x86:

* Fixes for the new scalable MMU
 * Fixes for migration of nested hypervisors on AMD
 * Fix for clang integrated assembler
 * Fix for left shift by 64 (UBSAN)
 * Small cleanups
 * Straggler SEV-ES patch
 
 ARM:
 * VM init cleanups
 * PSCI relay cleanups
 * Kill CONFIG_KVM_ARM_PMU
 * Fixup __init annotations
 * Fixup reg_to_encoding()
 * Fix spurious PMCR_EL0 access
 
 * selftests cleanups
 -----BEGIN PGP SIGNATURE-----
 
 iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAl/4YOMUHHBib256aW5p
 QHJlZGhhdC5jb20ACgkQv/vSX3jHroMg0Qf/eGDOoLL18KhA9MpoPXusR5kU+82G
 AcbFj9siNW46EF7mL+sw/xAx+gZaqSpIEmn/f6BzgiaUBdFTv9CKX3B54e43e59G
 HAKD0NpqwvDIi1b0T6bcgC92mY3Qx/IDCc7/9JYjBs/iORfqyWW6xVtkF/Gfymxt
 eK+MnfMqqNZODgR/cZnCH1E48fuwOvRMxLqilLi3OOMSUfs2cQOSLTNfYQYqjeaJ
 dsQ4YeyPJO5JHtfHFr6VPIo/jDhowniac9CNvOomWWVIx2zPYVSl9d8ub6ESEPNF
 GM7UeBCOMmZG/a3qFEZPAUJ7znW4yYE85Z6pjxlrGhd1I54MJi4dd+RApw==
 =5Nfj
 -----END PGP SIGNATURE-----

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull kvm fixes from Paolo Bonzini:
 "x86:
   - Fixes for the new scalable MMU
   - Fixes for migration of nested hypervisors on AMD
   - Fix for clang integrated assembler
   - Fix for left shift by 64 (UBSAN)
   - Small cleanups
   - Straggler SEV-ES patch

  ARM:
   - VM init cleanups
   - PSCI relay cleanups
   - Kill CONFIG_KVM_ARM_PMU
   - Fixup __init annotations
   - Fixup reg_to_encoding()
   - Fix spurious PMCR_EL0 access

  Misc:
   - selftests cleanups"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (38 commits)
  KVM: x86: __kvm_vcpu_halt can be static
  KVM: SVM: Add support for booting APs in an SEV-ES guest
  KVM: nSVM: cancel KVM_REQ_GET_NESTED_STATE_PAGES on nested vmexit
  KVM: nSVM: mark vmcb as dirty when forcingly leaving the guest mode
  KVM: nSVM: correctly restore nested_run_pending on migration
  KVM: x86/mmu: Clarify TDP MMU page list invariants
  KVM: x86/mmu: Ensure TDP MMU roots are freed after yield
  kvm: check tlbs_dirty directly
  KVM: x86: change in pv_eoi_get_pending() to make code more readable
  MAINTAINERS: Really update email address for Sean Christopherson
  KVM: x86: fix shift out of bounds reported by UBSAN
  KVM: selftests: Implement perf_test_util more conventionally
  KVM: selftests: Use vm_create_with_vcpus in create_vm
  KVM: selftests: Factor out guest mode code
  KVM/SVM: Remove leftover __svm_vcpu_run prototype from svm.c
  KVM: SVM: Add register operand to vmsave call in sev_es_vcpu_load
  KVM: x86/mmu: Optimize not-present/MMIO SPTE check in get_mmio_spte()
  KVM: x86/mmu: Use raw level to index into MMIO walks' sptes array
  KVM: x86/mmu: Get root level from walkers when retrieving MMIO SPTE
  KVM: x86/mmu: Use -1 to flag an undefined spte in get_mmio_spte()
  ...

Documentation/virt/kvm/api.rst

@@ -392,9 +392,14 @@ This ioctl is obsolete and has been removed.
 
 Errors:
 
-  =====      =============================
+  =======    ==============================================================
   EINTR      an unmasked signal is pending
-  =====      =============================
+  ENOEXEC    the vcpu hasn't been initialized or the guest tried to execute
+             instructions from device memory (arm64)
+  ENOSYS     data abort outside memslots with no syndrome info and
+             KVM_CAP_ARM_NISV_TO_USER not enabled (arm64)
+  EPERM      SVE feature set but not finalized (arm64)
+  =======    ==============================================================
 
 This ioctl is used to run a guest virtual cpu.  While there are no
 explicit parameters, there is an implicit parameter block that can be

MAINTAINERS

@@ -9776,7 +9776,7 @@ F:	tools/testing/selftests/kvm/s390x/
 
 KERNEL VIRTUAL MACHINE FOR X86 (KVM/x86)
 M:	Paolo Bonzini <pbonzini@redhat.com>
-R:	Sean Christopherson <sean.j.christopherson@intel.com>
+R:	Sean Christopherson <seanjc@google.com>
 R:	Vitaly Kuznetsov <vkuznets@redhat.com>
 R:	Wanpeng Li <wanpengli@tencent.com>
 R:	Jim Mattson <jmattson@google.com>

arch/arm64/include/asm/kvm_host.h

@@ -17,6 +17,7 @@
 #include <linux/jump_label.h>
 #include <linux/kvm_types.h>
 #include <linux/percpu.h>
+#include <linux/psci.h>
 #include <asm/arch_gicv3.h>
 #include <asm/barrier.h>
 #include <asm/cpufeature.h>
@@ -240,6 +241,28 @@ struct kvm_host_data {
 	struct kvm_pmu_events pmu_events;
 };
 
+struct kvm_host_psci_config {
+	/* PSCI version used by host. */
+	u32 version;
+
+	/* Function IDs used by host if version is v0.1. */
+	struct psci_0_1_function_ids function_ids_0_1;
+
+	bool psci_0_1_cpu_suspend_implemented;
+	bool psci_0_1_cpu_on_implemented;
+	bool psci_0_1_cpu_off_implemented;
+	bool psci_0_1_migrate_implemented;
+};
+
+extern struct kvm_host_psci_config kvm_nvhe_sym(kvm_host_psci_config);
+#define kvm_host_psci_config CHOOSE_NVHE_SYM(kvm_host_psci_config)
+
+extern s64 kvm_nvhe_sym(hyp_physvirt_offset);
+#define hyp_physvirt_offset CHOOSE_NVHE_SYM(hyp_physvirt_offset)
+
+extern u64 kvm_nvhe_sym(hyp_cpu_logical_map)[NR_CPUS];
+#define hyp_cpu_logical_map CHOOSE_NVHE_SYM(hyp_cpu_logical_map)
+
 struct vcpu_reset_state {
 	unsigned long	pc;
 	unsigned long	r0;

arch/arm64/kernel/cpufeature.c

@@ -2568,7 +2568,7 @@ static void verify_hyp_capabilities(void)
 	int parange, ipa_max;
 	unsigned int safe_vmid_bits, vmid_bits;
 
-	if (!IS_ENABLED(CONFIG_KVM) || !IS_ENABLED(CONFIG_KVM_ARM_HOST))
+	if (!IS_ENABLED(CONFIG_KVM))
 		return;
 
 	safe_mmfr1 = read_sanitised_ftr_reg(SYS_ID_AA64MMFR1_EL1);

arch/arm64/kernel/smp.c

@@ -434,7 +434,7 @@ static void __init hyp_mode_check(void)
 			   "CPU: CPUs started in inconsistent modes");
 	else
 		pr_info("CPU: All CPU(s) started at EL1\n");
-	if (IS_ENABLED(CONFIG_KVM))
+	if (IS_ENABLED(CONFIG_KVM) && !is_kernel_in_hyp_mode())
 		kvm_compute_layout();
 }
 

arch/arm64/kvm/Kconfig

@@ -49,14 +49,6 @@ if KVM
 
 source "virt/kvm/Kconfig"
 
-config KVM_ARM_PMU
-	bool "Virtual Performance Monitoring Unit (PMU) support"
-	depends on HW_PERF_EVENTS
-	default y
-	help
-	  Adds support for a virtual Performance Monitoring Unit (PMU) in
-	  virtual machines.
-
 endif # KVM
 
 endif # VIRTUALIZATION

arch/arm64/kvm/Makefile

@@ -24,4 +24,4 @@ kvm-y := $(KVM)/kvm_main.o $(KVM)/coalesced_mmio.o $(KVM)/eventfd.o \
 	 vgic/vgic-mmio-v3.o vgic/vgic-kvm-device.o \
 	 vgic/vgic-its.o vgic/vgic-debug.o
 
-kvm-$(CONFIG_KVM_ARM_PMU)  += pmu-emul.o
+kvm-$(CONFIG_HW_PERF_EVENTS)  += pmu-emul.o

arch/arm64/kvm/arch_timer.c

@@ -1129,9 +1129,10 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu)
 	if (!irqchip_in_kernel(vcpu->kvm))
 		goto no_vgic;
 
-	if (!vgic_initialized(vcpu->kvm))
+	/*
-		return -ENODEV;
+	 * At this stage, we have the guarantee that the vgic is both
-
+	 * available and initialized.
+	 */
 	if (!timer_irqs_are_valid(vcpu)) {
 		kvm_debug("incorrectly configured timer irqs\n");
 		return -EINVAL;

arch/arm64/kvm/arm.c

@@ -65,10 +65,6 @@ static bool vgic_present;
 static DEFINE_PER_CPU(unsigned char, kvm_arm_hardware_enabled);
 DEFINE_STATIC_KEY_FALSE(userspace_irqchip_in_use);
 
-extern u64 kvm_nvhe_sym(__cpu_logical_map)[NR_CPUS];
-extern u32 kvm_nvhe_sym(kvm_host_psci_version);
-extern struct psci_0_1_function_ids kvm_nvhe_sym(kvm_host_psci_0_1_function_ids);
-
 int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu)
 {
 	return kvm_vcpu_exiting_guest_mode(vcpu) == IN_GUEST_MODE;
@@ -584,11 +580,9 @@ static int kvm_vcpu_first_run_init(struct kvm_vcpu *vcpu)
 		 * Map the VGIC hardware resources before running a vcpu the
 		 * first time on this VM.
 		 */
-		if (unlikely(!vgic_ready(kvm))) {
+		ret = kvm_vgic_map_resources(kvm);
-			ret = kvm_vgic_map_resources(kvm);
+		if (ret)
-			if (ret)
+			return ret;
-				return ret;
-		}
 	} else {
 		/*
 		 * Tell the rest of the code that there are userspace irqchip
@@ -1574,12 +1568,12 @@ static struct notifier_block hyp_init_cpu_pm_nb = {
 	.notifier_call = hyp_init_cpu_pm_notifier,
 };
 
-static void __init hyp_cpu_pm_init(void)
+static void hyp_cpu_pm_init(void)
 {
 	if (!is_protected_kvm_enabled())
 		cpu_pm_register_notifier(&hyp_init_cpu_pm_nb);
 }
-static void __init hyp_cpu_pm_exit(void)
+static void hyp_cpu_pm_exit(void)
 {
 	if (!is_protected_kvm_enabled())
 		cpu_pm_unregister_notifier(&hyp_init_cpu_pm_nb);
@@ -1604,9 +1598,12 @@ static void init_cpu_logical_map(void)
 	 * allow any other CPUs from the `possible` set to boot.
 	 */
 	for_each_online_cpu(cpu)
-		kvm_nvhe_sym(__cpu_logical_map)[cpu] = cpu_logical_map(cpu);
+		hyp_cpu_logical_map[cpu] = cpu_logical_map(cpu);
 }
 
+#define init_psci_0_1_impl_state(config, what)	\
+	config.psci_0_1_ ## what ## _implemented = psci_ops.what
+
 static bool init_psci_relay(void)
 {
 	/*
@@ -1618,8 +1615,15 @@ static bool init_psci_relay(void)
 		return false;
 	}
 
-	kvm_nvhe_sym(kvm_host_psci_version) = psci_ops.get_version();
+	kvm_host_psci_config.version = psci_ops.get_version();
-	kvm_nvhe_sym(kvm_host_psci_0_1_function_ids) = get_psci_0_1_function_ids();
+
+	if (kvm_host_psci_config.version == PSCI_VERSION(0, 1)) {
+		kvm_host_psci_config.function_ids_0_1 = get_psci_0_1_function_ids();
+		init_psci_0_1_impl_state(kvm_host_psci_config, cpu_suspend);
+		init_psci_0_1_impl_state(kvm_host_psci_config, cpu_on);
+		init_psci_0_1_impl_state(kvm_host_psci_config, cpu_off);
+		init_psci_0_1_impl_state(kvm_host_psci_config, migrate);
+	}
 	return true;
 }
 

arch/arm64/kvm/hyp/include/hyp/adjust_pc.h

@@ -59,4 +59,13 @@ static inline void __adjust_pc(struct kvm_vcpu *vcpu)
 	}
 }
 
+/*
+ * Skip an instruction while host sysregs are live.
+ * Assumes host is always 64-bit.
+ */
+static inline void kvm_skip_host_instr(void)
+{
+	write_sysreg_el2(read_sysreg_el2(SYS_ELR) + 4, SYS_ELR);
+}
+
 #endif

arch/arm64/kvm/hyp/nvhe/hyp-main.c

@@ -157,11 +157,6 @@ static void default_host_smc_handler(struct kvm_cpu_context *host_ctxt)
 	__kvm_hyp_host_forward_smc(host_ctxt);
 }
 
