linux_old1/tools/perf/util/event.c

613 lines
15 KiB
C
Raw Normal View History

#include <linux/types.h>
#include "event.h"
#include "debug.h"
#include "session.h"
#include "sort.h"
#include "string.h"
#include "strlist.h"
#include "thread.h"
static pid_t event__synthesize_comm(pid_t pid, int full,
event__handler_t process,
struct perf_session *session)
{
event_t ev;
char filename[PATH_MAX];
char bf[BUFSIZ];
FILE *fp;
size_t size = 0;
DIR *tasks;
struct dirent dirent, *next;
pid_t tgid = 0;
snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
fp = fopen(filename, "r");
if (fp == NULL) {
out_race:
/*
* We raced with a task exiting - just return:
*/
pr_debug("couldn't open %s\n", filename);
return 0;
}
memset(&ev.comm, 0, sizeof(ev.comm));
while (!ev.comm.comm[0] || !ev.comm.pid) {
if (fgets(bf, sizeof(bf), fp) == NULL)
goto out_failure;
if (memcmp(bf, "Name:", 5) == 0) {
char *name = bf + 5;
while (*name && isspace(*name))
++name;
size = strlen(name) - 1;
memcpy(ev.comm.comm, name, size++);
} else if (memcmp(bf, "Tgid:", 5) == 0) {
char *tgids = bf + 5;
while (*tgids && isspace(*tgids))
++tgids;
tgid = ev.comm.pid = atoi(tgids);
}
}
ev.comm.header.type = PERF_RECORD_COMM;
size = ALIGN(size, sizeof(u64));
ev.comm.header.size = sizeof(ev.comm) - (sizeof(ev.comm.comm) - size);
if (!full) {
ev.comm.tid = pid;
process(&ev, session);
goto out_fclose;
}
snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
tasks = opendir(filename);
if (tasks == NULL)
goto out_race;
while (!readdir_r(tasks, &dirent, &next) && next) {
char *end;
pid = strtol(dirent.d_name, &end, 10);
if (*end)
continue;
ev.comm.tid = pid;
process(&ev, session);
}
closedir(tasks);
out_fclose:
fclose(fp);
return tgid;
out_failure:
pr_warning("couldn't get COMM and pgid, malformed %s\n", filename);
return -1;
}
static int event__synthesize_mmap_events(pid_t pid, pid_t tgid,
event__handler_t process,
struct perf_session *session)
{
char filename[PATH_MAX];
FILE *fp;
snprintf(filename, sizeof(filename), "/proc/%d/maps", pid);
fp = fopen(filename, "r");
if (fp == NULL) {
/*
* We raced with a task exiting - just return:
*/
pr_debug("couldn't open %s\n", filename);
return -1;
}
while (1) {
char bf[BUFSIZ], *pbf = bf;
event_t ev = {
.header = {
.type = PERF_RECORD_MMAP,
.misc = 0, /* Just like the kernel, see kernel/perf_event.c __perf_event_mmap */
},
};
int n;
size_t size;
if (fgets(bf, sizeof(bf), fp) == NULL)
break;
/* 00400000-0040c000 r-xp 00000000 fd:01 41038 /bin/cat */
n = hex2u64(pbf, &ev.mmap.start);
if (n < 0)
continue;
pbf += n + 1;
n = hex2u64(pbf, &ev.mmap.len);
if (n < 0)
continue;
pbf += n + 3;
if (*pbf == 'x') { /* vm_exec */
char *execname = strchr(bf, '/');
/* Catch VDSO */
if (execname == NULL)
execname = strstr(bf, "[vdso]");
if (execname == NULL)
continue;
size = strlen(execname);
