Merge tag 'kvm-arm-for-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into next

KVM/ARM changes for Linux 4.8

- GICv3 ITS emulation
- Simpler idmap management that fixes potential TLB conflicts
- Honor the kernel protection in HYP mode
- Removal of the old vgic implementation

Documentation/virtual/kvm/api.txt

@@ -2182,7 +2182,7 @@ after pausing the vcpu, but before it is resumed.
 4.71 KVM_SIGNAL_MSI
 
 Capability: KVM_CAP_SIGNAL_MSI
-Architectures: x86
+Architectures: x86 arm64
 Type: vm ioctl
 Parameters: struct kvm_msi (in)
 Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
@@ -2195,10 +2195,18 @@ struct kvm_msi {
 	__u32 address_hi;
 	__u32 data;
 	__u32 flags;
-	__u8  pad[16];
+	__u32 devid;
+	__u8  pad[12];
 };
 
-No flags are defined so far. The corresponding field must be 0.
+flags: KVM_MSI_VALID_DEVID: devid contains a valid value
+devid: If KVM_MSI_VALID_DEVID is set, contains a unique device identifier
+       for the device that wrote the MSI message.
+       For PCI, this is usually a BFD identifier in the lower 16 bits.
+
+The per-VM KVM_CAP_MSI_DEVID capability advertises the need to provide
+the device ID. If this capability is not set, userland cannot rely on
+the kernel to allow the KVM_MSI_VALID_DEVID flag being set.
 
 On x86, address_hi is ignored unless the KVM_CAP_X2APIC_API capability is
 enabled.  If it is enabled, address_hi bits 31-8 provide bits 31-8 of the

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

@@ -4,16 +4,22 @@ ARM Virtual Generic Interrupt Controller (VGIC)
 Device types supported:
   KVM_DEV_TYPE_ARM_VGIC_V2     ARM Generic Interrupt Controller v2.0
   KVM_DEV_TYPE_ARM_VGIC_V3     ARM Generic Interrupt Controller v3.0
+  KVM_DEV_TYPE_ARM_VGIC_ITS    ARM Interrupt Translation Service Controller
 
-Only one VGIC instance may be instantiated through either this API or the
-legacy KVM_CREATE_IRQCHIP api.  The created VGIC will act as the VM interrupt
-controller, requiring emulated user-space devices to inject interrupts to the
-VGIC instead of directly to CPUs.
+Only one VGIC instance of the V2/V3 types above may be instantiated through
+either this API or the legacy KVM_CREATE_IRQCHIP api.  The created VGIC will
+act as the VM interrupt controller, requiring emulated user-space devices to
+inject interrupts to the VGIC instead of directly to CPUs.
 
 Creating a guest GICv3 device requires a host GICv3 as well.
 GICv3 implementations with hardware compatibility support allow a guest GICv2
 as well.
 
+Creating a virtual ITS controller requires a host GICv3 (but does not depend
+on having physical ITS controllers).
+There can be multiple ITS controllers per guest, each of them has to have
+a separate, non-overlapping MMIO region.
+
 Groups:
   KVM_DEV_ARM_VGIC_GRP_ADDR
   Attributes:
@@ -39,6 +45,13 @@ Groups:
       Only valid for KVM_DEV_TYPE_ARM_VGIC_V3.
       This address needs to be 64K aligned.
 
+    KVM_VGIC_V3_ADDR_TYPE_ITS (rw, 64-bit)
+      Base address in the guest physical address space of the GICv3 ITS
+      control register frame. The ITS allows MSI(-X) interrupts to be
+      injected into guests. This extension is optional. If the kernel
+      does not support the ITS, the call returns -ENODEV.
+      Only valid for KVM_DEV_TYPE_ARM_VGIC_ITS.
+      This address needs to be 64K aligned and the region covers 128K.
 
   KVM_DEV_ARM_VGIC_GRP_DIST_REGS
   Attributes:
@@ -109,8 +122,8 @@ Groups:
   KVM_DEV_ARM_VGIC_GRP_CTRL
   Attributes:
     KVM_DEV_ARM_VGIC_CTRL_INIT
-      request the initialization of the VGIC, no additional parameter in
-      kvm_device_attr.addr.
+      request the initialization of the VGIC or ITS, no additional parameter
+      in kvm_device_attr.addr.
   Errors:
     -ENXIO: VGIC not properly configured as required prior to calling
      this attribute

arch/arm/include/asm/kvm_asm.h

@@ -66,6 +66,8 @@ extern void __kvm_tlb_flush_vmid(struct kvm *kvm);
 extern int __kvm_vcpu_run(struct kvm_vcpu *vcpu);
 
 extern void __init_stage2_translation(void);
+
+extern void __kvm_hyp_reset(unsigned long);
 #endif
 
 #endif /* __ARM_KVM_ASM_H__ */

arch/arm/include/asm/kvm_host.h

@@ -241,8 +241,7 @@ int kvm_arm_coproc_set_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *);
 int handle_exit(struct kvm_vcpu *vcpu, struct kvm_run *run,
 		int exception_index);
 
-static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
-				       phys_addr_t pgd_ptr,
+static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
 				       unsigned long hyp_stack_ptr,
 				       unsigned long vector_ptr)
 {
@@ -251,18 +250,13 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
 	 * code. The init code doesn't need to preserve these
 	 * registers as r0-r3 are already callee saved according to
 	 * the AAPCS.
-	 * Note that we slightly misuse the prototype by casing the
+	 * Note that we slightly misuse the prototype by casting the
 	 * stack pointer to a void *.
-	 *
-	 * We don't have enough registers to perform the full init in
-	 * one go.  Install the boot PGD first, and then install the
-	 * runtime PGD, stack pointer and vectors. The PGDs are always
-	 * passed as the third argument, in order to be passed into
-	 * r2-r3 to the init code (yes, this is compliant with the
-	 * PCS!).
-	 */
 
-	kvm_call_hyp(NULL, 0, boot_pgd_ptr);
+	 * The PGDs are always passed as the third argument, in order
+	 * to be passed into r2-r3 to the init code (yes, this is
+	 * compliant with the PCS!).
+	 */
 
 	kvm_call_hyp((void*)hyp_stack_ptr, vector_ptr, pgd_ptr);
 }
@@ -272,16 +266,13 @@ static inline void __cpu_init_stage2(void)
 	kvm_call_hyp(__init_stage2_translation);
 }
 
-static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr,
+static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr,
 					phys_addr_t phys_idmap_start)
 {
-	/*
-	 * TODO
-	 * kvm_call_reset(boot_pgd_ptr, phys_idmap_start);
-	 */
+	kvm_call_hyp((void *)virt_to_idmap(__kvm_hyp_reset), vector_ptr);
 }
 
-static inline int kvm_arch_dev_ioctl_check_extension(long ext)
+static inline int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
 {
 	return 0;
 }

arch/arm/include/asm/kvm_hyp.h

@@ -25,9 +25,6 @@
 
 #define __hyp_text __section(.hyp.text) notrace
 
-#define kern_hyp_va(v) (v)
-#define hyp_kern_va(v) (v)
-
 #define __ACCESS_CP15(CRn, Op1, CRm, Op2)	\
 	"mrc", "mcr", __stringify(p15, Op1, %0, CRn, CRm, Op2), u32
 #define __ACCESS_CP15_64(Op1, CRm)		\

arch/arm/include/asm/kvm_mmu.h

@@ -26,16 +26,7 @@
  * We directly use the kernel VA for the HYP, as we can directly share
  * the mapping (HTTBR "covers" TTBR1).
  */
-#define HYP_PAGE_OFFSET_MASK	UL(~0)
-#define HYP_PAGE_OFFSET		PAGE_OFFSET
-#define KERN_TO_HYP(kva)	(kva)
-
-/*
- * Our virtual mapping for the boot-time MMU-enable code. Must be
- * shared across all the page-tables. Conveniently, we use the vectors
- * page, where no kernel data will ever be shared with HYP.
- */
-#define TRAMPOLINE_VA		UL(CONFIG_VECTORS_BASE)
+#define kern_hyp_va(kva)	(kva)
 
 /*
  * KVM_MMU_CACHE_MIN_PAGES is the number of stage2 page table translation levels.
@@ -49,9 +40,8 @@
 #include <asm/pgalloc.h>
 #include <asm/stage2_pgtable.h>
 
-int create_hyp_mappings(void *from, void *to);
+int create_hyp_mappings(void *from, void *to, pgprot_t prot);
 int create_hyp_io_mappings(void *from, void *to, phys_addr_t);
-void free_boot_hyp_pgd(void);
 void free_hyp_pgds(void);
 
 void stage2_unmap_vm(struct kvm *kvm);
@@ -65,7 +55,6 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run);
 void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu);
 
 phys_addr_t kvm_mmu_get_httbr(void);
-phys_addr_t kvm_mmu_get_boot_httbr(void);
 phys_addr_t kvm_get_idmap_vector(void);
 phys_addr_t kvm_get_idmap_start(void);
 int kvm_mmu_init(void);

arch/arm/include/asm/pgtable.h

@@ -97,7 +97,9 @@ extern pgprot_t		pgprot_s2_device;
 #define PAGE_READONLY_EXEC	_MOD_PROT(pgprot_user, L_PTE_USER | L_PTE_RDONLY)
 #define PAGE_KERNEL		_MOD_PROT(pgprot_kernel, L_PTE_XN)
 #define PAGE_KERNEL_EXEC	pgprot_kernel
-#define PAGE_HYP		_MOD_PROT(pgprot_kernel, L_PTE_HYP)
+#define PAGE_HYP		_MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_XN)
+#define PAGE_HYP_EXEC		_MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY)
+#define PAGE_HYP_RO		_MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY | L_PTE_XN)
 #define PAGE_HYP_DEVICE		_MOD_PROT(pgprot_hyp_device, L_PTE_HYP)
 #define PAGE_S2			_MOD_PROT(pgprot_s2, L_PTE_S2_RDONLY)
 #define PAGE_S2_DEVICE		_MOD_PROT(pgprot_s2_device, L_PTE_S2_RDONLY)

arch/arm/include/asm/virt.h

@@ -80,6 +80,10 @@ static inline bool is_kernel_in_hyp_mode(void)
 	return false;
 }
 
+/* The section containing the hypervisor idmap text */
+extern char __hyp_idmap_text_start[];
+extern char __hyp_idmap_text_end[];
+
 /* The section containing the hypervisor text */
 extern char __hyp_text_start[];
 extern char __hyp_text_end[];

arch/arm/kvm/Kconfig

@@ -46,13 +46,6 @@ config KVM_ARM_HOST
 	---help---
 	  Provides host support for ARM processors.
 
-config KVM_NEW_VGIC
-	bool "New VGIC implementation"
-	depends on KVM
-	default y
-	---help---
-	  uses the new VGIC implementation
-
 source drivers/vhost/Kconfig
 
 endif # VIRTUALIZATION

arch/arm/kvm/Makefile

@@ -22,7 +22,6 @@ obj-y += kvm-arm.o init.o interrupts.o
 obj-y += arm.o handle_exit.o guest.o mmu.o emulate.o reset.o
 obj-y += coproc.o coproc_a15.o coproc_a7.o mmio.o psci.o perf.o
 
-ifeq ($(CONFIG_KVM_NEW_VGIC),y)
 obj-y += $(KVM)/arm/vgic/vgic.o
 obj-y += $(KVM)/arm/vgic/vgic-init.o
 obj-y += $(KVM)/arm/vgic/vgic-irqfd.o
@@ -30,9 +29,4 @@ obj-y += $(KVM)/arm/vgic/vgic-v2.o
 obj-y += $(KVM)/arm/vgic/vgic-mmio.o
 obj-y += $(KVM)/arm/vgic/vgic-mmio-v2.o
 obj-y += $(KVM)/arm/vgic/vgic-kvm-device.o
-else
-obj-y += $(KVM)/arm/vgic.o
-obj-y += $(KVM)/arm/vgic-v2.o
-obj-y += $(KVM)/arm/vgic-v2-emul.o
-endif
 obj-y += $(KVM)/arm/arch_timer.o

arch/arm/kvm/arm.c

@@ -20,6 +20,7 @@
 #include <linux/errno.h>
 #include <linux/err.h>
 #include <linux/kvm_host.h>
+#include <linux/list.h>
 #include <linux/module.h>
 #include <linux/vmalloc.h>
 #include <linux/fs.h>
@@ -122,7 +123,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	if (ret)
 		goto out_fail_alloc;
 
-	ret = create_hyp_mappings(kvm, kvm + 1);
+	ret = create_hyp_mappings(kvm, kvm + 1, PAGE_HYP);
 	if (ret)
 		goto out_free_stage2_pgd;
 
@@ -201,7 +202,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 		r = KVM_MAX_VCPUS;
 		break;
 	default:
-		r = kvm_arch_dev_ioctl_check_extension(ext);
+		r = kvm_arch_dev_ioctl_check_extension(kvm, ext);
 		break;
 	}
 	return r;
@@ -239,7 +240,7 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
 	if (err)
 		goto free_vcpu;
 
-	err = create_hyp_mappings(vcpu, vcpu + 1);
+	err = create_hyp_mappings(vcpu, vcpu + 1, PAGE_HYP);
 	if (err)
 		goto vcpu_uninit;
 
@@ -1038,7 +1039,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 
 static void cpu_init_hyp_mode(void *dummy)
 {
-	phys_addr_t boot_pgd_ptr;
 	phys_addr_t pgd_ptr;
 	unsigned long hyp_stack_ptr;
 	unsigned long stack_page;
@@ -1047,13 +1047,12 @@ static void cpu_init_hyp_mode(void *dummy)
 	/* Switch from the HYP stub to our own HYP init vector */
 	__hyp_set_vectors(kvm_get_idmap_vector());
 
-	boot_pgd_ptr = kvm_mmu_get_boot_httbr();
 	pgd_ptr = kvm_mmu_get_httbr();
 	stack_page = __this_cpu_read(kvm_arm_hyp_stack_page);
 	hyp_stack_ptr = stack_page + PAGE_SIZE;
 	vector_ptr = (unsigned long)kvm_ksym_ref(__kvm_hyp_vector);
 
-	__cpu_init_hyp_mode(boot_pgd_ptr, pgd_ptr, hyp_stack_ptr, vector_ptr);
+	__cpu_init_hyp_mode(pgd_ptr, hyp_stack_ptr, vector_ptr);
 	__cpu_init_stage2();
 
 	kvm_arm_init_debug();
@@ -1075,15 +1074,9 @@ static void cpu_hyp_reinit(void)
 
 static void cpu_hyp_reset(void)
 {
-	phys_addr_t boot_pgd_ptr;
-	phys_addr_t phys_idmap_start;
-
-	if (!is_kernel_in_hyp_mode()) {
-		boot_pgd_ptr = kvm_mmu_get_boot_httbr();
-		phys_idmap_start = kvm_get_idmap_start();
-
-		__cpu_reset_hyp_mode(boot_pgd_ptr, phys_idmap_start);
-	}
+	if (!is_kernel_in_hyp_mode())
+		__cpu_reset_hyp_mode(hyp_default_vectors,
+				     kvm_get_idmap_start());
 }
 
 static void _kvm_arch_hardware_enable(void *discard)
@@ -1293,14 +1286,14 @@ static int init_hyp_mode(void)
 	 * Map the Hyp-code called directly from the host
 	 */
 	err = create_hyp_mappings(kvm_ksym_ref(__hyp_text_start),
-				  kvm_ksym_ref(__hyp_text_end));
+				  kvm_ksym_ref(__hyp_text_end), PAGE_HYP_EXEC);
 	if (err) {
 		kvm_err("Cannot map world-switch code\n");
 		goto out_err;
 	}
 
 	err = create_hyp_mappings(kvm_ksym_ref(__start_rodata),
-				  kvm_ksym_ref(__end_rodata));
+				  kvm_ksym_ref(__end_rodata), PAGE_HYP_RO);
 	if (err) {
 		kvm_err("Cannot map rodata section\n");
 		goto out_err;
@@ -1311,7 +1304,8 @@ static int init_hyp_mode(void)
 	 */
 	for_each_possible_cpu(cpu) {
 		char *stack_page = (char *)per_cpu(kvm_arm_hyp_stack_page, cpu);
-		err = create_hyp_mappings(stack_page, stack_page + PAGE_SIZE);
+		err = create_hyp_mappings(stack_page, stack_page + PAGE_SIZE,
+					  PAGE_HYP);
 
 		if (err) {
 			kvm_err("Cannot map hyp stack\n");
@@ -1323,7 +1317,7 @@ static int init_hyp_mode(void)
 		kvm_cpu_context_t *cpu_ctxt;
 
 		cpu_ctxt = per_cpu_ptr(kvm_host_cpu_state, cpu);
-		err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1);
+		err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1, PAGE_HYP);
 
 		if (err) {
 			kvm_err("Cannot map host CPU state: %d\n", err);
@@ -1331,10 +1325,6 @@ static int init_hyp_mode(void)
 		}
 	}
 
-#ifndef CONFIG_HOTPLUG_CPU
-	free_boot_hyp_pgd();
-#endif
-
 	/* set size of VMID supported by CPU */
 	kvm_vmid_bits = kvm_get_vmid_bits();
 	kvm_info("%d-bit VMID\n", kvm_vmid_bits);

arch/arm/kvm/init.S

@@ -32,23 +32,13 @@
  *       r2,r3 = Hypervisor pgd pointer
  *
  * The init scenario is:
- * - We jump in HYP with four parameters: boot HYP pgd, runtime HYP pgd,
- *   runtime stack, runtime vectors
- * - Enable the MMU with the boot pgd
- * - Jump to a target into the trampoline page (remember, this is the same
- *   physical page!)
- * - Now switch to the runtime pgd (same VA, and still the same physical
- *   page!)
+ * - We jump in HYP with 3 parameters: runtime HYP pgd, runtime stack,
+ *   runtime vectors
  * - Invalidate TLBs
  * - Set stack and vectors
+ * - Setup the page tables
+ * - Enable the MMU
  * - Profit! (or eret, if you only care about the code).
- *
- * As we only have four registers available to pass parameters (and we
- * need six), we split the init in two phases:
- * - Phase 1: r0 = 0, r1 = 0, r2,r3 contain the boot PGD.
- *   Provides the basic HYP init, and enable the MMU.
- * - Phase 2: r0 = ToS, r1 = vectors, r2,r3 contain the runtime PGD.
- *   Switches to the runtime PGD, set stack and vectors.
  */
 
 	.text
@@ -68,8 +58,11 @@ __kvm_hyp_init:
 	W(b)	.
 
 __do_hyp_init:
-	cmp	r0, #0			@ We have a SP?
-	bne	phase2			@ Yes, second stage init
+	@ Set stack pointer
+	mov	sp, r0
+
+	@ Set HVBAR to point to the HYP vectors
+	mcr	p15, 4, r1, c12, c0, 0	@ HVBAR
 
 	@ Set the HTTBR to point to the hypervisor PGD pointer passed
 	mcrr	p15, 4, rr_lo_hi(r2, r3), c2
@@ -114,34 +107,25 @@ __do_hyp_init:
  THUMB(	ldr	r2, =(HSCTLR_M | HSCTLR_A | HSCTLR_TE)		)
 	orr	r1, r1, r2
 	orr	r0, r0, r1
-	isb
 	mcr	p15, 4, r0, c1, c0, 0	@ HSCR
+	isb
 
-	@ End of init phase-1
 	eret
 
-phase2:
-	@ Set stack pointer
-	mov	sp, r0
-
-	@ Set HVBAR to point to the HYP vectors
-	mcr	p15, 4, r1, c12, c0, 0	@ HVBAR
-
-	@ Jump to the trampoline page
-	ldr	r0, =TRAMPOLINE_VA
-	adr	r1, target
-	bfi	r0, r1, #0, #PAGE_SHIFT
-	ret	r0
+	@ r0 : stub vectors address
+ENTRY(__kvm_hyp_reset)
+	/* We're now in idmap, disable MMU */
+	mrc	p15, 4, r1, c1, c0, 0	@ HSCTLR
+	ldr	r2, =(HSCTLR_M | HSCTLR_A | HSCTLR_C | HSCTLR_I)
+	bic	r1, r1, r2
+	mcr	p15, 4, r1, c1, c0, 0	@ HSCTLR
 
-target:	@ We're now in the trampoline code, switch page tables
-	mcrr	p15, 4, rr_lo_hi(r2, r3), c2
+	/* Install stub vectors */
+	mcr	p15, 4, r0, c12, c0, 0	@ HVBAR
 	isb
 
-	@ Invalidate the old TLBs
-	mcr	p15, 4, r0, c8, c7, 0	@ TLBIALLH
-	dsb	ish
-
 	eret
+ENDPROC(__kvm_hyp_reset)
 
 	.ltorg
 

arch/arm/kvm/mmu.c

@@ -32,8 +32,6 @@
 
 #include "trace.h"
 
-extern char  __hyp_idmap_text_start[], __hyp_idmap_text_end[];
-
 static pgd_t *boot_hyp_pgd;
 static pgd_t *hyp_pgd;
 static pgd_t *merged_hyp_pgd;
@@ -483,28 +481,6 @@ static void unmap_hyp_range(pgd_t *pgdp, phys_addr_t start, u64 size)
 	} while (pgd++, addr = next, addr != end);
 }
 
-/**
- * free_boot_hyp_pgd - free HYP boot page tables
- *
- * Free the HYP boot page tables. The bounce page is also freed.
- */
-void free_boot_hyp_pgd(void)
-{
-	mutex_lock(&kvm_hyp_pgd_mutex);
-
-	if (boot_hyp_pgd) {
-		unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE);
-		unmap_hyp_range(boot_hyp_pgd, TRAMPOLINE_VA, PAGE_SIZE);
-		free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order);
-		boot_hyp_pgd = NULL;
-	}
-
-	if (hyp_pgd)
-		unmap_hyp_range(hyp_pgd, TRAMPOLINE_VA, PAGE_SIZE);
-
-	mutex_unlock(&kvm_hyp_pgd_mutex);
-}
-
 /**
  * free_hyp_pgds - free Hyp-mode page tables
  *
@@ -519,15 +495,20 @@ void free_hyp_pgds(void)
 {
 	unsigned long addr;
 
-	free_boot_hyp_pgd();
-
 	mutex_lock(&kvm_hyp_pgd_mutex);
 
+	if (boot_hyp_pgd) {
+		unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE);
+		free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order);
+		boot_hyp_pgd = NULL;
+	}
+
 	if (hyp_pgd) {
+		unmap_hyp_range(hyp_pgd, hyp_idmap_start, PAGE_SIZE);
 		for (addr = PAGE_OFFSET; virt_addr_valid(addr); addr += PGDIR_SIZE)
-			unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE);
+			unmap_hyp_range(hyp_pgd, kern_hyp_va(addr), PGDIR_SIZE);
 		for (addr = VMALLOC_START; is_vmalloc_addr((void*)addr); addr += PGDIR_SIZE)
-			unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE);
+			unmap_hyp_range(hyp_pgd, kern_hyp_va(addr), PGDIR_SIZE);
 
 		free_pages((unsigned long)hyp_pgd, hyp_pgd_order);
 		hyp_pgd = NULL;
@@ -679,17 +660,18 @@ static phys_addr_t kvm_kaddr_to_phys(void *kaddr)
  * create_hyp_mappings - duplicate a kernel virtual address range in Hyp mode
  * @from:	The virtual kernel start address of the range
  * @to:		The virtual kernel end address of the range (exclusive)
+ * @prot:	The protection to be applied to this range
  *
  * The same virtual address as the kernel virtual address is also used
  * in Hyp-mode mapping (modulo HYP_PAGE_OFFSET) to the same underlying
  * physical pages.
  */
-int create_hyp_mappings(void *from, void *to)
+int create_hyp_mappings(void *from, void *to, pgprot_t prot)
 {
 	phys_addr_t phys_addr;
 	unsigned long virt_addr;
-	unsigned long start = KERN_TO_HYP((unsigned long)from);
-	unsigned long end = KERN_TO_HYP((unsigned long)to);
+	unsigned long start = kern_hyp_va((unsigned long)from);
+	unsigned long end = kern_hyp_va((unsigned long)to);
 
 	if (is_kernel_in_hyp_mode())
 		return 0;
@@ -704,7 +686,7 @@ int create_hyp_mappings(void *from, void *to)
 		err = __create_hyp_mappings(hyp_pgd, virt_addr,
 					    virt_addr + PAGE_SIZE,
 					    __phys_to_pfn(phys_addr),
-					    PAGE_HYP);
+					    prot);
 		if (err)
 			return err;
 	}
@@ -723,8 +705,8 @@ int create_hyp_mappings(void *from, void *to)
  */
 int create_hyp_io_mappings(void *from, void *to, phys_addr_t phys_addr)
 {
-	unsigned long start = KERN_TO_HYP((unsigned long)from);
-	unsigned long end = KERN_TO_HYP((unsigned long)to);
+	unsigned long start = kern_hyp_va((unsigned long)from);
+	unsigned long end = kern_hyp_va((unsigned long)to);
 
 	if (is_kernel_in_hyp_mode())
 		return 0;
@@ -1687,14 +1669,6 @@ phys_addr_t kvm_mmu_get_httbr(void)
 		return virt_to_phys(hyp_pgd);
 }
 
-phys_addr_t kvm_mmu_get_boot_httbr(void)
-{
-	if (__kvm_cpu_uses_extended_idmap())
-		return virt_to_phys(merged_hyp_pgd);
-	else
-		return virt_to_phys(boot_hyp_pgd);
-}
-
 phys_addr_t kvm_get_idmap_vector(void)
 {
 	return hyp_idmap_vector;
@@ -1705,6 +1679,22 @@ phys_addr_t kvm_get_idmap_start(void)
 	return hyp_idmap_start;
 }
 
+static int kvm_map_idmap_text(pgd_t *pgd)
+{
+	int err;
+
+	/* Create the idmap in the boot page tables */
+	err = 	__create_hyp_mappings(pgd,
+				      hyp_idmap_start, hyp_idmap_end,
+				      __phys_to_pfn(hyp_idmap_start),
+				      PAGE_HYP_EXEC);
+	if (err)
+		kvm_err("Failed to idmap %lx-%lx\n",
+			hyp_idmap_start, hyp_idmap_end);
+
+	return err;
+}
+
 int kvm_mmu_init(void)
 {
 	int err;
@@ -1719,28 +1709,41 @@ int kvm_mmu_init(void)
 	 */
 	BUG_ON((hyp_idmap_start ^ (hyp_idmap_end - 1)) & PAGE_MASK);
 
-	hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order);
-	boot_hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order);
+	kvm_info("IDMAP page: %lx\n", hyp_idmap_start);
+	kvm_info("HYP VA range: %lx:%lx\n",
+		 kern_hyp_va(PAGE_OFFSET), kern_hyp_va(~0UL));
 
-	if (!hyp_pgd || !boot_hyp_pgd) {
-		kvm_err("Hyp mode PGD not allocated\n");
-		err = -ENOMEM;
+	if (hyp_idmap_start >= kern_hyp_va(PAGE_OFFSET) &&
+	    hyp_idmap_start <  kern_hyp_va(~0UL)) {
+		/*
+		 * The idmap page is intersecting with the VA space,
+		 * it is not safe to continue further.
+		 */
+		kvm_err("IDMAP intersecting with HYP VA, unable to continue\n");
+		err = -EINVAL;
 		goto out;
 	}
 
-	/* Create the idmap in the boot page tables */
-	err = 	__create_hyp_mappings(boot_hyp_pgd,
-				      hyp_idmap_start, hyp_idmap_end,
-				      __phys_to_pfn(hyp_idmap_start),
-				      PAGE_HYP);
-
-	if (err) {
-		kvm_err("Failed to idmap %lx-%lx\n",
-			hyp_idmap_start, hyp_idmap_end);
+	hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order);
+	if (!hyp_pgd) {
+		kvm_err("Hyp mode PGD not allocated\n");
+		err = -ENOMEM;
 		goto out;
 	}
 
 	if (__kvm_cpu_uses_extended_idmap()) {
+		boot_hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
+							 hyp_pgd_order);
+		if (!boot_hyp_pgd) {
+			kvm_err("Hyp boot PGD not allocated\n");
+			err = -ENOMEM;
+			goto out;
+		}
+
+		err = kvm_map_idmap_text(boot_hyp_pgd);
+		if (err)
+			goto out;
+
 		merged_hyp_pgd = (pgd_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
 		if (!merged_hyp_pgd) {
 			kvm_err("Failed to allocate extra HYP pgd\n");
@@ -1748,29 +1751,10 @@ int kvm_mmu_init(void)
 		}
 		__kvm_extend_hypmap(boot_hyp_pgd, hyp_pgd, merged_hyp_pgd,
 				    hyp_idmap_start);
-		return 0;
-	}
-
-	/* Map the very same page at the trampoline VA */
-	err = 	__create_hyp_mappings(boot_hyp_pgd,
-				      TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE,
-				      __phys_to_pfn(hyp_idmap_start),
-				      PAGE_HYP);
-	if (err) {
-		kvm_err("Failed to map trampoline @%lx into boot HYP pgd\n",
-			TRAMPOLINE_VA);
-		goto out;
-	}
-
-	/* Map the same page again into the runtime page tables */
-	err = 	__create_hyp_mappings(hyp_pgd,
-				      TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE,
-				      __phys_to_pfn(hyp_idmap_start),
-				      PAGE_HYP);
-	if (err) {
-		kvm_err("Failed to map trampoline @%lx into runtime HYP pgd\n",
-			TRAMPOLINE_VA);
-		goto out;
+	} else {
+		err = kvm_map_idmap_text(hyp_pgd);
+		if (err)
+			goto out;
 	}
 
 	return 0;

arch/arm64/include/asm/cpufeature.h

@@ -36,8 +36,9 @@
 #define ARM64_HAS_VIRT_HOST_EXTN		11
 #define ARM64_WORKAROUND_CAVIUM_27456		12
 #define ARM64_HAS_32BIT_EL0			13
+#define ARM64_HYP_OFFSET_LOW			14
 
-#define ARM64_NCAPS				14
+#define ARM64_NCAPS				15
 
 #ifndef __ASSEMBLY__
 

arch/arm64/include/asm/kvm_host.h

@@ -47,8 +47,7 @@
 
 int __attribute_const__ kvm_target_cpu(void);
 int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
-int kvm_arch_dev_ioctl_check_extension(long ext);
-unsigned long kvm_hyp_reset_entry(void);
+int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext);
 void __extended_idmap_trampoline(phys_addr_t boot_pgd, phys_addr_t idmap_start);
 
 struct kvm_arch {
@@ -348,8 +347,7 @@ int kvm_perf_teardown(void);
 
 struct kvm_vcpu *kvm_mpidr_to_vcpu(struct kvm *kvm, unsigned long mpidr);
 
-static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
-				       phys_addr_t pgd_ptr,
+static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
 				       unsigned long hyp_stack_ptr,
 				       unsigned long vector_ptr)
 {
@@ -357,19 +355,14 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
 	 * Call initialization code, and switch to the full blown
 	 * HYP code.
 	 */
-	__kvm_call_hyp((void *)boot_pgd_ptr, pgd_ptr,
-		       hyp_stack_ptr, vector_ptr);
+	__kvm_call_hyp((void *)pgd_ptr, hyp_stack_ptr, vector_ptr);
 }
 
-static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr,
+void __kvm_hyp_teardown(void);
+static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr,
 					phys_addr_t phys_idmap_start)
 {
-	/*
-	 * Call reset code, and switch back to stub hyp vectors.
-	 * Uses __kvm_call_hyp() to avoid kaslr's kvm_ksym_ref() translation.
-	 */
-	__kvm_call_hyp((void *)kvm_hyp_reset_entry(),
-		       boot_pgd_ptr, phys_idmap_start);
+	kvm_call_hyp(__kvm_hyp_teardown, phys_idmap_start);
 }
 
 static inline void kvm_arch_hardware_unsetup(void) {}

arch/arm64/include/asm/kvm_hyp.h

@@ -25,29 +25,6 @@
 
 #define __hyp_text __section(.hyp.text) notrace
 
-static inline unsigned long __kern_hyp_va(unsigned long v)
-{
-	asm volatile(ALTERNATIVE("and %0, %0, %1",
-				 "nop",
-				 ARM64_HAS_VIRT_HOST_EXTN)
-		     : "+r" (v) : "i" (HYP_PAGE_OFFSET_MASK));
-	return v;
-}
-
-#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v)))
-
-static inline unsigned long __hyp_kern_va(unsigned long v)
-{
-	u64 offset = PAGE_OFFSET - HYP_PAGE_OFFSET;
-	asm volatile(ALTERNATIVE("add %0, %0, %1",
-				 "nop",
-				 ARM64_HAS_VIRT_HOST_EXTN)
-		     : "+r" (v) : "r" (offset));
-	return v;
-}
-
-#define hyp_kern_va(v) (typeof(v))(__hyp_kern_va((unsigned long)(v)))
-
 #define read_sysreg_elx(r,nvh,vh)					\
 	({								\
 		u64 reg;						\

arch/arm64/include/asm/kvm_mmu.h

@@ -29,21 +29,48 @@
  *
  * Instead, give the HYP mode its own VA region at a fixed offset from
  * the kernel by just masking the top bits (which are all ones for a
- * kernel address).
+ * kernel address). We need to find out how many bits to mask.
  *
- * ARMv8.1 (using VHE) does have a TTBR1_EL2, and doesn't use these
- * macros (the entire kernel runs at EL2).
+ * We want to build a set of page tables that cover both parts of the
+ * idmap (the trampoline page used to initialize EL2), and our normal
+ * runtime VA space, at the same time.
+ *
+ * Given that the kernel uses VA_BITS for its entire address space,
+ * and that half of that space (VA_BITS - 1) is used for the linear
+ * mapping, we can also limit the EL2 space to (VA_BITS - 1).
+ *
+ * The main question is "Within the VA_BITS space, does EL2 use the
+ * top or the bottom half of that space to shadow the kernel's linear
+ * mapping?". As we need to idmap the trampoline page, this is
+ * determined by the range in which this page lives.
+ *
+ * If the page is in the bottom half, we have to use the top half. If
+ * the page is in the top half, we have to use the bottom half:
+ *
+ * T = __virt_to_phys(__hyp_idmap_text_start)
+ * if (T & BIT(VA_BITS - 1))
+ *	HYP_VA_MIN = 0  //idmap in upper half
+ * else
+ *	HYP_VA_MIN = 1 << (VA_BITS - 1)
+ * HYP_VA_MAX = HYP_VA_MIN + (1 << (VA_BITS - 1)) - 1
+ *
+ * This of course assumes that the trampoline page exists within the
+ * VA_BITS range. If it doesn't, then it means we're in the odd case
+ * where the kernel idmap (as well as HYP) uses more levels than the
+ * kernel runtime page tables (as seen when the kernel is configured
+ * for 4k pages, 39bits VA, and yet memory lives just above that
+ * limit, forcing the idmap to use 4 levels of page tables while the
+ * kernel itself only uses 3). In this particular case, it doesn't
+ * matter which side of VA_BITS we use, as we're guaranteed not to
+ * conflict with anything.
+ *
+ * When using VHE, there are no separate hyp mappings and all KVM
+ * functionality is already mapped as part of the main kernel
+ * mappings, and none of this applies in that case.
  */
-#define HYP_PAGE_OFFSET_SHIFT	VA_BITS
-#define HYP_PAGE_OFFSET_MASK	((UL(1) << HYP_PAGE_OFFSET_SHIFT) - 1)
-#define HYP_PAGE_OFFSET		(PAGE_OFFSET & HYP_PAGE_OFFSET_MASK)
 
-/*
- * Our virtual mapping for the idmap-ed MMU-enable code. Must be
- * shared across all the page-tables. Conveniently, we use the last
- * possible page, where no kernel mapping will ever exist.
- */
-#define TRAMPOLINE_VA		(HYP_PAGE_OFFSET_MASK & PAGE_MASK)
+#define HYP_PAGE_OFFSET_HIGH_MASK	((UL(1) << VA_BITS) - 1)
+#define HYP_PAGE_OFFSET_LOW_MASK	((UL(1) << (VA_BITS - 1)) - 1)
 
 #ifdef __ASSEMBLY__
 
@@ -53,13 +80,33 @@
 /*
  * Convert a kernel VA into a HYP VA.
  * reg: VA to be converted.
+ *
+ * This generates the following sequences:
+ * - High mask:
+ *		and x0, x0, #HYP_PAGE_OFFSET_HIGH_MASK
+ *		nop
+ * - Low mask:
+ *		and x0, x0, #HYP_PAGE_OFFSET_HIGH_MASK
+ *		and x0, x0, #HYP_PAGE_OFFSET_LOW_MASK
+ * - VHE:
+ *		nop
+ *		nop
+ *
+ * The "low mask" version works because the mask is a strict subset of
+ * the "high mask", hence performing the first mask for nothing.
+ * Should be completely invisible on any viable CPU.
  */
 .macro kern_hyp_va	reg
-alternative_if_not ARM64_HAS_VIRT_HOST_EXTN	
-	and	\reg, \reg, #HYP_PAGE_OFFSET_MASK
+alternative_if_not ARM64_HAS_VIRT_HOST_EXTN
+	and     \reg, \reg, #HYP_PAGE_OFFSET_HIGH_MASK
 alternative_else
 	nop
 alternative_endif
+alternative_if_not ARM64_HYP_OFFSET_LOW
+	nop
+alternative_else
+	and     \reg, \reg, #HYP_PAGE_OFFSET_LOW_MASK
+alternative_endif
 .endm
 
 #else
@@ -70,7 +117,22 @@ alternative_endif
 #include <asm/mmu_context.h>
 #include <asm/pgtable.h>
 
-#define KERN_TO_HYP(kva)	((unsigned long)kva - PAGE_OFFSET + HYP_PAGE_OFFSET)
+static inline unsigned long __kern_hyp_va(unsigned long v)
+{
+	asm volatile(ALTERNATIVE("and %0, %0, %1",
+				 "nop",
+				 ARM64_HAS_VIRT_HOST_EXTN)
+		     : "+r" (v)
+		     : "i" (HYP_PAGE_OFFSET_HIGH_MASK));
+	asm volatile(ALTERNATIVE("nop",
+				 "and %0, %0, %1",
+				 ARM64_HYP_OFFSET_LOW)
+		     : "+r" (v)
+		     : "i" (HYP_PAGE_OFFSET_LOW_MASK));
+	return v;
+}
+
+#define kern_hyp_va(v) 	(typeof(v))(__kern_hyp_va((unsigned long)(v)))
 
 /*
  * We currently only support a 40bit IPA.
@@ -81,9 +143,8 @@ alternative_endif
 
 #include <asm/stage2_pgtable.h>
 
-int create_hyp_mappings(void *from, void *to);
+int create_hyp_mappings(void *from, void *to, pgprot_t prot);
 int create_hyp_io_mappings(void *from, void *to, phys_addr_t);
-void free_boot_hyp_pgd(void);
 void free_hyp_pgds(void);
 
 void stage2_unmap_vm(struct kvm *kvm);
@@ -97,7 +158,6 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run);
 void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu);
 
 phys_addr_t kvm_mmu_get_httbr(void);
-phys_addr_t kvm_mmu_get_boot_httbr(void);
 phys_addr_t kvm_get_idmap_vector(void);
 phys_addr_t kvm_get_idmap_start(void);
 int kvm_mmu_init(void);

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

@@ -164,6 +164,7 @@
 #define PTE_CONT		(_AT(pteval_t, 1) << 52)	/* Contiguous range */
 #define PTE_PXN			(_AT(pteval_t, 1) << 53)	/* Privileged XN */
 #define PTE_UXN			(_AT(pteval_t, 1) << 54)	/* User XN */
+#define PTE_HYP_XN		(_AT(pteval_t, 1) << 54)	/* HYP XN */
 
 /*
  * AttrIndx[2:0] encoding (mapping attributes defined in the MAIR* registers).

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

@@ -55,7 +55,9 @@
 #define PAGE_KERNEL_EXEC	__pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE)
 #define PAGE_KERNEL_EXEC_CONT	__pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT)
 
-#define PAGE_HYP		__pgprot(_PAGE_DEFAULT | PTE_HYP)
+#define PAGE_HYP		__pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_HYP_XN)
+#define PAGE_HYP_EXEC		__pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY)
+#define PAGE_HYP_RO		__pgprot(_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)

arch/arm64/include/asm/virt.h

@@ -82,6 +82,10 @@ extern void verify_cpu_run_el(void);
 static inline void verify_cpu_run_el(void) {}
 #endif
 
+/* The section containing the hypervisor idmap text */
+extern char __hyp_idmap_text_start[];
+extern char __hyp_idmap_text_end[];
+
 /* The section containing the hypervisor text */
 extern char __hyp_text_start[];
 extern char __hyp_text_end[];

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

@@ -87,9 +87,11 @@ struct kvm_regs {
 /* Supported VGICv3 address types  */
 #define KVM_VGIC_V3_ADDR_TYPE_DIST	2
 #define KVM_VGIC_V3_ADDR_TYPE_REDIST	3
+#define KVM_VGIC_ITS_ADDR_TYPE		4
 
 #define KVM_VGIC_V3_DIST_SIZE		SZ_64K
 #define KVM_VGIC_V3_REDIST_SIZE		(2 * SZ_64K)
+#define KVM_VGIC_V3_ITS_SIZE		(2 * SZ_64K)
 
 #define KVM_ARM_VCPU_POWER_OFF		0 /* CPU is started in OFF state */
 #define KVM_ARM_VCPU_EL1_32BIT		1 /* CPU running a 32bit VM */

arch/arm64/kernel/cpufeature.c

@@ -726,6 +726,19 @@ static bool runs_at_el2(const struct arm64_cpu_capabilities *entry, int __unused
 	return is_kernel_in_hyp_mode();
 }
 
+static bool hyp_offset_low(const struct arm64_cpu_capabilities *entry,
+			   int __unused)
+{
+	phys_addr_t idmap_addr = virt_to_phys(__hyp_idmap_text_start);
+
+	/*
+	 * Activate the lower HYP offset only if:
+	 * - the idmap doesn't clash with it,
+	 * - the kernel is not running at EL2.
+	 */
+	return idmap_addr > GENMASK(VA_BITS - 2, 0) && !is_kernel_in_hyp_mode();
+}
+
 static const struct arm64_cpu_capabilities arm64_features[] = {
 	{
 		.desc = "GIC system register CPU interface",
@@ -803,6 +816,12 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
 		.field_pos = ID_AA64PFR0_EL0_SHIFT,
 		.min_field_value = ID_AA64PFR0_EL0_32BIT_64BIT,
 	},
+	{
+		.desc = "Reduced HYP mapping offset",
+		.capability = ARM64_HYP_OFFSET_LOW,
+		.def_scope = SCOPE_SYSTEM,
+		.matches = hyp_offset_low,
+	},
 	{},
 };
 

arch/arm64/kvm/Kconfig

@@ -36,6 +36,7 @@ config KVM
 	select HAVE_KVM_IRQFD
 	select KVM_ARM_VGIC_V3
 	select KVM_ARM_PMU if HW_PERF_EVENTS
+	select HAVE_KVM_MSI
 	---help---
 	  Support hosting virtualized guest machines.
 	  We don't support KVM with 16K page tables yet, due to the multiple
@@ -54,13 +55,6 @@ config KVM_ARM_PMU
 	  Adds support for a virtual Performance Monitoring Unit (PMU) in
 	  virtual machines.
 
-config KVM_NEW_VGIC
-	bool "New VGIC implementation"
-	depends on KVM
-	default y
-        ---help---
-          uses the new VGIC implementation
-
 source drivers/vhost/Kconfig
 
 endif # VIRTUALIZATION

arch/arm64/kvm/Makefile

@@ -20,7 +20,6 @@ kvm-$(CONFIG_KVM_ARM_HOST) += emulate.o inject_fault.o regmap.o
 kvm-$(CONFIG_KVM_ARM_HOST) += hyp.o hyp-init.o handle_exit.o
 kvm-$(CONFIG_KVM_ARM_HOST) += guest.o debug.o reset.o sys_regs.o sys_regs_generic_v8.o
 
-ifeq ($(CONFIG_KVM_NEW_VGIC),y)
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic.o
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-init.o
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-irqfd.o
@@ -30,12 +29,6 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio.o
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v2.o
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v3.o
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-kvm-device.o
-else
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v2.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v2-emul.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v3.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v3-emul.o
-endif
+kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
 kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
 kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o

arch/arm64/kvm/hyp-init.S

@@ -53,10 +53,9 @@ __invalid:
 	b	.
 
