2009-10-08 17:17:38 -04:00
# ifndef _PROBE_FINDER_H
# define _PROBE_FINDER_H
2010-02-25 08:35:42 -05:00
# include <stdbool.h>
2009-12-27 21:37:00 -02:00
# include "util.h"
2010-03-16 18:06:12 -04:00
# include "probe-event.h"
2009-12-27 21:37:00 -02:00
2009-12-16 17:16:19 -05:00
# define MAX_PATH_LEN 256
# define MAX_PROBE_BUFFER 1024
# define MAX_PROBES 128
2009-10-08 17:17:38 -04:00
static inline int is_c_varname ( const char * name )
{
/* TODO */
return isalpha ( name [ 0 ] ) | | name [ 0 ] = = ' _ ' ;
}
2010-03-22 13:10:26 -03:00
# ifdef DWARF_SUPPORT
2010-07-29 19:43:51 +05:30
/* Find probe_trace_events specified by perf_probe_event from debuginfo */
extern int find_probe_trace_events ( int fd , struct perf_probe_event * pev ,
struct probe_trace_event * * tevs ,
2010-04-21 15:56:40 -04:00
int max_tevs ) ;
2010-03-16 18:06:12 -04:00
2010-03-16 18:06:19 -04:00
/* Find a perf_probe_point from debuginfo */
2010-10-21 19:13:41 +09:00
extern int find_perf_probe_point ( unsigned long addr ,
2010-03-16 18:06:19 -04:00
struct perf_probe_point * ppt ) ;
2010-10-21 19:13:23 +09:00
/* Find a line range */
2010-01-06 09:45:34 -05:00
extern int find_line_range ( int fd , struct line_range * lr ) ;
2009-10-08 17:17:38 -04:00
2010-10-21 19:13:23 +09:00
/* Find available variables */
extern int find_available_vars_at ( int fd , struct perf_probe_event * pev ,
2010-10-21 19:13:35 +09:00
struct variable_list * * vls , int max_points ,
bool externs ) ;
2011-06-27 16:27:27 +09:00
# include "dwarf-aux.h"
2009-10-08 17:17:38 -04:00
struct probe_finder {
2010-03-16 18:06:12 -04:00
struct perf_probe_event * pev ; /* Target probe event */
2010-10-21 19:13:23 +09:00
/* Callback when a probe point is found */
int ( * callback ) ( Dwarf_Die * sp_die , struct probe_finder * pf ) ;
2009-10-08 17:17:38 -04:00
/* For function searching */
perf tools: Reorganize some structs to save space
Using 'pahole --packable' I found some structs that could be reorganized
to eliminate alignment holes, in some cases getting them to be cacheline
multiples.
[acme@doppio linux-2.6-tip]$ codiff perf.old ~/bin/perf
builtin-annotate.c:
struct perf_session | -8
struct perf_header | -8
2 structs changed
builtin-diff.c:
struct sample_data | -8
1 struct changed
diff__process_sample_event | -8
1 function changed, 8 bytes removed, diff: -8
builtin-sched.c:
struct sched_atom | -8
1 struct changed
builtin-timechart.c:
struct per_pid | -8
1 struct changed
cmd_timechart | -16
1 function changed, 16 bytes removed, diff: -16
builtin-probe.c:
struct perf_probe_point | -8
struct perf_probe_event | -8
2 structs changed
opt_add_probe_event | -3
1 function changed, 3 bytes removed, diff: -3
util/probe-finder.c:
struct probe_finder | -8
1 struct changed
find_kprobe_trace_events | -16
1 function changed, 16 bytes removed, diff: -16
/home/acme/bin/perf:
4 functions changed, 43 bytes removed, diff: -43
[acme@doppio linux-2.6-tip]$
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2010-04-05 12:53:45 -03:00
int lno ; /* Line number */
2010-02-25 08:35:42 -05:00
Dwarf_Addr addr ; /* Address */
2010-03-16 18:06:12 -04:00
const char * fname ; /* Real file name */
2010-02-25 08:35:42 -05:00
Dwarf_Die cu_die ; /* Current CU */
perf probe: Add fastpath to do lookup by function name
v3 -> v2:
- Make pubname_search_cb more generic
- Add fastpath to find_probes also
v2 -> v1:
- Don't compare file names with cu_find_realpath(...), instead, compare
them with the name returned by dwarf_decl_file(sp_die)
The vmlinux file may have thousands of CUs.
We can lookup function name from .debug_pubnames section
to avoid the slow loop on CUs.
