2015-10-14 20:41:13 +08:00
|
|
|
/*
|
|
|
|
* bpf-loader.c
|
|
|
|
*
|
|
|
|
* Copyright (C) 2015 Wang Nan <wangnan0@huawei.com>
|
|
|
|
* Copyright (C) 2015 Huawei Inc.
|
|
|
|
*/
|
|
|
|
|
2015-11-16 20:10:13 +08:00
|
|
|
#include <linux/bpf.h>
|
2015-10-14 20:41:13 +08:00
|
|
|
#include <bpf/libbpf.h>
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
#include <bpf/bpf.h>
|
2015-10-14 20:41:13 +08:00
|
|
|
#include <linux/err.h>
|
2015-11-16 20:10:08 +08:00
|
|
|
#include <linux/string.h>
|
2015-10-14 20:41:13 +08:00
|
|
|
#include "perf.h"
|
|
|
|
#include "debug.h"
|
|
|
|
#include "bpf-loader.h"
|
2015-11-16 20:10:13 +08:00
|
|
|
#include "bpf-prologue.h"
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
#include "probe-event.h"
|
|
|
|
#include "probe-finder.h" // for MAX_PROBES
|
2016-02-22 17:10:34 +08:00
|
|
|
#include "parse-events.h"
|
2015-10-14 20:41:20 +08:00
|
|
|
#include "llvm-utils.h"
|
2016-11-26 15:03:39 +08:00
|
|
|
#include "c++/clang-c.h"
|
2015-10-14 20:41:13 +08:00
|
|
|
|
|
|
|
#define DEFINE_PRINT_FN(name, level) \
|
|
|
|
static int libbpf_##name(const char *fmt, ...) \
|
|
|
|
{ \
|
|
|
|
va_list args; \
|
|
|
|
int ret; \
|
|
|
|
\
|
|
|
|
va_start(args, fmt); \
|
|
|
|
ret = veprintf(level, verbose, pr_fmt(fmt), args);\
|
|
|
|
va_end(args); \
|
|
|
|
return ret; \
|
|
|
|
}
|
|
|
|
|
2015-11-03 18:44:43 +08:00
|
|
|
DEFINE_PRINT_FN(warning, 1)
|
|
|
|
DEFINE_PRINT_FN(info, 1)
|
2015-10-14 20:41:13 +08:00
|
|
|
DEFINE_PRINT_FN(debug, 1)
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
struct bpf_prog_priv {
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
bool is_tp;
|
|
|
|
char *sys_name;
|
|
|
|
char *evt_name;
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
struct perf_probe_event pev;
|
2015-11-16 20:10:13 +08:00
|
|
|
bool need_prologue;
|
|
|
|
struct bpf_insn *insns_buf;
|
2015-11-17 16:32:47 +08:00
|
|
|
int nr_types;
|
|
|
|
int *type_mapping;
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
};
|
|
|
|
|
2015-11-06 21:49:43 +08:00
|
|
|
static bool libbpf_initialized;
|
|
|
|
|
|
|
|
struct bpf_object *
|
|
|
|
bpf__prepare_load_buffer(void *obj_buf, size_t obj_buf_sz, const char *name)
|
|
|
|
{
|
|
|
|
struct bpf_object *obj;
|
|
|
|
|
|
|
|
if (!libbpf_initialized) {
|
|
|
|
libbpf_set_print(libbpf_warning,
|
|
|
|
libbpf_info,
|
|
|
|
libbpf_debug);
|
|
|
|
libbpf_initialized = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
obj = bpf_object__open_buffer(obj_buf, obj_buf_sz, name);
|
|
|
|
if (IS_ERR(obj)) {
|
|
|
|
pr_debug("bpf: failed to load buffer\n");
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
}
|
|
|
|
|
|
|
|
return obj;
|
|
|
|
}
|
|
|
|
|
2015-10-14 20:41:20 +08:00
|
|
|
struct bpf_object *bpf__prepare_load(const char *filename, bool source)
|
2015-10-14 20:41:13 +08:00
|
|
|
{
|
|
|
|
struct bpf_object *obj;
|
|
|
|
|
|
|
|
if (!libbpf_initialized) {
|
|
|
|
libbpf_set_print(libbpf_warning,
|
|
|
|
libbpf_info,
|
|
|
|
libbpf_debug);
|
|
|
|
libbpf_initialized = true;
|
|
|
|
}
|
|
|
|
|
2015-10-14 20:41:20 +08:00
|
|
|
if (source) {
|
|
|
|
int err;
|
|
|
|
void *obj_buf;
|
|
|
|
size_t obj_buf_sz;
|
|
|
|
|
2016-11-26 15:03:39 +08:00
|
|
|
perf_clang__init();
|
|
|
|
err = perf_clang__compile_bpf(filename, &obj_buf, &obj_buf_sz);
|
|
|
|
perf_clang__cleanup();
|
|
|
|
if (err) {
|
|
|
|
pr_warning("bpf: builtin compilation failed: %d, try external compiler\n", err);
|
|
|
|
err = llvm__compile_bpf(filename, &obj_buf, &obj_buf_sz);
|
|
|
|
if (err)
|
|
|
|
return ERR_PTR(-BPF_LOADER_ERRNO__COMPILE);
|
|
|
|
} else
|
|
|
|
pr_debug("bpf: successfull builtin compilation\n");
|
2015-10-14 20:41:20 +08:00
|
|
|
obj = bpf_object__open_buffer(obj_buf, obj_buf_sz, filename);
|
2016-11-26 15:03:30 +08:00
|
|
|
|
|
|
|
if (!IS_ERR(obj) && llvm_param.dump_obj)
|
|
|
|
llvm__dump_obj(filename, obj_buf, obj_buf_sz);
|
|
|
|
|
2015-10-14 20:41:20 +08:00
|
|
|
free(obj_buf);
|
|
|
|
} else
|
|
|
|
obj = bpf_object__open(filename);
|
|
|
|
|
2015-11-06 21:49:37 +08:00
|
|
|
if (IS_ERR(obj)) {
|
2015-10-14 20:41:13 +08:00
|
|
|
pr_debug("bpf: failed to load %s\n", filename);
|
2015-11-06 21:49:37 +08:00
|
|
|
return obj;
|
2015-10-14 20:41:13 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
return obj;
|
|
|
|
}
|
|
|
|
|
|
|
|
void bpf__clear(void)
|
|
|
|
{
|
|
|
|
struct bpf_object *obj, *tmp;
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
bpf_object__for_each_safe(obj, tmp) {
|
|
|
|
bpf__unprobe(obj);
|
2015-10-14 20:41:13 +08:00
|
|
|
bpf_object__close(obj);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2016-02-19 19:43:51 +08:00
|
|
|
clear_prog_priv(struct bpf_program *prog __maybe_unused,
|
|
|
|
void *_priv)
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
{
|
|
|
|
struct bpf_prog_priv *priv = _priv;
|
|
|
|
|
|
|
|
cleanup_perf_probe_events(&priv->pev, 1);
|
2015-11-16 20:10:13 +08:00
|
|
|
zfree(&priv->insns_buf);
|
2015-11-17 16:32:47 +08:00
|
|
|
zfree(&priv->type_mapping);
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
zfree(&priv->sys_name);
|
|
|
|
zfree(&priv->evt_name);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
free(priv);
|
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:05 +08:00
|
|
|
static int
|
2015-11-27 16:47:37 +08:00
|
|
|
prog_config__exec(const char *value, struct perf_probe_event *pev)
|
2015-11-16 20:10:05 +08:00
|
|
|
{
|
|
|
|
pev->uprobes = true;
|
|
|
|
pev->target = strdup(value);
|
|
|
|
if (!pev->target)
|
|
|
|
return -ENOMEM;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:06 +08:00
|
|
|
static int
|
2015-11-27 16:47:37 +08:00
|
|
|
prog_config__module(const char *value, struct perf_probe_event *pev)
|
2015-11-16 20:10:06 +08:00
|
|
|
{
|
|
|
|
pev->uprobes = false;
|
|
|
|
pev->target = strdup(value);
|
|
|
|
if (!pev->target)
|
|
|
|
return -ENOMEM;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:08 +08:00
|
|
|
static int
|
2015-11-27 16:47:37 +08:00
|
|
|
prog_config__bool(const char *value, bool *pbool, bool invert)
|
2015-11-16 20:10:08 +08:00
|
|
|
{
|
|
|
|
int err;
|
|
|
|
bool bool_value;
|
|
|
|
|
|
|
|
if (!pbool)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
err = strtobool(value, &bool_value);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
*pbool = invert ? !bool_value : bool_value;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2015-11-27 16:47:37 +08:00
|
|
|
prog_config__inlines(const char *value,
|
|
|
|
struct perf_probe_event *pev __maybe_unused)
|
2015-11-16 20:10:08 +08:00
|
|
|
{
|
2015-11-27 16:47:37 +08:00
|
|
|
return prog_config__bool(value, &probe_conf.no_inlines, true);
|
2015-11-16 20:10:08 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2015-11-27 16:47:37 +08:00
|
|
|
prog_config__force(const char *value,
|
|
|
|
struct perf_probe_event *pev __maybe_unused)
|
2015-11-16 20:10:08 +08:00
|
|
|
{
|
2015-11-27 16:47:37 +08:00
|
|
|
return prog_config__bool(value, &probe_conf.force_add, false);
|
2015-11-16 20:10:08 +08:00
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:05 +08:00
|
|
|
static struct {
|
|
|
|
const char *key;
|
|
|
|
const char *usage;
|
|
|
|
const char *desc;
|
|
|
|
int (*func)(const char *, struct perf_probe_event *);
|
2015-11-27 16:47:37 +08:00
|
|
|
} bpf_prog_config_terms[] = {
|
2015-11-16 20:10:05 +08:00
|
|
|
{
|
|
|
|
.key = "exec",
|
|
|
|
.usage = "exec=<full path of file>",
|
|
|
|
.desc = "Set uprobe target",
|
2015-11-27 16:47:37 +08:00
|
|
|
.func = prog_config__exec,
|
2015-11-16 20:10:05 +08:00
|
|
|
},
|
2015-11-16 20:10:06 +08:00
|
|
|
{
|
|
|
|
.key = "module",
|
|
|
|
.usage = "module=<module name> ",
|
|
|
|
.desc = "Set kprobe module",
|
2015-11-27 16:47:37 +08:00
|
|
|
.func = prog_config__module,
|
2015-11-16 20:10:08 +08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.key = "inlines",
|
|
|
|
.usage = "inlines=[yes|no] ",
|
|
|
|
.desc = "Probe at inline symbol",
|
2015-11-27 16:47:37 +08:00
|
|
|
.func = prog_config__inlines,
|
2015-11-16 20:10:08 +08:00
|
|
|
},
|
|
|
|
{
|
|
|
|
.key = "force",
|
|
|
|
.usage = "force=[yes|no] ",
|
|
|
|
.desc = "Forcibly add events with existing name",
|
2015-11-27 16:47:37 +08:00
|
|
|
.func = prog_config__force,
|
2015-11-16 20:10:08 +08:00
|
|
|
},
|
2015-11-16 20:10:05 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
static int
|
2015-11-27 16:47:37 +08:00
|
|
|
do_prog_config(const char *key, const char *value,
|
|
|
|
struct perf_probe_event *pev)
|
2015-11-16 20:10:05 +08:00
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
pr_debug("config bpf program: %s=%s\n", key, value);
|
2015-11-27 16:47:37 +08:00
|
|
|
for (i = 0; i < ARRAY_SIZE(bpf_prog_config_terms); i++)
|
|
|
|
if (strcmp(key, bpf_prog_config_terms[i].key) == 0)
|
|
|
|
return bpf_prog_config_terms[i].func(value, pev);
|
2015-11-16 20:10:05 +08:00
|
|
|
|
2015-11-27 16:47:37 +08:00
|
|
|
pr_debug("BPF: ERROR: invalid program config option: %s=%s\n",
|
2015-11-16 20:10:05 +08:00
|
|
|
key, value);
|
|
|
|
|
2015-11-27 16:47:37 +08:00
|
|
|
pr_debug("\nHint: Valid options are:\n");
|
|
|
|
for (i = 0; i < ARRAY_SIZE(bpf_prog_config_terms); i++)
|
|
|
|
pr_debug("\t%s:\t%s\n", bpf_prog_config_terms[i].usage,
|
|
|
|
bpf_prog_config_terms[i].desc);
|
2015-11-16 20:10:05 +08:00
|
|
|
pr_debug("\n");
|
|
|
|
|
2015-11-27 16:47:37 +08:00
|
|
|
return -BPF_LOADER_ERRNO__PROGCONF_TERM;
|
2015-11-16 20:10:05 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static const char *
|
2015-11-27 16:47:37 +08:00
|
|
|
parse_prog_config_kvpair(const char *config_str, struct perf_probe_event *pev)
|
2015-11-16 20:10:05 +08:00
|
|
|
{
|
|
|
|
char *text = strdup(config_str);
|
|
|
|
char *sep, *line;
|
|
|
|
const char *main_str = NULL;
|
|
|
|
int err = 0;
|
|
|
|
|
|
|
|
if (!text) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory: dup config_str failed\n");
|
2015-11-16 20:10:05 +08:00
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
|
|
|
|
line = text;
|
|
|
|
while ((sep = strchr(line, ';'))) {
|
|
|
|
char *equ;
|
|
|
|
|
|
|
|
*sep = '\0';
|
|
|
|
equ = strchr(line, '=');
|
|
|
|
if (!equ) {
|
|
|
|
pr_warning("WARNING: invalid config in BPF object: %s\n",
|
|
|
|
line);
|
|
|
|
pr_warning("\tShould be 'key=value'.\n");
|
|
|
|
goto nextline;
|
|
|
|
}
|
|
|
|
*equ = '\0';
|
|
|
|
|
2015-11-27 16:47:37 +08:00
|
|
|
err = do_prog_config(line, equ + 1, pev);
|
2015-11-16 20:10:05 +08:00
|
|
|
if (err)
|
|
|
|
break;
|
|
|
|
nextline:
|
|
|
|
line = sep + 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!err)
|
|
|
|
main_str = config_str + (line - text);
|
|
|
|
free(text);
|
|
|
|
|
|
|
|
return err ? ERR_PTR(err) : main_str;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