execname[size - 1] = '\0'; /* Remove \n */
memcpy(ev.mmap.filename, execname, size);
size = ALIGN(size, sizeof(u64));
ev.mmap.len -= ev.mmap.start;
ev.mmap.header.size = (sizeof(ev.mmap) -
(sizeof(ev.mmap.filename) - size));
ev.mmap.pid = tgid;
ev.mmap.tid = pid;
process(&ev, session);
}
}
fclose(fp);
return 0;
}
int event__synthesize_modules(event__handler_t process,
struct perf_session *session)
{
struct rb_node *nd;
for (nd = rb_first(&session->kmaps.maps[MAP__FUNCTION]);
nd; nd = rb_next(nd)) {
event_t ev;
size_t size;
struct map *pos = rb_entry(nd, struct map, rb_node);
if (pos->dso->kernel)
continue;
size = ALIGN(pos->dso->long_name_len + 1, sizeof(u64));
memset(&ev, 0, sizeof(ev));
ev.mmap.header.misc = 1; /* kernel uses 0 for user space maps, see kernel/perf_event.c __perf_event_mmap */
ev.mmap.header.type = PERF_RECORD_MMAP;
ev.mmap.header.size = (sizeof(ev.mmap) -
(sizeof(ev.mmap.filename) - size));
ev.mmap.start = pos->start;
ev.mmap.len = pos->end - pos->start;
memcpy(ev.mmap.filename, pos->dso->long_name,
pos->dso->long_name_len + 1);
process(&ev, session);
}
return 0;
}
int event__synthesize_thread(pid_t pid, event__handler_t process,
struct perf_session *session)
{
pid_t tgid = event__synthesize_comm(pid, 1, process, session);
if (tgid == -1)
return -1;
return event__synthesize_mmap_events(pid, tgid, process, session);
}
void event__synthesize_threads(event__handler_t process,
struct perf_session *session)
{
DIR *proc;
struct dirent dirent, *next;
proc = opendir("/proc");
while (!readdir_r(proc, &dirent, &next) && next) {
char *end;
pid_t pid = strtol(dirent.d_name, &end, 10);
if (*end) /* only interested in proper numerical dirents */
continue;
event__synthesize_thread(pid, process, session);
}
closedir(proc);
}
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
struct process_symbol_args {
const char *name;
u64 start;
};
static int find_symbol_cb(void *arg, const char *name, char type, u64 start)
{
struct process_symbol_args *args = arg;
/*
* Must be a function or at least an alias, as in PARISC64, where "_text" is
* an 'A' to the same address as "_stext".
*/
if (!(symbol_type__is_a(type, MAP__FUNCTION) ||
type == 'A') || strcmp(name, args->name))
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
return 0;
args->start = start;
return 1;
}
int event__synthesize_kernel_mmap(event__handler_t process,
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
struct perf_session *session,
const char *symbol_name)
{
size_t size;
event_t ev = {
.header = {
.type = PERF_RECORD_MMAP,
.misc = 1, /* kernel uses 0 for user space maps, see kernel/perf_event.c __perf_event_mmap */
},
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
};
/*
* We should get this from /sys/kernel/sections/.text, but till that is
* available use this, and after it is use this as a fallback for older
* kernels.
