200
    [PATCH 12/35] perf top: Use MSEC_PER_SEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-23 23:50 +0200
    [PATCH 01/35] tools: Introduce tools/include/linux/time64.h for *SEC_PER_*SEC macros Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-23 23:50 +0200
    [PATCH 34/35] perf bpf: Fix typo: "ehough" -> "enough" Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 17/35] perf hists: Add support for header span Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 23/35] perf tools: Use __weak definition from linux/compiler.h Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 09/35] perf record: Use USEC_PER_MSEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 22/35] perf report: Allow configuring the default sort order in ~/.perfconfig Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 27/35] ftrace: probe: Add README entries for k/uprobe-events Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 14/35] perf hists: Add line argument into perf_hpp_fmt's header callback Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 21/35] tools: Copy coresight-pmu.h header file needed by perf tools Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 32/35] perf tools: Fix typo: "ehough" -> "enough" Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 02/35] perf bench numa: Use NSEC_PER_U?SEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 29/35] perf probe: Support hexadecimal casting Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 04/35] perf timechart: Use NSEC_PER_U?SEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 18/35] perf disassemble: Move check for kallsyms + !kcore Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 08/35] perf bench sched-messaging: Use USEC_PER_MSEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 33/35] perf test bpf: Fix typo: "ehough" -> "enough" Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 31/35] ftrace: kprobe: uprobe: Show u8/u16/u32/u64 types in decimal Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 15/35] perf tools tui: Display multiple header lines Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 05/35] perf bench sched-pipe: Use linux/time64.h, USEC_PER_SEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:00 +0200
    [PATCH 35/35] perf record: Fix spelling mistake "Finshed" -> "Finished" Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 00:30 +0200
    [PATCH 03/35] perf sched: Use linux/time64.h Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:00 +0200
    [PATCH 06/35] perf stat: Use *SEC_PER_*SEC macros Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:00 +0200
    [PATCH 16/35] perf tools stdio: Display multiple header lines Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:00 +0200
    [PATCH 25/35] perf hists browser: Remove superfluous null check on map Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:10 +0200
    [PATCH 07/35] perf bench mem: Use USEC_PER_SEC Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:10 +0200
    [PATCH 20/35] perf disassemble: Extract logic to find file to pass to objdump to a separate function Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:10 +0200
    [PATCH 19/35] perf disassemble: Simplify logic for picking the filename to disassemble Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:10 +0200
    [PATCH 24/35] perf tools: Skip running the feature tests for 'make install-doc' Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:10 +0200
    [PATCH 13/35] perf hists: Introduce nr_header_lines into struct perf_hpp_list Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:10 +0200
    [PATCH 30/35] perf probe: Use hexadecimal type by default if possible Arnaldo Carvalho de Melo <acme@kernel.org> - 2016-08-24 01:20 +0200
    Re: [GIT PULL 00/35] perf/core improvements and fixes Ingo Molnar <mingo@kernel.org> - 2016-08-24 11:10 +0200

Page 2 of 2 — ← Prev page 1 [2]

#1468928 — [PATCH 31/35] ftrace: kprobe: uprobe: Show u8/u16/u32/u64 types in decimal

From: Masami Hiramatsu <mhiramat@kernel.org>

Change kprobe/uprobe-tracer to show the arguments type-casted
with u8/u16/u32/u64 in decimal digits instead of hexadecimal.

To minimize compatibility issue, the arguments without type
casting are typed by x64 (or x32 for 32bit arch) by default.

Note: all arguments set by old perf probe without types are
shown in decimal by default.

Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Hemant Kumar <hemant@linux.vnet.ibm.com>
Cc: Naohiro Aota <naohiro.aota@hgst.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/147151076135.12957.14684546093034343894.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 Documentation/trace/kprobetrace.txt     | 5 ++++-
 Documentation/trace/uprobetracer.txt    | 5 ++++-
 kernel/trace/trace_probe.c              | 8 ++++----
 kernel/trace/trace_probe.h              | 2 +-
 tools/perf/Documentation/perf-probe.txt | 5 ++---
 5 files changed, 15 insertions(+), 10 deletions(-)

diff --git a/Documentation/trace/kprobetrace.txt b/Documentation/trace/kprobetrace.txt
index 9109c8e8a5a1..e4991fb1eedc 100644
--- a/Documentation/trace/kprobetrace.txt
+++ b/Documentation/trace/kprobetrace.txt
@@ -54,7 +54,10 @@ Types
 -----
 Several types are supported for fetch-args. Kprobe tracer will access memory
 by given type. Prefix 's' and 'u' means those types are signed and unsigned
-respectively. Traced arguments are shown in decimal (signed) or hex (unsigned).
+respectively. 'x' prefix implies it is unsigned. Traced arguments are shown
+in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32'
+or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and
+x86-64 uses x64).
 String type is a special type, which fetches a "null-terminated" string from
 kernel space. This means it will fail and store NULL if the string container
 has been paged out.
diff --git a/Documentation/trace/uprobetracer.txt b/Documentation/trace/uprobetracer.txt
index 7e6d28c54b75..94b6b4581763 100644
--- a/Documentation/trace/uprobetracer.txt
+++ b/Documentation/trace/uprobetracer.txt
@@ -50,7 +50,10 @@ Types
 -----
 Several types are supported for fetch-args. Uprobe tracer will access memory
 by given type. Prefix 's' and 'u' means those types are signed and unsigned
-respectively. Traced arguments are shown in decimal (signed) or hex (unsigned).
+respectively. 'x' prefix implies it is unsigned. Traced arguments are shown
+in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32'
+or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and
+x86-64 uses x64).
 String type is a special type, which fetches a "null-terminated" string from
 user space.
 Bitfield is another special type, which takes 3 parameters, bit-width, bit-
diff --git a/kernel/trace/trace_probe.c b/kernel/trace/trace_probe.c
index 725af9dcbdff..8c0553d9afd3 100644
--- a/kernel/trace/trace_probe.c
+++ b/kernel/trace/trace_probe.c
@@ -46,10 +46,10 @@ int PRINT_TYPE_FUNC_NAME(tname)(struct trace_seq *s, const char *name,	\
 const char PRINT_TYPE_FMT_NAME(tname)[] = fmt;				\
 NOKPROBE_SYMBOL(PRINT_TYPE_FUNC_NAME(tname));
 
-DEFINE_BASIC_PRINT_TYPE_FUNC(u8,  u8,  "0x%x")
-DEFINE_BASIC_PRINT_TYPE_FUNC(u16, u16, "0x%x")
-DEFINE_BASIC_PRINT_TYPE_FUNC(u32, u32, "0x%x")
-DEFINE_BASIC_PRINT_TYPE_FUNC(u64, u64, "0x%Lx")
+DEFINE_BASIC_PRINT_TYPE_FUNC(u8,  u8,  "%u")
+DEFINE_BASIC_PRINT_TYPE_FUNC(u16, u16, "%u")
+DEFINE_BASIC_PRINT_TYPE_FUNC(u32, u32, "%u")
+DEFINE_BASIC_PRINT_TYPE_FUNC(u64, u64, "%Lu")
 DEFINE_BASIC_PRINT_TYPE_FUNC(s8,  s8,  "%d")
 DEFINE_BASIC_PRINT_TYPE_FUNC(s16, s16, "%d")
 DEFINE_BASIC_PRINT_TYPE_FUNC(s32, s32, "%d")
diff --git a/kernel/trace/trace_probe.h b/kernel/trace/trace_probe.h
index f0c470a10edd..0c0ae54d44c6 100644
--- a/kernel/trace/trace_probe.h
+++ b/kernel/trace/trace_probe.h
@@ -208,7 +208,7 @@ DEFINE_FETCH_##method(u32)		\
 DEFINE_FETCH_##method(u64)
 
