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

Pull second set of KVM updates from Paolo Bonzini:
 "ARM:
   - Support for Group0 interrupts in guests
   - Cache management optimizations for ARMv8.4 systems
   - Userspace interface for RAS
   - Fault path optimization
   - Emulated physical timer fixes
   - Random cleanups

  x86:
   - fixes for L1TF
   - a new test case
   - non-support for SGX (inject the right exception in the guest)
   - fix lockdep false positive"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (49 commits)
  KVM: VMX: fixes for vmentry_l1d_flush module parameter
  kvm: selftest: add dirty logging test
  kvm: selftest: pass in extra memory when create vm
  kvm: selftest: include the tools headers
  kvm: selftest: unify the guest port macros
  tools: introduce test_and_clear_bit
  KVM: x86: SVM: Call x86_spec_ctrl_set_guest/host() with interrupts disabled
  KVM: vmx: Inject #UD for SGX ENCLS instruction in guest
  KVM: vmx: Add defines for SGX ENCLS exiting
  x86/kvm/vmx: Fix coding style in vmx_setup_l1d_flush()
  x86: kvm: avoid unused variable warning
  KVM: Documentation: rename the capability of KVM_CAP_ARM_SET_SERROR_ESR
  KVM: arm/arm64: Skip updating PTE entry if no change
  KVM: arm/arm64: Skip updating PMD entry if no change
  KVM: arm: Use true and false for boolean values
  KVM: arm/arm64: vgic: Do not use spin_lock_irqsave/restore with irq disabled
  KVM: arm/arm64: vgic: Move DEBUG_SPINLOCK_BUG_ON to vgic.h
  KVM: arm: vgic-v3: Add support for ICC_SGI0R and ICC_ASGI1R accesses
  KVM: arm64: vgic-v3: Add support for ICC_SGI0R_EL1 and ICC_ASGI1R_EL1 accesses
  KVM: arm/arm64: vgic-v3: Add core support for Group0 SGIs
  ...

Documentation/virtual/kvm/api.txt

@@ -835,11 +835,13 @@ struct kvm_clock_data {
 
 Capability: KVM_CAP_VCPU_EVENTS
 Extended by: KVM_CAP_INTR_SHADOW
-Architectures: x86
-Type: vm ioctl
+Architectures: x86, arm, arm64
+Type: vcpu ioctl
 Parameters: struct kvm_vcpu_event (out)
 Returns: 0 on success, -1 on error
 
+X86:
+
 Gets currently pending exceptions, interrupts, and NMIs as well as related
 states of the vcpu.
 
@@ -881,15 +883,64 @@ Only two fields are defined in the flags field:
 - KVM_VCPUEVENT_VALID_SMM may be set in the flags field to signal that
   smi contains a valid state.
 
+ARM/ARM64:
+
+If the guest accesses a device that is being emulated by the host kernel in
+such a way that a real device would generate a physical SError, KVM may make
+a virtual SError pending for that VCPU. This system error interrupt remains
+pending until the guest takes the exception by unmasking PSTATE.A.
+
+Running the VCPU may cause it to take a pending SError, or make an access that
+causes an SError to become pending. The event's description is only valid while
+the VPCU is not running.
+
+This API provides a way to read and write the pending 'event' state that is not
+visible to the guest. To save, restore or migrate a VCPU the struct representing
+the state can be read then written using this GET/SET API, along with the other
+guest-visible registers. It is not possible to 'cancel' an SError that has been
+made pending.
+
+A device being emulated in user-space may also wish to generate an SError. To do
+this the events structure can be populated by user-space. The current state
+should be read first, to ensure no existing SError is pending. If an existing
+SError is pending, the architecture's 'Multiple SError interrupts' rules should
+be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and
+Serviceability (RAS) Specification").
+
+SError exceptions always have an ESR value. Some CPUs have the ability to
+specify what the virtual SError's ESR value should be. These systems will
+advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will
+always have a non-zero value when read, and the agent making an SError pending
+should specify the ISS field in the lower 24 bits of exception.serror_esr. If
+the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events
+with exception.has_esr as zero, KVM will choose an ESR.
+
+Specifying exception.has_esr on a system that does not support it will return
+-EINVAL. Setting anything other than the lower 24bits of exception.serror_esr
+will return -EINVAL.
+
+struct kvm_vcpu_events {
+	struct {
+		__u8 serror_pending;
+		__u8 serror_has_esr;
+		/* Align it to 8 bytes */
+		__u8 pad[6];
+		__u64 serror_esr;
+	} exception;
+	__u32 reserved[12];
+};
+
 4.32 KVM_SET_VCPU_EVENTS
 
 Capability: KVM_CAP_VCPU_EVENTS
 Extended by: KVM_CAP_INTR_SHADOW
-Architectures: x86
-Type: vm ioctl
+Architectures: x86, arm, arm64
+Type: vcpu ioctl
 Parameters: struct kvm_vcpu_event (in)
 Returns: 0 on success, -1 on error
 
+X86:
+
 Set pending exceptions, interrupts, and NMIs as well as related states of the
 vcpu.
 
@@ -910,6 +961,13 @@ shall be written into the VCPU.
 
 KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available.
 
+ARM/ARM64:
+
+Set the pending SError exception state for this VCPU. It is not possible to
+'cancel' an Serror that has been made pending.
+
+See KVM_GET_VCPU_EVENTS for the data structure.
+
 
 4.33 KVM_GET_DEBUGREGS
 
@@ -4690,3 +4748,17 @@ This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush
 hypercalls:
 HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx,
 HvFlushVirtualAddressList, HvFlushVirtualAddressListEx.
+
+8.19 KVM_CAP_ARM_INJECT_SERROR_ESR
+
+Architectures: arm, arm64
+
+This capability indicates that userspace can specify (via the
+KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it
+takes a virtual SError interrupt exception.
+If KVM advertises this capability, userspace can only specify the ISS field for
+the ESR syndrome. Other parts of the ESR, such as the EC are generated by the
+CPU when the exception is taken. If this virtual SError is taken to EL1 using
+AArch64, this value will be reported in the ISS field of ESR_ELx.
+
+See KVM_CAP_VCPU_EVENTS for more details.

Documentation/virtual/kvm/devices/arm-vgic-v3.txt

@@ -100,6 +100,14 @@ Groups:
     Note that distributor fields are not banked, but return the same value
     regardless of the mpidr used to access the register.
 
+    GICD_IIDR.Revision is updated when the KVM implementation is changed in a
+    way directly observable by the guest or userspace.  Userspace should read
+    GICD_IIDR from KVM and write back the read value to confirm its expected
+    behavior is aligned with the KVM implementation.  Userspace should set
+    GICD_IIDR before setting any other registers to ensure the expected
+    behavior.
+
+
     The GICD_STATUSR and GICR_STATUSR registers are architecturally defined such
     that a write of a clear bit has no effect, whereas a write with a set bit
     clears that value.  To allow userspace to freely set the values of these two

Documentation/virtual/kvm/devices/arm-vgic.txt

@@ -49,9 +49,15 @@ Groups:
     index is specified with the vcpu_index field.  Note that most distributor
     fields are not banked, but return the same value regardless of the
     vcpu_index used to access the register.
-  Limitations:
-    - Priorities are not implemented, and registers are RAZ/WI
-    - Currently only implemented for KVM_DEV_TYPE_ARM_VGIC_V2.
+
+    GICD_IIDR.Revision is updated when the KVM implementation of an emulated
+    GICv2 is changed in a way directly observable by the guest or userspace.
+    Userspace should read GICD_IIDR from KVM and write back the read value to
+    confirm its expected behavior is aligned with the KVM implementation.
+    Userspace should set GICD_IIDR before setting any other registers (both
+    KVM_DEV_ARM_VGIC_GRP_DIST_REGS and KVM_DEV_ARM_VGIC_GRP_CPU_REGS) to ensure
+    the expected behavior. Unless GICD_IIDR has been set from userspace, writes
+    to the interrupt group registers (GICD_IGROUPR) are ignored.
   Errors:
     -ENXIO: Getting or setting this register is not yet supported
     -EBUSY: One or more VCPUs are running
@@ -94,9 +100,6 @@ Groups:
     use the lower 5 bits to communicate with the KVM device and must shift the
     value left by 3 places to obtain the actual priority mask level.
 
-  Limitations:
-    - Priorities are not implemented, and registers are RAZ/WI
-    - Currently only implemented for KVM_DEV_TYPE_ARM_VGIC_V2.
   Errors:
     -ENXIO: Getting or setting this register is not yet supported
     -EBUSY: One or more VCPUs are running

arch/arm/include/asm/kvm_emulate.h

@@ -107,9 +107,19 @@ static inline unsigned long *vcpu_hcr(const struct kvm_vcpu *vcpu)
 	return (unsigned long *)&vcpu->arch.hcr;
 }
 
+static inline void vcpu_clear_wfe_traps(struct kvm_vcpu *vcpu)
+{
+	vcpu->arch.hcr &= ~HCR_TWE;
+}
+
+static inline void vcpu_set_wfe_traps(struct kvm_vcpu *vcpu)
+{
+	vcpu->arch.hcr |= HCR_TWE;
+}
+
 static inline bool vcpu_mode_is_32bit(const struct kvm_vcpu *vcpu)
 {
-	return 1;
+	return true;
 }
 
 static inline unsigned long *vcpu_pc(struct kvm_vcpu *vcpu)

arch/arm/include/asm/kvm_host.h

@@ -216,6 +216,11 @@ int kvm_arm_get_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg);
 int kvm_arm_set_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg);
 unsigned long kvm_call_hyp(void *hypfn, ...);
 void force_vm_exit(const cpumask_t *mask);
+int __kvm_arm_vcpu_get_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events);
+
+int __kvm_arm_vcpu_set_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events);
 
 #define KVM_ARCH_WANT_MMU_NOTIFIER
 int kvm_unmap_hva(struct kvm *kvm, unsigned long hva);

arch/arm/include/asm/kvm_mmu.h

@@ -75,17 +75,9 @@ phys_addr_t kvm_get_idmap_vector(void);
 int kvm_mmu_init(void);
 void kvm_clear_hyp_idmap(void);
 
-static inline void kvm_set_pmd(pmd_t *pmd, pmd_t new_pmd)
-{
-	*pmd = new_pmd;
-	dsb(ishst);
-}
-
-static inline void kvm_set_pte(pte_t *pte, pte_t new_pte)
-{
-	*pte = new_pte;
-	dsb(ishst);
-}
+#define kvm_mk_pmd(ptep)	__pmd(__pa(ptep) | PMD_TYPE_TABLE)
+#define kvm_mk_pud(pmdp)	__pud(__pa(pmdp) | PMD_TYPE_TABLE)
+#define kvm_mk_pgd(pudp)	({ BUILD_BUG(); 0; })
 
 static inline pte_t kvm_s2pte_mkwrite(pte_t pte)
 {

arch/arm/include/uapi/asm/kvm.h

@@ -27,6 +27,7 @@
 #define __KVM_HAVE_GUEST_DEBUG
 #define __KVM_HAVE_IRQ_LINE
 #define __KVM_HAVE_READONLY_MEM
+#define __KVM_HAVE_VCPU_EVENTS
 
 #define KVM_COALESCED_MMIO_PAGE_OFFSET 1
 
@@ -125,6 +126,18 @@ struct kvm_sync_regs {
 struct kvm_arch_memory_slot {
 };
 
+/* for KVM_GET/SET_VCPU_EVENTS */
+struct kvm_vcpu_events {
+	struct {
+		__u8 serror_pending;
+		__u8 serror_has_esr;
+		/* Align it to 8 bytes */
+		__u8 pad[6];
+		__u64 serror_esr;
+	} exception;
+	__u32 reserved[12];
+};
+
 /* If you need to interpret the index values, here is the key: */
 #define KVM_REG_ARM_COPROC_MASK		0x000000000FFF0000
 #define KVM_REG_ARM_COPROC_SHIFT	16

arch/arm/kvm/coproc.c

@@ -246,6 +246,7 @@ static bool access_gic_sgi(struct kvm_vcpu *vcpu,
 			   const struct coproc_reg *r)
 {
 	u64 reg;
+	bool g1;
 
 	if (!p->is_write)
 		return read_from_write_only(vcpu, p);
@@ -253,7 +254,25 @@ static bool access_gic_sgi(struct kvm_vcpu *vcpu,
 	reg = (u64)*vcpu_reg(vcpu, p->Rt2) << 32;
 	reg |= *vcpu_reg(vcpu, p->Rt1) ;
 
-	vgic_v3_dispatch_sgi(vcpu, reg);
+	/*
+	 * In a system where GICD_CTLR.DS=1, a ICC_SGI0R access generates
+	 * Group0 SGIs only, while ICC_SGI1R can generate either group,
+	 * depending on the SGI configuration. ICC_ASGI1R is effectively
+	 * equivalent to ICC_SGI0R, as there is no "alternative" secure
+	 * group.
+	 */
+	switch (p->Op1) {
+	default:		/* Keep GCC quiet */
+	case 0:			/* ICC_SGI1R */
+		g1 = true;
+		break;
+	case 1:			/* ICC_ASGI1R */
+	case 2:			/* ICC_SGI0R */
+		g1 = false;
+		break;
+	}
+
+	vgic_v3_dispatch_sgi(vcpu, reg, g1);
 
 	return true;
 }
@@ -459,6 +478,10 @@ static const struct coproc_reg cp15_regs[] = {
 
 	/* ICC_SGI1R */
 	{ CRm64(12), Op1( 0), is64, access_gic_sgi},
+	/* ICC_ASGI1R */
+	{ CRm64(12), Op1( 1), is64, access_gic_sgi},
+	/* ICC_SGI0R */
+	{ CRm64(12), Op1( 2), is64, access_gic_sgi},
 
 	/* VBAR: swapped by interrupt.S. */
 	{ CRn(12), CRm( 0), Op1( 0), Op2( 0), is32,

arch/arm/kvm/guest.c

@@ -261,6 +261,29 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	return -EINVAL;
 }
 
+
+int __kvm_arm_vcpu_get_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events)
+{
+	events->exception.serror_pending = !!(*vcpu_hcr(vcpu) & HCR_VA);
+
+	return 0;
+}
+
+int __kvm_arm_vcpu_set_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events)
+{
+	bool serror_pending = events->exception.serror_pending;
+	bool has_esr = events->exception.serror_has_esr;
+
+	if (serror_pending && has_esr)
+		return -EINVAL;
+	else if (serror_pending)
+		kvm_inject_vabt(vcpu);
+
+	return 0;
+}
+
 int __attribute_const__ kvm_target_cpu(void)
 {
 	switch (read_cpuid_part()) {

arch/arm64/include/asm/cpucaps.h

@@ -50,7 +50,8 @@
 #define ARM64_HW_DBM				29
 #define ARM64_SSBD				30
 #define ARM64_MISMATCHED_CACHE_TYPE		31
+#define ARM64_HAS_STAGE2_FWB			32
 
-#define ARM64_NCAPS				32
+#define ARM64_NCAPS				33
 
 #endif /* __ASM_CPUCAPS_H */

arch/arm64/include/asm/kvm_arm.h

@@ -23,6 +23,7 @@
 #include <asm/types.h>
 
 /* Hyp Configuration Register (HCR) bits */
+#define HCR_FWB		(UL(1) << 46)
 #define HCR_TEA		(UL(1) << 37)
 #define HCR_TERR	(UL(1) << 36)
 #define HCR_TLOR	(UL(1) << 35)

arch/arm64/include/asm/kvm_emulate.h

@@ -63,6 +63,8 @@ static inline void vcpu_reset_hcr(struct kvm_vcpu *vcpu)
 		/* trap error record accesses */
 		vcpu->arch.hcr_el2 |= HCR_TERR;
 	}
+	if (cpus_have_const_cap(ARM64_HAS_STAGE2_FWB))
+		vcpu->arch.hcr_el2 |= HCR_FWB;
 
 	if (test_bit(KVM_ARM_VCPU_EL1_32BIT, vcpu->arch.features))
 		vcpu->arch.hcr_el2 &= ~HCR_RW;
@@ -81,6 +83,21 @@ static inline unsigned long *vcpu_hcr(struct kvm_vcpu *vcpu)
 	return (unsigned long *)&vcpu->arch.hcr_el2;
 }
 
+static inline void vcpu_clear_wfe_traps(struct kvm_vcpu *vcpu)
+{
+	vcpu->arch.hcr_el2 &= ~HCR_TWE;
+}
+
+static inline void vcpu_set_wfe_traps(struct kvm_vcpu *vcpu)
+{
+	vcpu->arch.hcr_el2 |= HCR_TWE;
+}
+
+static inline unsigned long vcpu_get_vsesr(struct kvm_vcpu *vcpu)
+{
+	return vcpu->arch.vsesr_el2;
+}
+
 static inline void vcpu_set_vsesr(struct kvm_vcpu *vcpu, u64 vsesr)
 {
 	vcpu->arch.vsesr_el2 = vsesr;

arch/arm64/include/asm/kvm_host.h

@@ -350,6 +350,11 @@ unsigned long kvm_arm_num_regs(struct kvm_vcpu *vcpu);
 int kvm_arm_copy_reg_indices(struct kvm_vcpu *vcpu, u64 __user *indices);
 int kvm_arm_get_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg);
 int kvm_arm_set_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg);
+int __kvm_arm_vcpu_get_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events);
+
+int __kvm_arm_vcpu_set_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events);
 
