Re: [PATCH 05/12] perf tools: Allow BPF program config probing options

From: Wangnan (F)
Date: Mon Nov 16 2015 - 04:13:16 EST




On 2015/11/13 23:46, Arnaldo Carvalho de Melo wrote:
Em Fri, Nov 13, 2015 at 12:29:14PM +0000, Wang Nan escreveu:
By extending the syntax of BPF object section names, this patch allows
user to config probing options like what they can do in 'perf probe'.

Test result:

For following BPF file bpf.c:

SEC("inlines=no\n"
"func=SyS_dup?")
int func(void *ctx)
{
return 1;
}

Cmdline:

# ./perf record -e ./test_probe_glob.c ls /
...
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# ./perf evlist
perf_bpf_probe:func_1
perf_bpf_probe:func

Change "inlines=no" to "inlines=yes":

Cmdline:

# ./perf record -e ./test_probe_glob.c ls /
...
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
# ./perf evlist
perf_bpf_probe:func_3
perf_bpf_probe:func_2
perf_bpf_probe:func_1
perf_bpf_probe:func

Signed-off-by: Wang Nan <wangnan0@xxxxxxxxxx>
Cc: Alexei Starovoitov <ast@xxxxxxxxxx>
Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@xxxxxxxxxxx>
Cc: Zefan Li <lizefan@xxxxxxxxxx>
Cc: pi3orama@xxxxxxx
---

[SNIP]

diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c
index 47b1e36..55785d5 100644
--- a/tools/perf/util/util.c
+++ b/tools/perf/util/util.c
@@ -695,3 +695,21 @@ fetch_kernel_version(unsigned int *puint, char *str,
*puint = (version << 16) + (patchlevel << 8) + sublevel;
return 0;
}
+
+int convert_str_to_bool(const char *str, bool *result)
strtobool() should be more compact and convey the same idea....

Hey, I googled for that name and guess what, the kernel has exactly this
function:

lib/string.c

/**
* strtobool - convert common user inputs into boolean values
* @s: input string
* @res: result
*
* This routine returns 0 iff the first character is one of 'Yy1Nn0'.
* Otherwise it will return -EINVAL. Value pointed to by res is
* updated upon finding a match.
*/

include/linux/string.h

So, please add it to tools/include/linux/string.h and
tools/lib/util/string.c, this way we use the same code as the kernel,
with the same function signature, etc.

Good suggestion. One small inconvenience: kernel's strtobool
only support 'yes|no', it doesn't match perf's config, which
requires parsing yes/no/on/off/true/false. In addition, it
accept '0/1' but perf config not.

It is very easy for me to change strtobool() to match them.
However, do you think we can provide a function with same
name with kernel but have different behavior?

However I think we can first clone that function and use it in
BPF without touching perf config, then decide whether we need
to change strtobool() or not.

Thank you.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/