 /* Default (unsigned long) fetch type */
-#define __DEFAULT_FETCH_TYPE(t) u##t
+#define __DEFAULT_FETCH_TYPE(t) x##t
 #define _DEFAULT_FETCH_TYPE(t) __DEFAULT_FETCH_TYPE(t)
 #define DEFAULT_FETCH_TYPE _DEFAULT_FETCH_TYPE(BITS_PER_LONG)
 #define DEFAULT_FETCH_TYPE_STR __stringify(DEFAULT_FETCH_TYPE)
diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt
index d217617108a1..f37d123d5dac 100644
--- a/tools/perf/Documentation/perf-probe.txt
+++ b/tools/perf/Documentation/perf-probe.txt
@@ -176,13 +176,12 @@ Each probe argument follows below syntax.
 
 'NAME' specifies the name of this argument (optional). You can use the name of local variable, local data structure member (e.g. var->field, var.field2), local array with fixed index (e.g. array[1], var->array[0], var->pointer[2]), or kprobe-tracer argument format (e.g. $retval, %ax, etc). Note that the name of this argument will be set as the last member name if you specify a local data structure member (e.g. field2 for 'var->field1.field2'.)
 '$vars' and '$params' special arguments are also available for NAME, '$vars' is expanded to the local variables (including function parameters) which can access at given probe point. '$params' is expanded to only the function parameters.
-'TYPE' casts the type of this argument (optional). If omitted, perf probe automatically set the type based on debuginfo. Currently, basic types (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal integers (x/x8/x16/x32/x64), signedness casting (u/s), "string" and bitfield are supported. (see TYPES for detail)
-
+'TYPE' casts the type of this argument (optional). If omitted, perf probe automatically set the type based on debuginfo (*). Currently, basic types (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal integers (x/x8/x16/x32/x64), signedness casting (u/s), "string" and bitfield are supported. (see TYPES for detail)
 On x86 systems %REG is always the short form of the register: for example %AX. %RAX or %EAX is not valid.
 
 TYPES
 -----
-Basic types (u8/u16/u32/u64/s8/s16/s32/s64) and hexadecimal integers (x8/x16/x32/x64) are integer types. Prefix 's' and 'u' means those types are signed and unsigned respectively, and 'x' means that is shown in hexadecimal format. Traced arguments are shown in decimal (signed) or hex (unsigned). You can also use 's' or 'u' to specify only signedness and leave its size auto-detected by perf probe. Moreover, you can use 'x' to explicitly specify to be shown in hexadecimal (the size is also auto-detected).
+Basic types (u8/u16/u32/u64/s8/s16/s32/s64) and hexadecimal integers (x8/x16/x32/x64) are integer types. Prefix 's' and 'u' means those types are signed and unsigned respectively, and 'x' means that is shown in hexadecimal format. Traced arguments are shown in decimal (sNN/uNN) or hex (xNN). You can also use 's' or 'u' to specify only signedness and leave its size auto-detected by perf probe. Moreover, you can use 'x' to explicitly specify to be shown in hexadecimal (the size is also auto-detected).
 String type is a special type, which fetches a "null-terminated" string from kernel space. This means it will fail and store NULL if the string container has been paged out. You can specify 'string' type only for the local variable or structure member which is an array of or a pointer to 'char' or 'unsigned char' type.
 Bitfield is another special type, which takes 3 parameters, bit-width, bit-offset, and container-size (usually 32). The syntax is;
 
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468930 — [PATCH 15/35] perf tools tui: Display multiple header lines

From: Jiri Olsa <jolsa@kernel.org>

Display multiple header lines in TUI browser, if it's configured within
struct perf_hpp_list::nr_header_lines.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1470583710-1649-4-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/ui/browsers/hists.c | 25 +++++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c
index 303ed62a8f53..4a7aac8ed5ee 100644
--- a/tools/perf/ui/browsers/hists.c
+++ b/tools/perf/ui/browsers/hists.c
@@ -1502,7 +1502,9 @@ static int advance_hpp_check(struct perf_hpp *hpp, int inc)
 	return hpp->size <= 0;
 }
 
-static int hists_browser__scnprintf_headers(struct hist_browser *browser, char *buf, size_t size)
+static int
+hists_browser__scnprintf_headers(struct hist_browser *browser, char *buf,
+				 size_t size, int line)
 {
 	struct hists *hists = browser->hists;
 	struct perf_hpp dummy_hpp = {
@@ -1523,7 +1525,7 @@ static int hists_browser__scnprintf_headers(struct hist_browser *browser, char *
 		if (perf_hpp__should_skip(fmt, hists)  || column++ < browser->b.horiz_scroll)
 			continue;
 
-		ret = fmt->header(fmt, &dummy_hpp, hists, 0);
+		ret = fmt->header(fmt, &dummy_hpp, hists, line);
 		if (advance_hpp_check(&dummy_hpp, ret))
 			break;
 
@@ -1628,14 +1630,21 @@ static void hists_browser__hierarchy_headers(struct hist_browser *browser)
 
 static void hists_browser__headers(struct hist_browser *browser)
 {
-	char headers[1024];
+	struct hists *hists = browser->hists;
+	struct perf_hpp_list *hpp_list = hists->hpp_list;
 
-	hists_browser__scnprintf_headers(browser, headers,
-					 sizeof(headers));
+	int line;
 
-	ui_browser__gotorc(&browser->b, 0, 0);
-	ui_browser__set_color(&browser->b, HE_COLORSET_ROOT);
-	ui_browser__write_nstring(&browser->b, headers, browser->b.width + 1);
+	for (line = 0; line < hpp_list->nr_header_lines; line++) {
+		char headers[1024];
+
+		hists_browser__scnprintf_headers(browser, headers,
+						 sizeof(headers), line);
+
+		ui_browser__gotorc(&browser->b, line, 0);
+		ui_browser__set_color(&browser->b, HE_COLORSET_ROOT);
+		ui_browser__write_nstring(&browser->b, headers, browser->b.width + 1);
+	}
 }
 
 static void hist_browser__show_headers(struct hist_browser *browser)
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468931 — [PATCH 05/35] perf bench sched-pipe: Use linux/time64.h, USEC_PER_SEC

From: Arnaldo Carvalho de Melo <acme@redhat.com>

Following kernel practices.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-wgfu1h1pnw8lc919o2tan58y@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/bench/sched-pipe.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/tools/perf/bench/sched-pipe.c b/tools/perf/bench/sched-pipe.c
index 1dc2d13cc272..2243f0150d76 100644
--- a/tools/perf/bench/sched-pipe.c
+++ b/tools/perf/bench/sched-pipe.c
@@ -25,6 +25,7 @@
 #include <sys/time.h>
 #include <sys/types.h>
 #include <sys/syscall.h>
+#include <linux/time64.h>
 
