Re: [PATCH v3] perf/arm: adjust hwevents mappings on boot

[Mark Rutland]

On Tue, Aug 16, 2022 at 03:02:21PM +0200, Peter Newman wrote:
> From: Stephane Eranian <eranian@xxxxxxxxxx>
> 
> The mapping of perf_events generic hardware events to actual PMU events on
> ARM PMUv3 may not always be correct. This is in particular true for the
> PERF_COUNT_HW_BRANCH_INSTRUCTIONS event. Although the mapping points to an
> architected event, it may not always be available. This can be seen with a
> simple:
> 
> $ perf stat -e branches sleep 0
>  Performance counter stats for 'sleep 0':
> 
>    <not supported>      branches
> 
>        0.001401081 seconds time elapsed
> 
> Yet the hardware does have an event that could be used for branches.  This
> patch fixes the problem by dynamically validating the generic hardware
> events against the supported architected events. If a mapping is wrong it
> can be replaced it with another. This is done for the event above at boot
> time.
> 
> And with that:
> 
> $ perf stat -e branches sleep 0
> 
>  Performance counter stats for 'sleep 0':
> 
>            166,739      branches
> 
>        0.000832163 seconds time elapsed
> 
> Signed-off-by: Stephane Eranian <eranian@xxxxxxxxxx>
> Co-developed-by: Peter Newman <peternewman@xxxxxxxxxx>
> Signed-off-by: Peter Newman <peternewman@xxxxxxxxxx>
> ---
> 
> v2: https://lore.kernel.org/lkml/20220324181458.3216262-1-eranian@xxxxxxxxxx/
> 
> since v2, removed prints per Will's suggestion
> 
>  arch/arm64/kernel/perf_event.c | 36 +++++++++++++++++++++++++++++++++-
>  1 file changed, 35 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c
> index cb69ff1e6138..945c31e3f3e3 100644
> --- a/arch/arm64/kernel/perf_event.c
> +++ b/arch/arm64/kernel/perf_event.c
> @@ -39,7 +39,7 @@
>   * be supported on any given implementation. Unsupported events will
>   * be disabled at run-time based on the PMCEID registers.

>   */
> -static const unsigned armv8_pmuv3_perf_map[PERF_COUNT_HW_MAX] = {
> +static unsigned armv8_pmuv3_perf_map[PERF_COUNT_HW_MAX] = {
>  	PERF_MAP_ALL_UNSUPPORTED,
>  	[PERF_COUNT_HW_CPU_CYCLES]		= ARMV8_PMUV3_PERFCTR_CPU_CYCLES,
>  	[PERF_COUNT_HW_INSTRUCTIONS]		= ARMV8_PMUV3_PERFCTR_INST_RETIRED,

On big.LITTLE systems this is array is shared by multiple PMUs, so this cannot
be altered based on a single PMU.

Rather than applying a fixup, could we special-case this at mapping time?

Does the following work for you?

Thanks
Mark.

---->8----