2009-09-25 00:02:49 +08:00
|
|
|
#ifndef __PERF_SORT_H
|
|
|
|
#define __PERF_SORT_H
|
|
|
|
#include "../builtin.h"
|
|
|
|
|
|
|
|
#include "util.h"
|
|
|
|
|
|
|
|
#include "color.h"
|
|
|
|
#include <linux/list.h>
|
|
|
|
#include "cache.h"
|
|
|
|
#include <linux/rbtree.h>
|
|
|
|
#include "symbol.h"
|
|
|
|
#include "string.h"
|
|
|
|
#include "callchain.h"
|
|
|
|
#include "strlist.h"
|
|
|
|
#include "values.h"
|
|
|
|
|
|
|
|
#include "../perf.h"
|
|
|
|
#include "debug.h"
|
|
|
|
#include "header.h"
|
|
|
|
|
2015-12-15 23:39:39 +08:00
|
|
|
#include <subcmd/parse-options.h>
|
2009-09-25 00:02:49 +08:00
|
|
|
#include "parse-events.h"
|
2013-10-31 09:17:39 +08:00
|
|
|
#include "hist.h"
|
2009-09-25 00:02:49 +08:00
|
|
|
#include "thread.h"
|
|
|
|
|
|
|
|
extern regex_t parent_regex;
|
2010-05-18 03:22:41 +08:00
|
|
|
extern const char *sort_order;
|
2014-03-04 09:46:34 +08:00
|
|
|
extern const char *field_order;
|
2010-05-18 03:22:41 +08:00
|
|
|
extern const char default_parent_pattern[];
|
|
|
|
extern const char *parent_pattern;
|
2016-08-13 07:41:01 +08:00
|
|
|
extern const char *default_sort_order;
|
2012-12-07 13:48:05 +08:00
|
|
|
extern regex_t ignore_callees_regex;
|
|
|
|
extern int have_ignore_callees;
|
2013-04-01 19:35:20 +08:00
|
|
|
extern enum sort_mode sort__mode;
|
2009-09-25 00:02:49 +08:00
|
|
|
extern struct sort_entry sort_comm;
|
|
|
|
extern struct sort_entry sort_dso;
|
|
|
|
extern struct sort_entry sort_sym;
|
|
|
|
extern struct sort_entry sort_parent;
|
2012-03-09 06:47:48 +08:00
|
|
|
extern struct sort_entry sort_dso_from;
|
|
|
|
extern struct sort_entry sort_dso_to;
|
|
|
|
extern struct sort_entry sort_sym_from;
|
|
|
|
extern struct sort_entry sort_sym_to;
|
2016-09-19 21:10:10 +08:00
|
|
|
extern struct sort_entry sort_srcline;
|
perf tools: Bind callchains to the first sort dimension column
Currently, the callchains are displayed using a constant left
margin. So depending on the current sort dimension
configuration, callchains may appear to be well attached to the
first sort dimension column field which is mostly the case,
except when the first dimension of sorting is done by comm,
because these are right aligned.
This patch binds the callchain to the first letter in the first
column, whatever type of column it is (dso, comm, symbol).
Before:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
After:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
Also, for clarity, we don't put anymore the callchain as is but:
- If we have a top level ancestor in the callchain, start it
with a first ascii hook.
Before:
0.80% perf [kernel] [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
After:
0.80% perf [kernel] [k] __lock_acquire
|
--- __lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
- Otherwise, if we have several top level ancestors, then
display these like we did before:
1.69% Xorg
|
|--21.21%-- vread_hpet
| 0x7fffd85b46fc
| 0x7fffd85b494d
| 0x7f4fafb4e54d
|
|--15.15%-- exaOffscreenAlloc
|
|--9.09%-- I830WaitLpRing
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
LKML-Reference: <1256246604-17156-2-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-23 05:23:23 +08:00
|
|
|
extern enum sort_type sort__first_dimension;
|
perf tools: Remove (null) value of "Sort order" for perf mem report
When '--sort' is not set, 'perf mem report" will print a null pointer as
the output value of sort order, so fix it.
