linux/tools/perf/util/probe-event.c

3533 lines
80 KiB
C
Raw Normal View History

/*
* probe-event.c : perf-probe definition to probe_events format converter
*
* Written by Masami Hiramatsu <mhiramat@redhat.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*
*/
#include <inttypes.h>
#include <sys/utsname.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <errno.h>
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <stdarg.h>
#include <limits.h>
#include <elf.h>
#include "util.h"
#include "event.h"
#include "strlist.h"
#include "strfilter.h"
#include "debug.h"
#include "cache.h"
perf probe: Support --line option to show probable source-code lines Add --line option to support showing probable source-code lines. perf probe --line SRC:LN[-LN|+NUM] or perf probe --line FUNC[:LN[-LN|+NUM]] This option shows source-code with line number if the line can be probed. Lines without line number (and blue color) means that the line can not be probed, because debuginfo doesn't have the information of those lines. The argument specifies the range of lines, "source.c:100-120" shows lines between 100th to l20th in source.c file. And "func:10+20" shows 20 lines from 10th line of func function. e.g. # ./perf probe --line kernel/sched.c:1080 <kernel/sched.c:1080> * * called with rq->lock held and irqs disabled */ static void hrtick_start(struct rq *rq, u64 delay) { struct hrtimer *timer = &rq->hrtick_timer; 1086 ktime_t time = ktime_add_ns(timer->base->get_time(), delay); hrtimer_set_expires(timer, time); 1090 if (rq == this_rq()) { 1091 hrtimer_restart(timer); 1092 } else if (!rq->hrtick_csd_pending) { 1093 __smp_call_function_single(cpu_of(rq), &rq->hrtick_csd, 1094 rq->hrtick_csd_pending = 1; If you specifying function name, this shows function-relative line number. # ./perf probe --line schedule <schedule:0> asmlinkage void __sched schedule(void) 1 { struct task_struct *prev, *next; unsigned long *switch_count; struct rq *rq; int cpu; need_resched: preempt_disable(); 9 cpu = smp_processor_id(); 10 rq = cpu_rq(cpu); 11 rcu_sched_qs(cpu); 12 prev = rq->curr; 13 switch_count = &prev->nivcsw; Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mike Galbraith <efault@gmx.de> LKML-Reference: <20100106144534.27218.77939.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 14:45:34 +00:00
#include "color.h"
#include "symbol.h"
#include "thread.h"
#include <api/fs/fs.h>
perf tools: Use __maybe_used for unused variables perf defines both __used and __unused variables to use for marking unused variables. The variable __used is defined to __attribute__((__unused__)), which contradicts the kernel definition to __attribute__((__used__)) for new gcc versions. On Android, __used is also defined in system headers and this leads to warnings like: warning: '__used__' attribute ignored __unused is not defined in the kernel and is not a standard definition. If __unused is included everywhere instead of __used, this leads to conflicts with glibc headers, since glibc has a variables with this name in its headers. The best approach is to use __maybe_unused, the definition used in the kernel for __attribute__((unused)). In this way there is only one definition in perf sources (instead of 2 definitions that point to the same thing: __used and __unused) and it works on both Linux and Android. This patch simply replaces all instances of __used and __unused with __maybe_unused. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Acked-by: Pekka Enberg <penberg@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1347315303-29906-7-git-send-email-irina.tirdea@intel.com [ committer note: fixed up conflict with a116e05 in builtin-sched.c ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2012-09-10 22:15:03 +00:00
#include "trace-event.h" /* For __maybe_unused */
#include "probe-event.h"
#include "probe-finder.h"
#include "probe-file.h"
#include "session.h"
#include "string2.h"
#include "sane_ctype.h"
#define PERFPROBE_GROUP "probe"
bool probe_event_dry_run; /* Dry run flag */
struct probe_conf probe_conf;
#define semantic_error(msg ...) pr_err("Semantic error :" msg)
int e_snprintf(char *str, size_t size, const char *format, ...)
{
int ret;
va_list ap;
va_start(ap, format);
ret = vsnprintf(str, size, format, ap);
va_end(ap);
if (ret >= (int)size)
ret = -E2BIG;
return ret;
}
static struct machine *host_machine;
2010-10-21 10:13:41 +00:00
/* Initialize symbol maps and path of vmlinux/modules */
int init_probe_symbol_maps(bool user_only)
{
int ret;
symbol_conf.sort_by_name = true;
symbol_conf.allow_aliases = true;
perf tools: Check recorded kernel version when finding vmlinux Currently vmlinux_path__init() only tries to find vmlinux file from current directory, /boot and some canonical directories with version number of the running kernel. This can be a problem when reporting old data recorded on a kernel version not running currently. We can use --symfs option for this but it's annoying for user to do it always. As we already have the info in the perf.data file, it can be changed to use it for the search automatically. Before: $ perf report ... # Samples: 4K of event 'cpu-clock' # Event count (approx.): 1067250000 # # Overhead Command Shared Object Symbol # ........ .......... ................. .............................. 71.87% swapper [kernel.kallsyms] [k] recover_probed_instruction After: # Overhead Command Shared Object Symbol # ........ .......... ................. .................... 71.87% swapper [kernel.kallsyms] [k] native_safe_halt This requires to change signature of symbol__init() to receive struct perf_session_env *. Reported-by: Minchan Kim <minchan@kernel.org> Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Namhyung Kim <namhyung.kim@lge.com> 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/1407825645-24586-14-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-08-12 06:40:45 +00:00
ret = symbol__init(NULL);
if (ret < 0) {
pr_debug("Failed to init symbol map.\n");
goto out;
}
if (host_machine || user_only) /* already initialized */
return 0;
if (symbol_conf.vmlinux_name)
pr_debug("Use vmlinux: %s\n", symbol_conf.vmlinux_name);
host_machine = machine__new_host();
if (!host_machine) {
pr_debug("machine__new_host() failed.\n");
symbol__exit();
ret = -1;
2010-10-21 10:13:41 +00:00
}
out:
if (ret < 0)
pr_warning("Failed to init vmlinux path.\n");
return ret;
}
void exit_probe_symbol_maps(void)
{
machine__delete(host_machine);
host_machine = NULL;
symbol__exit();
}
static struct ref_reloc_sym *kernel_get_ref_reloc_sym(void)
{
/* kmap->ref_reloc_sym should be set if host_machine is initialized */
struct kmap *kmap;
struct map *map = machine__kernel_map(host_machine);
if (map__load(map) < 0)
return NULL;
kmap = map__kmap(map);
if (!kmap)
return NULL;
return kmap->ref_reloc_sym;
}
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
static int kernel_get_symbol_address_by_name(const char *name, u64 *addr,
bool reloc, bool reladdr)
{
struct ref_reloc_sym *reloc_sym;
struct symbol *sym;
struct map *map;
/* ref_reloc_sym is just a label. Need a special fix*/
reloc_sym = kernel_get_ref_reloc_sym();
if (reloc_sym && strcmp(name, reloc_sym->name) == 0)
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
*addr = (reloc) ? reloc_sym->addr : reloc_sym->unrelocated_addr;
else {
sym = machine__find_kernel_symbol_by_name(host_machine, name, &map);
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
if (!sym)
return -ENOENT;
*addr = map->unmap_ip(map, sym->start) -
((reloc) ? 0 : map->reloc) -
((reladdr) ? map->start : 0);
}
return 0;
}
static struct map *kernel_get_module_map(const char *module)
{
struct maps *maps = machine__kernel_maps(host_machine);
struct map *pos;
/* A file path -- this is an offline module */
if (module && strchr(module, '/'))
return dso__new_map(module);
if (!module)
module = "kernel";
for (pos = maps__first(maps); pos; pos = map__next(pos)) {
/* short_name is "[module]" */
if (strncmp(pos->dso->short_name + 1, module,
pos->dso->short_name_len - 2) == 0 &&
module[pos->dso->short_name_len - 2] == '\0') {
map__get(pos);
return pos;
}
}
return NULL;
}
struct map *get_target_map(const char *target, struct nsinfo *nsi, bool user)
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
{
/* Init maps of given executable or kernel */
if (user) {
struct map *map;
map = dso__new_map(target);
if (map && map->dso)
map->dso->nsinfo = nsinfo__get(nsi);
return map;
} else {
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
return kernel_get_module_map(target);
}
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
}
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
static int convert_exec_to_group(const char *exec, char **result)
{
char *ptr1, *ptr2, *exec_copy;
char buf[64];
int ret;
exec_copy = strdup(exec);
if (!exec_copy)
return -ENOMEM;
ptr1 = basename(exec_copy);
if (!ptr1) {
ret = -EINVAL;
goto out;
}
for (ptr2 = ptr1; *ptr2 != '\0'; ptr2++) {
perf probe: Fix to cut off incompatible chars from group name Cut off the characters which can not use for group name of uprobes when making it based on executable filename. For example, if the exec name is libstdc++.so, without this fix perf probe generates "probe_libstdc++" as the group name, but it is failed to set because '+' can not be used for group name. With this fix perf accepts only alphabet, number or '_' for group name, thus perf generates "probe_libstdc" as the group name. E.g. with this fix, you can see the event name has no "+". ---- $ ./perf probe -x /usr/lib64/libstdc++.so.6 -D is_open p:probe_libstdc/is_open /usr/lib64/libstdc++.so.6.0.22:0x8ca80 p:probe_libstdc/is_open_1 /usr/lib64/libstdc++.so.6.0.22:0x8ca70 p:probe_libstdc/is_open_2 /usr/lib64/libstdc++.so.6.0.22:0x8ca60 p:probe_libstdc/is_open_3 /usr/lib64/libstdc++.so.6.0.22:0xb0ad0 p:probe_libstdc/is_open_4 /usr/lib64/libstdc++.so.6.0.22:0xecca9 ---- Committer note: Before this fix: # perf probe -x /usr/lib64/libstdc++.so.6 is_open Failed to write event: Invalid argument Error: Failed to add events. # After the fix: # perf probe -x /usr/lib64/libstdc++.so.6 is_open Added new events: probe_libstdc:is_open (on is_open in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_1 (on is_open in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_2 (on is_open in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_3 (on is_open in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_4 (on is_open in /usr/lib64/libstdc++.so.6.0.22) You can now use it in all perf tools, such as: perf record -e probe_libstdc:is_open_4 -aR sleep 1 # perf probe -l probe_libstdc:* probe_libstdc:is_open (on is_open@libstdc++-v3/include/fstream in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_1 (on is_open@libstdc++-v3/include/fstream in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_2 (on is_open@libstdc++-v3/include/fstream in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_3 (on is_open@src/c++98/basic_file.cc in /usr/lib64/libstdc++.so.6.0.22) probe_libstdc:is_open_4 (on stdio_filebuf:5@include/ext/stdio_filebuf.h in /usr/lib64/libstdc++.so.6.0.22) # Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Tested-by: Jiri Olsa <jolsa@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/147464491667.29804.9553638175441827970.stgit@devbox Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-09-23 15:35:16 +00:00
if (!isalnum(*ptr2) && *ptr2 != '_') {
*ptr2 = '\0';
break;
}
}
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
ret = e_snprintf(buf, 64, "%s_%s", PERFPROBE_GROUP, ptr1);
if (ret < 0)
goto out;
*result = strdup(buf);
ret = *result ? 0 : -ENOMEM;
out:
free(exec_copy);
return ret;
}
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
static void clear_perf_probe_point(struct perf_probe_point *pp)
{
free(pp->file);
free(pp->function);
free(pp->lazy_line);
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
static void clear_probe_trace_events(struct probe_trace_event *tevs, int ntevs)
{
int i;
for (i = 0; i < ntevs; i++)
clear_probe_trace_event(tevs + i);
}
static bool kprobe_blacklist__listed(unsigned long address);
static bool kprobe_warn_out_range(const char *symbol, unsigned long address)
{
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
u64 etext_addr = 0;
int ret;
/* Get the address of _etext for checking non-probable text symbol */
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
ret = kernel_get_symbol_address_by_name("_etext", &etext_addr,
false, false);
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
if (ret == 0 && etext_addr < address)
pr_warning("%s is out of .text, skip it.\n", symbol);
else if (kprobe_blacklist__listed(address))
pr_warning("%s is blacklisted function, skip it.\n", symbol);
else
return false;
return true;
}
perf probe: Fix module probe issue if no dwarf support Perf is not able to register probe in kernel module when dwarf supprt is not there(and so it goes for symtab). Perf passes full path of module where only module name is required which is causing the problem. This patch fixes this issue. Before applying patch: $ dpkg -s libdw-dev dpkg-query: package 'libdw-dev' is not installed and no information is... $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in /linux/samples/kprobes/kprobe_example.ko) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init /linux/samples/kprobes/kprobe_example.ko:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data ] $ sudo ./perf script # No output here After applying patch: $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in kprobe_example) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init kprobe_example:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data (2 samples) ] $ sudo ./perf script insmod 13990 [002] 5961.216833: probe:kprobe_init: ... insmod 13995 [002] 5962.889384: probe:kprobe_init: ... Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:40 +00:00
/*
* @module can be module name of module file path. In case of path,
* inspect elf and find out what is actual module name.
* Caller has to free mod_name after using it.
*/
static char *find_module_name(const char *module)
{
int fd;
Elf *elf;
GElf_Ehdr ehdr;
GElf_Shdr shdr;
Elf_Data *data;
Elf_Scn *sec;
char *mod_name = NULL;
int name_offset;
perf probe: Fix module probe issue if no dwarf support Perf is not able to register probe in kernel module when dwarf supprt is not there(and so it goes for symtab). Perf passes full path of module where only module name is required which is causing the problem. This patch fixes this issue. Before applying patch: $ dpkg -s libdw-dev dpkg-query: package 'libdw-dev' is not installed and no information is... $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in /linux/samples/kprobes/kprobe_example.ko) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init /linux/samples/kprobes/kprobe_example.ko:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data ] $ sudo ./perf script # No output here After applying patch: $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in kprobe_example) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init kprobe_example:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data (2 samples) ] $ sudo ./perf script insmod 13990 [002] 5961.216833: probe:kprobe_init: ... insmod 13995 [002] 5962.889384: probe:kprobe_init: ... Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:40 +00:00
fd = open(module, O_RDONLY);
if (fd < 0)
return NULL;
elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
if (elf == NULL)
goto elf_err;
if (gelf_getehdr(elf, &ehdr) == NULL)
goto ret_err;
sec = elf_section_by_name(elf, &ehdr, &shdr,
".gnu.linkonce.this_module", NULL);
if (!sec)
goto ret_err;
data = elf_getdata(sec, NULL);
if (!data || !data->d_buf)
goto ret_err;
/*
* NOTE:
* '.gnu.linkonce.this_module' section of kernel module elf directly
* maps to 'struct module' from linux/module.h. This section contains
* actual module name which will be used by kernel after loading it.
* But, we cannot use 'struct module' here since linux/module.h is not
* exposed to user-space. Offset of 'name' has remained same from long
* time, so hardcoding it here.
*/
if (ehdr.e_ident[EI_CLASS] == ELFCLASS32)
name_offset = 12;
else /* expect ELFCLASS64 by default */
name_offset = 24;
mod_name = strdup((char *)data->d_buf + name_offset);
perf probe: Fix module probe issue if no dwarf support Perf is not able to register probe in kernel module when dwarf supprt is not there(and so it goes for symtab). Perf passes full path of module where only module name is required which is causing the problem. This patch fixes this issue. Before applying patch: $ dpkg -s libdw-dev dpkg-query: package 'libdw-dev' is not installed and no information is... $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in /linux/samples/kprobes/kprobe_example.ko) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init /linux/samples/kprobes/kprobe_example.ko:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data ] $ sudo ./perf script # No output here After applying patch: $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in kprobe_example) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init kprobe_example:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data (2 samples) ] $ sudo ./perf script insmod 13990 [002] 5961.216833: probe:kprobe_init: ... insmod 13995 [002] 5962.889384: probe:kprobe_init: ... Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:40 +00:00
ret_err:
elf_end(elf);
elf_err:
close(fd);
return mod_name;
}
#ifdef HAVE_DWARF_SUPPORT
static int kernel_get_module_dso(const char *module, struct dso **pdso)
{
struct dso *dso;
struct map *map;
const char *vmlinux_name;
int ret = 0;
if (module) {
char module_name[128];
snprintf(module_name, sizeof(module_name), "[%s]", module);
map = map_groups__find_by_name(&host_machine->kmaps, module_name);
if (map) {
dso = map->dso;
goto found;
}
pr_debug("Failed to find module %s.\n", module);
return -ENOENT;
}
map = machine__kernel_map(host_machine);
dso = map->dso;
vmlinux_name = symbol_conf.vmlinux_name;
dso->load_errno = 0;
if (vmlinux_name)
ret = dso__load_vmlinux(dso, map, vmlinux_name, false);
else
ret = dso__load_vmlinux_path(dso, map);
found:
*pdso = dso;
return ret;
}
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
/*
* Some binaries like glibc have special symbols which are on the symbol
* table, but not in the debuginfo. If we can find the address of the
* symbol from map, we can translate the address back to the probe point.