 	/*
-	 * x0: HYP boot pgd
-	 * x1: HYP pgd
-	 * x2: HYP stack
-	 * x3: HYP vectors
+	 * x0: HYP pgd
+	 * x1: HYP stack
+	 * x2: HYP vectors
 	 */
 __do_hyp_init:
 
@@ -110,71 +109,27 @@ __do_hyp_init:
 	msr	sctlr_el2, x4
 	isb
 
-	/* Skip the trampoline dance if we merged the boot and runtime PGDs */
-	cmp	x0, x1
-	b.eq	merged
-
-	/* MMU is now enabled. Get ready for the trampoline dance */
-	ldr	x4, =TRAMPOLINE_VA
-	adr	x5, target
-	bfi	x4, x5, #0, #PAGE_SHIFT
-	br	x4
-
-target: /* We're now in the trampoline code, switch page tables */
-	msr	ttbr0_el2, x1
-	isb
-
-	/* Invalidate the old TLBs */
-	tlbi	alle2
-	dsb	sy
-
-merged:
 	/* Set the stack and new vectors */
+	kern_hyp_va	x1
+	mov	sp, x1
 	kern_hyp_va	x2
-	mov	sp, x2
-	kern_hyp_va	x3
-	msr	vbar_el2, x3
+	msr	vbar_el2, x2
 
 	/* Hello, World! */
 	eret
 ENDPROC(__kvm_hyp_init)
 
 	/*
-	 * Reset kvm back to the hyp stub. This is the trampoline dance in
-	 * reverse. If kvm used an extended idmap, __extended_idmap_trampoline
-	 * calls this code directly in the idmap. In this case switching to the
-	 * boot tables is a no-op.
-	 *
-	 * x0: HYP boot pgd
-	 * x1: HYP phys_idmap_start
+	 * Reset kvm back to the hyp stub.
 	 */
 ENTRY(__kvm_hyp_reset)
-	/* We're in trampoline code in VA, switch back to boot page tables */
-	msr	ttbr0_el2, x0
-	isb
-
-	/* Ensure the PA branch doesn't find a stale tlb entry or stale code. */
-	ic	iallu
-	tlbi	alle2
-	dsb	sy
-	isb
-
-	/* Branch into PA space */
-	adr	x0, 1f
-	bfi	x1, x0, #0, #PAGE_SHIFT
-	br	x1
-
 	/* We're now in idmap, disable MMU */
-1:	mrs	x0, sctlr_el2
+	mrs	x0, sctlr_el2
 	ldr	x1, =SCTLR_ELx_FLAGS
 	bic	x0, x0, x1		// Clear SCTL_M and etc
 	msr	sctlr_el2, x0
 	isb
 
-	/* Invalidate the old TLBs */
-	tlbi	alle2
-	dsb	sy
-
 	/* Install stub vectors */
 	adr_l	x0, __hyp_stub_vectors
 	msr	vbar_el2, x0

arch/arm64/kvm/hyp/entry.S

@@ -164,22 +164,3 @@ alternative_endif
 
 	eret
 ENDPROC(__fpsimd_guest_restore)
-
-/*
- * When using the extended idmap, we don't have a trampoline page we can use
- * while we switch pages tables during __kvm_hyp_reset. Accessing the idmap
- * directly would be ideal, but if we're using the extended idmap then the
- * idmap is located above HYP_PAGE_OFFSET, and the address will be masked by
- * kvm_call_hyp using kern_hyp_va.
- *
- * x0: HYP boot pgd
- * x1: HYP phys_idmap_start
- */
-ENTRY(__extended_idmap_trampoline)
-	mov	x4, x1
-	adr_l	x3, __kvm_hyp_reset
-
-	/* insert __kvm_hyp_reset()s offset into phys_idmap_start */
-	bfi	x4, x3, #0, #PAGE_SHIFT
-	br	x4
-ENDPROC(__extended_idmap_trampoline)

arch/arm64/kvm/hyp/hyp-entry.S

@@ -62,6 +62,21 @@ ENTRY(__vhe_hyp_call)
 	isb
 	ret
 ENDPROC(__vhe_hyp_call)
+
+/*
+ * Compute the idmap address of __kvm_hyp_reset based on the idmap
+ * start passed as a parameter, and jump there.
+ *
+ * x0: HYP phys_idmap_start
+ */
+ENTRY(__kvm_hyp_teardown)
+	mov	x4, x0
+	adr_l	x3, __kvm_hyp_reset
+
+	/* insert __kvm_hyp_reset()s offset into phys_idmap_start */
+	bfi	x4, x3, #0, #PAGE_SHIFT
+	br	x4
+ENDPROC(__kvm_hyp_teardown)
 	
 el1_sync:				// Guest trapped into EL2
 	save_x0_to_x3

arch/arm64/kvm/hyp/switch.c

@@ -299,9 +299,16 @@ static const char __hyp_panic_string[] = "HYP panic:\nPS:%08llx PC:%016llx ESR:%
 
 static void __hyp_text __hyp_call_panic_nvhe(u64 spsr, u64 elr, u64 par)
 {
-	unsigned long str_va = (unsigned long)__hyp_panic_string;
+	unsigned long str_va;
 
-	__hyp_do_panic(hyp_kern_va(str_va),
+	/*
+	 * Force the panic string to be loaded from the literal pool,
+	 * making sure it is a kernel address and not a PC-relative
+	 * reference.
+	 */
+	asm volatile("ldr %0, =__hyp_panic_string" : "=r" (str_va));
+
+	__hyp_do_panic(str_va,
 		       spsr,  elr,
 		       read_sysreg(esr_el2),   read_sysreg_el2(far),
 		       read_sysreg(hpfar_el2), par,

arch/arm64/kvm/reset.c

@@ -65,7 +65,7 @@ static bool cpu_has_32bit_el1(void)
  * We currently assume that the number of HW registers is uniform
  * across all CPUs (see cpuinfo_sanity_check).
  */
-int kvm_arch_dev_ioctl_check_extension(long ext)
+int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
 {
 	int r;
 
@@ -86,6 +86,12 @@ int kvm_arch_dev_ioctl_check_extension(long ext)
 	case KVM_CAP_VCPU_ATTRIBUTES:
 		r = 1;
 		break;
+	case KVM_CAP_MSI_DEVID:
+		if (!kvm)
+			r = -EINVAL;
+		else
+			r = kvm->arch.vgic.msis_require_devid;
+		break;
 	default:
 		r = 0;
 	}
@@ -132,31 +138,3 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu)
 	/* Reset timer */
 	return kvm_timer_vcpu_reset(vcpu, cpu_vtimer_irq);
 }
-
-extern char __hyp_idmap_text_start[];
-
-unsigned long kvm_hyp_reset_entry(void)
-{
-	if (!__kvm_cpu_uses_extended_idmap()) {
-		unsigned long offset;
-
-		/*
-		 * Find the address of __kvm_hyp_reset() in the trampoline page.
-		 * This is present in the running page tables, and the boot page
-		 * tables, so we call the code here to start the trampoline
-		 * dance in reverse.
-		 */
-		offset = (unsigned long)__kvm_hyp_reset
-			 - ((unsigned long)__hyp_idmap_text_start & PAGE_MASK);
-
-		return TRAMPOLINE_VA + offset;
-	} else {
-		/*
-		 * KVM is running with merged page tables, which don't have the
-		 * trampoline page mapped. We know the idmap is still mapped,
-		 * but can't be called into directly. Use
-		 * __extended_idmap_trampoline to do the call.
-		 */
-		return (unsigned long)kvm_ksym_ref(__extended_idmap_trampoline);
-	}
-}

arch/arm64/kvm/sys_regs.c

@@ -1546,7 +1546,7 @@ static void unhandled_cp_access(struct kvm_vcpu *vcpu,
 				struct sys_reg_params *params)
 {
 	u8 hsr_ec = kvm_vcpu_trap_get_class(vcpu);
-	int cp;
+	int cp = -1;
 
 	switch(hsr_ec) {
 	case ESR_ELx_EC_CP15_32:
@@ -1558,7 +1558,7 @@ static void unhandled_cp_access(struct kvm_vcpu *vcpu,
 		cp = 14;
 		break;
 	default:
-		WARN_ON((cp = -1));
+		WARN_ON(1);
 	}
 
 	kvm_err("Unsupported guest CP%d access at: %08lx\n",

include/kvm/arm_vgic.h

@@ -1,6 +1,5 @@
 /*
- * Copyright (C) 2012 ARM Ltd.
- * Author: Marc Zyngier <marc.zyngier@arm.com>
+ * Copyright (C) 2015, 2016 ARM Ltd.
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -12,16 +11,10 @@
  * GNU General Public License for more details.
  *
  * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
  */
-
-#ifndef __ASM_ARM_KVM_VGIC_H
-#define __ASM_ARM_KVM_VGIC_H
-
-#ifdef CONFIG_KVM_NEW_VGIC
-#include <kvm/vgic/vgic.h>
-#else
+#ifndef __KVM_ARM_VGIC_H
+#define __KVM_ARM_VGIC_H
 
 #include <linux/kernel.h>
 #include <linux/kvm.h>
@@ -29,248 +22,187 @@
 #include <linux/spinlock.h>
 #include <linux/types.h>
 #include <kvm/iodev.h>
-#include <linux/irqchip/arm-gic-common.h>
+#include <linux/list.h>
 
-#define VGIC_NR_IRQS_LEGACY	256
+#define VGIC_V3_MAX_CPUS	255
+#define VGIC_V2_MAX_CPUS	8
+#define VGIC_NR_IRQS_LEGACY     256
 #define VGIC_NR_SGIS		16
 #define VGIC_NR_PPIS		16
 #define VGIC_NR_PRIVATE_IRQS	(VGIC_NR_SGIS + VGIC_NR_PPIS)
+#define VGIC_MAX_PRIVATE	(VGIC_NR_PRIVATE_IRQS - 1)
+#define VGIC_MAX_SPI		1019
+#define VGIC_MAX_RESERVED	1023
+#define VGIC_MIN_LPI		8192
 
-#define VGIC_V2_MAX_LRS		(1 << 6)
-#define VGIC_V3_MAX_LRS		16
-#define VGIC_MAX_IRQS		1024
-#define VGIC_V2_MAX_CPUS	8
-#define VGIC_V3_MAX_CPUS	255
+enum vgic_type {
+	VGIC_V2,		/* Good ol' GICv2 */
+	VGIC_V3,		/* New fancy GICv3 */
+};
 
-#if (VGIC_NR_IRQS_LEGACY & 31)
-#error "VGIC_NR_IRQS must be a multiple of 32"
-#endif
+/* same for all guests, as depending only on the _host's_ GIC model */
+struct vgic_global {
+	/* type of the host GIC */
+	enum vgic_type		type;
 
-#if (VGIC_NR_IRQS_LEGACY > VGIC_MAX_IRQS)
-#error "VGIC_NR_IRQS must be <= 1024"
-#endif
+	/* Physical address of vgic virtual cpu interface */
+	phys_addr_t		vcpu_base;
 
-/*
- * The GIC distributor registers describing interrupts have two parts:
- * - 32 per-CPU interrupts (SGI + PPI)
- * - a bunch of shared interrupts (SPI)
- */
-struct vgic_bitmap {
-	/*
-	 * - One UL per VCPU for private interrupts (assumes UL is at
-	 *   least 32 bits)
-	 * - As many UL as necessary for shared interrupts.
-	 *
-	 * The private interrupts are accessed via the "private"
-	 * field, one UL per vcpu (the state for vcpu n is in
-	 * private[n]). The shared interrupts are accessed via the
-	 * "shared" pointer (IRQn state is at bit n-32 in the bitmap).
-	 */
-	unsigned long *private;
-	unsigned long *shared;
-};
+	/* virtual control interface mapping */
+	void __iomem		*vctrl_base;
 
-struct vgic_bytemap {
-	/*
-	 * - 8 u32 per VCPU for private interrupts
-	 * - As many u32 as necessary for shared interrupts.
-	 *
-	 * The private interrupts are accessed via the "private"
-	 * field, (the state for vcpu n is in private[n*8] to
-	 * private[n*8 + 7]). The shared interrupts are accessed via
-	 * the "shared" pointer (IRQn state is at byte (n-32)%4 of the
-	 * shared[(n-32)/4] word).
-	 */
-	u32 *private;
-	u32 *shared;
-};
+	/* Number of implemented list registers */
+	int			nr_lr;
 
-struct kvm_vcpu;
+	/* Maintenance IRQ number */
+	unsigned int		maint_irq;
 
-enum vgic_type {
-	VGIC_V2,		/* Good ol' GICv2 */
-	VGIC_V3,		/* New fancy GICv3 */
+	/* maximum number of VCPUs allowed (GICv2 limits us to 8) */
+	int			max_gic_vcpus;
+
+	/* Only needed for the legacy KVM_CREATE_IRQCHIP */
+	bool			can_emulate_gicv2;
 };
 
-#define LR_STATE_PENDING	(1 << 0)
-#define LR_STATE_ACTIVE		(1 << 1)
-#define LR_STATE_MASK		(3 << 0)
-#define LR_EOI_INT		(1 << 2)
-#define LR_HW			(1 << 3)
+extern struct vgic_global kvm_vgic_global_state;
 
-struct vgic_lr {
-	unsigned irq:10;
-	union {
-		unsigned hwirq:10;
-		unsigned source:3;
-	};
-	unsigned state:4;
-};
+#define VGIC_V2_MAX_LRS		(1 << 6)
+#define VGIC_V3_MAX_LRS		16
+#define VGIC_V3_LR_INDEX(lr)	(VGIC_V3_MAX_LRS - 1 - lr)
 
-struct vgic_vmcr {
-	u32	ctlr;
-	u32	abpr;
-	u32	bpr;
-	u32	pmr;
+enum vgic_irq_config {
+	VGIC_CONFIG_EDGE = 0,
+	VGIC_CONFIG_LEVEL
 };
 
-struct vgic_ops {
-	struct vgic_lr	(*get_lr)(const struct kvm_vcpu *, int);
-	void	(*set_lr)(struct kvm_vcpu *, int, struct vgic_lr);
-	u64	(*get_elrsr)(const struct kvm_vcpu *vcpu);
-	u64	(*get_eisr)(const struct kvm_vcpu *vcpu);
-	void	(*clear_eisr)(struct kvm_vcpu *vcpu);
-	u32	(*get_interrupt_status)(const struct kvm_vcpu *vcpu);
-	void	(*enable_underflow)(struct kvm_vcpu *vcpu);
-	void	(*disable_underflow)(struct kvm_vcpu *vcpu);
-	void	(*get_vmcr)(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcr);
-	void	(*set_vmcr)(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcr);
-	void	(*enable)(struct kvm_vcpu *vcpu);
+struct vgic_irq {
+	spinlock_t irq_lock;		/* Protects the content of the struct */
+	struct list_head lpi_list;	/* Used to link all LPIs together */
+	struct list_head ap_list;
+
+	struct kvm_vcpu *vcpu;		/* SGIs and PPIs: The VCPU
+					 * SPIs and LPIs: The VCPU whose ap_list
+					 * this is queued on.
+					 */
+
+	struct kvm_vcpu *target_vcpu;	/* The VCPU that this interrupt should
+					 * be sent to, as a result of the
+					 * targets reg (v2) or the
+					 * affinity reg (v3).
+					 */
+
+	u32 intid;			/* Guest visible INTID */
+	bool pending;
+	bool line_level;		/* Level only */
+	bool soft_pending;		/* Level only */
+	bool active;			/* not used for LPIs */
+	bool enabled;
+	bool hw;			/* Tied to HW IRQ */
+	struct kref refcount;		/* Used for LPIs */
+	u32 hwintid;			/* HW INTID number */
+	union {
+		u8 targets;			/* GICv2 target VCPUs mask */
+		u32 mpidr;			/* GICv3 target VCPU */
+	};
+	u8 source;			/* GICv2 SGIs only */
+	u8 priority;
+	enum vgic_irq_config config;	/* Level or edge */
 };
 
-struct vgic_params {
-	/* vgic type */
-	enum vgic_type	type;
-	/* Physical address of vgic virtual cpu interface */
-	phys_addr_t	vcpu_base;
-	/* Number of list registers */
-	u32		nr_lr;
-	/* Interrupt number */
-	unsigned int	maint_irq;
-	/* Virtual control interface base address */
-	void __iomem	*vctrl_base;
-	int		max_gic_vcpus;
-	/* Only needed for the legacy KVM_CREATE_IRQCHIP */
-	bool		can_emulate_gicv2;
-};
+struct vgic_register_region;
+struct vgic_its;
 
-struct vgic_vm_ops {
-	bool	(*queue_sgi)(struct kvm_vcpu *, int irq);
-	void	(*add_sgi_source)(struct kvm_vcpu *, int irq, int source);
-	int	(*init_model)(struct kvm *);
-	int	(*map_resources)(struct kvm *, const struct vgic_params *);
+enum iodev_type {
+	IODEV_CPUIF,
+	IODEV_DIST,
+	IODEV_REDIST,
+	IODEV_ITS
 };
 
 struct vgic_io_device {
-	gpa_t addr;
-	int len;
-	const struct vgic_io_range *reg_ranges;
-	struct kvm_vcpu *redist_vcpu;
+	gpa_t base_addr;
+	union {
+		struct kvm_vcpu *redist_vcpu;
+		struct vgic_its *its;
+	};
+	const struct vgic_register_region *regions;
+	enum iodev_type iodev_type;
+	int nr_regions;
 	struct kvm_io_device dev;
 };
 
-struct irq_phys_map {
-	u32			virt_irq;
-	u32			phys_irq;
-};
-
-struct irq_phys_map_entry {
-	struct list_head	entry;
-	struct rcu_head		rcu;
-	struct irq_phys_map	map;
+struct vgic_its {
+	/* The base address of the ITS control register frame */
+	gpa_t			vgic_its_base;
+
+	bool			enabled;
+	bool			initialized;
+	struct vgic_io_device	iodev;
+	struct kvm_device	*dev;
+
+	/* These registers correspond to GITS_BASER{0,1} */
+	u64			baser_device_table;
+	u64			baser_coll_table;
+
+	/* Protects the command queue */
+	struct mutex		cmd_lock;
+	u64			cbaser;
+	u32			creadr;
+	u32			cwriter;
+
+	/* Protects the device and collection lists */
+	struct mutex		its_lock;
+	struct list_head	device_list;
+	struct list_head	collection_list;
 };
 
 struct vgic_dist {
-	spinlock_t		lock;
 	bool			in_kernel;
 	bool			ready;
+	bool			initialized;
 
 	/* vGIC model the kernel emulates for the guest (GICv2 or GICv3) */
 	u32			vgic_model;
 
-	int			nr_cpus;
-	int			nr_irqs;
+	/* Do injected MSIs require an additional device ID? */
+	bool			msis_require_devid;
+
+	int			nr_spis;
 
+	/* TODO: Consider moving to global state */
 	/* Virtual control interface mapping */
 	void __iomem		*vctrl_base;
 
-	/* Distributor and vcpu interface mapping in the guest */
-	phys_addr_t		vgic_dist_base;
-	/* GICv2 and GICv3 use different mapped register blocks */
+	/* base addresses in guest physical address space: */
+	gpa_t			vgic_dist_base;		/* distributor */
 	union {
-		phys_addr_t		vgic_cpu_base;
-		phys_addr_t		vgic_redist_base;
+		/* either a GICv2 CPU interface */
+		gpa_t			vgic_cpu_base;
+		/* or a number of GICv3 redistributor regions */
+		gpa_t			vgic_redist_base;
 	};
 
-	/* Distributor enabled */
-	u32			enabled;
-
-	/* Interrupt enabled (one bit per IRQ) */
-	struct vgic_bitmap	irq_enabled;
-
-	/* Level-triggered interrupt external input is asserted */
-	struct vgic_bitmap	irq_level;
-
-	/*
-	 * Interrupt state is pending on the distributor
-	 */
-	struct vgic_bitmap	irq_pending;
-
-	/*
-	 * Tracks writes to GICD_ISPENDRn and GICD_ICPENDRn for level-triggered
-	 * interrupts.  Essentially holds the state of the flip-flop in
-	 * Figure 4-10 on page 4-101 in ARM IHI 0048B.b.
-	 * Once set, it is only cleared for level-triggered interrupts on
-	 * guest ACKs (when we queue it) or writes to GICD_ICPENDRn.
-	 */
-	struct vgic_bitmap	irq_soft_pend;
-
-	/* Level-triggered interrupt queued on VCPU interface */
-	struct vgic_bitmap	irq_queued;
-
-	/* Interrupt was active when unqueue from VCPU interface */
-	struct vgic_bitmap	irq_active;
-
-	/* Interrupt priority. Not used yet. */
-	struct vgic_bytemap	irq_priority;
+	/* distributor enabled */
+	bool			enabled;
 
-	/* Level/edge triggered */
-	struct vgic_bitmap	irq_cfg;
+	struct vgic_irq		*spis;
 
-	/*
-	 * Source CPU per SGI and target CPU:
-	 *
-	 * Each byte represent a SGI observable on a VCPU, each bit of
-	 * this byte indicating if the corresponding VCPU has
-	 * generated this interrupt. This is a GICv2 feature only.
-	 *
-	 * For VCPUn (n < 8), irq_sgi_sources[n*16] to [n*16 + 15] are
-	 * the SGIs observable on VCPUn.
-	 */
-	u8			*irq_sgi_sources;
+	struct vgic_io_device	dist_iodev;
 
-	/*
-	 * Target CPU for each SPI:
-	 *
-	 * Array of available SPI, each byte indicating the target
-	 * VCPU for SPI. IRQn (n >=32) is at irq_spi_cpu[n-32].
-	 */
-	u8			*irq_spi_cpu;
+	bool			has_its;
 
 	/*
-	 * Reverse lookup of irq_spi_cpu for faster compute pending:
-	 *
-	 * Array of bitmaps, one per VCPU, describing if IRQn is
-	 * routed to a particular VCPU.
+	 * Contains the attributes and gpa of the LPI configuration table.
+	 * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
+	 * one address across all redistributors.
+	 * GICv3 spec: 6.1.2 "LPI Configuration tables"
 	 */
-	struct vgic_bitmap	*irq_spi_target;
-
-	/* Target MPIDR for each IRQ (needed for GICv3 IROUTERn) only */
-	u32			*irq_spi_mpidr;
+	u64			propbaser;
 
-	/* Bitmap indicating which CPU has something pending */
-	unsigned long		*irq_pending_on_cpu;
-
-	/* Bitmap indicating which CPU has active IRQs */
-	unsigned long		*irq_active_on_cpu;
-
-	struct vgic_vm_ops	vm_ops;
-	struct vgic_io_device	dist_iodev;
-	struct vgic_io_device	*redist_iodevs;
-
-	/* Virtual irq to hwirq mapping */
-	spinlock_t		irq_phys_map_lock;
-	struct list_head	irq_phys_map_list;
+	/* Protects the lpi_list and the count value below. */
+	spinlock_t		lpi_list_lock;
+	struct list_head	lpi_list_head;
+	int			lpi_list_count;
 };
 
 struct vgic_v2_cpu_if {
@@ -298,78 +230,88 @@ struct vgic_v3_cpu_if {
 };
 
 struct vgic_cpu {
-	/* Pending/active/both interrupts on this VCPU */
-	DECLARE_BITMAP(pending_percpu, VGIC_NR_PRIVATE_IRQS);
-	DECLARE_BITMAP(active_percpu, VGIC_NR_PRIVATE_IRQS);
-	DECLARE_BITMAP(pend_act_percpu, VGIC_NR_PRIVATE_IRQS);
-
-	/* Pending/active/both shared interrupts, dynamically sized */
-	unsigned long	*pending_shared;
-	unsigned long   *active_shared;
-	unsigned long   *pend_act_shared;
-
 	/* CPU vif control registers for world switch */
 	union {
 		struct vgic_v2_cpu_if	vgic_v2;
 		struct vgic_v3_cpu_if	vgic_v3;
 	};
 
-	/* Protected by the distributor's irq_phys_map_lock */
-	struct list_head	irq_phys_map_list;
+	unsigned int used_lrs;
+	struct vgic_irq private_irqs[VGIC_NR_PRIVATE_IRQS];
 
-	u64		live_lrs;
-};
+	spinlock_t ap_list_lock;	/* Protects the ap_list */
+
+	/*
+	 * List of IRQs that this VCPU should consider because they are either
+	 * Active or Pending (hence the name; AP list), or because they recently
+	 * were one of the two and need to be migrated off this list to another
+	 * VCPU.
+	 */
+	struct list_head ap_list_head;
 
-#define LR_EMPTY	0xff
+	u64 live_lrs;
 
-#define INT_STATUS_EOI		(1 << 0)
-#define INT_STATUS_UNDERFLOW	(1 << 1)
+	/*
+	 * Members below are used with GICv3 emulation only and represent
+	 * parts of the redistributor.
+	 */
+	struct vgic_io_device	rd_iodev;
+	struct vgic_io_device	sgi_iodev;
 
-struct kvm;
-struct kvm_vcpu;
+	/* Contains the attributes and gpa of the LPI pending tables. */
+	u64 pendbaser;
+
+	bool lpis_enabled;
+};
 
 int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
-int kvm_vgic_hyp_init(void);
-int kvm_vgic_map_resources(struct kvm *kvm);
-int kvm_vgic_get_max_vcpus(void);
 void kvm_vgic_early_init(struct kvm *kvm);
 int kvm_vgic_create(struct kvm *kvm, u32 type);
 void kvm_vgic_destroy(struct kvm *kvm);
 void kvm_vgic_vcpu_early_init(struct kvm_vcpu *vcpu);
 void kvm_vgic_vcpu_destroy(struct kvm_vcpu *vcpu);
-void kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu);
-void kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu);
-int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int irq_num,
+int kvm_vgic_map_resources(struct kvm *kvm);
+int kvm_vgic_hyp_init(void);
+
+int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int intid,
 			bool level);
-int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid,
-			       unsigned int virt_irq, bool level);
-void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg);
-int kvm_vgic_vcpu_pending_irq(struct kvm_vcpu *vcpu);
-int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, int virt_irq, int phys_irq);
+int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid, unsigned int intid,
+			       bool level);
+int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, u32 virt_irq, u32 phys_irq);
 int kvm_vgic_unmap_phys_irq(struct kvm_vcpu *vcpu, unsigned int virt_irq);
 bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq);
 
+int kvm_vgic_vcpu_pending_irq(struct kvm_vcpu *vcpu);
+
 #define irqchip_in_kernel(k)	(!!((k)->arch.vgic.in_kernel))
-#define vgic_initialized(k)	(!!((k)->arch.vgic.nr_cpus))
+#define vgic_initialized(k)	((k)->arch.vgic.initialized)
 #define vgic_ready(k)		((k)->arch.vgic.ready)
 #define vgic_valid_spi(k, i)	(((i) >= VGIC_NR_PRIVATE_IRQS) && \
-				 ((i) < (k)->arch.vgic.nr_irqs))
+			((i) < (k)->arch.vgic.nr_spis + VGIC_NR_PRIVATE_IRQS))
+
+bool kvm_vcpu_has_pending_irqs(struct kvm_vcpu *vcpu);
+void kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu);
+void kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu);
 
-int vgic_v2_probe(const struct gic_kvm_info *gic_kvm_info,
-		  const struct vgic_ops **ops,
-		  const struct vgic_params **params);
 #ifdef CONFIG_KVM_ARM_VGIC_V3
-int vgic_v3_probe(const struct gic_kvm_info *gic_kvm_info,
-		  const struct vgic_ops **ops,
-		  const struct vgic_params **params);
+void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg);
 #else
-static inline int vgic_v3_probe(const struct gic_kvm_info *gic_kvm_info,
-				const struct vgic_ops **ops,
-				const struct vgic_params **params)
+static inline void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg)
 {
-	return -ENODEV;
 }
 #endif
 
-#endif	/* old VGIC include */
-#endif
+/**
+ * kvm_vgic_get_max_vcpus - Get the maximum number of VCPUs allowed by HW
+ *
+ * The host's GIC naturally limits the maximum amount of VCPUs a guest
+ * can use.
+ */
+static inline int kvm_vgic_get_max_vcpus(void)
+{
+	return kvm_vgic_global_state.max_gic_vcpus;
+}
+
+int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);
+
+#endif /* __KVM_ARM_VGIC_H */

include/kvm/vgic/vgic.h

@@ -1,246 +0,0 @@
-/*
- * Copyright (C) 2015, 2016 ARM Ltd.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-#ifndef __ASM_ARM_KVM_VGIC_VGIC_H
-#define __ASM_ARM_KVM_VGIC_VGIC_H
-
-#include <linux/kernel.h>
-#include <linux/kvm.h>
-#include <linux/irqreturn.h>
-#include <linux/spinlock.h>
-#include <linux/types.h>
-#include <kvm/iodev.h>
-
-#define VGIC_V3_MAX_CPUS	255
-#define VGIC_V2_MAX_CPUS	8
-#define VGIC_NR_IRQS_LEGACY     256
-#define VGIC_NR_SGIS		16
-#define VGIC_NR_PPIS		16
-#define VGIC_NR_PRIVATE_IRQS	(VGIC_NR_SGIS + VGIC_NR_PPIS)
-#define VGIC_MAX_PRIVATE	(VGIC_NR_PRIVATE_IRQS - 1)
-#define VGIC_MAX_SPI		1019
-#define VGIC_MAX_RESERVED	1023
-#define VGIC_MIN_LPI		8192
-
-enum vgic_type {
-	VGIC_V2,		/* Good ol' GICv2 */
-	VGIC_V3,		/* New fancy GICv3 */
-};
-
-/* same for all guests, as depending only on the _host's_ GIC model */
-struct vgic_global {
-	/* type of the host GIC */
-	enum vgic_type		type;
-
-	/* Physical address of vgic virtual cpu interface */
-	phys_addr_t		vcpu_base;
-
-	/* virtual control interface mapping */
-	void __iomem		*vctrl_base;
-
-	/* Number of implemented list registers */
-	int			nr_lr;
-
-	/* Maintenance IRQ number */
-	unsigned int		maint_irq;
-
-	/* maximum number of VCPUs allowed (GICv2 limits us to 8) */
-	int			max_gic_vcpus;
-
-	/* Only needed for the legacy KVM_CREATE_IRQCHIP */
-	bool			can_emulate_gicv2;
-};
-
-extern struct vgic_global kvm_vgic_global_state;
-
-#define VGIC_V2_MAX_LRS		(1 << 6)
-#define VGIC_V3_MAX_LRS		16
-#define VGIC_V3_LR_INDEX(lr)	(VGIC_V3_MAX_LRS - 1 - lr)
-
-enum vgic_irq_config {
-	VGIC_CONFIG_EDGE = 0,
-	VGIC_CONFIG_LEVEL
-};
-
-struct vgic_irq {
-	spinlock_t irq_lock;		/* Protects the content of the struct */
-	struct list_head ap_list;
-
-	struct kvm_vcpu *vcpu;		/* SGIs and PPIs: The VCPU
-					 * SPIs and LPIs: The VCPU whose ap_list
-					 * this is queued on.
-					 */
-
-	struct kvm_vcpu *target_vcpu;	/* The VCPU that this interrupt should
-					 * be sent to, as a result of the
-					 * targets reg (v2) or the
-					 * affinity reg (v3).
-					 */
-
-	u32 intid;			/* Guest visible INTID */
-	bool pending;
-	bool line_level;		/* Level only */
-	bool soft_pending;		/* Level only */
-	bool active;			/* not used for LPIs */
-	bool enabled;
-	bool hw;			/* Tied to HW IRQ */
-	u32 hwintid;			/* HW INTID number */
-	union {
-		u8 targets;			/* GICv2 target VCPUs mask */
-		u32 mpidr;			/* GICv3 target VCPU */
-	};
-	u8 source;			/* GICv2 SGIs only */
-	u8 priority;
-	enum vgic_irq_config config;	/* Level or edge */
-};
-
-struct vgic_register_region;
-
-struct vgic_io_device {
-	gpa_t base_addr;
-	struct kvm_vcpu *redist_vcpu;
-	const struct vgic_register_region *regions;
-	int nr_regions;
-	struct kvm_io_device dev;
-};
-
-struct vgic_dist {
-	bool			in_kernel;
-	bool			ready;
-	bool			initialized;
-
-	/* vGIC model the kernel emulates for the guest (GICv2 or GICv3) */
-	u32			vgic_model;
-
-	int			nr_spis;
-
-	/* TODO: Consider moving to global state */
-	/* Virtual control interface mapping */
-	void __iomem		*vctrl_base;
-
-	/* base addresses in guest physical address space: */
-	gpa_t			vgic_dist_base;		/* distributor */
-	union {
-		/* either a GICv2 CPU interface */
-		gpa_t			vgic_cpu_base;
-		/* or a number of GICv3 redistributor regions */
-		gpa_t			vgic_redist_base;
-	};
-
-	/* distributor enabled */
-	bool			enabled;
-
-	struct vgic_irq		*spis;
-
-	struct vgic_io_device	dist_iodev;
-	struct vgic_io_device	*redist_iodevs;
-};
-
-struct vgic_v2_cpu_if {
-	u32		vgic_hcr;
-	u32		vgic_vmcr;
-	u32		vgic_misr;	/* Saved only */
-	u64		vgic_eisr;	/* Saved only */
-	u64		vgic_elrsr;	/* Saved only */
-	u32		vgic_apr;
-	u32		vgic_lr[VGIC_V2_MAX_LRS];
-};
-
-struct vgic_v3_cpu_if {
-#ifdef CONFIG_KVM_ARM_VGIC_V3
-	u32		vgic_hcr;
-	u32		vgic_vmcr;
-	u32		vgic_sre;	/* Restored only, change ignored */
-	u32		vgic_misr;	/* Saved only */
-	u32		vgic_eisr;	/* Saved only */
-	u32		vgic_elrsr;	/* Saved only */
-	u32		vgic_ap0r[4];
-	u32		vgic_ap1r[4];
-	u64		vgic_lr[VGIC_V3_MAX_LRS];
-#endif
-};
-
-struct vgic_cpu {
-	/* CPU vif control registers for world switch */
-	union {
-		struct vgic_v2_cpu_if	vgic_v2;
-		struct vgic_v3_cpu_if	vgic_v3;
-	};
-
-	unsigned int used_lrs;
-	struct vgic_irq private_irqs[VGIC_NR_PRIVATE_IRQS];
-
-	spinlock_t ap_list_lock;	/* Protects the ap_list */
-
-	/*
-	 * List of IRQs that this VCPU should consider because they are either
-	 * Active or Pending (hence the name; AP list), or because they recently
-	 * were one of the two and need to be migrated off this list to another
-	 * VCPU.
-	 */
-	struct list_head ap_list_head;
-
-	u64 live_lrs;
-};
-
-int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
-void kvm_vgic_early_init(struct kvm *kvm);
-int kvm_vgic_create(struct kvm *kvm, u32 type);
-void kvm_vgic_destroy(struct kvm *kvm);
-void kvm_vgic_vcpu_early_init(struct kvm_vcpu *vcpu);
-void kvm_vgic_vcpu_destroy(struct kvm_vcpu *vcpu);
-int kvm_vgic_map_resources(struct kvm *kvm);
-int kvm_vgic_hyp_init(void);
-
-int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int intid,
-			bool level);
-int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid, unsigned int intid,
-			       bool level);
-int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, u32 virt_irq, u32 phys_irq);
-int kvm_vgic_unmap_phys_irq(struct kvm_vcpu *vcpu, unsigned int virt_irq);
-bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq);
-
-int kvm_vgic_vcpu_pending_irq(struct kvm_vcpu *vcpu);
-
-#define irqchip_in_kernel(k)	(!!((k)->arch.vgic.in_kernel))
-#define vgic_initialized(k)	((k)->arch.vgic.initialized)
-#define vgic_ready(k)		((k)->arch.vgic.ready)
-#define vgic_valid_spi(k, i)	(((i) >= VGIC_NR_PRIVATE_IRQS) && \
-			((i) < (k)->arch.vgic.nr_spis + VGIC_NR_PRIVATE_IRQS))
-
-bool kvm_vcpu_has_pending_irqs(struct kvm_vcpu *vcpu);
-void kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu);
-void kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu);
-
-#ifdef CONFIG_KVM_ARM_VGIC_V3
-void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg);
-#else
-static inline void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg)
-{
-}
-#endif
-
-/**
- * kvm_vgic_get_max_vcpus - Get the maximum number of VCPUs allowed by HW
- *
- * The host's GIC naturally limits the maximum amount of VCPUs a guest
- * can use.
- */
-static inline int kvm_vgic_get_max_vcpus(void)
-{
-	return kvm_vgic_global_state.max_gic_vcpus;
-}
-
-#endif /* __ASM_ARM_KVM_VGIC_VGIC_H */

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

@@ -112,34 +112,76 @@
 #define GICR_WAKER_ProcessorSleep	(1U << 1)
 #define GICR_WAKER_ChildrenAsleep	(1U << 2)
 
-#define GICR_PROPBASER_NonShareable	(0U << 10)
-#define GICR_PROPBASER_InnerShareable	(1U << 10)
-#define GICR_PROPBASER_OuterShareable	(2U << 10)
-#define GICR_PROPBASER_SHAREABILITY_MASK (3UL << 10)
-#define GICR_PROPBASER_nCnB		(0U << 7)
-#define GICR_PROPBASER_nC		(1U << 7)
-#define GICR_PROPBASER_RaWt		(2U << 7)
-#define GICR_PROPBASER_RaWb		(3U << 7)
-#define GICR_PROPBASER_WaWt		(4U << 7)
-#define GICR_PROPBASER_WaWb		(5U << 7)
-#define GICR_PROPBASER_RaWaWt		(6U << 7)
-#define GICR_PROPBASER_RaWaWb		(7U << 7)
-#define GICR_PROPBASER_CACHEABILITY_MASK (7U << 7)
-#define GICR_PROPBASER_IDBITS_MASK	(0x1f)
-
-#define GICR_PENDBASER_NonShareable	(0U << 10)
-#define GICR_PENDBASER_InnerShareable	(1U << 10)
-#define GICR_PENDBASER_OuterShareable	(2U << 10)
-#define GICR_PENDBASER_SHAREABILITY_MASK (3UL << 10)
-#define GICR_PENDBASER_nCnB		(0U << 7)
-#define GICR_PENDBASER_nC		(1U << 7)
-#define GICR_PENDBASER_RaWt		(2U << 7)
-#define GICR_PENDBASER_RaWb		(3U << 7)
-#define GICR_PENDBASER_WaWt		(4U << 7)
-#define GICR_PENDBASER_WaWb		(5U << 7)
-#define GICR_PENDBASER_RaWaWt		(6U << 7)
-#define GICR_PENDBASER_RaWaWb		(7U << 7)
-#define GICR_PENDBASER_CACHEABILITY_MASK (7U << 7)
+#define GIC_BASER_CACHE_nCnB		0ULL
+#define GIC_BASER_CACHE_SameAsInner	0ULL
+#define GIC_BASER_CACHE_nC		1ULL
+#define GIC_BASER_CACHE_RaWt		2ULL
+#define GIC_BASER_CACHE_RaWb		3ULL
+#define GIC_BASER_CACHE_WaWt		4ULL
+#define GIC_BASER_CACHE_WaWb		5ULL
+#define GIC_BASER_CACHE_RaWaWt		6ULL
+#define GIC_BASER_CACHE_RaWaWb		7ULL
+#define GIC_BASER_CACHE_MASK		7ULL
+#define GIC_BASER_NonShareable		0ULL
+#define GIC_BASER_InnerShareable	1ULL
+#define GIC_BASER_OuterShareable	2ULL
+#define GIC_BASER_SHAREABILITY_MASK	3ULL
+
+#define GIC_BASER_CACHEABILITY(reg, inner_outer, type)			\
+	(GIC_BASER_CACHE_##type << reg##_##inner_outer##_CACHEABILITY_SHIFT)
+
+#define GIC_BASER_SHAREABILITY(reg, type)				\
+	(GIC_BASER_##type << reg##_SHAREABILITY_SHIFT)
+
+#define GICR_PROPBASER_SHAREABILITY_SHIFT		(10)
+#define GICR_PROPBASER_INNER_CACHEABILITY_SHIFT		(7)
+#define GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT		(56)
+#define GICR_PROPBASER_SHAREABILITY_MASK				\
+	GIC_BASER_SHAREABILITY(GICR_PROPBASER, SHAREABILITY_MASK)
+#define GICR_PROPBASER_INNER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, MASK)
+#define GICR_PROPBASER_OUTER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GICR_PROPBASER, OUTER, MASK)
+#define GICR_PROPBASER_CACHEABILITY_MASK GICR_PROPBASER_INNER_CACHEABILITY_MASK
+
+#define GICR_PROPBASER_InnerShareable					\
+	GIC_BASER_SHAREABILITY(GICR_PROPBASER, InnerShareable)
+
+#define GICR_PROPBASER_nCnB	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, nCnB)
+#define GICR_PROPBASER_nC 	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, nC)
+#define GICR_PROPBASER_RaWt	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWt)
+#define GICR_PROPBASER_RaWb	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWt)
+#define GICR_PROPBASER_WaWt	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, WaWt)
+#define GICR_PROPBASER_WaWb	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, WaWb)
+#define GICR_PROPBASER_RaWaWt	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWaWt)
+#define GICR_PROPBASER_RaWaWb	GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWaWb)
+
+#define GICR_PROPBASER_IDBITS_MASK			(0x1f)
+
+#define GICR_PENDBASER_SHAREABILITY_SHIFT		(10)
+#define GICR_PENDBASER_INNER_CACHEABILITY_SHIFT		(7)
+#define GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT		(56)
+#define GICR_PENDBASER_SHAREABILITY_MASK				\
+	GIC_BASER_SHAREABILITY(GICR_PENDBASER, SHAREABILITY_MASK)
+#define GICR_PENDBASER_INNER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, MASK)
+#define GICR_PENDBASER_OUTER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GICR_PENDBASER, OUTER, MASK)
+#define GICR_PENDBASER_CACHEABILITY_MASK GICR_PENDBASER_INNER_CACHEABILITY_MASK
+
+#define GICR_PENDBASER_InnerShareable					\
+	GIC_BASER_SHAREABILITY(GICR_PENDBASER, InnerShareable)
+
+#define GICR_PENDBASER_nCnB	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, nCnB)
+#define GICR_PENDBASER_nC 	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, nC)
+#define GICR_PENDBASER_RaWt	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWt)
+#define GICR_PENDBASER_RaWb	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWt)
+#define GICR_PENDBASER_WaWt	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, WaWt)
+#define GICR_PENDBASER_WaWb	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, WaWb)
+#define GICR_PENDBASER_RaWaWt	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWaWt)
+#define GICR_PENDBASER_RaWaWb	GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWaWb)
+
+#define GICR_PENDBASER_PTZ				BIT_ULL(62)
 
 /*
  * Re-Distributor registers, offsets from SGI_base
@@ -175,59 +217,90 @@
 #define GITS_CWRITER			0x0088
 #define GITS_CREADR			0x0090
 #define GITS_BASER			0x0100
+#define GITS_IDREGS_BASE		0xffd0
+#define GITS_PIDR0			0xffe0
+#define GITS_PIDR1			0xffe4
 #define GITS_PIDR2			GICR_PIDR2
+#define GITS_PIDR4			0xffd0
+#define GITS_CIDR0			0xfff0
+#define GITS_CIDR1			0xfff4
+#define GITS_CIDR2			0xfff8
+#define GITS_CIDR3			0xfffc
 
 #define GITS_TRANSLATER			0x10040
 
 #define GITS_CTLR_ENABLE		(1U << 0)
 #define GITS_CTLR_QUIESCENT		(1U << 31)
 
+#define GITS_TYPER_PLPIS		(1UL << 0)
+#define GITS_TYPER_IDBITS_SHIFT		8
 #define GITS_TYPER_DEVBITS_SHIFT	13
 #define GITS_TYPER_DEVBITS(r)		((((r) >> GITS_TYPER_DEVBITS_SHIFT) & 0x1f) + 1)
 #define GITS_TYPER_PTA			(1UL << 19)
-
-#define GITS_CBASER_VALID		(1UL << 63)
-#define GITS_CBASER_nCnB		(0UL << 59)
-#define GITS_CBASER_nC			(1UL << 59)
-#define GITS_CBASER_RaWt		(2UL << 59)
-#define GITS_CBASER_RaWb		(3UL << 59)
-#define GITS_CBASER_WaWt		(4UL << 59)
-#define GITS_CBASER_WaWb		(5UL << 59)
-#define GITS_CBASER_RaWaWt		(6UL << 59)
-#define GITS_CBASER_RaWaWb		(7UL << 59)
-#define GITS_CBASER_CACHEABILITY_MASK	(7UL << 59)
-#define GITS_CBASER_NonShareable	(0UL << 10)
-#define GITS_CBASER_InnerShareable	(1UL << 10)
-#define GITS_CBASER_OuterShareable	(2UL << 10)
-#define GITS_CBASER_SHAREABILITY_MASK	(3UL << 10)
+#define GITS_TYPER_HWCOLLCNT_SHIFT	24
+
+#define GITS_CBASER_VALID			(1UL << 63)
+#define GITS_CBASER_SHAREABILITY_SHIFT		(10)
+#define GITS_CBASER_INNER_CACHEABILITY_SHIFT	(59)
+#define GITS_CBASER_OUTER_CACHEABILITY_SHIFT	(53)
+#define GITS_CBASER_SHAREABILITY_MASK					\
+	GIC_BASER_SHAREABILITY(GITS_CBASER, SHAREABILITY_MASK)
+#define GITS_CBASER_INNER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, MASK)
+#define GITS_CBASER_OUTER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GITS_CBASER, OUTER, MASK)
+#define GITS_CBASER_CACHEABILITY_MASK GITS_CBASER_INNER_CACHEABILITY_MASK
+
+#define GITS_CBASER_InnerShareable					\
+	GIC_BASER_SHAREABILITY(GITS_CBASER, InnerShareable)
+
+#define GITS_CBASER_nCnB	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, nCnB)
+#define GITS_CBASER_nC		GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, nC)
+#define GITS_CBASER_RaWt	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWt)
+#define GITS_CBASER_RaWb	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWt)
+#define GITS_CBASER_WaWt	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, WaWt)
+#define GITS_CBASER_WaWb	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, WaWb)
+#define GITS_CBASER_RaWaWt	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWaWt)
+#define GITS_CBASER_RaWaWb	GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWaWb)
 
