Re: [PATCH v4 3/4] perf utils: Add helper function is_pmu_core to detect PMU CORE devices

From: Jonathan Cameron
Date: Tue Aug 22 2017 - 06:29:23 EST


On Thu, 20 Jul 2017 11:26:07 +0530
Ganapatrao Kulkarni <ganapatrao.kulkarni@xxxxxxxxxx> wrote:

> On some platforms, PMU core devices sysfs name is not cpu.
> Adding function is_pmu_core to detect as core device using
> core device specific hints in sysfs.
>
> For arm64 platforms, all core devices have file "cpus" in sysfs.
>
> Signed-off-by: Ganapatrao Kulkarni <ganapatrao.kulkarni@xxxxxxxxxx>

One really trivial point inline. Otherwise looks good to me.

Jonathan

> ---
> tools/perf/util/pmu.c | 44 ++++++++++++++++++++++++++++++++++++++++----
> 1 file changed, 40 insertions(+), 4 deletions(-)
>
> diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
> index aefdbd1..0057d1c 100644
> --- a/tools/perf/util/pmu.c
> +++ b/tools/perf/util/pmu.c
> @@ -506,6 +506,39 @@ static struct cpu_map *pmu_cpumask(const char *name)
> }
>
> /*
> + * PMU CORE devices have different name other than cpu in sysfs on some
> + * platforms. looking for possible sysfs files to identify as core device.
> + */
> +static int is_pmu_core(const char *name)
> +{
> + struct stat st;
> + char path[PATH_MAX];
> + const char *sysfs = sysfs__mountpoint();
> + const char **template;
> + const char *templates[] = {
> + "%s/bus/event_source/devices/%s/cpus",
> + NULL
> + };
> +
> + if (!sysfs)
> + return 0;
> +
> + /* Look for cpu sysfs */
> + snprintf(path, PATH_MAX, "%s/bus/event_source/devices/cpu", sysfs);
> + if ((stat(path, &st) == 0) &&
> + (strncmp(name, "cpu", strlen("cpu")) == 0))
> + return 1;
> +
> + for (template = templates; *template; template++) {
> + snprintf(path, PATH_MAX, *template, sysfs, name);
> + if (stat(path, &st) == 0)
> + return 1;
> + }

Adding generic infrastructure for one entry where we already know the other
case doesn't fit the pattern, seems a little premature.

If we can't make these two fit it seems like we should just hard code
the second case as well for now.

Refactoring to this may be make sense once we have a few different cases
that do share a common form (of including the name)

You could actually put both options in templates and rely on the fact that
snprintf will ignore excess parameters, but that feels fragile.

> +
> + return 0;
> +}
> +
> +/*
> * Return the CPU id as a raw string.
> *
> * Each architecture should provide a more precise id string that
> @@ -558,15 +591,18 @@ static void pmu_add_cpu_aliases(struct list_head *head, const char *name,
> */
> i = 0;
> while (1) {
> - const char *pname;
>
> pe = &map->table[i++];
> if (!pe->name)
> break;
>
> - pname = pe->pmu ? pe->pmu : "cpu";
> - if (strncmp(pname, name, strlen(pname)))
> - continue;
> + if (!is_pmu_core(name)) {
> + /* check for uncore devices */
> + if (pe->pmu == NULL)
> + continue;
> + if (strncmp(pe->pmu, name, strlen(pe->pmu)))
> + continue;
> + }
>
> /* need type casts to override 'const' */
> __perf_pmu__new_alias(head, NULL, (char *)pe->name,