parse_prog_config(const char *config_str, const char **p_main_str,
|
|
|
|
bool *is_tp, struct perf_probe_event *pev)
|
2015-11-16 20:10:05 +08:00
|
|
|
{
|
|
|
|
int err;
|
2015-11-27 16:47:37 +08:00
|
|
|
const char *main_str = parse_prog_config_kvpair(config_str, pev);
|
2015-11-16 20:10:05 +08:00
|
|
|
|
|
|
|
if (IS_ERR(main_str))
|
|
|
|
return PTR_ERR(main_str);
|
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
*p_main_str = main_str;
|
|
|
|
if (!strchr(main_str, '=')) {
|
|
|
|
/* Is a tracepoint event? */
|
|
|
|
const char *s = strchr(main_str, ':');
|
|
|
|
|
|
|
|
if (!s) {
|
|
|
|
pr_debug("bpf: '%s' is not a valid tracepoint\n",
|
|
|
|
config_str);
|
|
|
|
return -BPF_LOADER_ERRNO__CONFIG;
|
|
|
|
}
|
|
|
|
|
|
|
|
*is_tp = true;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
*is_tp = false;
|
2015-11-16 20:10:05 +08:00
|
|
|
err = parse_perf_probe_command(main_str, pev);
|
|
|
|
if (err < 0) {
|
|
|
|
pr_debug("bpf: '%s' is not a valid config string\n",
|
|
|
|
config_str);
|
|
|
|
/* parse failed, don't need clear pev. */
|
|
|
|
return -BPF_LOADER_ERRNO__CONFIG;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
static int
|
|
|
|
config_bpf_program(struct bpf_program *prog)
|
|
|
|
{
|
|
|
|
struct perf_probe_event *pev = NULL;
|
|
|
|
struct bpf_prog_priv *priv = NULL;
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
const char *config_str, *main_str;
|
|
|
|
bool is_tp = false;
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
int err;
|
|
|
|
|
2015-11-16 20:10:08 +08:00
|
|
|
/* Initialize per-program probing setting */
|
|
|
|
probe_conf.no_inlines = false;
|
|
|
|
probe_conf.force_add = false;
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
config_str = bpf_program__title(prog, false);
|
2015-11-06 21:49:37 +08:00
|
|
|
if (IS_ERR(config_str)) {
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
pr_debug("bpf: unable to get title for program\n");
|
2015-11-06 21:49:37 +08:00
|
|
|
return PTR_ERR(config_str);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
priv = calloc(sizeof(*priv), 1);
|
|
|
|
if (!priv) {
|
|
|
|
pr_debug("bpf: failed to alloc priv\n");
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
pev = &priv->pev;
|
|
|
|
|
|
|
|
pr_debug("bpf: config program '%s'\n", config_str);
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
err = parse_prog_config(config_str, &main_str, &is_tp, pev);
|
2015-11-16 20:10:05 +08:00
|
|
|
if (err)
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
goto errout;
|
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
if (is_tp) {
|
|
|
|
char *s = strchr(main_str, ':');
|
|
|
|
|
|
|
|
priv->is_tp = true;
|
|
|
|
priv->sys_name = strndup(main_str, s - main_str);
|
|
|
|
priv->evt_name = strdup(s + 1);
|
|
|
|
goto set_priv;
|
|
|
|
}
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
if (pev->group && strcmp(pev->group, PERF_BPF_PROBE_GROUP)) {
|
|
|
|
pr_debug("bpf: '%s': group for event is set and not '%s'.\n",
|
|
|
|
config_str, PERF_BPF_PROBE_GROUP);
|
2015-11-06 21:58:09 +08:00
|
|
|
err = -BPF_LOADER_ERRNO__GROUP;
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
goto errout;
|
|
|
|
} else if (!pev->group)
|
|
|
|
pev->group = strdup(PERF_BPF_PROBE_GROUP);
|
|
|
|
|
|
|
|
if (!pev->group) {
|
|
|
|
pr_debug("bpf: strdup failed\n");
|
|
|
|
err = -ENOMEM;
|
|
|
|
goto errout;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!pev->event) {
|
2015-11-06 21:58:09 +08:00
|
|
|
pr_debug("bpf: '%s': event name is missing. Section name should be 'key=value'\n",
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
config_str);
|
2015-11-06 21:58:09 +08:00
|
|
|
err = -BPF_LOADER_ERRNO__EVENTNAME;
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
goto errout;
|
|
|
|
}
|
|
|
|
pr_debug("bpf: config '%s' is ok\n", config_str);
|
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
set_priv:
|
2016-06-03 23:38:21 +08:00
|
|
|
err = bpf_program__set_priv(prog, priv, clear_prog_priv);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
if (err) {
|
|
|
|
pr_debug("Failed to set priv for program '%s'\n", config_str);
|
|
|
|
goto errout;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
errout:
|
|
|
|
if (pev)
|
|
|
|
clear_perf_probe_event(pev);
|
|
|
|
free(priv);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int bpf__prepare_probe(void)
|
|
|
|
{
|
|
|
|
static int err = 0;
|
|
|
|
static bool initialized = false;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Make err static, so if init failed the first, bpf__prepare_probe()
|
|
|
|
* fails each time without calling init_probe_symbol_maps multiple
|
|
|
|
* times.
|
|
|
|
*/
|
|
|
|
if (initialized)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
initialized = true;
|
|
|
|
err = init_probe_symbol_maps(false);
|
|
|
|
if (err < 0)
|
|
|
|
pr_debug("Failed to init_probe_symbol_maps\n");
|
|
|
|
probe_conf.max_probes = MAX_PROBES;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:13 +08:00
|
|
|
static int
|
|
|
|
preproc_gen_prologue(struct bpf_program *prog, int n,
|
|
|
|
struct bpf_insn *orig_insns, int orig_insns_cnt,
|
|
|
|
struct bpf_prog_prep_result *res)
|
|
|
|
{
|
2016-06-03 23:36:39 +08:00
|
|
|
struct bpf_prog_priv *priv = bpf_program__priv(prog);
|
2015-11-16 20:10:13 +08:00
|
|
|
struct probe_trace_event *tev;
|
|
|
|
struct perf_probe_event *pev;
|
|
|
|
struct bpf_insn *buf;
|
|
|
|
size_t prologue_cnt = 0;
|
2015-11-17 16:32:47 +08:00
|
|
|
int i, err;
|
2015-11-16 20:10:13 +08:00
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
if (IS_ERR(priv) || !priv || priv->is_tp)
|
2015-11-16 20:10:13 +08:00
|
|
|
goto errout;
|
|
|
|
|
|
|
|
pev = &priv->pev;
|
|
|
|
|
2015-11-17 16:32:47 +08:00
|
|
|
if (n < 0 || n >= priv->nr_types)
|
2015-11-16 20:10:13 +08:00
|
|
|
goto errout;
|
|
|
|
|
2015-11-17 16:32:47 +08:00
|
|
|
/* Find a tev belongs to that type */
|
|
|
|
for (i = 0; i < pev->ntevs; i++) {
|
|
|
|
if (priv->type_mapping[i] == n)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (i >= pev->ntevs) {
|
|
|
|
pr_debug("Internal error: prologue type %d not found\n", n);
|
|
|
|
return -BPF_LOADER_ERRNO__PROLOGUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
tev = &pev->tevs[i];
|
2015-11-16 20:10:13 +08:00
|
|
|
|
|
|
|
buf = priv->insns_buf;
|
|
|
|
err = bpf__gen_prologue(tev->args, tev->nargs,
|
|
|
|
buf, &prologue_cnt,
|
|
|
|
BPF_MAXINSNS - orig_insns_cnt);
|
|
|
|
if (err) {
|
|
|
|
const char *title;
|
|
|
|
|
|
|
|
title = bpf_program__title(prog, false);
|
|
|
|
if (!title)
|
|
|
|
title = "[unknown]";
|
|
|
|
|
|
|
|
pr_debug("Failed to generate prologue for program %s\n",
|
|
|
|
title);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
memcpy(&buf[prologue_cnt], orig_insns,
|
|
|
|
sizeof(struct bpf_insn) * orig_insns_cnt);
|
|
|
|
|
|
|
|
res->new_insn_ptr = buf;
|
|
|
|
res->new_insn_cnt = prologue_cnt + orig_insns_cnt;
|
|
|
|
res->pfd = NULL;
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
errout:
|
|
|
|
pr_debug("Internal error in preproc_gen_prologue\n");
|
|
|
|
return -BPF_LOADER_ERRNO__PROLOGUE;
|
|
|
|
}
|
|
|
|
|
2015-11-17 16:32:47 +08:00
|
|
|
/*
|
|
|
|
* compare_tev_args is reflexive, transitive and antisymmetric.
|
|
|
|
* I can proof it but this margin is too narrow to contain.
|
|
|
|
*/
|
|
|
|
static int compare_tev_args(const void *ptev1, const void *ptev2)
|
|
|
|
{
|
|
|
|
int i, ret;
|
|
|
|
const struct probe_trace_event *tev1 =
|
|
|
|
*(const struct probe_trace_event **)ptev1;
|
|
|
|
const struct probe_trace_event *tev2 =
|
|
|
|
*(const struct probe_trace_event **)ptev2;
|
|
|
|
|
|
|
|
ret = tev2->nargs - tev1->nargs;
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
for (i = 0; i < tev1->nargs; i++) {
|
|
|
|
struct probe_trace_arg *arg1, *arg2;
|
|
|
|
struct probe_trace_arg_ref *ref1, *ref2;
|
|
|
|
|
|
|
|
arg1 = &tev1->args[i];
|
|
|
|
arg2 = &tev2->args[i];
|
|
|
|
|
|
|
|
ret = strcmp(arg1->value, arg2->value);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ref1 = arg1->ref;
|
|
|
|
ref2 = arg2->ref;
|
|
|
|
|
|
|
|
while (ref1 && ref2) {
|
|
|
|
ret = ref2->offset - ref1->offset;
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ref1 = ref1->next;
|
|
|
|
ref2 = ref2->next;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ref1 || ref2)
|
|
|
|
return ref2 ? 1 : -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Assign a type number to each tevs in a pev.
|
|
|
|
* mapping is an array with same slots as tevs in that pev.
|
|
|
|
* nr_types will be set to number of types.
|
|
|
|
*/
|
|
|
|
static int map_prologue(struct perf_probe_event *pev, int *mapping,
|
|
|
|
int *nr_types)
|
|
|
|
{
|
|
|
|
int i, type = 0;
|
|
|
|
struct probe_trace_event **ptevs;
|
|
|
|
|
|
|
|
size_t array_sz = sizeof(*ptevs) * pev->ntevs;
|
|
|
|
|
|
|
|
ptevs = malloc(array_sz);
|
|
|
|
if (!ptevs) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory: alloc ptevs failed\n");
|
2015-11-17 16:32:47 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
pr_debug("In map_prologue, ntevs=%d\n", pev->ntevs);
|
|
|
|
for (i = 0; i < pev->ntevs; i++)
|
|
|
|
ptevs[i] = &pev->tevs[i];
|
|
|
|
|
|
|
|
qsort(ptevs, pev->ntevs, sizeof(*ptevs),
|
|
|
|
compare_tev_args);
|
|
|
|
|
|
|
|
for (i = 0; i < pev->ntevs; i++) {
|
|
|
|
int n;
|
|
|
|
|
|
|
|
n = ptevs[i] - pev->tevs;
|
|
|
|
if (i == 0) {
|
|
|
|
mapping[n] = type;
|
|
|
|
pr_debug("mapping[%d]=%d\n", n, type);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (compare_tev_args(ptevs + i, ptevs + i - 1) == 0)
|
|
|
|
mapping[n] = type;
|
|
|
|
else
|
|
|
|
mapping[n] = ++type;
|
|
|
|
|
|
|
|
pr_debug("mapping[%d]=%d\n", n, mapping[n]);
|
|
|
|
}
|
|
|
|
free(ptevs);
|
|
|
|
*nr_types = type + 1;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:13 +08:00
|
|
|
static int hook_load_preprocessor(struct bpf_program *prog)
|
|
|
|
{
|
2016-06-03 23:36:39 +08:00
|
|
|
struct bpf_prog_priv *priv = bpf_program__priv(prog);
|
2015-11-16 20:10:13 +08:00
|
|
|
struct perf_probe_event *pev;
|
|
|
|
bool need_prologue = false;
|
|
|
|
int err, i;
|
|
|
|
|
2016-06-03 23:36:39 +08:00
|
|
|
if (IS_ERR(priv) || !priv) {
|
2015-11-16 20:10:13 +08:00
|
|
|
pr_debug("Internal error when hook preprocessor\n");
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
if (priv->is_tp) {
|
|
|
|
priv->need_prologue = false;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-11-16 20:10:13 +08:00
|
|
|
pev = &priv->pev;
|
|
|
|
for (i = 0; i < pev->ntevs; i++) {
|
|
|
|
struct probe_trace_event *tev = &pev->tevs[i];
|
|
|
|
|
|
|
|
if (tev->nargs > 0) {
|
|
|
|
need_prologue = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Since all tevs don't have argument, we don't need generate
|
|
|
|
* prologue.
