aha/tools/perf/util/thread.c
Arnaldo Carvalho de Melo 1b46cddfcc perf tools: Use rb_tree for maps
Threads can have many and kernel modules will be represented as a
tree of maps as well.

Ah, and for a perf.data with 146607 samples:

Before:

[root@doppio ~]# perf stat -r 5 perf report > /dev/null

 Performance counter stats for 'perf report' (5 runs):

     699.823680  task-clock-msecs         #      0.991 CPUs    ( +-   0.454% )
             74  context-switches         #      0.000 M/sec   ( +-   1.709% )
              2  CPU-migrations           #      0.000 M/sec   ( +-  17.008% )
          23114  page-faults              #      0.033 M/sec   ( +-   0.000% )
     1381257019  cycles                   #   1973.721 M/sec   ( +-   0.290% )
     1456894438  instructions             #      1.055 IPC     ( +-   0.007% )
       18779818  cache-references         #     26.835 M/sec   ( +-   0.380% )
         641799  cache-misses             #      0.917 M/sec   ( +-   1.200% )

    0.705972729  seconds time elapsed   ( +-   0.501% )

[root@doppio ~]#

After

 Performance counter stats for 'perf report' (5 runs):

     691.261451  task-clock-msecs         #      0.993 CPUs    ( +-   0.307% )
             72  context-switches         #      0.000 M/sec   ( +-   0.829% )
              6  CPU-migrations           #      0.000 M/sec   ( +-  18.409% )
          23127  page-faults              #      0.033 M/sec   ( +-   0.000% )
     1366395876  cycles                   #   1976.670 M/sec   ( +-   0.153% )
     1443136016  instructions             #      1.056 IPC     ( +-   0.012% )
       17956402  cache-references         #     25.976 M/sec   ( +-   0.325% )
         661924  cache-misses             #      0.958 M/sec   ( +-   1.335% )

    0.696127275  seconds time elapsed   ( +-   0.377% )

I.e. we see some speedup too.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
LKML-Reference: <20090928174846.GA3361@ghostprotocols.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-30 13:57:56 +02:00

212 lines
4.2 KiB
C

#include "../perf.h"
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include "thread.h"
#include "util.h"
#include "debug.h"
static struct thread *thread__new(pid_t pid)
{
struct thread *self = calloc(1, sizeof(*self));
if (self != NULL) {
self->pid = pid;
self->comm = malloc(32);
if (self->comm)
snprintf(self->comm, 32, ":%d", self->pid);
self->maps = RB_ROOT;
}
return self;
}
int thread__set_comm(struct thread *self, const char *comm)
{
if (self->comm)
free(self->comm);
self->comm = strdup(comm);
return self->comm ? 0 : -ENOMEM;
}
static size_t thread__fprintf(struct thread *self, FILE *fp)
{
struct rb_node *nd;
size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm);
for (nd = rb_first(&self->maps); nd; nd = rb_next(nd)) {
struct map *pos = rb_entry(nd, struct map, rb_node);
ret += map__fprintf(pos, fp);
}
return ret;
}
struct thread *
threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
{
struct rb_node **p = &threads->rb_node;
struct rb_node *parent = NULL;
struct thread *th;
/*
* Font-end cache - PID lookups come in blocks,
* so most of the time we dont have to look up
* the full rbtree:
*/
if (*last_match && (*last_match)->pid == pid)
return *last_match;
while (*p != NULL) {
parent = *p;
th = rb_entry(parent, struct thread, rb_node);
if (th->pid == pid) {
*last_match = th;
return th;
}
if (pid < th->pid)
p = &(*p)->rb_left;
else
p = &(*p)->rb_right;
}
th = thread__new(pid);
if (th != NULL) {
rb_link_node(&th->rb_node, parent, p);
rb_insert_color(&th->rb_node, threads);
*last_match = th;
}
return th;
}
struct thread *
register_idle_thread(struct rb_root *threads, struct thread **last_match)
{
struct thread *thread = threads__findnew(0, threads, last_match);
if (!thread || thread__set_comm(thread, "swapper")) {
fprintf(stderr, "problem inserting idle task.\n");
exit(-1);
}
return thread;
}
static void thread__remove_overlappings(struct thread *self, struct map *map)
{
struct rb_node *next = rb_first(&self->maps);
while (next) {
struct map *pos = rb_entry(next, struct map, rb_node);
next = rb_next(&pos->rb_node);
if (!map__overlap(pos, map))
continue;
if (verbose >= 2) {
printf("overlapping maps:\n");
map__fprintf(map, stdout);
map__fprintf(pos, stdout);
}
if (map->start <= pos->start && map->end > pos->start)
pos->start = map->end;
if (map->end >= pos->end && map->start < pos->end)
pos->end = map->start;
if (verbose >= 2) {
printf("after collision:\n");
map__fprintf(pos, stdout);
}
if (pos->start >= pos->end) {
rb_erase(&pos->rb_node, &self->maps);
free(pos);
}
}
}
void maps__insert(struct rb_root *maps, struct map *map)
{
struct rb_node **p = &maps->rb_node;
struct rb_node *parent = NULL;
const u64 ip = map->start;
struct map *m;
while (*p != NULL) {
parent = *p;
m = rb_entry(parent, struct map, rb_node);
if (ip < m->start)
p = &(*p)->rb_left;
else
p = &(*p)->rb_right;
}
rb_link_node(&map->rb_node, parent, p);
rb_insert_color(&map->rb_node, maps);
}
struct map *maps__find(struct rb_root *maps, u64 ip)
{
struct rb_node **p = &maps->rb_node;
struct rb_node *parent = NULL;
struct map *m;
while (*p != NULL) {
parent = *p;
m = rb_entry(parent, struct map, rb_node);
if (ip < m->start)
p = &(*p)->rb_left;
else if (ip > m->end)
p = &(*p)->rb_right;
else
return m;
}
return NULL;
}
void thread__insert_map(struct thread *self, struct map *map)
{
thread__remove_overlappings(self, map);
maps__insert(&self->maps, map);
}
int thread__fork(struct thread *self, struct thread *parent)
{
struct rb_node *nd;
if (self->comm)
free(self->comm);
self->comm = strdup(parent->comm);
if (!self->comm)
return -ENOMEM;
for (nd = rb_first(&parent->maps); nd; nd = rb_next(nd)) {
struct map *map = rb_entry(nd, struct map, rb_node);
struct map *new = map__clone(map);
if (!new)
return -ENOMEM;
thread__insert_map(self, new);
}
return 0;
}
size_t threads__fprintf(FILE *fp, struct rb_root *threads)
{
size_t ret = 0;
struct rb_node *nd;
for (nd = rb_first(threads); nd; nd = rb_next(nd)) {
struct thread *pos = rb_entry(nd, struct thread, rb_node);
ret += thread__fprintf(pos, fp);
}
return ret;
}