 #include <pthread.h>
 
@@ -153,24 +154,24 @@ int bench_sched_pipe(int argc, const char **argv, const char *prefix __maybe_unu
 		printf("# Executed %d pipe operations between two %s\n\n",
 			loops, threaded ? "threads" : "processes");
 
-		result_usec = diff.tv_sec * 1000000;
+		result_usec = diff.tv_sec * USEC_PER_SEC;
 		result_usec += diff.tv_usec;
 
 		printf(" %14s: %lu.%03lu [sec]\n\n", "Total time",
 		       diff.tv_sec,
-		       (unsigned long) (diff.tv_usec/1000));
+		       (unsigned long) (diff.tv_usec / USEC_PER_MSEC));
 
 		printf(" %14lf usecs/op\n",
 		       (double)result_usec / (double)loops);
 		printf(" %14d ops/sec\n",
 		       (int)((double)loops /
-			     ((double)result_usec / (double)1000000)));
+			     ((double)result_usec / (double)USEC_PER_SEC)));
 		break;
 
 	case BENCH_FORMAT_SIMPLE:
 		printf("%lu.%03lu\n",
 		       diff.tv_sec,
-		       (unsigned long) (diff.tv_usec / 1000));
+		       (unsigned long) (diff.tv_usec / USEC_PER_MSEC));
 		break;
 
 	default:
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468950 — [PATCH 35/35] perf record: Fix spelling mistake "Finshed" -> "Finished"

From: Colin Ian King <colin.king@canonical.com>

Trivial fix to spelling mistake in pr_debug message.

Signed-off-by: Colin King <colin.king@canonical.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20160822183008.26368-1-colin.king@canonical.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-record.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index a3792e8c3e03..03251c7f14ec 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -96,7 +96,7 @@ backward_rb_find_range(void *buf, int mask, u64 head, u64 *start, u64 *end)
 	*start = head;
 	while (true) {
 		if (evt_head - head >= (unsigned int)size) {
-			pr_debug("Finshed reading backward ring buffer: rewind\n");
+			pr_debug("Finished reading backward ring buffer: rewind\n");
 			if (evt_head - head > (unsigned int)size)
 				evt_head -= pheader->size;
 			*end = evt_head;
@@ -106,7 +106,7 @@ backward_rb_find_range(void *buf, int mask, u64 head, u64 *start, u64 *end)
 		pheader = (struct perf_event_header *)(buf + (evt_head & mask));
 
 		if (pheader->size == 0) {
-			pr_debug("Finshed reading backward ring buffer: get start\n");
+			pr_debug("Finished reading backward ring buffer: get start\n");
 			*end = evt_head;
 			return 0;
 		}
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468955 — [PATCH 03/35] perf sched: Use linux/time64.h

From: Arnaldo Carvalho de Melo <acme@redhat.com>

Probably the next step is to introduce linux/time.h and use
timespec_to_ns(), etc.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-4nqhskn27fn93cz3ukbc8drf@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-sched.c | 37 +++++++++++++++++++------------------
 1 file changed, 19 insertions(+), 18 deletions(-)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 0dfe8df2ab9b..f5503ca22e1c 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -26,6 +26,7 @@
 #include <pthread.h>
 #include <math.h>
 #include <api/fs/fs.h>
+#include <linux/time64.h>
 
 #define PR_SET_NAME		15               /* Set process name */
 #define MAX_CPUS		4096
@@ -199,7 +200,7 @@ static u64 get_nsecs(void)
 
 	clock_gettime(CLOCK_MONOTONIC, &ts);
 
-	return ts.tv_sec * 1000000000ULL + ts.tv_nsec;
+	return ts.tv_sec * NSEC_PER_SEC + ts.tv_nsec;
 }
 
 static void burn_nsecs(struct perf_sched *sched, u64 nsecs)
@@ -223,7 +224,7 @@ static void sleep_nsecs(u64 nsecs)
 
 static void calibrate_run_measurement_overhead(struct perf_sched *sched)
 {
-	u64 T0, T1, delta, min_delta = 1000000000ULL;
+	u64 T0, T1, delta, min_delta = NSEC_PER_SEC;
 	int i;
 
 	for (i = 0; i < 10; i++) {
@@ -240,7 +241,7 @@ static void calibrate_run_measurement_overhead(struct perf_sched *sched)
 
 static void calibrate_sleep_measurement_overhead(struct perf_sched *sched)
 {
-	u64 T0, T1, delta, min_delta = 1000000000ULL;
+	u64 T0, T1, delta, min_delta = NSEC_PER_SEC;
 	int i;
 
 	for (i = 0; i < 10; i++) {
@@ -452,8 +453,8 @@ static u64 get_cpu_usage_nsec_parent(void)
 	err = getrusage(RUSAGE_SELF, &ru);
 	BUG_ON(err);
 
-	sum =  ru.ru_utime.tv_sec*1e9 + ru.ru_utime.tv_usec*1e3;
-	sum += ru.ru_stime.tv_sec*1e9 + ru.ru_stime.tv_usec*1e3;
+	sum =  ru.ru_utime.tv_sec * NSEC_PER_SEC + ru.ru_utime.tv_usec * NSEC_PER_USEC;
+	sum += ru.ru_stime.tv_sec * NSEC_PER_SEC + ru.ru_stime.tv_usec * NSEC_PER_USEC;
 
 	return sum;
 }
@@ -667,12 +668,12 @@ static void run_one_test(struct perf_sched *sched)
 		sched->run_avg = delta;
 	sched->run_avg = (sched->run_avg * (sched->replay_repeat - 1) + delta) / sched->replay_repeat;
 
-	printf("#%-3ld: %0.3f, ", sched->nr_runs, (double)delta / 1000000.0);
+	printf("#%-3ld: %0.3f, ", sched->nr_runs, (double)delta / NSEC_PER_MSEC);
 
-	printf("ravg: %0.2f, ", (double)sched->run_avg / 1e6);
+	printf("ravg: %0.2f, ", (double)sched->run_avg / NSEC_PER_MSEC);
 
 	printf("cpu: %0.2f / %0.2f",
-		(double)sched->cpu_usage / 1e6, (double)sched->runavg_cpu_usage / 1e6);
+		(double)sched->cpu_usage / NSEC_PER_MSEC, (double)sched->runavg_cpu_usage / NSEC_PER_MSEC);
 
 #if 0
 	/*
@@ -680,8 +681,8 @@ static void run_one_test(struct perf_sched *sched)
 	 * accurate than the sched->sum_exec_runtime based statistics:
 	 */
 	printf(" [%0.2f / %0.2f]",
-		(double)sched->parent_cpu_usage/1e6,
-		(double)sched->runavg_parent_cpu_usage/1e6);
+		(double)sched->parent_cpu_usage / NSEC_PER_MSEC,
+		(double)sched->runavg_parent_cpu_usage / NSEC_PER_MSEC);
 #endif
 
 	printf("\n");
@@ -696,13 +697,13 @@ static void test_calibrations(struct perf_sched *sched)
 	u64 T0, T1;
 