*/
static int find_alternative_probe_point(struct debuginfo *dinfo,
struct perf_probe_point *pp,
struct perf_probe_point *result,
const char *target, struct nsinfo *nsi,
bool uprobes)
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
{
struct map *map = NULL;
struct symbol *sym;
u64 address = 0;
int ret = -ENOENT;
/* This can work only for function-name based one */
if (!pp->function || pp->file)
return -ENOTSUP;
map = get_target_map(target, nsi, uprobes);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
if (!map)
return -EINVAL;
/* Find the address of given function */
map__for_each_symbol_by_name(map, pp->function, sym) {
if (uprobes)
address = sym->start;
else
address = map->unmap_ip(map, sym->start) - map->reloc;
break;
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
}
if (!address) {
ret = -ENOENT;
goto out;
}
pr_debug("Symbol %s address found : %" PRIx64 "\n",
pp->function, address);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
ret = debuginfo__find_probe_point(dinfo, (unsigned long)address,
result);
if (ret <= 0)
ret = (!ret) ? -ENOENT : ret;
else {
result->offset += pp->offset;
result->line += pp->line;
result->retprobe = pp->retprobe;
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
ret = 0;
}
out:
map__put(map);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
return ret;
}
static int get_alternative_probe_event(struct debuginfo *dinfo,
struct perf_probe_event *pev,
struct perf_probe_point *tmp)
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
{
int ret;
memcpy(tmp, &pev->point, sizeof(*tmp));
memset(&pev->point, 0, sizeof(pev->point));
ret = find_alternative_probe_point(dinfo, tmp, &pev->point, pev->target,
pev->nsi, pev->uprobes);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
if (ret < 0)
memcpy(&pev->point, tmp, sizeof(*tmp));
return ret;
}
static int get_alternative_line_range(struct debuginfo *dinfo,
struct line_range *lr,
const char *target, bool user)
{
struct perf_probe_point pp = { .function = lr->function,
.file = lr->file,
.line = lr->start };
struct perf_probe_point result;
int ret, len = 0;
memset(&result, 0, sizeof(result));
if (lr->end != INT_MAX)
len = lr->end - lr->start;
ret = find_alternative_probe_point(dinfo, &pp, &result,
target, NULL, user);
if (!ret) {
lr->function = result.function;
lr->file = result.file;
lr->start = result.line;
if (lr->end != INT_MAX)
lr->end = lr->start + len;
clear_perf_probe_point(&pp);
}
return ret;
}
/* Open new debuginfo of given module */
static struct debuginfo *open_debuginfo(const char *module, struct nsinfo *nsi,
bool silent)
{
const char *path = module;
char reason[STRERR_BUFSIZE];
struct debuginfo *ret = NULL;
struct dso *dso = NULL;
struct nscookie nsc;
int err;
if (!module || !strchr(module, '/')) {
err = kernel_get_module_dso(module, &dso);
if (err < 0) {
if (!dso || dso->load_errno == 0) {
if (!str_error_r(-err, reason, STRERR_BUFSIZE))
strcpy(reason, "(unknown)");
} else
dso__strerror_load(dso, reason, STRERR_BUFSIZE);
if (!silent)
pr_err("Failed to find the path for %s: %s\n",
module ?: "kernel", reason);
return NULL;
}
path = dso->long_name;
}
nsinfo__mountns_enter(nsi, &nsc);
ret = debuginfo__new(path);
if (!ret && !silent) {
pr_warning("The %s file has no debug information.\n", path);
if (!module || !strtailcmp(path, ".ko"))
pr_warning("Rebuild with CONFIG_DEBUG_INFO=y, ");
else
pr_warning("Rebuild with -g, ");
pr_warning("or install an appropriate debuginfo package.\n");
}
nsinfo__mountns_exit(&nsc);
return ret;
}
/* For caching the last debuginfo */
static struct debuginfo *debuginfo_cache;
static char *debuginfo_cache_path;
static struct debuginfo *debuginfo_cache__open(const char *module, bool silent)
{
const char *path = module;
/* If the module is NULL, it should be the kernel. */
if (!module)
path = "kernel";
if (debuginfo_cache_path && !strcmp(debuginfo_cache_path, path))
goto out;
/* Copy module path */
free(debuginfo_cache_path);
debuginfo_cache_path = strdup(path);
if (!debuginfo_cache_path) {
debuginfo__delete(debuginfo_cache);
debuginfo_cache = NULL;
goto out;
}
debuginfo_cache = open_debuginfo(module, NULL, silent);
if (!debuginfo_cache)
zfree(&debuginfo_cache_path);
out:
return debuginfo_cache;
}
static void debuginfo_cache__exit(void)
{
debuginfo__delete(debuginfo_cache);
debuginfo_cache = NULL;
zfree(&debuginfo_cache_path);
}
static int get_text_start_address(const char *exec, unsigned long *address,
struct nsinfo *nsi)
{
Elf *elf;
GElf_Ehdr ehdr;
GElf_Shdr shdr;
int fd, ret = -ENOENT;
struct nscookie nsc;
nsinfo__mountns_enter(nsi, &nsc);
fd = open(exec, O_RDONLY);
nsinfo__mountns_exit(&nsc);
if (fd < 0)
return -errno;
elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
if (elf == NULL) {
ret = -EINVAL;
goto out_close;
}
if (gelf_getehdr(elf, &ehdr) == NULL)
goto out;
if (!elf_section_by_name(elf, &ehdr, &shdr, ".text", NULL))
goto out;
*address = shdr.sh_addr - shdr.sh_offset;
ret = 0;
out:
elf_end(elf);
out_close:
close(fd);
return ret;
}
/*
* Convert trace point to probe point with debuginfo
*/
static int find_perf_probe_point_from_dwarf(struct probe_trace_point *tp,
struct perf_probe_point *pp,
bool is_kprobe)
{
struct debuginfo *dinfo = NULL;
unsigned long stext = 0;
u64 addr = tp->address;
int ret = -ENOENT;
/* convert the address to dwarf address */
if (!is_kprobe) {
if (!addr) {
ret = -EINVAL;
goto error;
}
ret = get_text_start_address(tp->module, &stext, NULL);
if (ret < 0)
goto error;
addr += stext;
} else if (tp->symbol) {
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
/* If the module is given, this returns relative address */
ret = kernel_get_symbol_address_by_name(tp->symbol, &addr,
false, !!tp->module);
if (ret != 0)
goto error;
addr += tp->offset;
}
pr_debug("try to find information at %" PRIx64 " in %s\n", addr,
tp->module ? : "kernel");
dinfo = debuginfo_cache__open(tp->module, verbose <= 0);
if (dinfo)
ret = debuginfo__find_probe_point(dinfo,
(unsigned long)addr, pp);
else
ret = -ENOENT;
if (ret > 0) {
pp->retprobe = tp->retprobe;
return 0;
}
error:
pr_debug("Failed to find corresponding probes from debuginfo.\n");
return ret ? : -ENOENT;
}
/* Adjust symbol name and address */
static int post_process_probe_trace_point(struct probe_trace_point *tp,
struct map *map, unsigned long offs)
{
struct symbol *sym;
perf probe: Fix probe definition for inlined functions In commit 613f050d68a8 ("perf probe: Fix to probe on gcc generated functions in modules"), the offset from symbol is, incorrectly, added to the trace point address. This leads to incorrect probe trace points for inlined functions and when using relative line number on symbols. Prior this patch: $ perf probe -m nf_nat -D in_range p:probe/in_range nf_nat:in_range.isra.9+0 $ perf probe -m i40e -D i40e_clean_rx_irq p:probe/i40e_clean_rx_irq i40e:i40e_napi_poll+2212 $ perf probe -m i40e -D i40e_clean_rx_irq:16 p:probe/i40e_clean_rx_irq i40e:i40e_lan_xmit_frame+626 After: $ perf probe -m nf_nat -D in_range p:probe/in_range nf_nat:in_range.isra.9+0 $ perf probe -m i40e -D i40e_clean_rx_irq p:probe/i40e_clean_rx_irq i40e:i40e_napi_poll+1106 $ perf probe -m i40e -D i40e_clean_rx_irq:16 p:probe/i40e_clean_rx_irq i40e:i40e_napi_poll+2665 Committer testing: Using 'pfunct', a tool found in the 'dwarves' package [1], one can ask what are the functions that while not being explicitely marked as inline, were inlined by the compiler: # pfunct --cc_inlined /lib/modules/4.12.0-rc4+/kernel/drivers/net/ethernet/intel/e1000e/e1000e.ko | head __ew32 e1000_regdump e1000e_dump_ps_pages e1000_desc_unused e1000e_systim_to_hwtstamp e1000e_rx_hwtstamp e1000e_update_rdt_wa e1000e_update_tdt_wa e1000_put_txbuf e1000_consume_page Then ask 'perf probe' to produce the kprobe_tracer probe definitions for two of them: # perf probe -m e1000e -D e1000e_rx_hwtstamp p:probe/e1000e_rx_hwtstamp e1000e:e1000_receive_skb+74 # perf probe -m e1000e -D e1000_consume_page p:probe/e1000_consume_page e1000e:e1000_clean_jumbo_rx_irq+876 p:probe/e1000_consume_page_1 e1000e:e1000_clean_jumbo_rx_irq+1506 p:probe/e1000_consume_page_2 e1000e:e1000_clean_rx_irq_ps+1074 Now lets concentrate on the 'e1000_consume_page' one, that was inlined twice in e1000_clean_jumbo_rx_irq(), lets see what readelf says about the DWARF tags for that function: $ readelf -wi /lib/modules/4.12.0-rc4+/kernel/drivers/net/ethernet/intel/e1000e/e1000e.ko <SNIP> <1><13e27b>: Abbrev Number: 121 (DW_TAG_subprogram) <13e27c> DW_AT_name : (indirect string, offset: 0xa8945): e1000_clean_jumbo_rx_irq <13e287> DW_AT_low_pc : 0x17a30 <3><13e6ef>: Abbrev Number: 119 (DW_TAG_inlined_subroutine) <13e6f0> DW_AT_abstract_origin: <0x13ed2c> <13e6f4> DW_AT_low_pc : 0x17be6 <SNIP> <1><13ed2c>: Abbrev Number: 142 (DW_TAG_subprogram) <13ed2e> DW_AT_name : (indirect string, offset: 0xa54c3): e1000_consume_page So, the first time in e1000_clean_jumbo_rx_irq() where e1000_consume_page() is inlined is at PC 0x17be6, which subtracted from e1000_clean_jumbo_rx_irq()'s address, gives us the offset we should use in the probe definition: 0x17be6 - 0x17a30 = 438 but above we have 876, which is twice as much. Lets see the second inline expansion of e1000_consume_page() in e1000_clean_jumbo_rx_irq(): <3><13e86e>: Abbrev Number: 119 (DW_TAG_inlined_subroutine) <13e86f> DW_AT_abstract_origin: <0x13ed2c> <13e873> DW_AT_low_pc : 0x17d21 0x17d21 - 0x17a30 = 753 So we where adding it at twice the offset from the containing function as we should. And then after this patch: # perf probe -m e1000e -D e1000e_rx_hwtstamp p:probe/e1000e_rx_hwtstamp e1000e:e1000_receive_skb+37 # perf probe -m e1000e -D e1000_consume_page p:probe/e1000_consume_page e1000e:e1000_clean_jumbo_rx_irq+438 p:probe/e1000_consume_page_1 e1000e:e1000_clean_jumbo_rx_irq+753 p:probe/e1000_consume_page_2 e1000e:e1000_clean_jumbo_rx_irq+1353 # Which matches the two first expansions and shows that because we were doubling the offset it would spill over the next function: readelf -sw /lib/modules/4.12.0-rc4+/kernel/drivers/net/ethernet/intel/e1000e/e1000e.ko 673: 0000000000017a30 1626 FUNC LOCAL DEFAULT 2 e1000_clean_jumbo_rx_irq 674: 0000000000018090 2013 FUNC LOCAL DEFAULT 2 e1000_clean_rx_irq_ps This is the 3rd inline expansion of e1000_consume_page() in e1000_clean_jumbo_rx_irq(): <3><13ec77>: Abbrev Number: 119 (DW_TAG_inlined_subroutine) <13ec78> DW_AT_abstract_origin: <0x13ed2c> <13ec7c> DW_AT_low_pc : 0x17f79 0x17f79 - 0x17a30 = 1353 So: 0x17a30 + 2 * 1353 = 0x184c2 And: 0x184c2 - 0x18090 = 1074 Which explains the bogus third expansion for e1000_consume_page() to end up at: p:probe/e1000_consume_page_2 e1000e:e1000_clean_rx_irq_ps+1074 All fixed now :-) [1] https://git.kernel.org/pub/scm/devel/pahole/pahole.git/ Signed-off-by: Björn Töpel <bjorn.topel@intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Acked-by: Magnus Karlsson <magnus.karlsson@intel.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: stable@vger.kernel.org Fixes: 613f050d68a8 ("perf probe: Fix to probe on gcc generated functions in modules") Link: http://lkml.kernel.org/r/20170621164134.5701-1-bjorn.topel@gmail.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-21 16:41:34 +00:00
u64 addr = tp->address - offs;
sym = map__find_symbol(map, addr);
if (!sym)
return -ENOENT;
if (strcmp(sym->name, tp->symbol)) {
/* If we have no realname, use symbol for it */
if (!tp->realname)
tp->realname = tp->symbol;
else
free(tp->symbol);
tp->symbol = strdup(sym->name);
if (!tp->symbol)
return -ENOMEM;
}
tp->offset = addr - sym->start;
tp->address -= offs;
return 0;
}
/*
* Rename DWARF symbols to ELF symbols -- gcc sometimes optimizes functions
* and generate new symbols with suffixes such as .constprop.N or .isra.N
* etc. Since those symbols are not recorded in DWARF, we have to find
* correct generated symbols from offline ELF binary.
* For online kernel or uprobes we don't need this because those are
* rebased on _text, or already a section relative address.
*/
static int
post_process_offline_probe_trace_events(struct probe_trace_event *tevs,
int ntevs, const char *pathname)
{
struct map *map;
unsigned long stext = 0;
int i, ret = 0;
/* Prepare a map for offline binary */
map = dso__new_map(pathname);
if (!map || get_text_start_address(pathname, &stext, NULL) < 0) {
pr_warning("Failed to get ELF symbols for %s\n", pathname);
return -EINVAL;
}
for (i = 0; i < ntevs; i++) {
ret = post_process_probe_trace_point(&tevs[i].point,
map, stext);
if (ret < 0)
break;
}
map__put(map);
return ret;
}
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
static int add_exec_to_probe_trace_events(struct probe_trace_event *tevs,
int ntevs, const char *exec,
struct nsinfo *nsi)
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
{
int i, ret = 0;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
unsigned long stext = 0;
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
if (!exec)
return 0;
ret = get_text_start_address(exec, &stext, nsi);
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
if (ret < 0)
return ret;
for (i = 0; i < ntevs && ret >= 0; i++) {
perf probe: Do not add offset twice to uprobe address Fix perf-probe not to add offset value twice to uprobe probe address when post processing. The tevs[i].point.address struct member is the address of symbol+offset, but current perf-probe adjusts the point.address by adding the offset. As a result, the probe address becomes symbol+offset+offset. This may cause unexpected code corruption. Urgent fix is needed. Without this fix: --- # ./perf probe -x ./perf dso__load_vmlinux+4 # ./perf probe -l probe_perf:dso__load_vmlinux (on 0x000000000006d2b8) # nm ./perf.orig | grep dso__load_vmlinux\$ 000000000046d0a0 T dso__load_vmlinux --- You can see the given offset is 3 but the actual probed address is dso__load_vmlinux+8. With this fix: --- # ./perf probe -x ./perf dso__load_vmlinux+4 # ./perf probe -l probe_perf:dso__load_vmlinux (on 0x000000000006d2b4) --- Now the problem is fixed. Note: This bug is introduced by commit fb7345bbf7fad9bf72ef63a19c707970b9685812 Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140205051858.6519.27314.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-05 05:18:58 +00:00
/* point.address is the addres of point.symbol + point.offset */
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
tevs[i].point.address -= stext;
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
tevs[i].point.module = strdup(exec);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (!tevs[i].point.module) {
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
ret = -ENOMEM;
break;
}
tevs[i].uprobes = true;
}
return ret;
}
static int
post_process_module_probe_trace_events(struct probe_trace_event *tevs,
int ntevs, const char *module,
struct debuginfo *dinfo)
{
Dwarf_Addr text_offs = 0;
int i, ret = 0;
perf probe: Fix offline module name missmatch issue Perf can add a probe on kernel module which has not been loaded yet. The current implementation finds the module name from path. But if the filename is different from the actual module name then perf fails to register a probe while loading module because of mismatch in the names. For example, samples/kobject/kobject-example.ko is loaded as kobject_example. Before applying patch: $ sudo ./perf probe -m /linux/samples/kobject/kobject-example.ko foo_show Added new event: probe:foo_show (on foo_show in kobject-example) You can now use it in all perf tools, such as: perf record -e probe:foo_show -aR sleep 1 $ cat /sys/kernel/debug/tracing/kprobe_events p:probe/foo_show kobject-example:foo_show $ insmod kobject-example.ko $ lsmod Module Size Used by kobject_example 16384 0 Generate read to /sys/kernel/kobject_example/foo while recording data with below command $ sudo ./perf record -e probe:foo_show -a [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.093 MB perf.data ] $./perf report --stdio -F overhead,comm,dso,sym Error: The perf.data.old file has no samples! After applying patch: $ sudo ./perf probe -m /linux/samples/kobject/kobject-example.ko foo_show Added new event: probe:foo_show (on foo_show in kobject_example) You can now use it in all perf tools, such as: perf record -e probe:foo_show -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/foo_show kobject_example:foo_show $ insmod kobject-example.ko $ lsmod Module Size Used by kobject_example 16384 0 Generate read to /sys/kernel/kobject_example/foo while recording data with below command $ sudo ./perf record -e probe:foo_show -a [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.097 MB perf.data (8 samples) ] $ sudo ./perf report --stdio -F overhead,comm,dso,sym ... # Samples: 8 of event 'probe:foo_show' # Event count (approx.): 8 # # Overhead Command Shared Object Symbol # ........ ....... ................. ............ # 100.00% cat [kobject_example] [k] foo_show Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:41 +00:00
char *mod_name = NULL;
struct map *map;
if (!module)
return 0;
map = get_target_map(module, NULL, false);
if (!map || debuginfo__get_text_offset(dinfo, &text_offs, true) < 0) {
pr_warning("Failed to get ELF symbols for %s\n", module);
return -EINVAL;
}
mod_name = find_module_name(module);
for (i = 0; i < ntevs; i++) {
ret = post_process_probe_trace_point(&tevs[i].point,
map, (unsigned long)text_offs);
if (ret < 0)
break;
perf probe: Fix offline module name missmatch issue Perf can add a probe on kernel module which has not been loaded yet. The current implementation finds the module name from path. But if the filename is different from the actual module name then perf fails to register a probe while loading module because of mismatch in the names. For example, samples/kobject/kobject-example.ko is loaded as kobject_example. Before applying patch: $ sudo ./perf probe -m /linux/samples/kobject/kobject-example.ko foo_show Added new event: probe:foo_show (on foo_show in kobject-example) You can now use it in all perf tools, such as: perf record -e probe:foo_show -aR sleep 1 $ cat /sys/kernel/debug/tracing/kprobe_events p:probe/foo_show kobject-example:foo_show $ insmod kobject-example.ko $ lsmod Module Size Used by kobject_example 16384 0 Generate read to /sys/kernel/kobject_example/foo while recording data with below command $ sudo ./perf record -e probe:foo_show -a [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.093 MB perf.data ] $./perf report --stdio -F overhead,comm,dso,sym Error: The perf.data.old file has no samples! After applying patch: $ sudo ./perf probe -m /linux/samples/kobject/kobject-example.ko foo_show Added new event: probe:foo_show (on foo_show in kobject_example) You can now use it in all perf tools, such as: perf record -e probe:foo_show -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/foo_show kobject_example:foo_show $ insmod kobject-example.ko $ lsmod Module Size Used by kobject_example 16384 0 Generate read to /sys/kernel/kobject_example/foo while recording data with below command $ sudo ./perf record -e probe:foo_show -a [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.097 MB perf.data (8 samples) ] $ sudo ./perf report --stdio -F overhead,comm,dso,sym ... # Samples: 8 of event 'probe:foo_show' # Event count (approx.): 8 # # Overhead Command Shared Object Symbol # ........ ....... ................. ............ # 100.00% cat [kobject_example] [k] foo_show Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:41 +00:00
tevs[i].point.module =
strdup(mod_name ? mod_name : module);
if (!tevs[i].point.module) {
ret = -ENOMEM;
break;
}
}
perf probe: Fix offline module name missmatch issue Perf can add a probe on kernel module which has not been loaded yet. The current implementation finds the module name from path. But if the filename is different from the actual module name then perf fails to register a probe while loading module because of mismatch in the names. For example, samples/kobject/kobject-example.ko is loaded as kobject_example. Before applying patch: $ sudo ./perf probe -m /linux/samples/kobject/kobject-example.ko foo_show Added new event: probe:foo_show (on foo_show in kobject-example) You can now use it in all perf tools, such as: perf record -e probe:foo_show -aR sleep 1 $ cat /sys/kernel/debug/tracing/kprobe_events p:probe/foo_show kobject-example:foo_show $ insmod kobject-example.ko $ lsmod Module Size Used by kobject_example 16384 0 Generate read to /sys/kernel/kobject_example/foo while recording data with below command $ sudo ./perf record -e probe:foo_show -a [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.093 MB perf.data ] $./perf report --stdio -F overhead,comm,dso,sym Error: The perf.data.old file has no samples! After applying patch: $ sudo ./perf probe -m /linux/samples/kobject/kobject-example.ko foo_show Added new event: probe:foo_show (on foo_show in kobject_example) You can now use it in all perf tools, such as: perf record -e probe:foo_show -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/foo_show kobject_example:foo_show $ insmod kobject-example.ko $ lsmod Module Size Used by kobject_example 16384 0 Generate read to /sys/kernel/kobject_example/foo while recording data with below command $ sudo ./perf record -e probe:foo_show -a [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.097 MB perf.data (8 samples) ] $ sudo ./perf report --stdio -F overhead,comm,dso,sym ... # Samples: 8 of event 'probe:foo_show' # Event count (approx.): 8 # # Overhead Command Shared Object Symbol # ........ ....... ................. ............ # 100.00% cat [kobject_example] [k] foo_show Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:41 +00:00
free(mod_name);
map__put(map);
return ret;
}
static int
post_process_kernel_probe_trace_events(struct probe_trace_event *tevs,
int ntevs)
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
{
struct ref_reloc_sym *reloc_sym;
char *tmp;
int i, skipped = 0;
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
/* Skip post process if the target is an offline kernel */
if (symbol_conf.ignore_vmlinux_buildid)
return post_process_offline_probe_trace_events(tevs, ntevs,
symbol_conf.vmlinux_name);
reloc_sym = kernel_get_ref_reloc_sym();
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
if (!reloc_sym) {
pr_warning("Relocated base symbol is not found!\n");
return -EINVAL;
}
for (i = 0; i < ntevs; i++) {
perf kretprobes: Offset from reloc_sym if kernel supports it We indicate support for accepting sym+offset with kretprobes through a line in ftrace README. Parse the same to identify support and choose the appropriate format for kprobe_events. As an example, without this perf patch, but with the ftrace changes: naveen@ubuntu:~/linux/tools/perf$ sudo cat /sys/kernel/debug/tracing/README | grep kretprobe place (kretprobe): [<module>:]<symbol>[+<offset>]|<memaddr> naveen@ubuntu:~/linux/tools/perf$ naveen@ubuntu:~/linux/tools/perf$ sudo ./perf probe -v do_open%return probe-definition(0): do_open%return symbol:do_open file:(null) line:0 offset:0 return:1 lazy:(null) 0 arguments Looking at the vmlinux_path (8 entries long) Using /boot/vmlinux for symbols Open Debuginfo file: /boot/vmlinux Try to find probe point from debuginfo. Matched function: do_open [2d0c7d8] Probe point found: do_open+0 Matched function: do_open [35d76b5] found inline addr: 0xc0000000004ba984 Failed to find "do_open%return", because do_open is an inlined function and has no return point. An error occurred in debuginfo analysis (-22). Trying to use symbols. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: r:probe/do_open do_open+0 Writing event: r:probe/do_open_1 do_open+0 Added new events: probe:do_open (on do_open%return) probe:do_open_1 (on do_open%return) You can now use it in all perf tools, such as: perf record -e probe:do_open_1 -aR sleep 1 naveen@ubuntu:~/linux/tools/perf$ sudo cat /sys/kernel/debug/kprobes/list c000000000041370 k kretprobe_trampoline+0x0 [OPTIMIZED] c0000000004433d0 r do_open+0x0 [DISABLED] c0000000004433d0 r do_open+0x0 [DISABLED] And after this patch (and the subsequent powerpc patch): naveen@ubuntu:~/linux/tools/perf$ sudo ./perf probe -v do_open%return probe-definition(0): do_open%return symbol:do_open file:(null) line:0 offset:0 return:1 lazy:(null) 0 arguments Looking at the vmlinux_path (8 entries long) Using /boot/vmlinux for symbols Open Debuginfo file: /boot/vmlinux Try to find probe point from debuginfo. Matched function: do_open [2d0c7d8] Probe point found: do_open+0 Matched function: do_open [35d76b5] found inline addr: 0xc0000000004ba984 Failed to find "do_open%return", because do_open is an inlined function and has no return point. An error occurred in debuginfo analysis (-22). Trying to use symbols. Opening /sys/kernel/debug/tracing//README write=0 Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: r:probe/do_open _text+4469712 Writing event: r:probe/do_open_1 _text+4956248 Added new events: probe:do_open (on do_open%return) probe:do_open_1 (on do_open%return) You can now use it in all perf tools, such as: perf record -e probe:do_open_1 -aR sleep 1 naveen@ubuntu:~/linux/tools/perf$ sudo cat /sys/kernel/debug/kprobes/list c000000000041370 k kretprobe_trampoline+0x0 [OPTIMIZED] c0000000004433d0 r do_open+0x0 [DISABLED] c0000000004ba058 r do_open+0x8 [DISABLED] Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/496ef9f33c1ab16286ece9dd62aa672807aef91c.1488961018.git.naveen.n.rao@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-03-08 08:26:09 +00:00