*/
struct process_symbol_args args = { .name = symbol_name, };
if (kallsyms__parse("/proc/kallsyms", &args, find_symbol_cb) <= 0)
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
return -ENOENT;
size = snprintf(ev.mmap.filename, sizeof(ev.mmap.filename),
"[kernel.kallsyms.%s]", symbol_name) + 1;
size = ALIGN(size, sizeof(u64));
ev.mmap.header.size = (sizeof(ev.mmap) - (sizeof(ev.mmap.filename) - size));
ev.mmap.pgoff = args.start;
ev.mmap.start = session->vmlinux_maps[MAP__FUNCTION]->start;
ev.mmap.len = session->vmlinux_maps[MAP__FUNCTION]->end - ev.mmap.start ;
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
return process(&ev, session);
}
static void thread__comm_adjust(struct thread *self)
{
char *comm = self->comm;
if (!symbol_conf.col_width_list_str && !symbol_conf.field_sep &&
(!symbol_conf.comm_list ||
strlist__has_entry(symbol_conf.comm_list, comm))) {
unsigned int slen = strlen(comm);
if (slen > comms__col_width) {
comms__col_width = slen;
threads__col_width = slen + 6;
}
}
}
static int thread__set_comm_adjust(struct thread *self, const char *comm)
{
int ret = thread__set_comm(self, comm);
if (ret)
return ret;
thread__comm_adjust(self);
return 0;
}
int event__process_comm(event_t *self, struct perf_session *session)
{
struct thread *thread = perf_session__findnew(session, self->comm.pid);
dump_printf(": %s:%d\n", self->comm.comm, self->comm.pid);
if (thread == NULL || thread__set_comm_adjust(thread, self->comm.comm)) {
dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n");
return -1;
}
return 0;
}
int event__process_lost(event_t *self, struct perf_session *session)
{
dump_printf(": id:%Ld: lost:%Ld\n", self->lost.id, self->lost.lost);
session->events_stats.lost += self->lost.lost;
return 0;
}
int event__process_mmap(event_t *self, struct perf_session *session)
{
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
struct thread *thread;
struct map *map;
dump_printf(" %d/%d: [%#Lx(%#Lx) @ %#Lx]: %s\n",
self->mmap.pid, self->mmap.tid, self->mmap.start,
self->mmap.len, self->mmap.pgoff, self->mmap.filename);
if (self->mmap.pid == 0) {
static const char kmmap_prefix[] = "[kernel.kallsyms.";
if (self->mmap.filename[0] == '/') {
char short_module_name[1024];
char *name = strrchr(self->mmap.filename, '/'), *dot;
if (name == NULL)
goto out_problem;
++name; /* skip / */
dot = strrchr(name, '.');
if (dot == NULL)
goto out_problem;
snprintf(short_module_name, sizeof(short_module_name),
"[%.*s]", (int)(dot - name), name);
strxfrchar(short_module_name, '-', '_');
map = perf_session__new_module_map(session,
self->mmap.start,