 #define KVM_ARCH_WANT_MMU_NOTIFIER
 int kvm_unmap_hva(struct kvm *kvm, unsigned long hva);
@@ -378,16 +383,23 @@ void handle_exit_early(struct kvm_vcpu *vcpu, struct kvm_run *run,
 int kvm_perf_init(void);
 int kvm_perf_teardown(void);
 
+void kvm_set_sei_esr(struct kvm_vcpu *vcpu, u64 syndrome);
+
 struct kvm_vcpu *kvm_mpidr_to_vcpu(struct kvm *kvm, unsigned long mpidr);
 
-void __kvm_set_tpidr_el2(u64 tpidr_el2);
 DECLARE_PER_CPU(kvm_cpu_context_t, kvm_host_cpu_state);
 
 static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
 				       unsigned long hyp_stack_ptr,
 				       unsigned long vector_ptr)
 {
-	u64 tpidr_el2;
+	/*
+	 * Calculate the raw per-cpu offset without a translation from the
+	 * kernel's mapping to the linear mapping, and store it in tpidr_el2
+	 * so that we can use adr_l to access per-cpu variables in EL2.
+	 */
+	u64 tpidr_el2 = ((u64)this_cpu_ptr(&kvm_host_cpu_state) -
+			 (u64)kvm_ksym_ref(kvm_host_cpu_state));
 
 	/*
 	 * Call initialization code, and switch to the full blown HYP code.
@@ -396,17 +408,7 @@ static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
 	 * cpus_have_const_cap() wrapper.
 	 */
 	BUG_ON(!static_branch_likely(&arm64_const_caps_ready));
-	__kvm_call_hyp((void *)pgd_ptr, hyp_stack_ptr, vector_ptr);
-
-	/*
-	 * Calculate the raw per-cpu offset without a translation from the
-	 * kernel's mapping to the linear mapping, and store it in tpidr_el2
-	 * so that we can use adr_l to access per-cpu variables in EL2.
-	 */
-	tpidr_el2 = (u64)this_cpu_ptr(&kvm_host_cpu_state)
-		- (u64)kvm_ksym_ref(kvm_host_cpu_state);
-
-	kvm_call_hyp(__kvm_set_tpidr_el2, tpidr_el2);
+	__kvm_call_hyp((void *)pgd_ptr, hyp_stack_ptr, vector_ptr, tpidr_el2);
 }
 
 static inline bool kvm_arch_check_sve_has_vhe(void)

arch/arm64/include/asm/kvm_mmu.h

@@ -169,8 +169,12 @@ phys_addr_t kvm_get_idmap_vector(void);
 int kvm_mmu_init(void);
 void kvm_clear_hyp_idmap(void);
 
-#define	kvm_set_pte(ptep, pte)		set_pte(ptep, pte)
-#define	kvm_set_pmd(pmdp, pmd)		set_pmd(pmdp, pmd)
+#define kvm_mk_pmd(ptep)					\
+	__pmd(__phys_to_pmd_val(__pa(ptep)) | PMD_TYPE_TABLE)
+#define kvm_mk_pud(pmdp)					\
+	__pud(__phys_to_pud_val(__pa(pmdp)) | PMD_TYPE_TABLE)
+#define kvm_mk_pgd(pudp)					\
+	__pgd(__phys_to_pgd_val(__pa(pudp)) | PUD_TYPE_TABLE)
 
 static inline pte_t kvm_s2pte_mkwrite(pte_t pte)
 {
@@ -267,6 +271,15 @@ static inline void __clean_dcache_guest_page(kvm_pfn_t pfn, unsigned long size)
 {
 	void *va = page_address(pfn_to_page(pfn));
 
+	/*
+	 * With FWB, we ensure that the guest always accesses memory using
+	 * cacheable attributes, and we don't have to clean to PoC when
+	 * faulting in pages. Furthermore, FWB implies IDC, so cleaning to
+	 * PoU is not required either in this case.
+	 */
+	if (cpus_have_const_cap(ARM64_HAS_STAGE2_FWB))
+		return;
+
 	kvm_flush_dcache_to_poc(va, size);
 }
 
@@ -287,20 +300,26 @@ static inline void __invalidate_icache_guest_page(kvm_pfn_t pfn,
 
 static inline void __kvm_flush_dcache_pte(pte_t pte)
 {
-	struct page *page = pte_page(pte);
-	kvm_flush_dcache_to_poc(page_address(page), PAGE_SIZE);
+	if (!cpus_have_const_cap(ARM64_HAS_STAGE2_FWB)) {
+		struct page *page = pte_page(pte);
+		kvm_flush_dcache_to_poc(page_address(page), PAGE_SIZE);
+	}
 }
 
 static inline void __kvm_flush_dcache_pmd(pmd_t pmd)
 {
-	struct page *page = pmd_page(pmd);
-	kvm_flush_dcache_to_poc(page_address(page), PMD_SIZE);
+	if (!cpus_have_const_cap(ARM64_HAS_STAGE2_FWB)) {
+		struct page *page = pmd_page(pmd);
+		kvm_flush_dcache_to_poc(page_address(page), PMD_SIZE);
+	}
 }
 
 static inline void __kvm_flush_dcache_pud(pud_t pud)
 {
-	struct page *page = pud_page(pud);
-	kvm_flush_dcache_to_poc(page_address(page), PUD_SIZE);
+	if (!cpus_have_const_cap(ARM64_HAS_STAGE2_FWB)) {
+		struct page *page = pud_page(pud);
+		kvm_flush_dcache_to_poc(page_address(page), PUD_SIZE);
+	}
 }
 
 #define kvm_virt_to_phys(x)		__pa_symbol(x)

arch/arm64/include/asm/memory.h

@@ -155,6 +155,13 @@
 #define MT_S2_NORMAL		0xf
 #define MT_S2_DEVICE_nGnRE	0x1
 
+/*
+ * Memory types for Stage-2 translation when ID_AA64MMFR2_EL1.FWB is 0001
+ * Stage-2 enforces Normal-WB and Device-nGnRE
+ */
+#define MT_S2_FWB_NORMAL	6
+#define MT_S2_FWB_DEVICE_nGnRE	1
+
 #ifdef CONFIG_ARM64_4K_PAGES
 #define IOREMAP_MAX_ORDER	(PUD_SHIFT)
 #else

arch/arm64/include/asm/pgtable-prot.h

@@ -67,8 +67,28 @@
 #define PAGE_HYP_RO		__pgprot(_HYP_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY | PTE_HYP_XN)
 #define PAGE_HYP_DEVICE		__pgprot(PROT_DEVICE_nGnRE | PTE_HYP)
 
-#define PAGE_S2			__pgprot(_PROT_DEFAULT | PTE_S2_MEMATTR(MT_S2_NORMAL) | PTE_S2_RDONLY | PTE_S2_XN)
-#define PAGE_S2_DEVICE		__pgprot(_PROT_DEFAULT | PTE_S2_MEMATTR(MT_S2_DEVICE_nGnRE) | PTE_S2_RDONLY | PTE_S2_XN)
+#define PAGE_S2_MEMATTR(attr)						\
+	({								\
+		u64 __val;						\
+		if (cpus_have_const_cap(ARM64_HAS_STAGE2_FWB))		\
+			__val = PTE_S2_MEMATTR(MT_S2_FWB_ ## attr);	\
+		else							\
+			__val = PTE_S2_MEMATTR(MT_S2_ ## attr);		\
+		__val;							\
+	 })
+
+#define PAGE_S2_XN							\
+	({								\
+		u64 __val;						\
+		if (cpus_have_const_cap(ARM64_HAS_CACHE_DIC))		\
+			__val = 0;					\
+		else							\
+			__val = PTE_S2_XN;				\
+		__val;							\
+	})
+
+#define PAGE_S2			__pgprot(_PROT_DEFAULT | PAGE_S2_MEMATTR(NORMAL) | PTE_S2_RDONLY | PAGE_S2_XN)
+#define PAGE_S2_DEVICE		__pgprot(_PROT_DEFAULT | PAGE_S2_MEMATTR(DEVICE_nGnRE) | PTE_S2_RDONLY | PAGE_S2_XN)
 
 #define PAGE_NONE		__pgprot(((_PAGE_DEFAULT) & ~PTE_VALID) | PTE_PROT_NONE | PTE_RDONLY | PTE_NG | PTE_PXN | PTE_UXN)
 #define PAGE_SHARED		__pgprot(_PAGE_DEFAULT | PTE_USER | PTE_NG | PTE_PXN | PTE_UXN | PTE_WRITE)

arch/arm64/include/asm/sysreg.h

@@ -314,6 +314,8 @@
 #define SYS_ICC_DIR_EL1			sys_reg(3, 0, 12, 11, 1)
 #define SYS_ICC_RPR_EL1			sys_reg(3, 0, 12, 11, 3)
 #define SYS_ICC_SGI1R_EL1		sys_reg(3, 0, 12, 11, 5)
+#define SYS_ICC_ASGI1R_EL1		sys_reg(3, 0, 12, 11, 6)
+#define SYS_ICC_SGI0R_EL1		sys_reg(3, 0, 12, 11, 7)
 #define SYS_ICC_IAR1_EL1		sys_reg(3, 0, 12, 12, 0)
 #define SYS_ICC_EOIR1_EL1		sys_reg(3, 0, 12, 12, 1)
 #define SYS_ICC_HPPIR1_EL1		sys_reg(3, 0, 12, 12, 2)
@@ -579,6 +581,7 @@
 #define ID_AA64MMFR1_VMIDBITS_16	2
 
 /* id_aa64mmfr2 */
+#define ID_AA64MMFR2_FWB_SHIFT		40
 #define ID_AA64MMFR2_AT_SHIFT		32
 #define ID_AA64MMFR2_LVA_SHIFT		16
 #define ID_AA64MMFR2_IESB_SHIFT		12

arch/arm64/include/uapi/asm/kvm.h

@@ -39,6 +39,7 @@
 #define __KVM_HAVE_GUEST_DEBUG
 #define __KVM_HAVE_IRQ_LINE
 #define __KVM_HAVE_READONLY_MEM
+#define __KVM_HAVE_VCPU_EVENTS
 
 #define KVM_COALESCED_MMIO_PAGE_OFFSET 1
 
@@ -154,6 +155,18 @@ struct kvm_sync_regs {
 struct kvm_arch_memory_slot {
 };
 
+/* for KVM_GET/SET_VCPU_EVENTS */
+struct kvm_vcpu_events {
+	struct {
+		__u8 serror_pending;
+		__u8 serror_has_esr;
+		/* Align it to 8 bytes */
+		__u8 pad[6];
+		__u64 serror_esr;
+	} exception;
+	__u32 reserved[12];
+};
+
 /* If you need to interpret the index values, here is the key: */
 #define KVM_REG_ARM_COPROC_MASK		0x000000000FFF0000
 #define KVM_REG_ARM_COPROC_SHIFT	16

arch/arm64/kernel/cpufeature.c

@@ -192,6 +192,7 @@ static const struct arm64_ftr_bits ftr_id_aa64mmfr1[] = {
 };
 
 static const struct arm64_ftr_bits ftr_id_aa64mmfr2[] = {
+	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR2_FWB_SHIFT, 4, 0),
 	ARM64_FTR_BITS(FTR_VISIBLE, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR2_AT_SHIFT, 4, 0),
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR2_LVA_SHIFT, 4, 0),
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR2_IESB_SHIFT, 4, 0),
@@ -1026,6 +1027,14 @@ static void cpu_copy_el2regs(const struct arm64_cpu_capabilities *__unused)
 }
 #endif
 
+static void cpu_has_fwb(const struct arm64_cpu_capabilities *__unused)
+{
+	u64 val = read_sysreg_s(SYS_CLIDR_EL1);
+
+	/* Check that CLIDR_EL1.LOU{U,IS} are both 0 */
+	WARN_ON(val & (7 << 27 | 7 << 21));
+}
+
 static const struct arm64_cpu_capabilities arm64_features[] = {
 	{
 		.desc = "GIC system register CPU interface",
@@ -1182,6 +1191,17 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
 		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
 		.matches = has_cache_dic,
 	},
+	{
+		.desc = "Stage-2 Force Write-Back",
+		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
+		.capability = ARM64_HAS_STAGE2_FWB,
+		.sys_reg = SYS_ID_AA64MMFR2_EL1,
+		.sign = FTR_UNSIGNED,
+		.field_pos = ID_AA64MMFR2_FWB_SHIFT,
+		.min_field_value = 1,
+		.matches = has_cpuid_feature,
+		.cpu_enable = cpu_has_fwb,
+	},
 #ifdef CONFIG_ARM64_HW_AFDBM
 	{
 		/*

arch/arm64/kvm/guest.c

@@ -289,6 +289,39 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	return -EINVAL;
 }
 
+int __kvm_arm_vcpu_get_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events)
+{
+	events->exception.serror_pending = !!(vcpu->arch.hcr_el2 & HCR_VSE);
+	events->exception.serror_has_esr = cpus_have_const_cap(ARM64_HAS_RAS_EXTN);
+
+	if (events->exception.serror_pending && events->exception.serror_has_esr)
+		events->exception.serror_esr = vcpu_get_vsesr(vcpu);
+
+	return 0;
+}
+
+int __kvm_arm_vcpu_set_events(struct kvm_vcpu *vcpu,
+			      struct kvm_vcpu_events *events)
+{
+	bool serror_pending = events->exception.serror_pending;
+	bool has_esr = events->exception.serror_has_esr;
+
+	if (serror_pending && has_esr) {
+		if (!cpus_have_const_cap(ARM64_HAS_RAS_EXTN))
+			return -EINVAL;
+
+		if (!((events->exception.serror_esr) & ~ESR_ELx_ISS_MASK))
+			kvm_set_sei_esr(vcpu, events->exception.serror_esr);
+		else
+			return -EINVAL;
+	} else if (serror_pending) {
+		kvm_inject_vabt(vcpu);
+	}
+
+	return 0;
+}
+
 int __attribute_const__ kvm_target_cpu(void)
 {
 	unsigned long implementor = read_cpuid_implementor();

arch/arm64/kvm/hyp-init.S

@@ -57,6 +57,7 @@ __invalid:
 	 * x0: HYP pgd
 	 * x1: HYP stack
 	 * x2: HYP vectors
+	 * x3: per-CPU offset
 	 */
 __do_hyp_init:
 	/* Check for a stub HVC call */
@@ -119,9 +120,8 @@ CPU_BE(	orr	x4, x4, #SCTLR_ELx_EE)
 	mov	sp, x1
 	msr	vbar_el2, x2
 
-	/* copy tpidr_el1 into tpidr_el2 for use by HYP */
-	mrs	x1, tpidr_el1
-	msr	tpidr_el2, x1
+	/* Set tpidr_el2 for use by HYP */
+	msr	tpidr_el2, x3
 
 	/* Hello, World! */
 	eret

arch/arm64/kvm/hyp/sysreg-sr.c

@@ -288,8 +288,3 @@ void kvm_vcpu_put_sysregs(struct kvm_vcpu *vcpu)
 
 	vcpu->arch.sysregs_loaded_on_cpu = false;
 }
-
-void __hyp_text __kvm_set_tpidr_el2(u64 tpidr_el2)
-{
-	asm("msr tpidr_el2, %0": : "r" (tpidr_el2));
-}

arch/arm64/kvm/inject_fault.c

@@ -164,9 +164,9 @@ void kvm_inject_undefined(struct kvm_vcpu *vcpu)
 		inject_undef64(vcpu);
 }
 
-static void pend_guest_serror(struct kvm_vcpu *vcpu, u64 esr)
+void kvm_set_sei_esr(struct kvm_vcpu *vcpu, u64 esr)
 {
-	vcpu_set_vsesr(vcpu, esr);
+	vcpu_set_vsesr(vcpu, esr & ESR_ELx_ISS_MASK);
 	*vcpu_hcr(vcpu) |= HCR_VSE;
 }
 
@@ -184,5 +184,5 @@ static void pend_guest_serror(struct kvm_vcpu *vcpu, u64 esr)
  */
 void kvm_inject_vabt(struct kvm_vcpu *vcpu)
 {
-	pend_guest_serror(vcpu, ESR_ELx_ISV);
+	kvm_set_sei_esr(vcpu, ESR_ELx_ISV);
 }

arch/arm64/kvm/reset.c

@@ -77,8 +77,12 @@ int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
 	case KVM_CAP_ARM_PMU_V3:
 		r = kvm_arm_support_pmu_v3();
 		break;
+	case KVM_CAP_ARM_INJECT_SERROR_ESR:
+		r = cpus_have_const_cap(ARM64_HAS_RAS_EXTN);
+		break;
 	case KVM_CAP_SET_GUEST_DEBUG:
 	case KVM_CAP_VCPU_ATTRIBUTES:
+	case KVM_CAP_VCPU_EVENTS:
 		r = 1;
 		break;
 	default:

arch/arm64/kvm/sys_regs.c

@@ -194,7 +194,16 @@ static bool access_dcsw(struct kvm_vcpu *vcpu,
 	if (!p->is_write)
 		return read_from_write_only(vcpu, p, r);
 