|
|
|
|
*/
|
|
|
|
if (!need_prologue) {
|
|
|
|
priv->need_prologue = false;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
priv->need_prologue = true;
|
|
|
|
priv->insns_buf = malloc(sizeof(struct bpf_insn) * BPF_MAXINSNS);
|
|
|
|
if (!priv->insns_buf) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory: alloc insns_buf failed\n");
|
2015-11-16 20:10:13 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
2015-11-17 16:32:47 +08:00
|
|
|
priv->type_mapping = malloc(sizeof(int) * pev->ntevs);
|
|
|
|
if (!priv->type_mapping) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory: alloc type_mapping failed\n");
|
2015-11-17 16:32:47 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
memset(priv->type_mapping, -1,
|
|
|
|
sizeof(int) * pev->ntevs);
|
|
|
|
|
|
|
|
err = map_prologue(pev, priv->type_mapping, &priv->nr_types);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
err = bpf_program__set_prep(prog, priv->nr_types,
|
2015-11-16 20:10:13 +08:00
|
|
|
preproc_gen_prologue);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
int bpf__probe(struct bpf_object *obj)
|
|
|
|
{
|
|
|
|
int err = 0;
|
|
|
|
struct bpf_program *prog;
|
|
|
|
struct bpf_prog_priv *priv;
|
|
|
|
struct perf_probe_event *pev;
|
|
|
|
|
|
|
|
err = bpf__prepare_probe();
|
|
|
|
if (err) {
|
|
|
|
pr_debug("bpf__prepare_probe failed\n");
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
bpf_object__for_each_program(prog, obj) {
|
|
|
|
err = config_bpf_program(prog);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
2016-06-03 23:36:39 +08:00
|
|
|
priv = bpf_program__priv(prog);
|
|
|
|
if (IS_ERR(priv) || !priv) {
|
|
|
|
err = PTR_ERR(priv);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
goto out;
|
2016-06-03 23:36:39 +08:00
|
|
|
}
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
|
|
|
|
if (priv->is_tp) {
|
|
|
|
bpf_program__set_tracepoint(prog);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
bpf_program__set_kprobe(prog);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
pev = &priv->pev;
|
|
|
|
|
|
|
|
err = convert_perf_probe_events(pev, 1);
|
|
|
|
if (err < 0) {
|
2017-02-07 15:34:12 +08:00
|
|
|
pr_debug("bpf_probe: failed to convert perf probe events\n");
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = apply_perf_probe_events(pev, 1);
|
|
|
|
if (err < 0) {
|
2017-02-07 15:34:12 +08:00
|
|
|
pr_debug("bpf_probe: failed to apply perf probe events\n");
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
goto out;
|
|
|
|
}
|
2015-11-16 20:10:13 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* After probing, let's consider prologue, which
|
|
|
|
* adds program fetcher to BPF programs.
|
|
|
|
*
|
|
|
|
* hook_load_preprocessorr() hooks pre-processor
|
|
|
|
* to bpf_program, let it generate prologue
|
|
|
|
* dynamically during loading.
|
|
|
|
*/
|
|
|
|
err = hook_load_preprocessor(prog);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
}
|
|
|
|
out:
|
|
|
|
return err < 0 ? err : 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#define EVENTS_WRITE_BUFSIZE 4096
|
|
|
|
int bpf__unprobe(struct bpf_object *obj)
|
|
|
|
{
|
|
|
|
int err, ret = 0;
|
|
|
|
struct bpf_program *prog;
|
|
|
|
|
|
|
|
bpf_object__for_each_program(prog, obj) {
|
2016-06-03 23:36:39 +08:00
|
|
|
struct bpf_prog_priv *priv = bpf_program__priv(prog);
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
int i;
|
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
if (IS_ERR(priv) || !priv || priv->is_tp)
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
continue;
|
|
|
|
|
|
|
|
for (i = 0; i < priv->pev.ntevs; i++) {
|
|
|
|
struct probe_trace_event *tev = &priv->pev.tevs[i];
|
|
|
|
char name_buf[EVENTS_WRITE_BUFSIZE];
|
|
|
|
struct strfilter *delfilter;
|
|
|
|
|
|
|
|
snprintf(name_buf, EVENTS_WRITE_BUFSIZE,
|
|
|
|
"%s:%s", tev->group, tev->event);
|
|
|
|
name_buf[EVENTS_WRITE_BUFSIZE - 1] = '\0';
|
|
|
|
|
|
|
|
delfilter = strfilter__new(name_buf, NULL);
|
|
|
|
if (!delfilter) {
|
|
|
|
pr_debug("Failed to create filter for unprobing\n");
|
|
|
|
ret = -ENOMEM;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = del_perf_probe_events(delfilter);
|
|
|
|
strfilter__delete(delfilter);
|
|
|
|
if (err) {
|
|
|
|
pr_debug("Failed to delete %s\n", name_buf);
|
|
|
|
ret = err;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
perf tools: Load eBPF object into kernel
This patch utilizes bpf_object__load() provided by libbpf to load all
objects into kernel.
Committer notes:
Testing it:
When using an incorrect kernel version number, i.e., having this in your
eBPF proggie:
int _version __attribute__((section("version"), used)) = 0x40100;
For a 4.3.0-rc6+ kernel, say, this happens and needs checking at event
parsing time, to provide a better error report to the user:
# perf record --event /tmp/foo.o sleep 1
libbpf: load bpf program failed: Invalid argument
libbpf: -- BEGIN DUMP LOG ---
libbpf:
libbpf: -- END LOG --
libbpf: failed to load program 'fork=_do_fork'
libbpf: failed to load object '/tmp/foo.o'
event syntax error: '/tmp/foo.o'
\___ Invalid argument: Are you root and runing a CONFIG_BPF_SYSCALL kernel?
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
If we instead make it match, i.e. use 0x40300 on this v4.3.0-rc6+
kernel, the whole process goes thru:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.202 MB perf.data ]
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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: Paul Mackerras <paulus@samba.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/1444826502-49291-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:16 +08:00
|
|
|
int bpf__load(struct bpf_object *obj)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
|
|
|
err = bpf_object__load(obj);
|
|
|
|
if (err) {
|
|
|
|
pr_debug("bpf: load objects failed\n");
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-07-13 18:44:04 +08:00
|
|
|
int bpf__foreach_event(struct bpf_object *obj,
|
|
|
|
bpf_prog_iter_callback_t func,
|
|
|
|
void *arg)
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
{
|
|
|
|
struct bpf_program *prog;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
bpf_object__for_each_program(prog, obj) {
|
2016-06-03 23:36:39 +08:00
|
|
|
struct bpf_prog_priv *priv = bpf_program__priv(prog);
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
struct probe_trace_event *tev;
|
|
|
|
struct perf_probe_event *pev;
|
|
|
|
int i, fd;
|
|
|
|
|
2016-06-03 23:36:39 +08:00
|
|
|
if (IS_ERR(priv) || !priv) {
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
pr_debug("bpf: failed to get private field\n");
|
2015-11-06 21:58:09 +08:00
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
}
|
|
|
|
|
perf bpf: Support BPF program attach to tracepoints
To support 98b5c2c65c29 ("perf, bpf: allow bpf programs attach to
tracepoints"), this patch allows BPF scripts to select tracepoints in
their section name.
Example:
# cat test_tracepoint.c
/*********************************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("raw_syscalls:sys_enter")
int func(void *ctx)
{
/*
* /sys/kernel/debug/tracing/events/raw_syscalls/sys_enter/format:
* ...
* field:long id; offset:8; size:8; signed:1;
* ...
* ctx + 8 select 'id'
*/
u64 id = *((u64 *)(ctx + 8));
if (id == 1)
return 1;
return 0;
}
SEC("_write=sys_write")
int _write(void *ctx)
{
return 1;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/*********************************************/
# perf record -e ./test_tracepoint.c dd if=/dev/zero of=/dev/null count=5
5+0 records in
5+0 records out
2560 bytes (2.6 kB) copied, 6.2281e-05 s, 41.1 MB/s
[ perf record: Woken up 1 times to write data ]
# perf script
dd 13436 [005] 1596.490869: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, 7ffe82470d60, ffffffffffffe020, fffff
dd 13436 [005] 1596.490871: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490873: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490874: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490876: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490876: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490878: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490879: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490881: raw_syscalls:sys_enter: NR 1 (1, 178d000, 200, ffffffffffffe000, ffffffffffffe020, f
dd 13436 [005] 1596.490882: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490900: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1f, 40acb8, 7f44bac74700, 7f44baa4fba
dd 13436 [005] 1596.490901: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490917: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffffa, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490918: perf_bpf_probe:_write: (ffffffff812351e0)
dd 13436 [005] 1596.490932: raw_syscalls:sys_enter: NR 1 (2, 7ffe8246e640, 1a, fffffff9, 7f44bac74700, 7f44baa4f
dd 13436 [005] 1596.490933: perf_bpf_probe:_write: (ffffffff812351e0)
Committer note:
Further testing:
# trace --no-sys --event /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 raw_syscalls:sys_enter:NR 1 (1, 7f0490504000, c48, 7f0490503010, ffffffffffffffff, 0))
0.006 perf_bpf_probe:_write:(ffffffff81241bc0))
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468406646-21642-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-13 18:44:05 +08:00
|
|
|
if (priv->is_tp) {
|
|
|
|
fd = bpf_program__fd(prog);
|
|
|
|
err = (*func)(priv->sys_name, priv->evt_name, fd, arg);
|
|
|
|
if (err) {
|
|
|
|
pr_debug("bpf: tracepoint call back failed, stop iterate\n");
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
pev = &priv->pev;
|
|
|
|
for (i = 0; i < pev->ntevs; i++) {
|
|
|
|
tev = &pev->tevs[i];
|
|
|
|
|
2015-11-17 16:32:47 +08:00
|
|
|
if (priv->need_prologue) {
|
|
|
|
int type = priv->type_mapping[i];
|
|
|
|
|
|
|
|
fd = bpf_program__nth_fd(prog, type);
|
|
|
|
} else {
|
2015-11-16 20:10:13 +08:00
|
|
|
fd = bpf_program__fd(prog);
|
2015-11-17 16:32:47 +08:00
|
|
|
}
|
2015-11-16 20:10:13 +08:00
|
|
|
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
if (fd < 0) {
|
|
|
|
pr_debug("bpf: failed to get file descriptor\n");
|
|
|
|
return fd;
|
|
|
|
}
|
|
|
|
|
2016-07-13 18:44:04 +08:00
|
|
|
err = (*func)(tev->group, tev->event, fd, arg);
|
perf bpf: Collect perf_evsel in BPF object files
This patch creates a 'struct perf_evsel' for every probe in a BPF object
file(s) and fills 'struct evlist' with them. The previously introduced
dummy event is now removed. After this patch, the following command:
# perf record --event filter.o ls
Can trace on each of the probes defined in filter.o.
The core of this patch is bpf__foreach_tev(), which calls a callback
function for each 'struct probe_trace_event' event for a bpf program
with each associated file descriptors. The add_bpf_event() callback
creates evsels by calling parse_events_add_tracepoint().
Since bpf-loader.c will not be built if libbpf is turned off, an empty
bpf__foreach_tev() is defined in bpf-loader.h to avoid build errors.
Committer notes:
Before:
# /tmp/oldperf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.198 MB perf.data ]
# perf evlist
/tmp/foo.o
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
I.e. we create just the PERF_TYPE_SOFTWARE (type: 1),
PERF_COUNT_SW_DUMMY(config 0x9) event, now, with this patch:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.210 MB perf.data ]
# perf evlist -v
perf_bpf_probe:fork: type: 2, size: 112, config: 0x6bd, { sample_period,
sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1,
inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest:
1, mmap2: 1, comm_exec: 1
#
We now have a PERF_TYPE_SOFTWARE (type: 1), but the config states 0x6bd,
which is how, after setting up the event via the kprobes interface, the
'perf_bpf_probe:fork' event is accessible via the perf_event_open
syscall. This is all transient, as soon as the 'perf record' session
ends, these probes will go away.
