Re: [PATCH v4 3/4] perf/core: Remove pmu linear searching code

[Oliver Upton <oliver.upton@linux.dev>]

On Thu, May 25, 2023 at 10:46:01AM +0530, Ravi Bangoria wrote:
> On 25-May-23 3:11 AM, Nathan Chancellor wrote:
> > My apologies if this has already been reported or fixed already, I did a
> > search of lore.kernel.org and did not find anything. This patch as
> > commit 9551fbb64d09 ("perf/core: Remove pmu linear searching code") in
> > -next breaks starting QEMU with KVM enabled on two of my arm64 machines:
> > 
> > $ qemu-system-aarch64 \
> >     -display none \
> >     -nodefaults \
> >     -machine virt,gic-version=max \
> >     -append 'console=ttyAMA0 earlycon' \
> >     -kernel arch/arm64/boot/Image.gz \
> >     -initrd rootfs.cpio \
> >     -cpu host \
> >     -enable-kvm \
> >     -m 512m \
> >     -smp 8 \
> >     -serial mon:stdio
> > qemu-system-aarch64: PMU: KVM_SET_DEVICE_ATTR: No such device
> > qemu-system-aarch64: failed to set irq for PMU
> > 
> > In the kernel log, I see
> > 
> > [   42.944952] kvm: pmu event creation failed -2
> > 
> > I am not sure if this issue is unexpected as a result of this change or
> > if there is something that needs to change on the arm64 KVM side (it
> > appears the kernel message comes from arch/arm64/kvm/pmu-emul.c).
> 
> Thanks for reporting it.
> 
> Based on these detail, I feel the pmu registration failed in the host,
> most probably because pmu driver did not pass pmu name while calling
> perf_pmu_register(). Consequently kvm also failed while trying to use
> it for guest. Can you please check host kernel logs.

The PMUv3 driver does pass a name, but it relies on getting back an
allocated pmu id as @type is -1 in the call to perf_pmu_register().

What actually broke is how KVM probes for a default core PMU to use for
a guest. kvm_pmu_probe_armpmu() creates a counter w/ PERF_TYPE_RAW and
reads the pmu from the returned perf_event. The linear search had the
effect of eventually stumbling on the correct core PMU and succeeding.

Perf folks: is this WAI for heterogenous systems?

Either way, the whole KVM end of this scheme is a bit clunky, and I
believe it to be unneccessary at this point as we maintain a list of
core PMU instances that KVM is able to virtualize. We can just walk
that to find a default PMU to use.

Not seeing any issues on -next with the below diff. If this works for
folks I can actually wrap it up in a patch and send it out.

diff --git a/arch/arm64/kvm/pmu-emul.c b/arch/arm64/kvm/pmu-emul.c
index 45727d50d18d..cbc0b662b7f8 100644
--- a/arch/arm64/kvm/pmu-emul.c
+++ b/arch/arm64/kvm/pmu-emul.c
@@ -694,47 +694,26 @@ void kvm_host_pmu_init(struct arm_pmu *pmu)
 
 static struct arm_pmu *kvm_pmu_probe_armpmu(void)
 {
-	struct perf_event_attr attr = { };
-	struct perf_event *event;
-	struct arm_pmu *pmu = NULL;
-
-	/*
-	 * Create a dummy event that only counts user cycles. As we'll never
-	 * leave this function with the event being live, it will never
-	 * count anything. But it allows us to probe some of the PMU
-	 * details. Yes, this is terrible.
-	 */
-	attr.type = PERF_TYPE_RAW;
-	attr.size = sizeof(attr);
-	attr.pinned = 1;
-	attr.disabled = 0;
-	attr.exclude_user = 0;
-	attr.exclude_kernel = 1;
-	attr.exclude_hv = 1;
-	attr.exclude_host = 1;
-	attr.config = ARMV8_PMUV3_PERFCTR_CPU_CYCLES;
-	attr.sample_period = GENMASK(63, 0);
+	struct arm_pmu *arm_pmu = NULL, *tmp;
+	struct arm_pmu_entry *entry;
+	int cpu;
 
-	event = perf_event_create_kernel_counter(&attr, -1, current,
-						 kvm_pmu_perf_overflow, &attr);
+	mutex_lock(&arm_pmus_lock);
+	cpu = get_cpu();
 
-	if (IS_ERR(event)) {
-		pr_err_once("kvm: pmu event creation failed %ld\n",
-			    PTR_ERR(event));
-		return NULL;
-	}
+	list_for_each_entry(entry, &arm_pmus, entry) {
+		tmp = entry->arm_pmu;
 
-	if (event->pmu) {
-		pmu = to_arm_pmu(event->pmu);
-		if (pmu->pmuver == ID_AA64DFR0_EL1_PMUVer_NI ||
-		    pmu->pmuver == ID_AA64DFR0_EL1_PMUVer_IMP_DEF)
-			pmu = NULL;
+		if (cpumask_test_cpu(cpu, &tmp->supported_cpus)) {
+			arm_pmu = tmp;
+			break;
+		}
 	}
 
-	perf_event_disable(event);
-	perf_event_release_kernel(event);
+	put_cpu();
+	mutex_unlock(&arm_pmus_lock);
 
-	return pmu;
+	return arm_pmu;
 }
 
 u64 kvm_pmu_get_pmceid(struct kvm_vcpu *vcpu, bool pmceid1)

-- 
Thanks,
Oliver