if (!tevs[i].point.address)
continue;
if (tevs[i].point.retprobe && !kretprobe_offset_is_supported())
continue;
/* If we found a wrong one, mark it by NULL symbol */
if (kprobe_warn_out_range(tevs[i].point.symbol,
tevs[i].point.address)) {
tmp = NULL;
skipped++;
} else {
tmp = strdup(reloc_sym->name);
if (!tmp)
return -ENOMEM;
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
}
/* If we have no realname, use symbol for it */
if (!tevs[i].point.realname)
tevs[i].point.realname = tevs[i].point.symbol;
else
free(tevs[i].point.symbol);
tevs[i].point.symbol = tmp;
tevs[i].point.offset = tevs[i].point.address -
reloc_sym->unrelocated_addr;
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
}
return skipped;
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
}
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
void __weak
arch__post_process_probe_trace_events(struct perf_probe_event *pev __maybe_unused,
int ntevs __maybe_unused)
{
}
/* Post processing the probe events */
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
static int post_process_probe_trace_events(struct perf_probe_event *pev,
struct probe_trace_event *tevs,
int ntevs, const char *module,
bool uprobe, struct debuginfo *dinfo)
{
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
int ret;
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
if (uprobe)
ret = add_exec_to_probe_trace_events(tevs, ntevs, module,
pev->nsi);
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
else if (module)
/* Currently ref_reloc_sym based probe is not for drivers */
ret = post_process_module_probe_trace_events(tevs, ntevs,
module, dinfo);
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
else
ret = post_process_kernel_probe_trace_events(tevs, ntevs);
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
if (ret >= 0)
arch__post_process_probe_trace_events(pev, ntevs);
return ret;
}
/* Try to find perf_probe_event with debuginfo */
static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
struct probe_trace_event **tevs)
{
bool need_dwarf = perf_probe_event_need_dwarf(pev);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
struct perf_probe_point tmp;
struct debuginfo *dinfo;
int ntevs, ret = 0;
dinfo = open_debuginfo(pev->target, pev->nsi, !need_dwarf);
if (!dinfo) {
if (need_dwarf)
return -ENOENT;
pr_debug("Could not open debuginfo. Try to use symbols.\n");
return 0;
}
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
pr_debug("Try to find probe point from debuginfo.\n");
/* Searching trace events corresponding to a probe event */
ntevs = debuginfo__find_trace_events(dinfo, pev, tevs);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
if (ntevs == 0) { /* Not found, retry with an alternative */
ret = get_alternative_probe_event(dinfo, pev, &tmp);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
if (!ret) {
ntevs = debuginfo__find_trace_events(dinfo, pev, tevs);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
/*
* Write back to the original probe_event for
* setting appropriate (user given) event name
*/
clear_perf_probe_point(&pev->point);
memcpy(&pev->point, &tmp, sizeof(tmp));
}
}
if (ntevs > 0) { /* Succeeded to find trace events */
perf probe: Use ref_reloc_sym based address instead of the symbol name Since several local symbols can have same name (e.g. t_show), we need to use the relative address from the symbol referred by kmap->ref_reloc_sym instead of the target symbol name itself. Because the kernel address space layout randomize (kASLR) changes the absolute address of kernel symbols, we can't rely on the absolute address. Note that this works only with debuginfo. E.g. without this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 ---- OK, we have 4 different t_show()s. All functions have different arguments as below; ---- # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show t_show m=%di:u64 v=%si:u64 p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64 ---- However, all of them have been put on the *same* address. ---- # cat /sys/kernel/debug/kprobes/list ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ffffffff810d9720 k t_show+0x0 [DISABLED] ---- With this change; ---- # ./perf probe -a "t_show \$vars" Added new events: probe:t_show (on t_show with $vars) probe:t_show_1 (on t_show with $vars) probe:t_show_2 (on t_show with $vars) probe:t_show_3 (on t_show with $vars) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # cat /sys/kernel/debug/tracing/kprobe_events p:probe/t_show _stext+889880 m=%di:u64 v=%si:u64 p:probe/t_show_1 _stext+928568 m=%di:u64 v=%si:u64 t=%si:u64 p:probe/t_show_2 _stext+969512 m=%di:u64 v=%si:u64 fmt=%si:u64 p:probe/t_show_3 _stext+1001416 m=%di:u64 v=%si:u64 file=%si:u64 # cat /sys/kernel/debug/kprobes/list ffffffffb50d95e0 k t_show+0x0 [DISABLED] ffffffffb50e2d00 k t_show+0x0 [DISABLED] ffffffffb50f4990 k t_show+0x0 [DISABLED] ffffffffb50eccf0 k t_show+0x0 [DISABLED] ---- This time, each event is put in different address correctly. Note that currently this doesn't support address-based probe on modules (thus the probes on modules are symbol based), since it requires relative address probe syntax for kprobe-tracer, and it isn't implemented yet. One more note, this allows us to put events on correct address, but --list option should be updated to show correct corresponding source code. Changes from v2: - Refer kmap->ref_reloc_sym instead of "_stext". - Refer map->reloc to catch up the kASLR perf fix. Changes from v1: - Use _stext relative address instead of actual absolute address recorded in debuginfo. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053216.29635.22584.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:16 +00:00
pr_debug("Found %d probe_trace_events.\n", ntevs);
perf probe ppc64le: Fix probe location when using DWARF Powerpc has Global Entry Point and Local Entry Point for functions. LEP catches call from both the GEP and the LEP. Symbol table of ELF contains GEP and Offset from which we can calculate LEP, but debuginfo does not have LEP info. Currently, perf prioritize symbol table over dwarf to probe on LEP for ppc64le. But when user tries to probe with function parameter, we fall back to using dwarf(i.e. GEP) and when function called via LEP, probe will never hit. For example: $ objdump -d vmlinux ... do_sys_open(): c0000000002eb4a0: e8 00 4c 3c addis r2,r12,232 c0000000002eb4a4: 60 00 42 38 addi r2,r2,96 c0000000002eb4a8: a6 02 08 7c mflr r0 c0000000002eb4ac: d0 ff 41 fb std r26,-48(r1) $ sudo ./perf probe do_sys_open $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060896 filename_string=+0(%gpr4):string For second case, perf probed on GEP. So when function will be called via LEP, probe won't hit. $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.195 MB perf.data ] To resolve this issue, let's not prioritize symbol table, let perf decide what it wants to use. Perf is already converting GEP to LEP when it uses symbol table. When perf uses debuginfo, let it find LEP offset form symbol table. This way we fall back to probe on LEP for all cases. After patch: $ sudo ./perf probe 'do_sys_open filename:string' $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/do_sys_open _text+3060904 filename_string=+0(%gpr4):string $ sudo ./perf record -a -e probe:do_sys_open ls [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.197 MB perf.data (11 samples) ] Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1470723805-5081-2-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-08-09 06:23:25 +00:00
ret = post_process_probe_trace_events(pev, *tevs, ntevs,
pev->target, pev->uprobes, dinfo);
if (ret < 0 || ret == ntevs) {
pr_debug("Post processing failed or all events are skipped. (%d)\n", ret);
clear_probe_trace_events(*tevs, ntevs);
zfree(tevs);
ntevs = 0;
}
}
debuginfo__delete(dinfo);
if (ntevs == 0) { /* No error but failed to find probe point. */
pr_warning("Probe point '%s' not found.\n",
synthesize_perf_probe_point(&pev->point));
return -ENOENT;
} else if (ntevs < 0) {
/* Error path : ntevs < 0 */
pr_debug("An error occurred in debuginfo analysis (%d).\n", ntevs);
perf probe: Try to use symbol table if searching debug info failed A problem can occur in a statically linked perf when vmlinux can be found: # perf probe --add sys_epoll_pwait probe-definition(0): sys_epoll_pwait symbol:sys_epoll_pwait file:(null) line:0 offset:0 return:0 lazy:(null) 0 arguments Looking at the vmlinux_path (7 entries long) Using /lib/modules/4.2.0-rc1+/build/vmlinux for symbols Open Debuginfo file: /lib/modules/4.2.0-rc1+/build/vmlinux Try to find probe point from debuginfo. Symbol sys_epoll_pwait address found : ffffffff8122bd40 Matched function: SyS_epoll_pwait Failed to get call frame on 0xffffffff8122bd40 An error occurred in debuginfo analysis (-2). Error: Failed to add events. Reason: No such file or directory (Code: -2) The reason is caused by libdw that, if libdw is statically linked, it can't load libebl_{arch}.so reliable. In this case it is still possible to get the address from /proc/kalksyms. However, perf tries that only when libdw returns -EBADF. This patch gives it another chance to utilize symbol table, even if libdw returns an error code other than -EBADF. After applying this patch: # perf probe -nv --add sys_epoll_pwait probe-definition(0): sys_epoll_pwait symbol:sys_epoll_pwait file:(null) line:0 offset:0 return:0 lazy:(null) 0 arguments Looking at the vmlinux_path (7 entries long) Using /lib/modules/4.2.0-rc1+/build/vmlinux for symbols Open Debuginfo file: /lib/modules/4.2.0-rc1+/build/vmlinux Try to find probe point from debuginfo. Symbol sys_epoll_pwait address found : ffffffff8122bd40 Matched function: SyS_epoll_pwait Failed to get call frame on 0xffffffff8122bd40 An error occurred in debuginfo analysis (-2). Trying to use symbols. Opening /sys/kernel/debug/tracing/kprobe_events write=1 Added new event: Writing event: p:probe/sys_epoll_pwait _text+2276672 probe:sys_epoll_pwait (on sys_epoll_pwait) You can now use it in all perf tools, such as: perf record -e probe:sys_epoll_pwait -aR sleep 1 Although libdw returns an error (Failed to get call frame), perf tries symbol table and finally gets correct address. Signed-off-by: Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Brendan Gregg <brendan.d.gregg@gmail.com> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: David Ahern <dsahern@gmail.com> Cc: He Kuang <hekuang@huawei.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kaixu Xia <xiakaixu@huawei.com> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440151770-129878-2-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-21 10:09:02 +00:00
if (ntevs == -EBADF)
pr_warning("Warning: No dwarf info found in the vmlinux - "
"please rebuild kernel with CONFIG_DEBUG_INFO=y.\n");
if (!need_dwarf) {
pr_debug("Trying to use symbols.\n");
return 0;
}
}
return ntevs;
}
#define LINEBUF_SIZE 256
#define NR_ADDITIONAL_LINES 2
static int __show_one_line(FILE *fp, int l, bool skip, bool show_num)
{
char buf[LINEBUF_SIZE], sbuf[STRERR_BUFSIZE];
const char *color = show_num ? "" : PERF_COLOR_BLUE;
const char *prefix = NULL;
do {
if (fgets(buf, LINEBUF_SIZE, fp) == NULL)
goto error;
if (skip)
continue;
if (!prefix) {
prefix = show_num ? "%7d " : " ";
color_fprintf(stdout, color, prefix, l);
}
color_fprintf(stdout, color, "%s", buf);
} while (strchr(buf, '\n') == NULL);
return 1;
error:
if (ferror(fp)) {
pr_warning("File read error: %s\n",
str_error_r(errno, sbuf, sizeof(sbuf)));
return -1;
}
return 0;
}
static int _show_one_line(FILE *fp, int l, bool skip, bool show_num)
{
int rv = __show_one_line(fp, l, skip, show_num);
if (rv == 0) {
pr_warning("Source file is shorter than expected.\n");
rv = -1;
}
return rv;
}
#define show_one_line_with_num(f,l) _show_one_line(f,l,false,true)
#define show_one_line(f,l) _show_one_line(f,l,false,false)
#define skip_one_line(f,l) _show_one_line(f,l,true,false)
#define show_one_line_or_eof(f,l) __show_one_line(f,l,false,false)
/*
* Show line-range always requires debuginfo to find source file and
* line number.
*/
static int __show_line_range(struct line_range *lr, const char *module,
bool user)
{
int l = 1;
struct int_node *ln;
struct debuginfo *dinfo;
FILE *fp;
int ret;
char *tmp;
char sbuf[STRERR_BUFSIZE];
/* Search a line range */
dinfo = open_debuginfo(module, NULL, false);
if (!dinfo)
return -ENOENT;
ret = debuginfo__find_line_range(dinfo, lr);
if (!ret) { /* Not found, retry with an alternative */
ret = get_alternative_line_range(dinfo, lr, module, user);
if (!ret)
ret = debuginfo__find_line_range(dinfo, lr);
}
debuginfo__delete(dinfo);
if (ret == 0 || ret == -ENOENT) {
pr_warning("Specified source line is not found.\n");
return -ENOENT;
} else if (ret < 0) {
pr_warning("Debuginfo analysis failed.\n");
return ret;
}
/* Convert source file path */
tmp = lr->path;
ret = get_real_path(tmp, lr->comp_dir, &lr->path);
/* Free old path when new path is assigned */
if (tmp != lr->path)
free(tmp);
if (ret < 0) {
pr_warning("Failed to find source file path.\n");
return ret;
}
setup_pager();
if (lr->function)
fprintf(stdout, "<%s@%s:%d>\n", lr->function, lr->path,
lr->start - lr->offset);
else
fprintf(stdout, "<%s:%d>\n", lr->path, lr->start);
fp = fopen(lr->path, "r");
if (fp == NULL) {
pr_warning("Failed to open %s: %s\n", lr->path,
str_error_r(errno, sbuf, sizeof(sbuf)));
return -errno;
}
/* Skip to starting line number */
while (l < lr->start) {
ret = skip_one_line(fp, l++);
if (ret < 0)
goto end;
}
intlist__for_each_entry(ln, lr->line_list) {
for (; ln->i > l; l++) {
ret = show_one_line(fp, l - lr->offset);
if (ret < 0)
goto end;
}
ret = show_one_line_with_num(fp, l++ - lr->offset);
if (ret < 0)
goto end;
}
if (lr->end == INT_MAX)
lr->end = l + NR_ADDITIONAL_LINES;
while (l <= lr->end) {
ret = show_one_line_or_eof(fp, l++ - lr->offset);
if (ret <= 0)
break;
}
end:
fclose(fp);
return ret;
}
int show_line_range(struct line_range *lr, const char *module,
struct nsinfo *nsi, bool user)
{
int ret;
struct nscookie nsc;
ret = init_probe_symbol_maps(user);
if (ret < 0)
return ret;
nsinfo__mountns_enter(nsi, &nsc);
ret = __show_line_range(lr, module, user);
nsinfo__mountns_exit(&nsc);
exit_probe_symbol_maps();
return ret;
}
static int show_available_vars_at(struct debuginfo *dinfo,
struct perf_probe_event *pev,
struct strfilter *_filter)
{
char *buf;
int ret, i, nvars;
struct str_node *node;
struct variable_list *vls = NULL, *vl;
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
struct perf_probe_point tmp;
const char *var;
buf = synthesize_perf_probe_point(&pev->point);
if (!buf)
return -EINVAL;
pr_debug("Searching variables at %s\n", buf);
ret = debuginfo__find_available_vars_at(dinfo, pev, &vls);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
if (!ret) { /* Not found, retry with an alternative */
ret = get_alternative_probe_event(dinfo, pev, &tmp);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
if (!ret) {
ret = debuginfo__find_available_vars_at(dinfo, pev,
&vls);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
/* Release the old probe_point */
clear_perf_probe_point(&tmp);
}
}
if (ret <= 0) {
if (ret == 0 || ret == -ENOENT) {
pr_err("Failed to find the address of %s\n", buf);
ret = -ENOENT;
} else
pr_warning("Debuginfo analysis failed.\n");
goto end;
}
/* Some variables are found */
fprintf(stdout, "Available variables at %s\n", buf);
for (i = 0; i < ret; i++) {
vl = &vls[i];
/*
* A probe point might be converted to
* several trace points.
*/
fprintf(stdout, "\t@<%s+%lu>\n", vl->point.symbol,
vl->point.offset);
zfree(&vl->point.symbol);
nvars = 0;
if (vl->vars) {
strlist__for_each_entry(node, vl->vars) {
var = strchr(node->s, '\t') + 1;
if (strfilter__compare(_filter, var)) {
fprintf(stdout, "\t\t%s\n", node->s);
nvars++;
}
}
strlist__delete(vl->vars);
}
if (nvars == 0)
fprintf(stdout, "\t\t(No matched variables)\n");
}
free(vls);
end:
free(buf);
return ret;
}
/* Show available variables on given probe point */
int show_available_vars(struct perf_probe_event *pevs, int npevs,
struct strfilter *_filter)
{
int i, ret = 0;
struct debuginfo *dinfo;
ret = init_probe_symbol_maps(pevs->uprobes);
if (ret < 0)
return ret;
dinfo = open_debuginfo(pevs->target, pevs->nsi, false);
if (!dinfo) {
ret = -ENOENT;
goto out;
}
setup_pager();
for (i = 0; i < npevs && ret >= 0; i++)
ret = show_available_vars_at(dinfo, &pevs[i], _filter);
debuginfo__delete(dinfo);
out:
exit_probe_symbol_maps();
return ret;
}
#else /* !HAVE_DWARF_SUPPORT */
static void debuginfo_cache__exit(void)
{
}
static int
find_perf_probe_point_from_dwarf(struct probe_trace_point *tp __maybe_unused,
struct perf_probe_point *pp __maybe_unused,
bool is_kprobe __maybe_unused)
{
return -ENOSYS;
}
static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
struct probe_trace_event **tevs __maybe_unused)
{
if (perf_probe_event_need_dwarf(pev)) {
pr_warning("Debuginfo-analysis is not supported.\n");
return -ENOSYS;
}
return 0;
}
perf tools: Use __maybe_used for unused variables perf defines both __used and __unused variables to use for marking unused variables. The variable __used is defined to __attribute__((__unused__)), which contradicts the kernel definition to __attribute__((__used__)) for new gcc versions. On Android, __used is also defined in system headers and this leads to warnings like: warning: '__used__' attribute ignored __unused is not defined in the kernel and is not a standard definition. If __unused is included everywhere instead of __used, this leads to conflicts with glibc headers, since glibc has a variables with this name in its headers. The best approach is to use __maybe_unused, the definition used in the kernel for __attribute__((unused)). In this way there is only one definition in perf sources (instead of 2 definitions that point to the same thing: __used and __unused) and it works on both Linux and Android. This patch simply replaces all instances of __used and __unused with __maybe_unused. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Acked-by: Pekka Enberg <penberg@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1347315303-29906-7-git-send-email-irina.tirdea@intel.com [ committer note: fixed up conflict with a116e05 in builtin-sched.c ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2012-09-10 22:15:03 +00:00
int show_line_range(struct line_range *lr __maybe_unused,
perf probe: Do not access kallsyms when analyzing user binaries Do not access kallsyms to show available variables and show source lines in user binaries. This behavior always requires the root privilege when sysctl sets kernel.kptr_restrict=1, but we don't need it just for analyzing user binaries. Without this patch (by normal user, kptr_restrict=1): ---- $ perf probe -x ./perf -V add_cmdname Failed to init vmlinux path. Error: Failed to show vars. $ perf probe -x ./perf -L add_cmdname Failed to init vmlinux path. Error: Failed to show lines. ---- With this patch: ---- $ perf probe -x ./perf -V add_cmdname Available variables at add_cmdname @<perf_unknown_cmd_config+144> (No matched variables) @<list_commands_in_dir+160> (No matched variables) @<add_cmdname+0> char* name size_t len struct cmdnames* cmds $ perf probe -x ./perf -L add_cmdname <add_cmdname@/home/fedora/ksrc/linux-3/tools/perf/util/help.c:0> 0 void add_cmdname(struct cmdnames *cmds, const char *name, size_t len) 1 { 2 struct cmdname *ent = malloc(sizeof(*ent) + len + 1); 4 ent->len = len; 5 memcpy(ent->name, name, len); 6 ent->name[len] = 0; ... ---- Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: david lerner <dlernerdroid@gmail.com> Cc: linux-perf-user@vger.kernel.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140917084054.3722.73975.stgit@kbuild-f20.novalocal [ Added missing 'bool user' argument to the !DWARF show_line_range() stub ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-09-17 08:40:54 +00:00
const char *module __maybe_unused,
struct nsinfo *nsi __maybe_unused,
perf probe: Do not access kallsyms when analyzing user binaries Do not access kallsyms to show available variables and show source lines in user binaries. This behavior always requires the root privilege when sysctl sets kernel.kptr_restrict=1, but we don't need it just for analyzing user binaries. Without this patch (by normal user, kptr_restrict=1): ---- $ perf probe -x ./perf -V add_cmdname Failed to init vmlinux path. Error: Failed to show vars. $ perf probe -x ./perf -L add_cmdname Failed to init vmlinux path. Error: Failed to show lines. ---- With this patch: ---- $ perf probe -x ./perf -V add_cmdname Available variables at add_cmdname @<perf_unknown_cmd_config+144> (No matched variables) @<list_commands_in_dir+160> (No matched variables) @<add_cmdname+0> char* name size_t len struct cmdnames* cmds $ perf probe -x ./perf -L add_cmdname <add_cmdname@/home/fedora/ksrc/linux-3/tools/perf/util/help.c:0> 0 void add_cmdname(struct cmdnames *cmds, const char *name, size_t len) 1 { 2 struct cmdname *ent = malloc(sizeof(*ent) + len + 1); 4 ent->len = len; 5 memcpy(ent->name, name, len); 6 ent->name[len] = 0; ... ---- Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: david lerner <dlernerdroid@gmail.com> Cc: linux-perf-user@vger.kernel.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140917084054.3722.73975.stgit@kbuild-f20.novalocal [ Added missing 'bool user' argument to the !DWARF show_line_range() stub ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-09-17 08:40:54 +00:00
bool user __maybe_unused)
{
pr_warning("Debuginfo-analysis is not supported.\n");
return -ENOSYS;
}
perf tools: Use __maybe_used for unused variables perf defines both __used and __unused variables to use for marking unused variables. The variable __used is defined to __attribute__((__unused__)), which contradicts the kernel definition to __attribute__((__used__)) for new gcc versions. On Android, __used is also defined in system headers and this leads to warnings like: warning: '__used__' attribute ignored __unused is not defined in the kernel and is not a standard definition. If __unused is included everywhere instead of __used, this leads to conflicts with glibc headers, since glibc has a variables with this name in its headers. The best approach is to use __maybe_unused, the definition used in the kernel for __attribute__((unused)). In this way there is only one definition in perf sources (instead of 2 definitions that point to the same thing: __used and __unused) and it works on both Linux and Android. This patch simply replaces all instances of __used and __unused with __maybe_unused. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Acked-by: Pekka Enberg <penberg@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1347315303-29906-7-git-send-email-irina.tirdea@intel.com [ committer note: fixed up conflict with a116e05 in builtin-sched.c ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2012-09-10 22:15:03 +00:00
int show_available_vars(struct perf_probe_event *pevs __maybe_unused,
int npevs __maybe_unused,
struct strfilter *filter __maybe_unused)
{
pr_warning("Debuginfo-analysis is not supported.\n");
return -ENOSYS;
}
#endif
void line_range__clear(struct line_range *lr)
{
free(lr->function);
free(lr->file);
free(lr->path);
free(lr->comp_dir);
intlist__delete(lr->line_list);
memset(lr, 0, sizeof(*lr));
}
int line_range__init(struct line_range *lr)
{
memset(lr, 0, sizeof(*lr));
lr->line_list = intlist__new(NULL);
if (!lr->line_list)
return -ENOMEM;
else
return 0;
}
static int parse_line_num(char **ptr, int *val, const char *what)
{
const char *start = *ptr;
errno = 0;
*val = strtol(*ptr, ptr, 0);
if (errno || *ptr == start) {
semantic_error("'%s' is not a valid number.\n", what);
return -EINVAL;
}
return 0;
}
/* Check the name is good for event, group or function */
static bool is_c_func_name(const char *name)
{
if (!isalpha(*name) && *name != '_')
return false;
while (*++name != '\0') {
if (!isalpha(*name) && !isdigit(*name) && *name != '_')
return false;
}
return true;
}
/*
* Stuff 'lr' according to the line range described by 'arg'.