-static void skip_host_instruction(void)
-{
-	write_sysreg_el2(read_sysreg_el2(SYS_ELR) + 4, SYS_ELR);
-}
-
 static void handle_host_smc(struct kvm_cpu_context *host_ctxt)
 {
 	bool handled;
@@ -170,11 +165,8 @@ static void handle_host_smc(struct kvm_cpu_context *host_ctxt)
 	if (!handled)
 		default_host_smc_handler(host_ctxt);
 
-	/*
+	/* SMC was trapped, move ELR past the current PC. */
-	 * Unlike HVC, the return address of an SMC is the instruction's PC.
+	kvm_skip_host_instr();
-	 * Move the return address past the instruction.
-	 */
-	skip_host_instruction();
 }
 
 void handle_trap(struct kvm_cpu_context *host_ctxt)

arch/arm64/kvm/hyp/nvhe/hyp-smp.c

@@ -14,14 +14,14 @@
  * Other CPUs should not be allowed to boot because their features were
  * not checked against the finalized system capabilities.
  */
-u64 __ro_after_init __cpu_logical_map[NR_CPUS] = { [0 ... NR_CPUS-1] = INVALID_HWID };
+u64 __ro_after_init hyp_cpu_logical_map[NR_CPUS] = { [0 ... NR_CPUS-1] = INVALID_HWID };
 
 u64 cpu_logical_map(unsigned int cpu)
 {
-	if (cpu >= ARRAY_SIZE(__cpu_logical_map))
+	if (cpu >= ARRAY_SIZE(hyp_cpu_logical_map))
 		hyp_panic();
 
-	return __cpu_logical_map[cpu];
+	return hyp_cpu_logical_map[cpu];
 }
 
 unsigned long __hyp_per_cpu_offset(unsigned int cpu)

arch/arm64/kvm/hyp/nvhe/psci-relay.c

@@ -7,11 +7,8 @@
 #include <asm/kvm_asm.h>
 #include <asm/kvm_hyp.h>
 #include <asm/kvm_mmu.h>
-#include <kvm/arm_hypercalls.h>
 #include <linux/arm-smccc.h>
 #include <linux/kvm_host.h>
-#include <linux/psci.h>
-#include <kvm/arm_psci.h>
 #include <uapi/linux/psci.h>
 
 #include <nvhe/trap_handler.h>
@@ -22,9 +19,8 @@ void kvm_hyp_cpu_resume(unsigned long r0);
 void __noreturn __host_enter(struct kvm_cpu_context *host_ctxt);
 
 /* Config options set by the host. */
-__ro_after_init u32 kvm_host_psci_version;
+struct kvm_host_psci_config __ro_after_init kvm_host_psci_config;
-__ro_after_init struct psci_0_1_function_ids kvm_host_psci_0_1_function_ids;
+s64 __ro_after_init hyp_physvirt_offset;
-__ro_after_init s64 hyp_physvirt_offset;
 
 #define __hyp_pa(x) ((phys_addr_t)((x)) + hyp_physvirt_offset)
 
@@ -47,19 +43,16 @@ struct psci_boot_args {
 static DEFINE_PER_CPU(struct psci_boot_args, cpu_on_args) = PSCI_BOOT_ARGS_INIT;
 static DEFINE_PER_CPU(struct psci_boot_args, suspend_args) = PSCI_BOOT_ARGS_INIT;
 
-static u64 get_psci_func_id(struct kvm_cpu_context *host_ctxt)
+#define	is_psci_0_1(what, func_id)					\
-{
+	(kvm_host_psci_config.psci_0_1_ ## what ## _implemented &&	\
-	DECLARE_REG(u64, func_id, host_ctxt, 0);
+	 (func_id) == kvm_host_psci_config.function_ids_0_1.what)
-
-	return func_id;
-}
 
 static bool is_psci_0_1_call(u64 func_id)
 {
-	return (func_id == kvm_host_psci_0_1_function_ids.cpu_suspend) ||
+	return (is_psci_0_1(cpu_suspend, func_id) ||
-	       (func_id == kvm_host_psci_0_1_function_ids.cpu_on) ||
+		is_psci_0_1(cpu_on, func_id) ||
-	       (func_id == kvm_host_psci_0_1_function_ids.cpu_off) ||
+		is_psci_0_1(cpu_off, func_id) ||
-	       (func_id == kvm_host_psci_0_1_function_ids.migrate);
+		is_psci_0_1(migrate, func_id));
 }
 
 static bool is_psci_0_2_call(u64 func_id)
@@ -69,16 +62,6 @@ static bool is_psci_0_2_call(u64 func_id)
 	       (PSCI_0_2_FN64(0) <= func_id && func_id <= PSCI_0_2_FN64(31));
 }
 
-static bool is_psci_call(u64 func_id)
-{
-	switch (kvm_host_psci_version) {
-	case PSCI_VERSION(0, 1):
-		return is_psci_0_1_call(func_id);
-	default:
-		return is_psci_0_2_call(func_id);
-	}
-}
-
 static unsigned long psci_call(unsigned long fn, unsigned long arg0,
 			       unsigned long arg1, unsigned long arg2)
 {
@@ -248,15 +231,14 @@ asmlinkage void __noreturn kvm_host_psci_cpu_entry(bool is_cpu_on)
 
 static unsigned long psci_0_1_handler(u64 func_id, struct kvm_cpu_context *host_ctxt)
 {
-	if ((func_id == kvm_host_psci_0_1_function_ids.cpu_off) ||
+	if (is_psci_0_1(cpu_off, func_id) || is_psci_0_1(migrate, func_id))
-	    (func_id == kvm_host_psci_0_1_function_ids.migrate))
 		return psci_forward(host_ctxt);
-	else if (func_id == kvm_host_psci_0_1_function_ids.cpu_on)
+	if (is_psci_0_1(cpu_on, func_id))
 		return psci_cpu_on(func_id, host_ctxt);
-	else if (func_id == kvm_host_psci_0_1_function_ids.cpu_suspend)
+	if (is_psci_0_1(cpu_suspend, func_id))
 		return psci_cpu_suspend(func_id, host_ctxt);
-	else
+
-		return PSCI_RET_NOT_SUPPORTED;
+	return PSCI_RET_NOT_SUPPORTED;
 }
 
 static unsigned long psci_0_2_handler(u64 func_id, struct kvm_cpu_context *host_ctxt)
@@ -298,20 +280,23 @@ static unsigned long psci_1_0_handler(u64 func_id, struct kvm_cpu_context *host_
 
 bool kvm_host_psci_handler(struct kvm_cpu_context *host_ctxt)
 {
-	u64 func_id = get_psci_func_id(host_ctxt);
+	DECLARE_REG(u64, func_id, host_ctxt, 0);
 	unsigned long ret;
 
-	if (!is_psci_call(func_id))
+	switch (kvm_host_psci_config.version) {
-		return false;
-
-	switch (kvm_host_psci_version) {
 	case PSCI_VERSION(0, 1):
+		if (!is_psci_0_1_call(func_id))
+			return false;
 		ret = psci_0_1_handler(func_id, host_ctxt);
 		break;
 	case PSCI_VERSION(0, 2):
+		if (!is_psci_0_2_call(func_id))
+			return false;
 		ret = psci_0_2_handler(func_id, host_ctxt);
 		break;
 	default:
+		if (!is_psci_0_2_call(func_id))
+			return false;
 		ret = psci_1_0_handler(func_id, host_ctxt);
 		break;
 	}

arch/arm64/kvm/pmu-emul.c

@@ -850,8 +850,6 @@ int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu)
 		   return -EINVAL;
 	}
 
-	kvm_pmu_vcpu_reset(vcpu);
-
 	return 0;
 }
 

arch/arm64/kvm/sys_regs.c

@@ -594,6 +594,10 @@ static void reset_pmcr(struct kvm_vcpu *vcpu, const struct sys_reg_desc *r)
 {
 	u64 pmcr, val;
 
+	/* No PMU available, PMCR_EL0 may UNDEF... */
+	if (!kvm_arm_support_pmu_v3())
+		return;
+
 	pmcr = read_sysreg(pmcr_el0);
 	/*
 	 * Writable bits of PMCR_EL0 (ARMV8_PMU_PMCR_MASK) are reset to UNKNOWN
@@ -919,7 +923,7 @@ static bool access_pmuserenr(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 
 #define reg_to_encoding(x)						\
 	sys_reg((u32)(x)->Op0, (u32)(x)->Op1,				\
-		(u32)(x)->CRn, (u32)(x)->CRm, (u32)(x)->Op2);
+		(u32)(x)->CRn, (u32)(x)->CRm, (u32)(x)->Op2)
 
 /* Silly macro to expand the DBG{BCR,BVR,WVR,WCR}n_EL1 registers in one go */
 #define DBG_BCR_BVR_WCR_WVR_EL1(n)					\

arch/arm64/kvm/va_layout.c

@@ -34,17 +34,16 @@ static u64 __early_kern_hyp_va(u64 addr)
 }
 
 /*
- * Store a hyp VA <-> PA offset into a hyp-owned variable.
+ * Store a hyp VA <-> PA offset into a EL2-owned variable.
  */
 static void init_hyp_physvirt_offset(void)
 {
-	extern s64 kvm_nvhe_sym(hyp_physvirt_offset);
 	u64 kern_va, hyp_va;
 
 	/* Compute the offset from the hyp VA and PA of a random symbol. */
-	kern_va = (u64)kvm_ksym_ref(__hyp_text_start);
+	kern_va = (u64)lm_alias(__hyp_text_start);
 	hyp_va = __early_kern_hyp_va(kern_va);
-	CHOOSE_NVHE_SYM(hyp_physvirt_offset) = (s64)__pa(kern_va) - (s64)hyp_va;
+	hyp_physvirt_offset = (s64)__pa(kern_va) - (s64)hyp_va;
 }
 
 /*

arch/arm64/kvm/vgic/vgic-init.c

@@ -419,7 +419,8 @@ int vgic_lazy_init(struct kvm *kvm)
  * Map the MMIO regions depending on the VGIC model exposed to the guest
  * called on the first VCPU run.
  * Also map the virtual CPU interface into the VM.
- * v2/v3 derivatives call vgic_init if not already done.
+ * v2 calls vgic_init() if not already done.
+ * v3 and derivatives return an error if the VGIC is not initialized.
  * vgic_ready() returns true if this function has succeeded.
  * @kvm: kvm struct pointer
  */
@@ -428,7 +429,13 @@ int kvm_vgic_map_resources(struct kvm *kvm)
 	struct vgic_dist *dist = &kvm->arch.vgic;
 	int ret = 0;
 
+	if (likely(vgic_ready(kvm)))
+		return 0;
+
 	mutex_lock(&kvm->lock);
+	if (vgic_ready(kvm))
+		goto out;
+
 	if (!irqchip_in_kernel(kvm))
 		goto out;
 
@@ -439,6 +446,8 @@ int kvm_vgic_map_resources(struct kvm *kvm)
 
 	if (ret)
 		__kvm_vgic_destroy(kvm);
+	else
+		dist->ready = true;
 
 out:
 	mutex_unlock(&kvm->lock);

arch/arm64/kvm/vgic/vgic-v2.c

@@ -306,20 +306,15 @@ int vgic_v2_map_resources(struct kvm *kvm)
 	struct vgic_dist *dist = &kvm->arch.vgic;
 	int ret = 0;
 
-	if (vgic_ready(kvm))
-		goto out;
-
 	if (IS_VGIC_ADDR_UNDEF(dist->vgic_dist_base) ||
 	    IS_VGIC_ADDR_UNDEF(dist->vgic_cpu_base)) {
 		kvm_err("Need to set vgic cpu and dist addresses first\n");
-		ret = -ENXIO;
+		return -ENXIO;
-		goto out;
 	}
 
 	if (!vgic_v2_check_base(dist->vgic_dist_base, dist->vgic_cpu_base)) {
 		kvm_err("VGIC CPU and dist frames overlap\n");
-		ret = -EINVAL;
+		return -EINVAL;
-		goto out;
 	}
 
 	/*
@@ -329,13 +324,13 @@ int vgic_v2_map_resources(struct kvm *kvm)
 	ret = vgic_init(kvm);
 	if (ret) {
 		kvm_err("Unable to initialize VGIC dynamic data structures\n");
-		goto out;
+		return ret;
 	}
 
 	ret = vgic_register_dist_iodev(kvm, dist->vgic_dist_base, VGIC_V2);
 	if (ret) {
 		kvm_err("Unable to register VGIC MMIO regions\n");
-		goto out;
+		return ret;
 	}
 
 	if (!static_branch_unlikely(&vgic_v2_cpuif_trap)) {
@@ -344,14 +339,11 @@ int vgic_v2_map_resources(struct kvm *kvm)
 					    KVM_VGIC_V2_CPU_SIZE, true);
 		if (ret) {
 			kvm_err("Unable to remap VGIC CPU to VCPU\n");
-			goto out;
+			return ret;
 		}
 	}
 