perf symbols: The synthesized kernel modules MMAP must use the pathnames Since we use ->long_name in dsos__find now. Now 'perf buildid_list' is not duplicating those and managing to show the proper build-ids for the DSOs with hits: [root@doppio linux-2.6-tip]# perf buildid-list -H 74f9930ee94475b6b3238caf3725a50d59cb994b [kernel.kallsyms] 9ffdcac0a7935922d1f04b6cc9029dfef0f066ef /lib/modules/2.6.33-rc4-tip+/kernel/arch/x86/crypto/aes-x86_64.ko 3aaf89c32ebfc438ff546c93597d41788e3e65f3 /lib/modules/2.6.33-rc4-tip+/kernel/drivers/net/wireless/iwlwifi/iwl3945.ko 19f46033f73e1ec612937189bb118c5daba5a0c8 /lib/modules/2.6.33-rc4-tip+/kernel/net/mac80211/mac80211.ko 1772f014a7a7272859655acb0c64a20ab20b75ee /lib/modules/2.6.33-rc4-tip+/kernel/drivers/net/e1000e/e1000e.ko eb4ec8fa8b2a5eb18cad173c92f27ed8887ed1c1 /lib64/libc-2.10.2.so 5c68f7afeb33309c78037e374b0deee84dd441f6 /lib64/libpthread-2.10.2.so e9c9ad5c138ef882e4507d2605645b597da43873 /bin/dbus-daemon bcda7d09eb6c9ee380dae0ed3d591d4311decc31 /lib64/libdbus-1.so.3.4.0 7cc449a77f48b85d6088114000e970ced613bed8 /usr/lib64/libcrypto.so.0.9.8k fdd1ccd1ff7917ab020653147ab3bacf0a85b5b9 /lib64/libglib-2.0.so.0.2000.5 e4417ebb8762e5f2eee93c8011a71115ff5edad8 /lib64/libgobject-2.0.so.0.2000.5 931e49461f6df99104f0febcc52f6fed5e2efce6 /usr/sbin/sshd dab5f724c088f89fbd8304da553ed6cb30bbec96 /usr/lib64/libgdk-x11-2.0.so.0.1600.6 f2037a091ef36b591187a858d75e203690ea9409 /usr/sbin/openvpn a8e4f743b40fb1fd8b85e2f9b88d93b661472b8f /bin/find 81120aada06e68b1e85882925a0fc6d7345ef59a /home/acme/bin/perf [root@doppio linux-2.6-tip]# Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> 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: <1263568672-30323-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-15 23:17:51 +08:00
self->mmap.filename);
if (map == NULL)
goto out_problem;
perf symbols: The synthesized kernel modules MMAP must use the pathnames Since we use ->long_name in dsos__find now. Now 'perf buildid_list' is not duplicating those and managing to show the proper build-ids for the DSOs with hits: [root@doppio linux-2.6-tip]# perf buildid-list -H 74f9930ee94475b6b3238caf3725a50d59cb994b [kernel.kallsyms] 9ffdcac0a7935922d1f04b6cc9029dfef0f066ef /lib/modules/2.6.33-rc4-tip+/kernel/arch/x86/crypto/aes-x86_64.ko 3aaf89c32ebfc438ff546c93597d41788e3e65f3 /lib/modules/2.6.33-rc4-tip+/kernel/drivers/net/wireless/iwlwifi/iwl3945.ko 19f46033f73e1ec612937189bb118c5daba5a0c8 /lib/modules/2.6.33-rc4-tip+/kernel/net/mac80211/mac80211.ko 1772f014a7a7272859655acb0c64a20ab20b75ee /lib/modules/2.6.33-rc4-tip+/kernel/drivers/net/e1000e/e1000e.ko eb4ec8fa8b2a5eb18cad173c92f27ed8887ed1c1 /lib64/libc-2.10.2.so 5c68f7afeb33309c78037e374b0deee84dd441f6 /lib64/libpthread-2.10.2.so e9c9ad5c138ef882e4507d2605645b597da43873 /bin/dbus-daemon bcda7d09eb6c9ee380dae0ed3d591d4311decc31 /lib64/libdbus-1.so.3.4.0 7cc449a77f48b85d6088114000e970ced613bed8 /usr/lib64/libcrypto.so.0.9.8k fdd1ccd1ff7917ab020653147ab3bacf0a85b5b9 /lib64/libglib-2.0.so.0.2000.5 e4417ebb8762e5f2eee93c8011a71115ff5edad8 /lib64/libgobject-2.0.so.0.2000.5 931e49461f6df99104f0febcc52f6fed5e2efce6 /usr/sbin/sshd dab5f724c088f89fbd8304da553ed6cb30bbec96 /usr/lib64/libgdk-x11-2.0.so.0.1600.6 f2037a091ef36b591187a858d75e203690ea9409 /usr/sbin/openvpn a8e4f743b40fb1fd8b85e2f9b88d93b661472b8f /bin/find 81120aada06e68b1e85882925a0fc6d7345ef59a /home/acme/bin/perf [root@doppio linux-2.6-tip]# Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> 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: <1263568672-30323-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-15 23:17:51 +08:00