To see how it looks like, lets try doing a neverending session, one that
expects a control+C to end:
# perf record --event /tmp/foo.o -a
So, with that in place, we can use 'perf probe' to see what is in place:
# perf probe -l
perf_bpf_probe:fork (on _do_fork@acme/git/linux/kernel/fork.c)
We also can use debugfs:
[root@felicio ~]# cat /sys/kernel/debug/tracing/kprobe_events
p:perf_bpf_probe/fork _text+638512
Ok, now lets stop and see if we got some forks:
[root@felicio linux]# perf record --event /tmp/foo.o -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.325 MB perf.data (111 samples) ]
[root@felicio linux]# perf script
sshd 1271 [003] 81797.507678: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [000] 81797.524917: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.381603: perf_bpf_probe:fork: (ffffffff8109be30)
sshd 18309 [001] 81799.408635: perf_bpf_probe:fork: (ffffffff8109be30)
<SNIP>
Sure enough, we have 111 forks :-)
Callchains seems to work as well:
# perf report --stdio --no-child
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 562 of event 'perf_bpf_probe:fork'
# Event count (approx.): 562
#
# Overhead Command Shared Object Symbol
# ........ ........ ................ ............
#
44.66% sh [kernel.vmlinux] [k] _do_fork
|
---_do_fork
entry_SYSCALL_64_fastpath
__libc_fork
make_child
26.16% make [kernel.vmlinux] [k] _do_fork
<SNIP>
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:17 +08:00
|
|
|
if (err) {
|
|
|
|
pr_debug("bpf: call back failed, stop iterate\n");
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
enum bpf_map_op_type {
|
|
|
|
BPF_MAP_OP_SET_VALUE,
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
BPF_MAP_OP_SET_EVSEL,
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
enum bpf_map_key_type {
|
|
|
|
BPF_MAP_KEY_ALL,
|
2016-02-22 17:10:34 +08:00
|
|
|
BPF_MAP_KEY_RANGES,
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
struct bpf_map_op {
|
|
|
|
struct list_head list;
|
|
|
|
enum bpf_map_op_type op_type;
|
|
|
|
enum bpf_map_key_type key_type;
|
2016-02-22 17:10:34 +08:00
|
|
|
union {
|
|
|
|
struct parse_events_array array;
|
|
|
|
} k;
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
union {
|
|
|
|
u64 value;
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
struct perf_evsel *evsel;
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
} v;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct bpf_map_priv {
|
|
|
|
struct list_head ops_list;
|
|
|
|
};
|
|
|
|
|
|
|
|
static void
|
|
|
|
bpf_map_op__delete(struct bpf_map_op *op)
|
|
|
|
{
|
|
|
|
if (!list_empty(&op->list))
|
|
|
|
list_del(&op->list);
|
2016-02-22 17:10:34 +08:00
|
|
|
if (op->key_type == BPF_MAP_KEY_RANGES)
|
|
|
|
parse_events__clear_array(&op->k.array);
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
free(op);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
bpf_map_priv__purge(struct bpf_map_priv *priv)
|
|
|
|
{
|
|
|
|
struct bpf_map_op *pos, *n;
|
|
|
|
|
|
|
|
list_for_each_entry_safe(pos, n, &priv->ops_list, list) {
|
|
|
|
list_del_init(&pos->list);
|
|
|
|
bpf_map_op__delete(pos);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
bpf_map_priv__clear(struct bpf_map *map __maybe_unused,
|
|
|
|
void *_priv)
|
|
|
|
{
|
|
|
|
struct bpf_map_priv *priv = _priv;
|
|
|
|
|
|
|
|
bpf_map_priv__purge(priv);
|
|
|
|
free(priv);
|
|
|
|
}
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
static int
|
|
|
|
bpf_map_op_setkey(struct bpf_map_op *op, struct parse_events_term *term)
|
|
|
|
{
|
|
|
|
op->key_type = BPF_MAP_KEY_ALL;
|
|
|
|
if (!term)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (term->array.nr_ranges) {
|
|
|
|
size_t memsz = term->array.nr_ranges *
|
|
|
|
sizeof(op->k.array.ranges[0]);
|
|
|
|
|
|
|
|
op->k.array.ranges = memdup(term->array.ranges, memsz);
|
|
|
|
if (!op->k.array.ranges) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory to alloc indices for map\n");
|
2016-02-22 17:10:34 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
op->key_type = BPF_MAP_KEY_RANGES;
|
|
|
|
op->k.array.nr_ranges = term->array.nr_ranges;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
static struct bpf_map_op *
|
2016-02-22 17:10:34 +08:00
|
|
|
bpf_map_op__new(struct parse_events_term *term)
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
{
|
|
|
|
struct bpf_map_op *op;
|
2016-02-22 17:10:34 +08:00
|
|
|
int err;
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
|
|
|
|
op = zalloc(sizeof(*op));
|
|
|
|
if (!op) {
|
|
|
|
pr_debug("Failed to alloc bpf_map_op\n");
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
}
|
|
|
|
INIT_LIST_HEAD(&op->list);
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
err = bpf_map_op_setkey(op, term);
|
|
|
|
if (err) {
|
|
|
|
free(op);
|
|
|
|
return ERR_PTR(err);
|
|
|
|
}
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
return op;
|
|
|
|
}
|
|
|
|
|
2016-04-08 23:07:24 +08:00
|
|
|
static struct bpf_map_op *
|
|
|
|
bpf_map_op__clone(struct bpf_map_op *op)
|
|
|
|
{
|
|
|
|
struct bpf_map_op *newop;
|
|
|
|
|
|
|
|
newop = memdup(op, sizeof(*op));
|
|
|
|
if (!newop) {
|
|
|
|
pr_debug("Failed to alloc bpf_map_op\n");
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&newop->list);
|
|
|
|
if (op->key_type == BPF_MAP_KEY_RANGES) {
|
|
|
|
size_t memsz = op->k.array.nr_ranges *
|
|
|
|
sizeof(op->k.array.ranges[0]);
|
|
|
|
|
|
|
|
newop->k.array.ranges = memdup(op->k.array.ranges, memsz);
|
|
|
|
if (!newop->k.array.ranges) {
|
|
|
|
pr_debug("Failed to alloc indices for map\n");
|
|
|
|
free(newop);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return newop;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct bpf_map_priv *
|
|
|
|
bpf_map_priv__clone(struct bpf_map_priv *priv)
|
|
|
|
{
|
|
|
|
struct bpf_map_priv *newpriv;
|
|
|
|
struct bpf_map_op *pos, *newop;
|
|
|
|
|
|
|
|
newpriv = zalloc(sizeof(*newpriv));
|
|
|
|
if (!newpriv) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory to alloc map private\n");
|
2016-04-08 23:07:24 +08:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
INIT_LIST_HEAD(&newpriv->ops_list);
|
|
|
|
|
|
|
|
list_for_each_entry(pos, &priv->ops_list, list) {
|
|
|
|
newop = bpf_map_op__clone(pos);
|
|
|
|
if (!newop) {
|
|
|
|
bpf_map_priv__purge(newpriv);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
list_add_tail(&newop->list, &newpriv->ops_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
return newpriv;
|
|
|
|
}
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
static int
|
|
|
|
bpf_map__add_op(struct bpf_map *map, struct bpf_map_op *op)
|
|
|
|
{
|
2016-06-02 22:02:05 +08:00
|
|
|
const char *map_name = bpf_map__name(map);
|
2016-06-02 21:51:59 +08:00
|
|
|
struct bpf_map_priv *priv = bpf_map__priv(map);
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
|
2016-06-02 21:51:59 +08:00
|
|
|
if (IS_ERR(priv)) {
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
pr_debug("Failed to get private from map %s\n", map_name);
|
2016-06-02 21:51:59 +08:00
|
|
|
return PTR_ERR(priv);
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!priv) {
|
|
|
|
priv = zalloc(sizeof(*priv));
|
|
|
|
if (!priv) {
|
2016-10-14 00:18:11 +08:00
|
|
|
pr_debug("Not enough memory to alloc map private\n");
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
INIT_LIST_HEAD(&priv->ops_list);
|
|
|
|
|
2016-06-03 23:38:21 +08:00
|
|
|
if (bpf_map__set_priv(map, priv, bpf_map_priv__clear)) {
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
free(priv);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
list_add_tail(&op->list, &priv->ops_list);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
static struct bpf_map_op *
|
2016-02-22 17:10:34 +08:00
|
|
|
bpf_map__add_newop(struct bpf_map *map, struct parse_events_term *term)
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
{
|
|
|
|
struct bpf_map_op *op;
|
|
|
|
int err;
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
op = bpf_map_op__new(term);
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
if (IS_ERR(op))
|
|
|
|
return op;
|
|
|
|
|
|
|
|
err = bpf_map__add_op(map, op);
|
|
|
|
if (err) {
|
|
|
|
bpf_map_op__delete(op);
|
|
|
|
return ERR_PTR(err);
|
|
|
|
}
|
|
|
|
return op;
|
|
|
|
}
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
static int
|
|
|
|
__bpf_map__config_value(struct bpf_map *map,
|
|
|
|
struct parse_events_term *term)
|
|
|
|
{
|
|
|
|
struct bpf_map_op *op;
|
2016-06-02 22:02:05 +08:00
|
|
|
const char *map_name = bpf_map__name(map);
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *def = bpf_map__def(map);
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
|
2016-06-03 01:21:06 +08:00
|
|
|
if (IS_ERR(def)) {
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
pr_debug("Unable to get map definition from '%s'\n",
|
|
|
|
map_name);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
|
2016-06-03 01:21:06 +08:00
|
|
|
if (def->type != BPF_MAP_TYPE_ARRAY) {
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
pr_debug("Map %s type is not BPF_MAP_TYPE_ARRAY\n",
|
|
|
|
map_name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE;
|
|
|
|
}
|
2016-06-03 01:21:06 +08:00
|
|
|
if (def->key_size < sizeof(unsigned int)) {
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
pr_debug("Map %s has incorrect key size\n", map_name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_KEYSIZE;
|
|
|
|
}
|
2016-06-03 01:21:06 +08:00
|
|
|
switch (def->value_size) {
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
case 1:
|
|
|
|
case 2:
|
|
|
|
case 4:
|
|
|
|
case 8:
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
pr_debug("Map %s has incorrect value size\n", map_name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_VALUESIZE;
|
|
|
|
}
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
op = bpf_map__add_newop(map, term);
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
if (IS_ERR(op))
|
|
|
|
return PTR_ERR(op);
|
|
|
|
op->op_type = BPF_MAP_OP_SET_VALUE;
|
|
|
|
op->v.value = term->val.num;
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
return 0;
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
bpf_map__config_value(struct bpf_map *map,
|
|
|
|
struct parse_events_term *term,
|
|
|
|
struct perf_evlist *evlist __maybe_unused)
|
|
|
|
{
|
|
|
|
if (!term->err_val) {
|
|
|
|
pr_debug("Config value not set\n");
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_CONF;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (term->type_val != PARSE_EVENTS__TERM_TYPE_NUM) {
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
pr_debug("ERROR: wrong value type for 'value'\n");
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_VALUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return __bpf_map__config_value(map, term);
|
|
|
|
}
|
|
|
|
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
static int
|
|
|
|
__bpf_map__config_event(struct bpf_map *map,
|
|
|
|
struct parse_events_term *term,
|
|
|
|
struct perf_evlist *evlist)
|
|
|
|
{
|
|
|
|
struct perf_evsel *evsel;
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *def;
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
struct bpf_map_op *op;
|
2016-06-02 22:02:05 +08:00
|
|
|
const char *map_name = bpf_map__name(map);
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
|
|
|
|
evsel = perf_evlist__find_evsel_by_str(evlist, term->val.str);
|
|
|
|
if (!evsel) {
|
|
|
|
pr_debug("Event (for '%s') '%s' doesn't exist\n",
|
|
|
|
map_name, term->val.str);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_NOEVT;
|
|
|
|
}
|
|
|
|
|
2016-06-03 01:21:06 +08:00
|
|
|
def = bpf_map__def(map);
|
|
|
|
if (IS_ERR(def)) {
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
pr_debug("Unable to get map definition from '%s'\n",
|
|
|
|
map_name);
|
2016-06-03 01:21:06 +08:00
|
|
|
return PTR_ERR(def);
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* No need to check key_size and value_size:
|
|
|
|
* kernel has already checked them.