 #define GITS_BASER_NR_REGS		8
 
-#define GITS_BASER_VALID		(1UL << 63)
-#define GITS_BASER_nCnB			(0UL << 59)
-#define GITS_BASER_nC			(1UL << 59)
-#define GITS_BASER_RaWt			(2UL << 59)
-#define GITS_BASER_RaWb			(3UL << 59)
-#define GITS_BASER_WaWt			(4UL << 59)
-#define GITS_BASER_WaWb			(5UL << 59)
-#define GITS_BASER_RaWaWt		(6UL << 59)
-#define GITS_BASER_RaWaWb		(7UL << 59)
-#define GITS_BASER_CACHEABILITY_MASK	(7UL << 59)
-#define GITS_BASER_TYPE_SHIFT		(56)
+#define GITS_BASER_VALID			(1UL << 63)
+#define GITS_BASER_INDIRECT			(1ULL << 62)
+
+#define GITS_BASER_INNER_CACHEABILITY_SHIFT	(59)
+#define GITS_BASER_OUTER_CACHEABILITY_SHIFT	(53)
+#define GITS_BASER_INNER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GITS_BASER, INNER, MASK)
+#define GITS_BASER_CACHEABILITY_MASK		GITS_BASER_INNER_CACHEABILITY_MASK
+#define GITS_BASER_OUTER_CACHEABILITY_MASK				\
+	GIC_BASER_CACHEABILITY(GITS_BASER, OUTER, MASK)
+#define GITS_BASER_SHAREABILITY_MASK					\
+	GIC_BASER_SHAREABILITY(GITS_BASER, SHAREABILITY_MASK)
+
+#define GITS_BASER_nCnB		GIC_BASER_CACHEABILITY(GITS_BASER, INNER, nCnB)
+#define GITS_BASER_nC		GIC_BASER_CACHEABILITY(GITS_BASER, INNER, nC)
+#define GITS_BASER_RaWt		GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWt)
+#define GITS_BASER_RaWb		GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWt)
+#define GITS_BASER_WaWt		GIC_BASER_CACHEABILITY(GITS_BASER, INNER, WaWt)
+#define GITS_BASER_WaWb		GIC_BASER_CACHEABILITY(GITS_BASER, INNER, WaWb)
+#define GITS_BASER_RaWaWt	GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWaWt)
+#define GITS_BASER_RaWaWb	GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWaWb)
+
+#define GITS_BASER_TYPE_SHIFT			(56)
 #define GITS_BASER_TYPE(r)		(((r) >> GITS_BASER_TYPE_SHIFT) & 7)
-#define GITS_BASER_ENTRY_SIZE_SHIFT	(48)
+#define GITS_BASER_ENTRY_SIZE_SHIFT		(48)
 #define GITS_BASER_ENTRY_SIZE(r)	((((r) >> GITS_BASER_ENTRY_SIZE_SHIFT) & 0xff) + 1)
-#define GITS_BASER_NonShareable		(0UL << 10)
-#define GITS_BASER_InnerShareable	(1UL << 10)
-#define GITS_BASER_OuterShareable	(2UL << 10)
 #define GITS_BASER_SHAREABILITY_SHIFT	(10)
-#define GITS_BASER_SHAREABILITY_MASK	(3UL << GITS_BASER_SHAREABILITY_SHIFT)
+#define GITS_BASER_InnerShareable					\
+	GIC_BASER_SHAREABILITY(GITS_BASER, InnerShareable)
 #define GITS_BASER_PAGE_SIZE_SHIFT	(8)
 #define GITS_BASER_PAGE_SIZE_4K		(0UL << GITS_BASER_PAGE_SIZE_SHIFT)
 #define GITS_BASER_PAGE_SIZE_16K	(1UL << GITS_BASER_PAGE_SIZE_SHIFT)
 #define GITS_BASER_PAGE_SIZE_64K	(2UL << GITS_BASER_PAGE_SIZE_SHIFT)
 #define GITS_BASER_PAGE_SIZE_MASK	(3UL << GITS_BASER_PAGE_SIZE_SHIFT)
 #define GITS_BASER_PAGES_MAX		256
+#define GITS_BASER_NR_PAGES(r)		(((r) & 0xff) + 1)
 
 #define GITS_BASER_TYPE_NONE		0
 #define GITS_BASER_TYPE_DEVICE		1
@@ -243,7 +316,10 @@
  */
 #define GITS_CMD_MAPD			0x08
 #define GITS_CMD_MAPC			0x09
-#define GITS_CMD_MAPVI			0x0a
+#define GITS_CMD_MAPTI			0x0a
+/* older GIC documentation used MAPVI for this command */
+#define GITS_CMD_MAPVI			GITS_CMD_MAPTI
+#define GITS_CMD_MAPI			0x0b
 #define GITS_CMD_MOVI			0x01
 #define GITS_CMD_DISCARD		0x0f
 #define GITS_CMD_INV			0x0c
@@ -253,6 +329,22 @@
 #define GITS_CMD_CLEAR			0x04
 #define GITS_CMD_SYNC			0x05
 
+/*
+ * ITS error numbers
+ */
+#define E_ITS_MOVI_UNMAPPED_INTERRUPT		0x010107
+#define E_ITS_MOVI_UNMAPPED_COLLECTION		0x010109
+#define E_ITS_CLEAR_UNMAPPED_INTERRUPT		0x010507
+#define E_ITS_MAPD_DEVICE_OOR			0x010801
+#define E_ITS_MAPC_PROCNUM_OOR			0x010902
+#define E_ITS_MAPC_COLLECTION_OOR		0x010903
+#define E_ITS_MAPTI_UNMAPPED_DEVICE		0x010a04
+#define E_ITS_MAPTI_PHYSICALID_OOR		0x010a06
+#define E_ITS_INV_UNMAPPED_INTERRUPT		0x010c07
+#define E_ITS_INVALL_UNMAPPED_COLLECTION	0x010d09
+#define E_ITS_MOVALL_PROCNUM_OOR		0x010e01
+#define E_ITS_DISCARD_UNMAPPED_INTERRUPT	0x010f07
+
 /*
  * CPU interface registers
  */

include/linux/kvm_host.h

@@ -164,6 +164,8 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
 			    int len, struct kvm_io_device *dev);
 int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
 			      struct kvm_io_device *dev);
+struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
+					 gpa_t addr);
 
 #ifdef CONFIG_KVM_ASYNC_PF
 struct kvm_async_pf {

include/uapi/linux/kvm.h

@@ -868,6 +868,7 @@ struct kvm_ppc_smmu_info {
 #define KVM_CAP_MAX_VCPU_ID 128
 #define KVM_CAP_X2APIC_API 129
 #define KVM_CAP_S390_USER_INSTR0 130
+#define KVM_CAP_MSI_DEVID 131
 
 #ifdef KVM_CAP_IRQ_ROUTING
 
@@ -1026,12 +1027,14 @@ struct kvm_one_reg {
 	__u64 addr;
 };
 
+#define KVM_MSI_VALID_DEVID	(1U << 0)
 struct kvm_msi {
 	__u32 address_lo;
 	__u32 address_hi;
 	__u32 data;
 	__u32 flags;
-	__u8  pad[16];
+	__u32 devid;
+	__u8  pad[12];
 };
 
 struct kvm_arm_device_addr {
@@ -1076,6 +1079,8 @@ enum kvm_device_type {
 #define KVM_DEV_TYPE_FLIC		KVM_DEV_TYPE_FLIC
 	KVM_DEV_TYPE_ARM_VGIC_V3,
 #define KVM_DEV_TYPE_ARM_VGIC_V3	KVM_DEV_TYPE_ARM_VGIC_V3
+	KVM_DEV_TYPE_ARM_VGIC_ITS,
+#define KVM_DEV_TYPE_ARM_VGIC_ITS	KVM_DEV_TYPE_ARM_VGIC_ITS
 	KVM_DEV_TYPE_MAX,
 };
 

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

@@ -21,18 +21,11 @@
 
 #include <asm/kvm_hyp.h>
 
-#ifdef CONFIG_KVM_NEW_VGIC
-extern struct vgic_global kvm_vgic_global_state;
-#define vgic_v2_params kvm_vgic_global_state
-#else
-extern struct vgic_params vgic_v2_params;
-#endif
-
 static void __hyp_text save_maint_int_state(struct kvm_vcpu *vcpu,
 					    void __iomem *base)
 {
 	struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
-	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+	int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
 	u32 eisr0, eisr1;
 	int i;
 	bool expect_mi;
@@ -74,7 +67,7 @@ static void __hyp_text save_maint_int_state(struct kvm_vcpu *vcpu,
 static void __hyp_text save_elrsr(struct kvm_vcpu *vcpu, void __iomem *base)
 {
 	struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
-	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+	int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
 	u32 elrsr0, elrsr1;
 
 	elrsr0 = readl_relaxed(base + GICH_ELRSR0);
@@ -93,7 +86,7 @@ static void __hyp_text save_elrsr(struct kvm_vcpu *vcpu, void __iomem *base)
 static void __hyp_text save_lrs(struct kvm_vcpu *vcpu, void __iomem *base)
 {
 	struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
-	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+	int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
 	int i;
 
 	for (i = 0; i < nr_lr; i++) {
@@ -147,7 +140,7 @@ void __hyp_text __vgic_v2_restore_state(struct kvm_vcpu *vcpu)
 	struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
 	struct vgic_dist *vgic = &kvm->arch.vgic;
 	void __iomem *base = kern_hyp_va(vgic->vctrl_base);
-	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+	int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
 	int i;
 	u64 live_lrs = 0;
 

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

@@ -1,856 +0,0 @@
-/*
- * Contains GICv2 specific emulation code, was in vgic.c before.
- *
- * Copyright (C) 2012 ARM Ltd.
- * Author: Marc Zyngier <marc.zyngier@arm.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-
-#include <linux/cpu.h>
-#include <linux/kvm.h>
-#include <linux/kvm_host.h>
-#include <linux/interrupt.h>
-#include <linux/io.h>
-#include <linux/uaccess.h>
-
-#include <linux/irqchip/arm-gic.h>
-
-#include <asm/kvm_emulate.h>
-#include <asm/kvm_arm.h>
-#include <asm/kvm_mmu.h>
-
-#include "vgic.h"
-
-#define GICC_ARCH_VERSION_V2		0x2
-
-static void vgic_dispatch_sgi(struct kvm_vcpu *vcpu, u32 reg);
-static u8 *vgic_get_sgi_sources(struct vgic_dist *dist, int vcpu_id, int sgi)
-{
-	return dist->irq_sgi_sources + vcpu_id * VGIC_NR_SGIS + sgi;
-}
-
-static bool handle_mmio_misc(struct kvm_vcpu *vcpu,
-			     struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 reg;
-	u32 word_offset = offset & 3;
-
-	switch (offset & ~3) {
-	case 0:			/* GICD_CTLR */
-		reg = vcpu->kvm->arch.vgic.enabled;
-		vgic_reg_access(mmio, &reg, word_offset,
-				ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-		if (mmio->is_write) {
-			vcpu->kvm->arch.vgic.enabled = reg & 1;
-			vgic_update_state(vcpu->kvm);
-			return true;
-		}
-		break;
-
-	case 4:			/* GICD_TYPER */
-		reg  = (atomic_read(&vcpu->kvm->online_vcpus) - 1) << 5;
-		reg |= (vcpu->kvm->arch.vgic.nr_irqs >> 5) - 1;
-		vgic_reg_access(mmio, &reg, word_offset,
-				ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-		break;
-
-	case 8:			/* GICD_IIDR */
-		reg = (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
-		vgic_reg_access(mmio, &reg, word_offset,
-				ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-		break;
-	}
-
-	return false;
-}
-
-static bool handle_mmio_set_enable_reg(struct kvm_vcpu *vcpu,
-				       struct kvm_exit_mmio *mmio,
-				       phys_addr_t offset)
-{
-	return vgic_handle_enable_reg(vcpu->kvm, mmio, offset,
-				      vcpu->vcpu_id, ACCESS_WRITE_SETBIT);
-}
-
-static bool handle_mmio_clear_enable_reg(struct kvm_vcpu *vcpu,
-					 struct kvm_exit_mmio *mmio,
-					 phys_addr_t offset)
-{
-	return vgic_handle_enable_reg(vcpu->kvm, mmio, offset,
-				      vcpu->vcpu_id, ACCESS_WRITE_CLEARBIT);
-}
-
-static bool handle_mmio_set_pending_reg(struct kvm_vcpu *vcpu,
-					struct kvm_exit_mmio *mmio,
-					phys_addr_t offset)
-{
-	return vgic_handle_set_pending_reg(vcpu->kvm, mmio, offset,
-					   vcpu->vcpu_id);
-}
-
-static bool handle_mmio_clear_pending_reg(struct kvm_vcpu *vcpu,
-					  struct kvm_exit_mmio *mmio,
-					  phys_addr_t offset)
-{
-	return vgic_handle_clear_pending_reg(vcpu->kvm, mmio, offset,
-					     vcpu->vcpu_id);
-}
-
-static bool handle_mmio_set_active_reg(struct kvm_vcpu *vcpu,
-				       struct kvm_exit_mmio *mmio,
-				       phys_addr_t offset)
-{
-	return vgic_handle_set_active_reg(vcpu->kvm, mmio, offset,
-					  vcpu->vcpu_id);
-}
-
-static bool handle_mmio_clear_active_reg(struct kvm_vcpu *vcpu,
-					 struct kvm_exit_mmio *mmio,
-					 phys_addr_t offset)
-{
-	return vgic_handle_clear_active_reg(vcpu->kvm, mmio, offset,
-					    vcpu->vcpu_id);
-}
-
-static bool handle_mmio_priority_reg(struct kvm_vcpu *vcpu,
-				     struct kvm_exit_mmio *mmio,
-				     phys_addr_t offset)
-{
-	u32 *reg = vgic_bytemap_get_reg(&vcpu->kvm->arch.vgic.irq_priority,
-					vcpu->vcpu_id, offset);
-	vgic_reg_access(mmio, reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-	return false;
-}
-
-#define GICD_ITARGETSR_SIZE	32
-#define GICD_CPUTARGETS_BITS	8
-#define GICD_IRQS_PER_ITARGETSR	(GICD_ITARGETSR_SIZE / GICD_CPUTARGETS_BITS)
-static u32 vgic_get_target_reg(struct kvm *kvm, int irq)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	int i;
-	u32 val = 0;
-
-	irq -= VGIC_NR_PRIVATE_IRQS;
-
-	for (i = 0; i < GICD_IRQS_PER_ITARGETSR; i++)
-		val |= 1 << (dist->irq_spi_cpu[irq + i] + i * 8);
-
-	return val;
-}
-
-static void vgic_set_target_reg(struct kvm *kvm, u32 val, int irq)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	struct kvm_vcpu *vcpu;
-	int i, c;
-	unsigned long *bmap;
-	u32 target;
-
-	irq -= VGIC_NR_PRIVATE_IRQS;
-
-	/*
-	 * Pick the LSB in each byte. This ensures we target exactly
-	 * one vcpu per IRQ. If the byte is null, assume we target
-	 * CPU0.
-	 */
-	for (i = 0; i < GICD_IRQS_PER_ITARGETSR; i++) {
-		int shift = i * GICD_CPUTARGETS_BITS;
-
-		target = ffs((val >> shift) & 0xffU);
-		target = target ? (target - 1) : 0;
-		dist->irq_spi_cpu[irq + i] = target;
-		kvm_for_each_vcpu(c, vcpu, kvm) {
-			bmap = vgic_bitmap_get_shared_map(&dist->irq_spi_target[c]);
-			if (c == target)
-				set_bit(irq + i, bmap);
-			else
-				clear_bit(irq + i, bmap);
-		}
-	}
-}
-
-static bool handle_mmio_target_reg(struct kvm_vcpu *vcpu,
-				   struct kvm_exit_mmio *mmio,
-				   phys_addr_t offset)
-{
-	u32 reg;
-
-	/* We treat the banked interrupts targets as read-only */
-	if (offset < 32) {
-		u32 roreg;
-
-		roreg = 1 << vcpu->vcpu_id;
-		roreg |= roreg << 8;
-		roreg |= roreg << 16;
-
-		vgic_reg_access(mmio, &roreg, offset,
-				ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-		return false;
-	}
-
-	reg = vgic_get_target_reg(vcpu->kvm, offset & ~3U);
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-	if (mmio->is_write) {
-		vgic_set_target_reg(vcpu->kvm, reg, offset & ~3U);
-		vgic_update_state(vcpu->kvm);
-		return true;
-	}
-
-	return false;
-}
-
-static bool handle_mmio_cfg_reg(struct kvm_vcpu *vcpu,
-				struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 *reg;
-
-	reg = vgic_bitmap_get_reg(&vcpu->kvm->arch.vgic.irq_cfg,
-				  vcpu->vcpu_id, offset >> 1);
-
-	return vgic_handle_cfg_reg(reg, mmio, offset);
-}
-
-static bool handle_mmio_sgi_reg(struct kvm_vcpu *vcpu,
-				struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 reg;
-
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_VALUE);
-	if (mmio->is_write) {
-		vgic_dispatch_sgi(vcpu, reg);
-		vgic_update_state(vcpu->kvm);
-		return true;
-	}
-
-	return false;
-}
-
-/* Handle reads of GICD_CPENDSGIRn and GICD_SPENDSGIRn */
-static bool read_set_clear_sgi_pend_reg(struct kvm_vcpu *vcpu,
-					struct kvm_exit_mmio *mmio,
-					phys_addr_t offset)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	int sgi;
-	int min_sgi = (offset & ~0x3);
-	int max_sgi = min_sgi + 3;
-	int vcpu_id = vcpu->vcpu_id;
-	u32 reg = 0;
-
-	/* Copy source SGIs from distributor side */
-	for (sgi = min_sgi; sgi <= max_sgi; sgi++) {
-		u8 sources = *vgic_get_sgi_sources(dist, vcpu_id, sgi);
-
-		reg |= ((u32)sources) << (8 * (sgi - min_sgi));
-	}
-
-	mmio_data_write(mmio, ~0, reg);
-	return false;
-}
-
-static bool write_set_clear_sgi_pend_reg(struct kvm_vcpu *vcpu,
-					 struct kvm_exit_mmio *mmio,
-					 phys_addr_t offset, bool set)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	int sgi;
-	int min_sgi = (offset & ~0x3);
-	int max_sgi = min_sgi + 3;
-	int vcpu_id = vcpu->vcpu_id;
-	u32 reg;
-	bool updated = false;
-
-	reg = mmio_data_read(mmio, ~0);
-
-	/* Clear pending SGIs on the distributor */
-	for (sgi = min_sgi; sgi <= max_sgi; sgi++) {
-		u8 mask = reg >> (8 * (sgi - min_sgi));
-		u8 *src = vgic_get_sgi_sources(dist, vcpu_id, sgi);
-
-		if (set) {
-			if ((*src & mask) != mask)
-				updated = true;
-			*src |= mask;
-		} else {
-			if (*src & mask)
-				updated = true;
-			*src &= ~mask;
-		}
-	}
-
-	if (updated)
-		vgic_update_state(vcpu->kvm);
-
-	return updated;
-}
-
-static bool handle_mmio_sgi_set(struct kvm_vcpu *vcpu,
-				struct kvm_exit_mmio *mmio,
-				phys_addr_t offset)
-{
-	if (!mmio->is_write)
-		return read_set_clear_sgi_pend_reg(vcpu, mmio, offset);
-	else
-		return write_set_clear_sgi_pend_reg(vcpu, mmio, offset, true);
-}
-
-static bool handle_mmio_sgi_clear(struct kvm_vcpu *vcpu,
-				  struct kvm_exit_mmio *mmio,
-				  phys_addr_t offset)
-{
-	if (!mmio->is_write)
-		return read_set_clear_sgi_pend_reg(vcpu, mmio, offset);
-	else
-		return write_set_clear_sgi_pend_reg(vcpu, mmio, offset, false);
-}
-
-static const struct vgic_io_range vgic_dist_ranges[] = {
-	{
-		.base		= GIC_DIST_SOFTINT,
-		.len		= 4,
-		.handle_mmio	= handle_mmio_sgi_reg,
-	},
-	{
-		.base		= GIC_DIST_CTRL,
-		.len		= 12,
-		.bits_per_irq	= 0,
-		.handle_mmio	= handle_mmio_misc,
-	},
-	{
-		.base		= GIC_DIST_IGROUP,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		.base		= GIC_DIST_ENABLE_SET,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_enable_reg,
-	},
-	{
-		.base		= GIC_DIST_ENABLE_CLEAR,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_enable_reg,
-	},
-	{
-		.base		= GIC_DIST_PENDING_SET,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_pending_reg,
-	},
-	{
-		.base		= GIC_DIST_PENDING_CLEAR,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_pending_reg,
-	},
-	{
-		.base		= GIC_DIST_ACTIVE_SET,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_active_reg,
-	},
-	{
-		.base		= GIC_DIST_ACTIVE_CLEAR,
-		.len		= VGIC_MAX_IRQS / 8,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_active_reg,
-	},
-	{
-		.base		= GIC_DIST_PRI,
-		.len		= VGIC_MAX_IRQS,
-		.bits_per_irq	= 8,
-		.handle_mmio	= handle_mmio_priority_reg,
-	},
-	{
-		.base		= GIC_DIST_TARGET,
-		.len		= VGIC_MAX_IRQS,
-		.bits_per_irq	= 8,
-		.handle_mmio	= handle_mmio_target_reg,
-	},
-	{
-		.base		= GIC_DIST_CONFIG,
-		.len		= VGIC_MAX_IRQS / 4,
-		.bits_per_irq	= 2,
-		.handle_mmio	= handle_mmio_cfg_reg,
-	},
-	{
-		.base		= GIC_DIST_SGI_PENDING_CLEAR,
-		.len		= VGIC_NR_SGIS,
-		.handle_mmio	= handle_mmio_sgi_clear,
-	},
-	{
-		.base		= GIC_DIST_SGI_PENDING_SET,
-		.len		= VGIC_NR_SGIS,
-		.handle_mmio	= handle_mmio_sgi_set,
-	},
-	{}
-};
-
-static void vgic_dispatch_sgi(struct kvm_vcpu *vcpu, u32 reg)
-{
-	struct kvm *kvm = vcpu->kvm;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	int nrcpus = atomic_read(&kvm->online_vcpus);
-	u8 target_cpus;
-	int sgi, mode, c, vcpu_id;
-
-	vcpu_id = vcpu->vcpu_id;
-
-	sgi = reg & 0xf;
-	target_cpus = (reg >> 16) & 0xff;
-	mode = (reg >> 24) & 3;
-
-	switch (mode) {
-	case 0:
-		if (!target_cpus)
-			return;
-		break;
-
-	case 1:
-		target_cpus = ((1 << nrcpus) - 1) & ~(1 << vcpu_id) & 0xff;
-		break;
-
-	case 2:
-		target_cpus = 1 << vcpu_id;
-		break;
-	}
-
-	kvm_for_each_vcpu(c, vcpu, kvm) {
-		if (target_cpus & 1) {
-			/* Flag the SGI as pending */
-			vgic_dist_irq_set_pending(vcpu, sgi);
-			*vgic_get_sgi_sources(dist, c, sgi) |= 1 << vcpu_id;
-			kvm_debug("SGI%d from CPU%d to CPU%d\n",
-				  sgi, vcpu_id, c);
-		}
-
-		target_cpus >>= 1;
-	}
-}
-
-static bool vgic_v2_queue_sgi(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	unsigned long sources;
-	int vcpu_id = vcpu->vcpu_id;
-	int c;
-
-	sources = *vgic_get_sgi_sources(dist, vcpu_id, irq);
-
-	for_each_set_bit(c, &sources, dist->nr_cpus) {
-		if (vgic_queue_irq(vcpu, c, irq))
-			clear_bit(c, &sources);
-	}
-
-	*vgic_get_sgi_sources(dist, vcpu_id, irq) = sources;
-
-	/*
-	 * If the sources bitmap has been cleared it means that we
-	 * could queue all the SGIs onto link registers (see the
-	 * clear_bit above), and therefore we are done with them in
-	 * our emulated gic and can get rid of them.
-	 */
-	if (!sources) {
-		vgic_dist_irq_clear_pending(vcpu, irq);
-		vgic_cpu_irq_clear(vcpu, irq);
-		return true;
-	}
-
-	return false;
-}
-
-/**
- * kvm_vgic_map_resources - Configure global VGIC state before running any VCPUs
- * @kvm: pointer to the kvm struct
- *
- * Map the virtual CPU interface into the VM before running any VCPUs.  We
- * can't do this at creation time, because user space must first set the
- * virtual CPU interface address in the guest physical address space.
- */
-static int vgic_v2_map_resources(struct kvm *kvm,
-				 const struct vgic_params *params)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	int ret = 0;
-
-	if (!irqchip_in_kernel(kvm))
-		return 0;
-
-	mutex_lock(&kvm->lock);
-
-	if (vgic_ready(kvm))
-		goto out;
-
-	if (IS_VGIC_ADDR_UNDEF(dist->vgic_dist_base) ||
-	    IS_VGIC_ADDR_UNDEF(dist->vgic_cpu_base)) {
-		kvm_err("Need to set vgic cpu and dist addresses first\n");
-		ret = -ENXIO;
-		goto out;
-	}
-
-	vgic_register_kvm_io_dev(kvm, dist->vgic_dist_base,
-				 KVM_VGIC_V2_DIST_SIZE,
-				 vgic_dist_ranges, -1, &dist->dist_iodev);
-
-	/*
-	 * Initialize the vgic if this hasn't already been done on demand by
-	 * accessing the vgic state from userspace.
-	 */
-	ret = vgic_init(kvm);
-	if (ret) {
-		kvm_err("Unable to allocate maps\n");
-		goto out_unregister;
-	}
-
-	ret = kvm_phys_addr_ioremap(kvm, dist->vgic_cpu_base,
-				    params->vcpu_base, KVM_VGIC_V2_CPU_SIZE,
-				    true);
-	if (ret) {
-		kvm_err("Unable to remap VGIC CPU to VCPU\n");
-		goto out_unregister;
-	}
-
-	dist->ready = true;
-	goto out;
-
-out_unregister:
-	kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS, &dist->dist_iodev.dev);
-
-out:
-	if (ret)
-		kvm_vgic_destroy(kvm);
-	mutex_unlock(&kvm->lock);
-	return ret;
-}
-
-static void vgic_v2_add_sgi_source(struct kvm_vcpu *vcpu, int irq, int source)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	*vgic_get_sgi_sources(dist, vcpu->vcpu_id, irq) |= 1 << source;
-}
-
-static int vgic_v2_init_model(struct kvm *kvm)
-{
-	int i;
-
-	for (i = VGIC_NR_PRIVATE_IRQS; i < kvm->arch.vgic.nr_irqs; i += 4)
-		vgic_set_target_reg(kvm, 0, i);
-
-	return 0;
-}
-
-void vgic_v2_init_emulation(struct kvm *kvm)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-
-	dist->vm_ops.queue_sgi = vgic_v2_queue_sgi;
-	dist->vm_ops.add_sgi_source = vgic_v2_add_sgi_source;
-	dist->vm_ops.init_model = vgic_v2_init_model;
-	dist->vm_ops.map_resources = vgic_v2_map_resources;
-
-	kvm->arch.max_vcpus = VGIC_V2_MAX_CPUS;
-}
-
-static bool handle_cpu_mmio_misc(struct kvm_vcpu *vcpu,
-				 struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	bool updated = false;
-	struct vgic_vmcr vmcr;
-	u32 *vmcr_field;
-	u32 reg;
-
-	vgic_get_vmcr(vcpu, &vmcr);
-
-	switch (offset & ~0x3) {
-	case GIC_CPU_CTRL:
-		vmcr_field = &vmcr.ctlr;
-		break;
-	case GIC_CPU_PRIMASK:
-		vmcr_field = &vmcr.pmr;
-		break;
-	case GIC_CPU_BINPOINT:
-		vmcr_field = &vmcr.bpr;
-		break;
-	case GIC_CPU_ALIAS_BINPOINT:
-		vmcr_field = &vmcr.abpr;
-		break;
-	default:
-		BUG();
-	}
-
-	if (!mmio->is_write) {
-		reg = *vmcr_field;
-		mmio_data_write(mmio, ~0, reg);
-	} else {
-		reg = mmio_data_read(mmio, ~0);
-		if (reg != *vmcr_field) {
-			*vmcr_field = reg;
-			vgic_set_vmcr(vcpu, &vmcr);
-			updated = true;
-		}
-	}
-	return updated;
-}
-
-static bool handle_mmio_abpr(struct kvm_vcpu *vcpu,
-			     struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	return handle_cpu_mmio_misc(vcpu, mmio, GIC_CPU_ALIAS_BINPOINT);
-}
-
-static bool handle_cpu_mmio_ident(struct kvm_vcpu *vcpu,
-				  struct kvm_exit_mmio *mmio,
-				  phys_addr_t offset)
-{
-	u32 reg;
-
-	if (mmio->is_write)
-		return false;
-
-	/* GICC_IIDR */
-	reg = (PRODUCT_ID_KVM << 20) |
-	      (GICC_ARCH_VERSION_V2 << 16) |
-	      (IMPLEMENTER_ARM << 0);
-	mmio_data_write(mmio, ~0, reg);
-	return false;
-}
-
-/*
- * CPU Interface Register accesses - these are not accessed by the VM, but by
- * user space for saving and restoring VGIC state.
- */
-static const struct vgic_io_range vgic_cpu_ranges[] = {
-	{
-		.base		= GIC_CPU_CTRL,
-		.len		= 12,
-		.handle_mmio	= handle_cpu_mmio_misc,
-	},
-	{
-		.base		= GIC_CPU_ALIAS_BINPOINT,
-		.len		= 4,
-		.handle_mmio	= handle_mmio_abpr,
-	},
-	{
-		.base		= GIC_CPU_ACTIVEPRIO,
-		.len		= 16,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		.base		= GIC_CPU_IDENT,
-		.len		= 4,
-		.handle_mmio	= handle_cpu_mmio_ident,
-	},
-};
-
-static int vgic_attr_regs_access(struct kvm_device *dev,
-				 struct kvm_device_attr *attr,
-				 u32 *reg, bool is_write)
-{
-	const struct vgic_io_range *r = NULL, *ranges;
-	phys_addr_t offset;
-	int ret, cpuid, c;
-	struct kvm_vcpu *vcpu, *tmp_vcpu;
-	struct vgic_dist *vgic;
-	struct kvm_exit_mmio mmio;
-	u32 data;
-
-	offset = attr->attr & KVM_DEV_ARM_VGIC_OFFSET_MASK;
-	cpuid = (attr->attr & KVM_DEV_ARM_VGIC_CPUID_MASK) >>
-		KVM_DEV_ARM_VGIC_CPUID_SHIFT;
-
-	mutex_lock(&dev->kvm->lock);
-
-	ret = vgic_init(dev->kvm);
-	if (ret)
-		goto out;
-
-	if (cpuid >= atomic_read(&dev->kvm->online_vcpus)) {
-		ret = -EINVAL;
-		goto out;
-	}
-
-	vcpu = kvm_get_vcpu(dev->kvm, cpuid);
-	vgic = &dev->kvm->arch.vgic;
-
-	mmio.len = 4;
-	mmio.is_write = is_write;
-	mmio.data = &data;
-	if (is_write)
-		mmio_data_write(&mmio, ~0, *reg);
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-		mmio.phys_addr = vgic->vgic_dist_base + offset;
-		ranges = vgic_dist_ranges;
-		break;
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS:
-		mmio.phys_addr = vgic->vgic_cpu_base + offset;
-		ranges = vgic_cpu_ranges;
-		break;
-	default:
-		BUG();
-	}
-	r = vgic_find_range(ranges, 4, offset);
-
-	if (unlikely(!r || !r->handle_mmio)) {
-		ret = -ENXIO;
-		goto out;
-	}
-
-
-	spin_lock(&vgic->lock);
-
-	/*
-	 * Ensure that no other VCPU is running by checking the vcpu->cpu
-	 * field.  If no other VPCUs are running we can safely access the VGIC
-	 * state, because even if another VPU is run after this point, that
-	 * VCPU will not touch the vgic state, because it will block on
-	 * getting the vgic->lock in kvm_vgic_sync_hwstate().
-	 */
-	kvm_for_each_vcpu(c, tmp_vcpu, dev->kvm) {
-		if (unlikely(tmp_vcpu->cpu != -1)) {
-			ret = -EBUSY;
-			goto out_vgic_unlock;
-		}
-	}
-
-	/*
-	 * Move all pending IRQs from the LRs on all VCPUs so the pending
-	 * state can be properly represented in the register state accessible
-	 * through this API.
-	 */
-	kvm_for_each_vcpu(c, tmp_vcpu, dev->kvm)
-		vgic_unqueue_irqs(tmp_vcpu);
-
-	offset -= r->base;
-	r->handle_mmio(vcpu, &mmio, offset);
-
-	if (!is_write)
-		*reg = mmio_data_read(&mmio, ~0);
-
-	ret = 0;
-out_vgic_unlock:
-	spin_unlock(&vgic->lock);
-out:
-	mutex_unlock(&dev->kvm->lock);
-	return ret;
-}
-
-static int vgic_v2_create(struct kvm_device *dev, u32 type)
-{
-	return kvm_vgic_create(dev->kvm, type);
-}
-
-static void vgic_v2_destroy(struct kvm_device *dev)
-{
-	kfree(dev);
-}
-
-static int vgic_v2_set_attr(struct kvm_device *dev,
-			    struct kvm_device_attr *attr)
-{
-	int ret;
-
-	ret = vgic_set_common_attr(dev, attr);
-	if (ret != -ENXIO)
-		return ret;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS: {
-		u32 __user *uaddr = (u32 __user *)(long)attr->addr;
-		u32 reg;
-
-		if (get_user(reg, uaddr))
-			return -EFAULT;
-
-		return vgic_attr_regs_access(dev, attr, &reg, true);
-	}
-
-	}
-
-	return -ENXIO;
-}
-
-static int vgic_v2_get_attr(struct kvm_device *dev,
-			    struct kvm_device_attr *attr)
-{
-	int ret;
-
-	ret = vgic_get_common_attr(dev, attr);
-	if (ret != -ENXIO)
-		return ret;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS: {
-		u32 __user *uaddr = (u32 __user *)(long)attr->addr;
-		u32 reg = 0;
-
-		ret = vgic_attr_regs_access(dev, attr, &reg, false);
-		if (ret)
-			return ret;
-		return put_user(reg, uaddr);
-	}
-
-	}
-
-	return -ENXIO;
-}
-
-static int vgic_v2_has_attr(struct kvm_device *dev,
-			    struct kvm_device_attr *attr)
-{
-	phys_addr_t offset;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_ADDR:
-		switch (attr->attr) {
-		case KVM_VGIC_V2_ADDR_TYPE_DIST:
-		case KVM_VGIC_V2_ADDR_TYPE_CPU:
-			return 0;
-		}
-		break;
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-		offset = attr->attr & KVM_DEV_ARM_VGIC_OFFSET_MASK;
-		return vgic_has_attr_regs(vgic_dist_ranges, offset);
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS:
-		offset = attr->attr & KVM_DEV_ARM_VGIC_OFFSET_MASK;
-		return vgic_has_attr_regs(vgic_cpu_ranges, offset);
-	case KVM_DEV_ARM_VGIC_GRP_NR_IRQS:
-		return 0;
-	case KVM_DEV_ARM_VGIC_GRP_CTRL:
-		switch (attr->attr) {
-		case KVM_DEV_ARM_VGIC_CTRL_INIT:
-			return 0;
-		}
-	}
-	return -ENXIO;
-}
-
-struct kvm_device_ops kvm_arm_vgic_v2_ops = {
-	.name = "kvm-arm-vgic-v2",
-	.create = vgic_v2_create,
-	.destroy = vgic_v2_destroy,
-	.set_attr = vgic_v2_set_attr,
-	.get_attr = vgic_v2_get_attr,
-	.has_attr = vgic_v2_has_attr,
-};

virt/kvm/arm/vgic-v2.c

@@ -1,274 +0,0 @@
-/*
- * Copyright (C) 2012,2013 ARM Limited, All Rights Reserved.
- * Author: Marc Zyngier <marc.zyngier@arm.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-
-#include <linux/cpu.h>
-#include <linux/kvm.h>
-#include <linux/kvm_host.h>
-#include <linux/interrupt.h>
-#include <linux/io.h>
-
-#include <linux/irqchip/arm-gic.h>
-
-#include <asm/kvm_emulate.h>
-#include <asm/kvm_arm.h>
-#include <asm/kvm_mmu.h>
-
-static struct vgic_lr vgic_v2_get_lr(const struct kvm_vcpu *vcpu, int lr)
-{
-	struct vgic_lr lr_desc;
-	u32 val = vcpu->arch.vgic_cpu.vgic_v2.vgic_lr[lr];
-
-	lr_desc.irq	= val & GICH_LR_VIRTUALID;
-	if (lr_desc.irq <= 15)
-		lr_desc.source	= (val >> GICH_LR_PHYSID_CPUID_SHIFT) & 0x7;
-	else
-		lr_desc.source = 0;
-	lr_desc.state	= 0;
-
-	if (val & GICH_LR_PENDING_BIT)
-		lr_desc.state |= LR_STATE_PENDING;
-	if (val & GICH_LR_ACTIVE_BIT)
-		lr_desc.state |= LR_STATE_ACTIVE;
-	if (val & GICH_LR_EOI)
-		lr_desc.state |= LR_EOI_INT;
-	if (val & GICH_LR_HW) {
-		lr_desc.state |= LR_HW;
-		lr_desc.hwirq = (val & GICH_LR_PHYSID_CPUID) >> GICH_LR_PHYSID_CPUID_SHIFT;
-	}
-
-	return lr_desc;
-}
-
-static void vgic_v2_set_lr(struct kvm_vcpu *vcpu, int lr,
-			   struct vgic_lr lr_desc)
-{
-	u32 lr_val;
-
-	lr_val = lr_desc.irq;
-
-	if (lr_desc.state & LR_STATE_PENDING)
-		lr_val |= GICH_LR_PENDING_BIT;
-	if (lr_desc.state & LR_STATE_ACTIVE)
-		lr_val |= GICH_LR_ACTIVE_BIT;
-	if (lr_desc.state & LR_EOI_INT)
-		lr_val |= GICH_LR_EOI;
-
-	if (lr_desc.state & LR_HW) {
-		lr_val |= GICH_LR_HW;
-		lr_val |= (u32)lr_desc.hwirq << GICH_LR_PHYSID_CPUID_SHIFT;
-	}
-
-	if (lr_desc.irq < VGIC_NR_SGIS)
-		lr_val |= (lr_desc.source << GICH_LR_PHYSID_CPUID_SHIFT);
-
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_lr[lr] = lr_val;
-
-	if (!(lr_desc.state & LR_STATE_MASK))
-		vcpu->arch.vgic_cpu.vgic_v2.vgic_elrsr |= (1ULL << lr);
-	else
-		vcpu->arch.vgic_cpu.vgic_v2.vgic_elrsr &= ~(1ULL << lr);
-}
-
-static u64 vgic_v2_get_elrsr(const struct kvm_vcpu *vcpu)
-{
-	return vcpu->arch.vgic_cpu.vgic_v2.vgic_elrsr;
-}
-
-static u64 vgic_v2_get_eisr(const struct kvm_vcpu *vcpu)
-{
-	return vcpu->arch.vgic_cpu.vgic_v2.vgic_eisr;
-}
-
-static void vgic_v2_clear_eisr(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_eisr = 0;
-}
-
-static u32 vgic_v2_get_interrupt_status(const struct kvm_vcpu *vcpu)
-{
-	u32 misr = vcpu->arch.vgic_cpu.vgic_v2.vgic_misr;
-	u32 ret = 0;
-
-	if (misr & GICH_MISR_EOI)
-		ret |= INT_STATUS_EOI;
-	if (misr & GICH_MISR_U)
-		ret |= INT_STATUS_UNDERFLOW;
-
-	return ret;
-}
-
-static void vgic_v2_enable_underflow(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_hcr |= GICH_HCR_UIE;
-}
-
-static void vgic_v2_disable_underflow(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_hcr &= ~GICH_HCR_UIE;
-}
-
-static void vgic_v2_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
-{
-	u32 vmcr = vcpu->arch.vgic_cpu.vgic_v2.vgic_vmcr;
-
-	vmcrp->ctlr = (vmcr & GICH_VMCR_CTRL_MASK) >> GICH_VMCR_CTRL_SHIFT;
-	vmcrp->abpr = (vmcr & GICH_VMCR_ALIAS_BINPOINT_MASK) >> GICH_VMCR_ALIAS_BINPOINT_SHIFT;
-	vmcrp->bpr  = (vmcr & GICH_VMCR_BINPOINT_MASK) >> GICH_VMCR_BINPOINT_SHIFT;
-	vmcrp->pmr  = (vmcr & GICH_VMCR_PRIMASK_MASK) >> GICH_VMCR_PRIMASK_SHIFT;
-}
-
-static void vgic_v2_set_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
-{
-	u32 vmcr;
-
-	vmcr  = (vmcrp->ctlr << GICH_VMCR_CTRL_SHIFT) & GICH_VMCR_CTRL_MASK;
-	vmcr |= (vmcrp->abpr << GICH_VMCR_ALIAS_BINPOINT_SHIFT) & GICH_VMCR_ALIAS_BINPOINT_MASK;
-	vmcr |= (vmcrp->bpr << GICH_VMCR_BINPOINT_SHIFT) & GICH_VMCR_BINPOINT_MASK;
-	vmcr |= (vmcrp->pmr << GICH_VMCR_PRIMASK_SHIFT) & GICH_VMCR_PRIMASK_MASK;
-
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_vmcr = vmcr;
-}
-
-static void vgic_v2_enable(struct kvm_vcpu *vcpu)
-{
-	/*
-	 * By forcing VMCR to zero, the GIC will restore the binary
-	 * points to their reset values. Anything else resets to zero
-	 * anyway.
-	 */
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_vmcr = 0;
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_elrsr = ~0;
-
-	/* Get the show on the road... */
-	vcpu->arch.vgic_cpu.vgic_v2.vgic_hcr = GICH_HCR_EN;
-}
-
-static const struct vgic_ops vgic_v2_ops = {
-	.get_lr			= vgic_v2_get_lr,
-	.set_lr			= vgic_v2_set_lr,
-	.get_elrsr		= vgic_v2_get_elrsr,
-	.get_eisr		= vgic_v2_get_eisr,
-	.clear_eisr		= vgic_v2_clear_eisr,
-	.get_interrupt_status	= vgic_v2_get_interrupt_status,
-	.enable_underflow	= vgic_v2_enable_underflow,
-	.disable_underflow	= vgic_v2_disable_underflow,
-	.get_vmcr		= vgic_v2_get_vmcr,
-	.set_vmcr		= vgic_v2_set_vmcr,
-	.enable			= vgic_v2_enable,
-};
-
-struct vgic_params __section(.hyp.text) vgic_v2_params;
-
-static void vgic_cpu_init_lrs(void *params)
-{
-	struct vgic_params *vgic = params;
-	int i;
-
-	for (i = 0; i < vgic->nr_lr; i++)
-		writel_relaxed(0, vgic->vctrl_base + GICH_LR0 + (i * 4));
-}
-
-/**
- * vgic_v2_probe - probe for a GICv2 compatible interrupt controller
- * @gic_kvm_info:	pointer to the GIC description
- * @ops:		address of a pointer to the GICv2 operations
- * @params:		address of a pointer to HW-specific parameters
- *
- * Returns 0 if a GICv2 has been found, with the low level operations
- * in *ops and the HW parameters in *params. Returns an error code
- * otherwise.
- */
-int vgic_v2_probe(const struct gic_kvm_info *gic_kvm_info,
-		   const struct vgic_ops **ops,
-		   const struct vgic_params **params)
-{
-	int ret;
-	struct vgic_params *vgic = &vgic_v2_params;
-	const struct resource *vctrl_res = &gic_kvm_info->vctrl;
-	const struct resource *vcpu_res = &gic_kvm_info->vcpu;
-
-	memset(vgic, 0, sizeof(*vgic));
-
-	if (!gic_kvm_info->maint_irq) {
-		kvm_err("error getting vgic maintenance irq\n");
-		ret = -ENXIO;
-		goto out;
-	}
-	vgic->maint_irq = gic_kvm_info->maint_irq;
-
-	if (!gic_kvm_info->vctrl.start) {
-		kvm_err("GICH not present in the firmware table\n");
-		ret = -ENXIO;
-		goto out;
-	}
-
-	vgic->vctrl_base = ioremap(gic_kvm_info->vctrl.start,
-				   resource_size(&gic_kvm_info->vctrl));
-	if (!vgic->vctrl_base) {
-		kvm_err("Cannot ioremap GICH\n");
-		ret = -ENOMEM;
-		goto out;
-	}
-
-	vgic->nr_lr = readl_relaxed(vgic->vctrl_base + GICH_VTR);
-	vgic->nr_lr = (vgic->nr_lr & 0x3f) + 1;
-
-	ret = create_hyp_io_mappings(vgic->vctrl_base,
-				     vgic->vctrl_base + resource_size(vctrl_res),
-				     vctrl_res->start);
-	if (ret) {
-		kvm_err("Cannot map VCTRL into hyp\n");
-		goto out_unmap;
-	}
-
-	if (!PAGE_ALIGNED(vcpu_res->start)) {
-		kvm_err("GICV physical address 0x%llx not page aligned\n",
-			(unsigned long long)vcpu_res->start);
-		ret = -ENXIO;
-		goto out_unmap;
-	}
-
-	if (!PAGE_ALIGNED(resource_size(vcpu_res))) {
-		kvm_err("GICV size 0x%llx not a multiple of page size 0x%lx\n",
-			(unsigned long long)resource_size(vcpu_res),
-			PAGE_SIZE);
-		ret = -ENXIO;
-		goto out_unmap;
-	}
-
-	vgic->can_emulate_gicv2 = true;
-	kvm_register_device_ops(&kvm_arm_vgic_v2_ops, KVM_DEV_TYPE_ARM_VGIC_V2);
-
-	vgic->vcpu_base = vcpu_res->start;
-
-	kvm_info("GICH base=0x%llx, GICV base=0x%llx, IRQ=%d\n",
-		 gic_kvm_info->vctrl.start, vgic->vcpu_base, vgic->maint_irq);
-
-	vgic->type = VGIC_V2;
-	vgic->max_gic_vcpus = VGIC_V2_MAX_CPUS;
-
-	on_each_cpu(vgic_cpu_init_lrs, vgic, 1);
-
-	*ops = &vgic_v2_ops;
-	*params = vgic;
-	goto out;
-
-out_unmap:
-	iounmap(vgic->vctrl_base);
-out:
-	return ret;
-}