* The line range syntax is described by:
*
* SRC[:SLN[+NUM|-ELN]]
* FNC[@SRC][:SLN[+NUM|-ELN]]
*/
int parse_line_range_desc(const char *arg, struct line_range *lr)
perf probe: Support --line option to show probable source-code lines Add --line option to support showing probable source-code lines. perf probe --line SRC:LN[-LN|+NUM] or perf probe --line FUNC[:LN[-LN|+NUM]] This option shows source-code with line number if the line can be probed. Lines without line number (and blue color) means that the line can not be probed, because debuginfo doesn't have the information of those lines. The argument specifies the range of lines, "source.c:100-120" shows lines between 100th to l20th in source.c file. And "func:10+20" shows 20 lines from 10th line of func function. e.g. # ./perf probe --line kernel/sched.c:1080 <kernel/sched.c:1080> * * called with rq->lock held and irqs disabled */ static void hrtick_start(struct rq *rq, u64 delay) { struct hrtimer *timer = &rq->hrtick_timer; 1086 ktime_t time = ktime_add_ns(timer->base->get_time(), delay); hrtimer_set_expires(timer, time); 1090 if (rq == this_rq()) { 1091 hrtimer_restart(timer); 1092 } else if (!rq->hrtick_csd_pending) { 1093 __smp_call_function_single(cpu_of(rq), &rq->hrtick_csd, 1094 rq->hrtick_csd_pending = 1; If you specifying function name, this shows function-relative line number. # ./perf probe --line schedule <schedule:0> asmlinkage void __sched schedule(void) 1 { struct task_struct *prev, *next; unsigned long *switch_count; struct rq *rq; int cpu; need_resched: preempt_disable(); 9 cpu = smp_processor_id(); 10 rq = cpu_rq(cpu); 11 rcu_sched_qs(cpu); 12 prev = rq->curr; 13 switch_count = &prev->nivcsw; Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mike Galbraith <efault@gmx.de> LKML-Reference: <20100106144534.27218.77939.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 14:45:34 +00:00
{
char *range, *file, *name = strdup(arg);
int err;
if (!name)
return -ENOMEM;
lr->start = 0;
lr->end = INT_MAX;
range = strchr(name, ':');
if (range) {
*range++ = '\0';
err = parse_line_num(&range, &lr->start, "start line");
if (err)
goto err;
if (*range == '+' || *range == '-') {
const char c = *range++;
err = parse_line_num(&range, &lr->end, "end line");
if (err)
goto err;
if (c == '+') {
lr->end += lr->start;
/*
* Adjust the number of lines here.
* If the number of lines == 1, the
* the end of line should be equal to
* the start of line.
*/
lr->end--;
}
}
pr_debug("Line range is %d to %d\n", lr->start, lr->end);
err = -EINVAL;
if (lr->start > lr->end) {
perf probe: Support --line option to show probable source-code lines Add --line option to support showing probable source-code lines. perf probe --line SRC:LN[-LN|+NUM] or perf probe --line FUNC[:LN[-LN|+NUM]] This option shows source-code with line number if the line can be probed. Lines without line number (and blue color) means that the line can not be probed, because debuginfo doesn't have the information of those lines. The argument specifies the range of lines, "source.c:100-120" shows lines between 100th to l20th in source.c file. And "func:10+20" shows 20 lines from 10th line of func function. e.g. # ./perf probe --line kernel/sched.c:1080 <kernel/sched.c:1080> * * called with rq->lock held and irqs disabled */ static void hrtick_start(struct rq *rq, u64 delay) { struct hrtimer *timer = &rq->hrtick_timer; 1086 ktime_t time = ktime_add_ns(timer->base->get_time(), delay); hrtimer_set_expires(timer, time); 1090 if (rq == this_rq()) { 1091 hrtimer_restart(timer); 1092 } else if (!rq->hrtick_csd_pending) { 1093 __smp_call_function_single(cpu_of(rq), &rq->hrtick_csd, 1094 rq->hrtick_csd_pending = 1; If you specifying function name, this shows function-relative line number. # ./perf probe --line schedule <schedule:0> asmlinkage void __sched schedule(void) 1 { struct task_struct *prev, *next; unsigned long *switch_count; struct rq *rq; int cpu; need_resched: preempt_disable(); 9 cpu = smp_processor_id(); 10 rq = cpu_rq(cpu); 11 rcu_sched_qs(cpu); 12 prev = rq->curr; 13 switch_count = &prev->nivcsw; Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mike Galbraith <efault@gmx.de> LKML-Reference: <20100106144534.27218.77939.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 14:45:34 +00:00
semantic_error("Start line must be smaller"
" than end line.\n");
goto err;
}
if (*range != '\0') {
semantic_error("Tailing with invalid str '%s'.\n", range);
goto err;
}
}
file = strchr(name, '@');
if (file) {
*file = '\0';
lr->file = strdup(++file);
if (lr->file == NULL) {
err = -ENOMEM;
goto err;
}
lr->function = name;
} else if (strchr(name, '/') || strchr(name, '.'))
lr->file = name;
else if (is_c_func_name(name))/* We reuse it for checking funcname */
lr->function = name;
else { /* Invalid name */
semantic_error("'%s' is not a valid function name.\n", name);
err = -EINVAL;
goto err;
}
return 0;
err:
free(name);
return err;
perf probe: Support --line option to show probable source-code lines Add --line option to support showing probable source-code lines. perf probe --line SRC:LN[-LN|+NUM] or perf probe --line FUNC[:LN[-LN|+NUM]] This option shows source-code with line number if the line can be probed. Lines without line number (and blue color) means that the line can not be probed, because debuginfo doesn't have the information of those lines. The argument specifies the range of lines, "source.c:100-120" shows lines between 100th to l20th in source.c file. And "func:10+20" shows 20 lines from 10th line of func function. e.g. # ./perf probe --line kernel/sched.c:1080 <kernel/sched.c:1080> * * called with rq->lock held and irqs disabled */ static void hrtick_start(struct rq *rq, u64 delay) { struct hrtimer *timer = &rq->hrtick_timer; 1086 ktime_t time = ktime_add_ns(timer->base->get_time(), delay); hrtimer_set_expires(timer, time); 1090 if (rq == this_rq()) { 1091 hrtimer_restart(timer); 1092 } else if (!rq->hrtick_csd_pending) { 1093 __smp_call_function_single(cpu_of(rq), &rq->hrtick_csd, 1094 rq->hrtick_csd_pending = 1; If you specifying function name, this shows function-relative line number. # ./perf probe --line schedule <schedule:0> asmlinkage void __sched schedule(void) 1 { struct task_struct *prev, *next; unsigned long *switch_count; struct rq *rq; int cpu; need_resched: preempt_disable(); 9 cpu = smp_processor_id(); 10 rq = cpu_rq(cpu); 11 rcu_sched_qs(cpu); 12 prev = rq->curr; 13 switch_count = &prev->nivcsw; Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mike Galbraith <efault@gmx.de> LKML-Reference: <20100106144534.27218.77939.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-01-06 14:45:34 +00:00
}
static int parse_perf_probe_event_name(char **arg, struct perf_probe_event *pev)
{
char *ptr;
ptr = strpbrk_esc(*arg, ":");
if (ptr) {
*ptr = '\0';
if (!pev->sdt && !is_c_func_name(*arg))
goto ng_name;
pev->group = strdup_esc(*arg);
if (!pev->group)
return -ENOMEM;
*arg = ptr + 1;
} else
pev->group = NULL;
pev->event = strdup_esc(*arg);
if (pev->event == NULL)
return -ENOMEM;
if (!pev->sdt && !is_c_func_name(pev->event)) {
zfree(&pev->event);
ng_name:
zfree(&pev->group);
semantic_error("%s is bad for event name -it must "
"follow C symbol-naming rule.\n", *arg);
return -EINVAL;
}
return 0;
}
/* Parse probepoint definition. */
static int parse_perf_probe_point(char *arg, struct perf_probe_event *pev)
{
struct perf_probe_point *pp = &pev->point;
char *ptr, *tmp;
char c, nc = 0;
bool file_spec = false;
int ret;
/*
* <Syntax>
* perf probe [GRP:][EVENT=]SRC[:LN|;PTN]
* perf probe [GRP:][EVENT=]FUNC[@SRC][+OFFS|%return|:LN|;PAT]
* perf probe %[GRP:]SDT_EVENT
*/
if (!arg)
return -EINVAL;
if (is_sdt_event(arg)) {
pev->sdt = true;
if (arg[0] == '%')
arg++;
}
ptr = strpbrk_esc(arg, ";=@+%");
if (pev->sdt) {
if (ptr) {
perf probe: Support @BUILDID or @FILE suffix for SDT events Support @BUILDID or @FILE suffix for SDT events. This allows perf to add probes on SDTs/pre-cached events on given FILE or the file which has given BUILDID (also, this complements BUILDID.) For example, both gcc and libstdc++ has same SDTs as below. If you would like to add a probe on sdt_libstdcxx:catch on gcc, you can do as below. ---- # perf list sdt | tail -n 6 sdt_libstdcxx:catch@/usr/bin/gcc(0cc207fc4b27) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49) sdt_libstdcxx:rethrow@/usr/bin/gcc(0cc207fc4b27) [SDT event] sdt_libstdcxx:rethrow@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49) sdt_libstdcxx:throw@/usr/bin/gcc(0cc207fc4b27) [SDT event] sdt_libstdcxx:throw@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49) # perf probe -a %sdt_libstdcxx:catch@0cc Added new event: sdt_libstdcxx:catch (on %catch in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 ---- Committer note: Doing the full sequence of steps to get the results above: With a clean build-id cache: [root@jouet ~]# rm -rf ~/.debug/ [root@jouet ~]# perf list sdt List of pre-defined events (to be used in -e): [root@jouet ~]# No events whatsoever, then, we can add all events in gcc to the build-id cache, doing a --add + --dry-run: [root@jouet ~]# perf probe --dry-run --cache -x /usr/bin/gcc --add %sdt_libstdcxx:\* Added new events: sdt_libstdcxx:throw (on %* in /usr/bin/gcc) sdt_libstdcxx:rethrow (on %* in /usr/bin/gcc) sdt_libstdcxx:catch (on %* in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 [root@jouet ~]# It really didn't add any events, it just cached them: [root@jouet ~]# perf probe -l [root@jouet ~]# We can see that it was cached as: [root@jouet ~]# ls -la ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/ total 976 drwxr-xr-x. 2 root root 4096 Jul 13 21:47 . drwxr-xr-x. 3 root root 4096 Jul 13 21:47 .. -rwxr-xr-x. 4 root root 985912 Jun 22 18:52 elf -rw-r--r--. 1 root root 303 Jul 13 21:47 probes [root@jouet ~]# file ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/elf /root/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/elf: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, BuildID[sha1]=9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2, stripped [root@jouet ~]# cat ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/probes %sdt_libstdcxx:throw=throw p:sdt_libstdcxx/throw /usr/bin/gcc:0x71ffd %sdt_libstdcxx:rethrow=rethrow p:sdt_libstdcxx/rethrow /usr/bin/gcc:0x720b8 %sdt_libstdcxx:catch=catch p:sdt_libstdcxx/catch /usr/bin/gcc:0x7307f %sdt_libgcc:unwind=unwind p:sdt_libgcc/unwind /usr/bin/gcc:0x7eec0 #sdt_libstdcxx:*=%* [root@jouet ~]# Ok, now we can use 'perf probe' to refer to those cached entries as: Humm, nope, doing as above we end up with: [root@jouet ~]# perf probe -a %sdt_libstdcxx:catch Semantic error :* is bad for event name -it must follow C symbol-naming rule. Error: Failed to add events. [root@jouet ~]# But it worked at some point, lets try not using --dry-run: Resetting everything: # rm -rf ~/.debug/ # perf probe -d *:* # perf probe -l # perf list sdt List of pre-defined events (to be used in -e): # Ok, now it cached everything, even things we haven't asked it to (sdt_libgcc:unwind): [root@jouet ~]# perf probe -x /usr/bin/gcc --add %sdt_libstdcxx:\* Added new events: sdt_libstdcxx:throw (on %* in /usr/bin/gcc) sdt_libstdcxx:rethrow (on %* in /usr/bin/gcc) sdt_libstdcxx:catch (on %* in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 [root@jouet ~]# perf list sdt List of pre-defined events (to be used in -e): sdt_libgcc:unwind [SDT event] sdt_libstdcxx:catch [SDT event] sdt_libstdcxx:rethrow [SDT event] sdt_libstdcxx:throw [SDT event] [root@jouet ~]# And we have the events in place: [root@jouet ~]# perf probe -l sdt_libstdcxx:catch (on execute_cfa_program+1551@../../../libgcc/unwind-dw2.c in /usr/bin/gcc) sdt_libstdcxx:rethrow (on d_print_subexpr+280@libsupc++/cp-demangle.c in /usr/bin/gcc) sdt_libstdcxx:throw (on d_print_subexpr+93@libsupc++/cp-demangle.c in /usr/bin/gcc) [root@jouet ~]# And trying to use them at least has 'perf trace --event sdt*:*' working. Then, if we try to add the ones in libstdc++: [root@jouet ~]# perf probe -x /usr/lib64/libstdc++.so.6 -a %sdt_libstdcxx:\* Error: event "catch" already exists. Hint: Remove existing event by 'perf probe -d' or force duplicates by 'perf probe -f' or set 'force=yes' in BPF source. Error: Failed to add events. [root@jouet ~]# Doesn't work, dups, but at least this served to, unbeknownst to the user, add the SDT probes in /usr/lib64/libstdc++.so.6! [root@jouet ~]# perf list sdt List of pre-defined events (to be used in -e): sdt_libgcc:unwind [SDT event] sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] sdt_libstdcxx:rethrow@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:rethrow@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] sdt_libstdcxx:throw@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:throw@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] [root@jouet ~]# Now we should be able to get to the original cset comment, if we remove all SDTs events in place, not from the cache, from the kernel, where it was set up as: [root@jouet ~]# ls -la /sys/kernel/debug/tracing/events/sdt_libstdcxx/ total 0 drwxr-xr-x. 5 root root 0 Jul 13 22:00 . drwxr-xr-x. 80 root root 0 Jul 13 21:56 .. drwxr-xr-x. 2 root root 0 Jul 13 22:00 catch -rw-r--r--. 1 root root 0 Jul 13 22:00 enable -rw-r--r--. 1 root root 0 Jul 13 22:00 filter drwxr-xr-x. 2 root root 0 Jul 13 22:00 rethrow drwxr-xr-x. 2 root root 0 Jul 13 22:00 throw [root@jouet ~]# [root@jouet ~]# head -2 /sys/kernel/debug/tracing/events/sdt_libstdcxx/throw/format name: throw ID: 2059 [root@jouet ~]# Now to remove it: [root@jouet ~]# perf probe -d sdt_libstdc*:* Removed event: sdt_libstdcxx:catch Removed event: sdt_libstdcxx:rethrow Removed event: sdt_libstdcxx:throw [root@jouet ~]# Which caused: [root@jouet ~]# ls -la /sys/kernel/debug/tracing/events/sdt_libstdcxx/ ls: cannot access '/sys/kernel/debug/tracing/events/sdt_libstdcxx/': No such file or directory [root@jouet ~]# Ok, now we can do: [root@jouet ~]# perf list sdt_libstdcxx:catch List of pre-defined events (to be used in -e): sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] [root@jouet ~]# So, these are not really 'pre-defined events', i.e. we can't use them with 'perf record --event': [root@jouet ~]# perf record --event sdt_libstdcxx:catch* event syntax error: 'sdt_libstdcxx:catch*' \___ unknown tracepoint Error: File /sys/kernel/debug/tracing/events/sdt_libstdcxx/catch* not found. Hint: Perhaps this kernel misses some CONFIG_ setting to enable this feature?. <SNIP> [root@jouet ~]# To have it really pre-defined we must use perf probe to get its definition from the cache and set it up in the kernel, creating the tracepoint to _then_ use it with 'perf record --event': [root@jouet ~]# perf probe -a sdt_libstdcxx:catch Semantic error :There is non-digit char in line number. <SNIP> Oops, there is another gotcha here, we need that pesky '%' character: [root@jouet ~]# perf probe -a %sdt_libstdcxx:catch Added new events: sdt_libstdcxx:catch (on %catch in /usr/bin/gcc) sdt_libstdcxx:catch_1 (on %catch in /usr/lib64/libstdc++.so.6.0.22) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch_1 -aR sleep 1 [root@jouet ~]# But then we added _two_ events, one with the name we expected, the other one with a _ added, when doing the analysis we need to pay attention to who maps to who. And here is where we get to the point of this patch, which is to be able to disambiguate those definitions for 'catch' in the build-id cache, but first we need remove those events we just added: [root@jouet ~]# perf probe -d %sdt_libstdcxx:catch Oops, that didn't remove anything, we need to _remove_ that % char in this case: [root@jouet ~]# perf probe -d sdt_libstdcxx:catch Removed event: sdt_libstdcxx:catch And we need to remove the other event added, i.e. I forgot to add a * at the end: [root@jouet ~]# perf probe -d sdt_libstdcxx:catch* Removed event: sdt_libstdcxx:catch_1 [root@jouet ~]# Ok, disambiguating it using what is in this patch: [root@jouet ~]# perf list sdt_libstdcxx:catch List of pre-defined events (to be used in -e): sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] [root@jouet ~]# [root@jouet ~]# perf probe -a %sdt_libstdcxx:catch@9a07 Added new event: sdt_libstdcxx:catch (on %catch in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 [root@jouet ~]# perf probe -l sdt_libstdcxx:catch (on execute_cfa_program+1551@../../../libgcc/unwind-dw2.c in /usr/bin/gcc) [root@jouet ~]# Yeah, it works! But we need to try and simplify this :-) Update: Some aspects of this simplification take place in the following patches. Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com> Cc: Brendan Gregg <brendan.d.gregg@gmail.com> Cc: Hemant Kumar <hemant@linux.vnet.ibm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/146831793746.17065.13065062753978236612.stgit@devbox Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-12 10:05:37 +00:00
if (*ptr != '@') {
semantic_error("%s must be an SDT name.\n",
arg);
return -EINVAL;
}
/* This must be a target file name or build id */
tmp = build_id_cache__complement(ptr + 1);
if (tmp) {
pev->target = build_id_cache__origname(tmp);
free(tmp);
} else
pev->target = strdup_esc(ptr + 1);
perf probe: Support @BUILDID or @FILE suffix for SDT events Support @BUILDID or @FILE suffix for SDT events. This allows perf to add probes on SDTs/pre-cached events on given FILE or the file which has given BUILDID (also, this complements BUILDID.) For example, both gcc and libstdc++ has same SDTs as below. If you would like to add a probe on sdt_libstdcxx:catch on gcc, you can do as below. ---- # perf list sdt | tail -n 6 sdt_libstdcxx:catch@/usr/bin/gcc(0cc207fc4b27) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49) sdt_libstdcxx:rethrow@/usr/bin/gcc(0cc207fc4b27) [SDT event] sdt_libstdcxx:rethrow@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49) sdt_libstdcxx:throw@/usr/bin/gcc(0cc207fc4b27) [SDT event] sdt_libstdcxx:throw@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49) # perf probe -a %sdt_libstdcxx:catch@0cc Added new event: sdt_libstdcxx:catch (on %catch in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 ---- Committer note: Doing the full sequence of steps to get the results above: With a clean build-id cache: [root@jouet ~]# rm -rf ~/.debug/ [root@jouet ~]# perf list sdt List of pre-defined events (to be used in -e): [root@jouet ~]# No events whatsoever, then, we can add all events in gcc to the build-id cache, doing a --add + --dry-run: [root@jouet ~]# perf probe --dry-run --cache -x /usr/bin/gcc --add %sdt_libstdcxx:\* Added new events: sdt_libstdcxx:throw (on %* in /usr/bin/gcc) sdt_libstdcxx:rethrow (on %* in /usr/bin/gcc) sdt_libstdcxx:catch (on %* in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 [root@jouet ~]# It really didn't add any events, it just cached them: [root@jouet ~]# perf probe -l [root@jouet ~]# We can see that it was cached as: [root@jouet ~]# ls -la ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/ total 976 drwxr-xr-x. 2 root root 4096 Jul 13 21:47 . drwxr-xr-x. 3 root root 4096 Jul 13 21:47 .. -rwxr-xr-x. 4 root root 985912 Jun 22 18:52 elf -rw-r--r--. 1 root root 303 Jul 13 21:47 probes [root@jouet ~]# file ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/elf /root/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/elf: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, BuildID[sha1]=9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2, stripped [root@jouet ~]# cat ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/probes %sdt_libstdcxx:throw=throw p:sdt_libstdcxx/throw /usr/bin/gcc:0x71ffd %sdt_libstdcxx:rethrow=rethrow p:sdt_libstdcxx/rethrow /usr/bin/gcc:0x720b8 %sdt_libstdcxx:catch=catch p:sdt_libstdcxx/catch /usr/bin/gcc:0x7307f %sdt_libgcc:unwind=unwind p:sdt_libgcc/unwind /usr/bin/gcc:0x7eec0 #sdt_libstdcxx:*=%* [root@jouet ~]# Ok, now we can use 'perf probe' to refer to those cached entries as: Humm, nope, doing as above we end up with: [root@jouet ~]# perf probe -a %sdt_libstdcxx:catch Semantic error :* is bad for event name -it must follow C symbol-naming rule. Error: Failed to add events. [root@jouet ~]# But it worked at some point, lets try not using --dry-run: Resetting everything: # rm -rf ~/.debug/ # perf probe -d *:* # perf probe -l # perf list sdt List of pre-defined events (to be used in -e): # Ok, now it cached everything, even things we haven't asked it to (sdt_libgcc:unwind): [root@jouet ~]# perf probe -x /usr/bin/gcc --add %sdt_libstdcxx:\* Added new events: sdt_libstdcxx:throw (on %* in /usr/bin/gcc) sdt_libstdcxx:rethrow (on %* in /usr/bin/gcc) sdt_libstdcxx:catch (on %* in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 [root@jouet ~]# perf list sdt List of pre-defined events (to be used in -e): sdt_libgcc:unwind [SDT event] sdt_libstdcxx:catch [SDT event] sdt_libstdcxx:rethrow [SDT event] sdt_libstdcxx:throw [SDT event] [root@jouet ~]# And we have the events in place: [root@jouet ~]# perf probe -l sdt_libstdcxx:catch (on execute_cfa_program+1551@../../../libgcc/unwind-dw2.c in /usr/bin/gcc) sdt_libstdcxx:rethrow (on d_print_subexpr+280@libsupc++/cp-demangle.c in /usr/bin/gcc) sdt_libstdcxx:throw (on d_print_subexpr+93@libsupc++/cp-demangle.c in /usr/bin/gcc) [root@jouet ~]# And trying to use them at least has 'perf trace --event sdt*:*' working. Then, if we try to add the ones in libstdc++: [root@jouet ~]# perf probe -x /usr/lib64/libstdc++.so.6 -a %sdt_libstdcxx:\* Error: event "catch" already exists. Hint: Remove existing event by 'perf probe -d' or force duplicates by 'perf probe -f' or set 'force=yes' in BPF source. Error: Failed to add events. [root@jouet ~]# Doesn't work, dups, but at least this served to, unbeknownst to the user, add the SDT probes in /usr/lib64/libstdc++.so.6! [root@jouet ~]# perf list sdt List of pre-defined events (to be used in -e): sdt_libgcc:unwind [SDT event] sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] sdt_libstdcxx:rethrow@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:rethrow@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] sdt_libstdcxx:throw@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:throw@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] [root@jouet ~]# Now we should be able to get to the original cset comment, if we remove all SDTs events in place, not from the cache, from the kernel, where it was set up as: [root@jouet ~]# ls -la /sys/kernel/debug/tracing/events/sdt_libstdcxx/ total 0 drwxr-xr-x. 5 root root 0 Jul 13 22:00 . drwxr-xr-x. 80 root root 0 Jul 13 21:56 .. drwxr-xr-x. 2 root root 0 Jul 13 22:00 catch -rw-r--r--. 1 root root 0 Jul 13 22:00 enable -rw-r--r--. 1 root root 0 Jul 13 22:00 filter drwxr-xr-x. 2 root root 0 Jul 13 22:00 rethrow drwxr-xr-x. 2 root root 0 Jul 13 22:00 throw [root@jouet ~]# [root@jouet ~]# head -2 /sys/kernel/debug/tracing/events/sdt_libstdcxx/throw/format name: throw ID: 2059 [root@jouet ~]# Now to remove it: [root@jouet ~]# perf probe -d sdt_libstdc*:* Removed event: sdt_libstdcxx:catch Removed event: sdt_libstdcxx:rethrow Removed event: sdt_libstdcxx:throw [root@jouet ~]# Which caused: [root@jouet ~]# ls -la /sys/kernel/debug/tracing/events/sdt_libstdcxx/ ls: cannot access '/sys/kernel/debug/tracing/events/sdt_libstdcxx/': No such file or directory [root@jouet ~]# Ok, now we can do: [root@jouet ~]# perf list sdt_libstdcxx:catch List of pre-defined events (to be used in -e): sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] [root@jouet ~]# So, these are not really 'pre-defined events', i.e. we can't use them with 'perf record --event': [root@jouet ~]# perf record --event sdt_libstdcxx:catch* event syntax error: 'sdt_libstdcxx:catch*' \___ unknown tracepoint Error: File /sys/kernel/debug/tracing/events/sdt_libstdcxx/catch* not found. Hint: Perhaps this kernel misses some CONFIG_ setting to enable this feature?. <SNIP> [root@jouet ~]# To have it really pre-defined we must use perf probe to get its definition from the cache and set it up in the kernel, creating the tracepoint to _then_ use it with 'perf record --event': [root@jouet ~]# perf probe -a sdt_libstdcxx:catch Semantic error :There is non-digit char in line number. <SNIP> Oops, there is another gotcha here, we need that pesky '%' character: [root@jouet ~]# perf probe -a %sdt_libstdcxx:catch Added new events: sdt_libstdcxx:catch (on %catch in /usr/bin/gcc) sdt_libstdcxx:catch_1 (on %catch in /usr/lib64/libstdc++.so.6.0.22) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch_1 -aR sleep 1 [root@jouet ~]# But then we added _two_ events, one with the name we expected, the other one with a _ added, when doing the analysis we need to pay attention to who maps to who. And here is where we get to the point of this patch, which is to be able to disambiguate those definitions for 'catch' in the build-id cache, but first we need remove those events we just added: [root@jouet ~]# perf probe -d %sdt_libstdcxx:catch Oops, that didn't remove anything, we need to _remove_ that % char in this case: [root@jouet ~]# perf probe -d sdt_libstdcxx:catch Removed event: sdt_libstdcxx:catch And we need to remove the other event added, i.e. I forgot to add a * at the end: [root@jouet ~]# perf probe -d sdt_libstdcxx:catch* Removed event: sdt_libstdcxx:catch_1 [root@jouet ~]# Ok, disambiguating it using what is in this patch: [root@jouet ~]# perf list sdt_libstdcxx:catch List of pre-defined events (to be used in -e): sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event] sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event] [root@jouet ~]# [root@jouet ~]# perf probe -a %sdt_libstdcxx:catch@9a07 Added new event: sdt_libstdcxx:catch (on %catch in /usr/bin/gcc) You can now use it in all perf tools, such as: perf record -e sdt_libstdcxx:catch -aR sleep 1 [root@jouet ~]# perf probe -l sdt_libstdcxx:catch (on execute_cfa_program+1551@../../../libgcc/unwind-dw2.c in /usr/bin/gcc) [root@jouet ~]# Yeah, it works! But we need to try and simplify this :-) Update: Some aspects of this simplification take place in the following patches. Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com> Cc: Brendan Gregg <brendan.d.gregg@gmail.com> Cc: Hemant Kumar <hemant@linux.vnet.ibm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/146831793746.17065.13065062753978236612.stgit@devbox Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-12 10:05:37 +00:00
if (!pev->target)
return -ENOMEM;
*ptr = '\0';
}
ret = parse_perf_probe_event_name(&arg, pev);
if (ret == 0) {
if (asprintf(&pev->point.function, "%%%s", pev->event) < 0)
ret = -errno;
}
return ret;
}
if (ptr && *ptr == '=') { /* Event name */
*ptr = '\0';
tmp = ptr + 1;
ret = parse_perf_probe_event_name(&arg, pev);
if (ret < 0)
return ret;
arg = tmp;
}
/*
* Check arg is function or file name and copy it.
