[PATCH v16 1/4] lib: add generic support for reading PMU events

[Stephen Hemminger]

On Mon, 18 Nov 2024 08:37:03 +0100
Tomasz Duszynski <tduszynski at marvell.com> wrote:

> +Performance counter based profiling
> +-----------------------------------
> +
> +Majority of architectures support some performance monitoring unit (PMU).
> +Such unit provides programmable counters that monitor specific events.

Sentence wording is awkward, maybe combine the two sentences and/or look
to wikipedia or vendor for better definition.

Also add PMU to glossary in programmer's guide

> +
> +Different tools gather that information, like for example perf.
> +However, in some scenarios when CPU cores are isolated and run
> +dedicated tasks interrupting those tasks with perf may be undesirable.
> +
> +In such cases, an application can use the PMU library to read such events via ``rte_pmu_read()``.
> +
> +By default, userspace applications are not allowed to access PMU internals. That can be changed
> +by setting ``/sys/kernel/perf_event_paranoid`` to 2 (that should be a default value anyway) and

Wrong path, should be /proc/sys/kernel/perf_event_paranoid
The default in current kernels is 2.

> +adding ``CAP_PERFMON`` capability to DPDK application. Please refer to
> +``Documentation/admin-guide/perf-security.rst`` under Linux sources for more information. Fairly
> +recent kernel, i.e >= 5.9, is advised too.

Provide link to kernel.org page for that doc please.
https://www.kernel.org/doc/html/latest/admin-guide/perf-security.html


Since multiple things (even selinux) can restrict this. Maybe best to just add a generic
paragraph about what is needed, and leave the security description up to the kernel and
distro to handle.

> +
> +As of now implementation imposes certain limitations:
> +
> +* Only EAL lcores are supported
> +
> +* EAL lcores must not share a cpu
> +
> +* Each EAL lcore measures same group of events
> +