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

@@ -1,1074 +0,0 @@
-/*
- * GICv3 distributor and redistributor emulation
- *
- * GICv3 emulation is currently only supported on a GICv3 host (because
- * we rely on the hardware's CPU interface virtualization support), but
- * supports both hardware with or without the optional GICv2 backwards
- * compatibility features.
- *
- * Limitations of the emulation:
- * (RAZ/WI: read as zero, write ignore, RAO/WI: read as one, write ignore)
- * - We do not support LPIs (yet). TYPER.LPIS is reported as 0 and is RAZ/WI.
- * - We do not support the message based interrupts (MBIs) triggered by
- *   writes to the GICD_{SET,CLR}SPI_* registers. TYPER.MBIS is reported as 0.
- * - We do not support the (optional) backwards compatibility feature.
- *   GICD_CTLR.ARE resets to 1 and is RAO/WI. If the _host_ GIC supports
- *   the compatiblity feature, you can use a GICv2 in the guest, though.
- * - We only support a single security state. GICD_CTLR.DS is 1 and is RAO/WI.
- * - Priorities are not emulated (same as the GICv2 emulation). Linux
- *   as a guest is fine with this, because it does not use priorities.
- * - We only support Group1 interrupts. Again Linux uses only those.
- *
- * Copyright (C) 2014 ARM Ltd.
- * Author: Andre Przywara <andre.przywara@arm.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-
-#include <linux/cpu.h>
-#include <linux/kvm.h>
-#include <linux/kvm_host.h>
-#include <linux/interrupt.h>
-
-#include <linux/irqchip/arm-gic-v3.h>
-#include <kvm/arm_vgic.h>
-
-#include <asm/kvm_emulate.h>
-#include <asm/kvm_arm.h>
-#include <asm/kvm_mmu.h>
-
-#include "vgic.h"
-
-static bool handle_mmio_rao_wi(struct kvm_vcpu *vcpu,
-			       struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 reg = 0xffffffff;
-
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-
-	return false;
-}
-
-static bool handle_mmio_ctlr(struct kvm_vcpu *vcpu,
-			     struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 reg = 0;
-
-	/*
-	 * Force ARE and DS to 1, the guest cannot change this.
-	 * For the time being we only support Group1 interrupts.
-	 */
-	if (vcpu->kvm->arch.vgic.enabled)
-		reg = GICD_CTLR_ENABLE_SS_G1;
-	reg |= GICD_CTLR_ARE_NS | GICD_CTLR_DS;
-
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-	if (mmio->is_write) {
-		vcpu->kvm->arch.vgic.enabled = !!(reg & GICD_CTLR_ENABLE_SS_G1);
-		vgic_update_state(vcpu->kvm);
-		return true;
-	}
-	return false;
-}
-
-/*
- * As this implementation does not provide compatibility
- * with GICv2 (ARE==1), we report zero CPUs in bits [5..7].
- * Also LPIs and MBIs are not supported, so we set the respective bits to 0.
- * Also we report at most 2**10=1024 interrupt IDs (to match 1024 SPIs).
- */
-#define INTERRUPT_ID_BITS 10
-static bool handle_mmio_typer(struct kvm_vcpu *vcpu,
-			      struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 reg;
-
-	reg = (min(vcpu->kvm->arch.vgic.nr_irqs, 1024) >> 5) - 1;
-
-	reg |= (INTERRUPT_ID_BITS - 1) << 19;
-
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-
-	return false;
-}
-
-static bool handle_mmio_iidr(struct kvm_vcpu *vcpu,
-			     struct kvm_exit_mmio *mmio, phys_addr_t offset)
-{
-	u32 reg;
-
-	reg = (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-
-	return false;
-}
-
-static bool handle_mmio_set_enable_reg_dist(struct kvm_vcpu *vcpu,
-					    struct kvm_exit_mmio *mmio,
-					    phys_addr_t offset)
-{
-	if (likely(offset >= VGIC_NR_PRIVATE_IRQS / 8))
-		return vgic_handle_enable_reg(vcpu->kvm, mmio, offset,
-					      vcpu->vcpu_id,
-					      ACCESS_WRITE_SETBIT);
-
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_clear_enable_reg_dist(struct kvm_vcpu *vcpu,
-					      struct kvm_exit_mmio *mmio,
-					      phys_addr_t offset)
-{
-	if (likely(offset >= VGIC_NR_PRIVATE_IRQS / 8))
-		return vgic_handle_enable_reg(vcpu->kvm, mmio, offset,
-					      vcpu->vcpu_id,
-					      ACCESS_WRITE_CLEARBIT);
-
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_set_pending_reg_dist(struct kvm_vcpu *vcpu,
-					     struct kvm_exit_mmio *mmio,
-					     phys_addr_t offset)
-{
-	if (likely(offset >= VGIC_NR_PRIVATE_IRQS / 8))
-		return vgic_handle_set_pending_reg(vcpu->kvm, mmio, offset,
-						   vcpu->vcpu_id);
-
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_clear_pending_reg_dist(struct kvm_vcpu *vcpu,
-					       struct kvm_exit_mmio *mmio,
-					       phys_addr_t offset)
-{
-	if (likely(offset >= VGIC_NR_PRIVATE_IRQS / 8))
-		return vgic_handle_clear_pending_reg(vcpu->kvm, mmio, offset,
-						     vcpu->vcpu_id);
-
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_set_active_reg_dist(struct kvm_vcpu *vcpu,
-					    struct kvm_exit_mmio *mmio,
-					    phys_addr_t offset)
-{
-	if (likely(offset >= VGIC_NR_PRIVATE_IRQS / 8))
-		return vgic_handle_set_active_reg(vcpu->kvm, mmio, offset,
-						   vcpu->vcpu_id);
-
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_clear_active_reg_dist(struct kvm_vcpu *vcpu,
-					      struct kvm_exit_mmio *mmio,
-					      phys_addr_t offset)
-{
-	if (likely(offset >= VGIC_NR_PRIVATE_IRQS / 8))
-		return vgic_handle_clear_active_reg(vcpu->kvm, mmio, offset,
-						    vcpu->vcpu_id);
-
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_priority_reg_dist(struct kvm_vcpu *vcpu,
-					  struct kvm_exit_mmio *mmio,
-					  phys_addr_t offset)
-{
-	u32 *reg;
-
-	if (unlikely(offset < VGIC_NR_PRIVATE_IRQS)) {
-		vgic_reg_access(mmio, NULL, offset,
-				ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-		return false;
-	}
-
-	reg = vgic_bytemap_get_reg(&vcpu->kvm->arch.vgic.irq_priority,
-				   vcpu->vcpu_id, offset);
-	vgic_reg_access(mmio, reg, offset,
-		ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-	return false;
-}
-
-static bool handle_mmio_cfg_reg_dist(struct kvm_vcpu *vcpu,
-				     struct kvm_exit_mmio *mmio,
-				     phys_addr_t offset)
-{
-	u32 *reg;
-
-	if (unlikely(offset < VGIC_NR_PRIVATE_IRQS / 4)) {
-		vgic_reg_access(mmio, NULL, offset,
-				ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-		return false;
-	}
-
-	reg = vgic_bitmap_get_reg(&vcpu->kvm->arch.vgic.irq_cfg,
-				  vcpu->vcpu_id, offset >> 1);
-
-	return vgic_handle_cfg_reg(reg, mmio, offset);
-}
-
-/*
- * We use a compressed version of the MPIDR (all 32 bits in one 32-bit word)
- * when we store the target MPIDR written by the guest.
- */
-static u32 compress_mpidr(unsigned long mpidr)
-{
-	u32 ret;
-
-	ret = MPIDR_AFFINITY_LEVEL(mpidr, 0);
-	ret |= MPIDR_AFFINITY_LEVEL(mpidr, 1) << 8;
-	ret |= MPIDR_AFFINITY_LEVEL(mpidr, 2) << 16;
-	ret |= MPIDR_AFFINITY_LEVEL(mpidr, 3) << 24;
-
-	return ret;
-}
-
-static unsigned long uncompress_mpidr(u32 value)
-{
-	unsigned long mpidr;
-
-	mpidr  = ((value >>  0) & 0xFF) << MPIDR_LEVEL_SHIFT(0);
-	mpidr |= ((value >>  8) & 0xFF) << MPIDR_LEVEL_SHIFT(1);
-	mpidr |= ((value >> 16) & 0xFF) << MPIDR_LEVEL_SHIFT(2);
-	mpidr |= (u64)((value >> 24) & 0xFF) << MPIDR_LEVEL_SHIFT(3);
-
-	return mpidr;
-}
-
-/*
- * Lookup the given MPIDR value to get the vcpu_id (if there is one)
- * and store that in the irq_spi_cpu[] array.
- * This limits the number of VCPUs to 255 for now, extending the data
- * type (or storing kvm_vcpu pointers) should lift the limit.
- * Store the original MPIDR value in an extra array to support read-as-written.
- * Unallocated MPIDRs are translated to a special value and caught
- * before any array accesses.
- */
-static bool handle_mmio_route_reg(struct kvm_vcpu *vcpu,
-				  struct kvm_exit_mmio *mmio,
-				  phys_addr_t offset)
-{
-	struct kvm *kvm = vcpu->kvm;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	int spi;
-	u32 reg;
-	int vcpu_id;
-	unsigned long *bmap, mpidr;
-
-	/*
-	 * The upper 32 bits of each 64 bit register are zero,
-	 * as we don't support Aff3.
-	 */
-	if ((offset & 4)) {
-		vgic_reg_access(mmio, NULL, offset,
-				ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-		return false;
-	}
-
-	/* This region only covers SPIs, so no handling of private IRQs here. */
-	spi = offset / 8;
-
-	/* get the stored MPIDR for this IRQ */
-	mpidr = uncompress_mpidr(dist->irq_spi_mpidr[spi]);
-	reg = mpidr;
-
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-
-	if (!mmio->is_write)
-		return false;
-
-	/*
-	 * Now clear the currently assigned vCPU from the map, making room
-	 * for the new one to be written below
-	 */
-	vcpu = kvm_mpidr_to_vcpu(kvm, mpidr);
-	if (likely(vcpu)) {
-		vcpu_id = vcpu->vcpu_id;
-		bmap = vgic_bitmap_get_shared_map(&dist->irq_spi_target[vcpu_id]);
-		__clear_bit(spi, bmap);
-	}
-
-	dist->irq_spi_mpidr[spi] = compress_mpidr(reg);
-	vcpu = kvm_mpidr_to_vcpu(kvm, reg & MPIDR_HWID_BITMASK);
-
-	/*
-	 * The spec says that non-existent MPIDR values should not be
-	 * forwarded to any existent (v)CPU, but should be able to become
-	 * pending anyway. We simply keep the irq_spi_target[] array empty, so
-	 * the interrupt will never be injected.
-	 * irq_spi_cpu[irq] gets a magic value in this case.
-	 */
-	if (likely(vcpu)) {
-		vcpu_id = vcpu->vcpu_id;
-		dist->irq_spi_cpu[spi] = vcpu_id;
-		bmap = vgic_bitmap_get_shared_map(&dist->irq_spi_target[vcpu_id]);
-		__set_bit(spi, bmap);
-	} else {
-		dist->irq_spi_cpu[spi] = VCPU_NOT_ALLOCATED;
-	}
-
-	vgic_update_state(kvm);
-
-	return true;
-}
-
-/*
- * We should be careful about promising too much when a guest reads
- * this register. Don't claim to be like any hardware implementation,
- * but just report the GIC as version 3 - which is what a Linux guest
- * would check.
- */
-static bool handle_mmio_idregs(struct kvm_vcpu *vcpu,
-			       struct kvm_exit_mmio *mmio,
-			       phys_addr_t offset)
-{
-	u32 reg = 0;
-
-	switch (offset + GICD_IDREGS) {
-	case GICD_PIDR2:
-		reg = 0x3b;
-		break;
-	}
-
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-
-	return false;
-}
-
-static const struct vgic_io_range vgic_v3_dist_ranges[] = {
-	{
-		.base           = GICD_CTLR,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_ctlr,
-	},
-	{
-		.base           = GICD_TYPER,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_typer,
-	},
-	{
-		.base           = GICD_IIDR,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_iidr,
-	},
-	{
-		/* this register is optional, it is RAZ/WI if not implemented */
-		.base           = GICD_STATUSR,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_raz_wi,
-	},
-	{
-		/* this write only register is WI when TYPER.MBIS=0 */
-		.base		= GICD_SETSPI_NSR,
-		.len		= 0x04,
-		.bits_per_irq	= 0,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this write only register is WI when TYPER.MBIS=0 */
-		.base		= GICD_CLRSPI_NSR,
-		.len		= 0x04,
-		.bits_per_irq	= 0,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this is RAZ/WI when DS=1 */
-		.base		= GICD_SETSPI_SR,
-		.len		= 0x04,
-		.bits_per_irq	= 0,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this is RAZ/WI when DS=1 */
-		.base		= GICD_CLRSPI_SR,
-		.len		= 0x04,
-		.bits_per_irq	= 0,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		.base		= GICD_IGROUPR,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_rao_wi,
-	},
-	{
-		.base		= GICD_ISENABLER,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_enable_reg_dist,
-	},
-	{
-		.base		= GICD_ICENABLER,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_enable_reg_dist,
-	},
-	{
-		.base		= GICD_ISPENDR,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_pending_reg_dist,
-	},
-	{
-		.base		= GICD_ICPENDR,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_pending_reg_dist,
-	},
-	{
-		.base		= GICD_ISACTIVER,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_active_reg_dist,
-	},
-	{
-		.base		= GICD_ICACTIVER,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_active_reg_dist,
-	},
-	{
-		.base		= GICD_IPRIORITYR,
-		.len		= 0x400,
-		.bits_per_irq	= 8,
-		.handle_mmio	= handle_mmio_priority_reg_dist,
-	},
-	{
-		/* TARGETSRn is RES0 when ARE=1 */
-		.base		= GICD_ITARGETSR,
-		.len		= 0x400,
-		.bits_per_irq	= 8,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		.base		= GICD_ICFGR,
-		.len		= 0x100,
-		.bits_per_irq	= 2,
-		.handle_mmio	= handle_mmio_cfg_reg_dist,
-	},
-	{
-		/* this is RAZ/WI when DS=1 */
-		.base		= GICD_IGRPMODR,
-		.len		= 0x80,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this is RAZ/WI when DS=1 */
-		.base		= GICD_NSACR,
-		.len		= 0x100,
-		.bits_per_irq	= 2,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this is RAZ/WI when ARE=1 */
-		.base		= GICD_SGIR,
-		.len		= 0x04,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this is RAZ/WI when ARE=1 */
-		.base		= GICD_CPENDSGIR,
-		.len		= 0x10,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		/* this is RAZ/WI when ARE=1 */
-		.base           = GICD_SPENDSGIR,
-		.len            = 0x10,
-		.handle_mmio    = handle_mmio_raz_wi,
-	},
-	{
-		.base		= GICD_IROUTER + 0x100,
-		.len		= 0x1ee0,
-		.bits_per_irq	= 64,
-		.handle_mmio	= handle_mmio_route_reg,
-	},
-	{
-		.base           = GICD_IDREGS,
-		.len            = 0x30,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_idregs,
-	},
-	{},
-};
-
-static bool handle_mmio_ctlr_redist(struct kvm_vcpu *vcpu,
-				    struct kvm_exit_mmio *mmio,
-				    phys_addr_t offset)
-{
-	/* since we don't support LPIs, this register is zero for now */
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_typer_redist(struct kvm_vcpu *vcpu,
-				     struct kvm_exit_mmio *mmio,
-				     phys_addr_t offset)
-{
-	u32 reg;
-	u64 mpidr;
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-	int target_vcpu_id = redist_vcpu->vcpu_id;
-
-	/* the upper 32 bits contain the affinity value */
-	if ((offset & ~3) == 4) {
-		mpidr = kvm_vcpu_get_mpidr_aff(redist_vcpu);
-		reg = compress_mpidr(mpidr);
-
-		vgic_reg_access(mmio, &reg, offset,
-				ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-		return false;
-	}
-
-	reg = redist_vcpu->vcpu_id << 8;
-	if (target_vcpu_id == atomic_read(&vcpu->kvm->online_vcpus) - 1)
-		reg |= GICR_TYPER_LAST;
-	vgic_reg_access(mmio, &reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-static bool handle_mmio_set_enable_reg_redist(struct kvm_vcpu *vcpu,
-					      struct kvm_exit_mmio *mmio,
-					      phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	return vgic_handle_enable_reg(vcpu->kvm, mmio, offset,
-				      redist_vcpu->vcpu_id,
-				      ACCESS_WRITE_SETBIT);
-}
-
-static bool handle_mmio_clear_enable_reg_redist(struct kvm_vcpu *vcpu,
-						struct kvm_exit_mmio *mmio,
-						phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	return vgic_handle_enable_reg(vcpu->kvm, mmio, offset,
-				      redist_vcpu->vcpu_id,
-				      ACCESS_WRITE_CLEARBIT);
-}
-
-static bool handle_mmio_set_active_reg_redist(struct kvm_vcpu *vcpu,
-					      struct kvm_exit_mmio *mmio,
-					      phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	return vgic_handle_set_active_reg(vcpu->kvm, mmio, offset,
-					  redist_vcpu->vcpu_id);
-}
-
-static bool handle_mmio_clear_active_reg_redist(struct kvm_vcpu *vcpu,
-						struct kvm_exit_mmio *mmio,
-						phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	return vgic_handle_clear_active_reg(vcpu->kvm, mmio, offset,
-					     redist_vcpu->vcpu_id);
-}
-
-static bool handle_mmio_set_pending_reg_redist(struct kvm_vcpu *vcpu,
-					       struct kvm_exit_mmio *mmio,
-					       phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	return vgic_handle_set_pending_reg(vcpu->kvm, mmio, offset,
-					   redist_vcpu->vcpu_id);
-}
-
-static bool handle_mmio_clear_pending_reg_redist(struct kvm_vcpu *vcpu,
-						 struct kvm_exit_mmio *mmio,
-						 phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	return vgic_handle_clear_pending_reg(vcpu->kvm, mmio, offset,
-					     redist_vcpu->vcpu_id);
-}
-
-static bool handle_mmio_priority_reg_redist(struct kvm_vcpu *vcpu,
-					    struct kvm_exit_mmio *mmio,
-					    phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-	u32 *reg;
-
-	reg = vgic_bytemap_get_reg(&vcpu->kvm->arch.vgic.irq_priority,
-				   redist_vcpu->vcpu_id, offset);
-	vgic_reg_access(mmio, reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-	return false;
-}
-
-static bool handle_mmio_cfg_reg_redist(struct kvm_vcpu *vcpu,
-				       struct kvm_exit_mmio *mmio,
-				       phys_addr_t offset)
-{
-	struct kvm_vcpu *redist_vcpu = mmio->private;
-
-	u32 *reg = vgic_bitmap_get_reg(&vcpu->kvm->arch.vgic.irq_cfg,
-				       redist_vcpu->vcpu_id, offset >> 1);
-
-	return vgic_handle_cfg_reg(reg, mmio, offset);
-}
-
-#define SGI_base(x) ((x) + SZ_64K)
-
-static const struct vgic_io_range vgic_redist_ranges[] = {
-	{
-		.base           = GICR_CTLR,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_ctlr_redist,
-	},
-	{
-		.base           = GICR_TYPER,
-		.len            = 0x08,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_typer_redist,
-	},
-	{
-		.base           = GICR_IIDR,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_iidr,
-	},
-	{
-		.base           = GICR_WAKER,
-		.len            = 0x04,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_raz_wi,
-	},
-	{
-		.base           = GICR_IDREGS,
-		.len            = 0x30,
-		.bits_per_irq   = 0,
-		.handle_mmio    = handle_mmio_idregs,
-	},
-	{
-		.base		= SGI_base(GICR_IGROUPR0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_rao_wi,
-	},
-	{
-		.base		= SGI_base(GICR_ISENABLER0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_enable_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_ICENABLER0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_enable_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_ISPENDR0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_pending_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_ICPENDR0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_pending_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_ISACTIVER0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_set_active_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_ICACTIVER0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_clear_active_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_IPRIORITYR0),
-		.len		= 0x20,
-		.bits_per_irq	= 8,
-		.handle_mmio	= handle_mmio_priority_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_ICFGR0),
-		.len		= 0x08,
-		.bits_per_irq	= 2,
-		.handle_mmio	= handle_mmio_cfg_reg_redist,
-	},
-	{
-		.base		= SGI_base(GICR_IGRPMODR0),
-		.len		= 0x04,
-		.bits_per_irq	= 1,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{
-		.base		= SGI_base(GICR_NSACR),
-		.len		= 0x04,
-		.handle_mmio	= handle_mmio_raz_wi,
-	},
-	{},
-};
-
-static bool vgic_v3_queue_sgi(struct kvm_vcpu *vcpu, int irq)
-{
-	if (vgic_queue_irq(vcpu, 0, irq)) {
-		vgic_dist_irq_clear_pending(vcpu, irq);
-		vgic_cpu_irq_clear(vcpu, irq);
-		return true;
-	}
-
-	return false;
-}
-
-static int vgic_v3_map_resources(struct kvm *kvm,
-				 const struct vgic_params *params)
-{
-	int ret = 0;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	gpa_t rdbase = dist->vgic_redist_base;
-	struct vgic_io_device *iodevs = NULL;
-	int i;
-
-	if (!irqchip_in_kernel(kvm))
-		return 0;
-
-	mutex_lock(&kvm->lock);
-
-	if (vgic_ready(kvm))
-		goto out;
-
-	if (IS_VGIC_ADDR_UNDEF(dist->vgic_dist_base) ||
-	    IS_VGIC_ADDR_UNDEF(dist->vgic_redist_base)) {
-		kvm_err("Need to set vgic distributor addresses first\n");
-		ret = -ENXIO;
-		goto out;
-	}
-
-	/*
-	 * For a VGICv3 we require the userland to explicitly initialize
-	 * the VGIC before we need to use it.
-	 */
-	if (!vgic_initialized(kvm)) {
-		ret = -EBUSY;
-		goto out;
-	}
-
-	ret = vgic_register_kvm_io_dev(kvm, dist->vgic_dist_base,
-				       GIC_V3_DIST_SIZE, vgic_v3_dist_ranges,
-				       -1, &dist->dist_iodev);
-	if (ret)
-		goto out;
-
-	iodevs = kcalloc(dist->nr_cpus, sizeof(iodevs[0]), GFP_KERNEL);
-	if (!iodevs) {
-		ret = -ENOMEM;
-		goto out_unregister;
-	}
-
-	for (i = 0; i < dist->nr_cpus; i++) {
-		ret = vgic_register_kvm_io_dev(kvm, rdbase,
-					       SZ_128K, vgic_redist_ranges,
-					       i, &iodevs[i]);
-		if (ret)
-			goto out_unregister;
-		rdbase += GIC_V3_REDIST_SIZE;
-	}
-
-	dist->redist_iodevs = iodevs;
-	dist->ready = true;
-	goto out;
-
-out_unregister:
-	kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS, &dist->dist_iodev.dev);
-	if (iodevs) {
-		for (i = 0; i < dist->nr_cpus; i++) {
-			if (iodevs[i].dev.ops)
-				kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS,
-							  &iodevs[i].dev);
-		}
-	}
-
-out:
-	if (ret)
-		kvm_vgic_destroy(kvm);
-	mutex_unlock(&kvm->lock);
-	return ret;
-}
-
-static int vgic_v3_init_model(struct kvm *kvm)
-{
-	int i;
-	u32 mpidr;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	int nr_spis = dist->nr_irqs - VGIC_NR_PRIVATE_IRQS;
-
-	dist->irq_spi_mpidr = kcalloc(nr_spis, sizeof(dist->irq_spi_mpidr[0]),
-				      GFP_KERNEL);
-
-	if (!dist->irq_spi_mpidr)
-		return -ENOMEM;
-
-	/* Initialize the target VCPUs for each IRQ to VCPU 0 */
-	mpidr = compress_mpidr(kvm_vcpu_get_mpidr_aff(kvm_get_vcpu(kvm, 0)));
-	for (i = VGIC_NR_PRIVATE_IRQS; i < dist->nr_irqs; i++) {
-		dist->irq_spi_cpu[i - VGIC_NR_PRIVATE_IRQS] = 0;
-		dist->irq_spi_mpidr[i - VGIC_NR_PRIVATE_IRQS] = mpidr;
-		vgic_bitmap_set_irq_val(dist->irq_spi_target, 0, i, 1);
-	}
-
-	return 0;
-}
-
-/* GICv3 does not keep track of SGI sources anymore. */
-static void vgic_v3_add_sgi_source(struct kvm_vcpu *vcpu, int irq, int source)
-{
-}
-
-void vgic_v3_init_emulation(struct kvm *kvm)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-
-	dist->vm_ops.queue_sgi = vgic_v3_queue_sgi;
-	dist->vm_ops.add_sgi_source = vgic_v3_add_sgi_source;
-	dist->vm_ops.init_model = vgic_v3_init_model;
-	dist->vm_ops.map_resources = vgic_v3_map_resources;
-
-	kvm->arch.max_vcpus = KVM_MAX_VCPUS;
-}
-
-/*
- * Compare a given affinity (level 1-3 and a level 0 mask, from the SGI
- * generation register ICC_SGI1R_EL1) with a given VCPU.
- * If the VCPU's MPIDR matches, return the level0 affinity, otherwise
- * return -1.
- */
-static int match_mpidr(u64 sgi_aff, u16 sgi_cpu_mask, struct kvm_vcpu *vcpu)
-{
-	unsigned long affinity;
-	int level0;
-
-	/*
-	 * Split the current VCPU's MPIDR into affinity level 0 and the
-	 * rest as this is what we have to compare against.
-	 */
-	affinity = kvm_vcpu_get_mpidr_aff(vcpu);
-	level0 = MPIDR_AFFINITY_LEVEL(affinity, 0);
-	affinity &= ~MPIDR_LEVEL_MASK;
-
-	/* bail out if the upper three levels don't match */
-	if (sgi_aff != affinity)
-		return -1;
-
-	/* Is this VCPU's bit set in the mask ? */
-	if (!(sgi_cpu_mask & BIT(level0)))
-		return -1;
-
-	return level0;
-}
-
-#define SGI_AFFINITY_LEVEL(reg, level) \
-	((((reg) & ICC_SGI1R_AFFINITY_## level ##_MASK) \
-	>> ICC_SGI1R_AFFINITY_## level ##_SHIFT) << MPIDR_LEVEL_SHIFT(level))
-
-/**
- * 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
- *
- * 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.
- * This ICC_SGI1R_EL1 register contains the upper three affinity levels of the
- * target processors as well as a bitmask of 16 Aff0 CPUs.
- * If the interrupt routing mode bit is not set, we iterate over all VCPUs to
- * 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)
-{
-	struct kvm *kvm = vcpu->kvm;
-	struct kvm_vcpu *c_vcpu;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	u16 target_cpus;
-	u64 mpidr;
-	int sgi, c;
-	int vcpu_id = vcpu->vcpu_id;
-	bool broadcast;
-	int updated = 0;
-
-	sgi = (reg & ICC_SGI1R_SGI_ID_MASK) >> ICC_SGI1R_SGI_ID_SHIFT;
-	broadcast = reg & BIT(ICC_SGI1R_IRQ_ROUTING_MODE_BIT);
-	target_cpus = (reg & ICC_SGI1R_TARGET_LIST_MASK) >> ICC_SGI1R_TARGET_LIST_SHIFT;
-	mpidr = SGI_AFFINITY_LEVEL(reg, 3);
-	mpidr |= SGI_AFFINITY_LEVEL(reg, 2);
-	mpidr |= SGI_AFFINITY_LEVEL(reg, 1);
-
-	/*
-	 * We take the dist lock here, because we come from the sysregs
-	 * code path and not from the MMIO one (which already takes the lock).
-	 */
-	spin_lock(&dist->lock);
-
-	/*
-	 * We iterate over all VCPUs to find the MPIDRs matching the request.
-	 * If we have handled one CPU, we clear it's bit to detect early
-	 * if we are already finished. This avoids iterating through all
-	 * VCPUs when most of the times we just signal a single VCPU.
-	 */
-	kvm_for_each_vcpu(c, c_vcpu, kvm) {
-
-		/* Exit early if we have dealt with all requested CPUs */
-		if (!broadcast && target_cpus == 0)
-			break;
-
-		 /* Don't signal the calling VCPU */
-		if (broadcast && c == vcpu_id)
-			continue;
-
-		if (!broadcast) {
-			int level0;
-
-			level0 = match_mpidr(mpidr, target_cpus, c_vcpu);
-			if (level0 == -1)
-				continue;
-
-			/* remove this matching VCPU from the mask */
-			target_cpus &= ~BIT(level0);
-		}
-
-		/* Flag the SGI as pending */
-		vgic_dist_irq_set_pending(c_vcpu, sgi);
-		updated = 1;
-		kvm_debug("SGI%d from CPU%d to CPU%d\n", sgi, vcpu_id, c);
-	}
-	if (updated)
-		vgic_update_state(vcpu->kvm);
-	spin_unlock(&dist->lock);
-	if (updated)
-		vgic_kick_vcpus(vcpu->kvm);
-}
-
-static int vgic_v3_create(struct kvm_device *dev, u32 type)
-{
-	return kvm_vgic_create(dev->kvm, type);
-}
-
-static void vgic_v3_destroy(struct kvm_device *dev)
-{
-	kfree(dev);
-}
-
-static int vgic_v3_set_attr(struct kvm_device *dev,
-			    struct kvm_device_attr *attr)
-{
-	int ret;
-
-	ret = vgic_set_common_attr(dev, attr);
-	if (ret != -ENXIO)
-		return ret;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS:
-		return -ENXIO;
-	}
-
-	return -ENXIO;
-}
-
-static int vgic_v3_get_attr(struct kvm_device *dev,
-			    struct kvm_device_attr *attr)
-{
-	int ret;
-
-	ret = vgic_get_common_attr(dev, attr);
-	if (ret != -ENXIO)
-		return ret;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS:
-		return -ENXIO;
-	}
-
-	return -ENXIO;
-}
-
-static int vgic_v3_has_attr(struct kvm_device *dev,
-			    struct kvm_device_attr *attr)
-{
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_ADDR:
-		switch (attr->attr) {
-		case KVM_VGIC_V2_ADDR_TYPE_DIST:
-		case KVM_VGIC_V2_ADDR_TYPE_CPU:
-			return -ENXIO;
-		case KVM_VGIC_V3_ADDR_TYPE_DIST:
-		case KVM_VGIC_V3_ADDR_TYPE_REDIST:
-			return 0;
-		}
-		break;
-	case KVM_DEV_ARM_VGIC_GRP_DIST_REGS:
-	case KVM_DEV_ARM_VGIC_GRP_CPU_REGS:
-		return -ENXIO;
-	case KVM_DEV_ARM_VGIC_GRP_NR_IRQS:
-		return 0;
-	case KVM_DEV_ARM_VGIC_GRP_CTRL:
-		switch (attr->attr) {
-		case KVM_DEV_ARM_VGIC_CTRL_INIT:
-			return 0;
-		}
-	}
-	return -ENXIO;
-}
-
-struct kvm_device_ops kvm_arm_vgic_v3_ops = {
-	.name = "kvm-arm-vgic-v3",
-	.create = vgic_v3_create,
-	.destroy = vgic_v3_destroy,
-	.set_attr = vgic_v3_set_attr,
-	.get_attr = vgic_v3_get_attr,
-	.has_attr = vgic_v3_has_attr,
-};

virt/kvm/arm/vgic-v3.c

@@ -1,279 +0,0 @@
-/*
- * Copyright (C) 2013 ARM Limited, All Rights Reserved.
- * Author: Marc Zyngier <marc.zyngier@arm.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-
-#include <linux/cpu.h>
-#include <linux/kvm.h>
-#include <linux/kvm_host.h>
-#include <linux/interrupt.h>
-#include <linux/io.h>
-
-#include <linux/irqchip/arm-gic-v3.h>
-#include <linux/irqchip/arm-gic-common.h>
-
-#include <asm/kvm_emulate.h>
-#include <asm/kvm_arm.h>
-#include <asm/kvm_asm.h>
-#include <asm/kvm_mmu.h>
-
-static u32 ich_vtr_el2;
-
-static struct vgic_lr vgic_v3_get_lr(const struct kvm_vcpu *vcpu, int lr)
-{
-	struct vgic_lr lr_desc;
-	u64 val = vcpu->arch.vgic_cpu.vgic_v3.vgic_lr[lr];
-
-	if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3)
-		lr_desc.irq = val & ICH_LR_VIRTUAL_ID_MASK;
-	else
-		lr_desc.irq = val & GICH_LR_VIRTUALID;
-
-	lr_desc.source = 0;
-	if (lr_desc.irq <= 15 &&
-	    vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V2)
-		lr_desc.source = (val >> GICH_LR_PHYSID_CPUID_SHIFT) & 0x7;
-
-	lr_desc.state = 0;
-
-	if (val & ICH_LR_PENDING_BIT)
-		lr_desc.state |= LR_STATE_PENDING;
-	if (val & ICH_LR_ACTIVE_BIT)
-		lr_desc.state |= LR_STATE_ACTIVE;
-	if (val & ICH_LR_EOI)
-		lr_desc.state |= LR_EOI_INT;
-	if (val & ICH_LR_HW) {
-		lr_desc.state |= LR_HW;
-		lr_desc.hwirq = (val >> ICH_LR_PHYS_ID_SHIFT) & GENMASK(9, 0);
-	}
-
-	return lr_desc;
-}
-
-static void vgic_v3_set_lr(struct kvm_vcpu *vcpu, int lr,
-			   struct vgic_lr lr_desc)
-{
-	u64 lr_val;
-
-	lr_val = lr_desc.irq;
-
-	/*
-	 * Currently all guest IRQs are Group1, as Group0 would result
-	 * in a FIQ in the guest, which it wouldn't expect.
-	 * Eventually we want to make this configurable, so we may revisit
-	 * this in the future.
-	 */
-	switch (vcpu->kvm->arch.vgic.vgic_model) {
-	case KVM_DEV_TYPE_ARM_VGIC_V3:
-		lr_val |= ICH_LR_GROUP;
-		break;
-	case  KVM_DEV_TYPE_ARM_VGIC_V2:
-		if (lr_desc.irq < VGIC_NR_SGIS)
-			lr_val |= (u32)lr_desc.source << GICH_LR_PHYSID_CPUID_SHIFT;
-		break;
-	default:
-		BUG();
-	}
-
-	if (lr_desc.state & LR_STATE_PENDING)
-		lr_val |= ICH_LR_PENDING_BIT;
-	if (lr_desc.state & LR_STATE_ACTIVE)
-		lr_val |= ICH_LR_ACTIVE_BIT;
-	if (lr_desc.state & LR_EOI_INT)
-		lr_val |= ICH_LR_EOI;
-	if (lr_desc.state & LR_HW) {
-		lr_val |= ICH_LR_HW;
-		lr_val |= ((u64)lr_desc.hwirq) << ICH_LR_PHYS_ID_SHIFT;
-	}
-
-	vcpu->arch.vgic_cpu.vgic_v3.vgic_lr[lr] = lr_val;
-
-	if (!(lr_desc.state & LR_STATE_MASK))
-		vcpu->arch.vgic_cpu.vgic_v3.vgic_elrsr |= (1U << lr);
-	else
-		vcpu->arch.vgic_cpu.vgic_v3.vgic_elrsr &= ~(1U << lr);
-}
-
-static u64 vgic_v3_get_elrsr(const struct kvm_vcpu *vcpu)
-{
-	return vcpu->arch.vgic_cpu.vgic_v3.vgic_elrsr;
-}
-
-static u64 vgic_v3_get_eisr(const struct kvm_vcpu *vcpu)
-{
-	return vcpu->arch.vgic_cpu.vgic_v3.vgic_eisr;
-}
-
-static void vgic_v3_clear_eisr(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.vgic_cpu.vgic_v3.vgic_eisr = 0;
-}
-
-static u32 vgic_v3_get_interrupt_status(const struct kvm_vcpu *vcpu)
-{
-	u32 misr = vcpu->arch.vgic_cpu.vgic_v3.vgic_misr;
-	u32 ret = 0;
-
-	if (misr & ICH_MISR_EOI)
-		ret |= INT_STATUS_EOI;
-	if (misr & ICH_MISR_U)
-		ret |= INT_STATUS_UNDERFLOW;
-
-	return ret;
-}
-
-static void vgic_v3_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
-{
-	u32 vmcr = vcpu->arch.vgic_cpu.vgic_v3.vgic_vmcr;
-
-	vmcrp->ctlr = (vmcr & ICH_VMCR_CTLR_MASK) >> ICH_VMCR_CTLR_SHIFT;
-	vmcrp->abpr = (vmcr & ICH_VMCR_BPR1_MASK) >> ICH_VMCR_BPR1_SHIFT;
-	vmcrp->bpr  = (vmcr & ICH_VMCR_BPR0_MASK) >> ICH_VMCR_BPR0_SHIFT;
-	vmcrp->pmr  = (vmcr & ICH_VMCR_PMR_MASK) >> ICH_VMCR_PMR_SHIFT;
-}
-
-static void vgic_v3_enable_underflow(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.vgic_cpu.vgic_v3.vgic_hcr |= ICH_HCR_UIE;
-}
-
-static void vgic_v3_disable_underflow(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.vgic_cpu.vgic_v3.vgic_hcr &= ~ICH_HCR_UIE;
-}
-
-static void vgic_v3_set_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
-{
-	u32 vmcr;
-
-	vmcr  = (vmcrp->ctlr << ICH_VMCR_CTLR_SHIFT) & ICH_VMCR_CTLR_MASK;
-	vmcr |= (vmcrp->abpr << ICH_VMCR_BPR1_SHIFT) & ICH_VMCR_BPR1_MASK;
-	vmcr |= (vmcrp->bpr << ICH_VMCR_BPR0_SHIFT) & ICH_VMCR_BPR0_MASK;
-	vmcr |= (vmcrp->pmr << ICH_VMCR_PMR_SHIFT) & ICH_VMCR_PMR_MASK;
-
-	vcpu->arch.vgic_cpu.vgic_v3.vgic_vmcr = vmcr;
-}
-
-static void vgic_v3_enable(struct kvm_vcpu *vcpu)
-{
-	struct vgic_v3_cpu_if *vgic_v3 = &vcpu->arch.vgic_cpu.vgic_v3;
-
-	/*
-	 * By forcing VMCR to zero, the GIC will restore the binary
-	 * points to their reset values. Anything else resets to zero
-	 * anyway.
-	 */
-	vgic_v3->vgic_vmcr = 0;
-	vgic_v3->vgic_elrsr = ~0;
-
-	/*
-	 * If we are emulating a GICv3, we do it in an non-GICv2-compatible
-	 * way, so we force SRE to 1 to demonstrate this to the guest.
-	 * This goes with the spec allowing the value to be RAO/WI.
-	 */
-	if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3)
-		vgic_v3->vgic_sre = ICC_SRE_EL1_SRE;
-	else
-		vgic_v3->vgic_sre = 0;
-
-	/* Get the show on the road... */
-	vgic_v3->vgic_hcr = ICH_HCR_EN;
-}
-
-static const struct vgic_ops vgic_v3_ops = {
-	.get_lr			= vgic_v3_get_lr,
-	.set_lr			= vgic_v3_set_lr,
-	.get_elrsr		= vgic_v3_get_elrsr,
-	.get_eisr		= vgic_v3_get_eisr,
-	.clear_eisr		= vgic_v3_clear_eisr,
-	.get_interrupt_status	= vgic_v3_get_interrupt_status,
-	.enable_underflow	= vgic_v3_enable_underflow,
-	.disable_underflow	= vgic_v3_disable_underflow,
-	.get_vmcr		= vgic_v3_get_vmcr,
-	.set_vmcr		= vgic_v3_set_vmcr,
-	.enable			= vgic_v3_enable,
-};
-
-static struct vgic_params vgic_v3_params;
-
-static void vgic_cpu_init_lrs(void *params)
-{
-	kvm_call_hyp(__vgic_v3_init_lrs);
-}
-
-/**
- * vgic_v3_probe - probe for a GICv3 compatible interrupt controller
- * @gic_kvm_info:	pointer to the GIC description
- * @ops:		address of a pointer to the GICv3 operations
- * @params:		address of a pointer to HW-specific parameters
- *
- * Returns 0 if a GICv3 has been found, with the low level operations
- * in *ops and the HW parameters in *params. Returns an error code
- * otherwise.
- */
-int vgic_v3_probe(const struct gic_kvm_info *gic_kvm_info,
-		  const struct vgic_ops **ops,
-		  const struct vgic_params **params)
-{
-	int ret = 0;
-	struct vgic_params *vgic = &vgic_v3_params;
-	const struct resource *vcpu_res = &gic_kvm_info->vcpu;
-
-	vgic->maint_irq = gic_kvm_info->maint_irq;
-
-	ich_vtr_el2 = kvm_call_hyp(__vgic_v3_get_ich_vtr_el2);
-
-	/*
-	 * The ListRegs field is 5 bits, but there is a architectural
-	 * maximum of 16 list registers. Just ignore bit 4...
-	 */
-	vgic->nr_lr = (ich_vtr_el2 & 0xf) + 1;
-	vgic->can_emulate_gicv2 = false;
-
-	if (!vcpu_res->start) {
-		kvm_info("GICv3: no GICV resource entry\n");
-		vgic->vcpu_base = 0;
-	} else if (!PAGE_ALIGNED(vcpu_res->start)) {
-		pr_warn("GICV physical address 0x%llx not page aligned\n",
-			(unsigned long long)vcpu_res->start);
-		vgic->vcpu_base = 0;
-	} else if (!PAGE_ALIGNED(resource_size(vcpu_res))) {
-		pr_warn("GICV size 0x%llx not a multiple of page size 0x%lx\n",
-			(unsigned long long)resource_size(vcpu_res),
-			PAGE_SIZE);
-	} else {
-		vgic->vcpu_base = vcpu_res->start;
-		vgic->can_emulate_gicv2 = true;
-		kvm_register_device_ops(&kvm_arm_vgic_v2_ops,
-					KVM_DEV_TYPE_ARM_VGIC_V2);
-	}
-	if (vgic->vcpu_base == 0)
-		kvm_info("disabling GICv2 emulation\n");
-	kvm_register_device_ops(&kvm_arm_vgic_v3_ops, KVM_DEV_TYPE_ARM_VGIC_V3);
-
-	vgic->vctrl_base = NULL;
-	vgic->type = VGIC_V3;
-	vgic->max_gic_vcpus = VGIC_V3_MAX_CPUS;
-
-	kvm_info("GICV base=0x%llx, IRQ=%d\n",
-		 vgic->vcpu_base, vgic->maint_irq);
-
-	on_each_cpu(vgic_cpu_init_lrs, vgic, 1);
-
-	*ops = &vgic_v3_ops;
-	*params = vgic;
-
-	return ret;
-}