*
* We consider arg to be a file spec if and only if it satisfies
* all of the below criteria::
* - it does not include any of "+@%",
* - it includes one of ":;", and
* - it has a period '.' in the name.
*
* Otherwise, we consider arg to be a function specification.
*/
if (!strpbrk_esc(arg, "+@%")) {
ptr = strpbrk_esc(arg, ";:");
/* This is a file spec if it includes a '.' before ; or : */
if (ptr && memchr(arg, '.', ptr - arg))
file_spec = true;
}
ptr = strpbrk_esc(arg, ";:+@%");
if (ptr) {
nc = *ptr;
*ptr++ = '\0';
}
if (arg[0] == '\0')
tmp = NULL;
else {
tmp = strdup_esc(arg);
if (tmp == NULL)
return -ENOMEM;
}
if (file_spec)
pp->file = tmp;
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
else {
pp->function = tmp;
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
/*
* Keep pp->function even if this is absolute address,
* so it can mark whether abs_address is valid.
* Which make 'perf probe lib.bin 0x0' possible.
*
* Note that checking length of tmp is not needed
* because when we access tmp[1] we know tmp[0] is '0',
* so tmp[1] should always valid (but could be '\0').
*/
if (tmp && !strncmp(tmp, "0x", 2)) {
pp->abs_address = strtoul(pp->function, &tmp, 0);
if (*tmp != '\0') {
semantic_error("Invalid absolute address.\n");
return -EINVAL;
}
}
}
/* Parse other options */
while (ptr) {
arg = ptr;
c = nc;
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
if (c == ';') { /* Lazy pattern must be the last part */
pp->lazy_line = strdup(arg); /* let leave escapes */
if (pp->lazy_line == NULL)
return -ENOMEM;
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
break;
}
ptr = strpbrk_esc(arg, ";:+@%");
if (ptr) {
nc = *ptr;
*ptr++ = '\0';
}
switch (c) {
case ':': /* Line number */
pp->line = strtoul(arg, &tmp, 0);
if (*tmp != '\0') {
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
semantic_error("There is non-digit char"
" in line number.\n");
return -EINVAL;
}
break;
case '+': /* Byte offset from a symbol */
pp->offset = strtoul(arg, &tmp, 0);
if (*tmp != '\0') {
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
semantic_error("There is non-digit character"
" in offset.\n");
return -EINVAL;
}
break;
case '@': /* File name */
if (pp->file) {
semantic_error("SRC@SRC is not allowed.\n");
return -EINVAL;
}
pp->file = strdup_esc(arg);
if (pp->file == NULL)
return -ENOMEM;
break;
case '%': /* Probe places */
if (strcmp(arg, "return") == 0) {
pp->retprobe = 1;
} else { /* Others not supported yet */
semantic_error("%%%s is not supported.\n", arg);
return -ENOTSUP;
}
break;
default: /* Buggy case */
pr_err("This program has a bug at %s:%d.\n",
__FILE__, __LINE__);
return -ENOTSUP;
break;
}
}
/* Exclusion check */
if (pp->lazy_line && pp->line) {
semantic_error("Lazy pattern can't be used with"
" line number.\n");
return -EINVAL;
}
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
if (pp->lazy_line && pp->offset) {
semantic_error("Lazy pattern can't be used with offset.\n");
return -EINVAL;
}
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
if (pp->line && pp->offset) {
semantic_error("Offset can't be used with line number.\n");
return -EINVAL;
}
if (!pp->line && !pp->lazy_line && pp->file && !pp->function) {
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
semantic_error("File always requires line number or "
"lazy pattern.\n");
return -EINVAL;
}
if (pp->offset && !pp->function) {
semantic_error("Offset requires an entry function.\n");
return -EINVAL;
}
if ((pp->offset || pp->line || pp->lazy_line) && pp->retprobe) {
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
semantic_error("Offset/Line/Lazy pattern can't be used with "
"return probe.\n");
return -EINVAL;
}
pr_debug("symbol:%s file:%s line:%d offset:%lu return:%d lazy:%s\n",
perf probe: Add lazy line matching support Add lazy line matching support for specifying new probes. This also changes the syntax of perf probe a bit. Now perf probe accepts one of below probe event definitions. 1) Define event based on function name [EVENT=]FUNC[@SRC][:RLN|+OFF|%return|;PTN] [ARG ...] 2) Define event based on source file with line number [EVENT=]SRC:ALN [ARG ...] 3) Define event based on source file with lazy pattern [EVENT=]SRC;PTN [ARG ...] - New lazy matching pattern(PTN) follows ';' (semicolon). And it must be put the end of the definition. - So, @SRC is no longer the part which must be put at the end of the definition. Note that ';' (semicolon) can be interpreted as the end of a command by the shell. This means that you need to quote it. (anyway you will need to quote the lazy pattern itself too, because it may contains other sensitive characters, like '[',']' etc.). Lazy matching ------------- The lazy line matching is similar to glob matching except ignoring spaces in both of pattern and target. e.g. 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on. This provides some sort of flexibility and robustness to probe point definitions against minor code changes. (for example, actual 10th line of schedule() can be changed easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist.) Changes in v3: - Cast Dwarf_Addr to uintmax_t for printf-formats. Changes in v2: - Cast Dwarf_Addr to unsigned long long for printf-formats. Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Cc: systemtap <systemtap@sources.redhat.com> Cc: DLE <dle-develop@lists.sourceforge.net> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mike Galbraith <efault@gmx.de> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> LKML-Reference: <20100225133611.6725.45078.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-25 13:36:12 +00:00
pp->function, pp->file, pp->line, pp->offset, pp->retprobe,
pp->lazy_line);
return 0;
}
/* Parse perf-probe event argument */
static int parse_perf_probe_arg(char *str, struct perf_probe_arg *arg)
{
char *tmp, *goodname;
struct perf_probe_arg_field **fieldp;
pr_debug("parsing arg: %s into ", str);
tmp = strchr(str, '=');
if (tmp) {
arg->name = strndup(str, tmp - str);
if (arg->name == NULL)
return -ENOMEM;
pr_debug("name:%s ", arg->name);
str = tmp + 1;
}
tmp = strchr(str, ':');
if (tmp) { /* Type setting */
*tmp = '\0';
arg->type = strdup(tmp + 1);
if (arg->type == NULL)
return -ENOMEM;
pr_debug("type:%s ", arg->type);
}
tmp = strpbrk(str, "-.[");
if (!is_c_varname(str) || !tmp) {
/* A variable, register, symbol or special value */
arg->var = strdup(str);
if (arg->var == NULL)
return -ENOMEM;
pr_debug("%s\n", arg->var);
return 0;
}
/* Structure fields or array element */
arg->var = strndup(str, tmp - str);
if (arg->var == NULL)
return -ENOMEM;
goodname = arg->var;
pr_debug("%s, ", arg->var);
fieldp = &arg->field;
do {
*fieldp = zalloc(sizeof(struct perf_probe_arg_field));
if (*fieldp == NULL)
return -ENOMEM;
if (*tmp == '[') { /* Array */
str = tmp;
(*fieldp)->index = strtol(str + 1, &tmp, 0);
(*fieldp)->ref = true;
if (*tmp != ']' || tmp == str + 1) {
semantic_error("Array index must be a"
" number.\n");
return -EINVAL;
}
tmp++;
if (*tmp == '\0')
tmp = NULL;
} else { /* Structure */
if (*tmp == '.') {
str = tmp + 1;
(*fieldp)->ref = false;
} else if (tmp[1] == '>') {
str = tmp + 2;
(*fieldp)->ref = true;
} else {
semantic_error("Argument parse error: %s\n",
str);
return -EINVAL;
}
tmp = strpbrk(str, "-.[");
}
if (tmp) {
(*fieldp)->name = strndup(str, tmp - str);
if ((*fieldp)->name == NULL)
return -ENOMEM;
if (*str != '[')
goodname = (*fieldp)->name;
pr_debug("%s(%d), ", (*fieldp)->name, (*fieldp)->ref);
fieldp = &(*fieldp)->next;
}
} while (tmp);
(*fieldp)->name = strdup(str);
if ((*fieldp)->name == NULL)
return -ENOMEM;
if (*str != '[')
goodname = (*fieldp)->name;
pr_debug("%s(%d)\n", (*fieldp)->name, (*fieldp)->ref);
/* If no name is specified, set the last field name (not array index)*/
if (!arg->name) {
arg->name = strdup(goodname);
if (arg->name == NULL)
return -ENOMEM;
}
return 0;
}
/* Parse perf-probe event command */
int parse_perf_probe_command(const char *cmd, struct perf_probe_event *pev)
{
char **argv;
int argc, i, ret = 0;
argv = argv_split(cmd, &argc);
if (!argv) {
pr_debug("Failed to split arguments.\n");
return -ENOMEM;
}
if (argc - 1 > MAX_PROBE_ARGS) {
semantic_error("Too many probe arguments (%d).\n", argc - 1);
ret = -ERANGE;
goto out;
}
/* Parse probe point */
ret = parse_perf_probe_point(argv[0], pev);
if (ret < 0)
goto out;
/* Copy arguments and ensure return probe has no C argument */
pev->nargs = argc - 1;
pev->args = zalloc(sizeof(struct perf_probe_arg) * pev->nargs);
if (pev->args == NULL) {
ret = -ENOMEM;
goto out;
}
for (i = 0; i < pev->nargs && ret >= 0; i++) {
ret = parse_perf_probe_arg(argv[i + 1], &pev->args[i]);
if (ret >= 0 &&
is_c_varname(pev->args[i].var) && pev->point.retprobe) {
semantic_error("You can't specify local variable for"
" kretprobe.\n");
ret = -EINVAL;
}
}
out:
argv_free(argv);
return ret;
}
/* Returns true if *any* ARG is either C variable, $params or $vars. */
bool perf_probe_with_var(struct perf_probe_event *pev)
{
int i = 0;
for (i = 0; i < pev->nargs; i++)
if (is_c_varname(pev->args[i].var) ||
!strcmp(pev->args[i].var, PROBE_ARG_PARAMS) ||
!strcmp(pev->args[i].var, PROBE_ARG_VARS))
return true;
return false;
}
/* Return true if this perf_probe_event requires debuginfo */
bool perf_probe_event_need_dwarf(struct perf_probe_event *pev)
{
if (pev->point.file || pev->point.line || pev->point.lazy_line)
return true;
if (perf_probe_with_var(pev))
return true;
return false;
}
/* Parse probe_events event into struct probe_point */
int parse_probe_trace_command(const char *cmd, struct probe_trace_event *tev)
{
struct probe_trace_point *tp = &tev->point;
char pr;
char *p;
char *argv0_str = NULL, *fmt, *fmt1_str, *fmt2_str, *fmt3_str;
int ret, i, argc;
char **argv;
pr_debug("Parsing probe_events: %s\n", cmd);
argv = argv_split(cmd, &argc);
if (!argv) {
pr_debug("Failed to split arguments.\n");
return -ENOMEM;
}
if (argc < 2) {
semantic_error("Too few probe arguments.\n");
ret = -ERANGE;
goto out;
}
/* Scan event and group name. */
argv0_str = strdup(argv[0]);
if (argv0_str == NULL) {
ret = -ENOMEM;
goto out;
}
fmt1_str = strtok_r(argv0_str, ":", &fmt);
fmt2_str = strtok_r(NULL, "/", &fmt);
fmt3_str = strtok_r(NULL, " \t", &fmt);
if (fmt1_str == NULL || strlen(fmt1_str) != 1 || fmt2_str == NULL
|| fmt3_str == NULL) {
semantic_error("Failed to parse event name: %s\n", argv[0]);
ret = -EINVAL;
goto out;
}
pr = fmt1_str[0];
tev->group = strdup(fmt2_str);
tev->event = strdup(fmt3_str);
if (tev->group == NULL || tev->event == NULL) {
ret = -ENOMEM;
goto out;
}
pr_debug("Group:%s Event:%s probe:%c\n", tev->group, tev->event, pr);
tp->retprobe = (pr == 'r');
/* Scan module name(if there), function name and offset */
p = strchr(argv[1], ':');
if (p) {
tp->module = strndup(argv[1], p - argv[1]);
if (!tp->module) {
ret = -ENOMEM;
goto out;
}
tev->uprobes = (tp->module[0] == '/');
p++;
} else
p = argv[1];
fmt1_str = strtok_r(p, "+", &fmt);
/* only the address started with 0x */
if (fmt1_str[0] == '0') {
/*
* Fix a special case:
* if address == 0, kernel reports something like:
* p:probe_libc/abs_0 /lib/libc-2.18.so:0x (null) arg1=%ax
* Newer kernel may fix that, but we want to
* support old kernel also.
*/
if (strcmp(fmt1_str, "0x") == 0) {
if (!argv[2] || strcmp(argv[2], "(null)")) {
ret = -EINVAL;
goto out;
}
tp->address = 0;
free(argv[2]);
for (i = 2; argv[i + 1] != NULL; i++)
argv[i] = argv[i + 1];
argv[i] = NULL;
argc -= 1;
} else
tp->address = strtoul(fmt1_str, NULL, 0);
} else {
/* Only the symbol-based probe has offset */
tp->symbol = strdup(fmt1_str);
if (tp->symbol == NULL) {
ret = -ENOMEM;
goto out;
}
fmt2_str = strtok_r(NULL, "", &fmt);
if (fmt2_str == NULL)
tp->offset = 0;
else
tp->offset = strtoul(fmt2_str, NULL, 10);
}
tev->nargs = argc - 2;
tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs);
if (tev->args == NULL) {
ret = -ENOMEM;
goto out;
}
for (i = 0; i < tev->nargs; i++) {
p = strchr(argv[i + 2], '=');
if (p) /* We don't need which register is assigned. */
*p++ = '\0';
else
p = argv[i + 2];
tev->args[i].name = strdup(argv[i + 2]);
/* TODO: parse regs and offset */
tev->args[i].value = strdup(p);
if (tev->args[i].name == NULL || tev->args[i].value == NULL) {
ret = -ENOMEM;
goto out;
}
}
ret = 0;
out:
free(argv0_str);
argv_free(argv);
return ret;
}
/* Compose only probe arg */
char *synthesize_perf_probe_arg(struct perf_probe_arg *pa)
{
struct perf_probe_arg_field *field = pa->field;
struct strbuf buf;
char *ret = NULL;
int err;
if (strbuf_init(&buf, 64) < 0)
return NULL;
if (pa->name && pa->var)
err = strbuf_addf(&buf, "%s=%s", pa->name, pa->var);
else
err = strbuf_addstr(&buf, pa->name ?: pa->var);
if (err)
goto out;
while (field) {
if (field->name[0] == '[')
err = strbuf_addstr(&buf, field->name);
else
err = strbuf_addf(&buf, "%s%s", field->ref ? "->" : ".",
field->name);
field = field->next;
if (err)
goto out;
}
if (pa->type)
if (strbuf_addf(&buf, ":%s", pa->type) < 0)
goto out;
ret = strbuf_detach(&buf, NULL);
out:
strbuf_release(&buf);
return ret;
}
/* Compose only probe point (not argument) */
char *synthesize_perf_probe_point(struct perf_probe_point *pp)
{
struct strbuf buf;
char *tmp, *ret = NULL;
int len, err = 0;
if (strbuf_init(&buf, 64) < 0)
return NULL;
if (pp->function) {
if (strbuf_addstr(&buf, pp->function) < 0)
goto out;
if (pp->offset)
err = strbuf_addf(&buf, "+%lu", pp->offset);
else if (pp->line)
err = strbuf_addf(&buf, ":%d", pp->line);
else if (pp->retprobe)
err = strbuf_addstr(&buf, "%return");
if (err)
goto out;
}
if (pp->file) {
tmp = pp->file;
len = strlen(tmp);
if (len > 30) {
tmp = strchr(pp->file + len - 30, '/');
tmp = tmp ? tmp + 1 : pp->file + len - 30;
}
err = strbuf_addf(&buf, "@%s", tmp);
if (!err && !pp->function && pp->line)
err = strbuf_addf(&buf, ":%d", pp->line);
}
if (!err)
ret = strbuf_detach(&buf, NULL);
out:
strbuf_release(&buf);
return ret;
}
char *synthesize_perf_probe_command(struct perf_probe_event *pev)
{
struct strbuf buf;
char *tmp, *ret = NULL;
int i;
if (strbuf_init(&buf, 64))
return NULL;
if (pev->event)
if (strbuf_addf(&buf, "%s:%s=", pev->group ?: PERFPROBE_GROUP,
pev->event) < 0)
goto out;
tmp = synthesize_perf_probe_point(&pev->point);
if (!tmp || strbuf_addstr(&buf, tmp) < 0)
goto out;
free(tmp);
for (i = 0; i < pev->nargs; i++) {
tmp = synthesize_perf_probe_arg(pev->args + i);
if (!tmp || strbuf_addf(&buf, " %s", tmp) < 0)
goto out;
free(tmp);
}
ret = strbuf_detach(&buf, NULL);
out:
strbuf_release(&buf);
return ret;
}
static int __synthesize_probe_trace_arg_ref(struct probe_trace_arg_ref *ref,
struct strbuf *buf, int depth)
{
int err;
if (ref->next) {
depth = __synthesize_probe_trace_arg_ref(ref->next, buf,
depth + 1);
if (depth < 0)
return depth;
}
err = strbuf_addf(buf, "%+ld(", ref->offset);
return (err < 0) ? err : depth;
}
static int synthesize_probe_trace_arg(struct probe_trace_arg *arg,
struct strbuf *buf)
{
struct probe_trace_arg_ref *ref = arg->ref;
int depth = 0, err;
/* Argument name or separator */
if (arg->name)
err = strbuf_addf(buf, " %s=", arg->name);
else
err = strbuf_addch(buf, ' ');
if (err)
return err;
/* Special case: @XXX */
if (arg->value[0] == '@' && arg->ref)
ref = ref->next;
/* Dereferencing arguments */
if (ref) {
depth = __synthesize_probe_trace_arg_ref(ref, buf, 1);
if (depth < 0)
return depth;
}
/* Print argument value */
if (arg->value[0] == '@' && arg->ref)
err = strbuf_addf(buf, "%s%+ld", arg->value, arg->ref->offset);
else
err = strbuf_addstr(buf, arg->value);
/* Closing */
while (!err && depth--)
err = strbuf_addch(buf, ')');
/* Print argument type */
if (!err && arg->type)
err = strbuf_addf(buf, ":%s", arg->type);
return err;
}
char *synthesize_probe_trace_command(struct probe_trace_event *tev)
{
struct probe_trace_point *tp = &tev->point;
struct strbuf buf;
char *ret = NULL;
int i, err;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
/* Uprobes must have tp->module */
if (tev->uprobes && !tp->module)
return NULL;
if (strbuf_init(&buf, 32) < 0)
return NULL;
if (strbuf_addf(&buf, "%c:%s/%s ", tp->retprobe ? 'r' : 'p',
tev->group, tev->event) < 0)
goto error;
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
/*
* If tp->address == 0, then this point must be a
* absolute address uprobe.