-	kvm_set_way_flush(vcpu);
+	/*
+	 * Only track S/W ops if we don't have FWB. It still indicates
+	 * that the guest is a bit broken (S/W operations should only
+	 * be done by firmware, knowing that there is only a single
+	 * CPU left in the system, and certainly not from non-secure
+	 * software).
+	 */
+	if (!cpus_have_const_cap(ARM64_HAS_STAGE2_FWB))
+		kvm_set_way_flush(vcpu);
+
 	return true;
 }
 
@@ -243,10 +252,43 @@ static bool access_gic_sgi(struct kvm_vcpu *vcpu,
 			   struct sys_reg_params *p,
 			   const struct sys_reg_desc *r)
 {
+	bool g1;
+
 	if (!p->is_write)
 		return read_from_write_only(vcpu, p, r);
 
-	vgic_v3_dispatch_sgi(vcpu, p->regval);
+	/*
+	 * In a system where GICD_CTLR.DS=1, a ICC_SGI0R_EL1 access generates
+	 * Group0 SGIs only, while ICC_SGI1R_EL1 can generate either group,
+	 * depending on the SGI configuration. ICC_ASGI1R_EL1 is effectively
+	 * equivalent to ICC_SGI0R_EL1, as there is no "alternative" secure
+	 * group.
+	 */
+	if (p->is_aarch32) {
+		switch (p->Op1) {
+		default:		/* Keep GCC quiet */
+		case 0:			/* ICC_SGI1R */
+			g1 = true;
+			break;
+		case 1:			/* ICC_ASGI1R */
+		case 2:			/* ICC_SGI0R */
+			g1 = false;
+			break;
+		}
+	} else {
+		switch (p->Op2) {
+		default:		/* Keep GCC quiet */
+		case 5:			/* ICC_SGI1R_EL1 */
+			g1 = true;
+			break;
+		case 6:			/* ICC_ASGI1R_EL1 */
+		case 7:			/* ICC_SGI0R_EL1 */
+			g1 = false;
+			break;
+		}
+	}
+
+	vgic_v3_dispatch_sgi(vcpu, p->regval, g1);
 
 	return true;
 }
@@ -1303,6 +1345,8 @@ static const struct sys_reg_desc sys_reg_descs[] = {
 	{ SYS_DESC(SYS_ICC_DIR_EL1), read_from_write_only },
 	{ SYS_DESC(SYS_ICC_RPR_EL1), write_to_read_only },
 	{ SYS_DESC(SYS_ICC_SGI1R_EL1), access_gic_sgi },
+	{ SYS_DESC(SYS_ICC_ASGI1R_EL1), access_gic_sgi },
+	{ SYS_DESC(SYS_ICC_SGI0R_EL1), access_gic_sgi },
 	{ SYS_DESC(SYS_ICC_IAR1_EL1), write_to_read_only },
 	{ SYS_DESC(SYS_ICC_EOIR1_EL1), read_from_write_only },
 	{ SYS_DESC(SYS_ICC_HPPIR1_EL1), write_to_read_only },
@@ -1613,8 +1657,6 @@ static const struct sys_reg_desc cp14_64_regs[] = {
  * register).
  */
 static const struct sys_reg_desc cp15_regs[] = {
-	{ Op1( 0), CRn( 0), CRm(12), Op2( 0), access_gic_sgi },
-
 	{ Op1( 0), CRn( 1), CRm( 0), Op2( 0), access_vm_reg, NULL, c1_SCTLR },
 	{ Op1( 0), CRn( 2), CRm( 0), Op2( 0), access_vm_reg, NULL, c2_TTBR0 },
 	{ Op1( 0), CRn( 2), CRm( 0), Op2( 1), access_vm_reg, NULL, c2_TTBR1 },
@@ -1737,8 +1779,10 @@ static const struct sys_reg_desc cp15_regs[] = {
 static const struct sys_reg_desc cp15_64_regs[] = {
 	{ Op1( 0), CRn( 0), CRm( 2), Op2( 0), access_vm_reg, NULL, c2_TTBR0 },
 	{ Op1( 0), CRn( 0), CRm( 9), Op2( 0), access_pmu_evcntr },
-	{ Op1( 0), CRn( 0), CRm(12), Op2( 0), access_gic_sgi },
+	{ Op1( 0), CRn( 0), CRm(12), Op2( 0), access_gic_sgi }, /* ICC_SGI1R */
 	{ Op1( 1), CRn( 0), CRm( 2), Op2( 0), access_vm_reg, NULL, c2_TTBR1 },
+	{ Op1( 1), CRn( 0), CRm(12), Op2( 0), access_gic_sgi }, /* ICC_ASGI1R */
+	{ Op1( 2), CRn( 0), CRm(12), Op2( 0), access_gic_sgi }, /* ICC_SGI0R */
 	{ Op1( 2), CRn( 0), CRm(14), Op2( 0), access_cntp_cval },
 };
 

arch/x86/include/asm/vmx.h

@@ -74,6 +74,7 @@
 #define SECONDARY_EXEC_ENABLE_INVPCID		0x00001000
 #define SECONDARY_EXEC_ENABLE_VMFUNC            0x00002000
 #define SECONDARY_EXEC_SHADOW_VMCS              0x00004000
+#define SECONDARY_EXEC_ENCLS_EXITING		0x00008000
 #define SECONDARY_EXEC_RDSEED_EXITING		0x00010000
 #define SECONDARY_EXEC_ENABLE_PML               0x00020000
 #define SECONDARY_EXEC_XSAVES			0x00100000
@@ -213,6 +214,8 @@ enum vmcs_field {
 	VMWRITE_BITMAP_HIGH             = 0x00002029,
 	XSS_EXIT_BITMAP                 = 0x0000202C,
 	XSS_EXIT_BITMAP_HIGH            = 0x0000202D,
+	ENCLS_EXITING_BITMAP		= 0x0000202E,
+	ENCLS_EXITING_BITMAP_HIGH	= 0x0000202F,
 	TSC_MULTIPLIER                  = 0x00002032,
 	TSC_MULTIPLIER_HIGH             = 0x00002033,
 	GUEST_PHYSICAL_ADDRESS          = 0x00002400,

arch/x86/kvm/svm.c

@@ -5586,8 +5586,6 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu)
 
 	clgi();
 
-	local_irq_enable();
-
 	/*
 	 * If this vCPU has touched SPEC_CTRL, restore the guest's value if
 	 * it's non-zero. Since vmentry is serialising on affected CPUs, there
@@ -5596,6 +5594,8 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu)
 	 */
 	x86_spec_ctrl_set_guest(svm->spec_ctrl, svm->virt_spec_ctrl);
 
+	local_irq_enable();
+
 	asm volatile (
 		"push %%" _ASM_BP "; \n\t"
 		"mov %c[rbx](%[svm]), %%" _ASM_BX " \n\t"
@@ -5718,12 +5718,12 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu)
 	if (unlikely(!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL)))
 		svm->spec_ctrl = native_read_msr(MSR_IA32_SPEC_CTRL);
 
-	x86_spec_ctrl_restore_host(svm->spec_ctrl, svm->virt_spec_ctrl);
-
 	reload_tss(vcpu);
 
 	local_irq_disable();
 
+	x86_spec_ctrl_restore_host(svm->spec_ctrl, svm->virt_spec_ctrl);
+
 	vcpu->arch.cr2 = svm->vmcb->save.cr2;
 	vcpu->arch.regs[VCPU_REGS_RAX] = svm->vmcb->save.rax;
 	vcpu->arch.regs[VCPU_REGS_RSP] = svm->vmcb->save.rsp;

arch/x86/kvm/vmx.c

@@ -198,12 +198,14 @@ static enum vmx_l1d_flush_state __read_mostly vmentry_l1d_flush_param = VMENTER_
 
 static const struct {
 	const char *option;
-	enum vmx_l1d_flush_state cmd;
+	bool for_parse;
 } vmentry_l1d_param[] = {
-	{"auto",	VMENTER_L1D_FLUSH_AUTO},
-	{"never",	VMENTER_L1D_FLUSH_NEVER},
-	{"cond",	VMENTER_L1D_FLUSH_COND},
-	{"always",	VMENTER_L1D_FLUSH_ALWAYS},
+	[VMENTER_L1D_FLUSH_AUTO]	 = {"auto", true},
+	[VMENTER_L1D_FLUSH_NEVER]	 = {"never", true},
+	[VMENTER_L1D_FLUSH_COND]	 = {"cond", true},
+	[VMENTER_L1D_FLUSH_ALWAYS]	 = {"always", true},
+	[VMENTER_L1D_FLUSH_EPT_DISABLED] = {"EPT disabled", false},
+	[VMENTER_L1D_FLUSH_NOT_REQUIRED] = {"not required", false},
 };
 
 #define L1D_CACHE_ORDER 4
@@ -219,15 +221,15 @@ static int vmx_setup_l1d_flush(enum vmx_l1d_flush_state l1tf)
 		return 0;
 	}
 
-       if (boot_cpu_has(X86_FEATURE_ARCH_CAPABILITIES)) {
-	       u64 msr;
+	if (boot_cpu_has(X86_FEATURE_ARCH_CAPABILITIES)) {
+		u64 msr;
 
-	       rdmsrl(MSR_IA32_ARCH_CAPABILITIES, msr);
-	       if (msr & ARCH_CAP_SKIP_VMENTRY_L1DFLUSH) {
-		       l1tf_vmx_mitigation = VMENTER_L1D_FLUSH_NOT_REQUIRED;
-		       return 0;
-	       }
-       }
+		rdmsrl(MSR_IA32_ARCH_CAPABILITIES, msr);
+		if (msr & ARCH_CAP_SKIP_VMENTRY_L1DFLUSH) {
+			l1tf_vmx_mitigation = VMENTER_L1D_FLUSH_NOT_REQUIRED;
+			return 0;
+		}
+	}
 
 	/* If set to auto use the default l1tf mitigation method */
 	if (l1tf == VMENTER_L1D_FLUSH_AUTO) {
@@ -287,8 +289,9 @@ static int vmentry_l1d_flush_parse(const char *s)
 
 	if (s) {
 		for (i = 0; i < ARRAY_SIZE(vmentry_l1d_param); i++) {
-			if (sysfs_streq(s, vmentry_l1d_param[i].option))
-				return vmentry_l1d_param[i].cmd;
+			if (vmentry_l1d_param[i].for_parse &&
+			    sysfs_streq(s, vmentry_l1d_param[i].option))
+				return i;
 		}
 	}
 	return -EINVAL;
@@ -298,13 +301,13 @@ static int vmentry_l1d_flush_set(const char *s, const struct kernel_param *kp)
 {
 	int l1tf, ret;
 
-	if (!boot_cpu_has(X86_BUG_L1TF))
-		return 0;
-
 	l1tf = vmentry_l1d_flush_parse(s);
 	if (l1tf < 0)
 		return l1tf;
 
+	if (!boot_cpu_has(X86_BUG_L1TF))
+		return 0;
+
 	/*
 	 * Has vmx_init() run already? If not then this is the pre init
 	 * parameter parsing. In that case just store the value and let
@@ -324,6 +327,9 @@ static int vmentry_l1d_flush_set(const char *s, const struct kernel_param *kp)
 
 static int vmentry_l1d_flush_get(char *s, const struct kernel_param *kp)
 {
+	if (WARN_ON_ONCE(l1tf_vmx_mitigation >= ARRAY_SIZE(vmentry_l1d_param)))
+		return sprintf(s, "???\n");
+
 	return sprintf(s, "%s\n", vmentry_l1d_param[l1tf_vmx_mitigation].option);
 }
 
@@ -1684,6 +1690,12 @@ static inline bool cpu_has_vmx_virtual_intr_delivery(void)
 		SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY;
 }
 
+static inline bool cpu_has_vmx_encls_vmexit(void)
+{
+	return vmcs_config.cpu_based_2nd_exec_ctrl &
+		SECONDARY_EXEC_ENCLS_EXITING;
+}
+
 /*
  * Comment's format: document - errata name - stepping - processor name.
  * Refer from
@@ -4551,7 +4563,8 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf)
 			SECONDARY_EXEC_RDRAND_EXITING |
 			SECONDARY_EXEC_ENABLE_PML |
 			SECONDARY_EXEC_TSC_SCALING |
-			SECONDARY_EXEC_ENABLE_VMFUNC;
+			SECONDARY_EXEC_ENABLE_VMFUNC |
+			SECONDARY_EXEC_ENCLS_EXITING;
 		if (adjust_vmx_controls(min2, opt2,
 					MSR_IA32_VMX_PROCBASED_CTLS2,
 					&_cpu_based_2nd_exec_control) < 0)
@@ -6648,6 +6661,9 @@ static void vmx_vcpu_setup(struct vcpu_vmx *vmx)
 		vmcs_write64(PML_ADDRESS, page_to_phys(vmx->pml_pg));
 		vmcs_write16(GUEST_PML_INDEX, PML_ENTITY_NUM - 1);
 	}
+
+	if (cpu_has_vmx_encls_vmexit())
+		vmcs_write64(ENCLS_EXITING_BITMAP, -1ull);
 }
 
 static void vmx_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
@@ -9314,6 +9330,17 @@ fail:
 	return 1;
 }
 
+static int handle_encls(struct kvm_vcpu *vcpu)
+{
+	/*
+	 * SGX virtualization is not yet supported.  There is no software
+	 * enable bit for SGX, so we have to trap ENCLS and inject a #UD
+	 * to prevent the guest from executing ENCLS.
+	 */
+	kvm_queue_exception(vcpu, UD_VECTOR);
+	return 1;
+}
+
 /*
  * The exit handlers return 1 if the exit was handled fully and guest execution
  * may resume.  Otherwise they set the kvm_run parameter to indicate what needs
@@ -9371,6 +9398,7 @@ static int (*const kvm_vmx_exit_handlers[])(struct kvm_vcpu *vcpu) = {
 	[EXIT_REASON_INVPCID]                 = handle_invpcid,
 	[EXIT_REASON_VMFUNC]                  = handle_vmfunc,
 	[EXIT_REASON_PREEMPTION_TIMER]	      = handle_preemption_timer,
+	[EXIT_REASON_ENCLS]		      = handle_encls,
 };
 
 static const int kvm_vmx_max_exit_handlers =
@@ -9741,6 +9769,9 @@ static bool nested_vmx_exit_reflected(struct kvm_vcpu *vcpu, u32 exit_reason)
 	case EXIT_REASON_VMFUNC:
 		/* VM functions are emulated through L2->L0 vmexits. */
 		return false;
+	case EXIT_REASON_ENCLS:
+		/* SGX is never exposed to L1 */
+		return false;
 	default:
 		return true;
 	}
@@ -12101,6 +12132,9 @@ static int prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12,
 		if (exec_control & SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES)
 			vmcs_write64(APIC_ACCESS_ADDR, -1ull);
 
+		if (exec_control & SECONDARY_EXEC_ENCLS_EXITING)
+			vmcs_write64(ENCLS_EXITING_BITMAP, -1ull);
+
 		vmcs_write32(SECONDARY_VM_EXEC_CONTROL, exec_control);
 	}
 

arch/x86/kvm/x86.c

@@ -6576,14 +6576,12 @@ static void kvm_set_mmio_spte_mask(void)
 	/* Set the present bit. */
 	mask |= 1ull;
 
-#ifdef CONFIG_X86_64
 	/*
 	 * If reserved bit is not supported, clear the present bit to disable
 	 * mmio page fault.
 	 */
-	if (maxphyaddr == 52)
+	if (IS_ENABLED(CONFIG_X86_64) && maxphyaddr == 52)
 		mask &= ~1ull;
-#endif
 
 	kvm_mmu_set_mmio_spte_mask(mask, mask);
 }

include/kvm/arm_vgic.h

@@ -133,6 +133,7 @@ struct vgic_irq {
 	u8 source;			/* GICv2 SGIs only */
 	u8 active_source;		/* GICv2 SGIs only */
 	u8 priority;
+	u8 group;			/* 0 == group 0, 1 == group 1 */
 	enum vgic_irq_config config;	/* Level or edge */
 
 	/*
@@ -217,6 +218,12 @@ struct vgic_dist {
 	/* vGIC model the kernel emulates for the guest (GICv2 or GICv3) */
 	u32			vgic_model;
 
+	/* Implementation revision as reported in the GICD_IIDR */
+	u32			implementation_rev;
+
+	/* Userspace can write to GICv2 IGROUPR */
+	bool			v2_groups_user_writable;
+
 	/* Do injected MSIs require an additional device ID? */
 	bool			msis_require_devid;
 
@@ -366,7 +373,7 @@ void kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu);
 void kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu);
 void kvm_vgic_reset_mapped_irq(struct kvm_vcpu *vcpu, u32 vintid);
 
-void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg);
+void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg, bool allow_group1);
 
 /**
  * kvm_vgic_get_max_vcpus - Get the maximum number of VCPUs allowed by HW

include/linux/irqchip/arm-gic-v3.h

@@ -61,6 +61,16 @@
 #define GICD_CTLR_ENABLE_G1A		(1U << 1)
 #define GICD_CTLR_ENABLE_G1		(1U << 0)
 