virt/kvm/arm/vgic.c

@@ -1,2440 +0,0 @@
-/*
- * Copyright (C) 2012 ARM Ltd.
- * Author: Marc Zyngier <marc.zyngier@arm.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
- */
-
-#include <linux/cpu.h>
-#include <linux/kvm.h>
-#include <linux/kvm_host.h>
-#include <linux/interrupt.h>
-#include <linux/io.h>
-#include <linux/irq.h>
-#include <linux/rculist.h>
-#include <linux/uaccess.h>
-
-#include <asm/kvm_emulate.h>
-#include <asm/kvm_arm.h>
-#include <asm/kvm_mmu.h>
-#include <trace/events/kvm.h>
-#include <asm/kvm.h>
-#include <kvm/iodev.h>
-#include <linux/irqchip/arm-gic-common.h>
-
-#define CREATE_TRACE_POINTS
-#include "trace.h"
-
-/*
- * How the whole thing works (courtesy of Christoffer Dall):
- *
- * - At any time, the dist->irq_pending_on_cpu is the oracle that knows if
- *   something is pending on the CPU interface.
- * - Interrupts that are pending on the distributor are stored on the
- *   vgic.irq_pending vgic bitmap (this bitmap is updated by both user land
- *   ioctls and guest mmio ops, and other in-kernel peripherals such as the
- *   arch. timers).
- * - Every time the bitmap changes, the irq_pending_on_cpu oracle is
- *   recalculated
- * - To calculate the oracle, we need info for each cpu from
- *   compute_pending_for_cpu, which considers:
- *   - PPI: dist->irq_pending & dist->irq_enable
- *   - SPI: dist->irq_pending & dist->irq_enable & dist->irq_spi_target
- *   - irq_spi_target is a 'formatted' version of the GICD_ITARGETSRn
- *     registers, stored on each vcpu. We only keep one bit of
- *     information per interrupt, making sure that only one vcpu can
- *     accept the interrupt.
- * - If any of the above state changes, we must recalculate the oracle.
- * - The same is true when injecting an interrupt, except that we only
- *   consider a single interrupt at a time. The irq_spi_cpu array
- *   contains the target CPU for each SPI.
- *
- * The handling of level interrupts adds some extra complexity. We
- * need to track when the interrupt has been EOIed, so we can sample
- * the 'line' again. This is achieved as such:
- *
- * - When a level interrupt is moved onto a vcpu, the corresponding
- *   bit in irq_queued is set. As long as this bit is set, the line
- *   will be ignored for further interrupts. The interrupt is injected
- *   into the vcpu with the GICH_LR_EOI bit set (generate a
- *   maintenance interrupt on EOI).
- * - When the interrupt is EOIed, the maintenance interrupt fires,
- *   and clears the corresponding bit in irq_queued. This allows the
- *   interrupt line to be sampled again.
- * - Note that level-triggered interrupts can also be set to pending from
- *   writes to GICD_ISPENDRn and lowering the external input line does not
- *   cause the interrupt to become inactive in such a situation.
- *   Conversely, writes to GICD_ICPENDRn do not cause the interrupt to become
- *   inactive as long as the external input line is held high.
- *
- *
- * Initialization rules: there are multiple stages to the vgic
- * initialization, both for the distributor and the CPU interfaces.
- *
- * Distributor:
- *
- * - kvm_vgic_early_init(): initialization of static data that doesn't
- *   depend on any sizing information or emulation type. No allocation
- *   is allowed there.
- *
- * - vgic_init(): allocation and initialization of the generic data
- *   structures that depend on sizing information (number of CPUs,
- *   number of interrupts). Also initializes the vcpu specific data
- *   structures. Can be executed lazily for GICv2.
- *   [to be renamed to kvm_vgic_init??]
- *
- * CPU Interface:
- *
- * - kvm_vgic_cpu_early_init(): initialization of static data that
- *   doesn't depend on any sizing information or emulation type. No
- *   allocation is allowed there.
- */
-
-#include "vgic.h"
-
-static void vgic_retire_disabled_irqs(struct kvm_vcpu *vcpu);
-static void vgic_retire_lr(int lr_nr, struct kvm_vcpu *vcpu);
-static struct vgic_lr vgic_get_lr(const struct kvm_vcpu *vcpu, int lr);
-static void vgic_set_lr(struct kvm_vcpu *vcpu, int lr, struct vgic_lr lr_desc);
-static u64 vgic_get_elrsr(struct kvm_vcpu *vcpu);
-static struct irq_phys_map *vgic_irq_map_search(struct kvm_vcpu *vcpu,
-						int virt_irq);
-static int compute_pending_for_cpu(struct kvm_vcpu *vcpu);
-
-static const struct vgic_ops *vgic_ops;
-static const struct vgic_params *vgic;
-
-static void add_sgi_source(struct kvm_vcpu *vcpu, int irq, int source)
-{
-	vcpu->kvm->arch.vgic.vm_ops.add_sgi_source(vcpu, irq, source);
-}
-
-static bool queue_sgi(struct kvm_vcpu *vcpu, int irq)
-{
-	return vcpu->kvm->arch.vgic.vm_ops.queue_sgi(vcpu, irq);
-}
-
-int kvm_vgic_map_resources(struct kvm *kvm)
-{
-	return kvm->arch.vgic.vm_ops.map_resources(kvm, vgic);
-}
-
-/*
- * struct vgic_bitmap contains a bitmap made of unsigned longs, but
- * extracts u32s out of them.
- *
- * This does not work on 64-bit BE systems, because the bitmap access
- * will store two consecutive 32-bit words with the higher-addressed
- * register's bits at the lower index and the lower-addressed register's
- * bits at the higher index.
- *
- * Therefore, swizzle the register index when accessing the 32-bit word
- * registers to access the right register's value.
- */
-#if defined(CONFIG_CPU_BIG_ENDIAN) && BITS_PER_LONG == 64
-#define REG_OFFSET_SWIZZLE	1
-#else
-#define REG_OFFSET_SWIZZLE	0
-#endif
-
-static int vgic_init_bitmap(struct vgic_bitmap *b, int nr_cpus, int nr_irqs)
-{
-	int nr_longs;
-
-	nr_longs = nr_cpus + BITS_TO_LONGS(nr_irqs - VGIC_NR_PRIVATE_IRQS);
-
-	b->private = kzalloc(sizeof(unsigned long) * nr_longs, GFP_KERNEL);
-	if (!b->private)
-		return -ENOMEM;
-
-	b->shared = b->private + nr_cpus;
-
-	return 0;
-}
-
-static void vgic_free_bitmap(struct vgic_bitmap *b)
-{
-	kfree(b->private);
-	b->private = NULL;
-	b->shared = NULL;
-}
-
-/*
- * Call this function to convert a u64 value to an unsigned long * bitmask
- * in a way that works on both 32-bit and 64-bit LE and BE platforms.
- *
- * Warning: Calling this function may modify *val.
- */
-static unsigned long *u64_to_bitmask(u64 *val)
-{
-#if defined(CONFIG_CPU_BIG_ENDIAN) && BITS_PER_LONG == 32
-	*val = (*val >> 32) | (*val << 32);
-#endif
-	return (unsigned long *)val;
-}
-
-u32 *vgic_bitmap_get_reg(struct vgic_bitmap *x, int cpuid, u32 offset)
-{
-	offset >>= 2;
-	if (!offset)
-		return (u32 *)(x->private + cpuid) + REG_OFFSET_SWIZZLE;
-	else
-		return (u32 *)(x->shared) + ((offset - 1) ^ REG_OFFSET_SWIZZLE);
-}
-
-static int vgic_bitmap_get_irq_val(struct vgic_bitmap *x,
-				   int cpuid, int irq)
-{
-	if (irq < VGIC_NR_PRIVATE_IRQS)
-		return test_bit(irq, x->private + cpuid);
-
-	return test_bit(irq - VGIC_NR_PRIVATE_IRQS, x->shared);
-}
-
-void vgic_bitmap_set_irq_val(struct vgic_bitmap *x, int cpuid,
-			     int irq, int val)
-{
-	unsigned long *reg;
-
-	if (irq < VGIC_NR_PRIVATE_IRQS) {
-		reg = x->private + cpuid;
-	} else {
-		reg = x->shared;
-		irq -= VGIC_NR_PRIVATE_IRQS;
-	}
-
-	if (val)
-		set_bit(irq, reg);
-	else
-		clear_bit(irq, reg);
-}
-
-static unsigned long *vgic_bitmap_get_cpu_map(struct vgic_bitmap *x, int cpuid)
-{
-	return x->private + cpuid;
-}
-
-unsigned long *vgic_bitmap_get_shared_map(struct vgic_bitmap *x)
-{
-	return x->shared;
-}
-
-static int vgic_init_bytemap(struct vgic_bytemap *x, int nr_cpus, int nr_irqs)
-{
-	int size;
-
-	size  = nr_cpus * VGIC_NR_PRIVATE_IRQS;
-	size += nr_irqs - VGIC_NR_PRIVATE_IRQS;
-
-	x->private = kzalloc(size, GFP_KERNEL);
-	if (!x->private)
-		return -ENOMEM;
-
-	x->shared = x->private + nr_cpus * VGIC_NR_PRIVATE_IRQS / sizeof(u32);
-	return 0;
-}
-
-static void vgic_free_bytemap(struct vgic_bytemap *b)
-{
-	kfree(b->private);
-	b->private = NULL;
-	b->shared = NULL;
-}
-
-u32 *vgic_bytemap_get_reg(struct vgic_bytemap *x, int cpuid, u32 offset)
-{
-	u32 *reg;
-
-	if (offset < VGIC_NR_PRIVATE_IRQS) {
-		reg = x->private;
-		offset += cpuid * VGIC_NR_PRIVATE_IRQS;
-	} else {
-		reg = x->shared;
-		offset -= VGIC_NR_PRIVATE_IRQS;
-	}
-
-	return reg + (offset / sizeof(u32));
-}
-
-#define VGIC_CFG_LEVEL	0
-#define VGIC_CFG_EDGE	1
-
-static bool vgic_irq_is_edge(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	int irq_val;
-
-	irq_val = vgic_bitmap_get_irq_val(&dist->irq_cfg, vcpu->vcpu_id, irq);
-	return irq_val == VGIC_CFG_EDGE;
-}
-
-static int vgic_irq_is_enabled(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return vgic_bitmap_get_irq_val(&dist->irq_enabled, vcpu->vcpu_id, irq);
-}
-
-static int vgic_irq_is_queued(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return vgic_bitmap_get_irq_val(&dist->irq_queued, vcpu->vcpu_id, irq);
-}
-
-static int vgic_irq_is_active(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return vgic_bitmap_get_irq_val(&dist->irq_active, vcpu->vcpu_id, irq);
-}
-
-static void vgic_irq_set_queued(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_queued, vcpu->vcpu_id, irq, 1);
-}
-
-static void vgic_irq_clear_queued(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_queued, vcpu->vcpu_id, irq, 0);
-}
-
-static void vgic_irq_set_active(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_active, vcpu->vcpu_id, irq, 1);
-}
-
-static void vgic_irq_clear_active(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_active, vcpu->vcpu_id, irq, 0);
-}
-
-static int vgic_dist_irq_get_level(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return vgic_bitmap_get_irq_val(&dist->irq_level, vcpu->vcpu_id, irq);
-}
-
-static void vgic_dist_irq_set_level(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_level, vcpu->vcpu_id, irq, 1);
-}
-
-static void vgic_dist_irq_clear_level(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_level, vcpu->vcpu_id, irq, 0);
-}
-
-static int vgic_dist_irq_soft_pend(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return vgic_bitmap_get_irq_val(&dist->irq_soft_pend, vcpu->vcpu_id, irq);
-}
-
-static void vgic_dist_irq_clear_soft_pend(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_soft_pend, vcpu->vcpu_id, irq, 0);
-	if (!vgic_dist_irq_get_level(vcpu, irq)) {
-		vgic_dist_irq_clear_pending(vcpu, irq);
-		if (!compute_pending_for_cpu(vcpu))
-			clear_bit(vcpu->vcpu_id, dist->irq_pending_on_cpu);
-	}
-}
-
-static int vgic_dist_irq_is_pending(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return vgic_bitmap_get_irq_val(&dist->irq_pending, vcpu->vcpu_id, irq);
-}
-
-void vgic_dist_irq_set_pending(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_pending, vcpu->vcpu_id, irq, 1);
-}
-
-void vgic_dist_irq_clear_pending(struct kvm_vcpu *vcpu, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	vgic_bitmap_set_irq_val(&dist->irq_pending, vcpu->vcpu_id, irq, 0);
-}
-
-static void vgic_cpu_irq_set(struct kvm_vcpu *vcpu, int irq)
-{
-	if (irq < VGIC_NR_PRIVATE_IRQS)
-		set_bit(irq, vcpu->arch.vgic_cpu.pending_percpu);
-	else
-		set_bit(irq - VGIC_NR_PRIVATE_IRQS,
-			vcpu->arch.vgic_cpu.pending_shared);
-}
-
-void vgic_cpu_irq_clear(struct kvm_vcpu *vcpu, int irq)
-{
-	if (irq < VGIC_NR_PRIVATE_IRQS)
-		clear_bit(irq, vcpu->arch.vgic_cpu.pending_percpu);
-	else
-		clear_bit(irq - VGIC_NR_PRIVATE_IRQS,
-			  vcpu->arch.vgic_cpu.pending_shared);
-}
-
-static bool vgic_can_sample_irq(struct kvm_vcpu *vcpu, int irq)
-{
-	return !vgic_irq_is_queued(vcpu, irq);
-}
-
-/**
- * vgic_reg_access - access vgic register
- * @mmio:   pointer to the data describing the mmio access
- * @reg:    pointer to the virtual backing of vgic distributor data
- * @offset: least significant 2 bits used for word offset
- * @mode:   ACCESS_ mode (see defines above)
- *
- * Helper to make vgic register access easier using one of the access
- * modes defined for vgic register access
- * (read,raz,write-ignored,setbit,clearbit,write)
- */
-void vgic_reg_access(struct kvm_exit_mmio *mmio, u32 *reg,
-		     phys_addr_t offset, int mode)
-{
-	int word_offset = (offset & 3) * 8;
-	u32 mask = (1UL << (mmio->len * 8)) - 1;
-	u32 regval;
-
-	/*
-	 * Any alignment fault should have been delivered to the guest
-	 * directly (ARM ARM B3.12.7 "Prioritization of aborts").
-	 */
-
-	if (reg) {
-		regval = *reg;
-	} else {
-		BUG_ON(mode != (ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED));
-		regval = 0;
-	}
-
-	if (mmio->is_write) {
-		u32 data = mmio_data_read(mmio, mask) << word_offset;
-		switch (ACCESS_WRITE_MASK(mode)) {
-		case ACCESS_WRITE_IGNORED:
-			return;
-
-		case ACCESS_WRITE_SETBIT:
-			regval |= data;
-			break;
-
-		case ACCESS_WRITE_CLEARBIT:
-			regval &= ~data;
-			break;
-
-		case ACCESS_WRITE_VALUE:
-			regval = (regval & ~(mask << word_offset)) | data;
-			break;
-		}
-		*reg = regval;
-	} else {
-		switch (ACCESS_READ_MASK(mode)) {
-		case ACCESS_READ_RAZ:
-			regval = 0;
-			/* fall through */
-
-		case ACCESS_READ_VALUE:
-			mmio_data_write(mmio, mask, regval >> word_offset);
-		}
-	}
-}
-
-bool handle_mmio_raz_wi(struct kvm_vcpu *vcpu, struct kvm_exit_mmio *mmio,
-			phys_addr_t offset)
-{
-	vgic_reg_access(mmio, NULL, offset,
-			ACCESS_READ_RAZ | ACCESS_WRITE_IGNORED);
-	return false;
-}
-
-bool vgic_handle_enable_reg(struct kvm *kvm, struct kvm_exit_mmio *mmio,
-			    phys_addr_t offset, int vcpu_id, int access)
-{
-	u32 *reg;
-	int mode = ACCESS_READ_VALUE | access;
-	struct kvm_vcpu *target_vcpu = kvm_get_vcpu(kvm, vcpu_id);
-
-	reg = vgic_bitmap_get_reg(&kvm->arch.vgic.irq_enabled, vcpu_id, offset);
-	vgic_reg_access(mmio, reg, offset, mode);
-	if (mmio->is_write) {
-		if (access & ACCESS_WRITE_CLEARBIT) {
-			if (offset < 4) /* Force SGI enabled */
-				*reg |= 0xffff;
-			vgic_retire_disabled_irqs(target_vcpu);
-		}
-		vgic_update_state(kvm);
-		return true;
-	}
-
-	return false;
-}
-
-bool vgic_handle_set_pending_reg(struct kvm *kvm,
-				 struct kvm_exit_mmio *mmio,
-				 phys_addr_t offset, int vcpu_id)
-{
-	u32 *reg, orig;
-	u32 level_mask;
-	int mode = ACCESS_READ_VALUE | ACCESS_WRITE_SETBIT;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-
-	reg = vgic_bitmap_get_reg(&dist->irq_cfg, vcpu_id, offset);
-	level_mask = (~(*reg));
-
-	/* Mark both level and edge triggered irqs as pending */
-	reg = vgic_bitmap_get_reg(&dist->irq_pending, vcpu_id, offset);
-	orig = *reg;
-	vgic_reg_access(mmio, reg, offset, mode);
-
-	if (mmio->is_write) {
-		/* Set the soft-pending flag only for level-triggered irqs */
-		reg = vgic_bitmap_get_reg(&dist->irq_soft_pend,
-					  vcpu_id, offset);
-		vgic_reg_access(mmio, reg, offset, mode);
-		*reg &= level_mask;
-
-		/* Ignore writes to SGIs */
-		if (offset < 2) {
-			*reg &= ~0xffff;
-			*reg |= orig & 0xffff;
-		}
-
-		vgic_update_state(kvm);
-		return true;
-	}
-
-	return false;
-}
-
-bool vgic_handle_clear_pending_reg(struct kvm *kvm,
-				   struct kvm_exit_mmio *mmio,
-				   phys_addr_t offset, int vcpu_id)
-{
-	u32 *level_active;
-	u32 *reg, orig;
-	int mode = ACCESS_READ_VALUE | ACCESS_WRITE_CLEARBIT;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-
-	reg = vgic_bitmap_get_reg(&dist->irq_pending, vcpu_id, offset);
-	orig = *reg;
-	vgic_reg_access(mmio, reg, offset, mode);
-	if (mmio->is_write) {
-		/* Re-set level triggered level-active interrupts */
-		level_active = vgic_bitmap_get_reg(&dist->irq_level,
-					  vcpu_id, offset);
-		reg = vgic_bitmap_get_reg(&dist->irq_pending, vcpu_id, offset);
-		*reg |= *level_active;
-
-		/* Ignore writes to SGIs */
-		if (offset < 2) {
-			*reg &= ~0xffff;
-			*reg |= orig & 0xffff;
-		}
-
-		/* Clear soft-pending flags */
-		reg = vgic_bitmap_get_reg(&dist->irq_soft_pend,
-					  vcpu_id, offset);
-		vgic_reg_access(mmio, reg, offset, mode);
-
-		vgic_update_state(kvm);
-		return true;
-	}
-	return false;
-}
-
-bool vgic_handle_set_active_reg(struct kvm *kvm,
-				struct kvm_exit_mmio *mmio,
-				phys_addr_t offset, int vcpu_id)
-{
-	u32 *reg;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-
-	reg = vgic_bitmap_get_reg(&dist->irq_active, vcpu_id, offset);
-	vgic_reg_access(mmio, reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_SETBIT);
-
-	if (mmio->is_write) {
-		vgic_update_state(kvm);
-		return true;
-	}
-
-	return false;
-}
-
-bool vgic_handle_clear_active_reg(struct kvm *kvm,
-				  struct kvm_exit_mmio *mmio,
-				  phys_addr_t offset, int vcpu_id)
-{
-	u32 *reg;
-	struct vgic_dist *dist = &kvm->arch.vgic;
-
-	reg = vgic_bitmap_get_reg(&dist->irq_active, vcpu_id, offset);
-	vgic_reg_access(mmio, reg, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_CLEARBIT);
-
-	if (mmio->is_write) {
-		vgic_update_state(kvm);
-		return true;
-	}
-
-	return false;
-}
-
-static u32 vgic_cfg_expand(u16 val)
-{
-	u32 res = 0;
-	int i;
-
-	/*
-	 * Turn a 16bit value like abcd...mnop into a 32bit word
-	 * a0b0c0d0...m0n0o0p0, which is what the HW cfg register is.
-	 */
-	for (i = 0; i < 16; i++)
-		res |= ((val >> i) & VGIC_CFG_EDGE) << (2 * i + 1);
-
-	return res;
-}
-
-static u16 vgic_cfg_compress(u32 val)
-{
-	u16 res = 0;
-	int i;
-
-	/*
-	 * Turn a 32bit word a0b0c0d0...m0n0o0p0 into 16bit value like
-	 * abcd...mnop which is what we really care about.
-	 */
-	for (i = 0; i < 16; i++)
-		res |= ((val >> (i * 2 + 1)) & VGIC_CFG_EDGE) << i;
-
-	return res;
-}
-
-/*
- * The distributor uses 2 bits per IRQ for the CFG register, but the
- * LSB is always 0. As such, we only keep the upper bit, and use the
- * two above functions to compress/expand the bits
- */
-bool vgic_handle_cfg_reg(u32 *reg, struct kvm_exit_mmio *mmio,
-			 phys_addr_t offset)
-{
-	u32 val;
-
-	if (offset & 4)
-		val = *reg >> 16;
-	else
-		val = *reg & 0xffff;
-
-	val = vgic_cfg_expand(val);
-	vgic_reg_access(mmio, &val, offset,
-			ACCESS_READ_VALUE | ACCESS_WRITE_VALUE);
-	if (mmio->is_write) {
-		/* Ignore writes to read-only SGI and PPI bits */
-		if (offset < 8)
-			return false;
-
-		val = vgic_cfg_compress(val);
-		if (offset & 4) {
-			*reg &= 0xffff;
-			*reg |= val << 16;
-		} else {
-			*reg &= 0xffff << 16;
-			*reg |= val;
-		}
-	}
-
-	return false;
-}
-
-/**
- * vgic_unqueue_irqs - move pending/active IRQs from LRs to the distributor
- * @vgic_cpu: Pointer to the vgic_cpu struct holding the LRs
- *
- * Move any IRQs that have already been assigned to LRs back to the
- * emulated distributor state so that the complete emulated state can be read
- * from the main emulation structures without investigating the LRs.
- */
-void vgic_unqueue_irqs(struct kvm_vcpu *vcpu)
-{
-	u64 elrsr = vgic_get_elrsr(vcpu);
-	unsigned long *elrsr_ptr = u64_to_bitmask(&elrsr);
-	int i;
-
-	for_each_clear_bit(i, elrsr_ptr, vgic->nr_lr) {
-		struct vgic_lr lr = vgic_get_lr(vcpu, i);
-
-		/*
-		 * There are three options for the state bits:
-		 *
-		 * 01: pending
-		 * 10: active
-		 * 11: pending and active
-		 */
-		BUG_ON(!(lr.state & LR_STATE_MASK));
-
-		/* Reestablish SGI source for pending and active IRQs */
-		if (lr.irq < VGIC_NR_SGIS)
-			add_sgi_source(vcpu, lr.irq, lr.source);
-
-		/*
-		 * If the LR holds an active (10) or a pending and active (11)
-		 * interrupt then move the active state to the
-		 * distributor tracking bit.
-		 */
-		if (lr.state & LR_STATE_ACTIVE)
-			vgic_irq_set_active(vcpu, lr.irq);
-
-		/*
-		 * Reestablish the pending state on the distributor and the
-		 * CPU interface and mark the LR as free for other use.
-		 */
-		vgic_retire_lr(i, vcpu);
-
-		/* Finally update the VGIC state. */
-		vgic_update_state(vcpu->kvm);
-	}
-}
-
-const
-struct vgic_io_range *vgic_find_range(const struct vgic_io_range *ranges,
-				      int len, gpa_t offset)
-{
-	while (ranges->len) {
-		if (offset >= ranges->base &&
-		    (offset + len) <= (ranges->base + ranges->len))
-			return ranges;
-		ranges++;
-	}
-
-	return NULL;
-}
-
-static bool vgic_validate_access(const struct vgic_dist *dist,
-				 const struct vgic_io_range *range,
-				 unsigned long offset)
-{
-	int irq;
-
-	if (!range->bits_per_irq)
-		return true;	/* Not an irq-based access */
-
-	irq = offset * 8 / range->bits_per_irq;
-	if (irq >= dist->nr_irqs)
-		return false;
-
-	return true;
-}
-
-/*
- * Call the respective handler function for the given range.
- * We split up any 64 bit accesses into two consecutive 32 bit
- * handler calls and merge the result afterwards.
- * We do this in a little endian fashion regardless of the host's
- * or guest's endianness, because the GIC is always LE and the rest of
- * the code (vgic_reg_access) also puts it in a LE fashion already.
- * At this point we have already identified the handle function, so
- * range points to that one entry and offset is relative to this.
- */
-static bool call_range_handler(struct kvm_vcpu *vcpu,
-			       struct kvm_exit_mmio *mmio,
-			       unsigned long offset,
-			       const struct vgic_io_range *range)
-{
-	struct kvm_exit_mmio mmio32;
-	bool ret;
-
-	if (likely(mmio->len <= 4))
-		return range->handle_mmio(vcpu, mmio, offset);
-
-	/*
-	 * Any access bigger than 4 bytes (that we currently handle in KVM)
-	 * is actually 8 bytes long, caused by a 64-bit access
-	 */
-
-	mmio32.len = 4;
-	mmio32.is_write = mmio->is_write;
-	mmio32.private = mmio->private;
-
-	mmio32.phys_addr = mmio->phys_addr + 4;
-	mmio32.data = &((u32 *)mmio->data)[1];
-	ret = range->handle_mmio(vcpu, &mmio32, offset + 4);
-
-	mmio32.phys_addr = mmio->phys_addr;
-	mmio32.data = &((u32 *)mmio->data)[0];
-	ret |= range->handle_mmio(vcpu, &mmio32, offset);
-
-	return ret;
-}
-
-/**
- * vgic_handle_mmio_access - handle an in-kernel MMIO access
- * This is called by the read/write KVM IO device wrappers below.
- * @vcpu:	pointer to the vcpu performing the access
- * @this:	pointer to the KVM IO device in charge
- * @addr:	guest physical address of the access
- * @len:	size of the access
- * @val:	pointer to the data region
- * @is_write:	read or write access
- *
- * returns true if the MMIO access could be performed
- */
-static int vgic_handle_mmio_access(struct kvm_vcpu *vcpu,
-				   struct kvm_io_device *this, gpa_t addr,
-				   int len, void *val, bool is_write)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	struct vgic_io_device *iodev = container_of(this,
-						    struct vgic_io_device, dev);
-	const struct vgic_io_range *range;
-	struct kvm_exit_mmio mmio;
-	bool updated_state;
-	gpa_t offset;
-
-	offset = addr - iodev->addr;
-	range = vgic_find_range(iodev->reg_ranges, len, offset);
-	if (unlikely(!range || !range->handle_mmio)) {
-		pr_warn("Unhandled access %d %08llx %d\n", is_write, addr, len);
-		return -ENXIO;
-	}
-
-	mmio.phys_addr = addr;
-	mmio.len = len;
-	mmio.is_write = is_write;
-	mmio.data = val;
-	mmio.private = iodev->redist_vcpu;
-
-	spin_lock(&dist->lock);
-	offset -= range->base;
-	if (vgic_validate_access(dist, range, offset)) {
-		updated_state = call_range_handler(vcpu, &mmio, offset, range);
-	} else {
-		if (!is_write)
-			memset(val, 0, len);
-		updated_state = false;
-	}
-	spin_unlock(&dist->lock);
-
-	if (updated_state)
-		vgic_kick_vcpus(vcpu->kvm);
-
-	return 0;
-}
-
-static int vgic_handle_mmio_read(struct kvm_vcpu *vcpu,
-				 struct kvm_io_device *this,
-				 gpa_t addr, int len, void *val)
-{
-	return vgic_handle_mmio_access(vcpu, this, addr, len, val, false);
-}
-
-static int vgic_handle_mmio_write(struct kvm_vcpu *vcpu,
-				  struct kvm_io_device *this,
-				  gpa_t addr, int len, const void *val)
-{
-	return vgic_handle_mmio_access(vcpu, this, addr, len, (void *)val,
-				       true);
-}
-
-static struct kvm_io_device_ops vgic_io_ops = {
-	.read	= vgic_handle_mmio_read,
-	.write	= vgic_handle_mmio_write,
-};
-
-/**
- * vgic_register_kvm_io_dev - register VGIC register frame on the KVM I/O bus
- * @kvm:            The VM structure pointer
- * @base:           The (guest) base address for the register frame
- * @len:            Length of the register frame window
- * @ranges:         Describing the handler functions for each register
- * @redist_vcpu_id: The VCPU ID to pass on to the handlers on call
- * @iodev:          Points to memory to be passed on to the handler
- *
- * @iodev stores the parameters of this function to be usable by the handler
- * respectively the dispatcher function (since the KVM I/O bus framework lacks
- * an opaque parameter). Initialization is done in this function, but the
- * reference should be valid and unique for the whole VGIC lifetime.
- * If the register frame is not mapped for a specific VCPU, pass -1 to
- * @redist_vcpu_id.
- */
-int vgic_register_kvm_io_dev(struct kvm *kvm, gpa_t base, int len,
-			     const struct vgic_io_range *ranges,
-			     int redist_vcpu_id,
-			     struct vgic_io_device *iodev)
-{
-	struct kvm_vcpu *vcpu = NULL;
-	int ret;
-
-	if (redist_vcpu_id >= 0)
-		vcpu = kvm_get_vcpu(kvm, redist_vcpu_id);
-
-	iodev->addr		= base;
-	iodev->len		= len;
-	iodev->reg_ranges	= ranges;
-	iodev->redist_vcpu	= vcpu;
-
-	kvm_iodevice_init(&iodev->dev, &vgic_io_ops);
-
-	mutex_lock(&kvm->slots_lock);
-
-	ret = kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, base, len,
-				      &iodev->dev);
-	mutex_unlock(&kvm->slots_lock);
-
-	/* Mark the iodev as invalid if registration fails. */
-	if (ret)
-		iodev->dev.ops = NULL;
-
-	return ret;
-}
-
-static int vgic_nr_shared_irqs(struct vgic_dist *dist)
-{
-	return dist->nr_irqs - VGIC_NR_PRIVATE_IRQS;
-}
-
-static int compute_active_for_cpu(struct kvm_vcpu *vcpu)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	unsigned long *active, *enabled, *act_percpu, *act_shared;
-	unsigned long active_private, active_shared;
-	int nr_shared = vgic_nr_shared_irqs(dist);
-	int vcpu_id;
-
-	vcpu_id = vcpu->vcpu_id;
-	act_percpu = vcpu->arch.vgic_cpu.active_percpu;
-	act_shared = vcpu->arch.vgic_cpu.active_shared;
-
-	active = vgic_bitmap_get_cpu_map(&dist->irq_active, vcpu_id);
-	enabled = vgic_bitmap_get_cpu_map(&dist->irq_enabled, vcpu_id);
-	bitmap_and(act_percpu, active, enabled, VGIC_NR_PRIVATE_IRQS);
-
-	active = vgic_bitmap_get_shared_map(&dist->irq_active);
-	enabled = vgic_bitmap_get_shared_map(&dist->irq_enabled);
-	bitmap_and(act_shared, active, enabled, nr_shared);
-	bitmap_and(act_shared, act_shared,
-		   vgic_bitmap_get_shared_map(&dist->irq_spi_target[vcpu_id]),
-		   nr_shared);
-
-	active_private = find_first_bit(act_percpu, VGIC_NR_PRIVATE_IRQS);
-	active_shared = find_first_bit(act_shared, nr_shared);
-
-	return (active_private < VGIC_NR_PRIVATE_IRQS ||
-		active_shared < nr_shared);
-}
-
-static int compute_pending_for_cpu(struct kvm_vcpu *vcpu)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	unsigned long *pending, *enabled, *pend_percpu, *pend_shared;
-	unsigned long pending_private, pending_shared;
-	int nr_shared = vgic_nr_shared_irqs(dist);
-	int vcpu_id;
-
-	vcpu_id = vcpu->vcpu_id;
-	pend_percpu = vcpu->arch.vgic_cpu.pending_percpu;
-	pend_shared = vcpu->arch.vgic_cpu.pending_shared;
-
-	if (!dist->enabled) {
-		bitmap_zero(pend_percpu, VGIC_NR_PRIVATE_IRQS);
-		bitmap_zero(pend_shared, nr_shared);
-		return 0;
-	}
-
-	pending = vgic_bitmap_get_cpu_map(&dist->irq_pending, vcpu_id);
-	enabled = vgic_bitmap_get_cpu_map(&dist->irq_enabled, vcpu_id);
-	bitmap_and(pend_percpu, pending, enabled, VGIC_NR_PRIVATE_IRQS);
-
-	pending = vgic_bitmap_get_shared_map(&dist->irq_pending);
-	enabled = vgic_bitmap_get_shared_map(&dist->irq_enabled);
-	bitmap_and(pend_shared, pending, enabled, nr_shared);
-	bitmap_and(pend_shared, pend_shared,
-		   vgic_bitmap_get_shared_map(&dist->irq_spi_target[vcpu_id]),
-		   nr_shared);
-
-	pending_private = find_first_bit(pend_percpu, VGIC_NR_PRIVATE_IRQS);
-	pending_shared = find_first_bit(pend_shared, nr_shared);
-	return (pending_private < VGIC_NR_PRIVATE_IRQS ||
-		pending_shared < vgic_nr_shared_irqs(dist));
-}
-
-/*
- * Update the interrupt state and determine which CPUs have pending
- * or active interrupts. Must be called with distributor lock held.
- */
-void vgic_update_state(struct kvm *kvm)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	struct kvm_vcpu *vcpu;
-	int c;
-
-	kvm_for_each_vcpu(c, vcpu, kvm) {
-		if (compute_pending_for_cpu(vcpu))
-			set_bit(c, dist->irq_pending_on_cpu);
-
-		if (compute_active_for_cpu(vcpu))
-			set_bit(c, dist->irq_active_on_cpu);
-		else
-			clear_bit(c, dist->irq_active_on_cpu);
-	}
-}
-
-static struct vgic_lr vgic_get_lr(const struct kvm_vcpu *vcpu, int lr)
-{
-	return vgic_ops->get_lr(vcpu, lr);
-}
-
-static void vgic_set_lr(struct kvm_vcpu *vcpu, int lr,
-			       struct vgic_lr vlr)
-{
-	vgic_ops->set_lr(vcpu, lr, vlr);
-}
-
-static inline u64 vgic_get_elrsr(struct kvm_vcpu *vcpu)
-{
-	return vgic_ops->get_elrsr(vcpu);
-}
-
-static inline u64 vgic_get_eisr(struct kvm_vcpu *vcpu)
-{
-	return vgic_ops->get_eisr(vcpu);
-}
-
-static inline void vgic_clear_eisr(struct kvm_vcpu *vcpu)
-{
-	vgic_ops->clear_eisr(vcpu);
-}
-
-static inline u32 vgic_get_interrupt_status(struct kvm_vcpu *vcpu)
-{
-	return vgic_ops->get_interrupt_status(vcpu);
-}
-
-static inline void vgic_enable_underflow(struct kvm_vcpu *vcpu)
-{
-	vgic_ops->enable_underflow(vcpu);
-}
-
-static inline void vgic_disable_underflow(struct kvm_vcpu *vcpu)
-{
-	vgic_ops->disable_underflow(vcpu);
-}
-
-void vgic_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcr)
-{
-	vgic_ops->get_vmcr(vcpu, vmcr);
-}
-
-void vgic_set_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcr)
-{
-	vgic_ops->set_vmcr(vcpu, vmcr);
-}
-
-static inline void vgic_enable(struct kvm_vcpu *vcpu)
-{
-	vgic_ops->enable(vcpu);
-}
-
-static void vgic_retire_lr(int lr_nr, struct kvm_vcpu *vcpu)
-{
-	struct vgic_lr vlr = vgic_get_lr(vcpu, lr_nr);
-
-	vgic_irq_clear_queued(vcpu, vlr.irq);
-
-	/*
-	 * We must transfer the pending state back to the distributor before
-	 * retiring the LR, otherwise we may loose edge-triggered interrupts.
-	 */
-	if (vlr.state & LR_STATE_PENDING) {
-		vgic_dist_irq_set_pending(vcpu, vlr.irq);
-		vlr.hwirq = 0;
-	}
-
-	vlr.state = 0;
-	vgic_set_lr(vcpu, lr_nr, vlr);
-}
-
-static bool dist_active_irq(struct kvm_vcpu *vcpu)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	return test_bit(vcpu->vcpu_id, dist->irq_active_on_cpu);
-}
-
-bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq)
-{
-	int i;
-
-	for (i = 0; i < vgic->nr_lr; i++) {
-		struct vgic_lr vlr = vgic_get_lr(vcpu, i);
-
-		if (vlr.irq == virt_irq && vlr.state & LR_STATE_ACTIVE)
-			return true;
-	}
-
-	return vgic_irq_is_active(vcpu, virt_irq);
-}
-
-/*
- * An interrupt may have been disabled after being made pending on the
- * CPU interface (the classic case is a timer running while we're
- * rebooting the guest - the interrupt would kick as soon as the CPU
- * interface gets enabled, with deadly consequences).
- *
- * The solution is to examine already active LRs, and check the
- * interrupt is still enabled. If not, just retire it.
- */
-static void vgic_retire_disabled_irqs(struct kvm_vcpu *vcpu)
-{
-	u64 elrsr = vgic_get_elrsr(vcpu);
-	unsigned long *elrsr_ptr = u64_to_bitmask(&elrsr);
-	int lr;
-
-	for_each_clear_bit(lr, elrsr_ptr, vgic->nr_lr) {
-		struct vgic_lr vlr = vgic_get_lr(vcpu, lr);
-
-		if (!vgic_irq_is_enabled(vcpu, vlr.irq))
-			vgic_retire_lr(lr, vcpu);
-	}
-}
-
-static void vgic_queue_irq_to_lr(struct kvm_vcpu *vcpu, int irq,
-				 int lr_nr, struct vgic_lr vlr)
-{
-	if (vgic_irq_is_active(vcpu, irq)) {
-		vlr.state |= LR_STATE_ACTIVE;
-		kvm_debug("Set active, clear distributor: 0x%x\n", vlr.state);
-		vgic_irq_clear_active(vcpu, irq);
-		vgic_update_state(vcpu->kvm);
-	} else {
-		WARN_ON(!vgic_dist_irq_is_pending(vcpu, irq));
-		vlr.state |= LR_STATE_PENDING;
-		kvm_debug("Set pending: 0x%x\n", vlr.state);
-	}
-
-	if (!vgic_irq_is_edge(vcpu, irq))
-		vlr.state |= LR_EOI_INT;
-
-	if (vlr.irq >= VGIC_NR_SGIS) {
-		struct irq_phys_map *map;
-		map = vgic_irq_map_search(vcpu, irq);
-
-		if (map) {
-			vlr.hwirq = map->phys_irq;
-			vlr.state |= LR_HW;
-			vlr.state &= ~LR_EOI_INT;
-
-			/*
-			 * Make sure we're not going to sample this
-			 * again, as a HW-backed interrupt cannot be
-			 * in the PENDING_ACTIVE stage.
-			 */
-			vgic_irq_set_queued(vcpu, irq);
-		}
-	}
-
-	vgic_set_lr(vcpu, lr_nr, vlr);
-}
-
-/*
- * Queue an interrupt to a CPU virtual interface. Return true on success,
- * or false if it wasn't possible to queue it.
- * sgi_source must be zero for any non-SGI interrupts.
- */
-bool vgic_queue_irq(struct kvm_vcpu *vcpu, u8 sgi_source_id, int irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	u64 elrsr = vgic_get_elrsr(vcpu);
-	unsigned long *elrsr_ptr = u64_to_bitmask(&elrsr);
-	struct vgic_lr vlr;
-	int lr;
-
-	/* Sanitize the input... */
-	BUG_ON(sgi_source_id & ~7);
-	BUG_ON(sgi_source_id && irq >= VGIC_NR_SGIS);
-	BUG_ON(irq >= dist->nr_irqs);
-
-	kvm_debug("Queue IRQ%d\n", irq);
-
-	/* Do we have an active interrupt for the same CPUID? */
-	for_each_clear_bit(lr, elrsr_ptr, vgic->nr_lr) {
-		vlr = vgic_get_lr(vcpu, lr);
-		if (vlr.irq == irq && vlr.source == sgi_source_id) {
-			kvm_debug("LR%d piggyback for IRQ%d\n", lr, vlr.irq);
-			vgic_queue_irq_to_lr(vcpu, irq, lr, vlr);
-			return true;
-		}
-	}
-
-	/* Try to use another LR for this interrupt */
-	lr = find_first_bit(elrsr_ptr, vgic->nr_lr);
-	if (lr >= vgic->nr_lr)
-		return false;
-
-	kvm_debug("LR%d allocated for IRQ%d %x\n", lr, irq, sgi_source_id);
-
-	vlr.irq = irq;
-	vlr.source = sgi_source_id;
-	vlr.state = 0;
-	vgic_queue_irq_to_lr(vcpu, irq, lr, vlr);
-
-	return true;
-}
-
-static bool vgic_queue_hwirq(struct kvm_vcpu *vcpu, int irq)
-{
-	if (!vgic_can_sample_irq(vcpu, irq))
-		return true; /* level interrupt, already queued */
-
-	if (vgic_queue_irq(vcpu, 0, irq)) {
-		if (vgic_irq_is_edge(vcpu, irq)) {
-			vgic_dist_irq_clear_pending(vcpu, irq);
-			vgic_cpu_irq_clear(vcpu, irq);
-		} else {
-			vgic_irq_set_queued(vcpu, irq);
-		}
-
-		return true;
-	}
-
-	return false;
-}
-
-/*
- * Fill the list registers with pending interrupts before running the
- * guest.
- */
-static void __kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu)
-{
-	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	unsigned long *pa_percpu, *pa_shared;
-	int i, vcpu_id;
-	int overflow = 0;
-	int nr_shared = vgic_nr_shared_irqs(dist);
-
-	vcpu_id = vcpu->vcpu_id;
-
-	pa_percpu = vcpu->arch.vgic_cpu.pend_act_percpu;
-	pa_shared = vcpu->arch.vgic_cpu.pend_act_shared;
-
-	bitmap_or(pa_percpu, vgic_cpu->pending_percpu, vgic_cpu->active_percpu,
-		  VGIC_NR_PRIVATE_IRQS);
-	bitmap_or(pa_shared, vgic_cpu->pending_shared, vgic_cpu->active_shared,
-		  nr_shared);
-	/*
-	 * We may not have any pending interrupt, or the interrupts
-	 * may have been serviced from another vcpu. In all cases,
-	 * move along.
-	 */
-	if (!kvm_vgic_vcpu_pending_irq(vcpu) && !dist_active_irq(vcpu))
-		goto epilog;
-
-	/* SGIs */
-	for_each_set_bit(i, pa_percpu, VGIC_NR_SGIS) {
-		if (!queue_sgi(vcpu, i))
-			overflow = 1;
-	}
-
-	/* PPIs */
-	for_each_set_bit_from(i, pa_percpu, VGIC_NR_PRIVATE_IRQS) {
-		if (!vgic_queue_hwirq(vcpu, i))
-			overflow = 1;
-	}
-
-	/* SPIs */
-	for_each_set_bit(i, pa_shared, nr_shared) {
-		if (!vgic_queue_hwirq(vcpu, i + VGIC_NR_PRIVATE_IRQS))
-			overflow = 1;
-	}
-
-
-
-
-epilog:
-	if (overflow) {
-		vgic_enable_underflow(vcpu);
-	} else {
-		vgic_disable_underflow(vcpu);
-		/*
-		 * We're about to run this VCPU, and we've consumed
-		 * everything the distributor had in store for
-		 * us. Claim we don't have anything pending. We'll
-		 * adjust that if needed while exiting.
-		 */
-		clear_bit(vcpu_id, dist->irq_pending_on_cpu);
-	}
-}
-
-static int process_queued_irq(struct kvm_vcpu *vcpu,
-				   int lr, struct vgic_lr vlr)
-{
-	int pending = 0;
-
-	/*
-	 * If the IRQ was EOIed (called from vgic_process_maintenance) or it
-	 * went from active to non-active (called from vgic_sync_hwirq) it was
-	 * also ACKed and we we therefore assume we can clear the soft pending
-	 * state (should it had been set) for this interrupt.
-	 *
-	 * Note: if the IRQ soft pending state was set after the IRQ was
-	 * acked, it actually shouldn't be cleared, but we have no way of
-	 * knowing that unless we start trapping ACKs when the soft-pending
-	 * state is set.
-	 */
-	vgic_dist_irq_clear_soft_pend(vcpu, vlr.irq);
-
-	/*
-	 * Tell the gic to start sampling this interrupt again.
-	 */
-	vgic_irq_clear_queued(vcpu, vlr.irq);
-
-	/* Any additional pending interrupt? */
-	if (vgic_irq_is_edge(vcpu, vlr.irq)) {
-		BUG_ON(!(vlr.state & LR_HW));
-		pending = vgic_dist_irq_is_pending(vcpu, vlr.irq);
-	} else {
-		if (vgic_dist_irq_get_level(vcpu, vlr.irq)) {
-			vgic_cpu_irq_set(vcpu, vlr.irq);
-			pending = 1;
-		} else {
-			vgic_dist_irq_clear_pending(vcpu, vlr.irq);
-			vgic_cpu_irq_clear(vcpu, vlr.irq);
-		}
-	}
-
-	/*
-	 * Despite being EOIed, the LR may not have
-	 * been marked as empty.
-	 */
-	vlr.state = 0;
-	vlr.hwirq = 0;
-	vgic_set_lr(vcpu, lr, vlr);
-
-	return pending;
-}
-
-static bool vgic_process_maintenance(struct kvm_vcpu *vcpu)
-{
-	u32 status = vgic_get_interrupt_status(vcpu);
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	struct kvm *kvm = vcpu->kvm;
-	int level_pending = 0;
-
-	kvm_debug("STATUS = %08x\n", status);
-
-	if (status & INT_STATUS_EOI) {
-		/*
-		 * Some level interrupts have been EOIed. Clear their
-		 * active bit.
-		 */
-		u64 eisr = vgic_get_eisr(vcpu);
-		unsigned long *eisr_ptr = u64_to_bitmask(&eisr);
-		int lr;
-
-		for_each_set_bit(lr, eisr_ptr, vgic->nr_lr) {
-			struct vgic_lr vlr = vgic_get_lr(vcpu, lr);
-
-			WARN_ON(vgic_irq_is_edge(vcpu, vlr.irq));
-			WARN_ON(vlr.state & LR_STATE_MASK);
-
-
-			/*
-			 * kvm_notify_acked_irq calls kvm_set_irq()
-			 * to reset the IRQ level, which grabs the dist->lock
-			 * so we call this before taking the dist->lock.
-			 */
-			kvm_notify_acked_irq(kvm, 0,
-					     vlr.irq - VGIC_NR_PRIVATE_IRQS);
-
-			spin_lock(&dist->lock);
-			level_pending |= process_queued_irq(vcpu, lr, vlr);
-			spin_unlock(&dist->lock);
-		}
-	}
-
-	if (status & INT_STATUS_UNDERFLOW)
-		vgic_disable_underflow(vcpu);
-
-	/*
-	 * In the next iterations of the vcpu loop, if we sync the vgic state
-	 * after flushing it, but before entering the guest (this happens for
-	 * pending signals and vmid rollovers), then make sure we don't pick
-	 * up any old maintenance interrupts here.
-	 */
-	vgic_clear_eisr(vcpu);
-
-	return level_pending;
-}
-
-/*
- * Save the physical active state, and reset it to inactive.
- *
- * Return true if there's a pending forwarded interrupt to queue.
- */
-static bool vgic_sync_hwirq(struct kvm_vcpu *vcpu, int lr, struct vgic_lr vlr)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	bool level_pending;
-
-	if (!(vlr.state & LR_HW))
-		return false;
-
-	if (vlr.state & LR_STATE_ACTIVE)
-		return false;
-
-	spin_lock(&dist->lock);
-	level_pending = process_queued_irq(vcpu, lr, vlr);
-	spin_unlock(&dist->lock);
-	return level_pending;
-}
-
-/* Sync back the VGIC state after a guest run */
-static void __kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	u64 elrsr;
-	unsigned long *elrsr_ptr;
-	int lr, pending;
-	bool level_pending;
-
-	level_pending = vgic_process_maintenance(vcpu);
-
-	/* Deal with HW interrupts, and clear mappings for empty LRs */
-	for (lr = 0; lr < vgic->nr_lr; lr++) {
-		struct vgic_lr vlr = vgic_get_lr(vcpu, lr);
-
-		level_pending |= vgic_sync_hwirq(vcpu, lr, vlr);
-		BUG_ON(vlr.irq >= dist->nr_irqs);
-	}
-
-	/* Check if we still have something up our sleeve... */
-	elrsr = vgic_get_elrsr(vcpu);
-	elrsr_ptr = u64_to_bitmask(&elrsr);
-	pending = find_first_zero_bit(elrsr_ptr, vgic->nr_lr);
-	if (level_pending || pending < vgic->nr_lr)
-		set_bit(vcpu->vcpu_id, dist->irq_pending_on_cpu);
-}
-
-void kvm_vgic_flush_hwstate(struct kvm_vcpu *vcpu)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	if (!irqchip_in_kernel(vcpu->kvm))
-		return;
-
-	spin_lock(&dist->lock);
-	__kvm_vgic_flush_hwstate(vcpu);
-	spin_unlock(&dist->lock);
-}
-
-void kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu)
-{
-	if (!irqchip_in_kernel(vcpu->kvm))
-		return;
-
-	__kvm_vgic_sync_hwstate(vcpu);
-}
-
-int kvm_vgic_vcpu_pending_irq(struct kvm_vcpu *vcpu)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-
-	if (!irqchip_in_kernel(vcpu->kvm))
-		return 0;
-
-	return test_bit(vcpu->vcpu_id, dist->irq_pending_on_cpu);
-}
-
-void vgic_kick_vcpus(struct kvm *kvm)
-{
-	struct kvm_vcpu *vcpu;
-	int c;
-
-	/*
-	 * We've injected an interrupt, time to find out who deserves
-	 * a good kick...
-	 */
-	kvm_for_each_vcpu(c, vcpu, kvm) {
-		if (kvm_vgic_vcpu_pending_irq(vcpu))
-			kvm_vcpu_kick(vcpu);
-	}
-}
-
-static int vgic_validate_injection(struct kvm_vcpu *vcpu, int irq, int level)
-{
-	int edge_triggered = vgic_irq_is_edge(vcpu, irq);
-
-	/*
-	 * Only inject an interrupt if:
-	 * - edge triggered and we have a rising edge
-	 * - level triggered and we change level
-	 */
-	if (edge_triggered) {
-		int state = vgic_dist_irq_is_pending(vcpu, irq);
-		return level > state;
-	} else {
-		int state = vgic_dist_irq_get_level(vcpu, irq);
-		return level != state;
-	}
-}
-
-static int vgic_update_irq_pending(struct kvm *kvm, int cpuid,
-				   unsigned int irq_num, bool level)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	struct kvm_vcpu *vcpu;
-	int edge_triggered, level_triggered;
-	int enabled;
-	bool ret = true, can_inject = true;
-
-	trace_vgic_update_irq_pending(cpuid, irq_num, level);
-
-	if (irq_num >= min(kvm->arch.vgic.nr_irqs, 1020))
-		return -EINVAL;
-
-	spin_lock(&dist->lock);
-
-	vcpu = kvm_get_vcpu(kvm, cpuid);
-	edge_triggered = vgic_irq_is_edge(vcpu, irq_num);
-	level_triggered = !edge_triggered;
-
-	if (!vgic_validate_injection(vcpu, irq_num, level)) {
-		ret = false;
-		goto out;
-	}
-
-	if (irq_num >= VGIC_NR_PRIVATE_IRQS) {
-		cpuid = dist->irq_spi_cpu[irq_num - VGIC_NR_PRIVATE_IRQS];
-		if (cpuid == VCPU_NOT_ALLOCATED) {
-			/* Pretend we use CPU0, and prevent injection */
-			cpuid = 0;
-			can_inject = false;
-		}
-		vcpu = kvm_get_vcpu(kvm, cpuid);
-	}
-
-	kvm_debug("Inject IRQ%d level %d CPU%d\n", irq_num, level, cpuid);
-
-	if (level) {
-		if (level_triggered)
-			vgic_dist_irq_set_level(vcpu, irq_num);
-		vgic_dist_irq_set_pending(vcpu, irq_num);
-	} else {
-		if (level_triggered) {
-			vgic_dist_irq_clear_level(vcpu, irq_num);
-			if (!vgic_dist_irq_soft_pend(vcpu, irq_num)) {
-				vgic_dist_irq_clear_pending(vcpu, irq_num);
-				vgic_cpu_irq_clear(vcpu, irq_num);
-				if (!compute_pending_for_cpu(vcpu))
-					clear_bit(cpuid, dist->irq_pending_on_cpu);
-			}
-		}
-
-		ret = false;
-		goto out;
-	}
-
-	enabled = vgic_irq_is_enabled(vcpu, irq_num);
-
-	if (!enabled || !can_inject) {
-		ret = false;
-		goto out;
-	}
-
-	if (!vgic_can_sample_irq(vcpu, irq_num)) {
-		/*
-		 * Level interrupt in progress, will be picked up
-		 * when EOId.
-		 */
-		ret = false;
-		goto out;
-	}
-
-	if (level) {
-		vgic_cpu_irq_set(vcpu, irq_num);
-		set_bit(cpuid, dist->irq_pending_on_cpu);
-	}
-
-out:
-	spin_unlock(&dist->lock);
-
-	if (ret) {
-		/* kick the specified vcpu */
-		kvm_vcpu_kick(kvm_get_vcpu(kvm, cpuid));
-	}
-
-	return 0;
-}
-
-static int vgic_lazy_init(struct kvm *kvm)
-{
-	int ret = 0;
-
-	if (unlikely(!vgic_initialized(kvm))) {
-		/*
-		 * We only provide the automatic initialization of the VGIC
-		 * for the legacy case of a GICv2. Any other type must
-		 * be explicitly initialized once setup with the respective
-		 * KVM device call.
-		 */
-		if (kvm->arch.vgic.vgic_model != KVM_DEV_TYPE_ARM_VGIC_V2)
-			return -EBUSY;
-
-		mutex_lock(&kvm->lock);
-		ret = vgic_init(kvm);
-		mutex_unlock(&kvm->lock);
-	}
-
-	return ret;
-}
-
-/**
- * kvm_vgic_inject_irq - Inject an IRQ from a device to the vgic
- * @kvm:     The VM structure pointer
- * @cpuid:   The CPU for PPIs
- * @irq_num: The IRQ number that is assigned to the device. This IRQ
- *           must not be mapped to a HW interrupt.
- * @level:   Edge-triggered:  true:  to trigger the interrupt
- *			      false: to ignore the call
- *	     Level-sensitive  true:  raise the input signal
- *			      false: lower the input signal
- *
- * The GIC is not concerned with devices being active-LOW or active-HIGH for
- * level-sensitive interrupts.  You can think of the level parameter as 1
- * being HIGH and 0 being LOW and all devices being active-HIGH.
- */
-int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int irq_num,
-			bool level)
-{
-	struct irq_phys_map *map;
-	int ret;
-
-	ret = vgic_lazy_init(kvm);
-	if (ret)
-		return ret;
-
-	map = vgic_irq_map_search(kvm_get_vcpu(kvm, cpuid), irq_num);
-	if (map)
-		return -EINVAL;
-
-	return vgic_update_irq_pending(kvm, cpuid, irq_num, level);
-}
-
-/**
- * kvm_vgic_inject_mapped_irq - Inject a physically mapped IRQ to the vgic
- * @kvm:     The VM structure pointer
- * @cpuid:   The CPU for PPIs
- * @virt_irq: The virtual IRQ to be injected
- * @level:   Edge-triggered:  true:  to trigger the interrupt
- *			      false: to ignore the call
- *	     Level-sensitive  true:  raise the input signal
- *			      false: lower the input signal
- *
- * The GIC is not concerned with devices being active-LOW or active-HIGH for
- * level-sensitive interrupts.  You can think of the level parameter as 1
- * being HIGH and 0 being LOW and all devices being active-HIGH.
- */
-int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid,
-			       unsigned int virt_irq, bool level)
-{
-	int ret;
-
-	ret = vgic_lazy_init(kvm);
-	if (ret)
-		return ret;
-
-	return vgic_update_irq_pending(kvm, cpuid, virt_irq, level);
-}
-
-static irqreturn_t vgic_maintenance_handler(int irq, void *data)
-{
-	/*
-	 * We cannot rely on the vgic maintenance interrupt to be
-	 * delivered synchronously. This means we can only use it to
-	 * exit the VM, and we perform the handling of EOIed
-	 * interrupts on the exit path (see vgic_process_maintenance).
-	 */
-	return IRQ_HANDLED;
-}
-
-static struct list_head *vgic_get_irq_phys_map_list(struct kvm_vcpu *vcpu,
-						    int virt_irq)
-{
-	if (virt_irq < VGIC_NR_PRIVATE_IRQS)
-		return &vcpu->arch.vgic_cpu.irq_phys_map_list;
-	else
-		return &vcpu->kvm->arch.vgic.irq_phys_map_list;
-}
-
-/**
- * kvm_vgic_map_phys_irq - map a virtual IRQ to a physical IRQ
- * @vcpu: The VCPU pointer
- * @virt_irq: The virtual IRQ number for the guest
- * @phys_irq: The hardware IRQ number of the host
- *
- * Establish a mapping between a guest visible irq (@virt_irq) and a
- * hardware irq (@phys_irq). On injection, @virt_irq will be associated with
- * the physical interrupt represented by @phys_irq. This mapping can be
- * established multiple times as long as the parameters are the same.
- *
- * Returns 0 on success or an error value otherwise.
- */
-int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, int virt_irq, int phys_irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	struct list_head *root = vgic_get_irq_phys_map_list(vcpu, virt_irq);
-	struct irq_phys_map *map;
-	struct irq_phys_map_entry *entry;
-	int ret = 0;
-
-	/* Create a new mapping */
-	entry = kzalloc(sizeof(*entry), GFP_KERNEL);
-	if (!entry)
-		return -ENOMEM;
-
-	spin_lock(&dist->irq_phys_map_lock);
-
-	/* Try to match an existing mapping */
-	map = vgic_irq_map_search(vcpu, virt_irq);
-	if (map) {
-		/* Make sure this mapping matches */
-		if (map->phys_irq != phys_irq)
-			ret = -EINVAL;
-
-		/* Found an existing, valid mapping */
-		goto out;
-	}
-
-	map           = &entry->map;
-	map->virt_irq = virt_irq;
-	map->phys_irq = phys_irq;
-
-	list_add_tail_rcu(&entry->entry, root);
-
-out:
-	spin_unlock(&dist->irq_phys_map_lock);
-	/* If we've found a hit in the existing list, free the useless
-	 * entry */
-	if (ret || map != &entry->map)
-		kfree(entry);
-	return ret;
-}
-
-static struct irq_phys_map *vgic_irq_map_search(struct kvm_vcpu *vcpu,
-						int virt_irq)
-{
-	struct list_head *root = vgic_get_irq_phys_map_list(vcpu, virt_irq);
-	struct irq_phys_map_entry *entry;
-	struct irq_phys_map *map;
-
-	rcu_read_lock();
-
-	list_for_each_entry_rcu(entry, root, entry) {
-		map = &entry->map;
-		if (map->virt_irq == virt_irq) {
-			rcu_read_unlock();
-			return map;
-		}
-	}
-
-	rcu_read_unlock();
-
-	return NULL;
-}
-
-static void vgic_free_phys_irq_map_rcu(struct rcu_head *rcu)
-{
-	struct irq_phys_map_entry *entry;
-
-	entry = container_of(rcu, struct irq_phys_map_entry, rcu);
-	kfree(entry);
-}
-
-/**
- * kvm_vgic_unmap_phys_irq - Remove a virtual to physical IRQ mapping
- * @vcpu: The VCPU pointer
- * @virt_irq: The virtual IRQ number to be unmapped
- *
- * Remove an existing mapping between virtual and physical interrupts.
- */
-int kvm_vgic_unmap_phys_irq(struct kvm_vcpu *vcpu, unsigned int virt_irq)
-{
-	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
-	struct irq_phys_map_entry *entry;
-	struct list_head *root;
-
-	root = vgic_get_irq_phys_map_list(vcpu, virt_irq);
-
-	spin_lock(&dist->irq_phys_map_lock);
-
-	list_for_each_entry(entry, root, entry) {
-		if (entry->map.virt_irq == virt_irq) {
-			list_del_rcu(&entry->entry);
-			call_rcu(&entry->rcu, vgic_free_phys_irq_map_rcu);
-			break;
-		}
-	}
-
-	spin_unlock(&dist->irq_phys_map_lock);
-
-	return 0;
-}
-
-static void vgic_destroy_irq_phys_map(struct kvm *kvm, struct list_head *root)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	struct irq_phys_map_entry *entry;
-
-	spin_lock(&dist->irq_phys_map_lock);
-
-	list_for_each_entry(entry, root, entry) {
-		list_del_rcu(&entry->entry);
-		call_rcu(&entry->rcu, vgic_free_phys_irq_map_rcu);
-	}
-
-	spin_unlock(&dist->irq_phys_map_lock);
-}
-
-void kvm_vgic_vcpu_destroy(struct kvm_vcpu *vcpu)
-{
-	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
-
-	kfree(vgic_cpu->pending_shared);
-	kfree(vgic_cpu->active_shared);
-	kfree(vgic_cpu->pend_act_shared);
-	vgic_destroy_irq_phys_map(vcpu->kvm, &vgic_cpu->irq_phys_map_list);
-	vgic_cpu->pending_shared = NULL;
-	vgic_cpu->active_shared = NULL;
-	vgic_cpu->pend_act_shared = NULL;
-}
-
-static int vgic_vcpu_init_maps(struct kvm_vcpu *vcpu, int nr_irqs)
-{
-	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
-	int nr_longs = BITS_TO_LONGS(nr_irqs - VGIC_NR_PRIVATE_IRQS);
-	int sz = nr_longs * sizeof(unsigned long);
-	vgic_cpu->pending_shared = kzalloc(sz, GFP_KERNEL);
-	vgic_cpu->active_shared = kzalloc(sz, GFP_KERNEL);
-	vgic_cpu->pend_act_shared = kzalloc(sz, GFP_KERNEL);
-
-	if (!vgic_cpu->pending_shared
-		|| !vgic_cpu->active_shared
-		|| !vgic_cpu->pend_act_shared) {
-		kvm_vgic_vcpu_destroy(vcpu);
-		return -ENOMEM;
-	}
-
-	return 0;
-}
-
-/**
- * kvm_vgic_vcpu_early_init - Earliest possible per-vcpu vgic init stage
- *
- * No memory allocation should be performed here, only static init.
- */
-void kvm_vgic_vcpu_early_init(struct kvm_vcpu *vcpu)
-{
-	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
-	INIT_LIST_HEAD(&vgic_cpu->irq_phys_map_list);
-}
-
-/**
- * kvm_vgic_get_max_vcpus - Get the maximum number of VCPUs allowed by HW
- *
- * The host's GIC naturally limits the maximum amount of VCPUs a guest
- * can use.
- */
-int kvm_vgic_get_max_vcpus(void)
-{
-	return vgic->max_gic_vcpus;
-}
-
-void kvm_vgic_destroy(struct kvm *kvm)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	struct kvm_vcpu *vcpu;
-	int i;
-
-	kvm_for_each_vcpu(i, vcpu, kvm)
-		kvm_vgic_vcpu_destroy(vcpu);
-
-	vgic_free_bitmap(&dist->irq_enabled);
-	vgic_free_bitmap(&dist->irq_level);
-	vgic_free_bitmap(&dist->irq_pending);
-	vgic_free_bitmap(&dist->irq_soft_pend);
-	vgic_free_bitmap(&dist->irq_queued);
-	vgic_free_bitmap(&dist->irq_cfg);
-	vgic_free_bytemap(&dist->irq_priority);
-	if (dist->irq_spi_target) {
-		for (i = 0; i < dist->nr_cpus; i++)
-			vgic_free_bitmap(&dist->irq_spi_target[i]);
-	}
-	kfree(dist->irq_sgi_sources);
-	kfree(dist->irq_spi_cpu);
-	kfree(dist->irq_spi_mpidr);
-	kfree(dist->irq_spi_target);
-	kfree(dist->irq_pending_on_cpu);
-	kfree(dist->irq_active_on_cpu);
-	vgic_destroy_irq_phys_map(kvm, &dist->irq_phys_map_list);
-	dist->irq_sgi_sources = NULL;
-	dist->irq_spi_cpu = NULL;
-	dist->irq_spi_target = NULL;
-	dist->irq_pending_on_cpu = NULL;
-	dist->irq_active_on_cpu = NULL;
-	dist->nr_cpus = 0;
-}
-
-/*
- * Allocate and initialize the various data structures. Must be called
- * with kvm->lock held!
- */
-int vgic_init(struct kvm *kvm)
-{
-	struct vgic_dist *dist = &kvm->arch.vgic;
-	struct kvm_vcpu *vcpu;
-	int nr_cpus, nr_irqs;
-	int ret, i, vcpu_id;
-
-	if (vgic_initialized(kvm))
-		return 0;
-
-	nr_cpus = dist->nr_cpus = atomic_read(&kvm->online_vcpus);
-	if (!nr_cpus)		/* No vcpus? Can't be good... */
-		return -ENODEV;
-
-	/*
-	 * If nobody configured the number of interrupts, use the
-	 * legacy one.
-	 */
-	if (!dist->nr_irqs)
-		dist->nr_irqs = VGIC_NR_IRQS_LEGACY;
-
-	nr_irqs = dist->nr_irqs;
-
-	ret  = vgic_init_bitmap(&dist->irq_enabled, nr_cpus, nr_irqs);
-	ret |= vgic_init_bitmap(&dist->irq_level, nr_cpus, nr_irqs);
-	ret |= vgic_init_bitmap(&dist->irq_pending, nr_cpus, nr_irqs);
-	ret |= vgic_init_bitmap(&dist->irq_soft_pend, nr_cpus, nr_irqs);
-	ret |= vgic_init_bitmap(&dist->irq_queued, nr_cpus, nr_irqs);
-	ret |= vgic_init_bitmap(&dist->irq_active, nr_cpus, nr_irqs);
-	ret |= vgic_init_bitmap(&dist->irq_cfg, nr_cpus, nr_irqs);
-	ret |= vgic_init_bytemap(&dist->irq_priority, nr_cpus, nr_irqs);
-
-	if (ret)
-		goto out;
-
-	dist->irq_sgi_sources = kzalloc(nr_cpus * VGIC_NR_SGIS, GFP_KERNEL);
-	dist->irq_spi_cpu = kzalloc(nr_irqs - VGIC_NR_PRIVATE_IRQS, GFP_KERNEL);
-	dist->irq_spi_target = kzalloc(sizeof(*dist->irq_spi_target) * nr_cpus,
-				       GFP_KERNEL);
-	dist->irq_pending_on_cpu = kzalloc(BITS_TO_LONGS(nr_cpus) * sizeof(long),
-					   GFP_KERNEL);
-	dist->irq_active_on_cpu = kzalloc(BITS_TO_LONGS(nr_cpus) * sizeof(long),
-					   GFP_KERNEL);
-	if (!dist->irq_sgi_sources ||
-	    !dist->irq_spi_cpu ||
-	    !dist->irq_spi_target ||
-	    !dist->irq_pending_on_cpu ||
-	    !dist->irq_active_on_cpu) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
-	for (i = 0; i < nr_cpus; i++)
-		ret |= vgic_init_bitmap(&dist->irq_spi_target[i],
-					nr_cpus, nr_irqs);
-
-	if (ret)
-		goto out;
-
-	ret = kvm->arch.vgic.vm_ops.init_model(kvm);
-	if (ret)
-		goto out;
-
-	kvm_for_each_vcpu(vcpu_id, vcpu, kvm) {
-		ret = vgic_vcpu_init_maps(vcpu, nr_irqs);
-		if (ret) {
-			kvm_err("VGIC: Failed to allocate vcpu memory\n");
-			break;
-		}
-
-		/*
-		 * Enable and configure all SGIs to be edge-triggere and
-		 * configure all PPIs as level-triggered.
-		 */
-		for (i = 0; i < VGIC_NR_PRIVATE_IRQS; i++) {
-			if (i < VGIC_NR_SGIS) {
-				/* SGIs */
-				vgic_bitmap_set_irq_val(&dist->irq_enabled,
-							vcpu->vcpu_id, i, 1);
-				vgic_bitmap_set_irq_val(&dist->irq_cfg,
-							vcpu->vcpu_id, i,
-							VGIC_CFG_EDGE);
-			} else if (i < VGIC_NR_PRIVATE_IRQS) {
-				/* PPIs */
-				vgic_bitmap_set_irq_val(&dist->irq_cfg,
-							vcpu->vcpu_id, i,
-							VGIC_CFG_LEVEL);
-			}
-		}
-
-		vgic_enable(vcpu);
-	}
-
-out:
-	if (ret)
-		kvm_vgic_destroy(kvm);
-
-	return ret;
-}
-
-static int init_vgic_model(struct kvm *kvm, int type)
-{
-	switch (type) {
-	case KVM_DEV_TYPE_ARM_VGIC_V2:
-		vgic_v2_init_emulation(kvm);
-		break;
-#ifdef CONFIG_KVM_ARM_VGIC_V3
-	case KVM_DEV_TYPE_ARM_VGIC_V3:
-		vgic_v3_init_emulation(kvm);
-		break;
-#endif
-	default:
-		return -ENODEV;
-	}
-
-	if (atomic_read(&kvm->online_vcpus) > kvm->arch.max_vcpus)
-		return -E2BIG;
-
-	return 0;
-}
-
-/**
- * kvm_vgic_early_init - Earliest possible vgic initialization stage
- *
- * No memory allocation should be performed here, only static init.
- */
-void kvm_vgic_early_init(struct kvm *kvm)
-{
-	spin_lock_init(&kvm->arch.vgic.lock);
-	spin_lock_init(&kvm->arch.vgic.irq_phys_map_lock);
-	INIT_LIST_HEAD(&kvm->arch.vgic.irq_phys_map_list);
-}
-
-int kvm_vgic_create(struct kvm *kvm, u32 type)
-{
-	int i, vcpu_lock_idx = -1, ret;
-	struct kvm_vcpu *vcpu;
-
-	mutex_lock(&kvm->lock);
-
-	if (irqchip_in_kernel(kvm)) {
-		ret = -EEXIST;
-		goto out;
-	}
-
-	/*
-	 * This function is also called by the KVM_CREATE_IRQCHIP handler,
-	 * which had no chance yet to check the availability of the GICv2
-	 * emulation. So check this here again. KVM_CREATE_DEVICE does
-	 * the proper checks already.
-	 */
-	if (type == KVM_DEV_TYPE_ARM_VGIC_V2 && !vgic->can_emulate_gicv2) {
-		ret = -ENODEV;
-		goto out;
-	}
-
-	/*
-	 * Any time a vcpu is run, vcpu_load is called which tries to grab the
-	 * vcpu->mutex.  By grabbing the vcpu->mutex of all VCPUs we ensure
-	 * that no other VCPUs are run while we create the vgic.
-	 */
-	ret = -EBUSY;
-	kvm_for_each_vcpu(i, vcpu, kvm) {
-		if (!mutex_trylock(&vcpu->mutex))
-			goto out_unlock;
-		vcpu_lock_idx = i;
-	}
-
-	kvm_for_each_vcpu(i, vcpu, kvm) {
-		if (vcpu->arch.has_run_once)
-			goto out_unlock;
-	}
-	ret = 0;
-
-	ret = init_vgic_model(kvm, type);
-	if (ret)
-		goto out_unlock;
-
-	kvm->arch.vgic.in_kernel = true;
-	kvm->arch.vgic.vgic_model = type;
-	kvm->arch.vgic.vctrl_base = vgic->vctrl_base;
-	kvm->arch.vgic.vgic_dist_base = VGIC_ADDR_UNDEF;
-	kvm->arch.vgic.vgic_cpu_base = VGIC_ADDR_UNDEF;
-	kvm->arch.vgic.vgic_redist_base = VGIC_ADDR_UNDEF;
-
-out_unlock:
-	for (; vcpu_lock_idx >= 0; vcpu_lock_idx--) {
-		vcpu = kvm_get_vcpu(kvm, vcpu_lock_idx);
-		mutex_unlock(&vcpu->mutex);
-	}
-
-out:
-	mutex_unlock(&kvm->lock);
-	return ret;
-}
-
-static int vgic_ioaddr_overlap(struct kvm *kvm)
-{
-	phys_addr_t dist = kvm->arch.vgic.vgic_dist_base;
-	phys_addr_t cpu = kvm->arch.vgic.vgic_cpu_base;
-
-	if (IS_VGIC_ADDR_UNDEF(dist) || IS_VGIC_ADDR_UNDEF(cpu))
-		return 0;
-	if ((dist <= cpu && dist + KVM_VGIC_V2_DIST_SIZE > cpu) ||
-	    (cpu <= dist && cpu + KVM_VGIC_V2_CPU_SIZE > dist))
-		return -EBUSY;
-	return 0;
-}
-
-static int vgic_ioaddr_assign(struct kvm *kvm, phys_addr_t *ioaddr,
-			      phys_addr_t addr, phys_addr_t size)
-{
-	int ret;
-
-	if (addr & ~KVM_PHYS_MASK)
-		return -E2BIG;
-
-	if (addr & (SZ_4K - 1))
-		return -EINVAL;
-
-	if (!IS_VGIC_ADDR_UNDEF(*ioaddr))
-		return -EEXIST;
-	if (addr + size < addr)
-		return -EINVAL;
-
-	*ioaddr = addr;
-	ret = vgic_ioaddr_overlap(kvm);
-	if (ret)
-		*ioaddr = VGIC_ADDR_UNDEF;
-
-	return ret;
-}
-
-/**
- * kvm_vgic_addr - set or get vgic VM base addresses
- * @kvm:   pointer to the vm struct
- * @type:  the VGIC addr type, one of KVM_VGIC_V[23]_ADDR_TYPE_XXX
- * @addr:  pointer to address value
- * @write: if true set the address in the VM address space, if false read the
- *          address
- *
- * Set or get the vgic base addresses for the distributor and the virtual CPU
- * interface in the VM physical address space.  These addresses are properties
- * of the emulated core/SoC and therefore user space initially knows this
- * information.
- */
-int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write)
-{
-	int r = 0;
-	struct vgic_dist *vgic = &kvm->arch.vgic;
-	int type_needed;
-	phys_addr_t *addr_ptr, block_size;
-	phys_addr_t alignment;
-
-	mutex_lock(&kvm->lock);
-	switch (type) {
-	case KVM_VGIC_V2_ADDR_TYPE_DIST:
-		type_needed = KVM_DEV_TYPE_ARM_VGIC_V2;
-		addr_ptr = &vgic->vgic_dist_base;
-		block_size = KVM_VGIC_V2_DIST_SIZE;
-		alignment = SZ_4K;
-		break;
-	case KVM_VGIC_V2_ADDR_TYPE_CPU:
-		type_needed = KVM_DEV_TYPE_ARM_VGIC_V2;
-		addr_ptr = &vgic->vgic_cpu_base;
-		block_size = KVM_VGIC_V2_CPU_SIZE;
-		alignment = SZ_4K;
-		break;
-#ifdef CONFIG_KVM_ARM_VGIC_V3
-	case KVM_VGIC_V3_ADDR_TYPE_DIST:
-		type_needed = KVM_DEV_TYPE_ARM_VGIC_V3;
-		addr_ptr = &vgic->vgic_dist_base;
-		block_size = KVM_VGIC_V3_DIST_SIZE;
-		alignment = SZ_64K;
-		break;
-	case KVM_VGIC_V3_ADDR_TYPE_REDIST:
-		type_needed = KVM_DEV_TYPE_ARM_VGIC_V3;
-		addr_ptr = &vgic->vgic_redist_base;
-		block_size = KVM_VGIC_V3_REDIST_SIZE;
-		alignment = SZ_64K;
-		break;
-#endif
-	default:
-		r = -ENODEV;
-		goto out;
-	}
-
-	if (vgic->vgic_model != type_needed) {
-		r = -ENODEV;
-		goto out;
-	}
-
-	if (write) {
-		if (!IS_ALIGNED(*addr, alignment))
-			r = -EINVAL;
-		else
-			r = vgic_ioaddr_assign(kvm, addr_ptr, *addr,
-					       block_size);
-	} else {
-		*addr = *addr_ptr;
-	}
-
-out:
-	mutex_unlock(&kvm->lock);
-	return r;
-}
-
-int vgic_set_common_attr(struct kvm_device *dev, struct kvm_device_attr *attr)
-{
-	int r;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_ADDR: {
-		u64 __user *uaddr = (u64 __user *)(long)attr->addr;
-		u64 addr;
-		unsigned long type = (unsigned long)attr->attr;
-
-		if (copy_from_user(&addr, uaddr, sizeof(addr)))
-			return -EFAULT;
-
-		r = kvm_vgic_addr(dev->kvm, type, &addr, true);
-		return (r == -ENODEV) ? -ENXIO : r;
-	}
-	case KVM_DEV_ARM_VGIC_GRP_NR_IRQS: {
-		u32 __user *uaddr = (u32 __user *)(long)attr->addr;
-		u32 val;
-		int ret = 0;
-
-		if (get_user(val, uaddr))
-			return -EFAULT;
-
-		/*
-		 * We require:
-		 * - at least 32 SPIs on top of the 16 SGIs and 16 PPIs
-		 * - at most 1024 interrupts
-		 * - a multiple of 32 interrupts
-		 */
-		if (val < (VGIC_NR_PRIVATE_IRQS + 32) ||
-		    val > VGIC_MAX_IRQS ||
-		    (val & 31))
-			return -EINVAL;
-
-		mutex_lock(&dev->kvm->lock);
-
-		if (vgic_ready(dev->kvm) || dev->kvm->arch.vgic.nr_irqs)
-			ret = -EBUSY;
-		else
-			dev->kvm->arch.vgic.nr_irqs = val;
-
-		mutex_unlock(&dev->kvm->lock);
-
-		return ret;
-	}
-	case KVM_DEV_ARM_VGIC_GRP_CTRL: {
-		switch (attr->attr) {
-		case KVM_DEV_ARM_VGIC_CTRL_INIT:
-			r = vgic_init(dev->kvm);
-			return r;
-		}
-		break;
-	}
-	}
-
-	return -ENXIO;
-}
-
-int vgic_get_common_attr(struct kvm_device *dev, struct kvm_device_attr *attr)
-{
-	int r = -ENXIO;
-
-	switch (attr->group) {
-	case KVM_DEV_ARM_VGIC_GRP_ADDR: {
-		u64 __user *uaddr = (u64 __user *)(long)attr->addr;
-		u64 addr;
-		unsigned long type = (unsigned long)attr->attr;
-
-		r = kvm_vgic_addr(dev->kvm, type, &addr, false);
-		if (r)
-			return (r == -ENODEV) ? -ENXIO : r;
-
-		if (copy_to_user(uaddr, &addr, sizeof(addr)))
-			return -EFAULT;
-		break;
-	}
-	case KVM_DEV_ARM_VGIC_GRP_NR_IRQS: {
-		u32 __user *uaddr = (u32 __user *)(long)attr->addr;
-
-		r = put_user(dev->kvm->arch.vgic.nr_irqs, uaddr);
-		break;
-	}
-
-	}
-
-	return r;
-}
-
-int vgic_has_attr_regs(const struct vgic_io_range *ranges, phys_addr_t offset)
-{
-	if (vgic_find_range(ranges, 4, offset))
-		return 0;
-	else
-		return -ENXIO;
-}
-
-static void vgic_init_maintenance_interrupt(void *info)
-{
-	enable_percpu_irq(vgic->maint_irq, 0);
-}
-
-static int vgic_cpu_notify(struct notifier_block *self,
-			   unsigned long action, void *cpu)
-{
-	switch (action) {
-	case CPU_STARTING:
-	case CPU_STARTING_FROZEN:
-		vgic_init_maintenance_interrupt(NULL);
-		break;
-	case CPU_DYING:
-	case CPU_DYING_FROZEN:
-		disable_percpu_irq(vgic->maint_irq);
-		break;
-	}
-
-	return NOTIFY_OK;
-}
-
-static struct notifier_block vgic_cpu_nb = {
-	.notifier_call = vgic_cpu_notify,
-};
-
-static int kvm_vgic_probe(void)
-{
-	const struct gic_kvm_info *gic_kvm_info;
-	int ret;
-
-	gic_kvm_info = gic_get_kvm_info();
-	if (!gic_kvm_info)
-		return -ENODEV;
-
-	switch (gic_kvm_info->type) {
-	case GIC_V2:
-		ret = vgic_v2_probe(gic_kvm_info, &vgic_ops, &vgic);
-		break;
-	case GIC_V3:
-		ret = vgic_v3_probe(gic_kvm_info, &vgic_ops, &vgic);
-		break;
-	default:
-		ret = -ENODEV;
-	}
-
-	return ret;
-}
-
-int kvm_vgic_hyp_init(void)
-{
-	int ret;
-
-	ret = kvm_vgic_probe();
-	if (ret) {
-		kvm_err("error: KVM vGIC probing failed\n");
-		return ret;
-	}
-
-	ret = request_percpu_irq(vgic->maint_irq, vgic_maintenance_handler,
-				 "vgic", kvm_get_running_vcpus());
-	if (ret) {
-		kvm_err("Cannot register interrupt %d\n", vgic->maint_irq);
-		return ret;
-	}
-
-	ret = __register_cpu_notifier(&vgic_cpu_nb);
-	if (ret) {
-		kvm_err("Cannot register vgic CPU notifier\n");
-		goto out_free_irq;
-	}
-
-	on_each_cpu(vgic_init_maintenance_interrupt, NULL, 1);
-
-	return 0;
-
-out_free_irq:
-	free_percpu_irq(vgic->maint_irq, kvm_get_running_vcpus());
-	return ret;
-}
-
-int kvm_irq_map_gsi(struct kvm *kvm,
-		    struct kvm_kernel_irq_routing_entry *entries,
-		    int gsi)
-{
-	return 0;
-}
-
-int kvm_irq_map_chip_pin(struct kvm *kvm, unsigned irqchip, unsigned pin)
-{
-	return pin;
-}
-
-int kvm_set_irq(struct kvm *kvm, int irq_source_id,
-		u32 irq, int level, bool line_status)
-{
-	unsigned int spi = irq + VGIC_NR_PRIVATE_IRQS;
-
-	trace_kvm_set_irq(irq, level, irq_source_id);
-
-	BUG_ON(!vgic_initialized(kvm));
-
-	return kvm_vgic_inject_irq(kvm, 0, spi, level);
-}
-
-/* MSI not implemented yet */
-int kvm_set_msi(struct kvm_kernel_irq_routing_entry *e,
-		struct kvm *kvm, int irq_source_id,
-		int level, bool line_status)
-{
-	return 0;
-}

