RE: [PATCH v3 2/3] perf: add arm64 smmuv3 pmu driver

From: Shameerali Kolothum Thodi
Date: Wed Oct 03 2018 - 07:28:32 EST




> -----Original Message-----
> From: Robin Murphy [mailto:robin.murphy@xxxxxxx]
> Sent: 03 October 2018 11:37
> To: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@xxxxxxxxxx>;
> lorenzo.pieralisi@xxxxxxx
> Cc: will.deacon@xxxxxxx; mark.rutland@xxxxxxx; Guohanjun (Hanjun Guo)
> <guohanjun@xxxxxxxxxx>; John Garry <john.garry@xxxxxxxxxx>;
> pabba@xxxxxxxxxxxxxx; vkilari@xxxxxxxxxxxxxx; rruigrok@xxxxxxxxxxxxxx;
> linux-acpi@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; linux-arm-
> kernel@xxxxxxxxxxxxxxxxxxx; Linuxarm <linuxarm@xxxxxxxxxx>;
> neil.m.leeder@xxxxxxxxx
> Subject: Re: [PATCH v3 2/3] perf: add arm64 smmuv3 pmu driver
>
> On 21/09/18 16:08, Shameer Kolothum wrote:
> [...]
> > +
> > + err = cpuhp_state_add_instance_nocalls(cpuhp_state_num,
> > + &smmu_pmu->node);
>
> In theory a hotplug event could happen as soon as the instance is
> registered...
>
> > + if (err) {
> > + dev_err(dev, "Error %d registering hotplug, PMU @%pa\n",
> > + err, &res_0->start);
> > + return err;
> > + }
> > +
> > + /* Pick one CPU to be the preferred one to use */
> > + smmu_pmu->on_cpu = get_cpu();
>
> ...so this looks too late, i.e. a race here can result in a bogus call
> to perf_pmu_migrate_context() with an uninitialised pmu.

Thanks Robin. I will reorder them.

Shameer

> Robin.
> > + WARN_ON(irq_set_affinity(smmu_pmu->irq, cpumask_of(smmu_pmu-
> >on_cpu)));
> > +
> > + err = perf_pmu_register(&smmu_pmu->pmu, name, -1);
> > + if (err) {
> > + dev_err(dev, "Error %d registering PMU @%pa\n",
> > + err, &res_0->start);
> > + goto out_unregister;
> > + }
> > +
> > + put_cpu();