 	T0 = get_nsecs();
-	burn_nsecs(sched, 1e6);
+	burn_nsecs(sched, NSEC_PER_MSEC);
 	T1 = get_nsecs();
 
 	printf("the run test took %" PRIu64 " nsecs\n", T1 - T0);
 
 	T0 = get_nsecs();
-	sleep_nsecs(1e6);
+	sleep_nsecs(NSEC_PER_MSEC);
 	T1 = get_nsecs();
 
 	printf("the sleep test took %" PRIu64 " nsecs\n", T1 - T0);
@@ -1213,10 +1214,10 @@ static void output_lat_thread(struct perf_sched *sched, struct work_atoms *work_
 	avg = work_list->total_lat / work_list->nb_atoms;
 
 	printf("|%11.3f ms |%9" PRIu64 " | avg:%9.3f ms | max:%9.3f ms | max at: %13.6f s\n",
-	      (double)work_list->total_runtime / 1e6,
-		 work_list->nb_atoms, (double)avg / 1e6,
-		 (double)work_list->max_lat / 1e6,
-		 (double)work_list->max_lat_at / 1e9);
+	      (double)work_list->total_runtime / NSEC_PER_MSEC,
+		 work_list->nb_atoms, (double)avg / NSEC_PER_MSEC,
+		 (double)work_list->max_lat / NSEC_PER_MSEC,
+		 (double)work_list->max_lat_at / NSEC_PER_SEC);
 }
 
 static int pid_cmp(struct work_atoms *l, struct work_atoms *r)
@@ -1491,7 +1492,7 @@ static int map_switch_event(struct perf_sched *sched, struct perf_evsel *evsel,
 	if (sched->map.cpus && !cpu_map__has(sched->map.cpus, this_cpu))
 		goto out;
 
-	color_fprintf(stdout, color, "  %12.6f secs ", (double)timestamp/1e9);
+	color_fprintf(stdout, color, "  %12.6f secs ", (double)timestamp / NSEC_PER_SEC);
 	if (new_shortname) {
 		const char *pid_color = color;
 
@@ -1753,7 +1754,7 @@ static int perf_sched__lat(struct perf_sched *sched)
 
 	printf(" -----------------------------------------------------------------------------------------------------------------\n");
 	printf("  TOTAL:                |%11.3f ms |%9" PRIu64 " |\n",
-		(double)sched->all_runtime / 1e6, sched->all_count);
+		(double)sched->all_runtime / NSEC_PER_MSEC, sched->all_count);
 
 	printf(" ---------------------------------------------------\n");
 
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468956 — [PATCH 06/35] perf stat: Use *SEC_PER_*SEC macros

From: Arnaldo Carvalho de Melo <acme@redhat.com>

To match how this is done in the kernel.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-gym6yshewpdegt153u8v2q5r@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index e33a66bffbdb..90882b1d6a91 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -173,7 +173,7 @@ static inline void diff_timespec(struct timespec *r, struct timespec *a,
 {
 	r->tv_sec = a->tv_sec - b->tv_sec;
 	if (a->tv_nsec < b->tv_nsec) {
-		r->tv_nsec = a->tv_nsec + 1000000000L - b->tv_nsec;
+		r->tv_nsec = a->tv_nsec + NSEC_PER_SEC - b->tv_nsec;
 		r->tv_sec--;
 	} else {
 		r->tv_nsec = a->tv_nsec - b->tv_nsec ;
@@ -365,7 +365,7 @@ static void process_interval(void)
 static void enable_counters(void)
 {
 	if (initial_delay)
-		usleep(initial_delay * 1000);
+		usleep(initial_delay * USEC_PER_MSEC);
 
 	/*
 	 * We need to enable counters only if:
@@ -542,8 +542,8 @@ static int __run_perf_stat(int argc, const char **argv)
 	bool is_pipe = STAT_RECORD ? perf_stat.file.is_pipe : false;
 
 	if (interval) {
-		ts.tv_sec  = interval / 1000;
-		ts.tv_nsec = (interval % 1000) * 1000000;
+		ts.tv_sec  = interval / USEC_PER_MSEC;
+		ts.tv_nsec = (interval % USEC_PER_MSEC) * NSEC_PER_MSEC;
 	} else {
 		ts.tv_sec  = 1;
 		ts.tv_nsec = 0;
@@ -972,7 +972,7 @@ static void print_metric_header(void *ctx, const char *color __maybe_unused,
 static void nsec_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 {
 	FILE *output = stat_config.output;
-	double msecs = avg / 1e6;
+	double msecs = avg / NSEC_PER_MSEC;
 	const char *fmt_v, *fmt_n;
 	char name[25];
 
@@ -1461,7 +1461,7 @@ static void print_footer(void)
 	if (!null_run)
 		fprintf(output, "\n");
 	fprintf(output, " %17.9f seconds time elapsed",
-			avg_stats(&walltime_nsecs_stats)/1e9);
+			avg_stats(&walltime_nsecs_stats) / NSEC_PER_SEC);
 	if (run_count > 1) {
 		fprintf(output, "                                        ");
 		print_noise_pct(stddev_stats(&walltime_nsecs_stats),
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468957 — [PATCH 16/35] perf tools stdio: Display multiple header lines

From: Jiri Olsa <jolsa@kernel.org>

Display multiple header lines in stdio output , if it's configured
within struct perf_hpp_list::nr_header_lines.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1470583710-1649-5-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/ui/stdio/hist.c | 34 +++++++++++++++++++++++++---------
 1 file changed, 25 insertions(+), 9 deletions(-)

diff --git a/tools/perf/ui/stdio/hist.c b/tools/perf/ui/stdio/hist.c
index 91b8e10cabd5..9b807cbb56b1 100644
--- a/tools/perf/ui/stdio/hist.c
+++ b/tools/perf/ui/stdio/hist.c
@@ -639,13 +639,10 @@ hists__fprintf_hierarchy_headers(struct hists *hists,
 	return print_hierarchy_header(hists, hpp, symbol_conf.field_sep, fp);
 }
 
-static int
-hists__fprintf_standard_headers(struct hists *hists,
-				struct perf_hpp *hpp,
-				FILE *fp)
+static void fprintf_line(struct hists *hists, struct perf_hpp *hpp,
+			 int line, FILE *fp)
 {
 	struct perf_hpp_fmt *fmt;
-	unsigned int width;
 	const char *sep = symbol_conf.field_sep;
 	bool first = true;
 
@@ -658,14 +655,33 @@ hists__fprintf_standard_headers(struct hists *hists,
 		else
 			first = false;
 
-		fmt->header(fmt, hpp, hists, 0);
+		fmt->header(fmt, hpp, hists, line);
 		fprintf(fp, "%s", hpp->buf);
 	}
+}
 
