KVM: arm/arm64: Factor out VMID into struct kvm_vmid

In preparation for nested virtualization where we are going to have more
than a single VMID per VM, let's factor out the VMID data into a
separate VMID data structure and change the VMID allocator to operate on
this new structure instead of using a struct kvm.

This also means that udate_vttbr now becomes update_vmid, and that the
vttbr itself is generated on the fly based on the stage 2 page table
base address and the vmid.

We cache the physical address of the pgd when allocating the pgd to
avoid doing the calculation on every entry to the guest and to avoid
calling into potentially non-hyp-mapped code from hyp/EL2.

If we wanted to merge the VMID allocator with the arm64 ASID allocator
at some point in the future, it should actually become easier to do that
after this patch.

Note that to avoid mapping the kvm_vmid_bits variable into hyp, we
simply forego the masking of the vmid value in kvm_get_vttbr and rely on
update_vmid to always assign a valid vmid value (within the supported
range).
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
[maz: minor cleanups]
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>

arch/arm/include/asm/kvm_host.h

@@ -57,10 +57,13 @@ int __attribute_const__ kvm_target_cpu(void);
 int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
 void kvm_reset_coprocs(struct kvm_vcpu *vcpu);
 
-struct kvm_arch {
-	/* VTTBR value associated with below pgd and vmid */
-	u64    vttbr;
+struct kvm_vmid {
+	/* The VMID generation used for the virt. memory system */
+	u64    vmid_gen;
+	u32    vmid;
+};
 
+struct kvm_arch {
 	/* The last vcpu id that ran on each physical CPU */
 	int __percpu *last_vcpu_ran;
 
@@ -70,11 +73,11 @@ struct kvm_arch {
 	 */
 
 	/* The VMID generation used for the virt. memory system */
-	u64    vmid_gen;
-	u32    vmid;
+	struct kvm_vmid vmid;
 
 	/* Stage-2 page table */
 	pgd_t *pgd;
+	phys_addr_t pgd_phys;
 
 	/* Interrupt controller */
 	struct vgic_dist	vgic;

arch/arm/include/asm/kvm_mmu.h

@@ -421,9 +421,14 @@ static inline int hyp_map_aux_data(void)
 
 static inline void kvm_set_ipa_limit(void) {}
 
-static inline bool kvm_cpu_has_cnp(void)
+static __always_inline u64 kvm_get_vttbr(struct kvm *kvm)
 {
-	return false;
+	struct kvm_vmid *vmid = &kvm->arch.vmid;
+	u64 vmid_field, baddr;
+
+	baddr = kvm->arch.pgd_phys;
+	vmid_field = (u64)vmid->vmid << VTTBR_VMID_SHIFT;
+	return kvm_phys_to_vttbr(baddr) | vmid_field;
 }
 
 #endif	/* !__ASSEMBLY__ */

arch/arm/kvm/hyp/switch.c

@@ -77,7 +77,7 @@ static void __hyp_text __deactivate_traps(struct kvm_vcpu *vcpu)
 static void __hyp_text __activate_vm(struct kvm_vcpu *vcpu)
 {
 	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
-	write_sysreg(kvm->arch.vttbr, VTTBR);
+	write_sysreg(kvm_get_vttbr(kvm), VTTBR);
 	write_sysreg(vcpu->arch.midr, VPIDR);
 }
 

arch/arm/kvm/hyp/tlb.c

@@ -41,7 +41,7 @@ void __hyp_text __kvm_tlb_flush_vmid(struct kvm *kvm)
 
 	/* Switch to requested VMID */
 	kvm = kern_hyp_va(kvm);
-	write_sysreg(kvm->arch.vttbr, VTTBR);
+	write_sysreg(kvm_get_vttbr(kvm), VTTBR);
 	isb();
 
 	write_sysreg(0, TLBIALLIS);
@@ -61,7 +61,7 @@ void __hyp_text __kvm_tlb_flush_local_vmid(struct kvm_vcpu *vcpu)
 	struct kvm *kvm = kern_hyp_va(kern_hyp_va(vcpu)->kvm);
 
 	/* Switch to requested VMID */
-	write_sysreg(kvm->arch.vttbr, VTTBR);
+	write_sysreg(kvm_get_vttbr(kvm), VTTBR);
 	isb();
 
 	write_sysreg(0, TLBIALL);

arch/arm64/include/asm/kvm_host.h

@@ -57,16 +57,19 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
 int kvm_arch_vm_ioctl_check_extension(struct kvm *kvm, long ext);
 void __extended_idmap_trampoline(phys_addr_t boot_pgd, phys_addr_t idmap_start);
 
-struct kvm_arch {
+struct kvm_vmid {
 	/* The VMID generation used for the virt. memory system */
 	u64    vmid_gen;
 	u32    vmid;
+};
+
+struct kvm_arch {
+	struct kvm_vmid vmid;
 