virt/kvm/arm/vgic.h

@@ -1,140 +0,0 @@
-/*
- * Copyright (C) 2012-2014 ARM Ltd.
- * Author: Marc Zyngier <marc.zyngier@arm.com>
- *
- * Derived from virt/kvm/arm/vgic.c
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-
-#ifndef __KVM_VGIC_H__
-#define __KVM_VGIC_H__
-
-#include <kvm/iodev.h>
-
-#define VGIC_ADDR_UNDEF		(-1)
-#define IS_VGIC_ADDR_UNDEF(_x)  ((_x) == VGIC_ADDR_UNDEF)
-
-#define PRODUCT_ID_KVM		0x4b	/* ASCII code K */
-#define IMPLEMENTER_ARM		0x43b
-
-#define ACCESS_READ_VALUE	(1 << 0)
-#define ACCESS_READ_RAZ		(0 << 0)
-#define ACCESS_READ_MASK(x)	((x) & (1 << 0))
-#define ACCESS_WRITE_IGNORED	(0 << 1)
-#define ACCESS_WRITE_SETBIT	(1 << 1)
-#define ACCESS_WRITE_CLEARBIT	(2 << 1)
-#define ACCESS_WRITE_VALUE	(3 << 1)
-#define ACCESS_WRITE_MASK(x)	((x) & (3 << 1))
-
-#define VCPU_NOT_ALLOCATED	((u8)-1)
-
-unsigned long *vgic_bitmap_get_shared_map(struct vgic_bitmap *x);
-
-void vgic_update_state(struct kvm *kvm);
-int vgic_init_common_maps(struct kvm *kvm);
-
-u32 *vgic_bitmap_get_reg(struct vgic_bitmap *x, int cpuid, u32 offset);
-u32 *vgic_bytemap_get_reg(struct vgic_bytemap *x, int cpuid, u32 offset);
-
-void vgic_dist_irq_set_pending(struct kvm_vcpu *vcpu, int irq);
-void vgic_dist_irq_clear_pending(struct kvm_vcpu *vcpu, int irq);
-void vgic_cpu_irq_clear(struct kvm_vcpu *vcpu, int irq);
-void vgic_bitmap_set_irq_val(struct vgic_bitmap *x, int cpuid,
-			     int irq, int val);
-
-void vgic_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcr);
-void vgic_set_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcr);
-
-bool vgic_queue_irq(struct kvm_vcpu *vcpu, u8 sgi_source_id, int irq);
-void vgic_unqueue_irqs(struct kvm_vcpu *vcpu);
-
-struct kvm_exit_mmio {
-	phys_addr_t	phys_addr;
-	void		*data;
-	u32		len;
-	bool		is_write;
-	void		*private;
-};
-
-void vgic_reg_access(struct kvm_exit_mmio *mmio, u32 *reg,
-		     phys_addr_t offset, int mode);
-bool handle_mmio_raz_wi(struct kvm_vcpu *vcpu, struct kvm_exit_mmio *mmio,
-			phys_addr_t offset);
-
-static inline
-u32 mmio_data_read(struct kvm_exit_mmio *mmio, u32 mask)
-{
-	return le32_to_cpu(*((u32 *)mmio->data)) & mask;
-}
-
-static inline
-void mmio_data_write(struct kvm_exit_mmio *mmio, u32 mask, u32 value)
-{
-	*((u32 *)mmio->data) = cpu_to_le32(value) & mask;
-}
-
-struct vgic_io_range {
-	phys_addr_t base;
-	unsigned long len;
-	int bits_per_irq;
-	bool (*handle_mmio)(struct kvm_vcpu *vcpu, struct kvm_exit_mmio *mmio,
-			    phys_addr_t offset);
-};
-
-int vgic_register_kvm_io_dev(struct kvm *kvm, gpa_t base, int len,
-			     const struct vgic_io_range *ranges,
-			     int redist_id,
-			     struct vgic_io_device *iodev);
-
-static inline bool is_in_range(phys_addr_t addr, unsigned long len,
-			       phys_addr_t baseaddr, unsigned long size)
-{
-	return (addr >= baseaddr) && (addr + len <= baseaddr + size);
-}
-
-const
-struct vgic_io_range *vgic_find_range(const struct vgic_io_range *ranges,
-				      int len, gpa_t offset);
-
-bool vgic_handle_enable_reg(struct kvm *kvm, struct kvm_exit_mmio *mmio,
-			    phys_addr_t offset, int vcpu_id, int access);
-
-bool vgic_handle_set_pending_reg(struct kvm *kvm, struct kvm_exit_mmio *mmio,
-				 phys_addr_t offset, int vcpu_id);
-
-bool vgic_handle_clear_pending_reg(struct kvm *kvm, struct kvm_exit_mmio *mmio,
-				   phys_addr_t offset, int vcpu_id);
-
-bool vgic_handle_set_active_reg(struct kvm *kvm,
-				struct kvm_exit_mmio *mmio,
-				phys_addr_t offset, int vcpu_id);
-
-bool vgic_handle_clear_active_reg(struct kvm *kvm,
-				  struct kvm_exit_mmio *mmio,
-				  phys_addr_t offset, int vcpu_id);
-
-bool vgic_handle_cfg_reg(u32 *reg, struct kvm_exit_mmio *mmio,
-			 phys_addr_t offset);
-
-void vgic_kick_vcpus(struct kvm *kvm);
-
-int vgic_has_attr_regs(const struct vgic_io_range *ranges, phys_addr_t offset);
-int vgic_set_common_attr(struct kvm_device *dev, struct kvm_device_attr *attr);
-int vgic_get_common_attr(struct kvm_device *dev, struct kvm_device_attr *attr);
-
-int vgic_init(struct kvm *kvm);
-void vgic_v2_init_emulation(struct kvm *kvm);
-void vgic_v3_init_emulation(struct kvm *kvm);
-
-#endif

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

@@ -157,6 +157,9 @@ static int kvm_vgic_dist_init(struct kvm *kvm, unsigned int nr_spis)
 	struct kvm_vcpu *vcpu0 = kvm_get_vcpu(kvm, 0);
 	int i;
 
+	INIT_LIST_HEAD(&dist->lpi_list_head);
+	spin_lock_init(&dist->lpi_list_lock);
+
 	dist->spis = kcalloc(nr_spis, sizeof(struct vgic_irq), GFP_KERNEL);
 	if (!dist->spis)
 		return  -ENOMEM;
@@ -177,6 +180,7 @@ static int kvm_vgic_dist_init(struct kvm *kvm, unsigned int nr_spis)
 		spin_lock_init(&irq->irq_lock);
 		irq->vcpu = NULL;
 		irq->target_vcpu = vcpu0;
+		kref_init(&irq->refcount);
 		if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V2)
 			irq->targets = 0;
 		else
@@ -211,6 +215,7 @@ static void kvm_vgic_vcpu_init(struct kvm_vcpu *vcpu)
 		irq->vcpu = NULL;
 		irq->target_vcpu = vcpu;
 		irq->targets = 1U << vcpu->vcpu_id;
+		kref_init(&irq->refcount);
 		if (vgic_irq_is_sgi(i)) {
 			/* SGIs */
 			irq->enabled = 1;
@@ -253,6 +258,9 @@ int vgic_init(struct kvm *kvm)
 	if (ret)
 		goto out;
 
