| Message ID | 20211117153842.302159-20-alexandru.elisei@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | KVM: arm64: Add Statistical Profiling Extension (SPE) support | expand |
Hello, This patch will be dropped in the next iteration, and instead I'll implement the same approach that PMU emulation emulation uses, which is currently being worked on [1]. Prospective reviewers can safely ignore this patch. [1] https://lore.kernel.org/linux-arm-kernel/20211213152309.158462-1-alexandru.elisei@arm.com Thanks, Alex On Wed, Nov 17, 2021 at 03:38:23PM +0000, Alexandru Elisei wrote: > The kernel allows heterogeneous systems where FEAT_SPE is not present on > all CPUs. This presents a challenge for KVM, as it will have to touch the > SPE registers when emulating SPE for a guest, and those accesses will cause > an undefined exception if SPE is not present on the CPU. > > Avoid this situation by keeping a cpumask of CPUs that the VCPU is > allowed run on, which for SPE is the reunion of all CPUs that support > SPE, and refuse to run the VCPU on a CPU which is not part of the > cpumask. > > Signed-off-by: Alexandru Elisei <alexandru.elisei@arm.com> > --- > arch/arm64/include/asm/kvm_host.h | 3 +++ > arch/arm64/kvm/arm.c | 15 +++++++++++++++ > arch/arm64/kvm/spe.c | 2 ++ > 3 files changed, 20 insertions(+) > > diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h > index 8b3faed48914..96ce98f6135d 100644 > --- a/arch/arm64/include/asm/kvm_host.h > +++ b/arch/arm64/include/asm/kvm_host.h > @@ -405,6 +405,9 @@ struct kvm_vcpu_arch { > u64 last_steal; > gpa_t base; > } steal; > + > + cpumask_var_t supported_cpus; > + bool cpu_not_supported; > }; > > /* Pointer to the vcpu's SVE FFR for sve_{save,load}_state() */ > diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c > index b2997b919be2..8a7c01d1df58 100644 > --- a/arch/arm64/kvm/arm.c > +++ b/arch/arm64/kvm/arm.c > @@ -351,6 +351,9 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu) > > vcpu->arch.mmu_page_cache.gfp_zero = __GFP_ZERO; > > + if (!zalloc_cpumask_var(&vcpu->arch.supported_cpus, GFP_KERNEL)) > + return -ENOMEM; > + > /* Set up the timer */ > kvm_timer_vcpu_init(vcpu); > > @@ -378,6 +381,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) > if (vcpu->arch.has_run_once && unlikely(!irqchip_in_kernel(vcpu->kvm))) > static_branch_dec(&userspace_irqchip_in_use); > > + free_cpumask_var(vcpu->arch.supported_cpus); > kvm_mmu_free_memory_cache(&vcpu->arch.mmu_page_cache); > kvm_timer_vcpu_terminate(vcpu); > kvm_pmu_vcpu_destroy(vcpu); > @@ -456,6 +460,10 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) > if (vcpu_has_ptrauth(vcpu)) > vcpu_ptrauth_disable(vcpu); > kvm_arch_vcpu_load_debug_state_flags(vcpu); > + > + if (!cpumask_empty(vcpu->arch.supported_cpus) && > + !cpumask_test_cpu(smp_processor_id(), vcpu->arch.supported_cpus)) > + vcpu->arch.cpu_not_supported = true; > } > > void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) > @@ -893,6 +901,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu) > */ > preempt_disable(); > > + if (unlikely(vcpu->arch.cpu_not_supported)) { > + vcpu->arch.cpu_not_supported = false; > + ret = -ENOEXEC; > + preempt_enable(); > + continue; > + } > + > kvm_pmu_flush_hwstate(vcpu); > > local_irq_disable(); > diff --git a/arch/arm64/kvm/spe.c b/arch/arm64/kvm/spe.c > index 