* try_to_find_absolute_address() should have made
* tp->symbol to "0x0".
*/
if (tev->uprobes && !tp->address) {
if (!tp->symbol || strcmp(tp->symbol, "0x0"))
goto error;
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
/* Use the tp->address for uprobes */
if (tev->uprobes)
err = strbuf_addf(&buf, "%s:0x%lx", tp->module, tp->address);
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
else if (!strncmp(tp->symbol, "0x", 2))
/* Absolute address. See try_to_find_absolute_address() */
err = strbuf_addf(&buf, "%s%s0x%lx", tp->module ?: "",
tp->module ? ":" : "", tp->address);
else
err = strbuf_addf(&buf, "%s%s%s+%lu", tp->module ?: "",
tp->module ? ":" : "", tp->symbol, tp->offset);
if (err)
goto error;
for (i = 0; i < tev->nargs; i++)
if (synthesize_probe_trace_arg(&tev->args[i], &buf) < 0)
goto error;
ret = strbuf_detach(&buf, NULL);
error:
strbuf_release(&buf);
return ret;
}
static int find_perf_probe_point_from_map(struct probe_trace_point *tp,
struct perf_probe_point *pp,
bool is_kprobe)
{
struct symbol *sym = NULL;
struct map *map = NULL;
u64 addr = tp->address;
int ret = -ENOENT;
if (!is_kprobe) {
map = dso__new_map(tp->module);
if (!map)
goto out;
sym = map__find_symbol(map, addr);
} else {
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
if (tp->symbol && !addr) {
if (kernel_get_symbol_address_by_name(tp->symbol,
&addr, true, false) < 0)
perf probe: Show correct source lines of probes on kmodules Perf probe always failed to find appropriate line numbers because of failing to find .text start address offset from debuginfo. e.g. ---- # ./perf probe -m pcspkr pcspkr_event:5 Added new events: probe:pcspkr_event (on pcspkr_event:5 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5 in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event_1 -aR sleep 1 # ./perf probe -l Failed to find debug information for address ffffffffa031f006 Failed to find debug information for address ffffffffa031f016 probe:pcspkr_event (on pcspkr_event+6 in pcspkr) probe:pcspkr_event_1 (on pcspkr_event+22 in pcspkr) ---- This fixes the above issue as below. 1. Get the relative address of the symbol in .text by using map->start. 2. Adjust the address by adding the offset of .text section in the kernel module binary. With this fix, perf probe -l shows lines correctly. ---- # ./perf probe -l probe:pcspkr_event (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) probe:pcspkr_event_1 (on pcspkr_event:5@drivers/input/misc/pcspkr.c in pcspkr) ---- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/20150930164132.3733.24643.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-09-30 16:41:33 +00:00
goto out;
}
if (addr) {
addr += tp->offset;
sym = machine__find_kernel_symbol(host_machine, addr, &map);
}
}
if (!sym)
goto out;
pp->retprobe = tp->retprobe;
pp->offset = addr - map->unmap_ip(map, sym->start);
pp->function = strdup(sym->name);
ret = pp->function ? 0 : -ENOMEM;
out:
if (map && !is_kprobe) {
map__put(map);
}
return ret;
}
static int convert_to_perf_probe_point(struct probe_trace_point *tp,
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
struct perf_probe_point *pp,
bool is_kprobe)
{
char buf[128];
int ret;
ret = find_perf_probe_point_from_dwarf(tp, pp, is_kprobe);
if (!ret)
return 0;
ret = find_perf_probe_point_from_map(tp, pp, is_kprobe);
if (!ret)
return 0;
pr_debug("Failed to find probe point from both of dwarf and map.\n");
if (tp->symbol) {
pp->function = strdup(tp->symbol);
pp->offset = tp->offset;
} else {
ret = e_snprintf(buf, 128, "0x%" PRIx64, (u64)tp->address);
if (ret < 0)
return ret;
pp->function = strdup(buf);
pp->offset = 0;
}
if (pp->function == NULL)
return -ENOMEM;
pp->retprobe = tp->retprobe;
return 0;
}
static int convert_to_perf_probe_event(struct probe_trace_event *tev,
struct perf_probe_event *pev, bool is_kprobe)
{
struct strbuf buf = STRBUF_INIT;
int i, ret;
/* Convert event/group name */
pev->event = strdup(tev->event);
pev->group = strdup(tev->group);
if (pev->event == NULL || pev->group == NULL)
return -ENOMEM;
/* Convert trace_point to probe_point */
ret = convert_to_perf_probe_point(&tev->point, &pev->point, is_kprobe);
if (ret < 0)
return ret;
/* Convert trace_arg to probe_arg */
pev->nargs = tev->nargs;
pev->args = zalloc(sizeof(struct perf_probe_arg) * pev->nargs);
if (pev->args == NULL)
return -ENOMEM;
for (i = 0; i < tev->nargs && ret >= 0; i++) {
if (tev->args[i].name)
pev->args[i].name = strdup(tev->args[i].name);
else {
if ((ret = strbuf_init(&buf, 32)) < 0)
goto error;
ret = synthesize_probe_trace_arg(&tev->args[i], &buf);
pev->args[i].name = strbuf_detach(&buf, NULL);
}
if (pev->args[i].name == NULL && ret >= 0)
ret = -ENOMEM;
}
error:
if (ret < 0)
clear_perf_probe_event(pev);
return ret;
}
void clear_perf_probe_event(struct perf_probe_event *pev)
{
struct perf_probe_arg_field *field, *next;
int i;
free(pev->event);
free(pev->group);
perf probe: Support multiple probes on different binaries Support multiple probes on different binaries with just one command. In the result, this example sets up the probes on icmp_rcv in kernel, on main and set_target in perf, and on pcspkr_event in pcspker.ko driver. ----- # perf probe -a icmp_rcv -x ./perf -a main -a set_target \ -m /lib/modules/4.0.0-rc5+/kernel/drivers/input/misc/pcspkr.ko \ -a pcspkr_event Added new event: probe:icmp_rcv (on icmp_rcv) You can now use it in all perf tools, such as: perf record -e probe:icmp_rcv -aR sleep 1 Added new event: probe_perf:main (on main in /home/mhiramat/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:main -aR sleep 1 Added new event: probe_perf:set_target (on set_target in /home/mhiramat/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:set_target -aR sleep 1 Added new event: probe:pcspkr_event (on pcspkr_event in pcspkr) You can now use it in all perf tools, such as: perf record -e probe:pcspkr_event -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@infradead.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150401102539.17137.46454.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-04-01 10:25:39 +00:00
free(pev->target);
perf probe: Fix to handle aliased symbols in glibc Fix perf probe to handle aliased symbols correctly in glibc. In the glibc, several symbols are defined as an alias of __libc_XXX, e.g. malloc is an alias of __libc_malloc. In such cases, dwarf has no subroutine instances of the alias functions (e.g. no "malloc" instance), but the map has that symbol and its address. Thus, if we search the alieased symbol in debuginfo, we always fail to find it, but it is in the map. To solve this problem, this fails back to address-based alternative search, which searches the symbol in the map, translates its address to alternative (correct) function name by using debuginfo, and retry to find the alternative function point from debuginfo. This adds fail-back process to --vars, --lines and --add options. So, now you can use those on malloc@libc :) Without this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Failed to find the address of malloc Error: Failed to show vars. # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Probe point 'malloc' not found in debuginfo. Error: Failed to add events. ----- With this patch; ----- # ./perf probe -x /usr/lib64/libc-2.17.so -V malloc Available variables at malloc @<__libc_malloc+0> size_t bytes # ./perf probe -x /usr/lib64/libc-2.17.so -a "malloc bytes" Added new event: probe_libc:malloc (on malloc in /usr/lib64/libc-2.17.so with bytes) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc -aR sleep 1 ----- Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naohiro Aota <naota@elisp.net> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20150306073120.6904.13779.stgit@localhost.localdomain Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-06 07:31:20 +00:00
clear_perf_probe_point(&pev->point);
for (i = 0; i < pev->nargs; i++) {
free(pev->args[i].name);
free(pev->args[i].var);
free(pev->args[i].type);
field = pev->args[i].field;
while (field) {
next = field->next;
zfree(&field->name);
free(field);
field = next;
}
}
free(pev->args);
memset(pev, 0, sizeof(*pev));
}
#define strdup_or_goto(str, label) \
({ char *__p = NULL; if (str && !(__p = strdup(str))) goto label; __p; })
static int perf_probe_point__copy(struct perf_probe_point *dst,
struct perf_probe_point *src)
{
dst->file = strdup_or_goto(src->file, out_err);
dst->function = strdup_or_goto(src->function, out_err);
dst->lazy_line = strdup_or_goto(src->lazy_line, out_err);
dst->line = src->line;
dst->retprobe = src->retprobe;
dst->offset = src->offset;
return 0;
out_err:
clear_perf_probe_point(dst);
return -ENOMEM;
}
static int perf_probe_arg__copy(struct perf_probe_arg *dst,
struct perf_probe_arg *src)
{
struct perf_probe_arg_field *field, **ppfield;
dst->name = strdup_or_goto(src->name, out_err);
dst->var = strdup_or_goto(src->var, out_err);
dst->type = strdup_or_goto(src->type, out_err);
field = src->field;
ppfield = &(dst->field);
while (field) {
*ppfield = zalloc(sizeof(*field));
if (!*ppfield)
goto out_err;
(*ppfield)->name = strdup_or_goto(field->name, out_err);
(*ppfield)->index = field->index;
(*ppfield)->ref = field->ref;
field = field->next;
ppfield = &((*ppfield)->next);
}
return 0;
out_err:
return -ENOMEM;
}
int perf_probe_event__copy(struct perf_probe_event *dst,
struct perf_probe_event *src)
{
int i;
dst->event = strdup_or_goto(src->event, out_err);
dst->group = strdup_or_goto(src->group, out_err);
dst->target = strdup_or_goto(src->target, out_err);
dst->uprobes = src->uprobes;
if (perf_probe_point__copy(&dst->point, &src->point) < 0)
goto out_err;
dst->args = zalloc(sizeof(struct perf_probe_arg) * src->nargs);
if (!dst->args)
goto out_err;
dst->nargs = src->nargs;
for (i = 0; i < src->nargs; i++)
if (perf_probe_arg__copy(&dst->args[i], &src->args[i]) < 0)
goto out_err;
return 0;
out_err:
clear_perf_probe_event(dst);
return -ENOMEM;
}
void clear_probe_trace_event(struct probe_trace_event *tev)
{
struct probe_trace_arg_ref *ref, *next;
int i;
free(tev->event);
free(tev->group);
free(tev->point.symbol);
free(tev->point.realname);
free(tev->point.module);
for (i = 0; i < tev->nargs; i++) {
free(tev->args[i].name);
free(tev->args[i].value);
free(tev->args[i].type);
ref = tev->args[i].ref;
while (ref) {
next = ref->next;
free(ref);
ref = next;
}
}
free(tev->args);
memset(tev, 0, sizeof(*tev));
}
struct kprobe_blacklist_node {
struct list_head list;
unsigned long start;
unsigned long end;
char *symbol;
};
static void kprobe_blacklist__delete(struct list_head *blacklist)
{
struct kprobe_blacklist_node *node;
while (!list_empty(blacklist)) {
node = list_first_entry(blacklist,
struct kprobe_blacklist_node, list);
list_del(&node->list);
free(node->symbol);
free(node);
}
}
static int kprobe_blacklist__load(struct list_head *blacklist)
{
struct kprobe_blacklist_node *node;
const char *__debugfs = debugfs__mountpoint();
char buf[PATH_MAX], *p;
FILE *fp;
int ret;
if (__debugfs == NULL)
return -ENOTSUP;
ret = e_snprintf(buf, PATH_MAX, "%s/kprobes/blacklist", __debugfs);
if (ret < 0)
return ret;
fp = fopen(buf, "r");
if (!fp)
return -errno;
ret = 0;
while (fgets(buf, PATH_MAX, fp)) {
node = zalloc(sizeof(*node));
if (!node) {
ret = -ENOMEM;
break;
}
INIT_LIST_HEAD(&node->list);
list_add_tail(&node->list, blacklist);
if (sscanf(buf, "0x%lx-0x%lx", &node->start, &node->end) != 2) {
ret = -EINVAL;
break;
}
p = strchr(buf, '\t');
if (p) {
p++;
if (p[strlen(p) - 1] == '\n')
p[strlen(p) - 1] = '\0';
} else
p = (char *)"unknown";
node->symbol = strdup(p);
if (!node->symbol) {
ret = -ENOMEM;
break;
}
pr_debug2("Blacklist: 0x%lx-0x%lx, %s\n",
node->start, node->end, node->symbol);
ret++;
}
if (ret < 0)
kprobe_blacklist__delete(blacklist);
fclose(fp);
return ret;
}
static struct kprobe_blacklist_node *
kprobe_blacklist__find_by_address(struct list_head *blacklist,
unsigned long address)
{
struct kprobe_blacklist_node *node;
list_for_each_entry(node, blacklist, list) {
perf probe: Fix kprobe blacklist checking condition The commit 9aaf5a5f479b ("perf probe: Check kprobes blacklist when adding new events"), 'perf probe' supports checking the blacklist of the fuctions which can not be probed. But the checking condition is wrong, that the end_addr of the symbol which is the start_addr of the next symbol can't be included. Committer notes: IOW make it match its kernel counterpart in kernel/kprobes.c: bool within_kprobe_blacklist(unsigned long addr) Each entry have as its end address not its end address, but the first address _outside_ that symbol, which for related functions, is the first address of the next symbol, like these from kernel/trace/trace_probe.c: 0xffffffffbd198df0-0xffffffffbd198e40 print_type_u8 0xffffffffbd198e40-0xffffffffbd198e90 print_type_u16 0xffffffffbd198e90-0xffffffffbd198ee0 print_type_u32 0xffffffffbd198ee0-0xffffffffbd198f30 print_type_u64 0xffffffffbd198f30-0xffffffffbd198f80 print_type_s8 0xffffffffbd198f80-0xffffffffbd198fd0 print_type_s16 0xffffffffbd198fd0-0xffffffffbd199020 print_type_s32 0xffffffffbd199020-0xffffffffbd199070 print_type_s64 0xffffffffbd199070-0xffffffffbd1990c0 print_type_x8 0xffffffffbd1990c0-0xffffffffbd199110 print_type_x16 0xffffffffbd199110-0xffffffffbd199160 print_type_x32 0xffffffffbd199160-0xffffffffbd1991b0 print_type_x64 But not always: 0xffffffffbd1997b0-0xffffffffbd1997c0 fetch_kernel_stack_address (kernel/trace/trace_probe.c) 0xffffffffbd1c57f0-0xffffffffbd1c58b0 __context_tracking_enter (kernel/context_tracking.c) Signed-off-by: Li Bin <huawei.libin@huawei.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Wang Nan <wangnan0@huawei.com> Cc: zhangmengting@huawei.com Fixes: 9aaf5a5f479b ("perf probe: Check kprobes blacklist when adding new events") Link: http://lkml.kernel.org/r/1504011443-7269-1-git-send-email-huawei.libin@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-08-29 12:57:23 +00:00
if (node->start <= address && address < node->end)
return node;
}
return NULL;
}
static LIST_HEAD(kprobe_blacklist);
static void kprobe_blacklist__init(void)
{
if (!list_empty(&kprobe_blacklist))
return;
if (kprobe_blacklist__load(&kprobe_blacklist) < 0)
pr_debug("No kprobe blacklist support, ignored\n");
}
static void kprobe_blacklist__release(void)
{
kprobe_blacklist__delete(&kprobe_blacklist);
}
static bool kprobe_blacklist__listed(unsigned long address)
{
return !!kprobe_blacklist__find_by_address(&kprobe_blacklist, address);
}
2015-06-16 11:50:57 +00:00
static int perf_probe_event__sprintf(const char *group, const char *event,
struct perf_probe_event *pev,
const char *module,
struct strbuf *result)
{
int i, ret;
char *buf;
if (asprintf(&buf, "%s:%s", group, event) < 0)
return -errno;
ret = strbuf_addf(result, " %-20s (on ", buf);
free(buf);
if (ret)
return ret;
/* Synthesize only event probe point */
buf = synthesize_perf_probe_point(&pev->point);
if (!buf)
return -ENOMEM;
ret = strbuf_addstr(result, buf);
free(buf);
if (!ret && module)
ret = strbuf_addf(result, " in %s", module);
if (!ret && pev->nargs > 0) {
ret = strbuf_add(result, " with", 5);
for (i = 0; !ret && i < pev->nargs; i++) {
buf = synthesize_perf_probe_arg(&pev->args[i]);
if (!buf)
return -ENOMEM;
ret = strbuf_addf(result, " %s", buf);
free(buf);
}
}
if (!ret)
ret = strbuf_addch(result, ')');
return ret;
}
/* Show an event */
int show_perf_probe_event(const char *group, const char *event,
struct perf_probe_event *pev,
const char *module, bool use_stdout)
{
struct strbuf buf = STRBUF_INIT;
int ret;
2015-06-16 11:50:57 +00:00
ret = perf_probe_event__sprintf(group, event, pev, module, &buf);
if (ret >= 0) {
if (use_stdout)
printf("%s\n", buf.buf);
else
pr_info("%s\n", buf.buf);
}
strbuf_release(&buf);
return ret;
}
static bool filter_probe_trace_event(struct probe_trace_event *tev,
struct strfilter *filter)
{
char tmp[128];
/* At first, check the event name itself */
if (strfilter__compare(filter, tev->event))
return true;
/* Next, check the combination of name and group */
if (e_snprintf(tmp, 128, "%s:%s", tev->group, tev->event) < 0)
return false;
return strfilter__compare(filter, tmp);
}
static int __show_perf_probe_events(int fd, bool is_kprobe,
struct strfilter *filter)
{
int ret = 0;
struct probe_trace_event tev;
struct perf_probe_event pev;
struct strlist *rawlist;
struct str_node *ent;
memset(&tev, 0, sizeof(tev));
memset(&pev, 0, sizeof(pev));
rawlist = probe_file__get_rawlist(fd);
if (!rawlist)
return -ENOMEM;
strlist__for_each_entry(ent, rawlist) {
ret = parse_probe_trace_command(ent->s, &tev);
if (ret >= 0) {
if (!filter_probe_trace_event(&tev, filter))
goto next;
ret = convert_to_perf_probe_event(&tev, &pev,
is_kprobe);
if (ret < 0)
goto next;
2015-06-16 11:50:57 +00:00
ret = show_perf_probe_event(pev.group, pev.event,
&pev, tev.point.module,
true);
}
next:
clear_perf_probe_event(&pev);
clear_probe_trace_event(&tev);
if (ret < 0)
break;
}
strlist__delete(rawlist);
/* Cleanup cached debuginfo if needed */
debuginfo_cache__exit();
return ret;
}
/* List up current perf-probe events */
int show_perf_probe_events(struct strfilter *filter)
{
int kp_fd, up_fd, ret;
setup_pager();
if (probe_conf.cache)
return probe_cache__show_all_caches(filter);
ret = init_probe_symbol_maps(false);
if (ret < 0)
return ret;
ret = probe_file__open_both(&kp_fd, &up_fd, 0);
if (ret < 0)
return ret;
if (kp_fd >= 0)
ret = __show_perf_probe_events(kp_fd, true, filter);
if (up_fd >= 0 && ret >= 0)
ret = __show_perf_probe_events(up_fd, false, filter);
if (kp_fd > 0)
close(kp_fd);
if (up_fd > 0)
close(up_fd);
exit_probe_symbol_maps();
return ret;
}
static int get_new_event_name(char *buf, size_t len, const char *base,
perf probe: Add __return suffix for return events Add __return suffix for function return events automatically. Without this, user have to give --force option and will see the number suffix for each event like "function_1", which is not easy to recognize. Instead, this adds __return suffix to it automatically. E.g. ===== # ./perf probe -x /lib64/libc-2.25.so 'malloc*%return' Added new events: probe_libc:malloc_printerr__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_consolidate__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_check__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_hook_ini__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_trim__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_usable_size__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_stats__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_info__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:mallochook__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_get_state__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_set_state__return (on malloc*%return in /usr/lib64/libc-2.25.so) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc_set_state__return -aR sleep 1 ===== Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Clarke <pc@us.ibm.com> Cc: bhargavb <bhargavaramudu@gmail.com> Cc: linux-rt-users@vger.kernel.org Link: http://lkml.kernel.org/r/151275046418.24652.6696011972866498489.stgit@devbox Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-12-08 16:27:44 +00:00
struct strlist *namelist, bool ret_event,
bool allow_suffix)
{
int i, ret;
char *p, *nbase;
if (*base == '.')