-	dist->ready = true;
+	return 0;
-
-out:
-	return ret;
 }
 
 DEFINE_STATIC_KEY_FALSE(vgic_v2_cpuif_trap);

arch/arm64/kvm/vgic/vgic-v3.c

@@ -500,29 +500,23 @@ int vgic_v3_map_resources(struct kvm *kvm)
 	int ret = 0;
 	int c;
 
-	if (vgic_ready(kvm))
-		goto out;
-
 	kvm_for_each_vcpu(c, vcpu, kvm) {
 		struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
 
 		if (IS_VGIC_ADDR_UNDEF(vgic_cpu->rd_iodev.base_addr)) {
 			kvm_debug("vcpu %d redistributor base not set\n", c);
-			ret = -ENXIO;
+			return -ENXIO;
-			goto out;
 		}
 	}
 
 	if (IS_VGIC_ADDR_UNDEF(dist->vgic_dist_base)) {
 		kvm_err("Need to set vgic distributor addresses first\n");
-		ret = -ENXIO;
+		return -ENXIO;
-		goto out;
 	}
 
 	if (!vgic_v3_check_base(kvm)) {
 		kvm_err("VGIC redist and dist frames overlap\n");
-		ret = -EINVAL;
+		return -EINVAL;
-		goto out;
 	}
 
 	/*
@@ -530,22 +524,19 @@ int vgic_v3_map_resources(struct kvm *kvm)
 	 * the VGIC before we need to use it.
 	 */
 	if (!vgic_initialized(kvm)) {
-		ret = -EBUSY;
+		return -EBUSY;
-		goto out;
 	}
 
 	ret = vgic_register_dist_iodev(kvm, dist->vgic_dist_base, VGIC_V3);
 	if (ret) {
 		kvm_err("Unable to register VGICv3 dist MMIO regions\n");
-		goto out;
+		return ret;
 	}
 
 	if (kvm_vgic_global_state.has_gicv4_1)
 		vgic_v4_configure_vsgis(kvm);
-	dist->ready = true;
 
-out:
+	return 0;
-	return ret;
 }
 
 DEFINE_STATIC_KEY_FALSE(vgic_v3_cpuif_trap);

arch/x86/include/asm/kvm_host.h

@@ -1010,9 +1010,21 @@ struct kvm_arch {
 	 */
 	bool tdp_mmu_enabled;
 
-	/* List of struct tdp_mmu_pages being used as roots */
+	/*
+	 * List of struct kvmp_mmu_pages being used as roots.
+	 * All struct kvm_mmu_pages in the list should have
+	 * tdp_mmu_page set.
+	 * All struct kvm_mmu_pages in the list should have a positive
+	 * root_count except when a thread holds the MMU lock and is removing
+	 * an entry from the list.
+	 */
 	struct list_head tdp_mmu_roots;
-	/* List of struct tdp_mmu_pages not being used as roots */
+
+	/*
+	 * List of struct kvmp_mmu_pages not being used as roots.
+	 * All struct kvm_mmu_pages in the list should have
+	 * tdp_mmu_page set and a root_count of 0.
+	 */
 	struct list_head tdp_mmu_pages;
 };
 
@@ -1287,6 +1299,8 @@ struct kvm_x86_ops {
 	void (*migrate_timers)(struct kvm_vcpu *vcpu);
 	void (*msr_filter_changed)(struct kvm_vcpu *vcpu);
 	int (*complete_emulated_msr)(struct kvm_vcpu *vcpu, int err);
+
+	void (*vcpu_deliver_sipi_vector)(struct kvm_vcpu *vcpu, u8 vector);
 };
 
 struct kvm_x86_nested_ops {
@@ -1468,6 +1482,7 @@ int kvm_fast_pio(struct kvm_vcpu *vcpu, int size, unsigned short port, int in);
 int kvm_emulate_cpuid(struct kvm_vcpu *vcpu);
 int kvm_emulate_halt(struct kvm_vcpu *vcpu);
 int kvm_vcpu_halt(struct kvm_vcpu *vcpu);
+int kvm_emulate_ap_reset_hold(struct kvm_vcpu *vcpu);
 int kvm_emulate_wbinvd(struct kvm_vcpu *vcpu);
 
 void kvm_get_segment(struct kvm_vcpu *vcpu, struct kvm_segment *var, int seg);

arch/x86/kvm/lapic.c

@@ -674,7 +674,7 @@ static bool pv_eoi_get_pending(struct kvm_vcpu *vcpu)
 			   (unsigned long long)vcpu->arch.pv_eoi.msr_val);
 		return false;
 	}
-	return val & 0x1;
+	return val & KVM_PV_EOI_ENABLED;
 }
 
 static void pv_eoi_set_pending(struct kvm_vcpu *vcpu)
@@ -2898,7 +2898,7 @@ void kvm_apic_accept_events(struct kvm_vcpu *vcpu)
 			/* evaluate pending_events before reading the vector */
 			smp_rmb();
 			sipi_vector = apic->sipi_vector;
-			kvm_vcpu_deliver_sipi_vector(vcpu, sipi_vector);
+			kvm_x86_ops.vcpu_deliver_sipi_vector(vcpu, sipi_vector);
 			vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
 		}
 	}

arch/x86/kvm/mmu.h

@@ -49,7 +49,7 @@ static inline u64 rsvd_bits(int s, int e)
 	if (e < s)
 		return 0;
 
-	return ((1ULL << (e - s + 1)) - 1) << s;
+	return ((2ULL << (e - s)) - 1) << s;
 }
 
 void kvm_mmu_set_mmio_spte_mask(u64 mmio_value, u64 access_mask);

arch/x86/kvm/mmu/mmu.c

@@ -3493,26 +3493,25 @@ static bool mmio_info_in_cache(struct kvm_vcpu *vcpu, u64 addr, bool direct)
  * Return the level of the lowest level SPTE added to sptes.
  * That SPTE may be non-present.
  */
-static int get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes)
+static int get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes, int *root_level)
 {
 	struct kvm_shadow_walk_iterator iterator;
-	int leaf = vcpu->arch.mmu->root_level;
+	int leaf = -1;
 	u64 spte;
 
-
 	walk_shadow_page_lockless_begin(vcpu);
 
-	for (shadow_walk_init(&iterator, vcpu, addr);
+	for (shadow_walk_init(&iterator, vcpu, addr),
+	     *root_level = iterator.level;
 	     shadow_walk_okay(&iterator);
 	     __shadow_walk_next(&iterator, spte)) {
 		leaf = iterator.level;
 		spte = mmu_spte_get_lockless(iterator.sptep);
 
-		sptes[leaf - 1] = spte;
+		sptes[leaf] = spte;
 
 		if (!is_shadow_present_pte(spte))
 			break;
-
 	}
 
 	walk_shadow_page_lockless_end(vcpu);
@@ -3520,14 +3519,12 @@ static int get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes)
 	return leaf;
 }
 
-/* return true if reserved bit is detected on spte. */
+/* return true if reserved bit(s) are detected on a valid, non-MMIO SPTE. */
 static bool get_mmio_spte(struct kvm_vcpu *vcpu, u64 addr, u64 *sptep)
 {
-	u64 sptes[PT64_ROOT_MAX_LEVEL];
+	u64 sptes[PT64_ROOT_MAX_LEVEL + 1];
 	struct rsvd_bits_validate *rsvd_check;
-	int root = vcpu->arch.mmu->shadow_root_level;
+	int root, leaf, level;
-	int leaf;
-	int level;
 	bool reserved = false;
 
 	if (!VALID_PAGE(vcpu->arch.mmu->root_hpa)) {
@@ -3536,35 +3533,45 @@ static bool get_mmio_spte(struct kvm_vcpu *vcpu, u64 addr, u64 *sptep)
 	}
 
 	if (is_tdp_mmu_root(vcpu->kvm, vcpu->arch.mmu->root_hpa))
-		leaf = kvm_tdp_mmu_get_walk(vcpu, addr, sptes);
+		leaf = kvm_tdp_mmu_get_walk(vcpu, addr, sptes, &root);
 	else
-		leaf = get_walk(vcpu, addr, sptes);
+		leaf = get_walk(vcpu, addr, sptes, &root);
+
+	if (unlikely(leaf < 0)) {
+		*sptep = 0ull;
+		return reserved;
+	}
+
+	*sptep = sptes[leaf];
+
+	/*
+	 * Skip reserved bits checks on the terminal leaf if it's not a valid
+	 * SPTE.  Note, this also (intentionally) skips MMIO SPTEs, which, by
+	 * design, always have reserved bits set.  The purpose of the checks is
+	 * to detect reserved bits on non-MMIO SPTEs. i.e. buggy SPTEs.
+	 */
+	if (!is_shadow_present_pte(sptes[leaf]))
+		leaf++;
 
 	rsvd_check = &vcpu->arch.mmu->shadow_zero_check;
 
-	for (level = root; level >= leaf; level--) {
+	for (level = root; level >= leaf; level--)
-		if (!is_shadow_present_pte(sptes[level - 1]))
-			break;
 		/*
 		 * Use a bitwise-OR instead of a logical-OR to aggregate the
 		 * reserved bit and EPT's invalid memtype/XWR checks to avoid
 		 * adding a Jcc in the loop.
 		 */
-		reserved |= __is_bad_mt_xwr(rsvd_check, sptes[level - 1]) |
+		reserved |= __is_bad_mt_xwr(rsvd_check, sptes[level]) |
-			    __is_rsvd_bits_set(rsvd_check, sptes[level - 1],
+			    __is_rsvd_bits_set(rsvd_check, sptes[level], level);
-					       level);
-	}
 
 	if (reserved) {
 		pr_err("%s: detect reserved bits on spte, addr 0x%llx, dump hierarchy:\n",
 		       __func__, addr);
 		for (level = root; level >= leaf; level--)
 			pr_err("------ spte 0x%llx level %d.\n",
-			       sptes[level - 1], level);
+			       sptes[level], level);
 	}
 
-	*sptep = sptes[leaf - 1];
-
 	return reserved;
 }
 

arch/x86/kvm/mmu/tdp_mmu.c

@@ -44,7 +44,48 @@ void kvm_mmu_uninit_tdp_mmu(struct kvm *kvm)
 	WARN_ON(!list_empty(&kvm->arch.tdp_mmu_roots));
 }
 
-#define for_each_tdp_mmu_root(_kvm, _root)			    \
+static void tdp_mmu_put_root(struct kvm *kvm, struct kvm_mmu_page *root)
+{
+	if (kvm_mmu_put_root(kvm, root))
+		kvm_tdp_mmu_free_root(kvm, root);
+}
+
+static inline bool tdp_mmu_next_root_valid(struct kvm *kvm,
+					   struct kvm_mmu_page *root)
+{
+	lockdep_assert_held(&kvm->mmu_lock);
+
+	if (list_entry_is_head(root, &kvm->arch.tdp_mmu_roots, link))
+		return false;
+
+	kvm_mmu_get_root(kvm, root);
+	return true;
+
+}
+
+static inline struct kvm_mmu_page *tdp_mmu_next_root(struct kvm *kvm,
+						     struct kvm_mmu_page *root)
+{
+	struct kvm_mmu_page *next_root;
+
+	next_root = list_next_entry(root, link);
+	tdp_mmu_put_root(kvm, root);
+	return next_root;
+}
+
+/*
+ * Note: this iterator gets and puts references to the roots it iterates over.
+ * This makes it safe to release the MMU lock and yield within the loop, but
+ * if exiting the loop early, the caller must drop the reference to the most
+ * recent root. (Unless keeping a live reference is desirable.)
+ */
+#define for_each_tdp_mmu_root_yield_safe(_kvm, _root)				\
+	for (_root = list_first_entry(&_kvm->arch.tdp_mmu_roots,	\
+				      typeof(*_root), link);		\
+	     tdp_mmu_next_root_valid(_kvm, _root);			\
+	     _root = tdp_mmu_next_root(_kvm, _root))
+
+#define for_each_tdp_mmu_root(_kvm, _root)				\
 	list_for_each_entry(_root, &_kvm->arch.tdp_mmu_roots, link)
 
 bool is_tdp_mmu_root(struct kvm *kvm, hpa_t hpa)
@@ -447,18 +488,9 @@ bool kvm_tdp_mmu_zap_gfn_range(struct kvm *kvm, gfn_t start, gfn_t end)
 	struct kvm_mmu_page *root;
 	bool flush = false;
 
-	for_each_tdp_mmu_root(kvm, root) {
+	for_each_tdp_mmu_root_yield_safe(kvm, root)
-		/*
-		 * Take a reference on the root so that it cannot be freed if
-		 * this thread releases the MMU lock and yields in this loop.
-		 */
-		kvm_mmu_get_root(kvm, root);
-
 		flush |= zap_gfn_range(kvm, root, start, end, true);
 