+#define GICD_IIDR_IMPLEMENTER_SHIFT	0
+#define GICD_IIDR_IMPLEMENTER_MASK	(0xfff << GICD_IIDR_IMPLEMENTER_SHIFT)
+#define GICD_IIDR_REVISION_SHIFT	12
+#define GICD_IIDR_REVISION_MASK		(0xf << GICD_IIDR_REVISION_SHIFT)
+#define GICD_IIDR_VARIANT_SHIFT		16
+#define GICD_IIDR_VARIANT_MASK		(0xf << GICD_IIDR_VARIANT_SHIFT)
+#define GICD_IIDR_PRODUCT_ID_SHIFT	24
+#define GICD_IIDR_PRODUCT_ID_MASK	(0xff << GICD_IIDR_PRODUCT_ID_SHIFT)
+
+
 /*
  * In systems with a single security state (what we emulate in KVM)
  * the meaning of the interrupt group enable bits is slightly different

include/linux/irqchip/arm-gic.h

@@ -71,6 +71,16 @@
 					(GICD_INT_DEF_PRI << 8) |\
 					GICD_INT_DEF_PRI)
 
+#define GICD_IIDR_IMPLEMENTER_SHIFT	0
+#define GICD_IIDR_IMPLEMENTER_MASK	(0xfff << GICD_IIDR_IMPLEMENTER_SHIFT)
+#define GICD_IIDR_REVISION_SHIFT	12
+#define GICD_IIDR_REVISION_MASK		(0xf << GICD_IIDR_REVISION_SHIFT)
+#define GICD_IIDR_VARIANT_SHIFT		16
+#define GICD_IIDR_VARIANT_MASK		(0xf << GICD_IIDR_VARIANT_SHIFT)
+#define GICD_IIDR_PRODUCT_ID_SHIFT	24
+#define GICD_IIDR_PRODUCT_ID_MASK	(0xff << GICD_IIDR_PRODUCT_ID_SHIFT)
+
+
 #define GICH_HCR			0x0
 #define GICH_VTR			0x4
 #define GICH_VMCR			0x8
@@ -94,6 +104,7 @@
 #define GICH_LR_PENDING_BIT		(1 << 28)
 #define GICH_LR_ACTIVE_BIT		(1 << 29)
 #define GICH_LR_EOI			(1 << 19)
+#define GICH_LR_GROUP1			(1 << 30)
 #define GICH_LR_HW			(1 << 31)
 
 #define GICH_VMCR_ENABLE_GRP0_SHIFT	0

include/uapi/linux/kvm.h

@@ -951,6 +951,7 @@ struct kvm_ppc_resize_hpt {
 #define KVM_CAP_HYPERV_TLBFLUSH 155
 #define KVM_CAP_S390_HPAGE_1M 156
 #define KVM_CAP_NESTED_STATE 157
+#define KVM_CAP_ARM_INJECT_SERROR_ESR 158
 
 #ifdef KVM_CAP_IRQ_ROUTING
 

tools/include/linux/bitmap.h

@@ -96,6 +96,23 @@ static inline int test_and_set_bit(int nr, unsigned long *addr)
 	return (old & mask) != 0;
 }
 
+/**
+ * test_and_clear_bit - Clear a bit and return its old value
+ * @nr: Bit to clear
+ * @addr: Address to count from
+ */
+static inline int test_and_clear_bit(int nr, unsigned long *addr)
+{
+	unsigned long mask = BIT_MASK(nr);
+	unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
+	unsigned long old;
+
+	old = *p;
+	*p = old & ~mask;
+
+	return (old & mask) != 0;
+}
+
 /**
  * bitmap_alloc - Allocate bitmap
  * @nbits: Number of bits

tools/testing/selftests/kvm/Makefile

@@ -11,13 +11,16 @@ TEST_GEN_PROGS_x86_64 += sync_regs_test
 TEST_GEN_PROGS_x86_64 += vmx_tsc_adjust_test
 TEST_GEN_PROGS_x86_64 += cr4_cpuid_sync_test
 TEST_GEN_PROGS_x86_64 += state_test
+TEST_GEN_PROGS_x86_64 += dirty_log_test
 
 TEST_GEN_PROGS += $(TEST_GEN_PROGS_$(UNAME_M))
 LIBKVM += $(LIBKVM_$(UNAME_M))
 
 INSTALL_HDR_PATH = $(top_srcdir)/usr
 LINUX_HDR_PATH = $(INSTALL_HDR_PATH)/include/
-CFLAGS += -O2 -g -std=gnu99 -I$(LINUX_HDR_PATH) -Iinclude -I$(<D) -I..
+LINUX_TOOL_INCLUDE = $(top_srcdir)tools/include
+CFLAGS += -O2 -g -std=gnu99 -I$(LINUX_TOOL_INCLUDE) -I$(LINUX_HDR_PATH) -Iinclude -I$(<D) -I..
+LDFLAGS += -lpthread
 
 # After inclusion, $(OUTPUT) is defined and
 # $(TEST_GEN_PROGS) starts with $(OUTPUT)/

tools/testing/selftests/kvm/cr4_cpuid_sync_test.c

@@ -23,20 +23,6 @@
 #define X86_FEATURE_OSXSAVE	(1<<27)
 #define VCPU_ID			1
 
-enum {
-	GUEST_UPDATE_CR4 = 0x1000,
-	GUEST_FAILED,
-	GUEST_DONE,
-};
-
-static void exit_to_hv(uint16_t port)
-{
-	__asm__ __volatile__("in %[port], %%al"
-			     :
-			     : [port]"d"(port)
-			     : "rax");
-}
-
 static inline bool cr4_cpuid_is_sync(void)
 {
 	int func, subfunc;
@@ -64,17 +50,15 @@ static void guest_code(void)
 	set_cr4(cr4);
 
 	/* verify CR4.OSXSAVE == CPUID.OSXSAVE */
-	if (!cr4_cpuid_is_sync())
-		exit_to_hv(GUEST_FAILED);
+	GUEST_ASSERT(cr4_cpuid_is_sync());
 
 	/* notify hypervisor to change CR4 */
-	exit_to_hv(GUEST_UPDATE_CR4);
+	GUEST_SYNC(0);
 
 	/* check again */
-	if (!cr4_cpuid_is_sync())
-		exit_to_hv(GUEST_FAILED);
+	GUEST_ASSERT(cr4_cpuid_is_sync());
 
-	exit_to_hv(GUEST_DONE);
+	GUEST_DONE();
 }
 
 int main(int argc, char *argv[])
@@ -95,7 +79,7 @@ int main(int argc, char *argv[])
 	setbuf(stdout, NULL);
 
 	/* Create VM */
-	vm = vm_create_default(VCPU_ID, guest_code);
+	vm = vm_create_default(VCPU_ID, 0, guest_code);
 	vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
 	run = vcpu_state(vm, VCPU_ID);
 
@@ -104,16 +88,16 @@ int main(int argc, char *argv[])
 
 		if (run->exit_reason == KVM_EXIT_IO) {
 			switch (run->io.port) {
-			case GUEST_UPDATE_CR4:
+			case GUEST_PORT_SYNC:
 				/* emulate hypervisor clearing CR4.OSXSAVE */
 				vcpu_sregs_get(vm, VCPU_ID, &sregs);
 				sregs.cr4 &= ~X86_CR4_OSXSAVE;
 				vcpu_sregs_set(vm, VCPU_ID, &sregs);
 				break;
-			case GUEST_FAILED:
+			case GUEST_PORT_ABORT:
 				TEST_ASSERT(false, "Guest CR4 bit (OSXSAVE) unsynchronized with CPUID bit.");
 				break;
-			case GUEST_DONE:
+			case GUEST_PORT_DONE:
 				goto done;
 			default:
 				TEST_ASSERT(false, "Unknown port 0x%x.",

tools/testing/selftests/kvm/dirty_log_test.c

@@ -0,0 +1,308 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * KVM dirty page logging test
+ *
+ * Copyright (C) 2018, Red Hat, Inc.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <time.h>
+#include <pthread.h>
+#include <linux/bitmap.h>
+#include <linux/bitops.h>
+
+#include "test_util.h"
+#include "kvm_util.h"
+
+#define  DEBUG                 printf
+
+#define  VCPU_ID                        1
+/* The memory slot index to track dirty pages */
+#define  TEST_MEM_SLOT_INDEX            1
+/*
+ * GPA offset of the testing memory slot. Must be bigger than the
+ * default vm mem slot, which is DEFAULT_GUEST_PHY_PAGES.
+ */
+#define  TEST_MEM_OFFSET                (1ULL << 30) /* 1G */
+/* Size of the testing memory slot */
+#define  TEST_MEM_PAGES                 (1ULL << 18) /* 1G for 4K pages */
+/* How many pages to dirty for each guest loop */
+#define  TEST_PAGES_PER_LOOP            1024
+/* How many host loops to run (one KVM_GET_DIRTY_LOG for each loop) */
+#define  TEST_HOST_LOOP_N               32
+/* Interval for each host loop (ms) */
+#define  TEST_HOST_LOOP_INTERVAL        10
+
+/*
+ * Guest variables.  We use these variables to share data between host
+ * and guest.  There are two copies of the variables, one in host memory
+ * (which is unused) and one in guest memory.  When the host wants to
+ * access these variables, it needs to call addr_gva2hva() to access the
+ * guest copy.
+ */
+uint64_t guest_random_array[TEST_PAGES_PER_LOOP];
+uint64_t guest_iteration;
+uint64_t guest_page_size;
+
+/*
+ * Writes to the first byte of a random page within the testing memory
+ * region continuously.
+ */
+void guest_code(void)
+{
+	int i = 0;
+	uint64_t volatile *array = guest_random_array;
+	uint64_t volatile *guest_addr;
+
+	while (true) {
+		for (i = 0; i < TEST_PAGES_PER_LOOP; i++) {
+			/*
+			 * Write to the first 8 bytes of a random page
+			 * on the testing memory region.
+			 */
+			guest_addr = (uint64_t *)
+			    (TEST_MEM_OFFSET +
+			     (array[i] % TEST_MEM_PAGES) * guest_page_size);
+			*guest_addr = guest_iteration;
+		}
+		/* Tell the host that we need more random numbers */
+		GUEST_SYNC(1);
+	}
+}
+
+/*
+ * Host variables.  These variables should only be used by the host
+ * rather than the guest.
+ */
+bool host_quit;
+
+/* Points to the test VM memory region on which we track dirty logs */
+void *host_test_mem;
+
+/* For statistics only */
+uint64_t host_dirty_count;
+uint64_t host_clear_count;
+uint64_t host_track_next_count;
+
+/*
+ * We use this bitmap to track some pages that should have its dirty
+ * bit set in the _next_ iteration.  For example, if we detected the
+ * page value changed to current iteration but at the same time the
+ * page bit is cleared in the latest bitmap, then the system must
+ * report that write in the next get dirty log call.
+ */
+unsigned long *host_bmap_track;
+
+void generate_random_array(uint64_t *guest_array, uint64_t size)
+{
+	uint64_t i;
+
+	for (i = 0; i < size; i++) {
+		guest_array[i] = random();
+	}
+}
+
+void *vcpu_worker(void *data)
+{
+	int ret;
+	uint64_t loops, *guest_array, pages_count = 0;
+	struct kvm_vm *vm = data;
+	struct kvm_run *run;
+	struct guest_args args;
+
+	run = vcpu_state(vm, VCPU_ID);
+
+	/* Retrieve the guest random array pointer and cache it */
+	guest_array = addr_gva2hva(vm, (vm_vaddr_t)guest_random_array);
+
+	DEBUG("VCPU starts\n");
+
+	generate_random_array(guest_array, TEST_PAGES_PER_LOOP);
+
+	while (!READ_ONCE(host_quit)) {
+		/* Let the guest to dirty these random pages */
+		ret = _vcpu_run(vm, VCPU_ID);
+		guest_args_read(vm, VCPU_ID, &args);
+		if (run->exit_reason == KVM_EXIT_IO &&
+		    args.port == GUEST_PORT_SYNC) {
+			pages_count += TEST_PAGES_PER_LOOP;
+			generate_random_array(guest_array, TEST_PAGES_PER_LOOP);
+		} else {
+			TEST_ASSERT(false,
+				    "Invalid guest sync status: "
+				    "exit_reason=%s\n",
+				    exit_reason_str(run->exit_reason));
+		}
+	}
+
+	DEBUG("VCPU exits, dirtied %"PRIu64" pages\n", pages_count);
+
+	return NULL;
+}
+
+void vm_dirty_log_verify(unsigned long *bmap, uint64_t iteration)
+{
+	uint64_t page;
+	uint64_t volatile *value_ptr;
+
+	for (page = 0; page < TEST_MEM_PAGES; page++) {
+		value_ptr = host_test_mem + page * getpagesize();
+
+		/* If this is a special page that we were tracking... */
+		if (test_and_clear_bit(page, host_bmap_track)) {
+			host_track_next_count++;
+			TEST_ASSERT(test_bit(page, bmap),
+				    "Page %"PRIu64" should have its dirty bit "
+				    "set in this iteration but it is missing",
+				    page);
+		}
+
+		if (test_bit(page, bmap)) {
+			host_dirty_count++;
+			/*
+			 * If the bit is set, the value written onto
+			 * the corresponding page should be either the
+			 * previous iteration number or the current one.
+			 */
+			TEST_ASSERT(*value_ptr == iteration ||
+				    *value_ptr == iteration - 1,
+				    "Set page %"PRIu64" value %"PRIu64
+				    " incorrect (iteration=%"PRIu64")",
+				    page, *value_ptr, iteration);
+		} else {
+			host_clear_count++;
+			/*
+			 * If cleared, the value written can be any
+			 * value smaller or equals to the iteration
+			 * number.  Note that the value can be exactly
+			 * (iteration-1) if that write can happen
+			 * like this:
+			 *
+			 * (1) increase loop count to "iteration-1"
+			 * (2) write to page P happens (with value
+			 *     "iteration-1")
+			 * (3) get dirty log for "iteration-1"; we'll
+			 *     see that page P bit is set (dirtied),
+			 *     and not set the bit in host_bmap_track
+			 * (4) increase loop count to "iteration"
+			 *     (which is current iteration)
+			 * (5) get dirty log for current iteration,
+			 *     we'll see that page P is cleared, with
+			 *     value "iteration-1".
+			 */
+			TEST_ASSERT(*value_ptr <= iteration,
+				    "Clear page %"PRIu64" value %"PRIu64
+				    " incorrect (iteration=%"PRIu64")",
+				    page, *value_ptr, iteration);
+			if (*value_ptr == iteration) {
+				/*
+				 * This page is _just_ modified; it
+				 * should report its dirtyness in the
+				 * next run
+				 */
+				set_bit(page, host_bmap_track);
+			}
+		}
+	}
+}
+
+void help(char *name)
+{
+	puts("");
+	printf("usage: %s [-i iterations] [-I interval] [-h]\n", name);
+	puts("");
+	printf(" -i: specify iteration counts (default: %"PRIu64")\n",
+	       TEST_HOST_LOOP_N);
+	printf(" -I: specify interval in ms (default: %"PRIu64" ms)\n",
+	       TEST_HOST_LOOP_INTERVAL);
+	puts("");
+	exit(0);
+}
+
+int main(int argc, char *argv[])
+{
+	pthread_t vcpu_thread;
+	struct kvm_vm *vm;
+	uint64_t volatile *psize, *iteration;
+	unsigned long *bmap, iterations = TEST_HOST_LOOP_N,
+	    interval = TEST_HOST_LOOP_INTERVAL;
+	int opt;
+
+	while ((opt = getopt(argc, argv, "hi:I:")) != -1) {
+		switch (opt) {
+		case 'i':
+			iterations = strtol(optarg, NULL, 10);
+			break;
+		case 'I':
+			interval = strtol(optarg, NULL, 10);
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			break;
+		}
+	}
+
+	TEST_ASSERT(iterations > 2, "Iteration must be bigger than zero\n");
+	TEST_ASSERT(interval > 0, "Interval must be bigger than zero");
+
+	DEBUG("Test iterations: %"PRIu64", interval: %"PRIu64" (ms)\n",
+	      iterations, interval);
+
+	srandom(time(0));
+
+	bmap = bitmap_alloc(TEST_MEM_PAGES);
+	host_bmap_track = bitmap_alloc(TEST_MEM_PAGES);
+
+	vm = vm_create_default(VCPU_ID, TEST_MEM_PAGES, guest_code);
+
+	/* Add an extra memory slot for testing dirty logging */
+	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
+				    TEST_MEM_OFFSET,
+				    TEST_MEM_SLOT_INDEX,
+				    TEST_MEM_PAGES,
+				    KVM_MEM_LOG_DIRTY_PAGES);
+	/* Cache the HVA pointer of the region */
+	host_test_mem = addr_gpa2hva(vm, (vm_paddr_t)TEST_MEM_OFFSET);
+
+	/* Do 1:1 mapping for the dirty track memory slot */
+	virt_map(vm, TEST_MEM_OFFSET, TEST_MEM_OFFSET,
+		 TEST_MEM_PAGES * getpagesize(), 0);
+
+	vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
+
+	/* Tell the guest about the page size on the system */
+	psize = addr_gva2hva(vm, (vm_vaddr_t)&guest_page_size);
+	*psize = getpagesize();
+
+	/* Start the iterations */
+	iteration = addr_gva2hva(vm, (vm_vaddr_t)&guest_iteration);
+	*iteration = 1;
+
+	/* Start dirtying pages */
+	pthread_create(&vcpu_thread, NULL, vcpu_worker, vm);
+
+	while (*iteration < iterations) {
+		/* Give the vcpu thread some time to dirty some pages */
+		usleep(interval * 1000);
+		kvm_vm_get_dirty_log(vm, TEST_MEM_SLOT_INDEX, bmap);
+		vm_dirty_log_verify(bmap, *iteration);
+		(*iteration)++;
+	}
+
+	/* Tell the vcpu thread to quit */
+	host_quit = true;
+	pthread_join(vcpu_thread, NULL);
+
+	DEBUG("Total bits checked: dirty (%"PRIu64"), clear (%"PRIu64"), "
+	      "track_next (%"PRIu64")\n", host_dirty_count, host_clear_count,
+	      host_track_next_count);
+
+	free(bmap);
+	free(host_bmap_track);
+	kvm_vm_free(vm);
+
+	return 0;
+}