1. Improvement data for find_line_range
./perf stat -e cycles -r 10 -- ./perf probe -k /home/mlin/vmlinux \
-s /home/mlin/linux-2.6 \
--line csum_partial_copy_to_user > tmp.log
before patch applied
=====================
847,988,276 cycles
0.355075856 seconds time elapsed
after patch applied
=====================
206,102,622 cycles
0.086883555 seconds time elapsed
2. Improvement data for find_probes
./perf stat -e cycles -r 10 -- ./perf probe -k /home/mlin/vmlinux \
-s /home/mlin/linux-2.6 \
--vars csum_partial_copy_to_user > tmp.log
before patch applied
=====================
848,490,844 cycles
0.355307901 seconds time elapsed
after patch applied
=====================
205,684,469 cycles
0.086694010 seconds time elapsed
Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: linux-kernel <linux-kernel@vger.kernel.org>
LKML-Reference: <1301041668.14111.52.camel@minggr.sh.intel.com>
Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2011-03-25 16:27:48 +08:00
Dwarf_Die sp_die ;
2010-03-16 18:06:12 -04:00
struct list_head lcache ; /* Line cache for lazy match */
2009-10-08 17:17:38 -04:00
/* For variable searching */
2010-05-10 13:12:07 -04:00
# if _ELFUTILS_PREREQ(0, 142)
2010-04-12 13:17:29 -04:00
Dwarf_CFI * cfi ; /* Call Frame Information */
2010-05-10 13:12:07 -04:00
# endif
2010-02-25 08:35:42 -05:00
Dwarf_Op * fb_ops ; /* Frame base attribute */
2010-03-16 18:06:12 -04:00
struct perf_probe_arg * pvar ; /* Current target variable */
2010-07-29 19:43:51 +05:30
struct probe_trace_arg * tvar ; /* Current result variable */
2009-10-08 17:17:38 -04:00
} ;
2010-01-06 09:45:34 -05:00
2010-10-21 19:13:23 +09:00
struct trace_event_finder {
struct probe_finder pf ;
struct probe_trace_event * tevs ; /* Found trace events */
int ntevs ; /* Number of trace events */
int max_tevs ; /* Max number of trace events */
} ;
struct available_var_finder {
struct probe_finder pf ;
struct variable_list * vls ; /* Found variable lists */
int nvls ; /* Number of variable lists */
int max_vls ; /* Max no. of variable lists */
2010-10-21 19:13:35 +09:00
bool externs ; /* Find external vars too */
bool child ; /* Search child scopes */
2010-10-21 19:13:23 +09:00
} ;
2010-01-06 09:45:34 -05:00
struct line_finder {
2010-02-25 08:35:42 -05:00
struct line_range * lr ; /* Target line range */
const char * fname ; /* File name */
int lno_s ; /* Start line number */
int lno_e ; /* End line number */
Dwarf_Die cu_die ; /* Current CU */
perf probe: Add fastpath to do lookup by function name
v3 -> v2:
- Make pubname_search_cb more generic
- Add fastpath to find_probes also
v2 -> v1:
- Don't compare file names with cu_find_realpath(...), instead, compare
them with the name returned by dwarf_decl_file(sp_die)
The vmlinux file may have thousands of CUs.
We can lookup function name from .debug_pubnames section
to avoid the slow loop on CUs.
1. Improvement data for find_line_range
./perf stat -e cycles -r 10 -- ./perf probe -k /home/mlin/vmlinux \
-s /home/mlin/linux-2.6 \
--line csum_partial_copy_to_user > tmp.log
before patch applied
=====================
847,988,276 cycles
0.355075856 seconds time elapsed
after patch applied
=====================
206,102,622 cycles
0.086883555 seconds time elapsed
2. Improvement data for find_probes
./perf stat -e cycles -r 10 -- ./perf probe -k /home/mlin/vmlinux \
-s /home/mlin/linux-2.6 \
--vars csum_partial_copy_to_user > tmp.log
before patch applied
=====================
848,490,844 cycles
0.355307901 seconds time elapsed
after patch applied
=====================
205,684,469 cycles
0.086694010 seconds time elapsed
Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: linux-kernel <linux-kernel@vger.kernel.org>
LKML-Reference: <1301041668.14111.52.camel@minggr.sh.intel.com>
Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2011-03-25 16:27:48 +08:00
Dwarf_Die sp_die ;
2010-01-06 09:45:34 -05:00
int found ;
} ;
2010-03-22 13:10:26 -03:00
# endif /* DWARF_SUPPORT */
2009-10-08 17:17:38 -04:00
# endif /*_PROBE_FINDER_H */