-		kvm_mmu_put_root(kvm, root);
-	}
-
 	return flush;
 }
 
@@ -619,13 +651,7 @@ static int kvm_tdp_mmu_handle_hva_range(struct kvm *kvm, unsigned long start,
 	int ret = 0;
 	int as_id;
 
-	for_each_tdp_mmu_root(kvm, root) {
+	for_each_tdp_mmu_root_yield_safe(kvm, root) {
-		/*
-		 * Take a reference on the root so that it cannot be freed if
-		 * this thread releases the MMU lock and yields in this loop.
-		 */
-		kvm_mmu_get_root(kvm, root);
-
 		as_id = kvm_mmu_page_as_id(root);
 		slots = __kvm_memslots(kvm, as_id);
 		kvm_for_each_memslot(memslot, slots) {
@@ -647,8 +673,6 @@ static int kvm_tdp_mmu_handle_hva_range(struct kvm *kvm, unsigned long start,
 			ret |= handler(kvm, memslot, root, gfn_start,
 				       gfn_end, data);
 		}
-
-		kvm_mmu_put_root(kvm, root);
 	}
 
 	return ret;
@@ -838,21 +862,13 @@ bool kvm_tdp_mmu_wrprot_slot(struct kvm *kvm, struct kvm_memory_slot *slot,
 	int root_as_id;
 	bool spte_set = false;
 
-	for_each_tdp_mmu_root(kvm, root) {
+	for_each_tdp_mmu_root_yield_safe(kvm, root) {
 		root_as_id = kvm_mmu_page_as_id(root);
 		if (root_as_id != slot->as_id)
 			continue;
 
-		/*
-		 * Take a reference on the root so that it cannot be freed if
-		 * this thread releases the MMU lock and yields in this loop.
-		 */
-		kvm_mmu_get_root(kvm, root);
-
 		spte_set |= wrprot_gfn_range(kvm, root, slot->base_gfn,
 			     slot->base_gfn + slot->npages, min_level);
-
-		kvm_mmu_put_root(kvm, root);
 	}
 
 	return spte_set;
@@ -906,21 +922,13 @@ bool kvm_tdp_mmu_clear_dirty_slot(struct kvm *kvm, struct kvm_memory_slot *slot)
 	int root_as_id;
 	bool spte_set = false;
 
-	for_each_tdp_mmu_root(kvm, root) {
+	for_each_tdp_mmu_root_yield_safe(kvm, root) {
 		root_as_id = kvm_mmu_page_as_id(root);
 		if (root_as_id != slot->as_id)
 			continue;
 
-		/*
-		 * Take a reference on the root so that it cannot be freed if
-		 * this thread releases the MMU lock and yields in this loop.
-		 */
-		kvm_mmu_get_root(kvm, root);
-
 		spte_set |= clear_dirty_gfn_range(kvm, root, slot->base_gfn,
 				slot->base_gfn + slot->npages);
-
-		kvm_mmu_put_root(kvm, root);
 	}
 
 	return spte_set;
@@ -1029,21 +1037,13 @@ bool kvm_tdp_mmu_slot_set_dirty(struct kvm *kvm, struct kvm_memory_slot *slot)
 	int root_as_id;
 	bool spte_set = false;
 
-	for_each_tdp_mmu_root(kvm, root) {
+	for_each_tdp_mmu_root_yield_safe(kvm, root) {
 		root_as_id = kvm_mmu_page_as_id(root);
 		if (root_as_id != slot->as_id)
 			continue;
 
-		/*
-		 * Take a reference on the root so that it cannot be freed if
-		 * this thread releases the MMU lock and yields in this loop.
-		 */
-		kvm_mmu_get_root(kvm, root);
-
 		spte_set |= set_dirty_gfn_range(kvm, root, slot->base_gfn,
 				slot->base_gfn + slot->npages);
-
-		kvm_mmu_put_root(kvm, root);
 	}
 	return spte_set;
 }
@@ -1089,21 +1089,13 @@ void kvm_tdp_mmu_zap_collapsible_sptes(struct kvm *kvm,
 	struct kvm_mmu_page *root;
 	int root_as_id;
 
-	for_each_tdp_mmu_root(kvm, root) {
+	for_each_tdp_mmu_root_yield_safe(kvm, root) {
 		root_as_id = kvm_mmu_page_as_id(root);
 		if (root_as_id != slot->as_id)
 			continue;
 
-		/*
-		 * Take a reference on the root so that it cannot be freed if
-		 * this thread releases the MMU lock and yields in this loop.
-		 */
-		kvm_mmu_get_root(kvm, root);
-
 		zap_collapsible_spte_range(kvm, root, slot->base_gfn,
 					   slot->base_gfn + slot->npages);
-
-		kvm_mmu_put_root(kvm, root);
 	}
 }
 
@@ -1160,16 +1152,19 @@ bool kvm_tdp_mmu_write_protect_gfn(struct kvm *kvm,
  * Return the level of the lowest level SPTE added to sptes.
  * That SPTE may be non-present.
  */
-int kvm_tdp_mmu_get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes)
+int kvm_tdp_mmu_get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes,
+			 int *root_level)
 {
 	struct tdp_iter iter;
 	struct kvm_mmu *mmu = vcpu->arch.mmu;
-	int leaf = vcpu->arch.mmu->shadow_root_level;
 	gfn_t gfn = addr >> PAGE_SHIFT;
+	int leaf = -1;
+
+	*root_level = vcpu->arch.mmu->shadow_root_level;
 
 	tdp_mmu_for_each_pte(iter, mmu, gfn, gfn + 1) {
 		leaf = iter.level;
-		sptes[leaf - 1] = iter.old_spte;
+		sptes[leaf] = iter.old_spte;
 	}
 
 	return leaf;

arch/x86/kvm/mmu/tdp_mmu.h

@@ -44,5 +44,7 @@ void kvm_tdp_mmu_zap_collapsible_sptes(struct kvm *kvm,
 bool kvm_tdp_mmu_write_protect_gfn(struct kvm *kvm,
 				   struct kvm_memory_slot *slot, gfn_t gfn);
 
-int kvm_tdp_mmu_get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes);
+int kvm_tdp_mmu_get_walk(struct kvm_vcpu *vcpu, u64 addr, u64 *sptes,
+			 int *root_level);
+
 #endif /* __KVM_X86_MMU_TDP_MMU_H */

arch/x86/kvm/svm/nested.c

@@ -199,6 +199,7 @@ static bool nested_svm_vmrun_msrpm(struct vcpu_svm *svm)
 static bool svm_get_nested_state_pages(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
+
 	if (!nested_svm_vmrun_msrpm(svm)) {
 		vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR;
 		vcpu->run->internal.suberror =
@@ -595,6 +596,8 @@ int nested_svm_vmexit(struct vcpu_svm *svm)
 	svm->nested.vmcb12_gpa = 0;
 	WARN_ON_ONCE(svm->nested.nested_run_pending);
 
+	kvm_clear_request(KVM_REQ_GET_NESTED_STATE_PAGES, &svm->vcpu);
+
 	/* in case we halted in L2 */
 	svm->vcpu.arch.mp_state = KVM_MP_STATE_RUNNABLE;
 
@@ -754,6 +757,7 @@ void svm_leave_nested(struct vcpu_svm *svm)
 		leave_guest_mode(&svm->vcpu);
 		copy_vmcb_control_area(&vmcb->control, &hsave->control);
 		nested_svm_uninit_mmu_context(&svm->vcpu);
+		vmcb_mark_all_dirty(svm->vmcb);
 	}
 
 	kvm_clear_request(KVM_REQ_GET_NESTED_STATE_PAGES, &svm->vcpu);
@@ -1194,6 +1198,10 @@ static int svm_set_nested_state(struct kvm_vcpu *vcpu,
 	 * in the registers, the save area of the nested state instead
 	 * contains saved L1 state.
 	 */
+
+	svm->nested.nested_run_pending =
+		!!(kvm_state->flags & KVM_STATE_NESTED_RUN_PENDING);
+
 	copy_vmcb_control_area(&hsave->control, &svm->vmcb->control);
 	hsave->save = *save;
 

arch/x86/kvm/svm/sev.c

@@ -1563,6 +1563,7 @@ static int sev_es_validate_vmgexit(struct vcpu_svm *svm)
 			goto vmgexit_err;
 		break;
 	case SVM_VMGEXIT_NMI_COMPLETE:
+	case SVM_VMGEXIT_AP_HLT_LOOP:
 	case SVM_VMGEXIT_AP_JUMP_TABLE:
 	case SVM_VMGEXIT_UNSUPPORTED_EVENT:
 		break;
@@ -1888,6 +1889,9 @@ int sev_handle_vmgexit(struct vcpu_svm *svm)
 	case SVM_VMGEXIT_NMI_COMPLETE:
 		ret = svm_invoke_exit_handler(svm, SVM_EXIT_IRET);
 		break;
+	case SVM_VMGEXIT_AP_HLT_LOOP:
+		ret = kvm_emulate_ap_reset_hold(&svm->vcpu);
+		break;
 	case SVM_VMGEXIT_AP_JUMP_TABLE: {
 		struct kvm_sev_info *sev = &to_kvm_svm(svm->vcpu.kvm)->sev_info;
 
@@ -2001,7 +2005,7 @@ void sev_es_vcpu_load(struct vcpu_svm *svm, int cpu)
 	 * of which one step is to perform a VMLOAD. Since hardware does not
 	 * perform a VMSAVE on VMRUN, the host savearea must be updated.
 	 */
-	asm volatile(__ex("vmsave") : : "a" (__sme_page_pa(sd->save_area)) : "memory");
+	asm volatile(__ex("vmsave %0") : : "a" (__sme_page_pa(sd->save_area)) : "memory");
 
 	/*
 	 * Certain MSRs are restored on VMEXIT, only save ones that aren't
@@ -2040,3 +2044,21 @@ void sev_es_vcpu_put(struct vcpu_svm *svm)
 		wrmsrl(host_save_user_msrs[i].index, svm->host_user_msrs[i]);
 	}
 }
+
+void sev_vcpu_deliver_sipi_vector(struct kvm_vcpu *vcpu, u8 vector)
+{
+	struct vcpu_svm *svm = to_svm(vcpu);
+
+	/* First SIPI: Use the values as initially set by the VMM */
+	if (!svm->received_first_sipi) {
+		svm->received_first_sipi = true;
+		return;
+	}
+
+	/*
+	 * Subsequent SIPI: Return from an AP Reset Hold VMGEXIT, where
+	 * the guest will set the CS and RIP. Set SW_EXIT_INFO_2 to a
+	 * non-zero value.
+	 */
+	ghcb_set_sw_exit_info_2(svm->ghcb, 1);
+}

arch/x86/kvm/svm/svm.c

@@ -3677,8 +3677,6 @@ static fastpath_t svm_exit_handlers_fastpath(struct kvm_vcpu *vcpu)
 	return EXIT_FASTPATH_NONE;
 }
 
-void __svm_vcpu_run(unsigned long vmcb_pa, unsigned long *regs);
-
 static noinstr void svm_vcpu_enter_exit(struct kvm_vcpu *vcpu,
 					struct vcpu_svm *svm)
 {
@@ -4384,6 +4382,14 @@ static bool svm_apic_init_signal_blocked(struct kvm_vcpu *vcpu)
 		   (vmcb_is_intercept(&svm->vmcb->control, INTERCEPT_INIT));
 }
 
+static void svm_vcpu_deliver_sipi_vector(struct kvm_vcpu *vcpu, u8 vector)
+{
+	if (!sev_es_guest(vcpu->kvm))
+		return kvm_vcpu_deliver_sipi_vector(vcpu, vector);
+
+	sev_vcpu_deliver_sipi_vector(vcpu, vector);
+}
+
 static void svm_vm_destroy(struct kvm *kvm)
 {
 	avic_vm_destroy(kvm);
@@ -4526,6 +4532,8 @@ static struct kvm_x86_ops svm_x86_ops __initdata = {
 
 	.msr_filter_changed = svm_msr_filter_changed,
 	.complete_emulated_msr = svm_complete_emulated_msr,
+
+	.vcpu_deliver_sipi_vector = svm_vcpu_deliver_sipi_vector,
 };
 
 static struct kvm_x86_init_ops svm_init_ops __initdata = {

arch/x86/kvm/svm/svm.h

@@ -185,6 +185,7 @@ struct vcpu_svm {
 	struct vmcb_save_area *vmsa;
 	struct ghcb *ghcb;
 	struct kvm_host_map ghcb_map;
+	bool received_first_sipi;
 
 	/* SEV-ES scratch area support */
 	void *ghcb_sa;
@@ -591,6 +592,7 @@ void sev_es_init_vmcb(struct vcpu_svm *svm);
 void sev_es_create_vcpu(struct vcpu_svm *svm);
 void sev_es_vcpu_load(struct vcpu_svm *svm, int cpu);
 void sev_es_vcpu_put(struct vcpu_svm *svm);
+void sev_vcpu_deliver_sipi_vector(struct kvm_vcpu *vcpu, u8 vector);
 