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

@@ -55,6 +55,7 @@ struct kvm_vm *vm_create(enum vm_guest_mode mode, uint64_t phy_pages, int perm);
 void kvm_vm_free(struct kvm_vm *vmp);
 void kvm_vm_restart(struct kvm_vm *vmp, int perm);
 void kvm_vm_release(struct kvm_vm *vmp);
+void kvm_vm_get_dirty_log(struct kvm_vm *vm, int slot, void *log);
 
 int kvm_memcmp_hva_gva(void *hva,
 	struct kvm_vm *vm, const vm_vaddr_t gva, size_t len);
@@ -80,6 +81,8 @@ void vm_mem_region_set_flags(struct kvm_vm *vm, uint32_t slot, uint32_t flags);
 void vm_vcpu_add(struct kvm_vm *vm, uint32_t vcpuid, int pgd_memslot, int gdt_memslot);
 vm_vaddr_t vm_vaddr_alloc(struct kvm_vm *vm, size_t sz, vm_vaddr_t vaddr_min,
 	uint32_t data_memslot, uint32_t pgd_memslot);
+void virt_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
+	      size_t size, uint32_t pgd_memslot);
 void *addr_gpa2hva(struct kvm_vm *vm, vm_paddr_t gpa);
 void *addr_gva2hva(struct kvm_vm *vm, vm_vaddr_t gva);
 vm_paddr_t addr_hva2gpa(struct kvm_vm *vm, void *hva);
@@ -127,7 +130,8 @@ kvm_get_supported_cpuid_entry(uint32_t function)
 	return kvm_get_supported_cpuid_index(function, 0);
 }
 
-struct kvm_vm *vm_create_default(uint32_t vcpuid, void *guest_code);
+struct kvm_vm *vm_create_default(uint32_t vcpuid, uint64_t extra_mem_size,
+				 void *guest_code);
 void vm_vcpu_add_default(struct kvm_vm *vm, uint32_t vcpuid, void *guest_code);
 
 typedef void (*vmx_guest_code_t)(vm_vaddr_t vmxon_vaddr,
@@ -144,4 +148,43 @@ allocate_kvm_dirty_log(struct kvm_userspace_memory_region *region);
 
 int vm_create_device(struct kvm_vm *vm, struct kvm_create_device *cd);
 
+#define GUEST_PORT_SYNC         0x1000
+#define GUEST_PORT_ABORT        0x1001
+#define GUEST_PORT_DONE         0x1002
+
+static inline void __exit_to_l0(uint16_t port, uint64_t arg0, uint64_t arg1)
+{
+	__asm__ __volatile__("in %[port], %%al"
+			     :
+			     : [port]"d"(port), "D"(arg0), "S"(arg1)
+			     : "rax");
+}
+
+/*
+ * Allows to pass three arguments to the host: port is 16bit wide,
+ * arg0 & arg1 are 64bit wide
+ */
+#define GUEST_SYNC_ARGS(_port, _arg0, _arg1) \
+	__exit_to_l0(_port, (uint64_t) (_arg0), (uint64_t) (_arg1))
+
+#define GUEST_ASSERT(_condition) do {				\
+		if (!(_condition))				\
+			GUEST_SYNC_ARGS(GUEST_PORT_ABORT,	\
+					"Failed guest assert: "	\
+					#_condition, __LINE__);	\
+	} while (0)
+
+#define GUEST_SYNC(stage)  GUEST_SYNC_ARGS(GUEST_PORT_SYNC, "hello", stage)
+
+#define GUEST_DONE()  GUEST_SYNC_ARGS(GUEST_PORT_DONE, 0, 0)
+
+struct guest_args {
+	uint64_t arg0;
+	uint64_t arg1;
+	uint16_t port;
+} __attribute__ ((packed));
+
+void guest_args_read(struct kvm_vm *vm, uint32_t vcpu_id,
+		     struct guest_args *args);
+
 #endif /* SELFTEST_KVM_UTIL_H */

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

@@ -28,8 +28,6 @@ int test_seq_read(const char *path, char **bufp, size_t *sizep);
 void test_assert(bool exp, const char *exp_str,
 		 const char *file, unsigned int line, const char *fmt, ...);
 
-#define ARRAY_SIZE(array) (sizeof(array) / sizeof((array)[0]))
-
 #define TEST_ASSERT(e, fmt, ...) \
 	test_assert((e), #e, __FILE__, __LINE__, fmt, ##__VA_ARGS__)
 

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

@@ -14,6 +14,7 @@
 #include <sys/mman.h>
 #include <sys/types.h>
 #include <sys/stat.h>
+#include <linux/kernel.h>
 
 #define KVM_DEV_PATH "/dev/kvm"
 
@@ -168,6 +169,16 @@ void kvm_vm_restart(struct kvm_vm *vmp, int perm)
 	}
 }
 
+void kvm_vm_get_dirty_log(struct kvm_vm *vm, int slot, void *log)
+{
+	struct kvm_dirty_log args = { .dirty_bitmap = log, .slot = slot };
+	int ret;
+
+	ret = ioctl(vm->fd, KVM_GET_DIRTY_LOG, &args);
+	TEST_ASSERT(ret == 0, "%s: KVM_GET_DIRTY_LOG failed: %s",
+		    strerror(-ret));
+}
+
 /* Userspace Memory Region Find
  *
  * Input Args:
@@ -923,6 +934,39 @@ vm_vaddr_t vm_vaddr_alloc(struct kvm_vm *vm, size_t sz, vm_vaddr_t vaddr_min,
 	return vaddr_start;
 }
 
+/*
+ * Map a range of VM virtual address to the VM's physical address
+ *
+ * Input Args:
+ *   vm - Virtual Machine
+ *   vaddr - Virtuall address to map
+ *   paddr - VM Physical Address
+ *   size - The size of the range to map
+ *   pgd_memslot - Memory region slot for new virtual translation tables
+ *
+ * Output Args: None
+ *
+ * Return: None
+ *
+ * Within the VM given by vm, creates a virtual translation for the
+ * page range starting at vaddr to the page range starting at paddr.
+ */
+void virt_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
+	      size_t size, uint32_t pgd_memslot)
+{
+	size_t page_size = vm->page_size;
+	size_t npages = size / page_size;
+
+	TEST_ASSERT(vaddr + size > vaddr, "Vaddr overflow");
+	TEST_ASSERT(paddr + size > paddr, "Paddr overflow");
+
+	while (npages--) {
+		virt_pg_map(vm, vaddr, paddr, pgd_memslot);
+		vaddr += page_size;
+		paddr += page_size;
+	}
+}
+
 /* Address VM Physical to Host Virtual
  *
  * Input Args:
@@ -1536,3 +1580,17 @@ void *addr_gva2hva(struct kvm_vm *vm, vm_vaddr_t gva)
 {
 	return addr_gpa2hva(vm, addr_gva2gpa(vm, gva));
 }
+
+void guest_args_read(struct kvm_vm *vm, uint32_t vcpu_id,
+		     struct guest_args *args)
+{
+	struct kvm_run *run = vcpu_state(vm, vcpu_id);
+	struct kvm_regs regs;
+
+	memset(&regs, 0, sizeof(regs));
+	vcpu_regs_get(vm, vcpu_id, &regs);
+
+	args->port = run->io.port;
+	args->arg0 = regs.rdi;
+	args->arg1 = regs.rsi;
+}

tools/testing/selftests/kvm/lib/x86.c

@@ -702,6 +702,9 @@ void vcpu_set_cpuid(struct kvm_vm *vm,
  *
  * Input Args:
  *   vcpuid - The id of the single VCPU to add to the VM.
+ *   extra_mem_pages - The size of extra memories to add (this will
+ *                     decide how much extra space we will need to
+ *                     setup the page tables using mem slot 0)
  *   guest_code - The vCPU's entry point
  *
  * Output Args: None
@@ -709,12 +712,23 @@ void vcpu_set_cpuid(struct kvm_vm *vm,
  * Return:
  *   Pointer to opaque structure that describes the created VM.
  */
-struct kvm_vm *vm_create_default(uint32_t vcpuid, void *guest_code)
+struct kvm_vm *vm_create_default(uint32_t vcpuid, uint64_t extra_mem_pages,
+				 void *guest_code)
 {
 	struct kvm_vm *vm;
+	/*
+	 * For x86 the maximum page table size for a memory region
+	 * will be when only 4K pages are used.  In that case the
+	 * total extra size for page tables (for extra N pages) will
+	 * be: N/512+N/512^2+N/512^3+... which is definitely smaller
+	 * than N/512*2.
+	 */
+	uint64_t extra_pg_pages = extra_mem_pages / 512 * 2;
 
 	/* Create VM */
-	vm = vm_create(VM_MODE_FLAT48PG, DEFAULT_GUEST_PHY_PAGES, O_RDWR);
+	vm = vm_create(VM_MODE_FLAT48PG,
+		       DEFAULT_GUEST_PHY_PAGES + extra_pg_pages,
+		       O_RDWR);
 
 	/* Setup guest code */
 	kvm_vm_elf_load(vm, program_invocation_name, 0, 0);

tools/testing/selftests/kvm/set_sregs_test.c

@@ -36,7 +36,7 @@ int main(int argc, char *argv[])
 	setbuf(stdout, NULL);
 
 	/* Create VM */
-	vm = vm_create_default(VCPU_ID, NULL);
+	vm = vm_create_default(VCPU_ID, 0, NULL);
 
 	vcpu_sregs_get(vm, VCPU_ID, &sregs);
 	sregs.apic_base = 1 << 10;

tools/testing/selftests/kvm/state_test.c

@@ -21,28 +21,6 @@
 #include "vmx.h"
 
 #define VCPU_ID		5
-#define PORT_SYNC	0x1000
-#define PORT_ABORT	0x1001
-#define PORT_DONE	0x1002
-
-static inline void __exit_to_l0(uint16_t port, uint64_t arg0, uint64_t arg1)
-{
-	__asm__ __volatile__("in %[port], %%al"
-			     :
-			     : [port]"d"(port), "D"(arg0), "S"(arg1)
-			     : "rax");
-}
-
-#define exit_to_l0(_port, _arg0, _arg1) \
-	__exit_to_l0(_port, (uint64_t) (_arg0), (uint64_t) (_arg1))
-
-#define GUEST_ASSERT(_condition) do { \
-	if (!(_condition)) \
-		exit_to_l0(PORT_ABORT, "Failed guest assert: " #_condition, __LINE__);\
-} while (0)
-
-#define GUEST_SYNC(stage) \
-	exit_to_l0(PORT_SYNC, "hello", stage);
 
 static bool have_nested_state;
 
@@ -137,7 +115,7 @@ void guest_code(struct vmx_pages *vmx_pages)
 	if (vmx_pages)
 		l1_guest_code(vmx_pages);
 
-	exit_to_l0(PORT_DONE, 0, 0);
+	GUEST_DONE();
 }
 
 int main(int argc, char *argv[])
@@ -154,7 +132,7 @@ int main(int argc, char *argv[])
 	struct kvm_cpuid_entry2 *entry = kvm_get_supported_cpuid_entry(1);
 
 	/* Create VM */
-	vm = vm_create_default(VCPU_ID, guest_code);
+	vm = vm_create_default(VCPU_ID, 0, guest_code);
 	vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
 	run = vcpu_state(vm, VCPU_ID);
 
@@ -178,13 +156,13 @@ int main(int argc, char *argv[])
 		memset(&regs1, 0, sizeof(regs1));
 		vcpu_regs_get(vm, VCPU_ID, &regs1);
 		switch (run->io.port) {
-		case PORT_ABORT:
+		case GUEST_PORT_ABORT:
 			TEST_ASSERT(false, "%s at %s:%d", (const char *) regs1.rdi,
 				    __FILE__, regs1.rsi);
 			/* NOT REACHED */
-		case PORT_SYNC:
+		case GUEST_PORT_SYNC:
 			break;
-		case PORT_DONE:
+		case GUEST_PORT_DONE:
 			goto done;
 		default:
 			TEST_ASSERT(false, "Unknown port 0x%x.", run->io.port);

tools/testing/selftests/kvm/sync_regs_test.c

@@ -22,28 +22,11 @@
 #include "x86.h"
 
 #define VCPU_ID 5
-#define PORT_HOST_SYNC 0x1000
-
-static void __exit_to_l0(uint16_t port, uint64_t arg0, uint64_t arg1)
-{
-	        __asm__ __volatile__("in %[port], %%al"
-				     :
-				     : [port]"d"(port), "D"(arg0), "S"(arg1)
-				     : "rax");
-}
-
-#define exit_to_l0(_port, _arg0, _arg1) \
-        __exit_to_l0(_port, (uint64_t) (_arg0), (uint64_t) (_arg1))
-
-#define GUEST_ASSERT(_condition) do { \
-	if (!(_condition)) \
-		exit_to_l0(PORT_ABORT, "Failed guest assert: " #_condition, 0);\
-} while (0)
 
 void guest_code(void)
 {
 	for (;;) {
-		exit_to_l0(PORT_HOST_SYNC, "hello", 0);
+		GUEST_SYNC(0);
 		asm volatile ("inc %r11");
 	}
 }
@@ -111,7 +94,7 @@ int main(int argc, char *argv[])
 	}
 
 	/* Create VM */
-	vm = vm_create_default(VCPU_ID, guest_code);
+	vm = vm_create_default(VCPU_ID, 0, guest_code);
 
 	run = vcpu_state(vm, VCPU_ID);
 

tools/testing/selftests/kvm/vmx_tsc_adjust_test.c

@@ -62,27 +62,12 @@ struct kvm_single_msr {
 /* The virtual machine object. */
 static struct kvm_vm *vm;
 
-#define exit_to_l0(_port, _arg) do_exit_to_l0(_port, (unsigned long) (_arg))
-static void do_exit_to_l0(uint16_t port, unsigned long arg)
-{
-	__asm__ __volatile__("in %[port], %%al"
-		:
-		: [port]"d"(port), "D"(arg)
-		: "rax");
-}
-
-
-#define GUEST_ASSERT(_condition) do {					     \
-	if (!(_condition))						     \
-		exit_to_l0(PORT_ABORT, "Failed guest assert: " #_condition); \
-} while (0)
-
 static void check_ia32_tsc_adjust(int64_t max)
 {
 	int64_t adjust;
 
 	adjust = rdmsr(MSR_IA32_TSC_ADJUST);
-	exit_to_l0(PORT_REPORT, adjust);
+	GUEST_SYNC(adjust);
 	GUEST_ASSERT(adjust <= max);
 }
 
@@ -132,7 +117,7 @@ static void l1_guest_code(struct vmx_pages *vmx_pages)
 
 	check_ia32_tsc_adjust(-2 * TSC_ADJUST_VALUE);
 
-	exit_to_l0(PORT_DONE, 0);
+	GUEST_DONE();
 }
 
 void report(int64_t val)
@@ -152,7 +137,7 @@ int main(int argc, char *argv[])
 		exit(KSFT_SKIP);
 	}
 
-	vm = vm_create_default(VCPU_ID, (void *) l1_guest_code);
+	vm = vm_create_default(VCPU_ID, 0, (void *) l1_guest_code);
 	vcpu_set_cpuid(vm, VCPU_ID, kvm_get_supported_cpuid());
 
 	/* Allocate VMX pages and shared descriptors (vmx_pages). */
@@ -161,26 +146,26 @@ int main(int argc, char *argv[])
 
 	for (;;) {
 		volatile struct kvm_run *run = vcpu_state(vm, VCPU_ID);
-		struct kvm_regs regs;
+		struct guest_args args;
 
 		vcpu_run(vm, VCPU_ID);
-		vcpu_regs_get(vm, VCPU_ID, &regs);
+		guest_args_read(vm, VCPU_ID, &args);
 		TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
-			    "Got exit_reason other than KVM_EXIT_IO: %u (%s), rip=%lx\n",
+			    "Got exit_reason other than KVM_EXIT_IO: %u (%s)\n",
 			    run->exit_reason,
-			    exit_reason_str(run->exit_reason), regs.rip);
+			    exit_reason_str(run->exit_reason));
 