base++;
nbase = strdup(base);
if (!nbase)
return -ENOMEM;
perf probe: Cut off the version suffix from event name Cut off the version suffix (e.g. @GLIBC_2.2.5 etc.) from automatic generated event name. This fixes wildcard event adding like below case; ===== # perf probe -x /lib64/libc-2.25.so malloc* Internal error: "malloc_get_state@GLIBC_2" is wrong event name. Error: Failed to add events. ===== This failure was caused by a versioned suffix symbol. With this fix, perf probe automatically cuts the suffix after @ as below. ===== # ./perf probe -x /lib64/libc-2.25.so malloc* Added new events: probe_libc:malloc_printerr (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_consolidate (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_check (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_hook_ini (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_trim (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_usable_size (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_stats (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_info (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:mallochook (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_get_state (on malloc* in /usr/lib64/libc-2.25.so) probe_libc:malloc_set_state (on malloc* in /usr/lib64/libc-2.25.so) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc_set_state -aR sleep 1 ===== Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Reported-by: bhargavb <bhargavaramudu@gmail.com> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Clarke <pc@us.ibm.com> Cc: linux-rt-users@vger.kernel.org Link: http://lkml.kernel.org/r/None Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-12-11 18:19:25 +00:00
/* Cut off the dot suffixes (e.g. .const, .isra) and version suffixes */
p = strpbrk(nbase, ".@");
if (p && p != nbase)
*p = '\0';
/* Try no suffix number */
perf probe: Add __return suffix for return events Add __return suffix for function return events automatically. Without this, user have to give --force option and will see the number suffix for each event like "function_1", which is not easy to recognize. Instead, this adds __return suffix to it automatically. E.g. ===== # ./perf probe -x /lib64/libc-2.25.so 'malloc*%return' Added new events: probe_libc:malloc_printerr__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_consolidate__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_check__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_hook_ini__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_trim__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_usable_size__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_stats__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_info__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:mallochook__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_get_state__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_set_state__return (on malloc*%return in /usr/lib64/libc-2.25.so) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc_set_state__return -aR sleep 1 ===== Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Clarke <pc@us.ibm.com> Cc: bhargavb <bhargavaramudu@gmail.com> Cc: linux-rt-users@vger.kernel.org Link: http://lkml.kernel.org/r/151275046418.24652.6696011972866498489.stgit@devbox Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-12-08 16:27:44 +00:00
ret = e_snprintf(buf, len, "%s%s", nbase, ret_event ? "__return" : "");
if (ret < 0) {
pr_debug("snprintf() failed: %d\n", ret);
goto out;
}
if (!strlist__has_entry(namelist, buf))
goto out;
if (!allow_suffix) {
perf bpf: Allow BPF program config probing options By extending the syntax of BPF object section names, this patch allows users to config probing options like what they can do in 'perf probe'. The error message in 'perf probe' is also updated. Test result: For following BPF file test_probe_glob.c: # cat test_probe_glob.c __attribute__((section("inlines=no;func=SyS_dup?"), used)) int func(void *ctx) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # # ./perf record -e ./test_probe_glob.c ls / ... [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.013 MB perf.data ] # ./perf evlist perf_bpf_probe:func_1 perf_bpf_probe:func After changing "inlines=no" to "inlines=yes": # ./perf record -e ./test_probe_glob.c ls / ... [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.013 MB perf.data ] # ./perf evlist perf_bpf_probe:func_3 perf_bpf_probe:func_2 perf_bpf_probe:func_1 perf_bpf_probe:func Then test 'force': Use following program: # cat test_probe_force.c __attribute__((section("func=sys_write"), used)) int funca(void *ctx) { return 1; } __attribute__((section("force=yes;func=sys_write"), used)) int funcb(void *ctx) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # # perf record -e ./test_probe_force.c usleep 1 Error: event "func" already exists. Hint: Remove existing event by 'perf probe -d' or force duplicates by 'perf probe -f' or set 'force=yes' in BPF source. event syntax error: './test_probe_force.c' \___ Probe point exist. Try 'perf probe -d "*"' and set 'force=yes' (add -v to see detail) ... Then replace 'force=no' to 'force=yes': # vim test_probe_force.c # perf record -e ./test_probe_force.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.017 MB perf.data ] # perf evlist perf_bpf_probe:func_1 perf_bpf_probe:func # Signed-off-by: Wang Nan <wangnan0@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 12:10:08 +00:00
pr_warning("Error: event \"%s\" already exists.\n"
" Hint: Remove existing event by 'perf probe -d'\n"
" or force duplicates by 'perf probe -f'\n"
" or set 'force=yes' in BPF source.\n",
buf);
ret = -EEXIST;
goto out;
}
/* Try to add suffix */
for (i = 1; i < MAX_EVENT_INDEX; i++) {
ret = e_snprintf(buf, len, "%s_%d", nbase, i);
if (ret < 0) {
pr_debug("snprintf() failed: %d\n", ret);
goto out;
}
if (!strlist__has_entry(namelist, buf))
break;
}
if (i == MAX_EVENT_INDEX) {
pr_warning("Too many events are on the same function.\n");
ret = -ERANGE;
}
out:
free(nbase);
/* Final validation */
if (ret >= 0 && !is_c_func_name(buf)) {
pr_warning("Internal error: \"%s\" is an invalid event name.\n",
buf);
ret = -EINVAL;
}
return ret;
}
/* Warn if the current kernel's uprobe implementation is old */
static void warn_uprobe_event_compat(struct probe_trace_event *tev)
{
int i;
char *buf = synthesize_probe_trace_command(tev);
/* Old uprobe event doesn't support memory dereference */
if (!tev->uprobes || tev->nargs == 0 || !buf)
goto out;
for (i = 0; i < tev->nargs; i++)
if (strglobmatch(tev->args[i].value, "[$@+-]*")) {
pr_warning("Please upgrade your kernel to at least "
"3.14 to have access to feature %s\n",
tev->args[i].value);
break;
}
out:
free(buf);
}
/* Set new name from original perf_probe_event and namelist */
static int probe_trace_event__set_name(struct probe_trace_event *tev,
struct perf_probe_event *pev,
struct strlist *namelist,
bool allow_suffix)
{
const char *event, *group;
char buf[64];
int ret;
/* If probe_event or trace_event already have the name, reuse it */
if (pev->event && !pev->sdt)
event = pev->event;
else if (tev->event)
event = tev->event;
else {
/* Or generate new one from probe point */
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
if (pev->point.function &&
(strncmp(pev->point.function, "0x", 2) != 0) &&
!strisglob(pev->point.function))
event = pev->point.function;
else
event = tev->point.realname;
}
if (pev->group && !pev->sdt)
group = pev->group;
else if (tev->group)
group = tev->group;
else
group = PERFPROBE_GROUP;
/* Get an unused new event name */
perf probe: Add __return suffix for return events Add __return suffix for function return events automatically. Without this, user have to give --force option and will see the number suffix for each event like "function_1", which is not easy to recognize. Instead, this adds __return suffix to it automatically. E.g. ===== # ./perf probe -x /lib64/libc-2.25.so 'malloc*%return' Added new events: probe_libc:malloc_printerr__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_consolidate__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_check__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_hook_ini__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_trim__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_usable_size__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_stats__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_info__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:mallochook__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_get_state__return (on malloc*%return in /usr/lib64/libc-2.25.so) probe_libc:malloc_set_state__return (on malloc*%return in /usr/lib64/libc-2.25.so) You can now use it in all perf tools, such as: perf record -e probe_libc:malloc_set_state__return -aR sleep 1 ===== Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Clarke <pc@us.ibm.com> Cc: bhargavb <bhargavaramudu@gmail.com> Cc: linux-rt-users@vger.kernel.org Link: http://lkml.kernel.org/r/151275046418.24652.6696011972866498489.stgit@devbox Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-12-08 16:27:44 +00:00
ret = get_new_event_name(buf, 64, event, namelist,
tev->point.retprobe, allow_suffix);
if (ret < 0)
return ret;
event = buf;
tev->event = strdup(event);
tev->group = strdup(group);
if (tev->event == NULL || tev->group == NULL)
return -ENOMEM;
/* Add added event name to namelist */
strlist__add(namelist, event);
return 0;
}
static int __open_probe_file_and_namelist(bool uprobe,
struct strlist **namelist)
{
int fd;
fd = probe_file__open(PF_FL_RW | (uprobe ? PF_FL_UPROBE : 0));
if (fd < 0)
return fd;
/* Get current event names */
*namelist = probe_file__get_namelist(fd);
if (!(*namelist)) {
pr_debug("Failed to get current event list.\n");
close(fd);
return -ENOMEM;
}
return fd;
}
static int __add_probe_trace_events(struct perf_probe_event *pev,
struct probe_trace_event *tevs,
int ntevs, bool allow_suffix)
{
int i, fd[2] = {-1, -1}, up, ret;
struct probe_trace_event *tev = NULL;
struct probe_cache *cache = NULL;
struct strlist *namelist[2] = {NULL, NULL};
struct nscookie nsc;
up = pev->uprobes ? 1 : 0;
fd[up] = __open_probe_file_and_namelist(up, &namelist[up]);
if (fd[up] < 0)
return fd[up];
ret = 0;
for (i = 0; i < ntevs; i++) {
tev = &tevs[i];
up = tev->uprobes ? 1 : 0;
if (fd[up] == -1) { /* Open the kprobe/uprobe_events */
fd[up] = __open_probe_file_and_namelist(up,
&namelist[up]);
if (fd[up] < 0)
goto close_out;
}
/* Skip if the symbol is out of .text or blacklisted */
if (!tev->point.symbol && !pev->uprobes)
continue;
/* Set new name for tev (and update namelist) */
ret = probe_trace_event__set_name(tev, pev, namelist[up],
allow_suffix);
if (ret < 0)
break;
nsinfo__mountns_enter(pev->nsi, &nsc);
ret = probe_file__add_event(fd[up], tev);
nsinfo__mountns_exit(&nsc);
if (ret < 0)
break;
/*
* Probes after the first probe which comes from same
* user input are always allowed to add suffix, because
* there might be several addresses corresponding to
* one code line.
*/
allow_suffix = true;
}
if (ret == -EINVAL && pev->uprobes)
warn_uprobe_event_compat(tev);
if (ret == 0 && probe_conf.cache) {
cache = probe_cache__new(pev->target, pev->nsi);
if (!cache ||
probe_cache__add_entry(cache, pev, tevs, ntevs) < 0 ||
probe_cache__commit(cache) < 0)
pr_warning("Failed to add event to probe cache\n");
probe_cache__delete(cache);
}
close_out:
for (up = 0; up < 2; up++) {
strlist__delete(namelist[up]);
if (fd[up] >= 0)
close(fd[up]);
}
return ret;
}
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
static int find_probe_functions(struct map *map, char *name,
struct symbol **syms)
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
{
int found = 0;
struct symbol *sym;
struct rb_node *tmp;
const char *norm, *ver;
char *buf = NULL;
bool cut_version = true;
if (map__load(map) < 0)
return 0;
/* If user gives a version, don't cut off the version from symbols */
if (strchr(name, '@'))
cut_version = false;
map__for_each_symbol(map, sym, tmp) {
norm = arch__normalize_symbol_name(sym->name);
if (!norm)
continue;
if (cut_version) {
/* We don't care about default symbol or not */
ver = strchr(norm, '@');
if (ver) {
buf = strndup(norm, ver - norm);
if (!buf)
return -ENOMEM;
norm = buf;
}
}
if (strglobmatch(norm, name)) {
found++;
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
if (syms && found < probe_conf.max_probes)
syms[found - 1] = sym;
}
if (buf)
zfree(&buf);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
}
return found;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
}
void __weak arch__fix_tev_from_maps(struct perf_probe_event *pev __maybe_unused,
struct probe_trace_event *tev __maybe_unused,
struct map *map __maybe_unused,
struct symbol *sym __maybe_unused) { }
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
/*
* Find probe function addresses from map.
* Return an error or the number of found probe_trace_event
*/
static int find_probe_trace_events_from_map(struct perf_probe_event *pev,
struct probe_trace_event **tevs)
{
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
struct map *map = NULL;
struct ref_reloc_sym *reloc_sym = NULL;
struct symbol *sym;
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
struct symbol **syms = NULL;
struct probe_trace_event *tev;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
struct perf_probe_point *pp = &pev->point;
struct probe_trace_point *tp;
int num_matched_functions;
int ret, i, j, skipped = 0;
perf probe: Fix module probe issue if no dwarf support Perf is not able to register probe in kernel module when dwarf supprt is not there(and so it goes for symtab). Perf passes full path of module where only module name is required which is causing the problem. This patch fixes this issue. Before applying patch: $ dpkg -s libdw-dev dpkg-query: package 'libdw-dev' is not installed and no information is... $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in /linux/samples/kprobes/kprobe_example.ko) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init /linux/samples/kprobes/kprobe_example.ko:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data ] $ sudo ./perf script # No output here After applying patch: $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in kprobe_example) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init kprobe_example:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data (2 samples) ] $ sudo ./perf script insmod 13990 [002] 5961.216833: probe:kprobe_init: ... insmod 13995 [002] 5962.889384: probe:kprobe_init: ... Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:40 +00:00
char *mod_name;
map = get_target_map(pev->target, pev->nsi, pev->uprobes);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (!map) {
ret = -EINVAL;
goto out;
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
}
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
syms = malloc(sizeof(struct symbol *) * probe_conf.max_probes);
if (!syms) {
ret = -ENOMEM;
goto out;
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
/*
* Load matched symbols: Since the different local symbols may have
* same name but different addresses, this lists all the symbols.
*/
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
num_matched_functions = find_probe_functions(map, pp->function, syms);
if (num_matched_functions <= 0) {
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
pr_err("Failed to find symbol %s in %s\n", pp->function,
pev->target ? : "kernel");
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
ret = -ENOENT;
goto out;
} else if (num_matched_functions > probe_conf.max_probes) {
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
pr_err("Too many functions matched in %s\n",
pev->target ? : "kernel");
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
ret = -E2BIG;
goto out;
perf probe: Support basic dwarf-based operations on uprobe events Support basic dwarf(debuginfo) based operations for uprobe events. With this change, perf probe can analyze debuginfo of user application binary to set up new uprobe event. This allows perf-probe --add(with local variables, line numbers) and --line works with -x option. (Actually, --vars has already accepted -x option) For example, the following command shows the probe-able lines of a given user space function. Something that so far was only available in the 'perf probe' tool for kernel space functions: # ./perf probe -x perf --line map__load <map__load@/home/fedora/ksrc/linux-2.6/tools/perf/util/map.c:0> 0 int map__load(struct map *map, symbol_filter_t filter) 1 { 2 const char *name = map->dso->long_name; int nr; 5 if (dso__loaded(map->dso, map->type)) 6 return 0; 8 nr = dso__load(map->dso, map, filter); 9 if (nr < 0) { 10 if (map->dso->has_build_id) { And this shows the available variables at the given line of the function. # ./perf probe -x perf --vars map__load:8 Available variables at map__load:8 @<map__load+96> char* name struct map* map symbol_filter_t filter @<map__find_symbol+112> char* name symbol_filter_t filter @<map__find_symbol_by_name+136> char* name symbol_filter_t filter @<map_groups__find_symbol_by_name+176> char* name struct map* map symbol_filter_t filter And lastly, we can now define probe(s) with all available variables on the given line: # ./perf probe -x perf --add 'map__load:8 $vars' Added new events: probe_perf:map__load (on map__load:8 with $vars) probe_perf:map__load_1 (on map__load:8 with $vars) probe_perf:map__load_2 (on map__load:8 with $vars) probe_perf:map__load_3 (on map__load:8 with $vars) You can now use it in all perf tools, such as: perf record -e probe_perf:map__load_3 -aR sleep 1 Changes from previous version: - Add examples in the patch description. - Use .text section start address and dwarf symbol address for calculating the offset of given symbol, instead of searching the symbol in symtab again. With this change, we can safely handle multiple local function instances (e.g. scnprintf in perf). Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: David A. Long <dave.long@linaro.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: systemtap@sourceware.org Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20131226054152.22364.47021.stgit@kbuild-fedora.novalocal Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-12-26 05:41:53 +00:00
}
/* Note that the symbols in the kmodule are not relocated */
perf kretprobes: Offset from reloc_sym if kernel supports it We indicate support for accepting sym+offset with kretprobes through a line in ftrace README. Parse the same to identify support and choose the appropriate format for kprobe_events. As an example, without this perf patch, but with the ftrace changes: naveen@ubuntu:~/linux/tools/perf$ sudo cat /sys/kernel/debug/tracing/README | grep kretprobe place (kretprobe): [<module>:]<symbol>[+<offset>]|<memaddr> naveen@ubuntu:~/linux/tools/perf$ naveen@ubuntu:~/linux/tools/perf$ sudo ./perf probe -v do_open%return probe-definition(0): do_open%return symbol:do_open file:(null) line:0 offset:0 return:1 lazy:(null) 0 arguments Looking at the vmlinux_path (8 entries long) Using /boot/vmlinux for symbols Open Debuginfo file: /boot/vmlinux Try to find probe point from debuginfo. Matched function: do_open [2d0c7d8] Probe point found: do_open+0 Matched function: do_open [35d76b5] found inline addr: 0xc0000000004ba984 Failed to find "do_open%return", because do_open is an inlined function and has no return point. An error occurred in debuginfo analysis (-22). Trying to use symbols. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: r:probe/do_open do_open+0 Writing event: r:probe/do_open_1 do_open+0 Added new events: probe:do_open (on do_open%return) probe:do_open_1 (on do_open%return) You can now use it in all perf tools, such as: perf record -e probe:do_open_1 -aR sleep 1 naveen@ubuntu:~/linux/tools/perf$ sudo cat /sys/kernel/debug/kprobes/list c000000000041370 k kretprobe_trampoline+0x0 [OPTIMIZED] c0000000004433d0 r do_open+0x0 [DISABLED] c0000000004433d0 r do_open+0x0 [DISABLED] And after this patch (and the subsequent powerpc patch): naveen@ubuntu:~/linux/tools/perf$ sudo ./perf probe -v do_open%return probe-definition(0): do_open%return symbol:do_open file:(null) line:0 offset:0 return:1 lazy:(null) 0 arguments Looking at the vmlinux_path (8 entries long) Using /boot/vmlinux for symbols Open Debuginfo file: /boot/vmlinux Try to find probe point from debuginfo. Matched function: do_open [2d0c7d8] Probe point found: do_open+0 Matched function: do_open [35d76b5] found inline addr: 0xc0000000004ba984 Failed to find "do_open%return", because do_open is an inlined function and has no return point. An error occurred in debuginfo analysis (-22). Trying to use symbols. Opening /sys/kernel/debug/tracing//README write=0 Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: r:probe/do_open _text+4469712 Writing event: r:probe/do_open_1 _text+4956248 Added new events: probe:do_open (on do_open%return) probe:do_open_1 (on do_open%return) You can now use it in all perf tools, such as: perf record -e probe:do_open_1 -aR sleep 1 naveen@ubuntu:~/linux/tools/perf$ sudo cat /sys/kernel/debug/kprobes/list c000000000041370 k kretprobe_trampoline+0x0 [OPTIMIZED] c0000000004433d0 r do_open+0x0 [DISABLED] c0000000004ba058 r do_open+0x8 [DISABLED] Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/496ef9f33c1ab16286ece9dd62aa672807aef91c.1488961018.git.naveen.n.rao@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-03-08 08:26:09 +00:00
if (!pev->uprobes && !pev->target &&
(!pp->retprobe || kretprobe_offset_is_supported())) {
perf probe: Fix failure to add multiple probes without debuginfo Perf tries to find probe function addresses from map when debuginfo could not be found. To the first added function, the value of ref_reloc_sym was set in maps__set_kallsyms_ref_reloc_sym() and can be obtained from host_machine->kmaps->maps. After that, new maps are added to host_machine->kmaps->maps in dso__load_kcore(), all these new added maps do not have a valid ref_reloc_sym. When adding a second function, get_target_map() may get a map without valid ref_reloc_sym, and raise the error "Relocated base symbol is not found". Fix this by using kernel_get_ref_reloc_sym() to get ref_reloc_sym. This problem can be reproduced as following: $ perf probe --add='sys_write' --add='sys_open' Relocated base symbol is not found! Error: Failed to add events. After this patch: $ perf probe --add='sys_write' --add='sys_open' Added new event: probe:sys_write (on sys_write) You can now use it in all perf tools, such as: perf record -e probe:sys_write -aR sleep 1 Added new event: probe:sys_open (on sys_open) You can now use it in all perf tools, such as: perf record -e probe:sys_open -aR sleep 1 Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Tested-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1426816616-2394-1-git-send-email-hekuang@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-03-20 01:56:56 +00:00
reloc_sym = kernel_get_ref_reloc_sym();
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (!reloc_sym) {
pr_warning("Relocated base symbol is not found!\n");
ret = -EINVAL;
goto out;
}
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
/* Setup result trace-probe-events */
*tevs = zalloc(sizeof(*tev) * num_matched_functions);
if (!*tevs) {
ret = -ENOMEM;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
goto out;
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
ret = 0;
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
for (j = 0; j < num_matched_functions; j++) {
sym = syms[j];
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
tev = (*tevs) + ret;
tp = &tev->point;
if (ret == num_matched_functions) {
pr_warning("Too many symbols are listed. Skip it.\n");
break;
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
ret++;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (pp->offset > sym->end - sym->start) {
pr_warning("Offset %ld is bigger than the size of %s\n",
pp->offset, sym->name);
ret = -ENOENT;
goto err_out;
}
/* Add one probe point */
tp->address = map->unmap_ip(map, sym->start) + pp->offset;
/* Check the kprobe (not in module) is within .text */
if (!pev->uprobes && !pev->target &&
kprobe_warn_out_range(sym->name, tp->address)) {
tp->symbol = NULL; /* Skip it */
skipped++;
} else if (reloc_sym) {
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
tp->symbol = strdup_or_goto(reloc_sym->name, nomem_out);
tp->offset = tp->address - reloc_sym->addr;
} else {
tp->symbol = strdup_or_goto(sym->name, nomem_out);
tp->offset = pp->offset;
}
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
tp->realname = strdup_or_goto(sym->name, nomem_out);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
tp->retprobe = pp->retprobe;
perf probe: Fix module probe issue if no dwarf support Perf is not able to register probe in kernel module when dwarf supprt is not there(and so it goes for symtab). Perf passes full path of module where only module name is required which is causing the problem. This patch fixes this issue. Before applying patch: $ dpkg -s libdw-dev dpkg-query: package 'libdw-dev' is not installed and no information is... $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in /linux/samples/kprobes/kprobe_example.ko) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init /linux/samples/kprobes/kprobe_example.ko:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data ] $ sudo ./perf script # No output here After applying patch: $ sudo ./perf probe -m /linux/samples/kprobes/kprobe_example.ko kprobe_init Added new event: probe:kprobe_init (on kprobe_init in kprobe_example) You can now use it in all perf tools, such as: perf record -e probe:kprobe_init -aR sleep 1 $ sudo cat /sys/kernel/debug/tracing/kprobe_events p:probe/kprobe_init kprobe_example:kprobe_init $ sudo ./perf record -a -e probe:kprobe_init [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.105 MB perf.data (2 samples) ] $ sudo ./perf script insmod 13990 [002] 5961.216833: probe:kprobe_init: ... insmod 13995 [002] 5962.889384: probe:kprobe_init: ... Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/r/1461680741-12517-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-26 14:25:40 +00:00
if (pev->target) {
if (pev->uprobes) {
tev->point.module = strdup_or_goto(pev->target,
nomem_out);
} else {
mod_name = find_module_name(pev->target);
tev->point.module =
strdup(mod_name ? mod_name : pev->target);
free(mod_name);
if (!tev->point.module)
goto nomem_out;
}
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
tev->uprobes = pev->uprobes;
tev->nargs = pev->nargs;
if (tev->nargs) {
tev->args = zalloc(sizeof(struct probe_trace_arg) *
tev->nargs);
if (tev->args == NULL)
goto nomem_out;
}
for (i = 0; i < tev->nargs; i++) {
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (pev->args[i].name)
tev->args[i].name =
strdup_or_goto(pev->args[i].name,
nomem_out);
tev->args[i].value = strdup_or_goto(pev->args[i].var,
nomem_out);
if (pev->args[i].type)
tev->args[i].type =
strdup_or_goto(pev->args[i].type,
nomem_out);
}
arch__fix_tev_from_maps(pev, tev, map, sym);
}
if (ret == skipped) {
ret = -ENOENT;
goto err_out;
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
out:
map__put(map);
perf probe: Fix segfault when glob matching function without debuginfo Commit 4c859351226c920b227fec040a3b447f0d482af3 ("perf probe: Support glob wildcards for function name") introduces segfault problems when debuginfo is not available: # perf probe 'sys_w*' Added new events: Segmentation fault The first problem resides in find_probe_trace_events_from_map(). In that function, find_probe_functions() is called to match each symbol against glob to find the number of matching functions, but still use map__for_each_symbol_by_name() to find 'struct symbol' for matching functions. Unfortunately, map__for_each_symbol_by_name() does exact matching by searching in an rbtree. It doesn't know glob matching, and not easy for it to support it because it use rbtree based binary search, but we are unable to ensure all names matched by the glob (any glob passed by user) reside in one subtree. This patch drops map__for_each_symbol_by_name(). Since there is no rbtree again, re-matching all symbols costs a lot. This patch avoid it by saving all matching results into an array (syms). The second problem is the lost of tp->realname. In __add_probe_trace_events(), if pev->point.function is glob, the event name should be set to tev->point.realname. This patch ensures its existence by strdup sym->name instead of leaving a NULL pointer there. After this patch: # perf probe 'sys_w*' Added new events: probe:sys_waitid (on sys_w*) probe:sys_wait4 (on sys_w*) probe:sys_waitpid (on sys_w*) probe:sys_write (on sys_w*) probe:sys_writev (on sys_w*) You can now use it in all perf tools, such as: perf record -e probe:sys_writev -aR sleep 1 Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Zefan Li <lizefan@huawei.com> Link: http://lkml.kernel.org/r/1432892747-232506-1-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-05-29 09:45:47 +00:00
free(syms);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
return ret;
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
nomem_out:
ret = -ENOMEM;
err_out:
clear_probe_trace_events(*tevs, num_matched_functions);
zfree(tevs);
goto out;
}
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
static int try_to_find_absolute_address(struct perf_probe_event *pev,
struct probe_trace_event **tevs)
{
struct perf_probe_point *pp = &pev->point;
struct probe_trace_event *tev;
struct probe_trace_point *tp;
int i, err;
if (!(pev->point.function && !strncmp(pev->point.function, "0x", 2)))
return -EINVAL;
if (perf_probe_event_need_dwarf(pev))
return -EINVAL;
/*
* This is 'perf probe /lib/libc.so 0xabcd'. Try to probe at
* absolute address.