 	/* stage2 entry level table */
 	pgd_t *pgd;
+	phys_addr_t pgd_phys;
 
-	/* VTTBR value associated with above pgd and vmid */
-	u64    vttbr;
 	/* VTCR_EL2 value for this VM */
 	u64    vtcr;
 

arch/arm64/include/asm/kvm_hyp.h

@@ -21,6 +21,7 @@
 #include <linux/compiler.h>
 #include <linux/kvm_host.h>
 #include <asm/alternative.h>
+#include <asm/kvm_mmu.h>
 #include <asm/sysreg.h>
 
 #define __hyp_text __section(.hyp.text) notrace
@@ -163,7 +164,7 @@ void __noreturn __hyp_do_panic(unsigned long, ...);
 static __always_inline void __hyp_text __load_guest_stage2(struct kvm *kvm)
 {
 	write_sysreg(kvm->arch.vtcr, vtcr_el2);
-	write_sysreg(kvm->arch.vttbr, vttbr_el2);
+	write_sysreg(kvm_get_vttbr(kvm), vttbr_el2);
 
 	/*
 	 * ARM erratum 1165522 requires the actual execution of the above

arch/arm64/include/asm/kvm_mmu.h

@@ -591,9 +591,15 @@ static inline u64 kvm_vttbr_baddr_mask(struct kvm *kvm)
 	return vttbr_baddr_mask(kvm_phys_shift(kvm), kvm_stage2_levels(kvm));
 }
 
-static inline bool kvm_cpu_has_cnp(void)
+static __always_inline u64 kvm_get_vttbr(struct kvm *kvm)
 {
-	return system_supports_cnp();
+	struct kvm_vmid *vmid = &kvm->arch.vmid;
+	u64 vmid_field, baddr;
+	u64 cnp = system_supports_cnp() ? VTTBR_CNP_BIT : 0;
+
+	baddr = kvm->arch.pgd_phys;
+	vmid_field = (u64)vmid->vmid << VTTBR_VMID_SHIFT;
+	return kvm_phys_to_vttbr(baddr) | vmid_field | cnp;
 }
 
 #endif /* __ASSEMBLY__ */

virt/kvm/arm/arm.c

@@ -65,7 +65,6 @@ static DEFINE_PER_CPU(struct kvm_vcpu *, kvm_arm_running_vcpu);
 /* The VMID used in the VTTBR */
 static atomic64_t kvm_vmid_gen = ATOMIC64_INIT(1);
 static u32 kvm_next_vmid;
-static unsigned int kvm_vmid_bits __read_mostly;
 static DEFINE_SPINLOCK(kvm_vmid_lock);
 
 static bool vgic_present;
@@ -142,7 +141,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	kvm_vgic_early_init(kvm);
 
 	/* Mark the initial VMID generation invalid */
-	kvm->arch.vmid_gen = 0;
+	kvm->arch.vmid.vmid_gen = 0;
 
 	/* The maximum number of VCPUs is limited by the host's GIC model */
 	kvm->arch.max_vcpus = vgic_present ?
@@ -472,37 +471,31 @@ void force_vm_exit(const cpumask_t *mask)
 
 /**
  * need_new_vmid_gen - check that the VMID is still valid
- * @kvm: The VM's VMID to check
+ * @vmid: The VMID to check
  *
  * return true if there is a new generation of VMIDs being used
  *
- * The hardware supports only 256 values with the value zero reserved for the
- * host, so we check if an assigned value belongs to a previous generation,
- * which which requires us to assign a new value. If we're the first to use a
- * VMID for the new generation, we must flush necessary caches and TLBs on all
- * CPUs.
+ * The hardware supports a limited set of values with the value zero reserved
+ * for the host, so we check if an assigned value belongs to a previous
+ * generation, which which requires us to assign a new value. If we're the
+ * first to use a VMID for the new generation, we must flush necessary caches
+ * and TLBs on all CPUs.
  */
-static bool need_new_vmid_gen(struct kvm *kvm)
+static bool need_new_vmid_gen(struct kvm_vmid *vmid)
 {
 	u64 current_vmid_gen = atomic64_read(&kvm_vmid_gen);
 	smp_rmb(); /* Orders read of kvm_vmid_gen and kvm->arch.vmid */
-	return unlikely(READ_ONCE(kvm->arch.vmid_gen) != current_vmid_gen);
+	return unlikely(READ_ONCE(vmid->vmid_gen) != current_vmid_gen);
 }
 