-		switch (run->io.port) {
-		case PORT_ABORT:
-			TEST_ASSERT(false, "%s", (const char *) regs.rdi);
+		switch (args.port) {
+		case GUEST_PORT_ABORT:
+			TEST_ASSERT(false, "%s", (const char *) args.arg0);
 			/* NOT REACHED */
-		case PORT_REPORT:
-			report(regs.rdi);
+		case GUEST_PORT_SYNC:
+			report(args.arg1);
 			break;
-		case PORT_DONE:
+		case GUEST_PORT_DONE:
 			goto done;
 		default:
-			TEST_ASSERT(false, "Unknown port 0x%x.", run->io.port);
+			TEST_ASSERT(false, "Unknown port 0x%x.", args.port);
 		}
 	}
 

virt/kvm/arm/arch_timer.c

@@ -295,9 +295,9 @@ static void phys_timer_emulate(struct kvm_vcpu *vcpu)
 	struct arch_timer_context *ptimer = vcpu_ptimer(vcpu);
 
 	/*
-	 * If the timer can fire now we have just raised the IRQ line and we
-	 * don't need to have a soft timer scheduled for the future.  If the
-	 * timer cannot fire at all, then we also don't need a soft timer.
+	 * If the timer can fire now, we don't need to have a soft timer
+	 * scheduled for the future.  If the timer cannot fire at all,
+	 * then we also don't need a soft timer.
 	 */
 	if (kvm_timer_should_fire(ptimer) || !kvm_timer_irq_can_fire(ptimer)) {
 		soft_timer_cancel(&timer->phys_timer, NULL);
@@ -332,10 +332,10 @@ static void kvm_timer_update_state(struct kvm_vcpu *vcpu)
 	level = kvm_timer_should_fire(vtimer);
 	kvm_timer_update_irq(vcpu, level, vtimer);
 
+	phys_timer_emulate(vcpu);
+
 	if (kvm_timer_should_fire(ptimer) != ptimer->irq.level)
 		kvm_timer_update_irq(vcpu, !ptimer->irq.level, ptimer);
-
-	phys_timer_emulate(vcpu);
 }
 
 static void vtimer_save_state(struct kvm_vcpu *vcpu)
@@ -487,6 +487,7 @@ void kvm_timer_vcpu_load(struct kvm_vcpu *vcpu)
 {
 	struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
 	struct arch_timer_context *vtimer = vcpu_vtimer(vcpu);
+	struct arch_timer_context *ptimer = vcpu_ptimer(vcpu);
 
 	if (unlikely(!timer->enabled))
 		return;
@@ -502,6 +503,10 @@ void kvm_timer_vcpu_load(struct kvm_vcpu *vcpu)
 
 	/* Set the background timer for the physical timer emulation. */
 	phys_timer_emulate(vcpu);
+
+	/* If the timer fired while we weren't running, inject it now */
+	if (kvm_timer_should_fire(ptimer) != ptimer->irq.level)
+		kvm_timer_update_irq(vcpu, !ptimer->irq.level, ptimer);
 }
 
 bool kvm_timer_should_notify_user(struct kvm_vcpu *vcpu)

virt/kvm/arm/arm.c

@@ -30,6 +30,7 @@
 #include <linux/kvm.h>
 #include <linux/kvm_irqfd.h>
 #include <linux/irqbypass.h>
+#include <linux/sched/stat.h>
 #include <trace/events/kvm.h>
 #include <kvm/arm_pmu.h>
 #include <kvm/arm_psci.h>
@@ -380,6 +381,11 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 	kvm_timer_vcpu_load(vcpu);
 	kvm_vcpu_load_sysregs(vcpu);
 	kvm_arch_vcpu_load_fp(vcpu);
+
+	if (single_task_running())
+		vcpu_clear_wfe_traps(vcpu);
+	else
+		vcpu_set_wfe_traps(vcpu);
 }
 
 void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
@@ -1044,6 +1050,32 @@ static int kvm_arm_vcpu_has_attr(struct kvm_vcpu *vcpu,
 	return ret;
 }
 
+static int kvm_arm_vcpu_get_events(struct kvm_vcpu *vcpu,
+				   struct kvm_vcpu_events *events)
+{
+	memset(events, 0, sizeof(*events));
+
+	return __kvm_arm_vcpu_get_events(vcpu, events);
+}
+
+static int kvm_arm_vcpu_set_events(struct kvm_vcpu *vcpu,
+				   struct kvm_vcpu_events *events)
+{
+	int i;
+
+	/* check whether the reserved field is zero */
+	for (i = 0; i < ARRAY_SIZE(events->reserved); i++)
+		if (events->reserved[i])
+			return -EINVAL;
+
+	/* check whether the pad field is zero */
+	for (i = 0; i < ARRAY_SIZE(events->exception.pad); i++)
+		if (events->exception.pad[i])
+			return -EINVAL;
+
+	return __kvm_arm_vcpu_set_events(vcpu, events);
+}
+
 long kvm_arch_vcpu_ioctl(struct file *filp,
 			 unsigned int ioctl, unsigned long arg)
 {
@@ -1124,6 +1156,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = kvm_arm_vcpu_has_attr(vcpu, &attr);
 		break;
 	}
+	case KVM_GET_VCPU_EVENTS: {
+		struct kvm_vcpu_events events;
+
+		if (kvm_arm_vcpu_get_events(vcpu, &events))
+			return -EINVAL;
+
+		if (copy_to_user(argp, &events, sizeof(events)))
+			return -EFAULT;
+
+		return 0;
+	}
+	case KVM_SET_VCPU_EVENTS: {
+		struct kvm_vcpu_events events;
+
+		if (copy_from_user(&events, argp, sizeof(events)))
+			return -EFAULT;
+
+		return kvm_arm_vcpu_set_events(vcpu, &events);
+	}
 	default:
 		r = -EINVAL;
 	}

virt/kvm/arm/mmu.c

@@ -177,6 +177,35 @@ static void clear_stage2_pmd_entry(struct kvm *kvm, pmd_t *pmd, phys_addr_t addr
 	put_page(virt_to_page(pmd));
 }
 
+static inline void kvm_set_pte(pte_t *ptep, pte_t new_pte)
+{
+	WRITE_ONCE(*ptep, new_pte);
+	dsb(ishst);
+}
+
+static inline void kvm_set_pmd(pmd_t *pmdp, pmd_t new_pmd)
+{
+	WRITE_ONCE(*pmdp, new_pmd);
+	dsb(ishst);
+}
+
+static inline void kvm_pmd_populate(pmd_t *pmdp, pte_t *ptep)
+{
+	kvm_set_pmd(pmdp, kvm_mk_pmd(ptep));
+}
+
+static inline void kvm_pud_populate(pud_t *pudp, pmd_t *pmdp)
+{
+	WRITE_ONCE(*pudp, kvm_mk_pud(pmdp));
+	dsb(ishst);
+}
+
+static inline void kvm_pgd_populate(pgd_t *pgdp, pud_t *pudp)
+{
+	WRITE_ONCE(*pgdp, kvm_mk_pgd(pudp));
+	dsb(ishst);
+}
+
 /*
  * Unmapping vs dcache management:
  *
@@ -196,6 +225,10 @@ static void clear_stage2_pmd_entry(struct kvm *kvm, pmd_t *pmd, phys_addr_t addr
  * This is why right after unmapping a page/section and invalidating
  * the corresponding TLBs, we call kvm_flush_dcache_p*() to make sure
  * the IO subsystem will never hit in the cache.
+ *
+ * This is all avoided on systems that have ARM64_HAS_STAGE2_FWB, as
+ * we then fully enforce cacheability of RAM, no matter what the guest
+ * does.
  */
 static void unmap_stage2_ptes(struct kvm *kvm, pmd_t *pmd,
 		       phys_addr_t addr, phys_addr_t end)
@@ -576,7 +609,6 @@ static void create_hyp_pte_mappings(pmd_t *pmd, unsigned long start,
 		pte = pte_offset_kernel(pmd, addr);
 		kvm_set_pte(pte, pfn_pte(pfn, prot));
 		get_page(virt_to_page(pte));
-		kvm_flush_dcache_to_poc(pte, sizeof(*pte));
 		pfn++;
 	} while (addr += PAGE_SIZE, addr != end);
 }
@@ -601,9 +633,8 @@ static int create_hyp_pmd_mappings(pud_t *pud, unsigned long start,
 				kvm_err("Cannot allocate Hyp pte\n");
 				return -ENOMEM;
 			}
-			pmd_populate_kernel(NULL, pmd, pte);
+			kvm_pmd_populate(pmd, pte);
 			get_page(virt_to_page(pmd));
-			kvm_flush_dcache_to_poc(pmd, sizeof(*pmd));
 		}
 
 		next = pmd_addr_end(addr, end);
@@ -634,9 +665,8 @@ static int create_hyp_pud_mappings(pgd_t *pgd, unsigned long start,
 				kvm_err("Cannot allocate Hyp pmd\n");
 				return -ENOMEM;
 			}
-			pud_populate(NULL, pud, pmd);
+			kvm_pud_populate(pud, pmd);
 			get_page(virt_to_page(pud));
-			kvm_flush_dcache_to_poc(pud, sizeof(*pud));
 		}
 
 		next = pud_addr_end(addr, end);
@@ -671,9 +701,8 @@ static int __create_hyp_mappings(pgd_t *pgdp, unsigned long ptrs_per_pgd,
 				err = -ENOMEM;
 				goto out;
 			}
-			pgd_populate(NULL, pgd, pud);
+			kvm_pgd_populate(pgd, pud);
 			get_page(virt_to_page(pgd));
-			kvm_flush_dcache_to_poc(pgd, sizeof(*pgd));
 		}
 
 		next = pgd_addr_end(addr, end);
@@ -1015,19 +1044,35 @@ static int stage2_set_pmd_huge(struct kvm *kvm, struct kvm_mmu_memory_cache
 	pmd = stage2_get_pmd(kvm, cache, addr);
 	VM_BUG_ON(!pmd);
 
-	/*
-	 * Mapping in huge pages should only happen through a fault.  If a
-	 * page is merged into a transparent huge page, the individual
-	 * subpages of that huge page should be unmapped through MMU
-	 * notifiers before we get here.
-	 *
-	 * Merging of CompoundPages is not supported; they should become
-	 * splitting first, unmapped, merged, and mapped back in on-demand.
-	 */
-	VM_BUG_ON(pmd_present(*pmd) && pmd_pfn(*pmd) != pmd_pfn(*new_pmd));
-
 	old_pmd = *pmd;
 	if (pmd_present(old_pmd)) {
+		/*
+		 * Multiple vcpus faulting on the same PMD entry, can
+		 * lead to them sequentially updating the PMD with the
+		 * same value. Following the break-before-make
+		 * (pmd_clear() followed by tlb_flush()) process can
+		 * hinder forward progress due to refaults generated
+		 * on missing translations.
+		 *
+		 * Skip updating the page table if the entry is
+		 * unchanged.
+		 */
+		if (pmd_val(old_pmd) == pmd_val(*new_pmd))
+			return 0;
+
+		/*
+		 * Mapping in huge pages should only happen through a
+		 * fault.  If a page is merged into a transparent huge
+		 * page, the individual subpages of that huge page
+		 * should be unmapped through MMU notifiers before we
+		 * get here.
+		 *
+		 * Merging of CompoundPages is not supported; they
+		 * should become splitting first, unmapped, merged,
+		 * and mapped back in on-demand.
+		 */
+		VM_BUG_ON(pmd_pfn(old_pmd) != pmd_pfn(*new_pmd));
+
 		pmd_clear(pmd);
 		kvm_tlb_flush_vmid_ipa(kvm, addr);
 	} else {
@@ -1090,7 +1135,7 @@ static int stage2_set_pte(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
 		if (!cache)
 			return 0; /* ignore calls from kvm_set_spte_hva */
 		pte = mmu_memory_cache_alloc(cache);
-		pmd_populate_kernel(NULL, pmd, pte);
+		kvm_pmd_populate(pmd, pte);
 		get_page(virt_to_page(pmd));
 	}
 
@@ -1102,6 +1147,10 @@ static int stage2_set_pte(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
 	/* Create 2nd stage page table mapping - Level 3 */
 	old_pte = *pte;
 	if (pte_present(old_pte)) {
+		/* Skip page table update if there is no change */
+		if (pte_val(old_pte) == pte_val(*new_pte))
+			return 0;
+
 		kvm_set_pte(pte, __pte(0));
 		kvm_tlb_flush_vmid_ipa(kvm, addr);
 	} else {

virt/kvm/arm/vgic/vgic-debug.c

@@ -36,9 +36,12 @@
 struct vgic_state_iter {
 	int nr_cpus;
 	int nr_spis;
+	int nr_lpis;
 	int dist_id;
 	int vcpu_id;
 	int intid;
+	int lpi_idx;
+	u32 *lpi_array;
 };
 
 static void iter_next(struct vgic_state_iter *iter)
@@ -52,6 +55,12 @@ static void iter_next(struct vgic_state_iter *iter)
 	if (iter->intid == VGIC_NR_PRIVATE_IRQS &&
 	    ++iter->vcpu_id < iter->nr_cpus)
 		iter->intid = 0;
+
+	if (iter->intid >= (iter->nr_spis + VGIC_NR_PRIVATE_IRQS)) {
+		if (iter->lpi_idx < iter->nr_lpis)
+			iter->intid = iter->lpi_array[iter->lpi_idx];
+		iter->lpi_idx++;
+	}
 }
 
 static void iter_init(struct kvm *kvm, struct vgic_state_iter *iter,
@@ -63,6 +72,11 @@ static void iter_init(struct kvm *kvm, struct vgic_state_iter *iter,
 
 	iter->nr_cpus = nr_cpus;
 	iter->nr_spis = kvm->arch.vgic.nr_spis;
+	if (kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) {
+		iter->nr_lpis = vgic_copy_lpi_list(kvm, NULL, &iter->lpi_array);
+		if (iter->nr_lpis < 0)
+			iter->nr_lpis = 0;
+	}
 
 	/* Fast forward to the right position if needed */
 	while (pos--)
@@ -73,7 +87,8 @@ static bool end_of_vgic(struct vgic_state_iter *iter)
 {
 	return iter->dist_id > 0 &&
 		iter->vcpu_id == iter->nr_cpus &&
-		(iter->intid - VGIC_NR_PRIVATE_IRQS) == iter->nr_spis;
+		iter->intid >= (iter->nr_spis + VGIC_NR_PRIVATE_IRQS) &&
+		iter->lpi_idx > iter->nr_lpis;
 }
 
 static void *vgic_debug_start(struct seq_file *s, loff_t *pos)
@@ -130,6 +145,7 @@ static void vgic_debug_stop(struct seq_file *s, void *v)
 
 	mutex_lock(&kvm->lock);
 	iter = kvm->arch.vgic.iter;
+	kfree(iter->lpi_array);
 	kfree(iter);
 	kvm->arch.vgic.iter = NULL;
 	mutex_unlock(&kvm->lock);
@@ -137,17 +153,20 @@ static void vgic_debug_stop(struct seq_file *s, void *v)
 
 static void print_dist_state(struct seq_file *s, struct vgic_dist *dist)
 {
+	bool v3 = dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3;
+
 	seq_printf(s, "Distributor\n");
 	seq_printf(s, "===========\n");
-	seq_printf(s, "vgic_model:\t%s\n",
-		   (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) ?
-		   "GICv3" : "GICv2");
+	seq_printf(s, "vgic_model:\t%s\n", v3 ? "GICv3" : "GICv2");
 	seq_printf(s, "nr_spis:\t%d\n", dist->nr_spis);
+	if (v3)
+		seq_printf(s, "nr_lpis:\t%d\n", dist->lpi_list_count);
 	seq_printf(s, "enabled:\t%d\n", dist->enabled);
 	seq_printf(s, "\n");
 
 	seq_printf(s, "P=pending_latch, L=line_level, A=active\n");
 	seq_printf(s, "E=enabled, H=hw, C=config (level=1, edge=0)\n");
+	seq_printf(s, "G=group\n");
 }
 
 static void print_header(struct seq_file *s, struct vgic_irq *irq,
@@ -162,8 +181,8 @@ static void print_header(struct seq_file *s, struct vgic_irq *irq,
 	}
 
 	seq_printf(s, "\n");
-	seq_printf(s, "%s%2d TYP   ID TGT_ID PLAEHC     HWID   TARGET SRC PRI VCPU_ID\n", hdr, id);
-	seq_printf(s, "---------------------------------------------------------------\n");
+	seq_printf(s, "%s%2d TYP   ID TGT_ID PLAEHCG     HWID   TARGET SRC PRI VCPU_ID\n", hdr, id);
+	seq_printf(s, "----------------------------------------------------------------\n");
 }
 
 static void print_irq_state(struct seq_file *s, struct vgic_irq *irq,
@@ -174,15 +193,17 @@ static void print_irq_state(struct seq_file *s, struct vgic_irq *irq,
 		type = "SGI";
 	else if (irq->intid < VGIC_NR_PRIVATE_IRQS)
 		type = "PPI";
-	else
+	else if (irq->intid < VGIC_MAX_SPI)
 		type = "SPI";
+	else
+		type = "LPI";
 
 	if (irq->intid ==0 || irq->intid == VGIC_NR_PRIVATE_IRQS)
 		print_header(s, irq, vcpu);
 
 	seq_printf(s, "       %s %4d "
 		      "    %2d "
-		      "%d%d%d%d%d%d "
+		      "%d%d%d%d%d%d%d "
 		      "%8d "
 		      "%8x "
 		      " %2x "
@@ -197,12 +218,12 @@ static void print_irq_state(struct seq_file *s, struct vgic_irq *irq,
 			irq->enabled,
 			irq->hw,
 			irq->config == VGIC_CONFIG_LEVEL,
+			irq->group,
 			irq->hwintid,
 			irq->mpidr,
 			irq->source,
 			irq->priority,
 			(irq->vcpu) ? irq->vcpu->vcpu_id : -1);
-
 }
 
 static int vgic_debug_show(struct seq_file *s, void *v)
@@ -221,17 +242,20 @@ static int vgic_debug_show(struct seq_file *s, void *v)
 	if (!kvm->arch.vgic.initialized)
 		return 0;
 