|
|
|
|
*/
|
2016-06-03 01:21:06 +08:00
|
|
|
if (def->type != BPF_MAP_TYPE_PERF_EVENT_ARRAY) {
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
pr_debug("Map %s type is not BPF_MAP_TYPE_PERF_EVENT_ARRAY\n",
|
|
|
|
map_name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE;
|
|
|
|
}
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
op = bpf_map__add_newop(map, term);
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
if (IS_ERR(op))
|
|
|
|
return PTR_ERR(op);
|
|
|
|
op->op_type = BPF_MAP_OP_SET_EVSEL;
|
|
|
|
op->v.evsel = evsel;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
bpf_map__config_event(struct bpf_map *map,
|
|
|
|
struct parse_events_term *term,
|
|
|
|
struct perf_evlist *evlist)
|
|
|
|
{
|
|
|
|
if (!term->err_val) {
|
|
|
|
pr_debug("Config value not set\n");
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_CONF;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (term->type_val != PARSE_EVENTS__TERM_TYPE_STR) {
|
|
|
|
pr_debug("ERROR: wrong value type for 'event'\n");
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_VALUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return __bpf_map__config_event(map, term, evlist);
|
|
|
|
}
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
struct bpf_obj_config__map_func {
|
|
|
|
const char *config_opt;
|
|
|
|
int (*config_func)(struct bpf_map *, struct parse_events_term *,
|
|
|
|
struct perf_evlist *);
|
|
|
|
};
|
|
|
|
|
|
|
|
struct bpf_obj_config__map_func bpf_obj_config__map_funcs[] = {
|
|
|
|
{"value", bpf_map__config_value},
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
{"event", bpf_map__config_event},
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
};
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
static int
|
|
|
|
config_map_indices_range_check(struct parse_events_term *term,
|
|
|
|
struct bpf_map *map,
|
|
|
|
const char *map_name)
|
|
|
|
{
|
|
|
|
struct parse_events_array *array = &term->array;
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *def;
|
2016-02-22 17:10:34 +08:00
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
if (!array->nr_ranges)
|
|
|
|
return 0;
|
|
|
|
if (!array->ranges) {
|
|
|
|
pr_debug("ERROR: map %s: array->nr_ranges is %d but range array is NULL\n",
|
|
|
|
map_name, (int)array->nr_ranges);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
|
2016-06-03 01:21:06 +08:00
|
|
|
def = bpf_map__def(map);
|
|
|
|
if (IS_ERR(def)) {
|
2016-02-22 17:10:34 +08:00
|
|
|
pr_debug("ERROR: Unable to get map definition from '%s'\n",
|
|
|
|
map_name);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < array->nr_ranges; i++) {
|
|
|
|
unsigned int start = array->ranges[i].start;
|
|
|
|
size_t length = array->ranges[i].length;
|
|
|
|
unsigned int idx = start + length - 1;
|
|
|
|
|
2016-06-03 01:21:06 +08:00
|
|
|
if (idx >= def->max_entries) {
|
2016-02-22 17:10:34 +08:00
|
|
|
pr_debug("ERROR: index %d too large\n", idx);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_IDX2BIG;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
static int
|
|
|
|
bpf__obj_config_map(struct bpf_object *obj,
|
|
|
|
struct parse_events_term *term,
|
|
|
|
struct perf_evlist *evlist,
|
|
|
|
int *key_scan_pos)
|
|
|
|
{
|
|
|
|
/* key is "map:<mapname>.<config opt>" */
|
|
|
|
char *map_name = strdup(term->config + sizeof("map:") - 1);
|
|
|
|
struct bpf_map *map;
|
|
|
|
int err = -BPF_LOADER_ERRNO__OBJCONF_OPT;
|
|
|
|
char *map_opt;
|
|
|
|
size_t i;
|
|
|
|
|
|
|
|
if (!map_name)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
map_opt = strchr(map_name, '.');
|
|
|
|
if (!map_opt) {
|
|
|
|
pr_debug("ERROR: Invalid map config: %s\n", map_name);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
*map_opt++ = '\0';
|
|
|
|
if (*map_opt == '\0') {
|
|
|
|
pr_debug("ERROR: Invalid map option: %s\n", term->config);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2016-06-03 23:22:51 +08:00
|
|
|
map = bpf_object__find_map_by_name(obj, map_name);
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
if (!map) {
|
|
|
|
pr_debug("ERROR: Map %s doesn't exist\n", map_name);
|
|
|
|
err = -BPF_LOADER_ERRNO__OBJCONF_MAP_NOTEXIST;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
*key_scan_pos += strlen(map_opt);
|
|
|
|
err = config_map_indices_range_check(term, map, map_name);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
*key_scan_pos -= strlen(map_opt);
|
|
|
|
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
for (i = 0; i < ARRAY_SIZE(bpf_obj_config__map_funcs); i++) {
|
|
|
|
struct bpf_obj_config__map_func *func =
|
|
|
|
&bpf_obj_config__map_funcs[i];
|
|
|
|
|
|
|
|
if (strcmp(map_opt, func->config_opt) == 0) {
|
|
|
|
err = func->config_func(map, term, evlist);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
pr_debug("ERROR: Invalid map config option '%s'\n", map_opt);
|
|
|
|
err = -BPF_LOADER_ERRNO__OBJCONF_MAP_OPT;
|
|
|
|
out:
|
|
|
|
free(map_name);
|
|
|
|
if (!err)
|
|
|
|
key_scan_pos += strlen(map_opt);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
int bpf__config_obj(struct bpf_object *obj,
|
|
|
|
struct parse_events_term *term,
|
|
|
|
struct perf_evlist *evlist,
|
|
|
|
int *error_pos)
|
|
|
|
{
|
|
|
|
int key_scan_pos = 0;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (!obj || !term || !term->config)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (!prefixcmp(term->config, "map:")) {
|
|
|
|
key_scan_pos = sizeof("map:") - 1;
|
|
|
|
err = bpf__obj_config_map(obj, term, evlist, &key_scan_pos);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
err = -BPF_LOADER_ERRNO__OBJCONF_OPT;
|
|
|
|
out:
|
|
|
|
if (error_pos)
|
|
|
|
*error_pos = key_scan_pos;
|
|
|
|
return err;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
typedef int (*map_config_func_t)(const char *name, int map_fd,
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *pdef,
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
struct bpf_map_op *op,
|
|
|
|
void *pkey, void *arg);
|
|
|
|
|
|
|
|
static int
|
|
|
|
foreach_key_array_all(map_config_func_t func,
|
|
|
|
void *arg, const char *name,
|
2016-06-03 01:21:06 +08:00
|
|
|
int map_fd, const struct bpf_map_def *pdef,
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
struct bpf_map_op *op)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
for (i = 0; i < pdef->max_entries; i++) {
|
|
|
|
err = func(name, map_fd, pdef, op, &i, arg);
|
|
|
|
if (err) {
|
|
|
|
pr_debug("ERROR: failed to insert value to %s[%u]\n",
|
|
|
|
name, i);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-02-22 17:10:34 +08:00
|
|
|
static int
|
|
|
|
foreach_key_array_ranges(map_config_func_t func, void *arg,
|
|
|
|
const char *name, int map_fd,
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *pdef,
|
2016-02-22 17:10:34 +08:00
|
|
|
struct bpf_map_op *op)
|
|
|
|
{
|
|
|
|
unsigned int i, j;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
for (i = 0; i < op->k.array.nr_ranges; i++) {
|
|
|
|
unsigned int start = op->k.array.ranges[i].start;
|
|
|
|
size_t length = op->k.array.ranges[i].length;
|
|
|
|
|
|
|
|
for (j = 0; j < length; j++) {
|
|
|
|
unsigned int idx = start + j;
|
|
|
|
|
|
|
|
err = func(name, map_fd, pdef, op, &idx, arg);
|
|
|
|
if (err) {
|
|
|
|
pr_debug("ERROR: failed to insert value to %s[%u]\n",
|
|
|
|
name, idx);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
static int
|
|
|
|
bpf_map_config_foreach_key(struct bpf_map *map,
|
|
|
|
map_config_func_t func,
|
|
|
|
void *arg)
|
|
|
|
{
|
|
|
|
int err, map_fd;
|
|
|
|
struct bpf_map_op *op;
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *def;
|
2016-06-02 22:02:05 +08:00
|
|
|
const char *name = bpf_map__name(map);
|
2016-06-02 21:51:59 +08:00
|
|
|
struct bpf_map_priv *priv = bpf_map__priv(map);
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
|
2016-06-02 21:51:59 +08:00
|
|
|
if (IS_ERR(priv)) {
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
pr_debug("ERROR: failed to get private from map %s\n", name);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
if (!priv || list_empty(&priv->ops_list)) {
|
|
|
|
pr_debug("INFO: nothing to config for map %s\n", name);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-06-03 01:21:06 +08:00
|
|
|
def = bpf_map__def(map);
|
|
|
|
if (IS_ERR(def)) {
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
pr_debug("ERROR: failed to get definition from map %s\n", name);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
2016-06-03 23:15:52 +08:00
|
|
|
map_fd = bpf_map__fd(map);
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
if (map_fd < 0) {
|
|
|
|
pr_debug("ERROR: failed to get fd from map %s\n", name);
|
|
|
|
return map_fd;
|
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry(op, &priv->ops_list, list) {
|
2016-06-03 01:21:06 +08:00
|
|
|
switch (def->type) {
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
case BPF_MAP_TYPE_ARRAY:
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
case BPF_MAP_TYPE_PERF_EVENT_ARRAY:
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
switch (op->key_type) {
|
|
|
|
case BPF_MAP_KEY_ALL:
|
|
|
|
err = foreach_key_array_all(func, arg, name,
|
2016-06-03 01:21:06 +08:00
|
|
|
map_fd, def, op);
|
2016-02-22 17:10:34 +08:00
|
|
|
break;
|
|
|
|
case BPF_MAP_KEY_RANGES:
|
|
|
|
err = foreach_key_array_ranges(func, arg, name,
|
2016-06-03 01:21:06 +08:00
|
|
|
map_fd, def,
|
2016-02-22 17:10:34 +08:00
|
|
|
op);
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
pr_debug("ERROR: keytype for map '%s' invalid\n",
|
|
|
|
name);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
2016-02-22 17:10:34 +08:00
|
|
|
if (err)
|
|
|
|
return err;
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
pr_debug("ERROR: type of '%s' incorrect\n", name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
apply_config_value_for_key(int map_fd, void *pkey,
|
|
|
|
size_t val_size, u64 val)
|
|
|
|
{
|
|
|
|
int err = 0;
|
|
|
|
|
|
|
|
switch (val_size) {
|
|
|
|
case 1: {
|
|
|
|
u8 _val = (u8)(val);
|
|
|
|
err = bpf_map_update_elem(map_fd, pkey, &_val, BPF_ANY);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
case 2: {
|
|
|
|
u16 _val = (u16)(val);
|
|
|
|
err = bpf_map_update_elem(map_fd, pkey, &_val, BPF_ANY);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
case 4: {
|
|
|
|
u32 _val = (u32)(val);
|
|
|
|
err = bpf_map_update_elem(map_fd, pkey, &_val, BPF_ANY);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
case 8: {
|
|
|
|
err = bpf_map_update_elem(map_fd, pkey, &val, BPF_ANY);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
default:
|
|
|
|
pr_debug("ERROR: invalid value size\n");
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_VALUESIZE;
|
|
|
|
}
|
|
|
|
if (err && errno)
|
|
|
|
err = -errno;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
static int
|
|
|
|
apply_config_evsel_for_key(const char *name, int map_fd, void *pkey,
|
|
|
|
struct perf_evsel *evsel)
|
|
|
|
{
|
|
|
|
struct xyarray *xy = evsel->fd;
|
|
|
|
struct perf_event_attr *attr;
|
|
|
|
unsigned int key, events;
|
|
|
|
bool check_pass = false;
|
|
|
|
int *evt_fd;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (!xy) {
|
|
|
|
pr_debug("ERROR: evsel not ready for map %s\n", name);
|
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (xy->row_size / xy->entry_size != 1) {
|
|
|
|
pr_debug("ERROR: Dimension of target event is incorrect for map %s\n",
|
|
|
|
name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_EVTDIM;
|
|
|
|
}
|
|
|
|
|
|
|
|
attr = &evsel->attr;
|
|
|
|
if (attr->inherit) {
|
|
|
|
pr_debug("ERROR: Can't put inherit event into map %s\n", name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_EVTINH;
|
|
|
|
}
|
|
|
|
|
perf tools: Introduce bpf-output event
Commit a43eec304259 ("bpf: introduce bpf_perf_event_output() helper")
adds a helper to enable a BPF program to output data to a perf ring
buffer through a new type of perf event, PERF_COUNT_SW_BPF_OUTPUT. This
patch enables perf to create events of that type. Now a perf user can
use the following cmdline to receive output data from BPF programs:
# perf record -a -e bpf-output/no-inherit,name=evt/ \
-e ./test_bpf_output.c/map:channel.event=evt/ ls /
# perf script
perf 1560 [004] 347747.086295: evt: ffffffff811fd201 sys_write ...
perf 1560 [004] 347747.086300: evt: ffffffff811fd201 sys_write ...
perf 1560 [004] 347747.086315: evt: ffffffff811fd201 sys_write ...
...
Test result:
# cat test_bpf_output.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
#define SEC(NAME) __attribute__((section(NAME), used))
static u64 (*ktime_get_ns)(void) =
(void *)BPF_FUNC_ktime_get_ns;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
(void *)BPF_FUNC_perf_event_output;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(u32),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
struct {
u64 ktime;
int cpuid;
} __attribute__((packed)) output_data;
char error_data[] = "Error: failed to output: %d\n";
output_data.cpuid = get_smp_processor_id();
output_data.ktime = ktime_get_ns();
int err = perf_event_output(ctx, &channel, get_smp_processor_id(),
&output_data, sizeof(output_data));
if (err)
trace_printk(error_data, sizeof(error_data), err);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************ END ***************************/
# perf record -a -e bpf-output/no-inherit,name=evt/ \
-e ./test_bpf_output.c/map:channel.event=evt/ ls /
# perf script | grep ls
ls 2242 [003] 347851.557563: evt: ffffffff811fd201 sys_write ...
ls 2242 [003] 347851.557571: evt: ffffffff811fd201 sys_write ...
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-11-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:37 +08:00
|
|
|
if (perf_evsel__is_bpf_output(evsel))
|
|
|
|
check_pass = true;
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
if (attr->type == PERF_TYPE_RAW)
|
|
|
|
check_pass = true;
|
|
|
|
if (attr->type == PERF_TYPE_HARDWARE)
|
|
|
|
check_pass = true;
|
|
|
|
if (!check_pass) {
|
|
|
|
pr_debug("ERROR: Event type is wrong for map %s\n", name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_EVTTYPE;
|
|
|
|
}
|
|
|
|
|
|
|
|
events = xy->entries / (xy->row_size / xy->entry_size);
|
|
|
|
key = *((unsigned int *)pkey);
|
|
|
|
if (key >= events) {
|
|
|
|
pr_debug("ERROR: there is no event %d for map %s\n",
|
|
|
|
key, name);
|
|
|
|
return -BPF_LOADER_ERRNO__OBJCONF_MAP_MAPSIZE;
|
|
|
|
}
|
|
|
|
evt_fd = xyarray__entry(xy, key, 0);
|
|
|
|
err = bpf_map_update_elem(map_fd, pkey, evt_fd, BPF_ANY);
|
|
|
|
if (err && errno)
|
|
|
|
err = -errno;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
static int
|
|
|
|
apply_obj_config_map_for_key(const char *name, int map_fd,
|
2016-06-03 01:21:06 +08:00
|
|
|
const struct bpf_map_def *pdef,
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
struct bpf_map_op *op,
|
|
|
|
void *pkey, void *arg __maybe_unused)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
|
|
|
switch (op->op_type) {
|
|
|
|
case BPF_MAP_OP_SET_VALUE:
|
|
|
|
err = apply_config_value_for_key(map_fd, pkey,
|
|
|
|
pdef->value_size,
|
|
|
|
op->v.value);
|
|
|
|
break;
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
case BPF_MAP_OP_SET_EVSEL:
|
|
|
|
err = apply_config_evsel_for_key(name, map_fd, pkey,
|
|
|
|
op->v.evsel);
|
|
|
|
break;
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
default:
|
|
|
|
pr_debug("ERROR: unknown value type for '%s'\n", name);
|
|
|
|
err = -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
apply_obj_config_map(struct bpf_map *map)
|
|
|
|
{
|
|
|
|
return bpf_map_config_foreach_key(map,
|
|
|
|
apply_obj_config_map_for_key,
|
|
|
|
NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
apply_obj_config_object(struct bpf_object *obj)
|
|
|
|
{
|
|
|
|
struct bpf_map *map;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
bpf_map__for_each(map, obj) {
|
|
|
|
err = apply_obj_config_map(map);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
int bpf__apply_obj_config(void)
|
|
|
|
{
|
|
|
|
struct bpf_object *obj, *tmp;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
bpf_object__for_each_safe(obj, tmp) {
|
|
|
|
err = apply_obj_config_object(obj);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-04-08 23:07:24 +08:00
|
|
|
#define bpf__for_each_map(pos, obj, objtmp) \
|
|
|
|
bpf_object__for_each_safe(obj, objtmp) \
|
|
|
|
bpf_map__for_each(pos, obj)
|
|
|
|
|
|
|
|
#define bpf__for_each_stdout_map(pos, obj, objtmp) \
|
|
|
|
bpf__for_each_map(pos, obj, objtmp) \
|
2016-06-02 22:02:05 +08:00
|
|
|
if (bpf_map__name(pos) && \
|
2016-04-08 23:07:24 +08:00
|
|
|
(strcmp("__bpf_stdout__", \
|
2016-06-02 22:02:05 +08:00
|
|
|
bpf_map__name(pos)) == 0))
|
2016-04-08 23:07:24 +08:00
|
|
|
|
|
|
|
int bpf__setup_stdout(struct perf_evlist *evlist __maybe_unused)
|
|
|
|
{
|
|
|
|
struct bpf_map_priv *tmpl_priv = NULL;
|
|
|
|
struct bpf_object *obj, *tmp;
|
perf bpf: Automatically create bpf-output event __bpf_stdout__
This patch removes the need to set a bpf-output event in cmdline. By
referencing a map named '__bpf_stdout__', perf automatically creates an
event for it.
