Re: [PATCH -tip 4/8] perf-probe: Use the actual address instead ofthe symbol name

From: Arnaldo Carvalho de Melo
Date: Thu Jan 23 2014 - 09:52:26 EST


Em Thu, Jan 23, 2014 at 02:29:55AM +0000, Masami Hiramatsu escreveu:
> Since several local symbols can have same name (e.g. t_show),
> we need to use the actual address instead of symbol name for
> those points. Note that this works only with debuginfo.
>
> E.g. without this change;
> ----

Please use spaces after dashed lines, this is even as (or more)
important as prefixing # lines, as this makes everything after the ---
line and the patch itself to be ignored.

I'm fixing this up this time, please add the spaces next time,

- Arnaldo

> # ./perf probe -a t_show \$vars
> Added new events:
> probe:t_show (on t_show with $vars)
> probe:t_show_1 (on t_show with $vars)
> probe:t_show_2 (on t_show with $vars)
> probe:t_show_3 (on t_show with $vars)
>
> You can now use it in all perf tools, such as:
>
> perf record -e probe:t_show_3 -aR sleep 1
> ----
> OK, we have 4 different t_show()s. All functions have
> different arguments as below;
> ----
> # cat /sys/kernel/debug/tracing/kprobe_events
> p:probe/t_show t_show m=%di:u64 v=%si:u64
> p:probe/t_show_1 t_show m=%di:u64 v=%si:u64 t=%si:u64
> p:probe/t_show_2 t_show m=%di:u64 v=%si:u64 fmt=%si:u64
> p:probe/t_show_3 t_show m=%di:u64 v=%si:u64 file=%si:u64
> ----
> However, all of them have been put on the *same* address.
> ----
> # cat /sys/kernel/debug/kprobes/list
> ffffffff810d9720 k t_show+0x0 [DISABLED]
> ffffffff810d9720 k t_show+0x0 [DISABLED]
> ffffffff810d9720 k t_show+0x0 [DISABLED]
> ffffffff810d9720 k t_show+0x0 [DISABLED]
> ----
>
> With this change;
> ----
> # ./perf probe -a t_show \$vars
> Added new events:
> probe:t_show (on t_show with $vars)
> probe:t_show_1 (on t_show with $vars)
> probe:t_show_2 (on t_show with $vars)
> probe:t_show_3 (on t_show with $vars)
>
> You can now use it in all perf tools, such as:
>
> perf record -e probe:t_show_3 -aR sleep 1
>
> # cat /sys/kernel/debug/tracing/kprobe_events
> p:probe/t_show 0xffffffff810d9720 m=%di:u64 v=%si:u64
> p:probe/t_show_1 0xffffffff810e2e40 m=%di:u64 v=%si:u64 t=%si:u64
> p:probe/t_show_2 0xffffffff810ece30 m=%di:u64 v=%si:u64 fmt=%si:u64
> p:probe/t_show_3 0xffffffff810f4ad0 m=%di:u64 v=%si:u64 file=%si:u64
>
> # cat /sys/kernel/debug/kprobes/list
> ffffffff810e2e40 k t_show+0x0 [DISABLED]
> ffffffff810ece30 k t_show+0x0 [DISABLED]
> ffffffff810f4ad0 k t_show+0x0 [DISABLED]
> ffffffff810d9720 k t_show+0x0 [DISABLED]
> ----
> This time, each event is put in different address
> correctly.
>
> Note that currently this doesn't support address-based
> probe on modules (thus the probes on modules are symbol
> based), since it requires relative address probe syntax
> for kprobe-tracer, and it doesn't implemented yet.
>
> One more note, this allows us to put events on correct
> address, but --list option should be updated to show
> correct corresponding source code.
>
> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@xxxxxxxxxxx>
> ---
> tools/perf/util/probe-event.c | 23 +++++++++++++++--------
> 1 file changed, 15 insertions(+), 8 deletions(-)
>
> diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
> index 2fb4486..92ab688 100644
> --- a/tools/perf/util/probe-event.c
> +++ b/tools/perf/util/probe-event.c
> @@ -1529,20 +1529,27 @@ char *synthesize_probe_trace_command(struct probe_trace_event *tev)
> if (buf == NULL)
> return NULL;
>
> - if (tev->uprobes)
> - len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s:%s",
> - tp->retprobe ? 'r' : 'p',
> - tev->group, tev->event,
> + len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s ", tp->retprobe ? 'r' : 'p',
> + tev->group, tev->event);
> + if (len <= 0)
> + goto error;
> +
> + /* Use the real address, except for kernel modules */
> + if (tp->address && !(tp->module && !tev->uprobes))
> + ret = e_snprintf(buf + len, MAX_CMDLEN, "%s%s0x%lx",
> + tp->module ?: "", tp->module ? ":" : "",
> + tp->address);
> + else if (tev->uprobes)
> + ret = e_snprintf(buf + len, MAX_CMDLEN, "%s:%s",
> tp->module, tp->symbol);
> else
> - len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s%s%s+%lu",
> - tp->retprobe ? 'r' : 'p',
> - tev->group, tev->event,
> + ret = e_snprintf(buf + len, MAX_CMDLEN, "%s%s%s+%lu",
> tp->module ?: "", tp->module ? ":" : "",
> tp->symbol, tp->offset);
>
> - if (len <= 0)
> + if (ret <= 0)
> goto error;
> + len += ret;
>
> for (i = 0; i < tev->nargs; i++) {
> ret = synthesize_probe_trace_arg(&tev->args[i], buf + len,
>
--
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/