Re: [PATCH v4 4/5] drivers/perf: Add support for ARMv8.2 Statistical Profiling Extension

[Will Deacon <will.deacon@arm.com>]

On Mon, Jun 05, 2017 at 10:55:16AM -0500, Kim Phillips wrote:
> On Mon, 5 Jun 2017 16:22:56 +0100
> Will Deacon <will.deacon@arm.com> wrote:
> 
> > +/* Perf callbacks */
> > +static int arm_spe_pmu_event_init(struct perf_event *event)
> > +{
> > +	u64 reg;
> > +	struct perf_event_attr *attr = &event->attr;
> > +	struct arm_spe_pmu *spe_pmu = to_spe_pmu(event->pmu);
> > +
> > +	/* This is, of course, deeply driver-specific */
> > +	if (attr->type != event->pmu->type)
> > +		return -ENOENT;
> > +
> > +	if (event->cpu >= 0 &&
> > +	    !cpumask_test_cpu(event->cpu, &spe_pmu->supported_cpus))
> > +		return -ENOENT;
> > +
> > +	if (arm_spe_event_to_pmsevfr(event) & PMSEVFR_EL1_RES0)
> > +		return -EOPNOTSUPP;
> > +
> > +	if (event->hw.sample_period < spe_pmu->min_period ||
> > +	    event->hw.sample_period & PMSIRR_EL1_IVAL_MASK)
> > +		return -EOPNOTSUPP;
> > +
> > +	if (attr->exclude_idle)
> > +		return -EOPNOTSUPP;
> > +
> > +	/*
> > +	 * Feedback-directed frequency throttling doesn't work when we
> > +	 * have a buffer of samples. We'd need to manually count the
> > +	 * samples in the buffer when it fills up and adjust the event
> > +	 * count to reflect that. Instead, force the user to specify a
> > +	 * sample period instead.
> > +	 */
> > +	if (attr->freq)
> > +		return -EINVAL;
> > +
> > +	reg = arm_spe_event_to_pmsfcr(event);
> > +	if ((reg & BIT(PMSFCR_EL1_FE_SHIFT)) &&
> > +	    !(spe_pmu->features & SPE_PMU_FEAT_FILT_EVT))
> > +		return -EOPNOTSUPP;
> > +
> > +	if ((reg & BIT(PMSFCR_EL1_FT_SHIFT)) &&
> > +	    !(spe_pmu->features & SPE_PMU_FEAT_FILT_TYP))
> > +		return -EOPNOTSUPP;
> > +
> > +	if ((reg & BIT(PMSFCR_EL1_FL_SHIFT)) &&
> > +	    !(spe_pmu->features & SPE_PMU_FEAT_FILT_LAT))
> > +		return -EOPNOTSUPP;
> > +
> > +	return 0;
> > +}
> 
> AFAICT, my comments from the last submission have still not been fully
> addressed:
> 
> http://lists.infradead.org/pipermail/linux-arm-kernel/2017-May/508027.html

To be frank, I really don't plan to address them and, even if I did, I would
trust Mark to NAK the change. If you're desperate for pr_debug, I'll add it
to keep you happy, but anything more than that needs to come in the form of
a separate patch submission addressing the wider problem of error reporting
from PMU drivers back to userspace. Patches welcome, but I suspect you're
still busy working on the tools code.

Do you have any constructive comments on the patch?

Will