-	fprintf(fp, "\n");
+static int
+hists__fprintf_standard_headers(struct hists *hists,
+				struct perf_hpp *hpp,
+				FILE *fp)
+{
+	struct perf_hpp_list *hpp_list = hists->hpp_list;
+	struct perf_hpp_fmt *fmt;
+	unsigned int width;
+	const char *sep = symbol_conf.field_sep;
+	bool first = true;
+	int line;
+
+	for (line = 0; line < hpp_list->nr_header_lines; line++) {
+		/* first # is displayed one level up */
+		if (line)
+			fprintf(fp, "# ");
+		fprintf_line(hists, hpp, line, fp);
+		fprintf(fp, "\n");
+	}
 
 	if (sep)
-		return 1;
+		return hpp_list->nr_header_lines;
 
 	first = true;
 
@@ -689,7 +705,7 @@ hists__fprintf_standard_headers(struct hists *hists,
 
 	fprintf(fp, "\n");
 	fprintf(fp, "#\n");
-	return 3;
+	return hpp_list->nr_header_lines + 2;
 }
 
 static int hists__fprintf_headers(struct hists *hists, FILE *fp)
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468958 — [PATCH 25/35] perf hists browser: Remove superfluous null check on map

From: Colin Ian King <colin.king@canonical.com>

'map' is being already checked if it is NULL at the start of
do_zoom_dso(), so the second subsequent check is superfluous and can be
removed.

Signed-off-by: Colin King <colin.king@canonical.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1471278343-14999-1-git-send-email-colin.king@canonical.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/ui/browsers/hists.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c
index 22e48a3ebe29..f0611c937d4b 100644
--- a/tools/perf/ui/browsers/hists.c
+++ b/tools/perf/ui/browsers/hists.c
@@ -2440,8 +2440,6 @@ do_zoom_dso(struct hist_browser *browser, struct popup_action *act)
 		browser->hists->dso_filter = NULL;
 		ui_helpline__pop();
 	} else {
-		if (map == NULL)
-			return 0;
 		ui_helpline__fpush("To zoom out press ESC or ENTER + \"Zoom out of %s DSO\"",
 				   __map__is_kernel(map) ? "the Kernel" : map->dso->short_name);
 		browser->hists->dso_filter = map->dso;
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468959 — [PATCH 07/35] perf bench mem: Use USEC_PER_SEC

From: Arnaldo Carvalho de Melo <acme@redhat.com>

Following kernel practices, using linux/time64.h

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-xdtmguafva17wp023sxojiib@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/bench/mem-functions.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/bench/mem-functions.c b/tools/perf/bench/mem-functions.c
index 2b54d0f2672a..c684910e5a48 100644
--- a/tools/perf/bench/mem-functions.c
+++ b/tools/perf/bench/mem-functions.c
@@ -21,6 +21,7 @@
 #include <string.h>
 #include <sys/time.h>
 #include <errno.h>
+#include <linux/time64.h>
 
 #define K 1024
 
@@ -89,7 +90,7 @@ static u64 get_cycles(void)
 
 static double timeval2double(struct timeval *ts)
 {
-	return (double)ts->tv_sec + (double)ts->tv_usec / (double)1000000;
+	return (double)ts->tv_sec + (double)ts->tv_usec / (double)USEC_PER_SEC;
 }
 
 #define print_bps(x) do {						\
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468960 — [PATCH 20/35] perf disassemble: Extract logic to find file to pass to objdump to a separate function

From: Arnaldo Carvalho de Melo <acme@redhat.com>

Disentangling this a bit further, more to come.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-7bjv2xazuyzs0xw01mlwosn5@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/annotate.c | 54 +++++++++++++++++++++++++++-------------------
 1 file changed, 32 insertions(+), 22 deletions(-)

diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index 8a9949399112..25a9259a6a6e 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -1162,29 +1162,19 @@ int symbol__strerror_disassemble(struct symbol *sym __maybe_unused, struct map *
 	return 0;
 }
 
-int symbol__disassemble(struct symbol *sym, struct map *map, size_t privsize)
+static int dso__disassemble_filename(struct dso *dso, char *filename, size_t filename_size)
 {
-	struct dso *dso = map->dso;
-	char *filename;
-	char command[PATH_MAX * 2];
-	FILE *file;
-	char symfs_filename[PATH_MAX];
-	struct kcore_extract kce;
-	bool delete_extract = false;
-	int stdout_fd[2];
-	int lineno = 0;
-	int nline;
-	pid_t pid;
-	int err = SYMBOL_ANNOTATE_ERRNO__NO_VMLINUX;
+	char linkname[PATH_MAX];
+	char *build_id_filename;
 
 	if (dso->symtab_type == DSO_BINARY_TYPE__KALLSYMS &&
 	    !dso__is_kcore(dso))
-		goto out;
+		return SYMBOL_ANNOTATE_ERRNO__NO_VMLINUX;
 
-	filename = dso__build_id_filename(dso, NULL, 0);
-	if (filename) {
-		symbol__join_symfs(symfs_filename, filename);
-		free(filename);
+	build_id_filename = dso__build_id_filename(dso, NULL, 0);
+	if (build_id_filename) {
+		__symbol__join_symfs(filename, filename_size, build_id_filename);
+		free(build_id_filename);
 	} else {
 		if (dso->has_build_id)
 			return ENOMEM;
@@ -1192,18 +1182,38 @@ int symbol__disassemble(struct symbol *sym, struct map *map, size_t privsize)
 	}
 
 	if (dso__is_kcore(dso) ||
-	    readlink(symfs_filename, command, sizeof(command)) < 0 ||
-	    strstr(command, DSO__NAME_KALLSYMS) ||
-	    access(symfs_filename, R_OK)) {
+	    readlink(filename, linkname, sizeof(linkname)) < 0 ||
+	    strstr(linkname, DSO__NAME_KALLSYMS) ||
+	    access(filename, R_OK)) {
 fallback:
 		/*
 		 * If we don't have build-ids or the build-id file isn't in the
 		 * cache, or is just a kallsyms file, well, lets hope that this
 		 * DSO is the same as when 'perf record' ran.
 		 */
-		symbol__join_symfs(symfs_filename, dso->long_name);
+		__symbol__join_symfs(filename, filename_size, dso->long_name);
 	}
 
+	return 0;
+}
+
+int symbol__disassemble(struct symbol *sym, struct map *map, size_t privsize)
+{
+	struct dso *dso = map->dso;
+	char command[PATH_MAX * 2];
+	FILE *file;
+	char symfs_filename[PATH_MAX];
+	struct kcore_extract kce;
+	bool delete_extract = false;
+	int stdout_fd[2];
+	int lineno = 0;
+	int nline;
+	pid_t pid;
+	int err = dso__disassemble_filename(dso, symfs_filename, sizeof(symfs_filename));
+
+	if (err)
+		return err;
+
 	pr_debug("%s: filename=%s, sym=%s, start=%#" PRIx64 ", end=%#" PRIx64 "\n", __func__,
 		 symfs_filename, sym->name, map->unmap_ip(map, sym->start),
 		 map->unmap_ip(map, sym->end));
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468963 — [PATCH 19/35] perf disassemble: Simplify logic for picking the filename to disassemble