*
* Only one tev can be generated by this.
*/
*tevs = zalloc(sizeof(*tev));
if (!*tevs)
return -ENOMEM;
tev = *tevs;
tp = &tev->point;
/*
* Don't use tp->offset, use address directly, because
* in synthesize_probe_trace_command() address cannot be
* zero.
*/
tp->address = pev->point.abs_address;
tp->retprobe = pp->retprobe;
tev->uprobes = pev->uprobes;
err = -ENOMEM;
/*
* Give it a '0x' leading symbol name.
* In __add_probe_trace_events, a NULL symbol is interpreted as
* invalud.
*/
if (asprintf(&tp->symbol, "0x%lx", tp->address) < 0)
goto errout;
/* For kprobe, check range */
if ((!tev->uprobes) &&
(kprobe_warn_out_range(tev->point.symbol,
tev->point.address))) {
err = -EACCES;
goto errout;
}
if (asprintf(&tp->realname, "abs_%lx", tp->address) < 0)
goto errout;
if (pev->target) {
tp->module = strdup(pev->target);
if (!tp->module)
goto errout;
}
if (tev->group) {
tev->group = strdup(pev->group);
if (!tev->group)
goto errout;
}
if (pev->event) {
tev->event = strdup(pev->event);
if (!tev->event)
goto errout;
}
tev->nargs = pev->nargs;
tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs);
if (!tev->args)
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
goto errout;
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
for (i = 0; i < tev->nargs; i++)
copy_to_probe_trace_arg(&tev->args[i], &pev->args[i]);
return 1;
errout:
clear_probe_trace_events(*tevs, 1);
*tevs = NULL;
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
return err;
}
/* Concatinate two arrays */
static void *memcat(void *a, size_t sz_a, void *b, size_t sz_b)
{
void *ret;
ret = malloc(sz_a + sz_b);
if (ret) {
memcpy(ret, a, sz_a);
memcpy(ret + sz_a, b, sz_b);
}
return ret;
}
static int
concat_probe_trace_events(struct probe_trace_event **tevs, int *ntevs,
struct probe_trace_event **tevs2, int ntevs2)
{
struct probe_trace_event *new_tevs;
int ret = 0;
if (*ntevs == 0) {
*tevs = *tevs2;
*ntevs = ntevs2;
*tevs2 = NULL;
return 0;
}
if (*ntevs + ntevs2 > probe_conf.max_probes)
ret = -E2BIG;
else {
/* Concatinate the array of probe_trace_event */
new_tevs = memcat(*tevs, (*ntevs) * sizeof(**tevs),
*tevs2, ntevs2 * sizeof(**tevs2));
if (!new_tevs)
ret = -ENOMEM;
else {
free(*tevs);
*tevs = new_tevs;
*ntevs += ntevs2;
}
}
if (ret < 0)
clear_probe_trace_events(*tevs2, ntevs2);
zfree(tevs2);
return ret;
}
/*
* Try to find probe_trace_event from given probe caches. Return the number
* of cached events found, if an error occurs return the error.
*/
static int find_cached_events(struct perf_probe_event *pev,
struct probe_trace_event **tevs,
const char *target)
{
struct probe_cache *cache;
struct probe_cache_entry *entry;
struct probe_trace_event *tmp_tevs = NULL;
int ntevs = 0;
int ret = 0;
cache = probe_cache__new(target, pev->nsi);
/* Return 0 ("not found") if the target has no probe cache. */
if (!cache)
return 0;
for_each_probe_cache_entry(entry, cache) {
/* Skip the cache entry which has no name */
if (!entry->pev.event || !entry->pev.group)
continue;
if ((!pev->group || strglobmatch(entry->pev.group, pev->group)) &&
strglobmatch(entry->pev.event, pev->event)) {
ret = probe_cache_entry__get_event(entry, &tmp_tevs);
if (ret > 0)
ret = concat_probe_trace_events(tevs, &ntevs,
&tmp_tevs, ret);
if (ret < 0)
break;
}
}
probe_cache__delete(cache);
if (ret < 0) {
clear_probe_trace_events(*tevs, ntevs);
zfree(tevs);
} else {
ret = ntevs;
if (ntevs > 0 && target && target[0] == '/')
pev->uprobes = true;
}
return ret;
}
/* Try to find probe_trace_event from all probe caches */
static int find_cached_events_all(struct perf_probe_event *pev,
struct probe_trace_event **tevs)
{
struct probe_trace_event *tmp_tevs = NULL;
struct strlist *bidlist;
struct str_node *nd;
char *pathname;
int ntevs = 0;
int ret;
/* Get the buildid list of all valid caches */
bidlist = build_id_cache__list_all(true);
if (!bidlist) {
ret = -errno;
pr_debug("Failed to get buildids: %d\n", ret);
return ret;
}
ret = 0;
strlist__for_each_entry(nd, bidlist) {
pathname = build_id_cache__origname(nd->s);
ret = find_cached_events(pev, &tmp_tevs, pathname);
/* In the case of cnt == 0, we just skip it */
if (ret > 0)
ret = concat_probe_trace_events(tevs, &ntevs,
&tmp_tevs, ret);
free(pathname);
if (ret < 0)
break;
}
strlist__delete(bidlist);
if (ret < 0) {
clear_probe_trace_events(*tevs, ntevs);
zfree(tevs);
} else
ret = ntevs;
return ret;
}
static int find_probe_trace_events_from_cache(struct perf_probe_event *pev,
struct probe_trace_event **tevs)
{
struct probe_cache *cache;
struct probe_cache_entry *entry;
struct probe_trace_event *tev;
struct str_node *node;
int ret, i;
if (pev->sdt) {
/* For SDT/cached events, we use special search functions */
if (!pev->target)
return find_cached_events_all(pev, tevs);
else
return find_cached_events(pev, tevs, pev->target);
}
cache = probe_cache__new(pev->target, pev->nsi);
if (!cache)
return 0;
entry = probe_cache__find(cache, pev);
if (!entry) {
/* SDT must be in the cache */
ret = pev->sdt ? -ENOENT : 0;
goto out;
}
ret = strlist__nr_entries(entry->tevlist);
if (ret > probe_conf.max_probes) {
pr_debug("Too many entries matched in the cache of %s\n",
pev->target ? : "kernel");
ret = -E2BIG;
goto out;
}
*tevs = zalloc(ret * sizeof(*tev));
if (!*tevs) {
ret = -ENOMEM;
goto out;
}
i = 0;
strlist__for_each_entry(node, entry->tevlist) {
tev = &(*tevs)[i++];
ret = parse_probe_trace_command(node->s, tev);
if (ret < 0)
goto out;
/* Set the uprobes attribute as same as original */
tev->uprobes = pev->uprobes;
}
ret = i;
out:
probe_cache__delete(cache);
return ret;
}
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
static int convert_to_probe_trace_events(struct perf_probe_event *pev,
struct probe_trace_event **tevs)
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
{
int ret;
if (!pev->group && !pev->sdt) {
/* Set group name if not given */
if (!pev->uprobes) {
pev->group = strdup(PERFPROBE_GROUP);
ret = pev->group ? 0 : -ENOMEM;
} else
ret = convert_exec_to_group(pev->target, &pev->group);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (ret != 0) {
pr_warning("Failed to make a group name.\n");
return ret;
}
}
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
ret = try_to_find_absolute_address(pev, tevs);
if (ret > 0)
return ret;
/* At first, we need to lookup cache entry */
ret = find_probe_trace_events_from_cache(pev, tevs);
if (ret > 0 || pev->sdt) /* SDT can be found only in the cache */
return ret == 0 ? -ENOENT : ret; /* Found in probe cache */
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
/* Convert perf_probe_event with debuginfo */
ret = try_to_find_probe_trace_events(pev, tevs);
perf probe: Allow to add events on the local functions Allow to add events on the local functions without debuginfo. (With the debuginfo, we can add events even on inlined functions) Currently, probing on local functions requires debuginfo to locate actual address. It is also possible without debuginfo since we have symbol maps. Without this change; ---- # ./perf probe -a t_show Added new event: probe:t_show (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show -aR sleep 1 # ./perf probe -x perf -a identity__map_ip no symbols found in /kbuild/ksrc/linux-3/tools/perf/perf, maybe install a debug package? Failed to load map. Error: Failed to add events. (-22) ---- As the above results, perf probe just put one event on the first found symbol for kprobe event. Moreover, for uprobe event, perf probe failed to find local functions. With this change; ---- # ./perf probe -a t_show Added new events: probe:t_show (on t_show) probe:t_show_1 (on t_show) probe:t_show_2 (on t_show) probe:t_show_3 (on t_show) You can now use it in all perf tools, such as: perf record -e probe:t_show_3 -aR sleep 1 # ./perf probe -x perf -a identity__map_ip Added new events: probe_perf:identity__map_ip (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_1 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_2 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) probe_perf:identity__map_ip_3 (on identity__map_ip in /kbuild/ksrc/linux-3/tools/perf/perf) You can now use it in all perf tools, such as: perf record -e probe_perf:identity__map_ip_3 -aR sleep 1 ---- Now we succeed to put events on every given local functions for both kprobes and uprobes. :) Note that this also introduces some symbol rbtree iteration macros; symbols__for_each, dso__for_each_symbol, and map__for_each_symbol. These are for walking through the symbol list in a map. Changes from v2: - Fix add_exec_to_probe_trace_events() not to convert address to tp->symbol any more. - Fix to set kernel probes based on ref_reloc_sym. Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: David Ahern <dsahern@gmail.com> Cc: "David A. Long" <dave.long@linaro.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/20140206053225.29635.15026.stgit@kbuild-fedora.yrl.intra.hitachi.co.jp Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2014-02-06 05:32:25 +00:00
if (ret != 0)
return ret; /* Found in debuginfo or got an error */
return find_probe_trace_events_from_map(pev, tevs);
}
int convert_perf_probe_events(struct perf_probe_event *pevs, int npevs)
{
int i, ret;
/* Loop 1: convert all events */
for (i = 0; i < npevs; i++) {
/* Init kprobe blacklist if needed */
if (!pevs[i].uprobes)
kprobe_blacklist__init();
/* Convert with or without debuginfo */
ret = convert_to_probe_trace_events(&pevs[i], &pevs[i].tevs);
if (ret < 0)
return ret;
pevs[i].ntevs = ret;
}
/* This just release blacklist only if allocated */
kprobe_blacklist__release();
return 0;
}
static int show_probe_trace_event(struct probe_trace_event *tev)
{
char *buf = synthesize_probe_trace_command(tev);
if (!buf) {
pr_debug("Failed to synthesize probe trace event.\n");
return -EINVAL;
}
/* Showing definition always go stdout */
printf("%s\n", buf);
free(buf);
return 0;
}
int show_probe_trace_events(struct perf_probe_event *pevs, int npevs)
{
struct strlist *namelist = strlist__new(NULL, NULL);
struct probe_trace_event *tev;
struct perf_probe_event *pev;
int i, j, ret = 0;
if (!namelist)
return -ENOMEM;
for (j = 0; j < npevs && !ret; j++) {
pev = &pevs[j];
for (i = 0; i < pev->ntevs && !ret; i++) {
tev = &pev->tevs[i];
/* Skip if the symbol is out of .text or blacklisted */
if (!tev->point.symbol && !pev->uprobes)
continue;
/* Set new name for tev (and update namelist) */
ret = probe_trace_event__set_name(tev, pev,
namelist, true);
if (!ret)
ret = show_probe_trace_event(tev);
}
}
strlist__delete(namelist);
return ret;
}
int apply_perf_probe_events(struct perf_probe_event *pevs, int npevs)
{
int i, ret = 0;
/* Loop 2: add all events */
for (i = 0; i < npevs; i++) {
ret = __add_probe_trace_events(&pevs[i], pevs[i].tevs,
pevs[i].ntevs,
probe_conf.force_add);
perf probe: Fix error propagation leading to segfault There are two hunks in this patch that stops probe processing as soon as one error is found, breaking out of loops, the other fix an error propagation that should return a negative error number but instead was returning the result of "ret < 0", which is 1 and thus made several error checks fail because they test agains < 0. The problem could be triggered by asking for a variable that was optimized out, fact that should stop the whole probe processing but instead was segfaulting while installing broken probes: [root@emilia ~]# probe perf_mmap:55 user_lock_limit Failed to find the location of user_lock_limit at this address. Perhaps, it has been optimized out. Failed to find 'user_lock_limit' in this function. Add new events: probe:perf_mmap (on perf_mmap:55 with user_lock_limit) probe:perf_mmap_1 (on perf_mmap:55 with user_lock_limit) Segmentation fault (core dumped) [root@emilia ~]# perf probe -l probe:perf_mmap (on perf_mmap:55@git/linux/kernel/perf_event.c with user_lock_limit) probe:perf_mmap_1 (on perf_mmap:55@git/linux/kernel/perf_event.c with user_lock_limit) [root@emilia ~]# After the fix: [root@emilia ~]# probe perf_mmap:55 user_lock_limit Failed to find the location of user_lock_limit at this address. Perhaps, it has been optimized out. Failed to find 'user_lock_limit' in this function. Error: Failed to add events. (-2) [root@emilia ~]# Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Tom Zanussi <tzanussi@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2011-02-21 16:23:57 +00:00
if (ret < 0)
break;
}
return ret;
}
void cleanup_perf_probe_events(struct perf_probe_event *pevs, int npevs)
{
int i, j;
struct perf_probe_event *pev;
/* Loop 3: cleanup and free trace events */
for (i = 0; i < npevs; i++) {
pev = &pevs[i];
for (j = 0; j < pevs[i].ntevs; j++)
clear_probe_trace_event(&pevs[i].tevs[j]);
zfree(&pevs[i].tevs);
pevs[i].ntevs = 0;
nsinfo__zput(pev->nsi);
clear_perf_probe_event(&pevs[i]);
}
}
int add_perf_probe_events(struct perf_probe_event *pevs, int npevs)
{
int ret;
ret = init_probe_symbol_maps(pevs->uprobes);
if (ret < 0)
return ret;
ret = convert_perf_probe_events(pevs, npevs);
if (ret == 0)
ret = apply_perf_probe_events(pevs, npevs);
cleanup_perf_probe_events(pevs, npevs);
exit_probe_symbol_maps();
return ret;
}
int del_perf_probe_events(struct strfilter *filter)
{
int ret, ret2, ufd = -1, kfd = -1;
char *str = strfilter__string(filter);
if (!str)
return -EINVAL;
/* Get current event names */
ret = probe_file__open_both(&kfd, &ufd, PF_FL_RW);
if (ret < 0)
goto out;
ret = probe_file__del_events(kfd, filter);
if (ret < 0 && ret != -ENOENT)
goto error;
ret2 = probe_file__del_events(ufd, filter);
if (ret2 < 0 && ret2 != -ENOENT) {
ret = ret2;
goto error;
}
ret = 0;
error:
if (kfd >= 0)
close(kfd);
if (ufd >= 0)
close(ufd);
out:
free(str);
return ret;
}
int show_available_funcs(const char *target, struct nsinfo *nsi,
struct strfilter *_filter, bool user)
{
struct rb_node *nd;
struct map *map;
int ret;
ret = init_probe_symbol_maps(user);
if (ret < 0)
return ret;
/* Get a symbol map */
map = get_target_map(target, nsi, user);
if (!map) {
pr_err("Failed to get a map for %s\n", (target) ? : "kernel");
return -EINVAL;
}
ret = map__load(map);
if (ret) {
if (ret == -2) {
char *str = strfilter__string(_filter);
pr_err("Failed to find symbols matched to \"%s\"\n",
str);
free(str);
} else
pr_err("Failed to load symbols in %s\n",
(target) ? : "kernel");
goto end;
}
if (!dso__sorted_by_name(map->dso))
dso__sort_by_name(map->dso);
/* Show all (filtered) symbols */
setup_pager();
for (nd = rb_first(&map->dso->symbol_names); nd; nd = rb_next(nd)) {
struct symbol_name_rb_node *pos = rb_entry(nd, struct symbol_name_rb_node, rb_node);
if (strfilter__compare(_filter, pos->sym.name))
printf("%s\n", pos->sym.name);
}
end:
map__put(map);
exit_probe_symbol_maps();
return ret;
}
perf probe: Support probing at absolute address It should be useful to allow 'perf probe' probe at absolute offset of a target. For example, when (u)probing at a instruction of a shared object in a embedded system where debuginfo is not avaliable but we know the offset of that instruction by manually digging. This patch enables following perf probe command syntax: # perf probe 0xffffffff811e6615 And # perf probe /lib/x86_64-linux-gnu/libc-2.19.so 0xeb860 In the above example, we don't need a anchor symbol, so it is possible to compute absolute addresses using other methods and then use 'perf probe' to create the probing points. v1 -> v2: Drop the leading '+' in cmdline; Allow uprobing at offset 0x0; Improve 'perf probe -l' result when uprobe at area without debuginfo. v2 -> v3: Split bugfix to a separated patch. Test result: # perf probe 0xffffffff8119d175 %ax # perf probe sys_write %ax # perf probe /lib64/libc-2.18.so 0x0 %ax # perf probe /lib64/libc-2.18.so 0x5 %ax # perf probe /lib64/libc-2.18.so 0xd8e40 %ax # perf probe /lib64/libc-2.18.so __write %ax # perf probe /lib64/libc-2.18.so 0xd8e49 %ax # cat /sys/kernel/debug/tracing/uprobe_events p:probe_libc/abs_0 /lib64/libc-2.18.so:0x (null) arg1=%ax p:probe_libc/abs_5 /lib64/libc-2.18.so:0x0000000000000005 arg1=%ax p:probe_libc/abs_d8e40 /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/__write /lib64/libc-2.18.so:0x00000000000d8e40 arg1=%ax p:probe_libc/abs_d8e49 /lib64/libc-2.18.so:0x00000000000d8e49 arg1=%ax # cat /sys/kernel/debug/tracing/kprobe_events p:probe/abs_ffffffff8119d175 0xffffffff8119d175 arg1=%ax p:probe/sys_write _text+1692016 arg1=%ax # perf probe -l Failed to find debug information for address 5 probe:abs_ffffffff8119d175 (on sys_write+5 with arg1) probe:sys_write (on sys_write with arg1) probe_libc:__write (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_0 (on 0x0 in /lib64/libc-2.18.so with arg1) probe_libc:abs_5 (on 0x5 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e40 (on @unix/syscall-template.S:81 in /lib64/libc-2.18.so with arg1) probe_libc:abs_d8e49 (on __GI___libc_write+9 in /lib64/libc-2.18.so with arg1) Signed-off-by: Wang Nan <wangnan0@huawei.com> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1440586666-235233-7-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-08-26 10:57:45 +00:00
int copy_to_probe_trace_arg(struct probe_trace_arg *tvar,
struct perf_probe_arg *pvar)
{
tvar->value = strdup(pvar->var);
if (tvar->value == NULL)
return -ENOMEM;
if (pvar->type) {
tvar->type = strdup(pvar->type);
if (tvar->type == NULL)
return -ENOMEM;
}
if (pvar->name) {
tvar->name = strdup(pvar->name);
if (tvar->name == NULL)
return -ENOMEM;
} else
tvar->name = NULL;
return 0;
}