 /* vmenter.S */
 

arch/x86/kvm/vmx/nested.c

@@ -4442,6 +4442,8 @@ void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 vm_exit_reason,
 	/* trying to cancel vmlaunch/vmresume is a bug */
 	WARN_ON_ONCE(vmx->nested.nested_run_pending);
 
+	kvm_clear_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu);
+
 	/* Service the TLB flush request for L2 before switching to L1. */
 	if (kvm_check_request(KVM_REQ_TLB_FLUSH_CURRENT, vcpu))
 		kvm_vcpu_flush_tlb_current(vcpu);

arch/x86/kvm/vmx/vmx.c

@@ -7707,6 +7707,8 @@ static struct kvm_x86_ops vmx_x86_ops __initdata = {
 	.msr_filter_changed = vmx_msr_filter_changed,
 	.complete_emulated_msr = kvm_complete_insn_gp,
 	.cpu_dirty_log_size = vmx_cpu_dirty_log_size,
+
+	.vcpu_deliver_sipi_vector = kvm_vcpu_deliver_sipi_vector,
 };
 
 static __init int hardware_setup(void)

arch/x86/kvm/x86.c

@@ -7976,17 +7976,22 @@ void kvm_arch_exit(void)
 	kmem_cache_destroy(x86_fpu_cache);
 }
 
-int kvm_vcpu_halt(struct kvm_vcpu *vcpu)
+static int __kvm_vcpu_halt(struct kvm_vcpu *vcpu, int state, int reason)
 {
 	++vcpu->stat.halt_exits;
 	if (lapic_in_kernel(vcpu)) {
-		vcpu->arch.mp_state = KVM_MP_STATE_HALTED;
+		vcpu->arch.mp_state = state;
 		return 1;
 	} else {
-		vcpu->run->exit_reason = KVM_EXIT_HLT;
+		vcpu->run->exit_reason = reason;
 		return 0;
 	}
 }
+
+int kvm_vcpu_halt(struct kvm_vcpu *vcpu)
+{
+	return __kvm_vcpu_halt(vcpu, KVM_MP_STATE_HALTED, KVM_EXIT_HLT);
+}
 EXPORT_SYMBOL_GPL(kvm_vcpu_halt);
 
 int kvm_emulate_halt(struct kvm_vcpu *vcpu)
@@ -8000,6 +8005,14 @@ int kvm_emulate_halt(struct kvm_vcpu *vcpu)
 }
 EXPORT_SYMBOL_GPL(kvm_emulate_halt);
 
+int kvm_emulate_ap_reset_hold(struct kvm_vcpu *vcpu)
+{
+	int ret = kvm_skip_emulated_instruction(vcpu);
+
+	return __kvm_vcpu_halt(vcpu, KVM_MP_STATE_AP_RESET_HOLD, KVM_EXIT_AP_RESET_HOLD) && ret;
+}
+EXPORT_SYMBOL_GPL(kvm_emulate_ap_reset_hold);
+
 #ifdef CONFIG_X86_64
 static int kvm_pv_clock_pairing(struct kvm_vcpu *vcpu, gpa_t paddr,
 			        unsigned long clock_type)
@@ -8789,7 +8802,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 
 	if (kvm_request_pending(vcpu)) {
 		if (kvm_check_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu)) {
-			if (unlikely(!kvm_x86_ops.nested_ops->get_nested_state_pages(vcpu))) {
+			if (WARN_ON_ONCE(!is_guest_mode(vcpu)))
+				;
+			else if (unlikely(!kvm_x86_ops.nested_ops->get_nested_state_pages(vcpu))) {
 				r = 0;
 				goto out;
 			}
@@ -9094,6 +9109,7 @@ static inline int vcpu_block(struct kvm *kvm, struct kvm_vcpu *vcpu)
 	kvm_apic_accept_events(vcpu);
 	switch(vcpu->arch.mp_state) {
 	case KVM_MP_STATE_HALTED:
+	case KVM_MP_STATE_AP_RESET_HOLD:
 		vcpu->arch.pv.pv_unhalted = false;
 		vcpu->arch.mp_state =
 			KVM_MP_STATE_RUNNABLE;
@@ -9520,8 +9536,9 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 		kvm_load_guest_fpu(vcpu);
 
 	kvm_apic_accept_events(vcpu);
-	if (vcpu->arch.mp_state == KVM_MP_STATE_HALTED &&
+	if ((vcpu->arch.mp_state == KVM_MP_STATE_HALTED ||
-					vcpu->arch.pv.pv_unhalted)
+	     vcpu->arch.mp_state == KVM_MP_STATE_AP_RESET_HOLD) &&
+	    vcpu->arch.pv.pv_unhalted)
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 	else
 		mp_state->mp_state = vcpu->arch.mp_state;
@@ -10152,6 +10169,7 @@ void kvm_vcpu_deliver_sipi_vector(struct kvm_vcpu *vcpu, u8 vector)
 	kvm_set_segment(vcpu, &cs, VCPU_SREG_CS);
 	kvm_rip_write(vcpu, 0);
 }
+EXPORT_SYMBOL_GPL(kvm_vcpu_deliver_sipi_vector);
 
 int kvm_arch_hardware_enable(void)
 {

include/kvm/arm_pmu.h

@@ -13,7 +13,7 @@
 #define ARMV8_PMU_CYCLE_IDX		(ARMV8_PMU_MAX_COUNTERS - 1)
 #define ARMV8_PMU_MAX_COUNTER_PAIRS	((ARMV8_PMU_MAX_COUNTERS + 1) >> 1)
 
-#ifdef CONFIG_KVM_ARM_PMU
+#ifdef CONFIG_HW_PERF_EVENTS
 
 struct kvm_pmc {
 	u8 idx;	/* index into the pmu->pmc array */

include/uapi/linux/kvm.h

@@ -251,6 +251,7 @@ struct kvm_hyperv_exit {
 #define KVM_EXIT_X86_RDMSR        29
 #define KVM_EXIT_X86_WRMSR        30
 #define KVM_EXIT_DIRTY_RING_FULL  31
+#define KVM_EXIT_AP_RESET_HOLD    32
 
 /* For KVM_EXIT_INTERNAL_ERROR */
 /* Emulate instruction failed. */
@@ -573,6 +574,7 @@ struct kvm_vapic_addr {
 #define KVM_MP_STATE_CHECK_STOP        6
 #define KVM_MP_STATE_OPERATING         7
 #define KVM_MP_STATE_LOAD              8
+#define KVM_MP_STATE_AP_RESET_HOLD     9
 
 struct kvm_mp_state {
 	__u32 mp_state;

tools/testing/selftests/kvm/Makefile

@@ -33,7 +33,7 @@ ifeq ($(ARCH),s390)
 	UNAME_M := s390x
 endif
 
-LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c lib/test_util.c
+LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c lib/test_util.c lib/guest_modes.c lib/perf_test_util.c
 LIBKVM_x86_64 = lib/x86_64/processor.c lib/x86_64/vmx.c lib/x86_64/svm.c lib/x86_64/ucall.c lib/x86_64/handlers.S
 LIBKVM_aarch64 = lib/aarch64/processor.c lib/aarch64/ucall.c
 LIBKVM_s390x = lib/s390x/processor.c lib/s390x/ucall.c lib/s390x/diag318_test_handler.c

tools/testing/selftests/kvm/demand_paging_test.c

@@ -7,23 +7,20 @@
  * Copyright (C) 2019, Google, Inc.
  */
 
-#define _GNU_SOURCE /* for program_invocation_name */
+#define _GNU_SOURCE /* for pipe2 */
 
 #include <stdio.h>
 #include <stdlib.h>
-#include <sys/syscall.h>
-#include <unistd.h>
-#include <asm/unistd.h>
 #include <time.h>
 #include <poll.h>
 #include <pthread.h>
-#include <linux/bitmap.h>
-#include <linux/bitops.h>
 #include <linux/userfaultfd.h>
+#include <sys/syscall.h>
 
-#include "perf_test_util.h"
+#include "kvm_util.h"
-#include "processor.h"
 #include "test_util.h"
+#include "perf_test_util.h"
+#include "guest_modes.h"
 
 #ifdef __NR_userfaultfd
 
@@ -39,12 +36,14 @@
 #define PER_VCPU_DEBUG(...) _no_printf(__VA_ARGS__)
 #endif
 
+static int nr_vcpus = 1;
+static uint64_t guest_percpu_mem_size = DEFAULT_PER_VCPU_MEM_SIZE;
 static char *guest_data_prototype;
 
 static void *vcpu_worker(void *data)
 {
 	int ret;
-	struct vcpu_args *vcpu_args = (struct vcpu_args *)data;
+	struct perf_test_vcpu_args *vcpu_args = (struct perf_test_vcpu_args *)data;
 	int vcpu_id = vcpu_args->vcpu_id;
 	struct kvm_vm *vm = perf_test_args.vm;
 	struct kvm_run *run;
@@ -248,9 +247,14 @@ static int setup_demand_paging(struct kvm_vm *vm,
 	return 0;
 }
 
-static void run_test(enum vm_guest_mode mode, bool use_uffd,
+struct test_params {
-		     useconds_t uffd_delay)
+	bool use_uffd;
+	useconds_t uffd_delay;
+};
+
+static void run_test(enum vm_guest_mode mode, void *arg)
 {
+	struct test_params *p = arg;
 	pthread_t *vcpu_threads;
 	pthread_t *uffd_handler_threads = NULL;
 	struct uffd_handler_args *uffd_args = NULL;
@@ -261,7 +265,7 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 	int vcpu_id;
 	int r;
 
-	vm = create_vm(mode, nr_vcpus, guest_percpu_mem_size);
+	vm = perf_test_create_vm(mode, nr_vcpus, guest_percpu_mem_size);
 
 	perf_test_args.wr_fract = 1;
 
@@ -273,9 +277,9 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 	vcpu_threads = malloc(nr_vcpus * sizeof(*vcpu_threads));
 	TEST_ASSERT(vcpu_threads, "Memory allocation failed");
 
-	add_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
+	perf_test_setup_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
 
-	if (use_uffd) {
+	if (p->use_uffd) {
 		uffd_handler_threads =
 			malloc(nr_vcpus * sizeof(*uffd_handler_threads));
 		TEST_ASSERT(uffd_handler_threads, "Memory allocation failed");
@@ -308,7 +312,7 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 			r = setup_demand_paging(vm,
 						&uffd_handler_threads[vcpu_id],
 						pipefds[vcpu_id * 2],
-						uffd_delay, &uffd_args[vcpu_id],
+						p->uffd_delay, &uffd_args[vcpu_id],
 						vcpu_hva, guest_percpu_mem_size);
 			if (r < 0)
 				exit(-r);
@@ -339,7 +343,7 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 
 	pr_info("All vCPU threads joined\n");
 
-	if (use_uffd) {
+	if (p->use_uffd) {
 		char c;
 
 		/* Tell the user fault fd handler threads to quit */
@@ -357,43 +361,23 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 		perf_test_args.vcpu_args[0].pages * nr_vcpus /
 		((double)ts_diff.tv_sec + (double)ts_diff.tv_nsec / 100000000.0));
 
-	ucall_uninit(vm);
+	perf_test_destroy_vm(vm);
-	kvm_vm_free(vm);
 
 	free(guest_data_prototype);
 	free(vcpu_threads);
-	if (use_uffd) {
+	if (p->use_uffd) {
 		free(uffd_handler_threads);
 		free(uffd_args);
 		free(pipefds);
 	}
 }
 
-struct guest_mode {
-	bool supported;
-	bool enabled;
-};
-static struct guest_mode guest_modes[NUM_VM_MODES];
-
-#define guest_mode_init(mode, supported, enabled) ({ \
-	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
-})
-
 static void help(char *name)
 {
-	int i;
-
 	puts("");
 	printf("usage: %s [-h] [-m mode] [-u] [-d uffd_delay_usec]\n"
 	       "          [-b memory] [-v vcpus]\n", name);
-	printf(" -m: specify the guest mode ID to test\n"
+	guest_modes_help();
-	       "     (default: test all supported modes)\n"
-	       "     This option may be used multiple times.\n"
-	       "     Guest mode IDs:\n");
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
-		       guest_modes[i].supported ? " (supported)" : "");
-	}
 	printf(" -u: use User Fault FD to handle vCPU page\n"
 	       "     faults.\n");
 	printf(" -d: add a delay in usec to the User Fault\n"
@@ -410,53 +394,22 @@ static void help(char *name)
 int main(int argc, char *argv[])
 {
 	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
-	bool mode_selected = false;
+	struct test_params p = {};
-	unsigned int mode;
+	int opt;
-	int opt, i;
-	bool use_uffd = false;
-	useconds_t uffd_delay = 0;
 