For example:
# perf record -e ./test_bpf_trace.c usleep 100000
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data (2 samples) ]
# perf script
usleep 4639 [000] 261895.307826: 0 __bpf_stdout__: ffffffff810eb9a1 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
usleep 4639 [000] 261895.407883: 0 __bpf_stdout__: ffffffff8105d609 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
perf record -e ./test_bpf_trace.c usleep 100000
equals to:
perf record -e bpf-output/no-inherit=1,name=__bpf_stdout__/ \
-e ./test_bpf_trace.c/map:__bpf_stdout__.event=__bpf_stdout__/ \
usleep 100000
Where test_bpf_trace.c is:
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
#define SEC(NAME) __attribute__((section(NAME), used))
static u64 (*ktime_get_ns)(void) =
(void *)BPF_FUNC_ktime_get_ns;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
(void *)BPF_FUNC_perf_event_output;
struct bpf_map_def SEC("maps") __bpf_stdout__ = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(u32),
.max_entries = __NR_CPUS__,
};
static inline int __attribute__((always_inline))
func(void *ctx, int type)
{
char output_str[] = "Raise a BPF event!";
char err_str[] = "BAD %d\n";
int err;
err = perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(),
&output_str, sizeof(output_str));
if (err)
trace_printk(err_str, sizeof(err_str), err);
return 1;
}
SEC("func_begin=sys_nanosleep")
int func_begin(void *ctx) {return func(ctx, 1);}
SEC("func_end=sys_nanosleep%return")
int func_end(void *ctx) { return func(ctx, 2);}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Committer note:
Testing with 'perf trace':
# trace -e nanosleep --ev test_bpf_stdout.c usleep 1
0.007 ( 0.007 ms): usleep/729 nanosleep(rqtp: 0x7ffc5bbc5fe0) ...
0.007 ( ): __bpf_stdout__:Raise a BPF event!..)
0.008 ( ): perf_bpf_probe:func_begin:(ffffffff81112460))
0.069 ( ): __bpf_stdout__:Raise a BPF event!..)
0.070 ( ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
0.072 ( 0.072 ms): usleep/729 ... [continued]: nanosleep()) = 0
#
Suggested-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460128045-97310-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-08 23:07:25 +08:00
|
|
|
struct perf_evsel *evsel = NULL;
|
2016-04-08 23:07:24 +08:00
|
|
|
struct bpf_map *map;
|
|
|
|
int err;
|
|
|
|
bool need_init = false;
|
|
|
|
|
|
|
|
bpf__for_each_stdout_map(map, obj, tmp) {
|
2016-06-02 21:51:59 +08:00
|
|
|
struct bpf_map_priv *priv = bpf_map__priv(map);
|
2016-04-08 23:07:24 +08:00
|
|
|
|
2016-06-02 21:51:59 +08:00
|
|
|
if (IS_ERR(priv))
|
2016-04-08 23:07:24 +08:00
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* No need to check map type: type should have been
|
|
|
|
* verified by kernel.
|
|
|
|
*/
|
|
|
|
if (!need_init && !priv)
|
|
|
|
need_init = !priv;
|
|
|
|
if (!tmpl_priv && priv)
|
|
|
|
tmpl_priv = priv;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!need_init)
|
|
|
|
return 0;
|
|
|
|
|
perf bpf: Automatically create bpf-output event __bpf_stdout__
This patch removes the need to set a bpf-output event in cmdline. By
referencing a map named '__bpf_stdout__', perf automatically creates an
event for it.
For example:
# perf record -e ./test_bpf_trace.c usleep 100000
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data (2 samples) ]
# perf script
usleep 4639 [000] 261895.307826: 0 __bpf_stdout__: ffffffff810eb9a1 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
usleep 4639 [000] 261895.407883: 0 __bpf_stdout__: ffffffff8105d609 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
perf record -e ./test_bpf_trace.c usleep 100000
equals to:
perf record -e bpf-output/no-inherit=1,name=__bpf_stdout__/ \
-e ./test_bpf_trace.c/map:__bpf_stdout__.event=__bpf_stdout__/ \
usleep 100000
Where test_bpf_trace.c is:
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
#define SEC(NAME) __attribute__((section(NAME), used))
static u64 (*ktime_get_ns)(void) =
(void *)BPF_FUNC_ktime_get_ns;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
(void *)BPF_FUNC_perf_event_output;
struct bpf_map_def SEC("maps") __bpf_stdout__ = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(u32),
.max_entries = __NR_CPUS__,
};
static inline int __attribute__((always_inline))
func(void *ctx, int type)
{
char output_str[] = "Raise a BPF event!";
char err_str[] = "BAD %d\n";
int err;
err = perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(),
&output_str, sizeof(output_str));
if (err)
trace_printk(err_str, sizeof(err_str), err);
return 1;
}
SEC("func_begin=sys_nanosleep")
int func_begin(void *ctx) {return func(ctx, 1);}
SEC("func_end=sys_nanosleep%return")
int func_end(void *ctx) { return func(ctx, 2);}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Committer note:
Testing with 'perf trace':
# trace -e nanosleep --ev test_bpf_stdout.c usleep 1
0.007 ( 0.007 ms): usleep/729 nanosleep(rqtp: 0x7ffc5bbc5fe0) ...
0.007 ( ): __bpf_stdout__:Raise a BPF event!..)
0.008 ( ): perf_bpf_probe:func_begin:(ffffffff81112460))
0.069 ( ): __bpf_stdout__:Raise a BPF event!..)
0.070 ( ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
0.072 ( 0.072 ms): usleep/729 ... [continued]: nanosleep()) = 0
#
Suggested-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460128045-97310-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-08 23:07:25 +08:00
|
|
|
if (!tmpl_priv) {
|
|
|
|
err = parse_events(evlist, "bpf-output/no-inherit=1,name=__bpf_stdout__/",
|
|
|
|
NULL);
|
|
|
|
if (err) {
|
|
|
|
pr_debug("ERROR: failed to create bpf-output event\n");
|
|
|
|
return -err;
|
|
|
|
}
|
|
|
|
|
|
|
|
evsel = perf_evlist__last(evlist);
|
|
|
|
}
|
2016-04-08 23:07:24 +08:00
|
|
|
|
|
|
|
bpf__for_each_stdout_map(map, obj, tmp) {
|
2016-06-02 21:51:59 +08:00
|
|
|
struct bpf_map_priv *priv = bpf_map__priv(map);
|
2016-04-08 23:07:24 +08:00
|
|
|
|
2016-06-02 21:51:59 +08:00
|
|
|
if (IS_ERR(priv))
|
2016-04-08 23:07:24 +08:00
|
|
|
return -BPF_LOADER_ERRNO__INTERNAL;
|
|
|
|
if (priv)
|
|
|
|
continue;
|
|
|
|
|
perf bpf: Automatically create bpf-output event __bpf_stdout__
This patch removes the need to set a bpf-output event in cmdline. By
referencing a map named '__bpf_stdout__', perf automatically creates an
event for it.
For example:
# perf record -e ./test_bpf_trace.c usleep 100000
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data (2 samples) ]
# perf script
usleep 4639 [000] 261895.307826: 0 __bpf_stdout__: ffffffff810eb9a1 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
usleep 4639 [000] 261895.407883: 0 __bpf_stdout__: ffffffff8105d609 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
perf record -e ./test_bpf_trace.c usleep 100000
equals to:
perf record -e bpf-output/no-inherit=1,name=__bpf_stdout__/ \
-e ./test_bpf_trace.c/map:__bpf_stdout__.event=__bpf_stdout__/ \
usleep 100000
Where test_bpf_trace.c is:
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
#define SEC(NAME) __attribute__((section(NAME), used))
static u64 (*ktime_get_ns)(void) =
(void *)BPF_FUNC_ktime_get_ns;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
(void *)BPF_FUNC_perf_event_output;
struct bpf_map_def SEC("maps") __bpf_stdout__ = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(u32),
.max_entries = __NR_CPUS__,
};
static inline int __attribute__((always_inline))
func(void *ctx, int type)
{
char output_str[] = "Raise a BPF event!";
char err_str[] = "BAD %d\n";
int err;
err = perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(),
&output_str, sizeof(output_str));
if (err)
trace_printk(err_str, sizeof(err_str), err);
return 1;
}
SEC("func_begin=sys_nanosleep")
int func_begin(void *ctx) {return func(ctx, 1);}
SEC("func_end=sys_nanosleep%return")
int func_end(void *ctx) { return func(ctx, 2);}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Committer note:
Testing with 'perf trace':
# trace -e nanosleep --ev test_bpf_stdout.c usleep 1
0.007 ( 0.007 ms): usleep/729 nanosleep(rqtp: 0x7ffc5bbc5fe0) ...
0.007 ( ): __bpf_stdout__:Raise a BPF event!..)
0.008 ( ): perf_bpf_probe:func_begin:(ffffffff81112460))
0.069 ( ): __bpf_stdout__:Raise a BPF event!..)
0.070 ( ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
0.072 ( 0.072 ms): usleep/729 ... [continued]: nanosleep()) = 0
#
Suggested-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460128045-97310-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-08 23:07:25 +08:00
|
|
|
if (tmpl_priv) {
|
|
|
|
priv = bpf_map_priv__clone(tmpl_priv);
|
|
|
|
if (!priv)
|
|
|
|
return -ENOMEM;
|
2016-04-08 23:07:24 +08:00
|
|
|
|
2016-06-03 23:38:21 +08:00
|
|
|
err = bpf_map__set_priv(map, priv, bpf_map_priv__clear);
|
perf bpf: Automatically create bpf-output event __bpf_stdout__
This patch removes the need to set a bpf-output event in cmdline. By
referencing a map named '__bpf_stdout__', perf automatically creates an
event for it.
For example:
# perf record -e ./test_bpf_trace.c usleep 100000
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data (2 samples) ]
# perf script
usleep 4639 [000] 261895.307826: 0 __bpf_stdout__: ffffffff810eb9a1 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
usleep 4639 [000] 261895.407883: 0 __bpf_stdout__: ffffffff8105d609 ...
BPF output: 0000: 52 61 69 73 65 20 61 20 Raise a
0008: 42 50 46 20 65 76 65 6e BPF even
0010: 74 21 00 00 t!..
BPF string: "Raise a BPF event!"
perf record -e ./test_bpf_trace.c usleep 100000
equals to:
perf record -e bpf-output/no-inherit=1,name=__bpf_stdout__/ \
-e ./test_bpf_trace.c/map:__bpf_stdout__.event=__bpf_stdout__/ \
usleep 100000
Where test_bpf_trace.c is:
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
#define SEC(NAME) __attribute__((section(NAME), used))
static u64 (*ktime_get_ns)(void) =
(void *)BPF_FUNC_ktime_get_ns;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
(void *)BPF_FUNC_perf_event_output;
struct bpf_map_def SEC("maps") __bpf_stdout__ = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(u32),
.max_entries = __NR_CPUS__,
};
static inline int __attribute__((always_inline))
func(void *ctx, int type)
{
char output_str[] = "Raise a BPF event!";
char err_str[] = "BAD %d\n";
int err;
err = perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(),
&output_str, sizeof(output_str));
if (err)
trace_printk(err_str, sizeof(err_str), err);
return 1;
}
SEC("func_begin=sys_nanosleep")
int func_begin(void *ctx) {return func(ctx, 1);}
SEC("func_end=sys_nanosleep%return")
int func_end(void *ctx) { return func(ctx, 2);}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Committer note:
Testing with 'perf trace':
# trace -e nanosleep --ev test_bpf_stdout.c usleep 1
0.007 ( 0.007 ms): usleep/729 nanosleep(rqtp: 0x7ffc5bbc5fe0) ...
0.007 ( ): __bpf_stdout__:Raise a BPF event!..)
0.008 ( ): perf_bpf_probe:func_begin:(ffffffff81112460))
0.069 ( ): __bpf_stdout__:Raise a BPF event!..)
