On 16/04/2025 14:41, Steven Price wrote:
When loading a realm VCPU much of the work is handled by the RMM so only
some of the actions are required. Rearrange kvm_arch_vcpu_load()
slightly so we can bail out early for a realm guest.
Signed-off-by: Steven Price <steven.price@xxxxxxx>
Reviewed-by: Gavin Shan <gshan@xxxxxxxxxx>
---
arch/arm64/kvm/arm.c | 13 +++++++++----
1 file changed, 9 insertions(+), 4 deletions(-)
diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index cf707130ef66..08d5e0d76749 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -644,10 +644,6 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
I think we use the pkvm hook to skip to the nommu goto, to avoid
the VMID allocation and context flush.
kvm_timer_vcpu_load(vcpu);
kvm_vgic_load(vcpu);
kvm_vcpu_load_debug(vcpu);
- if (has_vhe())
- kvm_vcpu_load_vhe(vcpu);
- kvm_arch_vcpu_load_fp(vcpu);
- kvm_vcpu_pmu_restore_guest(vcpu);
if (kvm_arm_is_pvtime_enabled(&vcpu->arch))
kvm_make_request(KVM_REQ_RECORD_STEAL, vcpu);
We could also move thise pvtime to the bottom too ?
@@ -671,6 +667,15 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
&vcpu->arch.vgic_cpu.vgic_v3);
}
+ /* No additional state needs to be loaded on Realmed VMs */
+ if (vcpu_is_rec(vcpu))
+ return;
+
+ if (has_vhe())
+ kvm_vcpu_load_vhe(vcpu);
+ kvm_arch_vcpu_load_fp(vcpu);
+ kvm_vcpu_pmu_restore_guest(vcpu);
+
With the above addressed:
Reviewed-by: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
if (!cpumask_test_cpu(cpu, vcpu->kvm->arch.supported_cpus))
vcpu_set_on_unsupported_cpu(vcpu);
}