name = strdup(short_module_name);
if (name == NULL)
goto out_problem;
perf symbols: The synthesized kernel modules MMAP must use the pathnames Since we use ->long_name in dsos__find now. Now 'perf buildid_list' is not duplicating those and managing to show the proper build-ids for the DSOs with hits: [root@doppio linux-2.6-tip]# perf buildid-list -H 74f9930ee94475b6b3238caf3725a50d59cb994b [kernel.kallsyms] 9ffdcac0a7935922d1f04b6cc9029dfef0f066ef /lib/modules/2.6.33-rc4-tip+/kernel/arch/x86/crypto/aes-x86_64.ko 3aaf89c32ebfc438ff546c93597d41788e3e65f3 /lib/modules/2.6.33-rc4-tip+/kernel/drivers/net/wireless/iwlwifi/iwl3945.ko 19f46033f73e1ec612937189bb118c5daba5a0c8 /lib/modules/2.6.33-rc4-tip+/kernel/net/mac80211/mac80211.ko 1772f014a7a7272859655acb0c64a20ab20b75ee /lib/modules/2.6.33-rc4-tip+/kernel/drivers/net/e1000e/e1000e.ko eb4ec8fa8b2a5eb18cad173c92f27ed8887ed1c1 /lib64/libc-2.10.2.so 5c68f7afeb33309c78037e374b0deee84dd441f6 /lib64/libpthread-2.10.2.so e9c9ad5c138ef882e4507d2605645b597da43873 /bin/dbus-daemon bcda7d09eb6c9ee380dae0ed3d591d4311decc31 /lib64/libdbus-1.so.3.4.0 7cc449a77f48b85d6088114000e970ced613bed8 /usr/lib64/libcrypto.so.0.9.8k fdd1ccd1ff7917ab020653147ab3bacf0a85b5b9 /lib64/libglib-2.0.so.0.2000.5 e4417ebb8762e5f2eee93c8011a71115ff5edad8 /lib64/libgobject-2.0.so.0.2000.5 931e49461f6df99104f0febcc52f6fed5e2efce6 /usr/sbin/sshd dab5f724c088f89fbd8304da553ed6cb30bbec96 /usr/lib64/libgdk-x11-2.0.so.0.1600.6 f2037a091ef36b591187a858d75e203690ea9409 /usr/sbin/openvpn a8e4f743b40fb1fd8b85e2f9b88d93b661472b8f /bin/find 81120aada06e68b1e85882925a0fc6d7345ef59a /home/acme/bin/perf [root@doppio linux-2.6-tip]# Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> 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: <1263568672-30323-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-15 23:17:51 +08:00
map->dso->short_name = name;
map->end = map->start + self->mmap.len;
} else if (memcmp(self->mmap.filename, kmmap_prefix,
sizeof(kmmap_prefix) - 1) == 0) {
const char *symbol_name = (self->mmap.filename +
sizeof(kmmap_prefix) - 1);
/*
* Should be there already, from the build-id table in
* the header.
*/
struct dso *kernel = __dsos__findnew(&dsos__kernel,
"[kernel.kallsyms]");
if (kernel == NULL)
goto out_problem;
kernel->kernel = 1;
if (__perf_session__create_kernel_maps(session, kernel) < 0)
goto out_problem;
session->vmlinux_maps[MAP__FUNCTION]->start = self->mmap.start;
session->vmlinux_maps[MAP__FUNCTION]->end = self->mmap.start + self->mmap.len;
/*
* Be a bit paranoid here, some perf.data file came with
* a zero sized synthesized MMAP event for the kernel.