Example:
Before this patch:
$ perf mem report
# To display the perf.data header info, please use --header/--header-only options.
#
# Samples: 18 of event 'cpu/mem-loads/pp'
# Total weight : 188
# Sort order : (null)
#
...
After this patch:
$ perf mem report
# To display the perf.data header info, please use --header/--header-only options.
#
# Samples: 18 of event 'cpu/mem-loads/pp'
# Total weight : 188
# Sort order : local_weight,mem,sym,dso,symbol_daddr,dso_daddr,snoop,tlb,locked
#
...
Signed-off-by: Yunlong Song <yunlong.song@huawei.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1427082605-12881-1-git-send-email-yunlong.song@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-23 11:50:05 +08:00
|
|
|
extern const char default_mem_sort_order[];
|
2009-09-25 00:02:49 +08:00
|
|
|
|
2012-10-04 20:49:41 +08:00
|
|
|
struct he_stat {
|
|
|
|
u64 period;
|
|
|
|
u64 period_sys;
|
|
|
|
u64 period_us;
|
|
|
|
u64 period_guest_sys;
|
|
|
|
u64 period_guest_us;
|
2013-01-24 23:10:29 +08:00
|
|
|
u64 weight;
|
2012-10-04 20:49:41 +08:00
|
|
|
u32 nr_events;
|
|
|
|
};
|
|
|
|
|
2012-10-05 22:44:42 +08:00
|
|
|
struct hist_entry_diff {
|
|
|
|
bool computed;
|
2015-04-19 12:04:10 +08:00
|
|
|
union {
|
|
|
|
/* PERF_HPP__DELTA */
|
|
|
|
double period_ratio_delta;
|
2012-10-05 22:44:42 +08:00
|
|
|
|
2015-04-19 12:04:10 +08:00
|
|
|
/* PERF_HPP__RATIO */
|
|
|
|
double period_ratio;
|
2012-10-05 22:44:43 +08:00
|
|
|
|
2015-04-19 12:04:10 +08:00
|
|
|
/* HISTC_WEIGHTED_DIFF */
|
|
|
|
s64 wdiff;
|
|
|
|
};
|
2012-10-05 22:44:42 +08:00
|
|
|
};
|
|
|
|
|
2016-07-05 14:56:04 +08:00
|
|
|
struct hist_entry_ops {
|
|
|
|
void *(*new)(size_t size);
|
|
|
|
void (*free)(void *ptr);
|
|
|
|
};
|
|
|
|
|
2010-07-27 04:13:40 +08:00
|
|
|
/**
|
|
|
|
* struct hist_entry - histogram entry
|
|
|
|
*
|
|
|
|
* @row_offset - offset from the first callchain expanded to appear on screen
|
|
|
|
* @nr_rows - rows expanded in callchain, recalculated on folding/unfolding
|
|
|
|
*/
|
2009-09-25 00:02:49 +08:00
|
|
|
struct hist_entry {
|
2011-10-06 04:50:23 +08:00
|
|
|
struct rb_node rb_node_in;
|
2009-09-25 00:02:49 +08:00
|
|
|
struct rb_node rb_node;
|
2012-10-26 00:42:45 +08:00
|
|
|
union {
|
|
|
|
struct list_head node;
|
|
|
|
struct list_head head;
|
|
|
|
} pairs;
|
2012-10-04 20:49:41 +08:00
|
|
|
struct he_stat stat;
|
2012-09-11 12:15:07 +08:00
|
|
|
struct he_stat *stat_acc;
|
2010-03-25 03:40:17 +08:00
|
|
|
struct map_symbol ms;
|
2010-04-04 09:44:37 +08:00
|
|
|
struct thread *thread;
|
2013-09-13 15:28:57 +08:00
|
|
|
struct comm *comm;
|
2009-09-25 00:02:49 +08:00
|
|
|
u64 ip;
|
2013-09-20 22:40:43 +08:00
|
|
|
u64 transaction;
|
2015-09-04 22:45:42 +08:00
|
|
|
s32 socket;
|
2010-06-04 22:27:10 +08:00
|
|
|
s32 cpu;
|
2014-05-28 00:28:05 +08:00
|
|
|
u8 cpumode;
|
2016-02-24 23:13:34 +08:00
|
|
|
u8 depth;
|
2010-07-27 04:13:40 +08:00
|
|
|
|
2012-12-02 04:18:20 +08:00
|
|
|
/* We are added by hists__add_dummy_entry. */
|
|
|
|
bool dummy;
|
2016-02-24 23:13:34 +08:00
|
|
|
bool leaf;
|
2012-12-02 04:18:20 +08:00
|
|
|
|
2009-09-25 00:02:49 +08:00
|
|
|
char level;
|
2010-04-04 09:44:37 +08:00
|
|
|
u8 filtered;
|
2015-04-22 15:18:12 +08:00
|
|
|
union {
|
|
|
|
/*
|
|
|
|
* Since perf diff only supports the stdio output, TUI
|
|
|
|
* fields are only accessed from perf report (or perf
|
|
|
|
* top). So make it an union to reduce memory usage.