-#ifdef __x86_64__
+	guest_modes_append_default();
-	guest_mode_init(VM_MODE_PXXV48_4K, true, true);
-#endif
-#ifdef __aarch64__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-	guest_mode_init(VM_MODE_P40V48_64K, true, true);
-	{
-		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
-
-		if (limit >= 52)
-			guest_mode_init(VM_MODE_P52V48_64K, true, true);
-		if (limit >= 48) {
-			guest_mode_init(VM_MODE_P48V48_4K, true, true);
-			guest_mode_init(VM_MODE_P48V48_64K, true, true);
-		}
-	}
-#endif
-#ifdef __s390x__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-#endif
 
 	while ((opt = getopt(argc, argv, "hm:ud:b:v:")) != -1) {
 		switch (opt) {
 		case 'm':
-			if (!mode_selected) {
+			guest_modes_cmdline(optarg);
-				for (i = 0; i < NUM_VM_MODES; ++i)
-					guest_modes[i].enabled = false;
-				mode_selected = true;
-			}
-			mode = strtoul(optarg, NULL, 10);
-			TEST_ASSERT(mode < NUM_VM_MODES,
-				    "Guest mode ID %d too big", mode);
-			guest_modes[mode].enabled = true;
 			break;
 		case 'u':
-			use_uffd = true;
+			p.use_uffd = true;
 			break;
 		case 'd':
-			uffd_delay = strtoul(optarg, NULL, 0);
+			p.uffd_delay = strtoul(optarg, NULL, 0);
-			TEST_ASSERT(uffd_delay >= 0,
+			TEST_ASSERT(p.uffd_delay >= 0, "A negative UFFD delay is not supported.");
-				    "A negative UFFD delay is not supported.");
 			break;
 		case 'b':
 			guest_percpu_mem_size = parse_size(optarg);
@@ -473,14 +426,7 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	for (i = 0; i < NUM_VM_MODES; ++i) {
+	for_each_guest_mode(run_test, &p);
-		if (!guest_modes[i].enabled)
-			continue;
-		TEST_ASSERT(guest_modes[i].supported,
-			    "Guest mode ID %d (%s) not supported.",
-			    i, vm_guest_mode_string(i));
-		run_test(i, use_uffd, uffd_delay);
-	}
 
 	return 0;
 }

tools/testing/selftests/kvm/dirty_log_perf_test.c

@@ -8,29 +8,28 @@
  * Copyright (C) 2020, Google, Inc.
  */
 
-#define _GNU_SOURCE /* for program_invocation_name */
-
 #include <stdio.h>
 #include <stdlib.h>
-#include <unistd.h>
 #include <time.h>
 #include <pthread.h>
 #include <linux/bitmap.h>
-#include <linux/bitops.h>
 
 #include "kvm_util.h"
-#include "perf_test_util.h"
-#include "processor.h"
 #include "test_util.h"
+#include "perf_test_util.h"
+#include "guest_modes.h"
 
 /* How many host loops to run by default (one KVM_GET_DIRTY_LOG for each loop)*/
 #define TEST_HOST_LOOP_N		2UL
 
+static int nr_vcpus = 1;
+static uint64_t guest_percpu_mem_size = DEFAULT_PER_VCPU_MEM_SIZE;
+
 /* Host variables */
 static u64 dirty_log_manual_caps;
 static bool host_quit;
 static uint64_t iteration;
-static uint64_t vcpu_last_completed_iteration[MAX_VCPUS];
+static uint64_t vcpu_last_completed_iteration[KVM_MAX_VCPUS];
 
 static void *vcpu_worker(void *data)
 {
@@ -42,7 +41,7 @@ static void *vcpu_worker(void *data)
 	struct timespec ts_diff;
 	struct timespec total = (struct timespec){0};
 	struct timespec avg;
-	struct vcpu_args *vcpu_args = (struct vcpu_args *)data;
+	struct perf_test_vcpu_args *vcpu_args = (struct perf_test_vcpu_args *)data;
 	int vcpu_id = vcpu_args->vcpu_id;
 
 	vcpu_args_set(vm, vcpu_id, 1, vcpu_id);
@@ -89,9 +88,15 @@ static void *vcpu_worker(void *data)
 	return NULL;
 }
 
-static void run_test(enum vm_guest_mode mode, unsigned long iterations,
+struct test_params {
-		     uint64_t phys_offset, int wr_fract)
+	unsigned long iterations;
+	uint64_t phys_offset;
+	int wr_fract;
+};
+
+static void run_test(enum vm_guest_mode mode, void *arg)
 {
+	struct test_params *p = arg;
 	pthread_t *vcpu_threads;
 	struct kvm_vm *vm;
 	unsigned long *bmap;
@@ -106,9 +111,9 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	struct kvm_enable_cap cap = {};
 	struct timespec clear_dirty_log_total = (struct timespec){0};
 
-	vm = create_vm(mode, nr_vcpus, guest_percpu_mem_size);
+	vm = perf_test_create_vm(mode, nr_vcpus, guest_percpu_mem_size);
 
-	perf_test_args.wr_fract = wr_fract;
+	perf_test_args.wr_fract = p->wr_fract;
 
 	guest_num_pages = (nr_vcpus * guest_percpu_mem_size) >> vm_get_page_shift(vm);
 	guest_num_pages = vm_adjust_num_guest_pages(mode, guest_num_pages);
@@ -124,7 +129,7 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	vcpu_threads = malloc(nr_vcpus * sizeof(*vcpu_threads));
 	TEST_ASSERT(vcpu_threads, "Memory allocation failed");
 
-	add_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
+	perf_test_setup_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
 
 	sync_global_to_guest(vm, perf_test_args);
 
@@ -150,13 +155,13 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 	/* Enable dirty logging */
 	clock_gettime(CLOCK_MONOTONIC, &start);
-	vm_mem_region_set_flags(vm, TEST_MEM_SLOT_INDEX,
+	vm_mem_region_set_flags(vm, PERF_TEST_MEM_SLOT_INDEX,
 				KVM_MEM_LOG_DIRTY_PAGES);
 	ts_diff = timespec_diff_now(start);
 	pr_info("Enabling dirty logging time: %ld.%.9lds\n\n",
 		ts_diff.tv_sec, ts_diff.tv_nsec);
 
-	while (iteration < iterations) {
+	while (iteration < p->iterations) {
 		/*
 		 * Incrementing the iteration number will start the vCPUs
 		 * dirtying memory again.
@@ -177,7 +182,7 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 			iteration, ts_diff.tv_sec, ts_diff.tv_nsec);
 
 		clock_gettime(CLOCK_MONOTONIC, &start);
-		kvm_vm_get_dirty_log(vm, TEST_MEM_SLOT_INDEX, bmap);
+		kvm_vm_get_dirty_log(vm, PERF_TEST_MEM_SLOT_INDEX, bmap);
 
 		ts_diff = timespec_diff_now(start);
 		get_dirty_log_total = timespec_add(get_dirty_log_total,
@@ -187,7 +192,7 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 		if (dirty_log_manual_caps) {
 			clock_gettime(CLOCK_MONOTONIC, &start);
-			kvm_vm_clear_dirty_log(vm, TEST_MEM_SLOT_INDEX, bmap, 0,
+			kvm_vm_clear_dirty_log(vm, PERF_TEST_MEM_SLOT_INDEX, bmap, 0,
 					       host_num_pages);
 
 			ts_diff = timespec_diff_now(start);
@@ -205,43 +210,30 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 	/* Disable dirty logging */
 	clock_gettime(CLOCK_MONOTONIC, &start);
-	vm_mem_region_set_flags(vm, TEST_MEM_SLOT_INDEX, 0);
+	vm_mem_region_set_flags(vm, PERF_TEST_MEM_SLOT_INDEX, 0);
 	ts_diff = timespec_diff_now(start);
 	pr_info("Disabling dirty logging time: %ld.%.9lds\n",
 		ts_diff.tv_sec, ts_diff.tv_nsec);
 
-	avg = timespec_div(get_dirty_log_total, iterations);
+	avg = timespec_div(get_dirty_log_total, p->iterations);
 	pr_info("Get dirty log over %lu iterations took %ld.%.9lds. (Avg %ld.%.9lds/iteration)\n",
-		iterations, get_dirty_log_total.tv_sec,
+		p->iterations, get_dirty_log_total.tv_sec,
 		get_dirty_log_total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 
 	if (dirty_log_manual_caps) {
-		avg = timespec_div(clear_dirty_log_total, iterations);
+		avg = timespec_div(clear_dirty_log_total, p->iterations);
 		pr_info("Clear dirty log over %lu iterations took %ld.%.9lds. (Avg %ld.%.9lds/iteration)\n",
-			iterations, clear_dirty_log_total.tv_sec,
+			p->iterations, clear_dirty_log_total.tv_sec,
 			clear_dirty_log_total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 	}
 
 	free(bmap);
 	free(vcpu_threads);
-	ucall_uninit(vm);
+	perf_test_destroy_vm(vm);
-	kvm_vm_free(vm);
 }
 
-struct guest_mode {
-	bool supported;
-	bool enabled;
-};
-static struct guest_mode guest_modes[NUM_VM_MODES];
-
-#define guest_mode_init(mode, supported, enabled) ({ \
-	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
-})
-
 static void help(char *name)
 {
-	int i;
-
 	puts("");
 	printf("usage: %s [-h] [-i iterations] [-p offset] "
 	       "[-m mode] [-b vcpu bytes] [-v vcpus]\n", name);
@@ -250,14 +242,7 @@ static void help(char *name)
 	       TEST_HOST_LOOP_N);
 	printf(" -p: specify guest physical test memory offset\n"
 	       "     Warning: a low offset can conflict with the loaded test code.\n");
-	printf(" -m: specify the guest mode ID to test "
+	guest_modes_help();
-	       "(default: test all supported modes)\n"
-	       "     This option may be used multiple times.\n"
-	       "     Guest mode IDs:\n");
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
-		       guest_modes[i].supported ? " (supported)" : "");
-	}
 	printf(" -b: specify the size of the memory region which should be\n"
 	       "     dirtied by each vCPU. e.g. 10M or 3G.\n"
 	       "     (default: 1G)\n");
@@ -272,74 +257,43 @@ static void help(char *name)
 
 int main(int argc, char *argv[])
 {
-	unsigned long iterations = TEST_HOST_LOOP_N;
+	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
-	bool mode_selected = false;
+	struct test_params p = {
-	uint64_t phys_offset = 0;
+		.iterations = TEST_HOST_LOOP_N,
-	unsigned int mode;
+		.wr_fract = 1,
-	int opt, i;
+	};
-	int wr_fract = 1;
+	int opt;
 
 	dirty_log_manual_caps =
 		kvm_check_cap(KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2);
 	dirty_log_manual_caps &= (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE |
 				  KVM_DIRTY_LOG_INITIALLY_SET);
 
-#ifdef __x86_64__
+	guest_modes_append_default();
-	guest_mode_init(VM_MODE_PXXV48_4K, true, true);
-#endif
-#ifdef __aarch64__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-	guest_mode_init(VM_MODE_P40V48_64K, true, true);
-
-	{
-		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
-
-		if (limit >= 52)
-			guest_mode_init(VM_MODE_P52V48_64K, true, true);
-		if (limit >= 48) {
-			guest_mode_init(VM_MODE_P48V48_4K, true, true);
-			guest_mode_init(VM_MODE_P48V48_64K, true, true);
-		}
-	}
-#endif
-#ifdef __s390x__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-#endif
 
 	while ((opt = getopt(argc, argv, "hi:p:m:b:f:v:")) != -1) {
 		switch (opt) {
 		case 'i':
-			iterations = strtol(optarg, NULL, 10);
+			p.iterations = strtol(optarg, NULL, 10);
 			break;
 		case 'p':
-			phys_offset = strtoull(optarg, NULL, 0);
+			p.phys_offset = strtoull(optarg, NULL, 0);
 			break;
 		case 'm':
-			if (!mode_selected) {
+			guest_modes_cmdline(optarg);
-				for (i = 0; i < NUM_VM_MODES; ++i)
-					guest_modes[i].enabled = false;
-				mode_selected = true;
-			}
-			mode = strtoul(optarg, NULL, 10);
-			TEST_ASSERT(mode < NUM_VM_MODES,
-				    "Guest mode ID %d too big", mode);
-			guest_modes[mode].enabled = true;
 			break;
 		case 'b':
 			guest_percpu_mem_size = parse_size(optarg);
 			break;
 		case 'f':
-			wr_fract = atoi(optarg);
+			p.wr_fract = atoi(optarg);
-			TEST_ASSERT(wr_fract >= 1,
+			TEST_ASSERT(p.wr_fract >= 1,
 				    "Write fraction cannot be less than one");
 			break;
 		case 'v':
 			nr_vcpus = atoi(optarg);
-			TEST_ASSERT(nr_vcpus > 0,
+			TEST_ASSERT(nr_vcpus > 0 && nr_vcpus <= max_vcpus,
-				    "Must have a positive number of vCPUs");
+				    "Invalid number of vcpus, must be between 1 and %d", max_vcpus);
-			TEST_ASSERT(nr_vcpus <= MAX_VCPUS,
-				    "This test does not currently support\n"
-				    "more than %d vCPUs.", MAX_VCPUS);
 			break;
 		case 'h':
 		default:
@@ -348,18 +302,11 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	TEST_ASSERT(iterations >= 2, "The test should have at least two iterations");
+	TEST_ASSERT(p.iterations >= 2, "The test should have at least two iterations");
 