-	if (iter->vcpu_id < iter->nr_cpus) {
+	if (iter->vcpu_id < iter->nr_cpus)
 		vcpu = kvm_get_vcpu(kvm, iter->vcpu_id);
-		irq = &vcpu->arch.vgic_cpu.private_irqs[iter->intid];
-	} else {
-		irq = &kvm->arch.vgic.spis[iter->intid - VGIC_NR_PRIVATE_IRQS];
+
+	irq = vgic_get_irq(kvm, vcpu, iter->intid);
+	if (!irq) {
+		seq_printf(s, "       LPI %4d freed\n", iter->intid);
+		return 0;
 	}
 
 	spin_lock_irqsave(&irq->irq_lock, flags);
 	print_irq_state(s, irq, vcpu);
 	spin_unlock_irqrestore(&irq->irq_lock, flags);
 
+	vgic_put_irq(kvm, irq);
 	return 0;
 }
 

virt/kvm/arm/vgic/vgic-init.c

@@ -175,10 +175,13 @@ static int kvm_vgic_dist_init(struct kvm *kvm, unsigned int nr_spis)
 		irq->vcpu = NULL;
 		irq->target_vcpu = vcpu0;
 		kref_init(&irq->refcount);
-		if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V2)
+		if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V2) {
 			irq->targets = 0;
-		else
+			irq->group = 0;
+		} else {
 			irq->mpidr = 0;
+			irq->group = 1;
+		}
 	}
 	return 0;
 }
@@ -227,6 +230,18 @@ int kvm_vgic_vcpu_init(struct kvm_vcpu *vcpu)
 			/* PPIs */
 			irq->config = VGIC_CONFIG_LEVEL;
 		}
+
+		/*
+		 * GICv3 can only be created via the KVM_DEVICE_CREATE API and
+		 * so we always know the emulation type at this point as it's
+		 * either explicitly configured as GICv3, or explicitly
+		 * configured as GICv2, or not configured yet which also
+		 * implies GICv2.
+		 */
+		if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3)
+			irq->group = 1;
+		else
+			irq->group = 0;
 	}
 
 	if (!irqchip_in_kernel(vcpu->kvm))
@@ -271,6 +286,10 @@ int vgic_init(struct kvm *kvm)
 	if (vgic_initialized(kvm))
 		return 0;
 
+	/* Are we also in the middle of creating a VCPU? */
+	if (kvm->created_vcpus != atomic_read(&kvm->online_vcpus))
+		return -EBUSY;
+
 	/* freeze the number of spis */
 	if (!dist->nr_spis)
 		dist->nr_spis = VGIC_NR_IRQS_LEGACY - VGIC_NR_PRIVATE_IRQS;
@@ -294,6 +313,7 @@ int vgic_init(struct kvm *kvm)
 
 	vgic_debug_init(kvm);
 
+	dist->implementation_rev = 2;
 	dist->initialized = true;
 
 out:

virt/kvm/arm/vgic/vgic-its.c

@@ -71,6 +71,7 @@ static struct vgic_irq *vgic_add_lpi(struct kvm *kvm, u32 intid,
 	kref_init(&irq->refcount);
 	irq->intid = intid;
 	irq->target_vcpu = vcpu;
+	irq->group = 1;
 
 	spin_lock_irqsave(&dist->lpi_list_lock, flags);
 
@@ -168,8 +169,14 @@ struct vgic_its_abi {
 	int (*commit)(struct vgic_its *its);
 };
 
+#define ABI_0_ESZ	8
+#define ESZ_MAX		ABI_0_ESZ
+
 static const struct vgic_its_abi its_table_abi_versions[] = {
-	[0] = {.cte_esz = 8, .dte_esz = 8, .ite_esz = 8,
+	[0] = {
+	 .cte_esz = ABI_0_ESZ,
+	 .dte_esz = ABI_0_ESZ,
+	 .ite_esz = ABI_0_ESZ,
 	 .save_tables = vgic_its_save_tables_v0,
 	 .restore_tables = vgic_its_restore_tables_v0,
 	 .commit = vgic_its_commit_v0,
@@ -183,7 +190,7 @@ inline const struct vgic_its_abi *vgic_its_get_abi(struct vgic_its *its)
 	return &its_table_abi_versions[its->abi_rev];
 }
 
-int vgic_its_set_abi(struct vgic_its *its, int rev)
+static int vgic_its_set_abi(struct vgic_its *its, u32 rev)
 {
 	const struct vgic_its_abi *abi;
 
@@ -312,9 +319,9 @@ static int update_lpi_config(struct kvm *kvm, struct vgic_irq *irq,
  * enumerate those LPIs without holding any lock.
  * Returns their number and puts the kmalloc'ed array into intid_ptr.
  */
-static int vgic_copy_lpi_list(struct kvm_vcpu *vcpu, u32 **intid_ptr)
+int vgic_copy_lpi_list(struct kvm *kvm, struct kvm_vcpu *vcpu, u32 **intid_ptr)
 {
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
+	struct vgic_dist *dist = &kvm->arch.vgic;
 	struct vgic_irq *irq;
 	unsigned long flags;
 	u32 *intids;
@@ -337,7 +344,7 @@ static int vgic_copy_lpi_list(struct kvm_vcpu *vcpu, u32 **intid_ptr)
 		if (i == irq_count)
 			break;
 		/* We don't need to "get" the IRQ, as we hold the list lock. */
-		if (irq->target_vcpu != vcpu)
+		if (vcpu && irq->target_vcpu != vcpu)
 			continue;
 		intids[i++] = irq->intid;
 	}
@@ -429,7 +436,7 @@ static int its_sync_lpi_pending_table(struct kvm_vcpu *vcpu)
 	unsigned long flags;
 	u8 pendmask;
 
-	nr_irqs = vgic_copy_lpi_list(vcpu, &intids);
+	nr_irqs = vgic_copy_lpi_list(vcpu->kvm, vcpu, &intids);
 	if (nr_irqs < 0)
 		return nr_irqs;
 
@@ -1154,7 +1161,7 @@ static int vgic_its_cmd_handle_invall(struct kvm *kvm, struct vgic_its *its,
 
 	vcpu = kvm_get_vcpu(kvm, collection->target_addr);
 
-	irq_count = vgic_copy_lpi_list(vcpu, &intids);
+	irq_count = vgic_copy_lpi_list(kvm, vcpu, &intids);
 	if (irq_count < 0)
 		return irq_count;
 
@@ -1202,7 +1209,7 @@ static int vgic_its_cmd_handle_movall(struct kvm *kvm, struct vgic_its *its,
 	vcpu1 = kvm_get_vcpu(kvm, target1_addr);
 	vcpu2 = kvm_get_vcpu(kvm, target2_addr);
 
-	irq_count = vgic_copy_lpi_list(vcpu1, &intids);
+	irq_count = vgic_copy_lpi_list(kvm, vcpu1, &intids);
 	if (irq_count < 0)
 		return irq_count;
 
@@ -1881,14 +1888,14 @@ typedef int (*entry_fn_t)(struct vgic_its *its, u32 id, void *entry,
  * Return: < 0 on error, 0 if last element was identified, 1 otherwise
  * (the last element may not be found on second level tables)
  */
-static int scan_its_table(struct vgic_its *its, gpa_t base, int size, int esz,
+static int scan_its_table(struct vgic_its *its, gpa_t base, int size, u32 esz,
 			  int start_id, entry_fn_t fn, void *opaque)
 {
 	struct kvm *kvm = its->dev->kvm;
 	unsigned long len = size;
 	int id = start_id;
 	gpa_t gpa = base;
-	char entry[esz];
+	char entry[ESZ_MAX];
 	int ret;
 
 	memset(entry, 0, esz);

virt/kvm/arm/vgic/vgic-mmio-v2.c

@@ -22,22 +22,33 @@
 #include "vgic.h"
 #include "vgic-mmio.h"
 
+/*
+ * The Revision field in the IIDR have the following meanings:
+ *
+ * Revision 1: Report GICv2 interrupts as group 0 instead of group 1
+ * Revision 2: Interrupt groups are guest-configurable and signaled using
+ * 	       their configured groups.
+ */
+
 static unsigned long vgic_mmio_read_v2_misc(struct kvm_vcpu *vcpu,
 					    gpa_t addr, unsigned int len)
 {
+	struct vgic_dist *vgic = &vcpu->kvm->arch.vgic;
 	u32 value;
 
 	switch (addr & 0x0c) {
 	case GIC_DIST_CTRL:
-		value = vcpu->kvm->arch.vgic.enabled ? GICD_ENABLE : 0;
+		value = vgic->enabled ? GICD_ENABLE : 0;
 		break;
 	case GIC_DIST_CTR:
-		value = vcpu->kvm->arch.vgic.nr_spis + VGIC_NR_PRIVATE_IRQS;
+		value = vgic->nr_spis + VGIC_NR_PRIVATE_IRQS;
 		value = (value >> 5) - 1;
 		value |= (atomic_read(&vcpu->kvm->online_vcpus) - 1) << 5;
 		break;
 	case GIC_DIST_IIDR:
-		value = (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
+		value = (PRODUCT_ID_KVM << GICD_IIDR_PRODUCT_ID_SHIFT) |
+			(vgic->implementation_rev << GICD_IIDR_REVISION_SHIFT) |
+			(IMPLEMENTER_ARM << GICD_IIDR_IMPLEMENTER_SHIFT);
 		break;
 	default:
 		return 0;
@@ -66,6 +77,42 @@ static void vgic_mmio_write_v2_misc(struct kvm_vcpu *vcpu,
 	}
 }
 
+static int vgic_mmio_uaccess_write_v2_misc(struct kvm_vcpu *vcpu,
+					   gpa_t addr, unsigned int len,
+					   unsigned long val)
+{
+	switch (addr & 0x0c) {
+	case GIC_DIST_IIDR:
+		if (val != vgic_mmio_read_v2_misc(vcpu, addr, len))
+			return -EINVAL;
+
+		/*
+		 * If we observe a write to GICD_IIDR we know that userspace
+		 * has been updated and has had a chance to cope with older
+		 * kernels (VGICv2 IIDR.Revision == 0) incorrectly reporting
+		 * interrupts as group 1, and therefore we now allow groups to
+		 * be user writable.  Doing this by default would break
+		 * migration from old kernels to new kernels with legacy
+		 * userspace.
+		 */
+		vcpu->kvm->arch.vgic.v2_groups_user_writable = true;
+		return 0;
+	}
+
+	vgic_mmio_write_v2_misc(vcpu, addr, len, val);
+	return 0;
+}
+
+static int vgic_mmio_uaccess_write_v2_group(struct kvm_vcpu *vcpu,
+					    gpa_t addr, unsigned int len,
+					    unsigned long val)
+{
+	if (vcpu->kvm->arch.vgic.v2_groups_user_writable)
+		vgic_mmio_write_group(vcpu, addr, len, val);
+
+	return 0;
+}
+
 static void vgic_mmio_write_sgir(struct kvm_vcpu *source_vcpu,
 				 gpa_t addr, unsigned int len,
 				 unsigned long val)
@@ -352,17 +399,22 @@ static void vgic_mmio_write_apr(struct kvm_vcpu *vcpu,
 
 		if (n > vgic_v3_max_apr_idx(vcpu))
 			return;
+
+		n = array_index_nospec(n, 4);
+
 		/* GICv3 only uses ICH_AP1Rn for memory mapped (GICv2) guests */
 		vgicv3->vgic_ap1r[n] = val;
 	}
 }
 
 static const struct vgic_register_region vgic_v2_dist_registers[] = {
-	REGISTER_DESC_WITH_LENGTH(GIC_DIST_CTRL,
-		vgic_mmio_read_v2_misc, vgic_mmio_write_v2_misc, 12,
-		VGIC_ACCESS_32bit),
+	REGISTER_DESC_WITH_LENGTH_UACCESS(GIC_DIST_CTRL,
+		vgic_mmio_read_v2_misc, vgic_mmio_write_v2_misc,
+		NULL, vgic_mmio_uaccess_write_v2_misc,
+		12, VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_BITS_PER_IRQ(GIC_DIST_IGROUP,
-		vgic_mmio_read_rao, vgic_mmio_write_wi, NULL, NULL, 1,
+		vgic_mmio_read_group, vgic_mmio_write_group,
+		NULL, vgic_mmio_uaccess_write_v2_group, 1,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_BITS_PER_IRQ(GIC_DIST_ENABLE_SET,
 		vgic_mmio_read_enable, vgic_mmio_write_senable, NULL, NULL, 1,

virt/kvm/arm/vgic/vgic-mmio-v3.c

@@ -59,19 +59,27 @@ bool vgic_supports_direct_msis(struct kvm *kvm)
 	return kvm_vgic_global_state.has_gicv4 && vgic_has_its(kvm);
 }
 
+/*
+ * The Revision field in the IIDR have the following meanings:
+ *
+ * Revision 2: Interrupt groups are guest-configurable and signaled using
+ * 	       their configured groups.
+ */
+
 static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
 					    gpa_t addr, unsigned int len)
 {
+	struct vgic_dist *vgic = &vcpu->kvm->arch.vgic;
 	u32 value = 0;
 
 	switch (addr & 0x0c) {
 	case GICD_CTLR:
-		if (vcpu->kvm->arch.vgic.enabled)
+		if (vgic->enabled)
 			value |= GICD_CTLR_ENABLE_SS_G1;
 		value |= GICD_CTLR_ARE_NS | GICD_CTLR_DS;
 		break;
 	case GICD_TYPER:
-		value = vcpu->kvm->arch.vgic.nr_spis + VGIC_NR_PRIVATE_IRQS;
+		value = vgic->nr_spis + VGIC_NR_PRIVATE_IRQS;
 		value = (value >> 5) - 1;
 		if (vgic_has_its(vcpu->kvm)) {
 			value |= (INTERRUPT_ID_BITS_ITS - 1) << 19;
@@ -81,7 +89,9 @@ static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
 		}
 		break;
 	case GICD_IIDR:
-		value = (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
+		value = (PRODUCT_ID_KVM << GICD_IIDR_PRODUCT_ID_SHIFT) |
+			(vgic->implementation_rev << GICD_IIDR_REVISION_SHIFT) |
+			(IMPLEMENTER_ARM << GICD_IIDR_IMPLEMENTER_SHIFT);
 		break;
 	default:
 		return 0;
@@ -110,6 +120,20 @@ static void vgic_mmio_write_v3_misc(struct kvm_vcpu *vcpu,
 	}
 }
 
+static int vgic_mmio_uaccess_write_v3_misc(struct kvm_vcpu *vcpu,
+					   gpa_t addr, unsigned int len,
+					   unsigned long val)
+{
+	switch (addr & 0x0c) {
+	case GICD_IIDR:
+		if (val != vgic_mmio_read_v3_misc(vcpu, addr, len))
+			return -EINVAL;
+	}
+
+	vgic_mmio_write_v3_misc(vcpu, addr, len, val);
+	return 0;
+}
+
 static unsigned long vgic_mmio_read_irouter(struct kvm_vcpu *vcpu,
 					    gpa_t addr, unsigned int len)
 {
@@ -246,9 +270,9 @@ static unsigned long vgic_v3_uaccess_read_pending(struct kvm_vcpu *vcpu,
 	return value;
 }
 