|
|
|
|
*/
|
|
|
|
struct hist_entry_diff diff;
|
|
|
|
struct /* for TUI */ {
|
|
|
|
u16 row_offset;
|
|
|
|
u16 nr_rows;
|
2015-04-22 15:18:13 +08:00
|
|
|
bool init_have_children;
|
2015-05-05 22:55:46 +08:00
|
|
|
bool unfolded;
|
|
|
|
bool has_children;
|
2016-02-26 20:13:19 +08:00
|
|
|
bool has_no_entry;
|
2015-04-22 15:18:12 +08:00
|
|
|
};
|
|
|
|
};
|
2012-05-30 21:33:24 +08:00
|
|
|
char *srcline;
|
2015-08-08 06:54:24 +08:00
|
|
|
char *srcfile;
|
2010-04-04 03:30:44 +08:00
|
|
|
struct symbol *parent;
|
2012-02-10 06:21:01 +08:00
|
|
|
struct branch_info *branch_info;
|
2012-10-04 20:49:35 +08:00
|
|
|
struct hists *hists;
|
2013-01-24 23:10:35 +08:00
|
|
|
struct mem_info *mem_info;
|
2015-12-24 10:16:17 +08:00
|
|
|
void *raw_data;
|
|
|
|
u32 raw_size;
|
2015-12-23 01:07:03 +08:00
|
|
|
void *trace_output;
|
2016-03-08 03:44:46 +08:00
|
|
|
struct perf_hpp_list *hpp_list;
|
2016-02-24 23:13:34 +08:00
|
|
|
struct hist_entry *parent_he;
|
2016-07-05 14:56:04 +08:00
|
|
|
struct hist_entry_ops *ops;
|
2016-02-24 23:13:34 +08:00
|
|
|
union {
|
|
|
|
/* this is for hierarchical entry structure */
|
|
|
|
struct {
|
|
|
|
struct rb_root hroot_in;
|
|
|
|
struct rb_root hroot_out;
|
|
|
|
}; /* non-leaf entries */
|
|
|
|
struct rb_root sorted_chain; /* leaf entry has callchains */
|
|
|
|
};
|
2013-01-24 23:10:35 +08:00
|
|
|
struct callchain_root callchain[0]; /* must be last member */
|
2009-09-25 00:02:49 +08:00
|
|
|
};
|
|
|
|
|
2012-10-26 00:42:45 +08:00
|
|
|
static inline bool hist_entry__has_pairs(struct hist_entry *he)
|
|
|
|
{
|
|
|
|
return !list_empty(&he->pairs.node);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline struct hist_entry *hist_entry__next_pair(struct hist_entry *he)
|
|
|
|
{
|
|
|
|
if (hist_entry__has_pairs(he))
|
|
|
|
return list_entry(he->pairs.node.next, struct hist_entry, pairs.node);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2012-12-13 21:09:00 +08:00
|
|
|
static inline void hist_entry__add_pair(struct hist_entry *pair,
|
|
|
|
struct hist_entry *he)
|
2012-10-26 00:42:45 +08:00
|
|
|
{
|
2012-12-13 21:09:00 +08:00
|
|
|
list_add_tail(&pair->pairs.node, &he->pairs.head);
|
2012-10-26 00:42:45 +08:00
|
|
|
}
|
|
|
|
|
2013-10-31 09:17:39 +08:00
|
|
|
static inline float hist_entry__get_percent_limit(struct hist_entry *he)
|
|
|
|
{
|
|
|
|
u64 period = he->stat.period;
|
|
|
|
u64 total_period = hists__total_period(he->hists);
|
|
|
|
|
|
|
|
if (unlikely(total_period == 0))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (symbol_conf.cumulate_callchain)
|
|
|
|
period = he->stat_acc->period;
|
|
|
|
|
|
|
|
return period * 100.0 / total_period;
|
|
|
|
}
|
|
|
|
|
2016-02-15 16:34:32 +08:00
|
|
|
static inline u64 cl_address(u64 address)
|
|
|
|
{
|
|
|
|
/* return the cacheline of the address */
|
|
|
|