-	pr_info("Test iterations: %"PRIu64"\n",	iterations);
+	pr_info("Test iterations: %"PRIu64"\n",	p.iterations);
 
-	for (i = 0; i < NUM_VM_MODES; ++i) {
+	for_each_guest_mode(run_test, &p);
-		if (!guest_modes[i].enabled)
-			continue;
-		TEST_ASSERT(guest_modes[i].supported,
-			    "Guest mode ID %d (%s) not supported.",
-			    i, vm_guest_mode_string(i));
-		run_test(i, iterations, phys_offset, wr_fract);
-	}
 
 	return 0;
 }

tools/testing/selftests/kvm/dirty_log_test.c

@@ -9,8 +9,6 @@
 
 #include <stdio.h>
 #include <stdlib.h>
-#include <unistd.h>
-#include <time.h>
 #include <pthread.h>
 #include <semaphore.h>
 #include <sys/types.h>
@@ -20,8 +18,9 @@
 #include <linux/bitops.h>
 #include <asm/barrier.h>
 
-#include "test_util.h"
 #include "kvm_util.h"
+#include "test_util.h"
+#include "guest_modes.h"
 #include "processor.h"
 
 #define VCPU_ID				1
@@ -673,9 +672,15 @@ static struct kvm_vm *create_vm(enum vm_guest_mode mode, uint32_t vcpuid,
 #define DIRTY_MEM_BITS 30 /* 1G */
 #define PAGE_SHIFT_4K  12
 
-static void run_test(enum vm_guest_mode mode, unsigned long iterations,
+struct test_params {
-		     unsigned long interval, uint64_t phys_offset)
+	unsigned long iterations;
+	unsigned long interval;
+	uint64_t phys_offset;
+};
+
+static void run_test(enum vm_guest_mode mode, void *arg)
 {
+	struct test_params *p = arg;
 	struct kvm_vm *vm;
 	unsigned long *bmap;
 
@@ -709,12 +714,12 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	host_page_size = getpagesize();
 	host_num_pages = vm_num_host_pages(mode, guest_num_pages);
 
-	if (!phys_offset) {
+	if (!p->phys_offset) {
 		guest_test_phys_mem = (vm_get_max_gfn(vm) -
 				       guest_num_pages) * guest_page_size;
 		guest_test_phys_mem &= ~(host_page_size - 1);
 	} else {
-		guest_test_phys_mem = phys_offset;
+		guest_test_phys_mem = p->phys_offset;
 	}
 
 #ifdef __s390x__
@@ -758,9 +763,9 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 	pthread_create(&vcpu_thread, NULL, vcpu_worker, vm);
 
-	while (iteration < iterations) {
+	while (iteration < p->iterations) {
 		/* Give the vcpu thread some time to dirty some pages */
-		usleep(interval * 1000);
+		usleep(p->interval * 1000);
 		log_mode_collect_dirty_pages(vm, TEST_MEM_SLOT_INDEX,
 					     bmap, host_num_pages);
 		vm_dirty_log_verify(mode, bmap);
@@ -783,20 +788,8 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	kvm_vm_free(vm);
 }
 
-struct guest_mode {
-	bool supported;
-	bool enabled;
-};
-static struct guest_mode guest_modes[NUM_VM_MODES];
-
-#define guest_mode_init(mode, supported, enabled) ({ \
-	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
-})
-
 static void help(char *name)
 {
-	int i;
-
 	puts("");
 	printf("usage: %s [-h] [-i iterations] [-I interval] "
 	       "[-p offset] [-m mode]\n", name);
@@ -813,51 +806,23 @@ static void help(char *name)
 	printf(" -M: specify the host logging mode "
 	       "(default: run all log modes).  Supported modes: \n\t");
 	log_modes_dump();
-	printf(" -m: specify the guest mode ID to test "
+	guest_modes_help();
-	       "(default: test all supported modes)\n"
-	       "     This option may be used multiple times.\n"
-	       "     Guest mode IDs:\n");
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
-		       guest_modes[i].supported ? " (supported)" : "");
-	}
 	puts("");
 	exit(0);
 }
 
 int main(int argc, char *argv[])
 {
-	unsigned long iterations = TEST_HOST_LOOP_N;
+	struct test_params p = {
-	unsigned long interval = TEST_HOST_LOOP_INTERVAL;
+		.iterations = TEST_HOST_LOOP_N,
-	bool mode_selected = false;
+		.interval = TEST_HOST_LOOP_INTERVAL,
-	uint64_t phys_offset = 0;
+	};
-	unsigned int mode;
+	int opt, i;
-	int opt, i, j;
 
 	sem_init(&dirty_ring_vcpu_stop, 0, 0);
 	sem_init(&dirty_ring_vcpu_cont, 0, 0);
 
-#ifdef __x86_64__
+	guest_modes_append_default();
-	guest_mode_init(VM_MODE_PXXV48_4K, true, true);
-#endif
-#ifdef __aarch64__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-	guest_mode_init(VM_MODE_P40V48_64K, true, true);
-
-	{
-		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
-
-		if (limit >= 52)
-			guest_mode_init(VM_MODE_P52V48_64K, true, true);
-		if (limit >= 48) {
-			guest_mode_init(VM_MODE_P48V48_4K, true, true);
-			guest_mode_init(VM_MODE_P48V48_64K, true, true);
-		}
-	}
-#endif
-#ifdef __s390x__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-#endif
 
 	while ((opt = getopt(argc, argv, "c:hi:I:p:m:M:")) != -1) {
 		switch (opt) {
@@ -865,24 +830,16 @@ int main(int argc, char *argv[])
 			test_dirty_ring_count = strtol(optarg, NULL, 10);
 			break;
 		case 'i':
-			iterations = strtol(optarg, NULL, 10);
+			p.iterations = strtol(optarg, NULL, 10);
 			break;
 		case 'I':
-			interval = strtol(optarg, NULL, 10);
+			p.interval = strtol(optarg, NULL, 10);
 			break;
 		case 'p':
-			phys_offset = strtoull(optarg, NULL, 0);
+			p.phys_offset = strtoull(optarg, NULL, 0);
 			break;
 		case 'm':
-			if (!mode_selected) {
+			guest_modes_cmdline(optarg);
-				for (i = 0; i < NUM_VM_MODES; ++i)
-					guest_modes[i].enabled = false;
-				mode_selected = true;
-			}
-			mode = strtoul(optarg, NULL, 10);
-			TEST_ASSERT(mode < NUM_VM_MODES,
-				    "Guest mode ID %d too big", mode);
-			guest_modes[mode].enabled = true;
 			break;
 		case 'M':
 			if (!strcmp(optarg, "all")) {
@@ -911,32 +868,24 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	TEST_ASSERT(iterations > 2, "Iterations must be greater than two");
+	TEST_ASSERT(p.iterations > 2, "Iterations must be greater than two");
-	TEST_ASSERT(interval > 0, "Interval must be greater than zero");
+	TEST_ASSERT(p.interval > 0, "Interval must be greater than zero");
 
 	pr_info("Test iterations: %"PRIu64", interval: %"PRIu64" (ms)\n",
-		iterations, interval);
+		p.iterations, p.interval);
 
 	srandom(time(0));
 
-	for (i = 0; i < NUM_VM_MODES; ++i) {
+	if (host_log_mode_option == LOG_MODE_ALL) {
-		if (!guest_modes[i].enabled)
+		/* Run each log mode */
-			continue;
+		for (i = 0; i < LOG_MODE_NUM; i++) {
-		TEST_ASSERT(guest_modes[i].supported,
+			pr_info("Testing Log Mode '%s'\n", log_modes[i].name);
-			    "Guest mode ID %d (%s) not supported.",
+			host_log_mode = i;
-			    i, vm_guest_mode_string(i));
+			for_each_guest_mode(run_test, &p);
-		if (host_log_mode_option == LOG_MODE_ALL) {
-			/* Run each log mode */
-			for (j = 0; j < LOG_MODE_NUM; j++) {
-				pr_info("Testing Log Mode '%s'\n",
-					log_modes[j].name);
-				host_log_mode = j;
-				run_test(i, iterations, interval, phys_offset);
-			}
-		} else {
-			host_log_mode = host_log_mode_option;
-			run_test(i, iterations, interval, phys_offset);
 		}
+	} else {
+		host_log_mode = host_log_mode_option;
+		for_each_guest_mode(run_test, &p);
 	}
 
 	return 0;

tools/testing/selftests/kvm/include/guest_modes.h

@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2020, Red Hat, Inc.
+ */
+#include "kvm_util.h"
+
+struct guest_mode {
+	bool supported;
+	bool enabled;
+};
+
+extern struct guest_mode guest_modes[NUM_VM_MODES];
+
+#define guest_mode_append(mode, supported, enabled) ({ \
+	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
+})
+
+void guest_modes_append_default(void);
+void for_each_guest_mode(void (*func)(enum vm_guest_mode, void *), void *arg);
+void guest_modes_help(void);
+void guest_modes_cmdline(const char *arg);

tools/testing/selftests/kvm/include/kvm_util.h

@@ -16,6 +16,7 @@
 
 #include "sparsebit.h"
 
+#define KVM_MAX_VCPUS 512
 
 /*
  * Callers of kvm_util only have an incomplete/opaque description of the
@@ -70,6 +71,14 @@ enum vm_guest_mode {
 #define vm_guest_mode_string(m) vm_guest_mode_string[m]
 extern const char * const vm_guest_mode_string[];
 
+struct vm_guest_mode_params {
+	unsigned int pa_bits;
+	unsigned int va_bits;
+	unsigned int page_size;
+	unsigned int page_shift;
+};
+extern const struct vm_guest_mode_params vm_guest_mode_params[];
+
 enum vm_mem_backing_src_type {
 	VM_MEM_SRC_ANONYMOUS,
 	VM_MEM_SRC_ANONYMOUS_THP,

tools/testing/selftests/kvm/include/perf_test_util.h

@@ -9,38 +9,15 @@
 #define SELFTEST_KVM_PERF_TEST_UTIL_H
 
 #include "kvm_util.h"
-#include "processor.h"
-
-#define MAX_VCPUS 512
-
-#define PAGE_SHIFT_4K  12
-#define PTES_PER_4K_PT 512
-
-#define TEST_MEM_SLOT_INDEX		1
 
 /* Default guest test virtual memory offset */
 #define DEFAULT_GUEST_TEST_MEM		0xc0000000
 
 #define DEFAULT_PER_VCPU_MEM_SIZE	(1 << 30) /* 1G */
 
-/*
+#define PERF_TEST_MEM_SLOT_INDEX	1
- * Guest physical memory offset of the testing memory slot.
- * This will be set to the topmost valid physical address minus
- * the test memory size.
- */
-static uint64_t guest_test_phys_mem;
 
-/*
+struct perf_test_vcpu_args {
- * Guest virtual memory offset of the testing memory slot.
- * Must not conflict with identity mapped test code.
- */
-static uint64_t guest_test_virt_mem = DEFAULT_GUEST_TEST_MEM;
-static uint64_t guest_percpu_mem_size = DEFAULT_PER_VCPU_MEM_SIZE;
-
-/* Number of VCPUs for the test */
-static int nr_vcpus = 1;
-
-struct vcpu_args {
 	uint64_t gva;
 	uint64_t pages;
 
@@ -54,141 +31,21 @@ struct perf_test_args {
 	uint64_t guest_page_size;
 	int wr_fract;
 
-	struct vcpu_args vcpu_args[MAX_VCPUS];
+	struct perf_test_vcpu_args vcpu_args[KVM_MAX_VCPUS];
 };
 
-static struct perf_test_args perf_test_args;
+extern struct perf_test_args perf_test_args;
 
 /*
- * Continuously write to the first 8 bytes of each page in the
+ * Guest physical memory offset of the testing memory slot.
- * specified region.
+ * This will be set to the topmost valid physical address minus
+ * the test memory size.
  */
-static void guest_code(uint32_t vcpu_id)
+extern uint64_t guest_test_phys_mem;
-{
-	struct vcpu_args *vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
-	uint64_t gva;
-	uint64_t pages;
-	int i;
 