 /**
- * update_vttbr - Update the VTTBR with a valid VMID before the guest runs
- * @kvm	The guest that we are about to run
- *
- * Called from kvm_arch_vcpu_ioctl_run before entering the guest to ensure the
- * VM has a valid VMID, otherwise assigns a new one and flushes corresponding
- * caches and TLBs.
+ * update_vmid - Update the vmid with a valid VMID for the current generation
+ * @kvm: The guest that struct vmid belongs to
+ * @vmid: The stage-2 VMID information struct
  */
-static void update_vttbr(struct kvm *kvm)
+static void update_vmid(struct kvm_vmid *vmid)
 {
-	phys_addr_t pgd_phys;
-	u64 vmid, cnp = kvm_cpu_has_cnp() ? VTTBR_CNP_BIT : 0;
-
-	if (!need_new_vmid_gen(kvm))
+	if (!need_new_vmid_gen(vmid))
 		return;
 
 	spin_lock(&kvm_vmid_lock);
@@ -512,7 +505,7 @@ static void update_vttbr(struct kvm *kvm)
 	 * already allocated a valid vmid for this vm, then this vcpu should
 	 * use the same vmid.
 	 */
-	if (!need_new_vmid_gen(kvm)) {
+	if (!need_new_vmid_gen(vmid)) {
 		spin_unlock(&kvm_vmid_lock);
 		return;
 	}
@@ -536,18 +529,12 @@ static void update_vttbr(struct kvm *kvm)
 		kvm_call_hyp(__kvm_flush_vm_context);
 	}
 
-	kvm->arch.vmid = kvm_next_vmid;
+	vmid->vmid = kvm_next_vmid;
 	kvm_next_vmid++;
-	kvm_next_vmid &= (1 << kvm_vmid_bits) - 1;
-
-	/* update vttbr to be used with the new vmid */
-	pgd_phys = virt_to_phys(kvm->arch.pgd);
-	BUG_ON(pgd_phys & ~kvm_vttbr_baddr_mask(kvm));
-	vmid = ((u64)(kvm->arch.vmid) << VTTBR_VMID_SHIFT) & VTTBR_VMID_MASK(kvm_vmid_bits);
-	kvm->arch.vttbr = kvm_phys_to_vttbr(pgd_phys) | vmid | cnp;
+	kvm_next_vmid &= (1 << kvm_get_vmid_bits()) - 1;
 
 	smp_wmb();
-	WRITE_ONCE(kvm->arch.vmid_gen, atomic64_read(&kvm_vmid_gen));
+	WRITE_ONCE(vmid->vmid_gen, atomic64_read(&kvm_vmid_gen));
 
 	spin_unlock(&kvm_vmid_lock);
 }
@@ -690,7 +677,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 		 */
 		cond_resched();
 
-		update_vttbr(vcpu->kvm);
+		update_vmid(&vcpu->kvm->arch.vmid);
 
 		check_vcpu_requests(vcpu);
 
@@ -739,7 +726,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 		 */
 		smp_store_mb(vcpu->mode, IN_GUEST_MODE);
 
-		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm) ||
+		if (ret <= 0 || need_new_vmid_gen(&vcpu->kvm->arch.vmid) ||
 		    kvm_request_pending(vcpu)) {
 			vcpu->mode = OUTSIDE_GUEST_MODE;
 			isb(); /* Ensure work in x_flush_hwstate is committed */
@@ -1417,10 +1404,6 @@ static inline void hyp_cpu_pm_exit(void)
 
 static int init_common_resources(void)
 {
-	/* set size of VMID supported by CPU */
-	kvm_vmid_bits = kvm_get_vmid_bits();
-	kvm_info("%d-bit VMID\n", kvm_vmid_bits);
-
 	kvm_set_ipa_limit();
 
 	return 0;

virt/kvm/arm/mmu.c

@@ -908,6 +908,7 @@ int create_hyp_exec_mappings(phys_addr_t phys_addr, size_t size,
  */
 int kvm_alloc_stage2_pgd(struct kvm *kvm)
 {
+	phys_addr_t pgd_phys;
 	pgd_t *pgd;
 
 	if (kvm->arch.pgd != NULL) {
@@ -920,7 +921,12 @@ int kvm_alloc_stage2_pgd(struct kvm *kvm)
 	if (!pgd)
 		return -ENOMEM;
 
+	pgd_phys = virt_to_phys(pgd);
+	if (WARN_ON(pgd_phys & ~kvm_vttbr_baddr_mask(kvm)))
+		return -EINVAL;
+
 	kvm->arch.pgd = pgd;
+	kvm->arch.pgd_phys = pgd_phys;
 	return 0;
 }
 
@@ -1008,6 +1014,7 @@ void kvm_free_stage2_pgd(struct kvm *kvm)
 		unmap_stage2_range(kvm, 0, kvm_phys_size(kvm));
 		pgd = READ_ONCE(kvm->arch.pgd);
 		kvm->arch.pgd = NULL;
+		kvm->arch.pgd_phys = 0;
 	}
 	spin_unlock(&kvm->mmu_lock);