+	if (vgic_has_its(kvm))
+		dist->msis_require_devid = true;
+
 	kvm_for_each_vcpu(i, vcpu, kvm)
 		kvm_vgic_vcpu_init(vcpu);
 
@@ -271,7 +279,6 @@ static void kvm_vgic_dist_destroy(struct kvm *kvm)
 	dist->initialized = false;
 
 	kfree(dist->spis);
-	kfree(dist->redist_iodevs);
 	dist->nr_spis = 0;
 
 	mutex_unlock(&kvm->lock);

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

@@ -0,0 +1,1500 @@
+/*
+ * GICv3 ITS emulation
+ *
+ * Copyright (C) 2015,2016 ARM Ltd.
+ * Author: Andre Przywara <andre.przywara@arm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/cpu.h>
+#include <linux/kvm.h>
+#include <linux/kvm_host.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/uaccess.h>
+
+#include <linux/irqchip/arm-gic-v3.h>
+
+#include <asm/kvm_emulate.h>
+#include <asm/kvm_arm.h>
+#include <asm/kvm_mmu.h>
+
+#include "vgic.h"
+#include "vgic-mmio.h"
+
+/*
+ * Creates a new (reference to a) struct vgic_irq for a given LPI.
+ * If this LPI is already mapped on another ITS, we increase its refcount
+ * and return a pointer to the existing structure.
+ * If this is a "new" LPI, we allocate and initialize a new struct vgic_irq.
+ * This function returns a pointer to the _unlocked_ structure.
+ */
+static struct vgic_irq *vgic_add_lpi(struct kvm *kvm, u32 intid)
+{
+	struct vgic_dist *dist = &kvm->arch.vgic;
+	struct vgic_irq *irq = vgic_get_irq(kvm, NULL, intid), *oldirq;
+
+	/* In this case there is no put, since we keep the reference. */
+	if (irq)
+		return irq;
+
+	irq = kzalloc(sizeof(struct vgic_irq), GFP_KERNEL);
+	if (!irq)
+		return NULL;
+
+	INIT_LIST_HEAD(&irq->lpi_list);
+	INIT_LIST_HEAD(&irq->ap_list);
+	spin_lock_init(&irq->irq_lock);
+
+	irq->config = VGIC_CONFIG_EDGE;
+	kref_init(&irq->refcount);
+	irq->intid = intid;
+
+	spin_lock(&dist->lpi_list_lock);
+
+	/*
+	 * There could be a race with another vgic_add_lpi(), so we need to
+	 * check that we don't add a second list entry with the same LPI.
+	 */
+	list_for_each_entry(oldirq, &dist->lpi_list_head, lpi_list) {
+		if (oldirq->intid != intid)
+			continue;
+
+		/* Someone was faster with adding this LPI, lets use that. */
+		kfree(irq);
+		irq = oldirq;
+
+		/*
+		 * This increases the refcount, the caller is expected to
+		 * call vgic_put_irq() on the returned pointer once it's
+		 * finished with the IRQ.
+		 */
+		vgic_get_irq_kref(irq);
+
+		goto out_unlock;
+	}
+
+	list_add_tail(&irq->lpi_list, &dist->lpi_list_head);
+	dist->lpi_list_count++;
+
+out_unlock:
+	spin_unlock(&dist->lpi_list_lock);
+
+	return irq;
+}
+
+struct its_device {
+	struct list_head dev_list;
+
+	/* the head for the list of ITTEs */
+	struct list_head itt_head;
+	u32 device_id;
+};
+
+#define COLLECTION_NOT_MAPPED ((u32)~0)
+
+struct its_collection {
+	struct list_head coll_list;
+
+	u32 collection_id;
+	u32 target_addr;
+};
+
+#define its_is_collection_mapped(coll) ((coll) && \
+				((coll)->target_addr != COLLECTION_NOT_MAPPED))
+
+struct its_itte {
+	struct list_head itte_list;
+
+	struct vgic_irq *irq;
+	struct its_collection *collection;
+	u32 lpi;
+	u32 event_id;
+};
+
+/*
+ * Find and returns a device in the device table for an ITS.
+ * Must be called with the its_lock mutex held.
+ */
+static struct its_device *find_its_device(struct vgic_its *its, u32 device_id)
+{
+	struct its_device *device;
+
+	list_for_each_entry(device, &its->device_list, dev_list)
+		if (device_id == device->device_id)
+			return device;
+
+	return NULL;
+}
+
+/*
+ * Find and returns an interrupt translation table entry (ITTE) for a given
+ * Device ID/Event ID pair on an ITS.
+ * Must be called with the its_lock mutex held.
+ */
+static struct its_itte *find_itte(struct vgic_its *its, u32 device_id,
+				  u32 event_id)
+{
+	struct its_device *device;
+	struct its_itte *itte;
+
+	device = find_its_device(its, device_id);
+	if (device == NULL)
+		return NULL;
+
+	list_for_each_entry(itte, &device->itt_head, itte_list)
+		if (itte->event_id == event_id)
+			return itte;
+
+	return NULL;
+}
+
+/* To be used as an iterator this macro misses the enclosing parentheses */
+#define for_each_lpi_its(dev, itte, its) \
+	list_for_each_entry(dev, &(its)->device_list, dev_list) \
+		list_for_each_entry(itte, &(dev)->itt_head, itte_list)
+
+/*
+ * We only implement 48 bits of PA at the moment, although the ITS
+ * supports more. Let's be restrictive here.
+ */
+#define BASER_ADDRESS(x)	((x) & GENMASK_ULL(47, 16))
+#define CBASER_ADDRESS(x)	((x) & GENMASK_ULL(47, 12))
+#define PENDBASER_ADDRESS(x)	((x) & GENMASK_ULL(47, 16))
+#define PROPBASER_ADDRESS(x)	((x) & GENMASK_ULL(47, 12))
+
+#define GIC_LPI_OFFSET 8192
+
+/*
+ * Finds and returns a collection in the ITS collection table.
+ * Must be called with the its_lock mutex held.
+ */
+static struct its_collection *find_collection(struct vgic_its *its, int coll_id)
+{
+	struct its_collection *collection;
+
+	list_for_each_entry(collection, &its->collection_list, coll_list) {
+		if (coll_id == collection->collection_id)
+			return collection;
+	}
+
+	return NULL;
+}
+
+#define LPI_PROP_ENABLE_BIT(p)	((p) & LPI_PROP_ENABLED)
+#define LPI_PROP_PRIORITY(p)	((p) & 0xfc)
+
+/*
+ * Reads the configuration data for a given LPI from guest memory and
+ * updates the fields in struct vgic_irq.
+ * If filter_vcpu is not NULL, applies only if the IRQ is targeting this
+ * VCPU. Unconditionally applies if filter_vcpu is NULL.
+ */
+static int update_lpi_config(struct kvm *kvm, struct vgic_irq *irq,
+			     struct kvm_vcpu *filter_vcpu)
+{
+	u64 propbase = PROPBASER_ADDRESS(kvm->arch.vgic.propbaser);
+	u8 prop;
+	int ret;
+
+	ret = kvm_read_guest(kvm, propbase + irq->intid - GIC_LPI_OFFSET,
+			     &prop, 1);
+
+	if (ret)
+		return ret;
+
+	spin_lock(&irq->irq_lock);
+
+	if (!filter_vcpu || filter_vcpu == irq->target_vcpu) {
+		irq->priority = LPI_PROP_PRIORITY(prop);
+		irq->enabled = LPI_PROP_ENABLE_BIT(prop);
+
+		vgic_queue_irq_unlock(kvm, irq);
+	} else {
+		spin_unlock(&irq->irq_lock);
+	}
+
+	return 0;
+}
+
+/*
+ * Create a snapshot of the current LPI list, so that we can enumerate all
+ * LPIs without holding any lock.
+ * Returns the array length and puts the kmalloc'ed array into intid_ptr.
+ */
+static int vgic_copy_lpi_list(struct kvm *kvm, u32 **intid_ptr)
+{
+	struct vgic_dist *dist = &kvm->arch.vgic;
+	struct vgic_irq *irq;
+	u32 *intids;
+	int irq_count = dist->lpi_list_count, i = 0;
+
+	/*
+	 * We use the current value of the list length, which may change
+	 * after the kmalloc. We don't care, because the guest shouldn't
+	 * change anything while the command handling is still running,
+	 * and in the worst case we would miss a new IRQ, which one wouldn't
+	 * expect to be covered by this command anyway.
+	 */
+	intids = kmalloc_array(irq_count, sizeof(intids[0]), GFP_KERNEL);
+	if (!intids)
+		return -ENOMEM;
+
+	spin_lock(&dist->lpi_list_lock);
+	list_for_each_entry(irq, &dist->lpi_list_head, lpi_list) {
+		/* We don't need to "get" the IRQ, as we hold the list lock. */
+		intids[i] = irq->intid;
+		if (++i == irq_count)
+			break;
+	}
+	spin_unlock(&dist->lpi_list_lock);
+
+	*intid_ptr = intids;
+	return irq_count;
+}
+
+/*
+ * Promotes the ITS view of affinity of an ITTE (which redistributor this LPI
+ * is targeting) to the VGIC's view, which deals with target VCPUs.
+ * Needs to be called whenever either the collection for a LPIs has
+ * changed or the collection itself got retargeted.
+ */
+static void update_affinity_itte(struct kvm *kvm, struct its_itte *itte)
+{
+	struct kvm_vcpu *vcpu;
+
+	if (!its_is_collection_mapped(itte->collection))
+		return;
+
+	vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
+
+	spin_lock(&itte->irq->irq_lock);
+	itte->irq->target_vcpu = vcpu;
+	spin_unlock(&itte->irq->irq_lock);
+}
+
+/*
+ * Updates the target VCPU for every LPI targeting this collection.
+ * Must be called with the its_lock mutex held.
+ */
+static void update_affinity_collection(struct kvm *kvm, struct vgic_its *its,
+				       struct its_collection *coll)
+{
+	struct its_device *device;
+	struct its_itte *itte;
+
+	for_each_lpi_its(device, itte, its) {
+		if (!itte->collection || coll != itte->collection)
+			continue;
+
+		update_affinity_itte(kvm, itte);
+	}
+}
+
+static u32 max_lpis_propbaser(u64 propbaser)
+{
+	int nr_idbits = (propbaser & 0x1f) + 1;
+
+	return 1U << min(nr_idbits, INTERRUPT_ID_BITS_ITS);
+}
+
+/*
+ * Scan the whole LPI pending table and sync the pending bit in there
+ * with our own data structures. This relies on the LPI being
+ * mapped before.
+ */
+static int its_sync_lpi_pending_table(struct kvm_vcpu *vcpu)
+{
+	gpa_t pendbase = PENDBASER_ADDRESS(vcpu->arch.vgic_cpu.pendbaser);
+	struct vgic_irq *irq;
+	int last_byte_offset = -1;
+	int ret = 0;
+	u32 *intids;
+	int nr_irqs, i;
+
+	nr_irqs = vgic_copy_lpi_list(vcpu->kvm, &intids);
+	if (nr_irqs < 0)
+		return nr_irqs;
+
+	for (i = 0; i < nr_irqs; i++) {
+		int byte_offset, bit_nr;
+		u8 pendmask;
+
+		byte_offset = intids[i] / BITS_PER_BYTE;
+		bit_nr = intids[i] % BITS_PER_BYTE;
+
+		/*
+		 * For contiguously allocated LPIs chances are we just read
+		 * this very same byte in the last iteration. Reuse that.
+		 */
+		if (byte_offset != last_byte_offset) {
+			ret = kvm_read_guest(vcpu->kvm, pendbase + byte_offset,
+					     &pendmask, 1);
+			if (ret) {
+				kfree(intids);
+				return ret;
+			}
+			last_byte_offset = byte_offset;
+		}
+
+		irq = vgic_get_irq(vcpu->kvm, NULL, intids[i]);
+		spin_lock(&irq->irq_lock);
+		irq->pending = pendmask & (1U << bit_nr);
+		vgic_queue_irq_unlock(vcpu->kvm, irq);
+		vgic_put_irq(vcpu->kvm, irq);
+	}
+
+	kfree(intids);
+
+	return ret;
+}
+
+static unsigned long vgic_mmio_read_its_ctlr(struct kvm *vcpu,
+					     struct vgic_its *its,
+					     gpa_t addr, unsigned int len)
+{
+	u32 reg = 0;
+
+	mutex_lock(&its->cmd_lock);
+	if (its->creadr == its->cwriter)
+		reg |= GITS_CTLR_QUIESCENT;
+	if (its->enabled)
+		reg |= GITS_CTLR_ENABLE;
+	mutex_unlock(&its->cmd_lock);
+
+	return reg;
+}
+
+static void vgic_mmio_write_its_ctlr(struct kvm *kvm, struct vgic_its *its,
+				     gpa_t addr, unsigned int len,
+				     unsigned long val)
+{
+	its->enabled = !!(val & GITS_CTLR_ENABLE);
+}
+
+static unsigned long vgic_mmio_read_its_typer(struct kvm *kvm,
+					      struct vgic_its *its,
+					      gpa_t addr, unsigned int len)
+{
+	u64 reg = GITS_TYPER_PLPIS;
+
+	/*
+	 * We use linear CPU numbers for redistributor addressing,
+	 * so GITS_TYPER.PTA is 0.
+	 * Also we force all PROPBASER registers to be the same, so
+	 * CommonLPIAff is 0 as well.
+	 * To avoid memory waste in the guest, we keep the number of IDBits and
+	 * DevBits low - as least for the time being.
+	 */
+	reg |= 0x0f << GITS_TYPER_DEVBITS_SHIFT;
+	reg |= 0x0f << GITS_TYPER_IDBITS_SHIFT;
+
+	return extract_bytes(reg, addr & 7, len);
+}
+
+static unsigned long vgic_mmio_read_its_iidr(struct kvm *kvm,
+					     struct vgic_its *its,
+					     gpa_t addr, unsigned int len)
+{
+	return (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
+}
+
+static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
+					       struct vgic_its *its,
+					       gpa_t addr, unsigned int len)
+{
+	switch (addr & 0xffff) {
+	case GITS_PIDR0:
+		return 0x92;	/* part number, bits[7:0] */
+	case GITS_PIDR1:
+		return 0xb4;	/* part number, bits[11:8] */
+	case GITS_PIDR2:
+		return GIC_PIDR2_ARCH_GICv3 | 0x0b;
+	case GITS_PIDR4:
+		return 0x40;	/* This is a 64K software visible page */
+	/* The following are the ID registers for (any) GIC. */
+	case GITS_CIDR0:
+		return 0x0d;
+	case GITS_CIDR1:
+		return 0xf0;
+	case GITS_CIDR2:
+		return 0x05;
+	case GITS_CIDR3:
+		return 0xb1;
+	}
+
+	return 0;
+}
+
+/*
+ * Find the target VCPU and the LPI number for a given devid/eventid pair
+ * and make this IRQ pending, possibly injecting it.
+ * Must be called with the its_lock mutex held.
+ */
+static void vgic_its_trigger_msi(struct kvm *kvm, struct vgic_its *its,
+				 u32 devid, u32 eventid)
+{
+	struct its_itte *itte;
+
+	if (!its->enabled)
+		return;
+
+	itte = find_itte(its, devid, eventid);
+	/* Triggering an unmapped IRQ gets silently dropped. */
+	if (itte && its_is_collection_mapped(itte->collection)) {
+		struct kvm_vcpu *vcpu;
+
+		vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
+		if (vcpu && vcpu->arch.vgic_cpu.lpis_enabled) {
+			spin_lock(&itte->irq->irq_lock);
+			itte->irq->pending = true;
+			vgic_queue_irq_unlock(kvm, itte->irq);
+		}
+	}
+}
+
+/*
+ * Queries the KVM IO bus framework to get the ITS pointer from the given
+ * doorbell address.
+ * We then call vgic_its_trigger_msi() with the decoded data.
+ */
+int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
+{
+	u64 address;
+	struct kvm_io_device *kvm_io_dev;
+	struct vgic_io_device *iodev;
+
+	if (!vgic_has_its(kvm))
+		return -ENODEV;
+
+	if (!(msi->flags & KVM_MSI_VALID_DEVID))
+		return -EINVAL;
+
+	address = (u64)msi->address_hi << 32 | msi->address_lo;
+
+	kvm_io_dev = kvm_io_bus_get_dev(kvm, KVM_MMIO_BUS, address);
+	if (!kvm_io_dev)
+		return -ENODEV;
+
+	iodev = container_of(kvm_io_dev, struct vgic_io_device, dev);
+
+	mutex_lock(&iodev->its->its_lock);
+	vgic_its_trigger_msi(kvm, iodev->its, msi->devid, msi->data);
+	mutex_unlock(&iodev->its->its_lock);
+
+	return 0;
+}
+
+/* Requires the its_lock to be held. */
+static void its_free_itte(struct kvm *kvm, struct its_itte *itte)
+{
+	list_del(&itte->itte_list);
+
+	/* This put matches the get in vgic_add_lpi. */
+	vgic_put_irq(kvm, itte->irq);
+
+	kfree(itte);
+}
+
+static u64 its_cmd_mask_field(u64 *its_cmd, int word, int shift, int size)
+{
+	return (le64_to_cpu(its_cmd[word]) >> shift) & (BIT_ULL(size) - 1);
+}
+
+#define its_cmd_get_command(cmd)	its_cmd_mask_field(cmd, 0,  0,  8)
+#define its_cmd_get_deviceid(cmd)	its_cmd_mask_field(cmd, 0, 32, 32)
+#define its_cmd_get_id(cmd)		its_cmd_mask_field(cmd, 1,  0, 32)
+#define its_cmd_get_physical_id(cmd)	its_cmd_mask_field(cmd, 1, 32, 32)
+#define its_cmd_get_collection(cmd)	its_cmd_mask_field(cmd, 2,  0, 16)
+#define its_cmd_get_target_addr(cmd)	its_cmd_mask_field(cmd, 2, 16, 32)
+#define its_cmd_get_validbit(cmd)	its_cmd_mask_field(cmd, 2, 63,  1)
+
+/*
+ * The DISCARD command frees an Interrupt Translation Table Entry (ITTE).
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_discard(struct kvm *kvm, struct vgic_its *its,
+				       u64 *its_cmd)
+{
+	u32 device_id = its_cmd_get_deviceid(its_cmd);
+	u32 event_id = its_cmd_get_id(its_cmd);
+	struct its_itte *itte;
+
+
+	itte = find_itte(its, device_id, event_id);
+	if (itte && itte->collection) {
+		/*
+		 * Though the spec talks about removing the pending state, we
+		 * don't bother here since we clear the ITTE anyway and the
+		 * pending state is a property of the ITTE struct.
+		 */
+		its_free_itte(kvm, itte);
+		return 0;
+	}
+
+	return E_ITS_DISCARD_UNMAPPED_INTERRUPT;
+}
+
+/*
+ * The MOVI command moves an ITTE to a different collection.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_movi(struct kvm *kvm, struct vgic_its *its,
+				    u64 *its_cmd)
+{
+	u32 device_id = its_cmd_get_deviceid(its_cmd);
+	u32 event_id = its_cmd_get_id(its_cmd);
+	u32 coll_id = its_cmd_get_collection(its_cmd);
+	struct kvm_vcpu *vcpu;
+	struct its_itte *itte;
+	struct its_collection *collection;
+
+	itte = find_itte(its, device_id, event_id);
+	if (!itte)
+		return E_ITS_MOVI_UNMAPPED_INTERRUPT;
+
+	if (!its_is_collection_mapped(itte->collection))
+		return E_ITS_MOVI_UNMAPPED_COLLECTION;
+
+	collection = find_collection(its, coll_id);
+	if (!its_is_collection_mapped(collection))
+		return E_ITS_MOVI_UNMAPPED_COLLECTION;
+
+	itte->collection = collection;
+	vcpu = kvm_get_vcpu(kvm, collection->target_addr);
+
+	spin_lock(&itte->irq->irq_lock);
+	itte->irq->target_vcpu = vcpu;
+	spin_unlock(&itte->irq->irq_lock);
+
+	return 0;
+}
+
+/*
+ * Check whether an ID can be stored into the corresponding guest table.
+ * For a direct table this is pretty easy, but gets a bit nasty for
+ * indirect tables. We check whether the resulting guest physical address
+ * is actually valid (covered by a memslot and guest accessbible).
+ * For this we have to read the respective first level entry.
+ */
+static bool vgic_its_check_id(struct vgic_its *its, u64 baser, int id)
+{
+	int l1_tbl_size = GITS_BASER_NR_PAGES(baser) * SZ_64K;
+	int index;
+	u64 indirect_ptr;
+	gfn_t gfn;
+
+	if (!(baser & GITS_BASER_INDIRECT)) {
+		phys_addr_t addr;
+
+		if (id >= (l1_tbl_size / GITS_BASER_ENTRY_SIZE(baser)))
+			return false;
+
+		addr = BASER_ADDRESS(baser) + id * GITS_BASER_ENTRY_SIZE(baser);
+		gfn = addr >> PAGE_SHIFT;
+
+		return kvm_is_visible_gfn(its->dev->kvm, gfn);
+	}
+
+	/* calculate and check the index into the 1st level */
+	index = id / (SZ_64K / GITS_BASER_ENTRY_SIZE(baser));
+	if (index >= (l1_tbl_size / sizeof(u64)))
+		return false;
+
+	/* Each 1st level entry is represented by a 64-bit value. */
+	if (kvm_read_guest(its->dev->kvm,
+			   BASER_ADDRESS(baser) + index * sizeof(indirect_ptr),
+			   &indirect_ptr, sizeof(indirect_ptr)))
+		return false;
+
+	indirect_ptr = le64_to_cpu(indirect_ptr);
+
+	/* check the valid bit of the first level entry */
+	if (!(indirect_ptr & BIT_ULL(63)))
+		return false;
+
+	/*
+	 * Mask the guest physical address and calculate the frame number.
+	 * Any address beyond our supported 48 bits of PA will be caught
+	 * by the actual check in the final step.
+	 */
+	indirect_ptr &= GENMASK_ULL(51, 16);
+
+	/* Find the address of the actual entry */
+	index = id % (SZ_64K / GITS_BASER_ENTRY_SIZE(baser));
+	indirect_ptr += index * GITS_BASER_ENTRY_SIZE(baser);
+	gfn = indirect_ptr >> PAGE_SHIFT;
+
+	return kvm_is_visible_gfn(its->dev->kvm, gfn);
+}
+
+static int vgic_its_alloc_collection(struct vgic_its *its,
+				     struct its_collection **colp,
+				     u32 coll_id)
+{
+	struct its_collection *collection;
+
+	if (!vgic_its_check_id(its, its->baser_coll_table, coll_id))
+		return E_ITS_MAPC_COLLECTION_OOR;
+
+	collection = kzalloc(sizeof(*collection), GFP_KERNEL);
+
+	collection->collection_id = coll_id;
+	collection->target_addr = COLLECTION_NOT_MAPPED;
+
+	list_add_tail(&collection->coll_list, &its->collection_list);
+	*colp = collection;
+
+	return 0;
+}
+
+static void vgic_its_free_collection(struct vgic_its *its, u32 coll_id)
+{
+	struct its_collection *collection;
+	struct its_device *device;
+	struct its_itte *itte;
+
+	/*
+	 * Clearing the mapping for that collection ID removes the
+	 * entry from the list. If there wasn't any before, we can
+	 * go home early.
+	 */
+	collection = find_collection(its, coll_id);
+	if (!collection)
+		return;
+
+	for_each_lpi_its(device, itte, its)
+		if (itte->collection &&
+		    itte->collection->collection_id == coll_id)
+			itte->collection = NULL;
+
+	list_del(&collection->coll_list);
+	kfree(collection);
+}
+
+/*
+ * The MAPTI and MAPI commands map LPIs to ITTEs.
+ * Must be called with its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
+				    u64 *its_cmd)
+{
+	u32 device_id = its_cmd_get_deviceid(its_cmd);
+	u32 event_id = its_cmd_get_id(its_cmd);
+	u32 coll_id = its_cmd_get_collection(its_cmd);
+	struct its_itte *itte;
+	struct its_device *device;
+	struct its_collection *collection, *new_coll = NULL;
+	int lpi_nr;
+
+	device = find_its_device(its, device_id);
+	if (!device)
+		return E_ITS_MAPTI_UNMAPPED_DEVICE;
+
+	if (its_cmd_get_command(its_cmd) == GITS_CMD_MAPTI)
+		lpi_nr = its_cmd_get_physical_id(its_cmd);
+	else
+		lpi_nr = event_id;
+	if (lpi_nr < GIC_LPI_OFFSET ||
+	    lpi_nr >= max_lpis_propbaser(kvm->arch.vgic.propbaser))
+		return E_ITS_MAPTI_PHYSICALID_OOR;
+
+	collection = find_collection(its, coll_id);
+	if (!collection) {
+		int ret = vgic_its_alloc_collection(its, &collection, coll_id);
+		if (ret)
+			return ret;
+		new_coll = collection;
+	}
+
+	itte = find_itte(its, device_id, event_id);
+	if (!itte) {
+		itte = kzalloc(sizeof(struct its_itte), GFP_KERNEL);
+		if (!itte) {
+			if (new_coll)
+				vgic_its_free_collection(its, coll_id);
+			return -ENOMEM;
+		}
+
+		itte->event_id	= event_id;
+		list_add_tail(&itte->itte_list, &device->itt_head);
+	}
+
+	itte->collection = collection;
+	itte->lpi = lpi_nr;
+	itte->irq = vgic_add_lpi(kvm, lpi_nr);
+	update_affinity_itte(kvm, itte);
+
+	/*
+	 * We "cache" the configuration table entries in out struct vgic_irq's.
+	 * However we only have those structs for mapped IRQs, so we read in
+	 * the respective config data from memory here upon mapping the LPI.
+	 */
+	update_lpi_config(kvm, itte->irq, NULL);
+
+	return 0;
+}
+
+/* Requires the its_lock to be held. */
+static void vgic_its_unmap_device(struct kvm *kvm, struct its_device *device)
+{
+	struct its_itte *itte, *temp;
+
+	/*
+	 * The spec says that unmapping a device with still valid
+	 * ITTEs associated is UNPREDICTABLE. We remove all ITTEs,
+	 * since we cannot leave the memory unreferenced.
+	 */
+	list_for_each_entry_safe(itte, temp, &device->itt_head, itte_list)
+		its_free_itte(kvm, itte);
+
+	list_del(&device->dev_list);
+	kfree(device);
+}
+
+/*
+ * MAPD maps or unmaps a device ID to Interrupt Translation Tables (ITTs).
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_mapd(struct kvm *kvm, struct vgic_its *its,
+				    u64 *its_cmd)
+{
+	u32 device_id = its_cmd_get_deviceid(its_cmd);
+	bool valid = its_cmd_get_validbit(its_cmd);
+	struct its_device *device;
+
+	if (!vgic_its_check_id(its, its->baser_device_table, device_id))
+		return E_ITS_MAPD_DEVICE_OOR;
+
+	device = find_its_device(its, device_id);
+
+	/*
+	 * The spec says that calling MAPD on an already mapped device
+	 * invalidates all cached data for this device. We implement this
+	 * by removing the mapping and re-establishing it.
+	 */
+	if (device)
+		vgic_its_unmap_device(kvm, device);
+
+	/*
+	 * The spec does not say whether unmapping a not-mapped device
+	 * is an error, so we are done in any case.
+	 */
+	if (!valid)
+		return 0;
+
+	device = kzalloc(sizeof(struct its_device), GFP_KERNEL);
+	if (!device)
+		return -ENOMEM;
+
+	device->device_id = device_id;
+	INIT_LIST_HEAD(&device->itt_head);
+
+	list_add_tail(&device->dev_list, &its->device_list);
+
+	return 0;
+}
+
+/*
+ * The MAPC command maps collection IDs to redistributors.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_mapc(struct kvm *kvm, struct vgic_its *its,
+				    u64 *its_cmd)
+{
+	u16 coll_id;
+	u32 target_addr;
+	struct its_collection *collection;
+	bool valid;
+
+	valid = its_cmd_get_validbit(its_cmd);
+	coll_id = its_cmd_get_collection(its_cmd);
+	target_addr = its_cmd_get_target_addr(its_cmd);
+
+	if (target_addr >= atomic_read(&kvm->online_vcpus))
+		return E_ITS_MAPC_PROCNUM_OOR;
+
+	if (!valid) {
+		vgic_its_free_collection(its, coll_id);
+	} else {
+		collection = find_collection(its, coll_id);
+
+		if (!collection) {
+			int ret;
+
+			ret = vgic_its_alloc_collection(its, &collection,
+							coll_id);
+			if (ret)
+				return ret;
+			collection->target_addr = target_addr;
+		} else {
+			collection->target_addr = target_addr;
+			update_affinity_collection(kvm, its, collection);
+		}
+	}
+
+	return 0;
+}
+
+/*
+ * The CLEAR command removes the pending state for a particular LPI.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_clear(struct kvm *kvm, struct vgic_its *its,
+				     u64 *its_cmd)
+{
+	u32 device_id = its_cmd_get_deviceid(its_cmd);
+	u32 event_id = its_cmd_get_id(its_cmd);
+	struct its_itte *itte;
+
+
+	itte = find_itte(its, device_id, event_id);
+	if (!itte)
+		return E_ITS_CLEAR_UNMAPPED_INTERRUPT;
+
+	itte->irq->pending = false;
+
+	return 0;
+}
+
+/*
+ * The INV command syncs the configuration bits from the memory table.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_inv(struct kvm *kvm, struct vgic_its *its,
+				   u64 *its_cmd)
+{
+	u32 device_id = its_cmd_get_deviceid(its_cmd);
+	u32 event_id = its_cmd_get_id(its_cmd);
+	struct its_itte *itte;
+
+
+	itte = find_itte(its, device_id, event_id);
+	if (!itte)
+		return E_ITS_INV_UNMAPPED_INTERRUPT;
+
+	return update_lpi_config(kvm, itte->irq, NULL);
+}
+
+/*
+ * The INVALL command requests flushing of all IRQ data in this collection.
+ * Find the VCPU mapped to that collection, then iterate over the VM's list
+ * of mapped LPIs and update the configuration for each IRQ which targets
+ * the specified vcpu. The configuration will be read from the in-memory
+ * configuration table.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_invall(struct kvm *kvm, struct vgic_its *its,
+				      u64 *its_cmd)
+{
+	u32 coll_id = its_cmd_get_collection(its_cmd);
+	struct its_collection *collection;
+	struct kvm_vcpu *vcpu;
+	struct vgic_irq *irq;
+	u32 *intids;
+	int irq_count, i;
+
+	collection = find_collection(its, coll_id);
+	if (!its_is_collection_mapped(collection))
+		return E_ITS_INVALL_UNMAPPED_COLLECTION;
+
+	vcpu = kvm_get_vcpu(kvm, collection->target_addr);
+
+	irq_count = vgic_copy_lpi_list(kvm, &intids);
+	if (irq_count < 0)
+		return irq_count;
+
+	for (i = 0; i < irq_count; i++) {
+		irq = vgic_get_irq(kvm, NULL, intids[i]);
+		if (!irq)
+			continue;
+		update_lpi_config(kvm, irq, vcpu);
+		vgic_put_irq(kvm, irq);
+	}
+
+	kfree(intids);
+
+	return 0;
+}
+
+/*
+ * The MOVALL command moves the pending state of all IRQs targeting one
+ * redistributor to another. We don't hold the pending state in the VCPUs,
+ * but in the IRQs instead, so there is really not much to do for us here.
+ * However the spec says that no IRQ must target the old redistributor
+ * afterwards, so we make sure that no LPI is using the associated target_vcpu.
+ * This command affects all LPIs in the system that target that redistributor.
+ */
+static int vgic_its_cmd_handle_movall(struct kvm *kvm, struct vgic_its *its,
+				      u64 *its_cmd)
+{
+	struct vgic_dist *dist = &kvm->arch.vgic;
+	u32 target1_addr = its_cmd_get_target_addr(its_cmd);
+	u32 target2_addr = its_cmd_mask_field(its_cmd, 3, 16, 32);
+	struct kvm_vcpu *vcpu1, *vcpu2;
+	struct vgic_irq *irq;
+
+	if (target1_addr >= atomic_read(&kvm->online_vcpus) ||
+	    target2_addr >= atomic_read(&kvm->online_vcpus))
+		return E_ITS_MOVALL_PROCNUM_OOR;
+
+	if (target1_addr == target2_addr)
+		return 0;
+
+	vcpu1 = kvm_get_vcpu(kvm, target1_addr);
+	vcpu2 = kvm_get_vcpu(kvm, target2_addr);
+
+	spin_lock(&dist->lpi_list_lock);
+
+	list_for_each_entry(irq, &dist->lpi_list_head, lpi_list) {
+		spin_lock(&irq->irq_lock);
+
+		if (irq->target_vcpu == vcpu1)
+			irq->target_vcpu = vcpu2;
+
+		spin_unlock(&irq->irq_lock);
+	}
+
+	spin_unlock(&dist->lpi_list_lock);
+
+	return 0;
+}
+
+/*
+ * The INT command injects the LPI associated with that DevID/EvID pair.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_int(struct kvm *kvm, struct vgic_its *its,
+				   u64 *its_cmd)
+{
+	u32 msi_data = its_cmd_get_id(its_cmd);
+	u64 msi_devid = its_cmd_get_deviceid(its_cmd);
+
+	vgic_its_trigger_msi(kvm, its, msi_devid, msi_data);
+
+	return 0;
+}
+
+/*
+ * This function is called with the its_cmd lock held, but the ITS data
+ * structure lock dropped.
+ */
+static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
+				   u64 *its_cmd)
+{
+	int ret = -ENODEV;
+
+	mutex_lock(&its->its_lock);
+	switch (its_cmd_get_command(its_cmd)) {
+	case GITS_CMD_MAPD:
+		ret = vgic_its_cmd_handle_mapd(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_MAPC:
+		ret = vgic_its_cmd_handle_mapc(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_MAPI:
+		ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_MAPTI:
+		ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_MOVI:
+		ret = vgic_its_cmd_handle_movi(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_DISCARD:
+		ret = vgic_its_cmd_handle_discard(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_CLEAR:
+		ret = vgic_its_cmd_handle_clear(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_MOVALL:
+		ret = vgic_its_cmd_handle_movall(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_INT:
+		ret = vgic_its_cmd_handle_int(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_INV:
+		ret = vgic_its_cmd_handle_inv(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_INVALL:
+		ret = vgic_its_cmd_handle_invall(kvm, its, its_cmd);
+		break;
+	case GITS_CMD_SYNC:
+		/* we ignore this command: we are in sync all of the time */
+		ret = 0;
+		break;
+	}
+	mutex_unlock(&its->its_lock);
+
+	return ret;
+}
+
+static u64 vgic_sanitise_its_baser(u64 reg)
+{
+	reg = vgic_sanitise_field(reg, GITS_BASER_SHAREABILITY_MASK,
+				  GITS_BASER_SHAREABILITY_SHIFT,
+				  vgic_sanitise_shareability);
+	reg = vgic_sanitise_field(reg, GITS_BASER_INNER_CACHEABILITY_MASK,
+				  GITS_BASER_INNER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_inner_cacheability);
+	reg = vgic_sanitise_field(reg, GITS_BASER_OUTER_CACHEABILITY_MASK,
+				  GITS_BASER_OUTER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_outer_cacheability);
+
+	/* Bits 15:12 contain bits 51:48 of the PA, which we don't support. */
+	reg &= ~GENMASK_ULL(15, 12);
+
+	/* We support only one (ITS) page size: 64K */
+	reg = (reg & ~GITS_BASER_PAGE_SIZE_MASK) | GITS_BASER_PAGE_SIZE_64K;
+
+	return reg;
+}
+
+static u64 vgic_sanitise_its_cbaser(u64 reg)
+{
+	reg = vgic_sanitise_field(reg, GITS_CBASER_SHAREABILITY_MASK,
+				  GITS_CBASER_SHAREABILITY_SHIFT,
+				  vgic_sanitise_shareability);
+	reg = vgic_sanitise_field(reg, GITS_CBASER_INNER_CACHEABILITY_MASK,
+				  GITS_CBASER_INNER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_inner_cacheability);
+	reg = vgic_sanitise_field(reg, GITS_CBASER_OUTER_CACHEABILITY_MASK,
+				  GITS_CBASER_OUTER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_outer_cacheability);
+
+	/*
+	 * Sanitise the physical address to be 64k aligned.
+	 * Also limit the physical addresses to 48 bits.
+	 */
+	reg &= ~(GENMASK_ULL(51, 48) | GENMASK_ULL(15, 12));
+
+	return reg;
+}
+
+static unsigned long vgic_mmio_read_its_cbaser(struct kvm *kvm,
+					       struct vgic_its *its,
+					       gpa_t addr, unsigned int len)
+{
+	return extract_bytes(its->cbaser, addr & 7, len);
+}
+
+static void vgic_mmio_write_its_cbaser(struct kvm *kvm, struct vgic_its *its,
+				       gpa_t addr, unsigned int len,
+				       unsigned long val)
+{
+	/* When GITS_CTLR.Enable is 1, this register is RO. */
+	if (its->enabled)
+		return;
+
+	mutex_lock(&its->cmd_lock);
+	its->cbaser = update_64bit_reg(its->cbaser, addr & 7, len, val);
+	its->cbaser = vgic_sanitise_its_cbaser(its->cbaser);
+	its->creadr = 0;
+	/*
+	 * CWRITER is architecturally UNKNOWN on reset, but we need to reset
+	 * it to CREADR to make sure we start with an empty command buffer.
+	 */
+	its->cwriter = its->creadr;
+	mutex_unlock(&its->cmd_lock);
+}
+
+#define ITS_CMD_BUFFER_SIZE(baser)	((((baser) & 0xff) + 1) << 12)
+#define ITS_CMD_SIZE			32
+#define ITS_CMD_OFFSET(reg)		((reg) & GENMASK(19, 5))
+
+/*
+ * By writing to CWRITER the guest announces new commands to be processed.
+ * To avoid any races in the first place, we take the its_cmd lock, which
+ * protects our ring buffer variables, so that there is only one user
+ * per ITS handling commands at a given time.
+ */
+static void vgic_mmio_write_its_cwriter(struct kvm *kvm, struct vgic_its *its,
+					gpa_t addr, unsigned int len,
+					unsigned long val)
+{
+	gpa_t cbaser;
+	u64 cmd_buf[4];
+	u32 reg;
+
+	if (!its)
+		return;
+
+	mutex_lock(&its->cmd_lock);
+
+	reg = update_64bit_reg(its->cwriter, addr & 7, len, val);
+	reg = ITS_CMD_OFFSET(reg);
+	if (reg >= ITS_CMD_BUFFER_SIZE(its->cbaser)) {
+		mutex_unlock(&its->cmd_lock);
+		return;
+	}
+
+	its->cwriter = reg;
+	cbaser = CBASER_ADDRESS(its->cbaser);
+
+	while (its->cwriter != its->creadr) {
+		int ret = kvm_read_guest(kvm, cbaser + its->creadr,
+					 cmd_buf, ITS_CMD_SIZE);
+		/*
+		 * If kvm_read_guest() fails, this could be due to the guest
+		 * programming a bogus value in CBASER or something else going
+		 * wrong from which we cannot easily recover.
+		 * According to section 6.3.2 in the GICv3 spec we can just
+		 * ignore that command then.
+		 */
+		if (!ret)
+			vgic_its_handle_command(kvm, its, cmd_buf);
+
+		its->creadr += ITS_CMD_SIZE;
+		if (its->creadr == ITS_CMD_BUFFER_SIZE(its->cbaser))
+			its->creadr = 0;
+	}
+
+	mutex_unlock(&its->cmd_lock);
+}
+
+static unsigned long vgic_mmio_read_its_cwriter(struct kvm *kvm,
+						struct vgic_its *its,
+						gpa_t addr, unsigned int len)
+{
+	return extract_bytes(its->cwriter, addr & 0x7, len);
+}
+
+static unsigned long vgic_mmio_read_its_creadr(struct kvm *kvm,
+					       struct vgic_its *its,
+					       gpa_t addr, unsigned int len)
+{
+	return extract_bytes(its->creadr, addr & 0x7, len);
+}
+
+#define BASER_INDEX(addr) (((addr) / sizeof(u64)) & 0x7)
+static unsigned long vgic_mmio_read_its_baser(struct kvm *kvm,
+					      struct vgic_its *its,
+					      gpa_t addr, unsigned int len)
+{
+	u64 reg;
+
+	switch (BASER_INDEX(addr)) {
+	case 0:
+		reg = its->baser_device_table;
+		break;
+	case 1:
+		reg = its->baser_coll_table;
+		break;
+	default:
+		reg = 0;
+		break;
+	}
+
+	return extract_bytes(reg, addr & 7, len);
+}
+
+#define GITS_BASER_RO_MASK	(GENMASK_ULL(52, 48) | GENMASK_ULL(58, 56))
+static void vgic_mmio_write_its_baser(struct kvm *kvm,
+				      struct vgic_its *its,
+				      gpa_t addr, unsigned int len,
+				      unsigned long val)
+{
+	u64 entry_size, device_type;
+	u64 reg, *regptr, clearbits = 0;
+
+	/* When GITS_CTLR.Enable is 1, we ignore write accesses. */
+	if (its->enabled)
+		return;
+
+	switch (BASER_INDEX(addr)) {
+	case 0:
+		regptr = &its->baser_device_table;
+		entry_size = 8;
+		device_type = GITS_BASER_TYPE_DEVICE;
+		break;
+	case 1:
+		regptr = &its->baser_coll_table;
+		entry_size = 8;
+		device_type = GITS_BASER_TYPE_COLLECTION;
+		clearbits = GITS_BASER_INDIRECT;
+		break;
+	default:
+		return;
+	}
+
+	reg = update_64bit_reg(*regptr, addr & 7, len, val);
+	reg &= ~GITS_BASER_RO_MASK;
+	reg &= ~clearbits;
+
+	reg |= (entry_size - 1) << GITS_BASER_ENTRY_SIZE_SHIFT;
+	reg |= device_type << GITS_BASER_TYPE_SHIFT;
+	reg = vgic_sanitise_its_baser(reg);
+
+	*regptr = reg;
+}
+
+#define REGISTER_ITS_DESC(off, rd, wr, length, acc)		\
+{								\
+	.reg_offset = off,					\
+	.len = length,						\
+	.access_flags = acc,					\
+	.its_read = rd,						\
+	.its_write = wr,					\
+}
+
+static void its_mmio_write_wi(struct kvm *kvm, struct vgic_its *its,
+			      gpa_t addr, unsigned int len, unsigned long val)
+{
+	/* Ignore */
+}
+
+static struct vgic_register_region its_registers[] = {
+	REGISTER_ITS_DESC(GITS_CTLR,
+		vgic_mmio_read_its_ctlr, vgic_mmio_write_its_ctlr, 4,
+		VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_IIDR,
+		vgic_mmio_read_its_iidr, its_mmio_write_wi, 4,
+		VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_TYPER,
+		vgic_mmio_read_its_typer, its_mmio_write_wi, 8,
+		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_CBASER,
+		vgic_mmio_read_its_cbaser, vgic_mmio_write_its_cbaser, 8,
+		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_CWRITER,
+		vgic_mmio_read_its_cwriter, vgic_mmio_write_its_cwriter, 8,
+		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_CREADR,
+		vgic_mmio_read_its_creadr, its_mmio_write_wi, 8,
+		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_BASER,
+		vgic_mmio_read_its_baser, vgic_mmio_write_its_baser, 0x40,
+		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+	REGISTER_ITS_DESC(GITS_IDREGS_BASE,
+		vgic_mmio_read_its_idregs, its_mmio_write_wi, 0x30,
+		VGIC_ACCESS_32bit),
+};
+
+/* This is called on setting the LPI enable bit in the redistributor. */
+void vgic_enable_lpis(struct kvm_vcpu *vcpu)
+{
+	if (!(vcpu->arch.vgic_cpu.pendbaser & GICR_PENDBASER_PTZ))
+		its_sync_lpi_pending_table(vcpu);
+}
+
+static int vgic_its_init_its(struct kvm *kvm, struct vgic_its *its)
+{
+	struct vgic_io_device *iodev = &its->iodev;
+	int ret;
+
+	if (its->initialized)
+		return 0;
+
+	if (IS_VGIC_ADDR_UNDEF(its->vgic_its_base))
+		return -ENXIO;
+
+	iodev->regions = its_registers;
+	iodev->nr_regions = ARRAY_SIZE(its_registers);
+	kvm_iodevice_init(&iodev->dev, &kvm_io_gic_ops);
+
+	iodev->base_addr = its->vgic_its_base;
+	iodev->iodev_type = IODEV_ITS;
+	iodev->its = its;
+	mutex_lock(&kvm->slots_lock);
+	ret = kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, iodev->base_addr,
+				      KVM_VGIC_V3_ITS_SIZE, &iodev->dev);
+	mutex_unlock(&kvm->slots_lock);
+
+	if (!ret)
+		its->initialized = true;
+
+	return ret;
+}
+
+#define INITIAL_BASER_VALUE						  \
+	(GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWb)		| \
+	 GIC_BASER_CACHEABILITY(GITS_BASER, OUTER, SameAsInner)		| \
+	 GIC_BASER_SHAREABILITY(GITS_BASER, InnerShareable)		| \
+	 ((8ULL - 1) << GITS_BASER_ENTRY_SIZE_SHIFT)			| \
+	 GITS_BASER_PAGE_SIZE_64K)
+
+#define INITIAL_PROPBASER_VALUE						  \
+	(GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWb)		| \
+	 GIC_BASER_CACHEABILITY(GICR_PROPBASER, OUTER, SameAsInner)	| \
+	 GIC_BASER_SHAREABILITY(GICR_PROPBASER, InnerShareable))
+
+static int vgic_its_create(struct kvm_device *dev, u32 type)
+{
+	struct vgic_its *its;
+
+	if (type != KVM_DEV_TYPE_ARM_VGIC_ITS)
+		return -ENODEV;
+
+	its = kzalloc(sizeof(struct vgic_its), GFP_KERNEL);
+	if (!its)
+		return -ENOMEM;
+
+	mutex_init(&its->its_lock);
+	mutex_init(&its->cmd_lock);
+
+	its->vgic_its_base = VGIC_ADDR_UNDEF;
+
+	INIT_LIST_HEAD(&its->device_list);
+	INIT_LIST_HEAD(&its->collection_list);
+
+	dev->kvm->arch.vgic.has_its = true;
+	its->initialized = false;
+	its->enabled = false;
+	its->dev = dev;
+
+	its->baser_device_table = INITIAL_BASER_VALUE			|
+		((u64)GITS_BASER_TYPE_DEVICE << GITS_BASER_TYPE_SHIFT);
+	its->baser_coll_table = INITIAL_BASER_VALUE |
+		((u64)GITS_BASER_TYPE_COLLECTION << GITS_BASER_TYPE_SHIFT);
+	dev->kvm->arch.vgic.propbaser = INITIAL_PROPBASER_VALUE;
+
+	dev->private = its;
+
+	return 0;
+}
+
+static void vgic_its_destroy(struct kvm_device *kvm_dev)
+{
+	struct kvm *kvm = kvm_dev->kvm;
+	struct vgic_its *its = kvm_dev->private;
+	struct its_device *dev;
+	struct its_itte *itte;
+	struct list_head *dev_cur, *dev_temp;
+	struct list_head *cur, *temp;
+
+	/*
+	 * We may end up here without the lists ever having been initialized.
+	 * Check this and bail out early to avoid dereferencing a NULL pointer.
+	 */
+	if (!its->device_list.next)
+		return;
+
+	mutex_lock(&its->its_lock);
+	list_for_each_safe(dev_cur, dev_temp, &its->device_list) {
+		dev = container_of(dev_cur, struct its_device, dev_list);
+		list_for_each_safe(cur, temp, &dev->itt_head) {
+			itte = (container_of(cur, struct its_itte, itte_list));
+			its_free_itte(kvm, itte);
+		}
+		list_del(dev_cur);
+		kfree(dev);
+	}
+
+	list_for_each_safe(cur, temp, &its->collection_list) {
+		list_del(cur);
+		kfree(container_of(cur, struct its_collection, coll_list));
+	}
+	mutex_unlock(&its->its_lock);
+
+	kfree(its);
+}
+
+static int vgic_its_has_attr(struct kvm_device *dev,
+			     struct kvm_device_attr *attr)
+{
+	switch (attr->group) {
+	case KVM_DEV_ARM_VGIC_GRP_ADDR:
+		switch (attr->attr) {
+		case KVM_VGIC_ITS_ADDR_TYPE:
+			return 0;
+		}
+		break;
+	case KVM_DEV_ARM_VGIC_GRP_CTRL:
+		switch (attr->attr) {
+		case KVM_DEV_ARM_VGIC_CTRL_INIT:
+			return 0;
+		}
+		break;
+	}
+	return -ENXIO;
+}
+
+static int vgic_its_set_attr(struct kvm_device *dev,
+			     struct kvm_device_attr *attr)
+{
+	struct vgic_its *its = dev->private;
+	int ret;
+
+	switch (attr->group) {
+	case KVM_DEV_ARM_VGIC_GRP_ADDR: {
+		u64 __user *uaddr = (u64 __user *)(long)attr->addr;
+		unsigned long type = (unsigned long)attr->attr;
+		u64 addr;
+
+		if (type != KVM_VGIC_ITS_ADDR_TYPE)
+			return -ENODEV;
+
+		if (its->initialized)
+			return -EBUSY;
+
+		if (copy_from_user(&addr, uaddr, sizeof(addr)))
+			return -EFAULT;
+
+		ret = vgic_check_ioaddr(dev->kvm, &its->vgic_its_base,
+					addr, SZ_64K);
+		if (ret)
+			return ret;
+
+		its->vgic_its_base = addr;
+
+		return 0;
+	}
+	case KVM_DEV_ARM_VGIC_GRP_CTRL:
+		switch (attr->attr) {
+		case KVM_DEV_ARM_VGIC_CTRL_INIT:
+			return vgic_its_init_its(dev->kvm, its);
+		}
+		break;
+	}
+	return -ENXIO;
+}
+
+static int vgic_its_get_attr(struct kvm_device *dev,
+			     struct kvm_device_attr *attr)
+{
+	switch (attr->group) {
+	case KVM_DEV_ARM_VGIC_GRP_ADDR: {
+		struct vgic_its *its = dev->private;
+		u64 addr = its->vgic_its_base;
+		u64 __user *uaddr = (u64 __user *)(long)attr->addr;
+		unsigned long type = (unsigned long)attr->attr;
+
+		if (type != KVM_VGIC_ITS_ADDR_TYPE)
+			return -ENODEV;
+
+		if (copy_to_user(uaddr, &addr, sizeof(addr)))
+			return -EFAULT;
+		break;
+	default:
+		return -ENXIO;
+	}
+	}
+
+	return 0;
+}
+
+static struct kvm_device_ops kvm_arm_vgic_its_ops = {
+	.name = "kvm-arm-vgic-its",
+	.create = vgic_its_create,
+	.destroy = vgic_its_destroy,
+	.set_attr = vgic_its_set_attr,
+	.get_attr = vgic_its_get_attr,
+	.has_attr = vgic_its_has_attr,
+};
+
+int kvm_vgic_register_its_device(void)
+{
+	return kvm_register_device_ops(&kvm_arm_vgic_its_ops,
+				       KVM_DEV_TYPE_ARM_VGIC_ITS);
+}