return (address & ~(cacheline_size - 1));
|
|
|
|
}
|
2013-10-31 09:17:39 +08:00
|
|
|
|
2016-02-15 16:34:33 +08:00
|
|
|
static inline u64 cl_offset(u64 address)
|
|
|
|
{
|
|
|
|
/* return the cacheline of the address */
|
|
|
|
return (address & (cacheline_size - 1));
|
|
|
|
}
|
|
|
|
|
2013-04-01 19:35:20 +08:00
|
|
|
enum sort_mode {
|
|
|
|
SORT_MODE__NORMAL,
|
|
|
|
SORT_MODE__BRANCH,
|
|
|
|
SORT_MODE__MEMORY,
|
2014-03-18 10:31:39 +08:00
|
|
|
SORT_MODE__TOP,
|
|
|
|
SORT_MODE__DIFF,
|
2015-12-23 01:07:10 +08:00
|
|
|
SORT_MODE__TRACEPOINT,
|
2013-04-01 19:35:20 +08:00
|
|
|
};
|
|
|
|
|
perf tools: Bind callchains to the first sort dimension column
Currently, the callchains are displayed using a constant left
margin. So depending on the current sort dimension
configuration, callchains may appear to be well attached to the
first sort dimension column field which is mostly the case,
except when the first dimension of sorting is done by comm,
because these are right aligned.
This patch binds the callchain to the first letter in the first
column, whatever type of column it is (dso, comm, symbol).
Before:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
After:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
Also, for clarity, we don't put anymore the callchain as is but:
- If we have a top level ancestor in the callchain, start it
with a first ascii hook.
Before:
0.80% perf [kernel] [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
After:
0.80% perf [kernel] [k] __lock_acquire
|
--- __lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
- Otherwise, if we have several top level ancestors, then
display these like we did before:
1.69% Xorg
|
|--21.21%-- vread_hpet
| 0x7fffd85b46fc
| 0x7fffd85b494d
| 0x7f4fafb4e54d
|
|--15.15%-- exaOffscreenAlloc
|
|--9.09%-- I830WaitLpRing
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
LKML-Reference: <1256246604-17156-2-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-23 05:23:23 +08:00
|
|
|
enum sort_type {
|
2012-12-27 17:11:46 +08:00
|
|
|
/* common sort keys */
|
perf tools: Bind callchains to the first sort dimension column
Currently, the callchains are displayed using a constant left
margin. So depending on the current sort dimension
configuration, callchains may appear to be well attached to the
first sort dimension column field which is mostly the case,
except when the first dimension of sorting is done by comm,
because these are right aligned.
This patch binds the callchain to the first letter in the first
column, whatever type of column it is (dso, comm, symbol).
Before:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
After:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
Also, for clarity, we don't put anymore the callchain as is but:
- If we have a top level ancestor in the callchain, start it
with a first ascii hook.