*/
if (session->vmlinux_maps[MAP__FUNCTION]->end == 0)
session->vmlinux_maps[MAP__FUNCTION]->end = ~0UL;
perf_session__set_kallsyms_ref_reloc_sym(session, symbol_name,
self->mmap.pgoff);
}
perf tools: Handle relocatable kernels DSOs don't have this problem because the kernel emits a PERF_MMAP for each new executable mapping it performs on monitored threads. To fix the kernel case we simulate the same behaviour, by having 'perf record' to synthesize a PERF_MMAP for the kernel, encoded like this: [root@doppio ~]# perf record -a -f sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ] [root@doppio ~]# perf report -D | head -10 0xd0 [0x40]: event: 1 . . ... raw event: size 64 bytes . 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........ . 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ............... . 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel . 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text] . 0xd0 [0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text] I.e. we identify such event as having: .pid = 0 .filename = [kernel.kallsyms.REFNAME] .start = REFNAME addr in /proc/kallsyms at 'perf record' time and use now a hardcoded value of '.text' for REFNAME. Then, later, in 'perf report', if there are any kernel hits and thus we need to resolve kernel symbols, we search for REFNAME and if its address changed, relocation happened and we thus must change the kernel mapping routines to one that uses .pgoff as the relocation to apply. This way we use the same mechanism used for the other DSOs and don't have to do a two pass in all the kernel symbols. Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 02:50:31 +08:00
return 0;
}
thread = perf_session__findnew(session, self->mmap.pid);
map = map__new(&self->mmap, MAP__FUNCTION,
session->cwd, session->cwdlen);
if (thread == NULL || map == NULL)
goto out_problem;
thread__insert_map(thread, map);
return 0;
out_problem:
dump_printf("problem processing PERF_RECORD_MMAP, skipping event.\n");
return 0;
}
int event__process_task(event_t *self, struct perf_session *session)
{
struct thread *thread = perf_session__findnew(session, self->fork.pid);
struct thread *parent = perf_session__findnew(session, self->fork.ppid);
dump_printf("(%d:%d):(%d:%d)\n", self->fork.pid, self->fork.tid,
self->fork.ppid, self->fork.ptid);
/*
* A thread clone will have the same PID for both parent and child.
*/
if (thread == parent)
return 0;
if (self->header.type == PERF_RECORD_EXIT)
return 0;
if (thread == NULL || parent == NULL ||
thread__fork(thread, parent) < 0) {
dump_printf("problem processing PERF_RECORD_FORK, skipping event.\n");
return -1;
}
return 0;
}
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
void thread__find_addr_map(struct thread *self,
struct perf_session *session, u8 cpumode,
enum map_type type, u64 addr,
struct addr_location *al)
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
{
struct map_groups *mg = &self->mg;
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
al->thread = self;
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
al->addr = addr;
if (cpumode == PERF_RECORD_MISC_KERNEL) {
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
al->level = 'k';
mg = &session->kmaps;
} else if (cpumode == PERF_RECORD_MISC_USER)
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
al->level = '.';
else {
al->level = 'H';
al->map = NULL;
return;
}
try_again:
al->map = map_groups__find(mg, type, al->addr);
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
if (al->map == NULL) {
/*
* If this is outside of all known maps, and is a negative
* address, try to look it up in the kernel dso, as it might be
* a vsyscall or vdso (which executes in user-mode).
*
* XXX This is nasty, we should have a symbol list in the
* "[vdso]" dso, but for now lets use the old trick of looking
* in the whole kernel symbol list.
*/
if ((long long)al->addr < 0 && mg != &session->kmaps) {
mg = &session->kmaps;
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
goto try_again;
}
} else
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
al->addr = al->map->map_ip(al->map, al->addr);
}
void thread__find_addr_location(struct thread *self,
struct perf_session *session, u8 cpumode,
enum map_type type, u64 addr,
struct addr_location *al,
symbol_filter_t filter)
{
thread__find_addr_map(self, session, cpumode, type, addr, al);
if (al->map != NULL)
al->sym = map__find_symbol(al->map, al->addr, filter);
else
al->sym = NULL;
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
}
static void dso__calc_col_width(struct dso *self)
{
if (!symbol_conf.col_width_list_str && !symbol_conf.field_sep &&
(!symbol_conf.dso_list ||
strlist__has_entry(symbol_conf.dso_list, self->name))) {
unsigned int slen = strlen(self->name);
if (slen > dsos__col_width)
dsos__col_width = slen;
}
self->slen_calculated = 1;
}
int event__preprocess_sample(const event_t *self, struct perf_session *session,
struct addr_location *al, symbol_filter_t filter)
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
{
u8 cpumode = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
struct thread *thread = perf_session__findnew(session, self->ip.pid);
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
if (thread == NULL)
return -1;
if (symbol_conf.comm_list &&
!strlist__has_entry(symbol_conf.comm_list, thread->comm))
goto out_filtered;
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
thread__find_addr_location(thread, session, cpumode, MAP__FUNCTION,
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
self->ip.ip, al, filter);
dump_printf(" ...... dso: %s\n",
al->map ? al->map->dso->long_name :
al->level == 'H' ? "[hypervisor]" : "<not found>");
/*
* We have to do this here as we may have a dso with no symbol hit that
* has a name longer than the ones with symbols sampled.