0.070 ( ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
0.072 ( 0.072 ms): usleep/729 ... [continued]: nanosleep()) = 0
#
Suggested-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460128045-97310-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-04-08 23:07:25 +08:00
|
|
|
if (err) {
|
|
|
|
bpf_map_priv__clear(map, priv);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
} else if (evsel) {
|
|
|
|
struct bpf_map_op *op;
|
|
|
|
|
|
|
|
op = bpf_map__add_newop(map, NULL);
|
|
|
|
if (IS_ERR(op))
|
|
|
|
return PTR_ERR(op);
|
|
|
|
op->op_type = BPF_MAP_OP_SET_EVSEL;
|
|
|
|
op->v.evsel = evsel;
|
2016-04-08 23:07:24 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-11-06 21:58:09 +08:00
|
|
|
#define ERRNO_OFFSET(e) ((e) - __BPF_LOADER_ERRNO__START)
|
|
|
|
#define ERRCODE_OFFSET(c) ERRNO_OFFSET(BPF_LOADER_ERRNO__##c)
|
|
|
|
#define NR_ERRNO (__BPF_LOADER_ERRNO__END - __BPF_LOADER_ERRNO__START)
|
|
|
|
|
|
|
|
static const char *bpf_loader_strerror_table[NR_ERRNO] = {
|
|
|
|
[ERRCODE_OFFSET(CONFIG)] = "Invalid config string",
|
|
|
|
[ERRCODE_OFFSET(GROUP)] = "Invalid group name",
|
|
|
|
[ERRCODE_OFFSET(EVENTNAME)] = "No event name found in config string",
|
|
|
|
[ERRCODE_OFFSET(INTERNAL)] = "BPF loader internal error",
|
|
|
|
[ERRCODE_OFFSET(COMPILE)] = "Error when compiling BPF scriptlet",
|
2015-11-27 16:47:37 +08:00
|
|
|
[ERRCODE_OFFSET(PROGCONF_TERM)] = "Invalid program config term in config string",
|
perf bpf: Add prologue for BPF programs for fetching arguments
This patch generates a prologue for a BPF program which fetches arguments for
it. With this patch, the program can have arguments as follow:
SEC("lock_page=__lock_page page->flags")
int lock_page(struct pt_regs *ctx, int err, unsigned long flags)
{
return 1;
}
This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx
pointer. r2 is used to indicate if dereferencing was done successfully.
This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer
for result. Result of each arguments first store on stack:
low address
BPF_REG_FP - 24 ARG3
BPF_REG_FP - 16 ARG2
BPF_REG_FP - 8 ARG1
BPF_REG_FP
high address
Then loaded into r3, r4 and r5.
The output prologue for offn(...off2(off1(reg)))) should be:
r6 <- r1 // save ctx into a callee saved register
r7 <- fp
r7 <- r7 - stack_offset // pointer to result slot
/* load r3 with the offset in pt_regs of 'reg' */
(r7) <- r3 // make slot valid
r3 <- r3 + off1 // prepare to read unsafe pointer
r2 <- 8
r1 <- r7 // result put onto stack
call probe_read // read unsafe pointer
jnei r0, 0, err // error checking
r3 <- (r7) // read result
r3 <- r3 + off2 // prepare to read unsafe pointer
r2 <- 8
r1 <- r7
call probe_read
jnei r0, 0, err
...
/* load r2, r3, r4 from stack */
goto success
err:
r2 <- 1
/* load r3, r4, r5 with 0 */
goto usercode
success:
r2 <- 0
usercode:
r1 <- r6 // restore ctx
// original user code
If all of arguments reside in register (dereferencing is not
required), gen_prologue_fastpath() will be used to create
fast prologue:
r3 <- (r1 + offset of reg1)
r4 <- (r1 + offset of reg2)
r5 <- (r1 + offset of reg3)
r2 <- 0
P.S.
eBPF calling convention is defined as:
* r0 - return value from in-kernel function, and exit value
for eBPF program
* r1 - r5 - arguments from eBPF program to in-kernel function
* r6 - r9 - callee saved registers that in-kernel function will
preserve
* r10 - read-only frame pointer to access stack
Committer note:
At least testing if it builds and loads:
# cat test_probe_arg.c
struct pt_regs;
__attribute__((section("lock_page=__lock_page page->flags"), used))
int func(struct pt_regs *ctx, int err, unsigned long flags)
{
return 1;
}
char _license[] __attribute__((section("license"), used)) = "GPL";
int _version __attribute__((section("version"), used)) = 0x40300;
# perf record -e ./test_probe_arg.c usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.016 MB perf.data ]
# perf evlist
perf_bpf_probe:lock_page
#
Signed-off-by: He Kuang <hekuang@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: Wang Nan <wangnan0@huawei.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
|
|
|
[ERRCODE_OFFSET(PROLOGUE)] = "Failed to generate prologue",
|
|
|
|
[ERRCODE_OFFSET(PROLOGUE2BIG)] = "Prologue too big for program",
|
|
|
|
[ERRCODE_OFFSET(PROLOGUEOOB)] = "Offset out of bound for prologue",
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
[ERRCODE_OFFSET(OBJCONF_OPT)] = "Invalid object config option",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_CONF)] = "Config value not set (missing '=')",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_OPT)] = "Invalid object map config option",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_NOTEXIST)] = "Target map doesn't exist",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_VALUE)] = "Incorrect value type for map",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_TYPE)] = "Incorrect map type",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_KEYSIZE)] = "Incorrect map key size",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_VALUESIZE)] = "Incorrect map value size",
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_NOEVT)] = "Event not found for map setting",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_MAPSIZE)] = "Invalid map size for event setting",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_EVTDIM)] = "Event dimension too large",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_EVTINH)] = "Doesn't support inherit event",
|
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_EVTTYPE)] = "Wrong event type for map",
|
2016-02-22 17:10:34 +08:00
|
|
|
[ERRCODE_OFFSET(OBJCONF_MAP_IDX2BIG)] = "Index too large",
|
2015-11-06 21:58:09 +08:00
|
|
|
};
|
|
|
|
|
2015-11-06 21:49:37 +08:00
|
|
|
static int
|
|
|
|
bpf_loader_strerror(int err, char *buf, size_t size)
|
|
|
|
{
|
|
|
|
char sbuf[STRERR_BUFSIZE];
|
|
|
|
const char *msg;
|
|
|
|
|
|
|
|
if (!buf || !size)
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
err = err > 0 ? err : -err;
|
|
|
|
|
|
|
|
if (err >= __LIBBPF_ERRNO__START)
|
|
|
|
return libbpf_strerror(err, buf, size);
|
|
|
|
|
2015-11-06 21:58:09 +08:00
|
|
|
if (err >= __BPF_LOADER_ERRNO__START && err < __BPF_LOADER_ERRNO__END) {
|
|
|
|
msg = bpf_loader_strerror_table[ERRNO_OFFSET(err)];
|
|
|
|
snprintf(buf, size, "%s", msg);
|
|
|
|
buf[size - 1] = '\0';
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (err >= __BPF_LOADER_ERRNO__END)
|
|
|
|
snprintf(buf, size, "Unknown bpf loader error %d", err);
|
|
|
|
else
|
|
|
|
snprintf(buf, size, "%s",
|
tools: Introduce str_error_r()
The tools so far have been using the strerror_r() GNU variant, that
returns a string, be it the buffer passed or something else.
But that, besides being tricky in cases where we expect that the
function using strerror_r() returns the error formatted in a provided
buffer (we have to check if it returned something else and copy that
instead), breaks the build on systems not using glibc, like Alpine
Linux, where musl libc is used.
So, introduce yet another wrapper, str_error_r(), that has the GNU
interface, but uses the portable XSI variant of strerror_r(), so that
users rest asured that the provided buffer is used and it is what is
returned.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-d4t42fnf48ytlk8rjxs822tf@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-07-06 22:56:20 +08:00
|
|
|
str_error_r(err, sbuf, sizeof(sbuf)));
|
2015-11-06 21:58:09 +08:00
|
|
|
|
2015-11-06 21:49:37 +08:00
|
|
|
buf[size - 1] = '\0';
|
2015-11-06 21:58:09 +08:00
|
|
|
return -1;
|
2015-11-06 21:49:37 +08:00
|
|
|
}
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
#define bpf__strerror_head(err, buf, size) \
|
|
|
|
char sbuf[STRERR_BUFSIZE], *emsg;\
|
|
|
|
if (!size)\
|
|
|
|
return 0;\
|
|
|
|
if (err < 0)\
|
|
|
|
err = -err;\
|
2015-11-06 21:49:37 +08:00
|
|
|
bpf_loader_strerror(err, sbuf, sizeof(sbuf));\
|
|
|
|
emsg = sbuf;\
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
switch (err) {\
|
|
|
|
default:\
|
|
|
|
scnprintf(buf, size, "%s", emsg);\
|
|
|
|
break;
|
|
|
|
|
|
|
|
#define bpf__strerror_entry(val, fmt...)\
|
|
|
|
case val: {\
|
|
|
|
scnprintf(buf, size, fmt);\
|
|
|
|
break;\
|
|
|
|
}
|
|
|
|
|
|
|
|
#define bpf__strerror_end(buf, size)\
|
|
|
|
}\
|
|
|
|
buf[size - 1] = '\0';
|
|
|
|
|
2015-11-06 21:58:09 +08:00
|
|
|
int bpf__strerror_prepare_load(const char *filename, bool source,
|
|
|
|
int err, char *buf, size_t size)
|
|
|
|
{
|
|
|
|
size_t n;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
n = snprintf(buf, size, "Failed to load %s%s: ",
|
|
|
|
filename, source ? " from source" : "");
|
|
|
|
if (n >= size) {
|
|
|
|
buf[size - 1] = '\0';
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
buf += n;
|
|
|
|
size -= n;
|
|
|
|
|
|
|
|
ret = bpf_loader_strerror(err, buf, size);
|
|
|
|
buf[size - 1] = '\0';
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
int bpf__strerror_probe(struct bpf_object *obj __maybe_unused,
|
|
|
|
int err, char *buf, size_t size)
|
|
|
|
{
|
|
|
|
bpf__strerror_head(err, buf, size);
|
2015-11-27 16:47:37 +08:00
|
|
|
case BPF_LOADER_ERRNO__PROGCONF_TERM: {
|
2015-11-16 20:10:05 +08:00
|
|
|
scnprintf(buf, size, "%s (add -v to see detail)", emsg);
|
|
|
|
break;
|
|
|
|
}
|
2015-11-16 20:10:08 +08:00
|
|
|
bpf__strerror_entry(EEXIST, "Probe point exist. Try 'perf probe -d \"*\"' and set 'force=yes'");
|
2015-11-06 21:58:09 +08:00
|
|
|
bpf__strerror_entry(EACCES, "You need to be root");
|
|
|
|
bpf__strerror_entry(EPERM, "You need to be root, and /proc/sys/kernel/kptr_restrict should be 0");
|
|
|
|
bpf__strerror_entry(ENOENT, "You need to check probing points in BPF file");
|
perf tools: Create probe points for BPF programs
This patch introduces bpf__{un,}probe() functions to enable callers to
create kprobe points based on section names a BPF program. It parses the
section names in the program and creates corresponding 'struct
perf_probe_event' structures. The parse_perf_probe_command() function is
used to do the main parsing work. The resuling 'struct perf_probe_event'
is stored into program private data for further using.
By utilizing the new probing API, this patch creates probe points during
event parsing.
To ensure probe points be removed correctly, register an atexit hook so
even perf quit through exit() bpf__clear() is still called, so probing
points are cleared. Note that bpf_clear() should be registered before
bpf__probe() is called, so failure of bpf__probe() can still trigger
bpf__clear() to remove probe points which are already probed.
strerror style error reporting scaffold is created by this patch.
bpf__strerror_probe() is the first error reporting function in
bpf-loader.c.
Committer note:
Trying it:
To build a test eBPF object file:
I am testing using a script I built from the 'perf test -v LLVM' output:
$ cat ~/bin/hello-ebpf
export KERNEL_INC_OPTIONS="-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/4.8.3/include -I/home/acme/git/linux/arch/x86/include -Iarch/x86/include/generated/uapi -Iarch/x86/include/generated -I/home/acme/git/linux/include -Iinclude -I/home/acme/git/linux/arch/x86/include/uapi -Iarch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -Iinclude/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h"
export WORKING_DIR=/lib/modules/4.2.0/build
export CLANG_SOURCE=-
export CLANG_OPTIONS=-xc
OBJ=/tmp/foo.o
rm -f $OBJ
echo '__attribute__((section("fork=do_fork"), used)) int fork(void *ctx) {return 0;} char _license[] __attribute__((section("license"), used)) = "GPL";int _version __attribute__((section("version"), used)) = 0x40100;' | \
clang -D__KERNEL__ $CLANG_OPTIONS $KERNEL_INC_OPTIONS -Wno-unused-value -Wno-pointer-sign -working-directory $WORKING_DIR -c "$CLANG_SOURCE" -target bpf -O2 -o /tmp/foo.o && file $OBJ
---
First asking to put a probe in a function not present in the kernel
(misses the initial _):
$ perf record --event /tmp/foo.o sleep 1
Probe point 'do_fork' not found.
event syntax error: '/tmp/foo.o'
\___ You need to check probing points in BPF file
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
---
Now, with "__attribute__((section("fork=_do_fork"), used)):
$ grep _do_fork /proc/kallsyms
ffffffff81099ab0 T _do_fork
$ perf record --event /tmp/foo.o sleep 1
Failed to open kprobe_events: Permission denied
event syntax error: '/tmp/foo.o'
\___ Permission denied
---
Cool, we need to provide some better hints, "kprobe_events" is too low
level, one doesn't strictly need to know the precise details of how
these things are put in place, so something that shows the command
needed to fix the permissions would be more helpful.
Lets try as root instead:
# perf record --event /tmp/foo.o sleep 1
Lowering default frequency rate to 1000.
Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# perf evlist
/tmp/foo.o
[root@felicio ~]# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 1000, sample_type: IP|TID|TIME|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1,
sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
---
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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/1444826502-49291-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:15 +08:00
|
|
|
bpf__strerror_end(buf, size);
|
|
|
|
return 0;
|
2015-10-14 20:41:13 +08:00
|
|
|
}
|
perf tools: Load eBPF object into kernel
This patch utilizes bpf_object__load() provided by libbpf to load all
objects into kernel.
Committer notes:
Testing it:
When using an incorrect kernel version number, i.e., having this in your
eBPF proggie:
int _version __attribute__((section("version"), used)) = 0x40100;
For a 4.3.0-rc6+ kernel, say, this happens and needs checking at event
parsing time, to provide a better error report to the user:
# perf record --event /tmp/foo.o sleep 1
libbpf: load bpf program failed: Invalid argument
libbpf: -- BEGIN DUMP LOG ---
libbpf:
libbpf: -- END LOG --
libbpf: failed to load program 'fork=_do_fork'
libbpf: failed to load object '/tmp/foo.o'
event syntax error: '/tmp/foo.o'
\___ Invalid argument: Are you root and runing a CONFIG_BPF_SYSCALL kernel?
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
If we instead make it match, i.e. use 0x40300 on this v4.3.0-rc6+
kernel, the whole process goes thru:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.202 MB perf.data ]
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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: Paul Mackerras <paulus@samba.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/1444826502-49291-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:16 +08:00
|
|
|
|
2015-11-06 21:58:09 +08:00
|
|
|
int bpf__strerror_load(struct bpf_object *obj,
|
perf tools: Load eBPF object into kernel
This patch utilizes bpf_object__load() provided by libbpf to load all
objects into kernel.
Committer notes:
Testing it:
When using an incorrect kernel version number, i.e., having this in your
eBPF proggie:
int _version __attribute__((section("version"), used)) = 0x40100;
For a 4.3.0-rc6+ kernel, say, this happens and needs checking at event
parsing time, to provide a better error report to the user:
# perf record --event /tmp/foo.o sleep 1
libbpf: load bpf program failed: Invalid argument
libbpf: -- BEGIN DUMP LOG ---
libbpf:
libbpf: -- END LOG --
libbpf: failed to load program 'fork=_do_fork'
libbpf: failed to load object '/tmp/foo.o'
event syntax error: '/tmp/foo.o'
\___ Invalid argument: Are you root and runing a CONFIG_BPF_SYSCALL kernel?
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
If we instead make it match, i.e. use 0x40300 on this v4.3.0-rc6+
kernel, the whole process goes thru:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.202 MB perf.data ]
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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: Paul Mackerras <paulus@samba.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/1444826502-49291-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:16 +08:00
|
|
|
int err, char *buf, size_t size)
|
|
|
|
{
|
|
|
|
bpf__strerror_head(err, buf, size);
|
2015-11-06 21:58:09 +08:00
|
|
|
case LIBBPF_ERRNO__KVER: {
|
2016-06-03 23:22:51 +08:00
|
|
|
unsigned int obj_kver = bpf_object__kversion(obj);
|
2015-11-06 21:58:09 +08:00
|
|
|
unsigned int real_kver;
|
|
|
|
|
|
|
|
if (fetch_kernel_version(&real_kver, NULL, 0)) {
|
|
|
|
scnprintf(buf, size, "Unable to fetch kernel version");
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (obj_kver != real_kver) {
|
|
|
|
scnprintf(buf, size,
|
|
|
|
"'version' ("KVER_FMT") doesn't match running kernel ("KVER_FMT")",
|
|
|
|
KVER_PARAM(obj_kver),
|
|
|
|
KVER_PARAM(real_kver));
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
scnprintf(buf, size, "Failed to load program for unknown reason");
|
|
|
|
break;
|
|
|
|
}
|
perf tools: Load eBPF object into kernel
This patch utilizes bpf_object__load() provided by libbpf to load all
objects into kernel.
Committer notes:
Testing it:
When using an incorrect kernel version number, i.e., having this in your
eBPF proggie:
int _version __attribute__((section("version"), used)) = 0x40100;
For a 4.3.0-rc6+ kernel, say, this happens and needs checking at event
parsing time, to provide a better error report to the user:
# perf record --event /tmp/foo.o sleep 1
libbpf: load bpf program failed: Invalid argument
libbpf: -- BEGIN DUMP LOG ---
libbpf:
libbpf: -- END LOG --
libbpf: failed to load program 'fork=_do_fork'
libbpf: failed to load object '/tmp/foo.o'
event syntax error: '/tmp/foo.o'
\___ Invalid argument: Are you root and runing a CONFIG_BPF_SYSCALL kernel?
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
If we instead make it match, i.e. use 0x40300 on this v4.3.0-rc6+
kernel, the whole process goes thru:
# perf record --event /tmp/foo.o -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.202 MB perf.data ]
# perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period,
sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD, disabled: 1,
inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1,
exclude_guest: 1, mmap2: 1, comm_exec: 1
#
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.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: Paul Mackerras <paulus@samba.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/1444826502-49291-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-10-14 20:41:16 +08:00
|
|
|
bpf__strerror_end(buf, size);
|
|
|
|
return 0;
|
|
|
|
}
|
perf bpf: Add API to set values to map entries in a bpf object
bpf__config_obj() is introduced as a core API to config BPF object after
loading. One configuration option of maps is introduced. After this
patch BPF object can accept assignments like:
map:my_map.value=1234
(map.my_map.value looks pretty. However, there's a small but hard to fix
problem related to flex's greedy matching. Please see [1]. Choose ':'
to avoid it in a simpler way.)
This patch is more complex than the work it does because the
consideration of extension. In designing BPF map configuration, the
following things should be considered:
1. Array indices selection: perf should allow user setting different
value for different slots in an array, with syntax like:
map:my_map.value[0,3...6]=1234;
2. A map should be set by different config terms, each for a part
of it. For example, set each slot to the pid of a thread;
3. Type of value: integer is not the only valid value type. A perf
counter can also be put into a map after commit 35578d798400
("bpf: Implement function bpf_perf_event_read() that get the
selected hardware PMU counter")
4. For a hash table, it should be possible to use a string or other
value as a key;
5. It is possible that map configuration is unable to be setup
during parsing. A perf counter is an example.
Therefore, this patch does the following:
1. Instead of updating map element during parsing, this patch stores
map config options in 'struct bpf_map_priv'. Following patches
will apply those configs at an appropriate time;
2. Link map operations in a list so a map can have multiple config
terms attached, so different parts can be configured separately;
3. Make 'struct bpf_map_priv' extensible so that the following patches
can add new types of keys and operations;
4. Use bpf_obj_config__map_funcs array to support more map config options.
Since the patch changing the event parser to parse BPF object config is
relative large, I've put it in another commit. Code in this patch can be
tested after applying the next patch.
[1] http://lkml.kernel.org/g/564ED621.4050500@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-4-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
[ Changes "maps:my_map.value" to "map:my_map.value", improved error messages ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:30 +08:00
|
|
|
|
|
|
|
int bpf__strerror_config_obj(struct bpf_object *obj __maybe_unused,
|
|
|
|
struct parse_events_term *term __maybe_unused,
|
|
|
|
struct perf_evlist *evlist __maybe_unused,
|
|
|
|
int *error_pos __maybe_unused, int err,
|
|
|
|
char *buf, size_t size)
|
|
|
|
{
|
|
|
|
bpf__strerror_head(err, buf, size);
|
|
|
|
bpf__strerror_entry(BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE,
|
|
|
|
"Can't use this config term with this map type");
|
|
|
|
bpf__strerror_end(buf, size);
|
|
|
|
return 0;
|
|
|
|
}
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
|
|
|
|
int bpf__strerror_apply_obj_config(int err, char *buf, size_t size)
|
|
|
|
{
|
|
|
|
bpf__strerror_head(err, buf, size);
|
perf tools: Enable passing event to BPF object
A new syntax is added to the parser so that the user can access
predefined perf events in BPF objects.
After this patch, BPF programs for perf are finally able to utilize
bpf_perf_event_read() introduced in commit 35578d798400 ("bpf: Implement
function bpf_perf_event_read() that get the selected hardware PMU
counter").
Test result:
# cat test_bpf_map_2.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
static int (*get_smp_processor_id)(void) =
(void *)BPF_FUNC_get_smp_processor_id;
static int (*perf_event_read)(struct bpf_map_def *, int) =
(void *)BPF_FUNC_perf_event_read;
struct bpf_map_def SEC("maps") pmu_map = {
.type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = __NR_CPUS__,
};
SEC("func_write=sys_write")
int func_write(void *ctx)
{
unsigned long long val;
char fmt[] = "sys_write: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
SEC("func_write_return=sys_write%return")
int func_write_return(void *ctx)
{
unsigned long long val = 0;
char fmt[] = "sys_write_return: pmu=%llu\n";
val = perf_event_read(&pmu_map, get_smp_processor_id());
trace_printk(fmt, sizeof(fmt), val);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (7 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17066 [000] d... 938449.863301: : sys_write: pmu=1157327
ls-17066 [000] dN.. 938449.863342: : sys_write_return: pmu=1225218
ls-17066 [000] d... 938449.863349: : sys_write: pmu=1241922
ls-17066 [000] dN.. 938449.863369: : sys_write_return: pmu=1267445
Normal case (system wide):
# echo "" > /sys/kernel/debug/tracing/trace
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.811 MB perf.data (120 samples) ]
# cat /sys/kernel/debug/tracing/trace | grep -v '18446744073709551594' | grep -v perf | head -n 20
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
gmain-30828 [002] d... 2740551.068992: : sys_write: pmu=84373
gmain-30828 [002] d... 2740551.068992: : sys_write_return: pmu=87696
gmain-30828 [002] d... 2740551.068996: : sys_write: pmu=100658
gmain-30828 [002] d... 2740551.068997: : sys_write_return: pmu=102572
Error case 1:
# perf record -e './test_bpf_map_2.c' ls /
[SNIP]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.014 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep ls
ls-17115 [007] d... 2724279.665625: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665651: : sys_write_return: pmu=18446744073709551614
ls-17115 [007] d... 2724279.665658: : sys_write: pmu=18446744073709551614
ls-17115 [007] dN.. 2724279.665677: : sys_write_return: pmu=18446744073709551614
(18446744073709551614 is 0xfffffffffffffffe (-2))
Error case 2:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=evt/' -a
event syntax error: '..ps:pmu_map.event=evt/'
\___ Event not found for map setting
Hint: Valid config terms:
map:[<arraymap>].value=[value]
map:[<eventmap>].event=[event]
[SNIP]
Error case 3:
# ls /proc/2348/task/
2348 2505 2506 2507 2508
# perf record -i -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' -p 2348
ERROR: Apply config to BPF failed: Cannot set event to BPF map in multi-thread tracing
Error case 4:
# perf record -e cycles -e './test_bpf_map_2.c/map:pmu_map.event=cycles/' ls /
ERROR: Apply config to BPF failed: Doesn't support inherit event (Hint: use -i to turn off inherit)
Error case 5:
# perf record -i -e raw_syscalls:sys_enter -e './test_bpf_map_2.c/map:pmu_map.event=raw_syscalls:sys_enter/' ls
ERROR: Apply config to BPF failed: Can only put raw, hardware and BPF output event into a BPF map
Error case 6:
# perf record -i -e './test_bpf_map_2.c/map:pmu_map.event=123/' ls /
event syntax error: '.._map.event=123/'
\___ Incorrect value type for map
[SNIP]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-7-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:33 +08:00
|
|
|
bpf__strerror_entry(BPF_LOADER_ERRNO__OBJCONF_MAP_EVTDIM,
|
|
|
|
"Cannot set event to BPF map in multi-thread tracing");
|
|
|
|
bpf__strerror_entry(BPF_LOADER_ERRNO__OBJCONF_MAP_EVTINH,
|
|
|
|
"%s (Hint: use -i to turn off inherit)", emsg);
|
|
|
|
bpf__strerror_entry(BPF_LOADER_ERRNO__OBJCONF_MAP_EVTTYPE,
|
|
|
|
"Can only put raw, hardware and BPF output event into a BPF map");
|
perf record: Apply config to BPF objects before recording
bpf__apply_obj_config() is introduced as the core API to apply object
config options to all BPF objects. This patch also does the real work
for setting values for BPF_MAP_TYPE_PERF_ARRAY maps by inserting value
stored in map's private field into the BPF map.
This patch is required because we are not always able to set all BPF
config during parsing. Further patch will set events created by perf to
BPF_MAP_TYPE_PERF_EVENT_ARRAY maps, which is not exist until
perf_evsel__open().
bpf_map_foreach_key() is introduced to iterate over each key needs to be
configured. This function would be extended to support more map types
and different key settings.
In perf record, before start recording, call bpf__apply_config() to turn
on all BPF config options.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_1.c/map:channel.value=11/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
# ./perf record -e './test_bpf_map_1.c/map:channel.value=101/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
# entries-in-buffer/entries-written: 1/1 #P:8
[SNIP]
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
usleep-18593 [007] d... 2394714.395539: : 11
usleep-19000 [006] d... 2394831.057840: : 101
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-6-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-02-22 17:10:32 +08:00
|
|
|
bpf__strerror_end(buf, size);
|
|
|
|
return 0;
|
|
|
|
}
|
2016-04-08 23:07:24 +08:00
|
|
|
|
|
|
|
int bpf__strerror_setup_stdout(struct perf_evlist *evlist __maybe_unused,
|
|
|
|
int err, char *buf, size_t size)
|
|
|
|
{
|
|
|
|
bpf__strerror_head(err, buf, size);
|
|
|
|
bpf__strerror_end(buf, size);
|
|
|
|
return 0;
|
|
|
|
}
|