Re: [PATCH 10/48] perf tools: Introduce bpf-output event

From: Wangnan (F)
Date: Tue Feb 23 2016 - 21:01:12 EST




On 2016/2/24 1:45, Arnaldo Carvalho de Melo wrote:
Em Mon, Feb 22, 2016 at 09:10:37AM +0000, Wang Nan escreveu:
Commit a43eec304259a6c637f4014a6d4767159b6a3aa3 (bpf: introduce
bpf_perf_event_output() helper) add a helper to enable BPF program
output data to perf ring buffer through a new type of perf event
PERF_COUNT_SW_BPF_OUTPUT. This patch enable perf to create perf
event of that type. Now perf user can use 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 ...
So, there is something strange here:

if (unlikely(event->oncpu != smp_processor_id()))
return -EOPNOTSUPP;

This is where I am hitting, with:

[acme@jouet linux]$ uname -r
4.5.0-rc4

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);

And then:

[root@jouet bpf]# tail /sys/kernel/debug/tracing/trace
perf-13040 [003] d... 12062.807729: : Error: failed to output: -95
perf-13040 [003] d... 12062.807731: : Error: failed to output: -95
perf-13040 [003] d... 12062.807732: : Error: failed to output: -95
perf-13040 [003] d... 12062.807735: : Error: failed to output: -95
perf-13040 [003] d... 12062.807737: : Error: failed to output: -95
perf-13040 [003] d... 12062.807744: : Error: failed to output: -95
gnome-terminal--3091 [001] d... 12062.807773: : Error: failed to output: -95
gnome-terminal--3091 [001] d... 12062.807784: : Error: failed to output: -95
gmain-2830 [002] d... 12062.811791: : Error: failed to output: -95
gmain-2830 [002] d... 12062.811810: : Error: failed to output: -95
[root@jouet bpf]#

Ideas? AFK for a while, will continue investigating.

I also noticed this output, but didn't digg into it because all events
I concerned is okay. I'll look into this today.

This already was submitted to Ingo, BTW.

I used, as in the changeset comment tests:

perf record -a -e bpf-output/no-inherit,name=evt/ -e ./test_bpf_output.c/map:channel.event=evt/ ls /

And perf script told me:

[root@jouet bpf]# perf script | tail
perf 13040 [003] 12062.708337: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708339: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708340: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708341: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708343: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708344: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708346: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708347: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708348: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
perf 13040 [003] 12062.708350: evt: ffffffff81234eb1 sys_write (/lib/modules/4.5.0-rc4/build/vmlinux)
[root@jouet bpf]#

Wonder where that /lib/modules/4.5.0-rc4/build/vmlinux came from...

[root@jouet bpf]# perf script | cut -d'(' -f2 | sort | uniq -c
1141 /lib/modules/4.5.0-rc4/build/vmlinux)

It's a standard directory for perf searching vmlinux. Isn't it?

tools/perf/util/symbol.c:

static const char * const vmlinux_paths_upd[] = {
"/boot/vmlinux-%s",
"/usr/lib/debug/boot/vmlinux-%s",
"/lib/modules/%s/build/vmlinux",
"/usr/lib/debug/lib/modules/%s/vmlinux",
"/usr/lib/debug/boot/vmlinux-%s.debug"
};

So what's your problem?

Thank you.