*/
if (al->map && !sort_dso.elide && !al->map->dso->slen_calculated)
dso__calc_col_width(al->map->dso);
if (symbol_conf.dso_list &&
(!al->map || !al->map->dso ||
!(strlist__has_entry(symbol_conf.dso_list, al->map->dso->short_name) ||
(al->map->dso->short_name != al->map->dso->long_name &&
strlist__has_entry(symbol_conf.dso_list, al->map->dso->long_name)))))
goto out_filtered;
if (symbol_conf.sym_list && al->sym &&
!strlist__has_entry(symbol_conf.sym_list, al->sym->name))
goto out_filtered;
al->filtered = false;
return 0;
out_filtered:
al->filtered = true;
perf tools: Consolidate symbol resolving across all tools Now we have a very high level routine for simple tools to process IP sample events: int event__preprocess_sample(const event_t *self, struct addr_location *al, symbol_filter_t filter) It receives the event itself and will insert new threads in the global threads list and resolve the map and symbol, filling all this info into the new addr_location struct, so that tools like annotate and report can further process the event by creating hist_entries in their specific way (with or without callgraphs, etc). It in turn uses the new next layer function: void thread__find_addr_location(struct thread *self, u8 cpumode, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter) This one will, given a thread (userspace or the kernel kthread one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE too in the near future) at the given cpumode, taking vdsos into account (userspace hit, but kernel symbol) and will fill all these details in the addr_location given. Tools that need a more compact API for plain function resolution, like 'kmem', can use this other one: struct symbol *thread__find_function(struct thread *self, u64 addr, symbol_filter_t filter) So, to resolve a kernel symbol, that is all the 'kmem' tool needs, its just a matter of calling: sym = thread__find_function(kthread, addr, NULL); The 'filter' parameter is needed because we do lazy parsing/loading of ELF symtabs or /proc/kallsyms. With this we remove more code duplication all around, which is always good, huh? :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-28 02:29:23 +08:00
return 0;
}
int event__parse_sample(event_t *event, u64 type, struct sample_data *data)
{
u64 *array = event->sample.array;
if (type & PERF_SAMPLE_IP) {
data->ip = event->ip.ip;
array++;
}
if (type & PERF_SAMPLE_TID) {
u32 *p = (u32 *)array;
data->pid = p[0];
data->tid = p[1];
array++;
}
if (type & PERF_SAMPLE_TIME) {
data->time = *array;
array++;
}
if (type & PERF_SAMPLE_ADDR) {
data->addr = *array;
array++;
}
if (type & PERF_SAMPLE_ID) {
data->id = *array;
array++;
}
if (type & PERF_SAMPLE_STREAM_ID) {
data->stream_id = *array;
array++;
}
if (type & PERF_SAMPLE_CPU) {
u32 *p = (u32 *)array;
data->cpu = *p;
array++;
}
if (type & PERF_SAMPLE_PERIOD) {
data->period = *array;
array++;
}
if (type & PERF_SAMPLE_READ) {
pr_debug("PERF_SAMPLE_READ is unsuported for now\n");
return -1;
}
if (type & PERF_SAMPLE_CALLCHAIN) {
data->callchain = (struct ip_callchain *)array;
array += 1 + data->callchain->nr;
}
if (type & PERF_SAMPLE_RAW) {
u32 *p = (u32 *)array;
data->raw_size = *p;
p++;
data->raw_data = p;
}
return 0;
}