-	/* Make sure vCPU args data structure is not corrupt. */
+struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
-	GUEST_ASSERT(vcpu_args->vcpu_id == vcpu_id);
+				uint64_t vcpu_memory_bytes);
-
+void perf_test_destroy_vm(struct kvm_vm *vm);
-	gva = vcpu_args->gva;
+void perf_test_setup_vcpus(struct kvm_vm *vm, int vcpus, uint64_t vcpu_memory_bytes);
-	pages = vcpu_args->pages;
-
-	while (true) {
-		for (i = 0; i < pages; i++) {
-			uint64_t addr = gva + (i * perf_test_args.guest_page_size);
-
-			if (i % perf_test_args.wr_fract == 0)
-				*(uint64_t *)addr = 0x0123456789ABCDEF;
-			else
-				READ_ONCE(*(uint64_t *)addr);
-		}
-
-		GUEST_SYNC(1);
-	}
-}
-
-static struct kvm_vm *create_vm(enum vm_guest_mode mode, int vcpus,
-				uint64_t vcpu_memory_bytes)
-{
-	struct kvm_vm *vm;
-	uint64_t pages = DEFAULT_GUEST_PHY_PAGES;
-	uint64_t guest_num_pages;
-
-	/* Account for a few pages per-vCPU for stacks */
-	pages += DEFAULT_STACK_PGS * vcpus;
-
-	/*
-	 * Reserve twice the ammount of memory needed to map the test region and
-	 * the page table / stacks region, at 4k, for page tables. Do the
-	 * calculation with 4K page size: the smallest of all archs. (e.g., 64K
-	 * page size guest will need even less memory for page tables).
-	 */
-	pages += (2 * pages) / PTES_PER_4K_PT;
-	pages += ((2 * vcpus * vcpu_memory_bytes) >> PAGE_SHIFT_4K) /
-		 PTES_PER_4K_PT;
-	pages = vm_adjust_num_guest_pages(mode, pages);
-
-	pr_info("Testing guest mode: %s\n", vm_guest_mode_string(mode));
-
-	vm = vm_create(mode, pages, O_RDWR);
-	kvm_vm_elf_load(vm, program_invocation_name, 0, 0);
-#ifdef __x86_64__
-	vm_create_irqchip(vm);
-#endif
-
-	perf_test_args.vm = vm;
-	perf_test_args.guest_page_size = vm_get_page_size(vm);
-	perf_test_args.host_page_size = getpagesize();
-
-	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.guest_page_size == 0,
-		    "Guest memory size is not guest page size aligned.");
-
-	guest_num_pages = (vcpus * vcpu_memory_bytes) /
-			  perf_test_args.guest_page_size;
-	guest_num_pages = vm_adjust_num_guest_pages(mode, guest_num_pages);
-
-	/*
-	 * If there should be more memory in the guest test region than there
-	 * can be pages in the guest, it will definitely cause problems.
-	 */
-	TEST_ASSERT(guest_num_pages < vm_get_max_gfn(vm),
-		    "Requested more guest memory than address space allows.\n"
-		    "    guest pages: %lx max gfn: %x vcpus: %d wss: %lx]\n",
-		    guest_num_pages, vm_get_max_gfn(vm), vcpus,
-		    vcpu_memory_bytes);
-
-	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.host_page_size == 0,
-		    "Guest memory size is not host page size aligned.");
-
-	guest_test_phys_mem = (vm_get_max_gfn(vm) - guest_num_pages) *
-			      perf_test_args.guest_page_size;
-	guest_test_phys_mem &= ~(perf_test_args.host_page_size - 1);
-
-#ifdef __s390x__
-	/* Align to 1M (segment size) */
-	guest_test_phys_mem &= ~((1 << 20) - 1);
-#endif
-
-	pr_info("guest physical test memory offset: 0x%lx\n", guest_test_phys_mem);
-
-	/* Add an extra memory slot for testing */
-	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
-				    guest_test_phys_mem,
-				    TEST_MEM_SLOT_INDEX,
-				    guest_num_pages, 0);
-
-	/* Do mapping for the demand paging memory slot */
-	virt_map(vm, guest_test_virt_mem, guest_test_phys_mem, guest_num_pages, 0);
-
-	ucall_init(vm, NULL);
-
-	return vm;
-}
-
-static void add_vcpus(struct kvm_vm *vm, int vcpus, uint64_t vcpu_memory_bytes)
-{
-	vm_paddr_t vcpu_gpa;
-	struct vcpu_args *vcpu_args;
-	int vcpu_id;
-
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
-		vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
-
-		vm_vcpu_add_default(vm, vcpu_id, guest_code);
-
-		vcpu_args->vcpu_id = vcpu_id;
-		vcpu_args->gva = guest_test_virt_mem +
-				 (vcpu_id * vcpu_memory_bytes);
-		vcpu_args->pages = vcpu_memory_bytes /
-				   perf_test_args.guest_page_size;
-
-		vcpu_gpa = guest_test_phys_mem + (vcpu_id * vcpu_memory_bytes);
-		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
-			 vcpu_id, vcpu_gpa, vcpu_gpa + vcpu_memory_bytes);
-	}
-}
 
 #endif /* SELFTEST_KVM_PERF_TEST_UTIL_H */

tools/testing/selftests/kvm/lib/guest_modes.c

@@ -0,0 +1,70 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020, Red Hat, Inc.
+ */
+#include "guest_modes.h"
+
+struct guest_mode guest_modes[NUM_VM_MODES];
+
+void guest_modes_append_default(void)
+{
+	guest_mode_append(VM_MODE_DEFAULT, true, true);
+
+#ifdef __aarch64__
+	guest_mode_append(VM_MODE_P40V48_64K, true, true);
+	{
+		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
+		if (limit >= 52)
+			guest_mode_append(VM_MODE_P52V48_64K, true, true);
+		if (limit >= 48) {
+			guest_mode_append(VM_MODE_P48V48_4K, true, true);
+			guest_mode_append(VM_MODE_P48V48_64K, true, true);
+		}
+	}
+#endif
+}
+
+void for_each_guest_mode(void (*func)(enum vm_guest_mode, void *), void *arg)
+{
+	int i;
+
+	for (i = 0; i < NUM_VM_MODES; ++i) {
+		if (!guest_modes[i].enabled)
+			continue;
+		TEST_ASSERT(guest_modes[i].supported,
+			    "Guest mode ID %d (%s) not supported.",
+			    i, vm_guest_mode_string(i));
+		func(i, arg);
+	}
+}
+
+void guest_modes_help(void)
+{
+	int i;
+
+	printf(" -m: specify the guest mode ID to test\n"
+	       "     (default: test all supported modes)\n"
+	       "     This option may be used multiple times.\n"
+	       "     Guest mode IDs:\n");
+	for (i = 0; i < NUM_VM_MODES; ++i) {
+		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
+		       guest_modes[i].supported ? " (supported)" : "");
+	}
+}
+
+void guest_modes_cmdline(const char *arg)
+{
+	static bool mode_selected;
+	unsigned int mode;
+	int i;
+
+	if (!mode_selected) {
+		for (i = 0; i < NUM_VM_MODES; ++i)
+			guest_modes[i].enabled = false;
+		mode_selected = true;
+	}
+
+	mode = strtoul(optarg, NULL, 10);
+	TEST_ASSERT(mode < NUM_VM_MODES, "Guest mode ID %d too big", mode);
+	guest_modes[mode].enabled = true;
+}

tools/testing/selftests/kvm/lib/kvm_util.c

@@ -153,14 +153,7 @@ const char * const vm_guest_mode_string[] = {
 _Static_assert(sizeof(vm_guest_mode_string)/sizeof(char *) == NUM_VM_MODES,
 	       "Missing new mode strings?");
 
-struct vm_guest_mode_params {
+const struct vm_guest_mode_params vm_guest_mode_params[] = {
-	unsigned int pa_bits;
-	unsigned int va_bits;
-	unsigned int page_size;
-	unsigned int page_shift;
-};
-
-static const struct vm_guest_mode_params vm_guest_mode_params[] = {
 	{ 52, 48,  0x1000, 12 },
 	{ 52, 48, 0x10000, 16 },
 	{ 48, 48,  0x1000, 12 },

tools/testing/selftests/kvm/lib/perf_test_util.c

@@ -0,0 +1,134 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020, Google LLC.
+ */
+
+#include "kvm_util.h"
+#include "perf_test_util.h"
+#include "processor.h"
+
+struct perf_test_args perf_test_args;
+
+uint64_t guest_test_phys_mem;
+
+/*
+ * Guest virtual memory offset of the testing memory slot.
+ * Must not conflict with identity mapped test code.
+ */
+static uint64_t guest_test_virt_mem = DEFAULT_GUEST_TEST_MEM;
+
+/*
+ * Continuously write to the first 8 bytes of each page in the
+ * specified region.
+ */
+static void guest_code(uint32_t vcpu_id)
+{
+	struct perf_test_vcpu_args *vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+	uint64_t gva;
+	uint64_t pages;
+	int i;
+
+	/* Make sure vCPU args data structure is not corrupt. */
+	GUEST_ASSERT(vcpu_args->vcpu_id == vcpu_id);
+
+	gva = vcpu_args->gva;
+	pages = vcpu_args->pages;
+
+	while (true) {
+		for (i = 0; i < pages; i++) {
+			uint64_t addr = gva + (i * perf_test_args.guest_page_size);
+
+			if (i % perf_test_args.wr_fract == 0)
+				*(uint64_t *)addr = 0x0123456789ABCDEF;
+			else
+				READ_ONCE(*(uint64_t *)addr);
+		}
+
+		GUEST_SYNC(1);
+	}
+}
+
+struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
+				   uint64_t vcpu_memory_bytes)
+{
+	struct kvm_vm *vm;
+	uint64_t guest_num_pages;
+
+	pr_info("Testing guest mode: %s\n", vm_guest_mode_string(mode));
+
+	perf_test_args.host_page_size = getpagesize();
+	perf_test_args.guest_page_size = vm_guest_mode_params[mode].page_size;
+
+	guest_num_pages = vm_adjust_num_guest_pages(mode,
+				(vcpus * vcpu_memory_bytes) / perf_test_args.guest_page_size);
+
+	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.host_page_size == 0,
+		    "Guest memory size is not host page size aligned.");
+	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.guest_page_size == 0,
+		    "Guest memory size is not guest page size aligned.");
+
+	vm = vm_create_with_vcpus(mode, vcpus,
+				  (vcpus * vcpu_memory_bytes) / perf_test_args.guest_page_size,
+				  0, guest_code, NULL);
+
+	perf_test_args.vm = vm;
+
+	/*
+	 * If there should be more memory in the guest test region than there
+	 * can be pages in the guest, it will definitely cause problems.
+	 */
+	TEST_ASSERT(guest_num_pages < vm_get_max_gfn(vm),
+		    "Requested more guest memory than address space allows.\n"
+		    "    guest pages: %lx max gfn: %x vcpus: %d wss: %lx]\n",
+		    guest_num_pages, vm_get_max_gfn(vm), vcpus,
+		    vcpu_memory_bytes);
+
+	guest_test_phys_mem = (vm_get_max_gfn(vm) - guest_num_pages) *
+			      perf_test_args.guest_page_size;
+	guest_test_phys_mem &= ~(perf_test_args.host_page_size - 1);
+#ifdef __s390x__
+	/* Align to 1M (segment size) */
+	guest_test_phys_mem &= ~((1 << 20) - 1);
+#endif
+	pr_info("guest physical test memory offset: 0x%lx\n", guest_test_phys_mem);
+
+	/* Add an extra memory slot for testing */
+	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
+				    guest_test_phys_mem,
+				    PERF_TEST_MEM_SLOT_INDEX,
+				    guest_num_pages, 0);
+
+	/* Do mapping for the demand paging memory slot */
+	virt_map(vm, guest_test_virt_mem, guest_test_phys_mem, guest_num_pages, 0);
+
+	ucall_init(vm, NULL);
+
+	return vm;
+}
+
+void perf_test_destroy_vm(struct kvm_vm *vm)
+{
+	ucall_uninit(vm);
+	kvm_vm_free(vm);
+}
+
+void perf_test_setup_vcpus(struct kvm_vm *vm, int vcpus, uint64_t vcpu_memory_bytes)
+{
+	vm_paddr_t vcpu_gpa;
+	struct perf_test_vcpu_args *vcpu_args;
+	int vcpu_id;
+
+	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
+		vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+
+		vcpu_args->vcpu_id = vcpu_id;
+		vcpu_args->gva = guest_test_virt_mem +
+				 (vcpu_id * vcpu_memory_bytes);
+		vcpu_args->pages = vcpu_memory_bytes /
+				   perf_test_args.guest_page_size;
+
+		vcpu_gpa = guest_test_phys_mem + (vcpu_id * vcpu_memory_bytes);
+		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
+			 vcpu_id, vcpu_gpa, vcpu_gpa + vcpu_memory_bytes);
+	}
+}

virt/kvm/kvm_main.c

@@ -485,9 +485,8 @@ static int kvm_mmu_notifier_invalidate_range_start(struct mmu_notifier *mn,
 	kvm->mmu_notifier_count++;
 	need_tlb_flush = kvm_unmap_hva_range(kvm, range->start, range->end,
 					     range->flags);
-	need_tlb_flush |= kvm->tlbs_dirty;
 	/* we've to flush the tlb before the pages can be freed */
-	if (need_tlb_flush)
+	if (need_tlb_flush || kvm->tlbs_dirty)
 		kvm_flush_remote_tlbs(kvm);
 
 	spin_unlock(&kvm->mmu_lock);