Re: [RFC V2 1/3] perf, tools: Support wildcards on pmu name in dynamic pmu events

From: Jiri Olsa
Date: Sat Mar 03 2018 - 09:34:46 EST


On Fri, Mar 02, 2018 at 06:41:30PM -0500, Agustin Vega-Frias wrote:

SNIP

>
> diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l
> index 655ecff..a1a01b1 100644
> --- a/tools/perf/util/parse-events.l
> +++ b/tools/perf/util/parse-events.l
> @@ -175,7 +175,7 @@ bpf_source [^,{}]+\.c[a-zA-Z0-9._]*
> num_dec [0-9]+
> num_hex 0x[a-fA-F0-9]+
> num_raw_hex [a-fA-F0-9]+
> -name [a-zA-Z_*?][a-zA-Z0-9_*?.]*
> +name [a-zA-Z_*?\[\]][a-zA-Z0-9_*?.\[\]]*
> name_minus [a-zA-Z_*?][a-zA-Z0-9\-_*?.:]*
> drv_cfg_term [a-zA-Z0-9_\.]+(=[a-zA-Z0-9_*?\.:]+)?
> /* If you add a modifier you need to update check_modifier() */
> diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
> index e81a20e..c528469 100644
> --- a/tools/perf/util/parse-events.y
> +++ b/tools/perf/util/parse-events.y
> @@ -8,6 +8,7 @@
>
> #define YYDEBUG 1
>
> +#include <fnmatch.h>
> #include <linux/compiler.h>
> #include <linux/list.h>
> #include <linux/types.h>
> @@ -241,7 +242,7 @@ PE_NAME opt_event_config
> if (!strncmp(name, "uncore_", 7) &&
> strncmp($1, "uncore_", 7))
> name += 7;
> - if (!strncmp($1, name, strlen($1))) {
> + if (!strncmp($1, name, strlen($1)) || !fnmatch($1, name, 0)) {

could we now get rid of the strncmp in here and keep the
glob matching only? I find it confusing now that following
commands give me same results:

- [root@krava perf]# ./perf stat -e 'cbox/clockticks/' --no-merge -a sleep 1

Performance counter stats for 'system wide':

<not supported> uncore_cbox_1/clockticks/
281,474,957,674,239 uncore_cbox_0/clockticks/

1.000958335 seconds time elapsed

- [root@krava perf]# ./perf stat -e '*cbox*/clockticks/' --no-merge -a sleep 1

Performance counter stats for 'system wide':

<not supported> uncore_cbox_1/clockticks/
5,427,337 uncore_cbox_0/clockticks/

1.000962724 seconds time elapsed

- [root@krava perf]# ./perf stat -e 'cbox*/clockticks/' --no-merge -a sleep 1

Performance counter stats for 'system wide':

<not supported> uncore_cbox_1/clockticks/
281,474,969,621,374 uncore_cbox_0/clockticks/

1.001026179 seconds time elapsed

and this one fails:

- [root@krava perf]# ./perf stat -e '*cbox/clockticks/' --no-merge -a sleep 1
event syntax error: '*cbox/clockticks/'
\___ Cannot find PMU `*cbox'. Missing kernel support?
Run 'perf list' for a list of valid events

Usage: perf stat [<options>] [<command>]

-e, --event <event> event selector. use 'perf list' to list available events


despite the fact that it makes as much sense as the previous one: perf stat -e 'cbox*/clockticks/'


I'd think let's keep just the glob matching, so it's clear
you what you use wildcards for.. thoughts?

thanks,
jirka