From: Arnaldo Carvalho de Melo <acme@redhat.com>

Lots of changes to support kcore, compressed modules, build-id files
left us with some spaguetti code, simplify it a bit, more to come.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-h70p7x451li3f2fhs44vzmm8@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/annotate.c | 41 ++++++++++++++++-------------------------
 1 file changed, 16 insertions(+), 25 deletions(-)

diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index 9882bc6ac566..8a9949399112 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -1166,7 +1166,6 @@ int symbol__disassemble(struct symbol *sym, struct map *map, size_t privsize)
 {
 	struct dso *dso = map->dso;
 	char *filename;
-	bool free_filename = true;
 	char command[PATH_MAX * 2];
 	FILE *file;
 	char symfs_filename[PATH_MAX];
@@ -1183,31 +1182,30 @@ int symbol__disassemble(struct symbol *sym, struct map *map, size_t privsize)
 		goto out;
 
 	filename = dso__build_id_filename(dso, NULL, 0);
-	if (filename)
+	if (filename) {
 		symbol__join_symfs(symfs_filename, filename);
-
-	if (filename == NULL) {
+		free(filename);
+	} else {
 		if (dso->has_build_id)
 			return ENOMEM;
 		goto fallback;
-	} else if (dso__is_kcore(dso) ||
-		   readlink(symfs_filename, command, sizeof(command)) < 0 ||
-		   strstr(command, DSO__NAME_KALLSYMS) ||
-		   access(symfs_filename, R_OK)) {
-		free(filename);
+	}
+
+	if (dso__is_kcore(dso) ||
+	    readlink(symfs_filename, command, sizeof(command)) < 0 ||
+	    strstr(command, DSO__NAME_KALLSYMS) ||
+	    access(symfs_filename, R_OK)) {
 fallback:
 		/*
 		 * If we don't have build-ids or the build-id file isn't in the
 		 * cache, or is just a kallsyms file, well, lets hope that this
 		 * DSO is the same as when 'perf record' ran.
 		 */
-		filename = (char *)dso->long_name;
-		symbol__join_symfs(symfs_filename, filename);
-		free_filename = false;
+		symbol__join_symfs(symfs_filename, dso->long_name);
 	}
 
 	pr_debug("%s: filename=%s, sym=%s, start=%#" PRIx64 ", end=%#" PRIx64 "\n", __func__,
-		 filename, sym->name, map->unmap_ip(map, sym->start),
+		 symfs_filename, sym->name, map->unmap_ip(map, sym->start),
 		 map->unmap_ip(map, sym->end));
 
 	pr_debug("annotating [%p] %30s : [%p] %30s\n",
@@ -1222,11 +1220,6 @@ fallback:
 			delete_extract = true;
 			strlcpy(symfs_filename, kce.extract_filename,
 				sizeof(symfs_filename));
-			if (free_filename) {
-				free(filename);
-				free_filename = false;
-			}
-			filename = symfs_filename;
 		}
 	} else if (dso__needs_decompress(dso)) {
 		char tmp[PATH_MAX];
@@ -1235,14 +1228,14 @@ fallback:
 		bool ret;
 
 		if (kmod_path__parse_ext(&m, symfs_filename))
-			goto out_free_filename;
+			goto out;
 
 		snprintf(tmp, PATH_MAX, "/tmp/perf-kmod-XXXXXX");
 
 		fd = mkstemp(tmp);
 		if (fd < 0) {
 			free(m.ext);
-			goto out_free_filename;
+			goto out;
 		}
 
 		ret = decompress_to_file(m.ext, symfs_filename, fd);
@@ -1254,7 +1247,7 @@ fallback:
 		close(fd);
 
 		if (!ret)
-			goto out_free_filename;
+			goto out;
 
 		strcpy(symfs_filename, tmp);
 	}
@@ -1270,7 +1263,7 @@ fallback:
 		 map__rip_2objdump(map, sym->end),
 		 symbol_conf.annotate_asm_raw ? "" : "--no-show-raw",
 		 symbol_conf.annotate_src ? "-S" : "",
-		 symfs_filename, filename);
+		 symfs_filename, symfs_filename);
 
 	pr_debug("Executing: %s\n", command);
 
@@ -1332,11 +1325,9 @@ out_remove_tmp:
 
 	if (dso__needs_decompress(dso))
 		unlink(symfs_filename);
-out_free_filename:
+
 	if (delete_extract)
 		kcore_extract__delete(&kce);
-	if (free_filename)
-		free(filename);
 out:
 	return err;
 
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468965 — [PATCH 24/35] perf tools: Skip running the feature tests for 'make install-doc'

From: Rui Teng <rui.teng@linux.vnet.ibm.com>

It is a requirement from the perf TODO list[1]:

''The feature tests should be performed only when a file that needs those
  tests, or at least only when some .c or .h file will be rebuilt. An
  initial step would be for 'make install-doc' not to run the feature
  tests, there it is not needed at all.''

By adding 'install-doc' to the NON_CONFIG_TARGETS, it will skip running
the feature tests for such target. The Auto-detecting system features
list will not be displayed:

  $ make install-doc
    BUILD:   Doing 'make -j2' parallel build
    SUBDIR   Documentation
  make[2]: Nothing to be done for 'install'.

[1] https://perf.wiki.kernel.org/index.php/Todo

Signed-off-by: Rui Teng <rui.teng@linux.vnet.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1470818948-17784-1-git-send-email-rui.teng@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Makefile.perf | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf
index aa7ab23fbc9f..828cfd766e31 100644
--- a/tools/perf/Makefile.perf
+++ b/tools/perf/Makefile.perf
@@ -165,7 +165,7 @@ SUBCMD_DIR	= $(srctree)/tools/lib/subcmd/
 # non-config cases
 config := 1
 
-NON_CONFIG_TARGETS := clean TAGS tags cscope help
+NON_CONFIG_TARGETS := clean TAGS tags cscope help install-doc
 
 ifdef MAKECMDGOALS
 ifeq ($(filter-out $(NON_CONFIG_TARGETS),$(MAKECMDGOALS)),)
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468966 — [PATCH 13/35] perf hists: Introduce nr_header_lines into struct perf_hpp_list

From: Jiri Olsa <jolsa@kernel.org>

Currently we support just single line headers, this is first step to
allow more.

Store the number of header lines in perf_hpp_list, which encompasses all
the display/sort entries and is thus suitable to hold this value.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1470583710-1649-2-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/ui/browsers/hists.c | 15 ++++++++++++---
 tools/perf/ui/hist.c           |  1 +
 tools/perf/util/hist.h         |  1 +
 3 files changed, 14 insertions(+), 3 deletions(-)

diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c
index 13d414384739..95f7cf1991fc 100644
--- a/tools/perf/ui/browsers/hists.c
+++ b/tools/perf/ui/browsers/hists.c
@@ -69,8 +69,11 @@ static u32 hist_browser__nr_entries(struct hist_browser *hb)
 static void hist_browser__update_rows(struct hist_browser *hb)
 {
 	struct ui_browser *browser = &hb->b;
-	u16 header_offset = hb->show_headers ? 1 : 0, index_row;
+	struct hists *hists = hb->hists;
+	struct perf_hpp_list *hpp_list = hists->hpp_list;
+	u16 header_offset, index_row;
 