Before:
0.80% perf [kernel] [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
After:
0.80% perf [kernel] [k] __lock_acquire
|
--- __lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
- Otherwise, if we have several top level ancestors, then
display these like we did before:
1.69% Xorg
|
|--21.21%-- vread_hpet
| 0x7fffd85b46fc
| 0x7fffd85b494d
| 0x7f4fafb4e54d
|
|--15.15%-- exaOffscreenAlloc
|
|--9.09%-- I830WaitLpRing
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
LKML-Reference: <1256246604-17156-2-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-23 05:23:23 +08:00
|
|
|
SORT_PID,
|
|
|
|
SORT_COMM,
|
|
|
|
SORT_DSO,
|
|
|
|
SORT_SYM,
|
2010-06-04 22:27:10 +08:00
|
|
|
SORT_PARENT,
|
|
|
|
SORT_CPU,
|
2015-09-04 22:45:43 +08:00
|
|
|
SORT_SOCKET,
|
2012-12-27 17:11:46 +08:00
|
|
|
SORT_SRCLINE,
|
2015-08-08 06:54:24 +08:00
|
|
|
SORT_SRCFILE,
|
2013-07-19 06:58:53 +08:00
|
|
|
SORT_LOCAL_WEIGHT,
|
|
|
|
SORT_GLOBAL_WEIGHT,
|
2013-09-20 22:40:43 +08:00
|
|
|
SORT_TRANSACTION,
|
2015-12-23 01:07:04 +08:00
|
|
|
SORT_TRACE,
|
2012-12-27 17:11:46 +08:00
|
|
|
|
|
|
|
/* branch stack specific sort keys */
|
|
|
|
__SORT_BRANCH_STACK,
|
|
|
|
SORT_DSO_FROM = __SORT_BRANCH_STACK,
|
2012-02-10 06:21:01 +08:00
|
|
|
SORT_DSO_TO,
|
|
|
|
SORT_SYM_FROM,
|
|
|
|
SORT_SYM_TO,
|
|
|
|
SORT_MISPREDICT,
|
2013-09-20 22:40:41 +08:00
|
|
|
SORT_ABORT,
|
|
|
|
SORT_IN_TX,
|
2015-07-18 23:24:46 +08:00
|
|
|
SORT_CYCLES,
|
2016-05-21 04:15:08 +08:00
|
|
|
SORT_SRCLINE_FROM,
|
|
|
|
SORT_SRCLINE_TO,
|
2013-04-03 20:26:11 +08:00
|
|
|
|
|
|
|
/* memory mode specific sort keys */
|
|
|
|
__SORT_MEMORY_MODE,
|
2013-07-19 06:58:53 +08:00
|
|
|
SORT_MEM_DADDR_SYMBOL = __SORT_MEMORY_MODE,
|
2013-04-03 20:26:11 +08:00
|
|
|
SORT_MEM_DADDR_DSO,
|
|
|
|
SORT_MEM_LOCKED,
|
|
|
|
SORT_MEM_TLB,
|
|
|
|
SORT_MEM_LVL,
|
|
|
|
SORT_MEM_SNOOP,
|
2014-06-01 21:38:29 +08:00
|
|
|
SORT_MEM_DCACHELINE,
|
2015-10-06 02:06:07 +08:00
|
|
|
SORT_MEM_IADDR_SYMBOL,
|
perf tools: Bind callchains to the first sort dimension column
Currently, the callchains are displayed using a constant left
margin. So depending on the current sort dimension
configuration, callchains may appear to be well attached to the
first sort dimension column field which is mostly the case,
except when the first dimension of sorting is done by comm,
because these are right aligned.
This patch binds the callchain to the first letter in the first
column, whatever type of column it is (dso, comm, symbol).
Before:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
After:
0.80% perf [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
| | __fsnotify_parent
Also, for clarity, we don't put anymore the callchain as is but:
- If we have a top level ancestor in the callchain, start it
with a first ascii hook.
Before:
0.80% perf [kernel] [k] __lock_acquire
__lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
After:
0.80% perf [kernel] [k] __lock_acquire
|
--- __lock_acquire
lock_acquire
|
|--58.33%-- _spin_lock
| |
| |--28.57%-- inotify_should_send_event
| | fsnotify
[..] [..]