7c6f94358cc1..f3863728bab6 100644 > --- a/arch/arm64/kvm/spe.c > +++ b/arch/arm64/kvm/spe.c > @@ -40,5 +40,7 @@ int kvm_spe_vcpu_enable_spe(struct kvm_vcpu *vcpu) > if (vcpu_has_feature(vcpu, KVM_ARM_VCPU_EL1_32BIT)) > return -EINVAL; > > + cpumask_copy(vcpu->arch.supported_cpus, &supported_cpus); > + > return 0; > } > -- > 2.33.1 > > _______________________________________________ > kvmarm mailing list > kvmarm@lists.cs.columbia.edu > https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h index 8b3faed48914..96ce98f6135d 100644 --- a/arch/arm64/include/asm/kvm_host.h +++ b/arch/arm64/include/asm/kvm_host.h @@ -405,6 +405,9 @@ struct kvm_vcpu_arch { u64 last_steal; gpa_t base; } steal; + + cpumask_var_t supported_cpus; + bool cpu_not_supported; }; /* Pointer to the vcpu's SVE FFR for sve_{save,load}_state() */ diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c index b2997b919be2..8a7c01d1df58 100644 --- a/arch/arm64/kvm/arm.c +++ b/arch/arm64/kvm/arm.c @@ -351,6 +351,9 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu) vcpu->arch.mmu_page_cache.gfp_zero = __GFP_ZERO; + if (!zalloc_cpumask_var(&vcpu->arch.supported_cpus, GFP_KERNEL)) + return -ENOMEM; + /* Set up the timer */ kvm_timer_vcpu_init(vcpu); @@ -378,6 +381,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) if (vcpu->arch.has_run_once && unlikely(!irqchip_in_kernel(vcpu->kvm))) static_branch_dec(&userspace_irqchip_in_use); + free_cpumask_var(vcpu->arch.supported_cpus); kvm_mmu_free_memory_cache(&vcpu->arch.mmu_page_cache); kvm_timer_vcpu_terminate(vcpu); kvm_pmu_vcpu_destroy(vcpu); @@ -456,6 +460,10 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) if (vcpu_has_ptrauth(vcpu)) vcpu_ptrauth_disable(vcpu); kvm_arch_vcpu_load_debug_state_flags(vcpu); + + if (!cpumask_empty(vcpu->arch.supported_cpus) && + !cpumask_test_cpu(smp_processor_id(), vcpu->arch.supported_cpus)) + vcpu->arch.cpu_not_supported = true; } void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) @@ -893,6 +901,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu) */ preempt_disable(); + if (unlikely(vcpu->arch.cpu_not_supported)) { + vcpu->arch.cpu_not_supported = false; + ret = -ENOEXEC; + preempt_enable(); + continue; + } + kvm_pmu_flush_hwstate(vcpu); local_irq_disable(); diff --git a/arch/arm64/kvm/spe.c b/arch/arm64/kvm/spe.c index 7c6f94358cc1..f3863728bab6 100644 --- a/arch/arm64/kvm/spe.c +++ b/arch/arm64/kvm/spe.c @@ -40,5 +40,7 @@ int kvm_spe_vcpu_enable_spe(struct kvm_vcpu *vcpu) if (vcpu_has_feature(vcpu, KVM_ARM_VCPU_EL1_32BIT)) return -EINVAL; + cpumask_copy(vcpu->arch.supported_cpus, &supported_cpus); + return 0; }
The kernel allows heterogeneous systems where FEAT_SPE is not present on all CPUs. This presents a challenge for KVM, as it will have to touch the SPE registers when emulating SPE for a guest, and those accesses will cause an undefined exception if SPE is not present on the CPU. Avoid this situation by keeping a cpumask of CPUs that the VCPU is allowed run on, which for SPE is the reunion of all CPUs that support SPE, and refuse to run the VCPU on a CPU which is not part of the cpumask. Signed-off-by: Alexandru Elisei <alexandru.elisei@arm.com> --- arch/arm64/include/asm/kvm_host.h | 3 +++ arch/arm64/kvm/arm.c | 15 +++++++++++++++ arch/arm64/kvm/spe.c | 2 ++ 3 files changed, 20 insertions(+)