-static void vgic_v3_uaccess_write_pending(struct kvm_vcpu *vcpu,
-					  gpa_t addr, unsigned int len,
-					  unsigned long val)
+static int vgic_v3_uaccess_write_pending(struct kvm_vcpu *vcpu,
+					 gpa_t addr, unsigned int len,
+					 unsigned long val)
 {
 	u32 intid = VGIC_ADDR_TO_INTID(addr, 1);
 	int i;
@@ -273,6 +297,8 @@ static void vgic_v3_uaccess_write_pending(struct kvm_vcpu *vcpu,
 
 		vgic_put_irq(vcpu->kvm, irq);
 	}
+
+	return 0;
 }
 
 /* We want to avoid outer shareable. */
@@ -444,14 +470,15 @@ static void vgic_mmio_write_pendbase(struct kvm_vcpu *vcpu,
 	}
 
 static const struct vgic_register_region vgic_v3_dist_registers[] = {
-	REGISTER_DESC_WITH_LENGTH(GICD_CTLR,
-		vgic_mmio_read_v3_misc, vgic_mmio_write_v3_misc, 16,
-		VGIC_ACCESS_32bit),
+	REGISTER_DESC_WITH_LENGTH_UACCESS(GICD_CTLR,
+		vgic_mmio_read_v3_misc, vgic_mmio_write_v3_misc,
+		NULL, vgic_mmio_uaccess_write_v3_misc,
+		16, VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH(GICD_STATUSR,
 		vgic_mmio_read_rao, vgic_mmio_write_wi, 4,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_BITS_PER_IRQ_SHARED(GICD_IGROUPR,
-		vgic_mmio_read_rao, vgic_mmio_write_wi, NULL, NULL, 1,
+		vgic_mmio_read_group, vgic_mmio_write_group, NULL, NULL, 1,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_BITS_PER_IRQ_SHARED(GICD_ISENABLER,
 		vgic_mmio_read_enable, vgic_mmio_write_senable, NULL, NULL, 1,
@@ -465,7 +492,7 @@ static const struct vgic_register_region vgic_v3_dist_registers[] = {
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_BITS_PER_IRQ_SHARED(GICD_ICPENDR,
 		vgic_mmio_read_pending, vgic_mmio_write_cpending,
-		vgic_mmio_read_raz, vgic_mmio_write_wi, 1,
+		vgic_mmio_read_raz, vgic_mmio_uaccess_write_wi, 1,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_BITS_PER_IRQ_SHARED(GICD_ISACTIVER,
 		vgic_mmio_read_active, vgic_mmio_write_sactive,
@@ -524,7 +551,7 @@ static const struct vgic_register_region vgic_v3_rdbase_registers[] = {
 
 static const struct vgic_register_region vgic_v3_sgibase_registers[] = {
 	REGISTER_DESC_WITH_LENGTH(GICR_IGROUPR0,
-		vgic_mmio_read_rao, vgic_mmio_write_wi, 4,
+		vgic_mmio_read_group, vgic_mmio_write_group, 4,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH(GICR_ISENABLER0,
 		vgic_mmio_read_enable, vgic_mmio_write_senable, 4,
@@ -538,7 +565,7 @@ static const struct vgic_register_region vgic_v3_sgibase_registers[] = {
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH_UACCESS(GICR_ICPENDR0,
 		vgic_mmio_read_pending, vgic_mmio_write_cpending,
-		vgic_mmio_read_raz, vgic_mmio_write_wi, 4,
+		vgic_mmio_read_raz, vgic_mmio_uaccess_write_wi, 4,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH_UACCESS(GICR_ISACTIVER0,
 		vgic_mmio_read_active, vgic_mmio_write_sactive,
@@ -873,7 +900,8 @@ static int match_mpidr(u64 sgi_aff, u16 sgi_cpu_mask, struct kvm_vcpu *vcpu)
 /**
  * vgic_v3_dispatch_sgi - handle SGI requests from VCPUs
  * @vcpu: The VCPU requesting a SGI
- * @reg: The value written into the ICC_SGI1R_EL1 register by that VCPU
+ * @reg: The value written into ICC_{ASGI1,SGI0,SGI1}R by that VCPU
+ * @allow_group1: Does the sysreg access allow generation of G1 SGIs
  *
  * With GICv3 (and ARE=1) CPUs trigger SGIs by writing to a system register.
  * This will trap in sys_regs.c and call this function.
@@ -883,7 +911,7 @@ static int match_mpidr(u64 sgi_aff, u16 sgi_cpu_mask, struct kvm_vcpu *vcpu)
  * check for matching ones. If this bit is set, we signal all, but not the
  * calling VCPU.
  */
-void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg)
+void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg, bool allow_group1)
 {
 	struct kvm *kvm = vcpu->kvm;
 	struct kvm_vcpu *c_vcpu;
@@ -932,9 +960,19 @@ void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg)
 		irq = vgic_get_irq(vcpu->kvm, c_vcpu, sgi);
 
 		spin_lock_irqsave(&irq->irq_lock, flags);
-		irq->pending_latch = true;
 
-		vgic_queue_irq_unlock(vcpu->kvm, irq, flags);
+		/*
+		 * An access targetting Group0 SGIs can only generate
+		 * those, while an access targetting Group1 SGIs can
+		 * generate interrupts of either group.
+		 */
+		if (!irq->group || allow_group1) {
+			irq->pending_latch = true;
+			vgic_queue_irq_unlock(vcpu->kvm, irq, flags);
+		} else {
+			spin_unlock_irqrestore(&irq->irq_lock, flags);
+		}
+
 		vgic_put_irq(vcpu->kvm, irq);
 	}
 }

virt/kvm/arm/vgic/vgic-mmio.c

@@ -40,6 +40,51 @@ void vgic_mmio_write_wi(struct kvm_vcpu *vcpu, gpa_t addr,
 	/* Ignore */
 }
 
+int vgic_mmio_uaccess_write_wi(struct kvm_vcpu *vcpu, gpa_t addr,
+			       unsigned int len, unsigned long val)
+{
+	/* Ignore */
+	return 0;
+}
+
+unsigned long vgic_mmio_read_group(struct kvm_vcpu *vcpu,
+				   gpa_t addr, unsigned int len)
+{
+	u32 intid = VGIC_ADDR_TO_INTID(addr, 1);
+	u32 value = 0;
+	int i;
+
+	/* Loop over all IRQs affected by this read */
+	for (i = 0; i < len * 8; i++) {
+		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
+
+		if (irq->group)
+			value |= BIT(i);
+
+		vgic_put_irq(vcpu->kvm, irq);
+	}
+
+	return value;
+}
+
+void vgic_mmio_write_group(struct kvm_vcpu *vcpu, gpa_t addr,
+			   unsigned int len, unsigned long val)
+{
+	u32 intid = VGIC_ADDR_TO_INTID(addr, 1);
+	int i;
+	unsigned long flags;
+
+	for (i = 0; i < len * 8; i++) {
+		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
+
+		spin_lock_irqsave(&irq->irq_lock, flags);
+		irq->group = !!(val & BIT(i));
+		vgic_queue_irq_unlock(vcpu->kvm, irq, flags);
+
+		vgic_put_irq(vcpu->kvm, irq);
+	}
+}
+
 /*
  * Read accesses to both GICD_ICENABLER and GICD_ISENABLER return the value
  * of the enabled bit, so there is only one function for both here.
@@ -363,11 +408,12 @@ void vgic_mmio_write_cactive(struct kvm_vcpu *vcpu,
 	mutex_unlock(&vcpu->kvm->lock);
 }
 
-void vgic_mmio_uaccess_write_cactive(struct kvm_vcpu *vcpu,
+int vgic_mmio_uaccess_write_cactive(struct kvm_vcpu *vcpu,
 				     gpa_t addr, unsigned int len,
 				     unsigned long val)
 {
 	__vgic_mmio_write_cactive(vcpu, addr, len, val);
+	return 0;
 }
 
 static void __vgic_mmio_write_sactive(struct kvm_vcpu *vcpu,
@@ -399,11 +445,12 @@ void vgic_mmio_write_sactive(struct kvm_vcpu *vcpu,
 	mutex_unlock(&vcpu->kvm->lock);
 }
 
-void vgic_mmio_uaccess_write_sactive(struct kvm_vcpu *vcpu,
+int vgic_mmio_uaccess_write_sactive(struct kvm_vcpu *vcpu,
 				     gpa_t addr, unsigned int len,
 				     unsigned long val)
 {
 	__vgic_mmio_write_sactive(vcpu, addr, len, val);
+	return 0;
 }
 
 unsigned long vgic_mmio_read_priority(struct kvm_vcpu *vcpu,
@@ -735,10 +782,9 @@ static int vgic_uaccess_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
 
 	r_vcpu = iodev->redist_vcpu ? iodev->redist_vcpu : vcpu;
 	if (region->uaccess_write)
-		region->uaccess_write(r_vcpu, addr, sizeof(u32), *val);
-	else
-		region->write(r_vcpu, addr, sizeof(u32), *val);
+		return region->uaccess_write(r_vcpu, addr, sizeof(u32), *val);
 
+	region->write(r_vcpu, addr, sizeof(u32), *val);
 	return 0;
 }
 

virt/kvm/arm/vgic/vgic-mmio.h

@@ -37,8 +37,8 @@ struct vgic_register_region {
 	unsigned long (*uaccess_read)(struct kvm_vcpu *vcpu, gpa_t addr,
 				      unsigned int len);
 	union {
-		void (*uaccess_write)(struct kvm_vcpu *vcpu, gpa_t addr,
-				      unsigned int len, unsigned long val);
+		int (*uaccess_write)(struct kvm_vcpu *vcpu, gpa_t addr,
+				     unsigned int len, unsigned long val);
 		int (*uaccess_its_write)(struct kvm *kvm, struct vgic_its *its,
 					 gpa_t addr, unsigned int len,
 					 unsigned long val);
@@ -134,6 +134,15 @@ unsigned long vgic_mmio_read_rao(struct kvm_vcpu *vcpu,
 void vgic_mmio_write_wi(struct kvm_vcpu *vcpu, gpa_t addr,
 			unsigned int len, unsigned long val);
 
+int vgic_mmio_uaccess_write_wi(struct kvm_vcpu *vcpu, gpa_t addr,
+			       unsigned int len, unsigned long val);
+
+unsigned long vgic_mmio_read_group(struct kvm_vcpu *vcpu, gpa_t addr,
+				   unsigned int len);
+
+void vgic_mmio_write_group(struct kvm_vcpu *vcpu, gpa_t addr,
+			   unsigned int len, unsigned long val);
+
 unsigned long vgic_mmio_read_enable(struct kvm_vcpu *vcpu,
 				    gpa_t addr, unsigned int len);
 
@@ -167,13 +176,13 @@ void vgic_mmio_write_sactive(struct kvm_vcpu *vcpu,
 			     gpa_t addr, unsigned int len,
 			     unsigned long val);
 
-void vgic_mmio_uaccess_write_cactive(struct kvm_vcpu *vcpu,
-				     gpa_t addr, unsigned int len,
-				     unsigned long val);
+int vgic_mmio_uaccess_write_cactive(struct kvm_vcpu *vcpu,
+				    gpa_t addr, unsigned int len,
+				    unsigned long val);
 
-void vgic_mmio_uaccess_write_sactive(struct kvm_vcpu *vcpu,
-				     gpa_t addr, unsigned int len,
-				     unsigned long val);
+int vgic_mmio_uaccess_write_sactive(struct kvm_vcpu *vcpu,
+				    gpa_t addr, unsigned int len,
+				    unsigned long val);
 
 unsigned long vgic_mmio_read_priority(struct kvm_vcpu *vcpu,
 				      gpa_t addr, unsigned int len);

virt/kvm/arm/vgic/vgic-v2.c

@@ -62,7 +62,8 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
 	struct vgic_v2_cpu_if *cpuif = &vgic_cpu->vgic_v2;
 	int lr;
-	unsigned long flags;
+
+	DEBUG_SPINLOCK_BUG_ON(!irqs_disabled());
 
 	cpuif->vgic_hcr &= ~GICH_HCR_UIE;
 
@@ -83,7 +84,7 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 
 		irq = vgic_get_irq(vcpu->kvm, vcpu, intid);
 
-		spin_lock_irqsave(&irq->irq_lock, flags);
+		spin_lock(&irq->irq_lock);
 
 		/* Always preserve the active bit */
 		irq->active = !!(val & GICH_LR_ACTIVE_BIT);
@@ -126,7 +127,7 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 				vgic_irq_set_phys_active(irq, false);
 		}
 
-		spin_unlock_irqrestore(&irq->irq_lock, flags);
+		spin_unlock(&irq->irq_lock);
 		vgic_put_irq(vcpu->kvm, irq);
 	}
 
@@ -159,6 +160,9 @@ void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 		}
 	}
 
+	if (irq->group)
+		val |= GICH_LR_GROUP1;
+
 	if (irq->hw) {
 		val |= GICH_LR_HW;
 		val |= irq->hwintid << GICH_LR_PHYSID_CPUID_SHIFT;

virt/kvm/arm/vgic/vgic-v3.c

@@ -46,7 +46,8 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 	struct vgic_v3_cpu_if *cpuif = &vgic_cpu->vgic_v3;
 	u32 model = vcpu->kvm->arch.vgic.vgic_model;
 	int lr;
-	unsigned long flags;
+
+	DEBUG_SPINLOCK_BUG_ON(!irqs_disabled());
 
 	cpuif->vgic_hcr &= ~ICH_HCR_UIE;
 
@@ -75,7 +76,7 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 		if (!irq)	/* An LPI could have been unmapped. */
 			continue;
 
-		spin_lock_irqsave(&irq->irq_lock, flags);
+		spin_lock(&irq->irq_lock);
 
 		/* Always preserve the active bit */
 		irq->active = !!(val & ICH_LR_ACTIVE_BIT);
@@ -118,7 +119,7 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 				vgic_irq_set_phys_active(irq, false);
 		}
 
-		spin_unlock_irqrestore(&irq->irq_lock, flags);
+		spin_unlock(&irq->irq_lock);
 		vgic_put_irq(vcpu->kvm, irq);
 	}
 
@@ -197,11 +198,7 @@ void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 	if (vgic_irq_is_mapped_level(irq) && (val & ICH_LR_PENDING_BIT))
 		irq->line_level = false;
 
-	/*
-	 * We currently only support Group1 interrupts, which is a
-	 * known defect. This needs to be addressed at some point.
-	 */
-	if (model == KVM_DEV_TYPE_ARM_VGIC_V3)
+	if (irq->group)
 		val |= ICH_LR_GROUP;
 
 	val |= (u64)irq->priority << ICH_LR_PRIORITY_SHIFT;

virt/kvm/arm/vgic/vgic.c

@@ -28,12 +28,6 @@
 #define CREATE_TRACE_POINTS
 #include "trace.h"
 
-#ifdef CONFIG_DEBUG_SPINLOCK
-#define DEBUG_SPINLOCK_BUG_ON(p) BUG_ON(p)
-#else
-#define DEBUG_SPINLOCK_BUG_ON(p)
-#endif
-
 struct vgic_global kvm_vgic_global_state __ro_after_init = {
 	.gicv3_cpuif = STATIC_KEY_FALSE_INIT,
 };
@@ -599,10 +593,11 @@ static void vgic_prune_ap_list(struct kvm_vcpu *vcpu)
 {
 	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
 	struct vgic_irq *irq, *tmp;
-	unsigned long flags;
+
+	DEBUG_SPINLOCK_BUG_ON(!irqs_disabled());
 
 retry:
-	spin_lock_irqsave(&vgic_cpu->ap_list_lock, flags);
+	spin_lock(&vgic_cpu->ap_list_lock);
 
 	list_for_each_entry_safe(irq, tmp, &vgic_cpu->ap_list_head, ap_list) {
 		struct kvm_vcpu *target_vcpu, *vcpuA, *vcpuB;
@@ -643,7 +638,7 @@ retry:
 		/* This interrupt looks like it has to be migrated. */
 
 		spin_unlock(&irq->irq_lock);
-		spin_unlock_irqrestore(&vgic_cpu->ap_list_lock, flags);
+		spin_unlock(&vgic_cpu->ap_list_lock);
 
 		/*
 		 * Ensure locking order by always locking the smallest
@@ -657,7 +652,7 @@ retry:
 			vcpuB = vcpu;
 		}
 
-		spin_lock_irqsave(&vcpuA->arch.vgic_cpu.ap_list_lock, flags);
+		spin_lock(&vcpuA->arch.vgic_cpu.ap_list_lock);
 		spin_lock_nested(&vcpuB->arch.vgic_cpu.ap_list_lock,
 				 SINGLE_DEPTH_NESTING);
 		spin_lock(&irq->irq_lock);
@@ -682,7 +677,7 @@ retry:
 
 		spin_unlock(&irq->irq_lock);
 		spin_unlock(&vcpuB->arch.vgic_cpu.ap_list_lock);
-		spin_unlock_irqrestore(&vcpuA->arch.vgic_cpu.ap_list_lock, flags);
+		spin_unlock(&vcpuA->arch.vgic_cpu.ap_list_lock);
 
 		if (target_vcpu_needs_kick) {
 			kvm_make_request(KVM_REQ_IRQ_PENDING, target_vcpu);
@@ -692,7 +687,7 @@ retry:
 		goto retry;
 	}
 
-	spin_unlock_irqrestore(&vgic_cpu->ap_list_lock, flags);
+	spin_unlock(&vgic_cpu->ap_list_lock);
 }
 
 static inline void vgic_fold_lr_state(struct kvm_vcpu *vcpu)

virt/kvm/arm/vgic/vgic.h

@@ -103,6 +103,12 @@
 #define KVM_VGIC_V3_RDIST_COUNT_MASK	GENMASK_ULL(63, 52)
 #define KVM_VGIC_V3_RDIST_COUNT_SHIFT	52
 
+#ifdef CONFIG_DEBUG_SPINLOCK
+#define DEBUG_SPINLOCK_BUG_ON(p) BUG_ON(p)
+#else
+#define DEBUG_SPINLOCK_BUG_ON(p)
+#endif
+
 /* Requires the irq_lock to be held by the caller. */
 static inline bool irq_is_pending(struct vgic_irq *irq)
 {
@@ -305,6 +311,7 @@ static inline bool vgic_dist_overlap(struct kvm *kvm, gpa_t base, size_t size)
 		(base < d->vgic_dist_base + KVM_VGIC_V3_DIST_SIZE);
 }
 
+int vgic_copy_lpi_list(struct kvm *kvm, struct kvm_vcpu *vcpu, u32 **intid_ptr);
 int vgic_its_resolve_lpi(struct kvm *kvm, struct vgic_its *its,
 			 u32 devid, u32 eventid, struct vgic_irq **irq);
 struct vgic_its *vgic_msi_to_its(struct kvm *kvm, struct kvm_msi *msi);