- Otherwise, if we have several top level ancestors, then
display these like we did before:
1.69% Xorg
|
|--21.21%-- vread_hpet
| 0x7fffd85b46fc
| 0x7fffd85b494d
| 0x7f4fafb4e54d
|
|--15.15%-- exaOffscreenAlloc
|
|--9.09%-- I830WaitLpRing
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
LKML-Reference: <1256246604-17156-2-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-23 05:23:23 +08:00
|
|
|
};
|
|
|
|
|
2009-09-25 00:02:49 +08:00
|
|
|
/*
|
|
|
|
* configurable sorting bits
|
|
|
|
*/
|
|
|
|
|
|
|
|
struct sort_entry {
|
2010-04-15 01:11:29 +08:00
|
|
|
const char *se_header;
|
2009-09-25 00:02:49 +08:00
|
|
|
|
2010-04-15 01:11:29 +08:00
|
|
|
int64_t (*se_cmp)(struct hist_entry *, struct hist_entry *);
|
|
|
|
int64_t (*se_collapse)(struct hist_entry *, struct hist_entry *);
|
2014-03-04 10:01:41 +08:00
|
|
|
int64_t (*se_sort)(struct hist_entry *, struct hist_entry *);
|
2013-11-06 02:32:36 +08:00
|
|
|
int (*se_snprintf)(struct hist_entry *he, char *bf, size_t size,
|
2010-04-15 01:11:29 +08:00
|
|
|
unsigned int width);
|
2016-02-24 23:13:37 +08:00
|
|
|
int (*se_filter)(struct hist_entry *he, int type, const void *arg);
|
2010-07-21 01:42:52 +08:00
|
|
|
u8 se_width_idx;
|
2009-09-25 00:02:49 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
extern struct sort_entry sort_thread;
|
|
|
|
extern struct list_head hist_entry__sort_list;
|
|
|
|
|
2015-12-23 01:07:01 +08:00
|
|
|
struct perf_evlist;
|
|
|
|
struct pevent;
|
|
|
|
int setup_sorting(struct perf_evlist *evlist);
|
2014-03-04 09:46:34 +08:00
|
|
|
int setup_output_field(void);
|
2014-05-07 17:42:24 +08:00
|
|
|
void reset_output_field(void);
|
2013-04-03 20:26:19 +08:00
|
|
|
void sort__setup_elide(FILE *fp);
|
perf tools: Move elide bool into perf_hpp_fmt struct
After output/sort fields refactoring, it's expensive
to check the elide bool in its current location inside
the 'struct sort_entry'.
The perf_hpp__should_skip function gets highly noticable in
workloads with high number of output/sort fields, like for:
$ perf report -i perf-test.data -F overhead,sample,period,comm,pid,dso,symbol,cpu --stdio
Performance report:
9.70% perf [.] perf_hpp__should_skip
Moving the elide bool into the 'struct perf_hpp_fmt', which
makes the perf_hpp__should_skip just single struct read.
Got speedup of around 22% for my test perf.data workload.
The change should not harm any other workload types.
Performance counter stats for (10 runs):
before:
358,319,732,626 cycles ( +- 0.55% )
467,129,581,515 instructions # 1.30 insns per cycle ( +- 0.00% )
150.943975206 seconds time elapsed ( +- 0.62% )
now:
278,785,972,990 cycles ( +- 0.12% )
370,146,797,640 instructions # 1.33 insns per cycle ( +- 0.00% )
116.416670507 seconds time elapsed ( +- 0.31% )
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20140601142622.GA9131@krava.brq.redhat.com
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
2014-05-23 23:15:47 +08:00
|
|
|
void perf_hpp__set_elide(int idx, bool elide);
|
2009-09-25 00:02:49 +08:00
|
|
|
|
2012-12-07 13:48:05 +08:00
|
|
|
int report_parse_ignore_callees_opt(const struct option *opt, const char *arg, int unset);
|
|
|
|
|
2014-08-22 21:58:38 +08:00
|
|
|
bool is_strict_order(const char *order);
|
2015-10-06 20:25:11 +08:00
|
|
|
|
|
|
|
int hpp_dimension__add_output(unsigned col);
|
2016-09-22 23:36:32 +08:00
|
|
|
void reset_dimensions(void);
|
2016-09-22 23:36:33 +08:00
|
|
|
int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
|
|
|
|
struct perf_evlist *evlist,
|
|
|
|
int level);
|
|
|
|
int output_field_add(struct perf_hpp_list *list, char *tok);
|
2009-09-25 00:02:49 +08:00
|
|
|
#endif /* __PERF_SORT_H */
|