+	header_offset = hb->show_headers ? hpp_list->nr_header_lines : 0;
 	browser->rows = browser->height - header_offset;
 	/*
 	 * Verify if we were at the last line and that line isn't
@@ -99,8 +102,11 @@ static void hist_browser__refresh_dimensions(struct ui_browser *browser)
 
 static void hist_browser__gotorc(struct hist_browser *browser, int row, int column)
 {
-	u16 header_offset = browser->show_headers ? 1 : 0;
+	struct hists *hists = browser->hists;
+	struct perf_hpp_list *hpp_list = hists->hpp_list;
+	u16 header_offset;
 
+	header_offset = browser->show_headers ? hpp_list->nr_header_lines : 0;
 	ui_browser__gotorc(&browser->b, row + header_offset, column);
 }
 
@@ -1656,10 +1662,13 @@ static unsigned int hist_browser__refresh(struct ui_browser *browser)
 	u16 header_offset = 0;
 	struct rb_node *nd;
 	struct hist_browser *hb = container_of(browser, struct hist_browser, b);
+	struct hists *hists = hb->hists;
 
 	if (hb->show_headers) {
+		struct perf_hpp_list *hpp_list = hists->hpp_list;
+
 		hist_browser__show_headers(hb);
-		header_offset = 1;
+		header_offset = hpp_list->nr_header_lines;
 	}
 
 	ui_browser__hists_init_top(browser);
diff --git a/tools/perf/ui/hist.c b/tools/perf/ui/hist.c
index 4274969ddc89..be7a17fec889 100644
--- a/tools/perf/ui/hist.c
+++ b/tools/perf/ui/hist.c
@@ -441,6 +441,7 @@ struct perf_hpp_fmt perf_hpp__format[] = {
 struct perf_hpp_list perf_hpp_list = {
 	.fields	= LIST_HEAD_INIT(perf_hpp_list.fields),
 	.sorts	= LIST_HEAD_INIT(perf_hpp_list.sorts),
+	.nr_header_lines = 1,
 };
 
 #undef HPP__COLOR_PRINT_FNS
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
index 0a1edf1ab450..8cc5d33d2405 100644
--- a/tools/perf/util/hist.h
+++ b/tools/perf/util/hist.h
@@ -259,6 +259,7 @@ struct perf_hpp_list {
 	struct list_head fields;
 	struct list_head sorts;
 
+	int nr_header_lines;
 	int need_collapse;
 	int parent;
 	int sym;
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1468967 — [PATCH 30/35] perf probe: Use hexadecimal type by default if possible

From: Masami Hiramatsu <mhiramat@kernel.org>

Use hexadecimal type by default if it is available on current running
kernel.

This keeps the default behavior of perf probe after changing the output
format of 'u8/16/32/64' to unsigned decimal number.

Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Hemant Kumar <hemant@linux.vnet.ibm.com>
Cc: Naohiro Aota <naohiro.aota@hgst.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/147151074685.12957.16415861010796255514.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/probe-finder.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
index f18cd6bbada9..ac4740f8ee3a 100644
--- a/tools/perf/util/probe-finder.c
+++ b/tools/perf/util/probe-finder.c
@@ -373,7 +373,8 @@ static int convert_variable_type(Dwarf_Die *vr_die,
 		 probe_type_is_available(PROBE_TYPE_X))
 		prefix = 'x';
 	else
-		prefix = die_is_signed_type(&type) ? 's' : 'u';
+		prefix = die_is_signed_type(&type) ? 's' :
+			 probe_type_is_available(PROBE_TYPE_X) ? 'x' : 'u';
 
 	ret = dwarf_bytesize(&type);
 	if (ret <= 0)
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1469237

* Arnaldo Carvalho de Melo <acme@kernel.org> wrote:

> Hi Ingo,
> 
> 	Please consider pulling, I first merged tip/perf/urgent into a
> tip/perf/core and rebased the patches I had in acme/perf/core.
> 
> - Arnaldo
> 
> Build stats at the end of this message.
> 
> The following changes since commit ce90c12d2453aa6be743719bb0a5d4040b92700f:
> 
>   Merge branch 'perf/urgent' into perf/core, to pick up fixes before merging new changes (2016-08-23 15:35:47 -0300)
> 
> are available in the git repository at:
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git tags/perf-core-for-mingo-20160823
> 
> for you to fetch changes up to 5e30d55c71de058e4156080fe32d426c22d094cb:
> 
>   perf record: Fix spelling mistake "Finshed" -> "Finished" (2016-08-23 17:06:40 -0300)
> 
> ----------------------------------------------------------------
> perf/core improvements and fixes:
> 
> User visible:
> 
> . Allow configuring the default 'perf report -s' sort order in ~/.perfconfig,
>   for instance, "sym,dso" may be more fitting for kernel developers. (Arnaldo Carvalho de Melo)
> 
> - Support x8/x16/x32/x64 hexadecimal "types" in ftrace and 'perf probe' (Masami Hiramatsu)
> 
> Infrastructure:
> 
> - Skip running the feature tests for 'make install-doc' (Rui Teng)
> 
> - Introduce tools/include/linux/time64.h with *SEC_PER_*SEC macros
>   to use in all of tools/ (Arnaldo Carvalho de Melo)
> 
> - Break down symbol__disassemble() into multiple functions, to ease
>   future work on better reporting the errors that may take place in
>   the various steps it performs (possibly decompressing kernel module
>   files, getting build-id keyed files, calling objdump, parsing its
>   output, etc) (Arnaldo Carvalho de Melo)
> 
> - Typo fixes in various places (Colin Ian King)
> 
> - Remove superfluous NULL check in the TUI code (Colin Ian King)
> 
> - Allow displaying multiple header lines in the TUI browser, prep
>   work for the 'perf c2c' browser (Jiri Olsa)
> 
> - Copy coresight-pmu.h header file needed by perf tools (Mathieu Poirier)
> 
> - Use __weak definition from linux/compiler.h (Rui Teng)
> 
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> 
> ----------------------------------------------------------------
> 
> Arnaldo Carvalho de Melo (16):
>       tools: Introduce tools/include/linux/time64.h for *SEC_PER_*SEC macros
>       perf bench numa: Use NSEC_PER_U?SEC
>       perf sched: Use linux/time64.h
>       perf timechart: Use NSEC_PER_U?SEC
>       perf bench sched-pipe: Use linux/time64.h, USEC_PER_SEC
>       perf stat: Use *SEC_PER_*SEC macros
>       perf bench mem: Use USEC_PER_SEC
>       perf bench sched-messaging: Use USEC_PER_MSEC
>       perf record: Use USEC_PER_MSEC
>       perf kvm: Use NSEC_PER_USEC
>       perf bench futex: Use NSEC_PER_USEC
>       perf top: Use MSEC_PER_SEC
>       perf disassemble: Move check for kallsyms + !kcore
>       perf disassemble: Simplify logic for picking the filename to disassemble
>       perf disassemble: Extract logic to find file to pass to objdump to a separate function
>       perf report: Allow configuring the default sort order in ~/.perfconfig
> 
> Colin Ian King (5):
>       perf hists browser: Remove superfluous null check on map
>       perf tools: Fix typo: "ehough" -> "enough"
>       perf test bpf: Fix typo: "ehough" -> "enough"
>       perf bpf: Fix typo: "ehough" -> "enough"
>       perf record: Fix spelling mistake "Finshed" -> "Finished"
> 
> Jiri Olsa (5):
>       perf hists: Introduce nr_header_lines into struct perf_hpp_list
>       perf hists: Add line argument into perf_hpp_fmt's header callback
>       perf tools tui: Display multiple header lines
>       perf tools stdio: Display multiple header lines
>       perf hists: Add support for header span
> 
> Masami Hiramatsu (6):
>       ftrace: kprobe: uprobe: Add x8/x16/x32/x64 for hexadecimal types
>       ftrace: probe: Add README entries for k/uprobe-events
>       perf probe: Add supported for type casting by the running kernel
>       perf probe: Support hexadecimal casting
>       perf probe: Use hexadecimal type by default if possible
>       ftrace: kprobe: uprobe: Show u8/u16/u32/u64 types in decimal
> 
> Mathieu Poirier (1):
>       tools: Copy coresight-pmu.h header file needed by perf tools
> 
> Rui Teng (2):
>       perf tools: Use __weak definition from linux/compiler.h
>       perf tools: Skip running the feature tests for 'make install-doc'
> 
>  Documentation/trace/kprobetrace.txt                |  9 ++-
>  Documentation/trace/uprobetracer.txt               |  9 ++-
>  kernel/trace/trace.c                               | 24 ++++++
>  kernel/trace/trace_kprobe.c                        |  4 +
>  kernel/trace/trace_probe.c                         | 30 ++++----
>  kernel/trace/trace_probe.h                         | 11 ++-
>  kernel/trace/trace_uprobe.c                        |  4 +
>  tools/include/linux/coresight-pmu.h                | 39 ++++++++++
>  tools/include/linux/time64.h                       | 12 +++
>  tools/perf/Documentation/perf-config.txt           |  4 +
>  tools/perf/Documentation/perf-probe.txt            |  5 +-
>  tools/perf/MANIFEST                                |  2 +
>  tools/perf/Makefile.perf                           |  5 +-
>  tools/perf/bench/futex-requeue.c                   |  5 +-
>  tools/perf/bench/futex-wake-parallel.c             |  5 +-
>  tools/perf/bench/futex-wake.c                      |  5 +-
>  tools/perf/bench/mem-functions.c                   |  3 +-
>  tools/perf/bench/numa.c                            | 53 ++++++-------
>  tools/perf/bench/sched-messaging.c                 |  5 +-
>  tools/perf/bench/sched-pipe.c                      |  9 ++-
>  tools/perf/builtin-diff.c                          |  4 +-
>  tools/perf/builtin-kvm.c                           | 11 +--
>  tools/perf/builtin-record.c                        |  8 +-
>  tools/perf/builtin-report.c                        |  4 +
>  tools/perf/builtin-sched.c                         | 37 ++++-----
>  tools/perf/builtin-script.c                        |  7 +-
>  tools/perf/builtin-stat.c                          | 19 ++---
>  tools/perf/builtin-timechart.c                     | 13 ++--
>  tools/perf/builtin-top.c                           |  3 +-
>  tools/perf/builtin-trace.c                         |  1 +
>  tools/perf/perf.h                                  |  7 --
>  tools/perf/tests/backward-ring-buffer.c            |  2 +-
>  tools/perf/tests/bpf.c                             |  2 +-
>  tools/perf/ui/browsers/hists.c                     | 50 +++++++++----
>  tools/perf/ui/gtk/hists.c                          |  2 +-
>  tools/perf/ui/hist.c                               |  4 +-
>  tools/perf/ui/stdio/hist.c                         | 45 +++++++----
>  tools/perf/util/annotate.c                         | 87 +++++++++++-----------
>  tools/perf/util/bpf-loader.c                       |  2 +-
>  tools/perf/util/debug.c                            | 10 +--
>  tools/perf/util/header.c                           |  3 +-
>  tools/perf/util/hist.h                             |  3 +-
>  tools/perf/util/probe-file.c                       | 57 ++++++++++++++
>  tools/perf/util/probe-file.h                       | 10 +++
>  tools/perf/util/probe-finder.c                     | 19 +++--
>  .../perf/util/scripting-engines/trace-event-perl.c |  5 +-
>  .../util/scripting-engines/trace-event-python.c    |  5 +-
>  tools/perf/util/sort.c                             |  9 ++-
>  tools/perf/util/sort.h                             |  2 +-
>  tools/perf/util/svghelper.c                        | 11 +--
>  tools/perf/util/util.c                             |  1 +
>  tools/perf/util/util.h                             |  4 -
>  52 files changed, 464 insertions(+), 226 deletions(-)
>  create mode 100644 tools/include/linux/coresight-pmu.h
>  create mode 100644 tools/include/linux/time64.h
> 
> Build Stats:
> 
> News:
> 
> The fedora:24-x-ARC-uClibc adds the ARC arch and the uClibc libc to the mix of
> targets tested, still with a minimal build that doesn't include even libelf,
> that will be added soon, using a more full featured pre-built toolchain provided
> by the Synopsys folks.
> 
>   [root@jouet ~]# time dm
>    1 70.304638757 alpine:3.4: Ok
>    2 24.806303766 android-ndk:r12b-arm: Ok
>    3 71.829633643 archlinux:latest: Ok
>    4 39.316551941 centos:5: Ok
>    5 59.282978228 centos:6: Ok
>    6 69.836088394 centos:7: Ok
>    7 63.476952272 debian:7: Ok
>    8 69.450110099 debian:8: Ok
>    9 72.484714796 debian:experimental: Ok
>   10 69.730035221 fedora:20: Ok
>   11 73.629813614 fedora:21: Ok
>   12 71.955425760 fedora:22: Ok
>   13 73.015579053 fedora:23: Ok
>   14 81.186943795 fedora:24: Ok
>   15 31.401503154 fedora:24-x-ARC-uClibc: Ok
>   16 78.366606942 fedora:rawhide: Ok
>   17 77.434661064 mageia:5: Ok
>   18 74.491020416 opensuse:13.2: Ok
>   19 71.673570141 opensuse:42.1: Ok
>   20 79.587415167 opensuse:tumbleweed: Ok
>   21 57.802976895 ubuntu:12.04.5: Ok
>   22 68.505810699 ubuntu:14.04.4: Ok
>   23 69.478057262 ubuntu:15.10: Ok
>   24 65.596669700 ubuntu:16.04: Ok
>   25 52.759886216 ubuntu:16.04-x-arm: Ok
>   26 52.389293077 ubuntu:16.04-x-arm64: Ok
>   27 52.319141734 ubuntu:16.04-x-powerpc64: Ok
>   28 54.378287607 ubuntu:16.04-x-powerpc64el: Ok
>   29 72.794919209 ubuntu:16.10: Ok
>   30 53.351899868 ubuntu:16.10-x-s390: Ok
>      1922.64s
> 
>   real	32m3.246s
>   user	0m1.858s
>   sys	0m2.363s
>   [root@jouet ~]#

Pulled, thanks a lot Arnaldo!

	Ingo

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | linux.kernel

csiph-web