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

@@ -21,8 +21,8 @@
 
 /* common helpers */
 
-static int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
-			     phys_addr_t addr, phys_addr_t alignment)
+int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
+		      phys_addr_t addr, phys_addr_t alignment)
 {
 	if (addr & ~KVM_PHYS_MASK)
 		return -E2BIG;
@@ -210,20 +210,27 @@ static void vgic_destroy(struct kvm_device *dev)
 	kfree(dev);
 }
 
-void kvm_register_vgic_device(unsigned long type)
+int kvm_register_vgic_device(unsigned long type)
 {
+	int ret = -ENODEV;
+
 	switch (type) {
 	case KVM_DEV_TYPE_ARM_VGIC_V2:
-		kvm_register_device_ops(&kvm_arm_vgic_v2_ops,
-					KVM_DEV_TYPE_ARM_VGIC_V2);
+		ret = kvm_register_device_ops(&kvm_arm_vgic_v2_ops,
+					      KVM_DEV_TYPE_ARM_VGIC_V2);
 		break;
 #ifdef CONFIG_KVM_ARM_VGIC_V3
 	case KVM_DEV_TYPE_ARM_VGIC_V3:
-		kvm_register_device_ops(&kvm_arm_vgic_v3_ops,
-					KVM_DEV_TYPE_ARM_VGIC_V3);
+		ret = kvm_register_device_ops(&kvm_arm_vgic_v3_ops,
+					      KVM_DEV_TYPE_ARM_VGIC_V3);
+		if (ret)
+			break;
+		ret = kvm_vgic_register_its_device();
 		break;
 #endif
 	}
+
+	return ret;
 }
 
 /** vgic_attr_regs_access: allows user space to read/write VGIC registers
@@ -428,4 +435,3 @@ struct kvm_device_ops kvm_arm_vgic_v3_ops = {
 };
 
 #endif /* CONFIG_KVM_ARM_VGIC_V3 */
-

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

@@ -102,6 +102,7 @@ static void vgic_mmio_write_sgir(struct kvm_vcpu *source_vcpu,
 		irq->source |= 1U << source_vcpu->vcpu_id;
 
 		vgic_queue_irq_unlock(source_vcpu->kvm, irq);
+		vgic_put_irq(source_vcpu->kvm, irq);
 	}
 }
 
@@ -116,6 +117,8 @@ static unsigned long vgic_mmio_read_target(struct kvm_vcpu *vcpu,
 		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
 
 		val |= (u64)irq->targets << (i * 8);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 
 	return val;
@@ -143,6 +146,7 @@ static void vgic_mmio_write_target(struct kvm_vcpu *vcpu,
 		irq->target_vcpu = kvm_get_vcpu(vcpu->kvm, target);
 
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -157,6 +161,8 @@ static unsigned long vgic_mmio_read_sgipend(struct kvm_vcpu *vcpu,
 		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
 
 		val |= (u64)irq->source << (i * 8);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 	return val;
 }
@@ -178,6 +184,7 @@ static void vgic_mmio_write_sgipendc(struct kvm_vcpu *vcpu,
 			irq->pending = false;
 
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -201,6 +208,7 @@ static void vgic_mmio_write_sgipends(struct kvm_vcpu *vcpu,
 		} else {
 			spin_unlock(&irq->irq_lock);
 		}
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -429,6 +437,7 @@ int vgic_v2_cpuif_uaccess(struct kvm_vcpu *vcpu, bool is_write,
 	struct vgic_io_device dev = {
 		.regions = vgic_v2_cpu_registers,
 		.nr_regions = ARRAY_SIZE(vgic_v2_cpu_registers),
+		.iodev_type = IODEV_CPUIF,
 	};
 
 	return vgic_uaccess(vcpu, &dev, is_write, offset, val);
@@ -440,6 +449,7 @@ int vgic_v2_dist_uaccess(struct kvm_vcpu *vcpu, bool is_write,
 	struct vgic_io_device dev = {
 		.regions = vgic_v2_dist_registers,
 		.nr_regions = ARRAY_SIZE(vgic_v2_dist_registers),
+		.iodev_type = IODEV_DIST,
 	};
 
 	return vgic_uaccess(vcpu, &dev, is_write, offset, val);

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

@@ -23,12 +23,35 @@
 #include "vgic-mmio.h"
 
 /* extract @num bytes at @offset bytes offset in data */
-static unsigned long extract_bytes(unsigned long data, unsigned int offset,
-				   unsigned int num)
+unsigned long extract_bytes(unsigned long data, unsigned int offset,
+			    unsigned int num)
 {
 	return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
 }
 
+/* allows updates of any half of a 64-bit register (or the whole thing) */
+u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
+		     unsigned long val)
+{
+	int lower = (offset & 4) * 8;
+	int upper = lower + 8 * len - 1;
+
+	reg &= ~GENMASK_ULL(upper, lower);
+	val &= GENMASK_ULL(len * 8 - 1, 0);
+
+	return reg | ((u64)val << lower);
+}
+
+bool vgic_has_its(struct kvm *kvm)
+{
+	struct vgic_dist *dist = &kvm->arch.vgic;
+
+	if (dist->vgic_model != KVM_DEV_TYPE_ARM_VGIC_V3)
+		return false;
+
+	return dist->has_its;
+}
+
 static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
 					    gpa_t addr, unsigned int len)
 {
@@ -43,7 +66,12 @@ static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
 	case GICD_TYPER:
 		value = vcpu->kvm->arch.vgic.nr_spis + VGIC_NR_PRIVATE_IRQS;
 		value = (value >> 5) - 1;
-		value |= (INTERRUPT_ID_BITS_SPIS - 1) << 19;
+		if (vgic_has_its(vcpu->kvm)) {
+			value |= (INTERRUPT_ID_BITS_ITS - 1) << 19;
+			value |= GICD_TYPER_LPIS;
+		} else {
+			value |= (INTERRUPT_ID_BITS_SPIS - 1) << 19;
+		}
 		break;
 	case GICD_IIDR:
 		value = (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
@@ -80,15 +108,17 @@ static unsigned long vgic_mmio_read_irouter(struct kvm_vcpu *vcpu,
 {
 	int intid = VGIC_ADDR_TO_INTID(addr, 64);
 	struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, NULL, intid);
+	unsigned long ret = 0;
 
 	if (!irq)
 		return 0;
 
 	/* The upper word is RAZ for us. */
-	if (addr & 4)
-		return 0;
+	if (!(addr & 4))
+		ret = extract_bytes(READ_ONCE(irq->mpidr), addr & 7, len);
 
-	return extract_bytes(READ_ONCE(irq->mpidr), addr & 7, len);
+	vgic_put_irq(vcpu->kvm, irq);
+	return ret;
 }
 
 static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
@@ -96,15 +126,17 @@ static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
 				    unsigned long val)
 {
 	int intid = VGIC_ADDR_TO_INTID(addr, 64);
-	struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, NULL, intid);
-
-	if (!irq)
-		return;
+	struct vgic_irq *irq;
 
 	/* The upper word is WI for us since we don't implement Aff3. */
 	if (addr & 4)
 		return;
 
+	irq = vgic_get_irq(vcpu->kvm, NULL, intid);
+
+	if (!irq)
+		return;
+
 	spin_lock(&irq->irq_lock);
 
 	/* We only care about and preserve Aff0, Aff1 and Aff2. */
@@ -112,6 +144,32 @@ static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
 	irq->target_vcpu = kvm_mpidr_to_vcpu(vcpu->kvm, irq->mpidr);
 
 	spin_unlock(&irq->irq_lock);
+	vgic_put_irq(vcpu->kvm, irq);
+}
+
+static unsigned long vgic_mmio_read_v3r_ctlr(struct kvm_vcpu *vcpu,
+					     gpa_t addr, unsigned int len)
+{
+	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+
+	return vgic_cpu->lpis_enabled ? GICR_CTLR_ENABLE_LPIS : 0;
+}
+
+
+static void vgic_mmio_write_v3r_ctlr(struct kvm_vcpu *vcpu,
+				     gpa_t addr, unsigned int len,
+				     unsigned long val)
+{
+	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+	bool was_enabled = vgic_cpu->lpis_enabled;
+
+	if (!vgic_has_its(vcpu->kvm))
+		return;
+
+	vgic_cpu->lpis_enabled = val & GICR_CTLR_ENABLE_LPIS;
+
+	if (!was_enabled && vgic_cpu->lpis_enabled)
+		vgic_enable_lpis(vcpu);
 }
 
 static unsigned long vgic_mmio_read_v3r_typer(struct kvm_vcpu *vcpu,
@@ -125,6 +183,8 @@ static unsigned long vgic_mmio_read_v3r_typer(struct kvm_vcpu *vcpu,
 	value |= ((target_vcpu_id & 0xffff) << 8);
 	if (target_vcpu_id == atomic_read(&vcpu->kvm->online_vcpus) - 1)
 		value |= GICR_TYPER_LAST;
+	if (vgic_has_its(vcpu->kvm))
+		value |= GICR_TYPER_PLPIS;
 
 	return extract_bytes(value, addr & 7, len);
 }
@@ -147,6 +207,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
 	return 0;
 }
 
+/* We want to avoid outer shareable. */
+u64 vgic_sanitise_shareability(u64 field)
+{
+	switch (field) {
+	case GIC_BASER_OuterShareable:
+		return GIC_BASER_InnerShareable;
+	default:
+		return field;
+	}
+}
+
+/* Avoid any inner non-cacheable mapping. */
+u64 vgic_sanitise_inner_cacheability(u64 field)
+{
+	switch (field) {
+	case GIC_BASER_CACHE_nCnB:
+	case GIC_BASER_CACHE_nC:
+		return GIC_BASER_CACHE_RaWb;
+	default:
+		return field;
+	}
+}
+
+/* Non-cacheable or same-as-inner are OK. */
+u64 vgic_sanitise_outer_cacheability(u64 field)
+{
+	switch (field) {
+	case GIC_BASER_CACHE_SameAsInner:
+	case GIC_BASER_CACHE_nC:
+		return field;
+	default:
+		return GIC_BASER_CACHE_nC;
+	}
+}
+
+u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
+			u64 (*sanitise_fn)(u64))
+{
+	u64 field = (reg & field_mask) >> field_shift;
+
+	field = sanitise_fn(field) << field_shift;
+	return (reg & ~field_mask) | field;
+}
+
+#define PROPBASER_RES0_MASK						\
+	(GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
+#define PENDBASER_RES0_MASK						\
+	(BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) |	\
+	 GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
+
+static u64 vgic_sanitise_pendbaser(u64 reg)
+{
+	reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
+				  GICR_PENDBASER_SHAREABILITY_SHIFT,
+				  vgic_sanitise_shareability);
+	reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
+				  GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_inner_cacheability);
+	reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
+				  GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_outer_cacheability);
+
+	reg &= ~PENDBASER_RES0_MASK;
+	reg &= ~GENMASK_ULL(51, 48);
+
+	return reg;
+}
+
+static u64 vgic_sanitise_propbaser(u64 reg)
+{
+	reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
+				  GICR_PROPBASER_SHAREABILITY_SHIFT,
+				  vgic_sanitise_shareability);
+	reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
+				  GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_inner_cacheability);
+	reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
+				  GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
+				  vgic_sanitise_outer_cacheability);
+
+	reg &= ~PROPBASER_RES0_MASK;
+	reg &= ~GENMASK_ULL(51, 48);
+	return reg;
+}
+
+static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
+					     gpa_t addr, unsigned int len)
+{
+	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
+
+	return extract_bytes(dist->propbaser, addr & 7, len);
+}
+
+static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
+				     gpa_t addr, unsigned int len,
+				     unsigned long val)
+{
+	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
+	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+	u64 propbaser = dist->propbaser;
+
+	/* Storing a value with LPIs already enabled is undefined */
+	if (vgic_cpu->lpis_enabled)
+		return;
+
+	propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
+	propbaser = vgic_sanitise_propbaser(propbaser);
+
+	dist->propbaser = propbaser;
+}
+
+static unsigned long vgic_mmio_read_pendbase(struct kvm_vcpu *vcpu,
+					     gpa_t addr, unsigned int len)
+{
+	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+
+	return extract_bytes(vgic_cpu->pendbaser, addr & 7, len);
+}
+
+static void vgic_mmio_write_pendbase(struct kvm_vcpu *vcpu,
+				     gpa_t addr, unsigned int len,
+				     unsigned long val)
+{
+	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+	u64 pendbaser = vgic_cpu->pendbaser;
+
+	/* Storing a value with LPIs already enabled is undefined */
+	if (vgic_cpu->lpis_enabled)
+		return;
+
+	pendbaser = update_64bit_reg(pendbaser, addr & 4, len, val);
+	pendbaser = vgic_sanitise_pendbaser(pendbaser);
+
+	vgic_cpu->pendbaser = pendbaser;
+}
+
 /*
  * The GICv3 per-IRQ registers are split to control PPIs and SGIs in the
  * redistributors, while SPIs are covered by registers in the distributor
@@ -218,7 +414,7 @@ static const struct vgic_register_region vgic_v3_dist_registers[] = {
 
 static const struct vgic_register_region vgic_v3_rdbase_registers[] = {
 	REGISTER_DESC_WITH_LENGTH(GICR_CTLR,
-		vgic_mmio_read_raz, vgic_mmio_write_wi, 4,
+		vgic_mmio_read_v3r_ctlr, vgic_mmio_write_v3r_ctlr, 4,
 		VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH(GICR_IIDR,
 		vgic_mmio_read_v3r_iidr, vgic_mmio_write_wi, 4,
@@ -227,10 +423,10 @@ static const struct vgic_register_region vgic_v3_rdbase_registers[] = {
 		vgic_mmio_read_v3r_typer, vgic_mmio_write_wi, 8,
 		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH(GICR_PROPBASER,
-		vgic_mmio_read_raz, vgic_mmio_write_wi, 8,
+		vgic_mmio_read_propbase, vgic_mmio_write_propbase, 8,
 		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH(GICR_PENDBASER,
-		vgic_mmio_read_raz, vgic_mmio_write_wi, 8,
+		vgic_mmio_read_pendbase, vgic_mmio_write_pendbase, 8,
 		VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
 	REGISTER_DESC_WITH_LENGTH(GICR_IDREGS,
 		vgic_mmio_read_v3_idregs, vgic_mmio_write_wi, 48,
@@ -285,24 +481,18 @@ unsigned int vgic_v3_init_dist_iodev(struct vgic_io_device *dev)
 
 int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
 {
-	int nr_vcpus = atomic_read(&kvm->online_vcpus);
 	struct kvm_vcpu *vcpu;
-	struct vgic_io_device *devices;
 	int c, ret = 0;
 
-	devices = kmalloc(sizeof(struct vgic_io_device) * nr_vcpus * 2,
-			  GFP_KERNEL);
-	if (!devices)
-		return -ENOMEM;
-
 	kvm_for_each_vcpu(c, vcpu, kvm) {
 		gpa_t rd_base = redist_base_address + c * SZ_64K * 2;
 		gpa_t sgi_base = rd_base + SZ_64K;
-		struct vgic_io_device *rd_dev = &devices[c * 2];
-		struct vgic_io_device *sgi_dev = &devices[c * 2 + 1];
+		struct vgic_io_device *rd_dev = &vcpu->arch.vgic_cpu.rd_iodev;
+		struct vgic_io_device *sgi_dev = &vcpu->arch.vgic_cpu.sgi_iodev;
 
 		kvm_iodevice_init(&rd_dev->dev, &kvm_io_gic_ops);
 		rd_dev->base_addr = rd_base;
+		rd_dev->iodev_type = IODEV_REDIST;
 		rd_dev->regions = vgic_v3_rdbase_registers;
 		rd_dev->nr_regions = ARRAY_SIZE(vgic_v3_rdbase_registers);
 		rd_dev->redist_vcpu = vcpu;
@@ -317,6 +507,7 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
 
 		kvm_iodevice_init(&sgi_dev->dev, &kvm_io_gic_ops);
 		sgi_dev->base_addr = sgi_base;
+		sgi_dev->iodev_type = IODEV_REDIST;
 		sgi_dev->regions = vgic_v3_sgibase_registers;
 		sgi_dev->nr_regions = ARRAY_SIZE(vgic_v3_sgibase_registers);
 		sgi_dev->redist_vcpu = vcpu;
@@ -335,14 +526,15 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
 	if (ret) {
 		/* The current c failed, so we start with the previous one. */
 		for (c--; c >= 0; c--) {
+			struct vgic_cpu *vgic_cpu;
+
+			vcpu = kvm_get_vcpu(kvm, c);
+			vgic_cpu = &vcpu->arch.vgic_cpu;
 			kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS,
-						  &devices[c * 2].dev);
+						  &vgic_cpu->rd_iodev.dev);
 			kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS,
-						  &devices[c * 2 + 1].dev);
+						  &vgic_cpu->sgi_iodev.dev);
 		}
-		kfree(devices);
-	} else {
-		kvm->arch.vgic.redist_iodevs = devices;
 	}
 
 	return ret;
@@ -451,5 +643,6 @@ void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg)
 		irq->pending = true;
 
 		vgic_queue_irq_unlock(vcpu->kvm, irq);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }

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

@@ -56,6 +56,8 @@ unsigned long vgic_mmio_read_enable(struct kvm_vcpu *vcpu,
 
 		if (irq->enabled)
 			value |= (1U << i);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 
 	return value;
@@ -74,6 +76,8 @@ void vgic_mmio_write_senable(struct kvm_vcpu *vcpu,
 		spin_lock(&irq->irq_lock);
 		irq->enabled = true;
 		vgic_queue_irq_unlock(vcpu->kvm, irq);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -92,6 +96,7 @@ void vgic_mmio_write_cenable(struct kvm_vcpu *vcpu,
 		irq->enabled = false;
 
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -108,6 +113,8 @@ unsigned long vgic_mmio_read_pending(struct kvm_vcpu *vcpu,
 
 		if (irq->pending)
 			value |= (1U << i);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 
 	return value;
@@ -129,6 +136,7 @@ void vgic_mmio_write_spending(struct kvm_vcpu *vcpu,
 			irq->soft_pending = true;
 
 		vgic_queue_irq_unlock(vcpu->kvm, irq);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -152,6 +160,7 @@ void vgic_mmio_write_cpending(struct kvm_vcpu *vcpu,
 		}
 
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -168,6 +177,8 @@ unsigned long vgic_mmio_read_active(struct kvm_vcpu *vcpu,
 
 		if (irq->active)
 			value |= (1U << i);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 
 	return value;
@@ -242,6 +253,7 @@ void vgic_mmio_write_cactive(struct kvm_vcpu *vcpu,
 	for_each_set_bit(i, &val, len * 8) {
 		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
 		vgic_mmio_change_active(vcpu, irq, false);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 	vgic_change_active_finish(vcpu, intid);
 }
@@ -257,6 +269,7 @@ void vgic_mmio_write_sactive(struct kvm_vcpu *vcpu,
 	for_each_set_bit(i, &val, len * 8) {
 		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
 		vgic_mmio_change_active(vcpu, irq, true);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 	vgic_change_active_finish(vcpu, intid);
 }
@@ -272,6 +285,8 @@ unsigned long vgic_mmio_read_priority(struct kvm_vcpu *vcpu,
 		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
 
 		val |= (u64)irq->priority << (i * 8);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 
 	return val;
@@ -298,6 +313,8 @@ void vgic_mmio_write_priority(struct kvm_vcpu *vcpu,
 		/* Narrow the priority range to what we actually support */
 		irq->priority = (val >> (i * 8)) & GENMASK(7, 8 - VGIC_PRI_BITS);
 		spin_unlock(&irq->irq_lock);
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -313,6 +330,8 @@ unsigned long vgic_mmio_read_config(struct kvm_vcpu *vcpu,
 
 		if (irq->config == VGIC_CONFIG_EDGE)
 			value |= (2U << (i * 2));
+
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 
 	return value;
@@ -326,7 +345,7 @@ void vgic_mmio_write_config(struct kvm_vcpu *vcpu,
 	int i;
 
 	for (i = 0; i < len * 4; i++) {
-		struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
+		struct vgic_irq *irq;
 
 		/*
 		 * The configuration cannot be changed for SGIs in general,
@@ -337,14 +356,18 @@ void vgic_mmio_write_config(struct kvm_vcpu *vcpu,
 		if (intid + i < VGIC_NR_PRIVATE_IRQS)
 			continue;
 
+		irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
 		spin_lock(&irq->irq_lock);
+
 		if (test_bit(i * 2 + 1, &val)) {
 			irq->config = VGIC_CONFIG_EDGE;
 		} else {
 			irq->config = VGIC_CONFIG_LEVEL;
 			irq->pending = irq->line_level | irq->soft_pending;
 		}
+
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -450,8 +473,7 @@ static int dispatch_mmio_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
 {
 	struct vgic_io_device *iodev = kvm_to_vgic_iodev(dev);
 	const struct vgic_register_region *region;
-	struct kvm_vcpu *r_vcpu;
-	unsigned long data;
+	unsigned long data = 0;
 
 	region = vgic_find_mmio_region(iodev->regions, iodev->nr_regions,
 				       addr - iodev->base_addr);
@@ -460,8 +482,21 @@ static int dispatch_mmio_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
 		return 0;
 	}
 
-	r_vcpu = iodev->redist_vcpu ? iodev->redist_vcpu : vcpu;
-	data = region->read(r_vcpu, addr, len);
+	switch (iodev->iodev_type) {
+	case IODEV_CPUIF:
+		data = region->read(vcpu, addr, len);
+		break;
+	case IODEV_DIST:
+		data = region->read(vcpu, addr, len);
+		break;
+	case IODEV_REDIST:
+		data = region->read(iodev->redist_vcpu, addr, len);
+		break;
+	case IODEV_ITS:
+		data = region->its_read(vcpu->kvm, iodev->its, addr, len);
+		break;
+	}
+
 	vgic_data_host_to_mmio_bus(val, len, data);
 	return 0;
 }
@@ -471,7 +506,6 @@ static int dispatch_mmio_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
 {
 	struct vgic_io_device *iodev = kvm_to_vgic_iodev(dev);
 	const struct vgic_register_region *region;
-	struct kvm_vcpu *r_vcpu;
 	unsigned long data = vgic_data_mmio_bus_to_host(val, len);
 
 	region = vgic_find_mmio_region(iodev->regions, iodev->nr_regions,
@@ -482,8 +516,21 @@ static int dispatch_mmio_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
 	if (!check_region(region, addr, len))
 		return 0;
 
-	r_vcpu = iodev->redist_vcpu ? iodev->redist_vcpu : vcpu;
-	region->write(r_vcpu, addr, len, data);
+	switch (iodev->iodev_type) {
+	case IODEV_CPUIF:
+		region->write(vcpu, addr, len, data);
+		break;
+	case IODEV_DIST:
+		region->write(vcpu, addr, len, data);
+		break;
+	case IODEV_REDIST:
+		region->write(iodev->redist_vcpu, addr, len, data);
+		break;
+	case IODEV_ITS:
+		region->its_write(vcpu->kvm, iodev->its, addr, len, data);
+		break;
+	}
+
 	return 0;
 }
 
@@ -513,6 +560,7 @@ int vgic_register_dist_iodev(struct kvm *kvm, gpa_t dist_base_address,
 	}
 
 	io_device->base_addr = dist_base_address;
+	io_device->iodev_type = IODEV_DIST;
 	io_device->redist_vcpu = NULL;
 
 	mutex_lock(&kvm->slots_lock);

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

@@ -21,10 +21,19 @@ struct vgic_register_region {
 	unsigned int len;
 	unsigned int bits_per_irq;
 	unsigned int access_flags;
-	unsigned long (*read)(struct kvm_vcpu *vcpu, gpa_t addr,
-			      unsigned int len);
-	void (*write)(struct kvm_vcpu *vcpu, gpa_t addr, unsigned int len,
-		      unsigned long val);
+	union {
+		unsigned long (*read)(struct kvm_vcpu *vcpu, gpa_t addr,
+				      unsigned int len);
+		unsigned long (*its_read)(struct kvm *kvm, struct vgic_its *its,
+					  gpa_t addr, unsigned int len);
+	};
+	union {
+		void (*write)(struct kvm_vcpu *vcpu, gpa_t addr,
+			      unsigned int len, unsigned long val);
+		void (*its_write)(struct kvm *kvm, struct vgic_its *its,
+				  gpa_t addr, unsigned int len,
+				  unsigned long val);
+	};
 };
 
 extern struct kvm_io_device_ops kvm_io_gic_ops;
@@ -87,6 +96,12 @@ unsigned long vgic_data_mmio_bus_to_host(const void *val, unsigned int len);
 void vgic_data_host_to_mmio_bus(void *buf, unsigned int len,
 				unsigned long data);
 
+unsigned long extract_bytes(unsigned long data, unsigned int offset,
+			    unsigned int num);
+
+u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
+		     unsigned long val);
+
 unsigned long vgic_mmio_read_raz(struct kvm_vcpu *vcpu,
 				 gpa_t addr, unsigned int len);
 
@@ -147,4 +162,12 @@ unsigned int vgic_v2_init_dist_iodev(struct vgic_io_device *dev);
 
 unsigned int vgic_v3_init_dist_iodev(struct vgic_io_device *dev);
 
+#ifdef CONFIG_KVM_ARM_VGIC_V3
+u64 vgic_sanitise_outer_cacheability(u64 reg);
+u64 vgic_sanitise_inner_cacheability(u64 reg);
+u64 vgic_sanitise_shareability(u64 reg);
+u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
+			u64 (*sanitise_fn)(u64));
+#endif
+
 #endif

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

@@ -124,6 +124,7 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 		}
 
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -332,20 +333,25 @@ int vgic_v2_probe(const struct gic_kvm_info *info)
 	vtr = readl_relaxed(kvm_vgic_global_state.vctrl_base + GICH_VTR);
 	kvm_vgic_global_state.nr_lr = (vtr & 0x3f) + 1;
 
+	ret = kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+	if (ret) {
+		kvm_err("Cannot register GICv2 KVM device\n");
+		iounmap(kvm_vgic_global_state.vctrl_base);
+		return ret;
+	}
+
 	ret = create_hyp_io_mappings(kvm_vgic_global_state.vctrl_base,
 				     kvm_vgic_global_state.vctrl_base +
 					 resource_size(&info->vctrl),
 				     info->vctrl.start);
-
 	if (ret) {
 		kvm_err("Cannot map VCTRL into hyp\n");
+		kvm_unregister_device_ops(KVM_DEV_TYPE_ARM_VGIC_V2);
 		iounmap(kvm_vgic_global_state.vctrl_base);
 		return ret;
 	}
 
 	kvm_vgic_global_state.can_emulate_gicv2 = true;
-	kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
-
 	kvm_vgic_global_state.vcpu_base = info->vcpu.start;
 	kvm_vgic_global_state.type = VGIC_V2;
 	kvm_vgic_global_state.max_gic_vcpus = VGIC_V2_MAX_CPUS;

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

@@ -81,6 +81,8 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 		else
 			intid = val & GICH_LR_VIRTUALID;
 		irq = vgic_get_irq(vcpu->kvm, vcpu, intid);
+		if (!irq)	/* An LPI could have been unmapped. */
+			continue;
 
 		spin_lock(&irq->irq_lock);
 
@@ -113,6 +115,7 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 		}
 
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(vcpu->kvm, irq);
 	}
 }
 
@@ -190,6 +193,11 @@ void vgic_v3_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
 	vmcrp->pmr  = (vmcr & ICH_VMCR_PMR_MASK) >> ICH_VMCR_PMR_SHIFT;
 }
 
+#define INITIAL_PENDBASER_VALUE						  \
+	(GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWb)		| \
+	GIC_BASER_CACHEABILITY(GICR_PENDBASER, OUTER, SameAsInner)	| \
+	GIC_BASER_SHAREABILITY(GICR_PENDBASER, InnerShareable))
+
 void vgic_v3_enable(struct kvm_vcpu *vcpu)
 {
 	struct vgic_v3_cpu_if *vgic_v3 = &vcpu->arch.vgic_cpu.vgic_v3;
@@ -207,10 +215,12 @@ void vgic_v3_enable(struct kvm_vcpu *vcpu)
 	 * way, so we force SRE to 1 to demonstrate this to the guest.
 	 * This goes with the spec allowing the value to be RAO/WI.
 	 */
-	if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3)
+	if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) {
 		vgic_v3->vgic_sre = ICC_SRE_EL1_SRE;
-	else
+		vcpu->arch.vgic_cpu.pendbaser = INITIAL_PENDBASER_VALUE;
+	} else {
 		vgic_v3->vgic_sre = 0;
+	}
 
 	/* Get the show on the road... */
 	vgic_v3->vgic_hcr = ICH_HCR_EN;
@@ -296,6 +306,7 @@ out:
 int vgic_v3_probe(const struct gic_kvm_info *info)
 {
 	u32 ich_vtr_el2 = kvm_call_hyp(__vgic_v3_get_ich_vtr_el2);
+	int ret;
 
 	/*
 	 * The ListRegs field is 5 bits, but there is a architectural
@@ -319,12 +330,22 @@ int vgic_v3_probe(const struct gic_kvm_info *info)
 	} else {
 		kvm_vgic_global_state.vcpu_base = info->vcpu.start;
 		kvm_vgic_global_state.can_emulate_gicv2 = true;
-		kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+		ret = kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+		if (ret) {
+			kvm_err("Cannot register GICv2 KVM device.\n");
+			return ret;
+		}
 		kvm_info("vgic-v2@%llx\n", info->vcpu.start);
 	}
+	ret = kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V3);
+	if (ret) {
+		kvm_err("Cannot register GICv3 KVM device.\n");
+		kvm_unregister_device_ops(KVM_DEV_TYPE_ARM_VGIC_V2);
+		return ret;
+	}
+
 	if (kvm_vgic_global_state.vcpu_base == 0)
 		kvm_info("disabling GICv2 emulation\n");
-	kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V3);
 
 	kvm_vgic_global_state.vctrl_base = NULL;
 	kvm_vgic_global_state.type = VGIC_V3;

virt/kvm/arm/vgic/vgic.c

@@ -33,10 +33,17 @@ struct vgic_global __section(.hyp.text) kvm_vgic_global_state;
 
 /*
  * Locking order is always:
- *   vgic_cpu->ap_list_lock
- *     vgic_irq->irq_lock
+ * its->cmd_lock (mutex)
+ *   its->its_lock (mutex)
+ *     vgic_cpu->ap_list_lock
+ *       kvm->lpi_list_lock
+ *         vgic_irq->irq_lock
  *
- * (that is, always take the ap_list_lock before the struct vgic_irq lock).
+ * If you need to take multiple locks, always take the upper lock first,
+ * then the lower ones, e.g. first take the its_lock, then the irq_lock.
+ * If you are already holding a lock and need to take a higher one, you
+ * have to drop the lower ranking lock first and re-aquire it after having
+ * taken the upper one.
  *
  * When taking more than one ap_list_lock at the same time, always take the
  * lowest numbered VCPU's ap_list_lock first, so:
@@ -45,6 +52,41 @@ struct vgic_global __section(.hyp.text) kvm_vgic_global_state;
  *     spin_lock(vcpuY->arch.vgic_cpu.ap_list_lock);
  */
 
+/*
+ * Iterate over the VM's list of mapped LPIs to find the one with a
+ * matching interrupt ID and return a reference to the IRQ structure.
+ */
+static struct vgic_irq *vgic_get_lpi(struct kvm *kvm, u32 intid)
+{
+	struct vgic_dist *dist = &kvm->arch.vgic;
+	struct vgic_irq *irq = NULL;
+
+	spin_lock(&dist->lpi_list_lock);
+
+	list_for_each_entry(irq, &dist->lpi_list_head, lpi_list) {
+		if (irq->intid != intid)
+			continue;
+
+		/*
+		 * This increases the refcount, the caller is expected to
+		 * call vgic_put_irq() later once it's finished with the IRQ.
+		 */
+		vgic_get_irq_kref(irq);
+		goto out_unlock;
+	}
+	irq = NULL;
+
+out_unlock:
+	spin_unlock(&dist->lpi_list_lock);
+
+	return irq;
+}
+
+/*
+ * This looks up the virtual interrupt ID to get the corresponding
+ * struct vgic_irq. It also increases the refcount, so any caller is expected
+ * to call vgic_put_irq() once it's finished with this IRQ.
+ */
 struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
 			      u32 intid)
 {
@@ -56,14 +98,43 @@ struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
 	if (intid <= VGIC_MAX_SPI)
 		return &kvm->arch.vgic.spis[intid - VGIC_NR_PRIVATE_IRQS];
 
-	/* LPIs are not yet covered */
+	/* LPIs */
 	if (intid >= VGIC_MIN_LPI)
-		return NULL;
+		return vgic_get_lpi(kvm, intid);
 
 	WARN(1, "Looking up struct vgic_irq for reserved INTID");
 	return NULL;
 }
 
+/*
+ * We can't do anything in here, because we lack the kvm pointer to
+ * lock and remove the item from the lpi_list. So we keep this function
+ * empty and use the return value of kref_put() to trigger the freeing.
+ */
+static void vgic_irq_release(struct kref *ref)
+{
+}
+
+void vgic_put_irq(struct kvm *kvm, struct vgic_irq *irq)
+{
+	struct vgic_dist *dist;
+
+	if (irq->intid < VGIC_MIN_LPI)
+		return;
+
+	if (!kref_put(&irq->refcount, vgic_irq_release))
+		return;
+
+	dist = &kvm->arch.vgic;
+
+	spin_lock(&dist->lpi_list_lock);
+	list_del(&irq->lpi_list);
+	dist->lpi_list_count--;
+	spin_unlock(&dist->lpi_list_lock);
+
+	kfree(irq);
+}
+
 /**
  * kvm_vgic_target_oracle - compute the target vcpu for an irq
  *
@@ -236,6 +307,11 @@ retry:
 		goto retry;
 	}
 
+	/*
+	 * Grab a reference to the irq to reflect the fact that it is
+	 * now in the ap_list.
+	 */
+	vgic_get_irq_kref(irq);
 	list_add_tail(&irq->ap_list, &vcpu->arch.vgic_cpu.ap_list_head);
 	irq->vcpu = vcpu;
 
@@ -269,14 +345,17 @@ static int vgic_update_irq_pending(struct kvm *kvm, int cpuid,
 	if (!irq)
 		return -EINVAL;
 
-	if (irq->hw != mapped_irq)
+	if (irq->hw != mapped_irq) {
+		vgic_put_irq(kvm, irq);
 		return -EINVAL;
+	}
 
 	spin_lock(&irq->irq_lock);
 
 	if (!vgic_validate_injection(irq, level)) {
 		/* Nothing to see here, move along... */
 		spin_unlock(&irq->irq_lock);
+		vgic_put_irq(kvm, irq);
 		return 0;
 	}
 
@@ -288,6 +367,7 @@ static int vgic_update_irq_pending(struct kvm *kvm, int cpuid,
 	}
 
 	vgic_queue_irq_unlock(kvm, irq);
+	vgic_put_irq(kvm, irq);
 
 	return 0;
 }
@@ -330,25 +410,28 @@ int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, u32 virt_irq, u32 phys_irq)
 	irq->hwintid = phys_irq;
 
 	spin_unlock(&irq->irq_lock);
+	vgic_put_irq(vcpu->kvm, irq);
 
 	return 0;
 }
 
 int kvm_vgic_unmap_phys_irq(struct kvm_vcpu *vcpu, unsigned int virt_irq)
 {
-	struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, virt_irq);
-
-	BUG_ON(!irq);
+	struct vgic_irq *irq;
 
 	if (!vgic_initialized(vcpu->kvm))
 		return -EAGAIN;
 
+	irq = vgic_get_irq(vcpu->kvm, vcpu, virt_irq);
+	BUG_ON(!irq);
+
 	spin_lock(&irq->irq_lock);
 
 	irq->hw = false;
 	irq->hwintid = 0;
 
 	spin_unlock(&irq->irq_lock);
+	vgic_put_irq(vcpu->kvm, irq);
 
 	return 0;
 }
@@ -386,6 +469,15 @@ retry:
 			list_del(&irq->ap_list);
 			irq->vcpu = NULL;
 			spin_unlock(&irq->irq_lock);
+
+			/*
+			 * This vgic_put_irq call matches the
+			 * vgic_get_irq_kref in vgic_queue_irq_unlock,
+			 * where we added the LPI to the ap_list. As
+			 * we remove the irq from the list, we drop
+			 * also drop the refcount.
+			 */
+			vgic_put_irq(vcpu->kvm, irq);
 			continue;
 		}
 
@@ -614,6 +706,15 @@ bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq)
 	spin_lock(&irq->irq_lock);
 	map_is_active = irq->hw && irq->active;
 	spin_unlock(&irq->irq_lock);
+	vgic_put_irq(vcpu->kvm, irq);
 
 	return map_is_active;
 }
+
+int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi)
+{
+	if (vgic_has_its(kvm))
+		return vgic_its_inject_msi(kvm, msi);
+	else
+		return -ENODEV;
+}

virt/kvm/arm/vgic/vgic.h

@@ -25,6 +25,7 @@
 #define IS_VGIC_ADDR_UNDEF(_x)  ((_x) == VGIC_ADDR_UNDEF)
 
 #define INTERRUPT_ID_BITS_SPIS	10
+#define INTERRUPT_ID_BITS_ITS	16
 #define VGIC_PRI_BITS		5
 
 #define vgic_irq_is_sgi(intid) ((intid) < VGIC_NR_SGIS)
@@ -38,9 +39,13 @@ struct vgic_vmcr {
 
 struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
 			      u32 intid);
+void vgic_put_irq(struct kvm *kvm, struct vgic_irq *irq);
 bool vgic_queue_irq_unlock(struct kvm *kvm, struct vgic_irq *irq);
 void vgic_kick_vcpus(struct kvm *kvm);
 
+int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
+		      phys_addr_t addr, phys_addr_t alignment);
+
 void vgic_v2_process_maintenance(struct kvm_vcpu *vcpu);
 void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu);
 void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr);
@@ -59,6 +64,14 @@ int vgic_v2_map_resources(struct kvm *kvm);
 int vgic_register_dist_iodev(struct kvm *kvm, gpa_t dist_base_address,
 			     enum vgic_type);
 
+static inline void vgic_get_irq_kref(struct vgic_irq *irq)
+{
+	if (irq->intid < VGIC_MIN_LPI)
+		return;
+
+	kref_get(&irq->refcount);
+}
+
 #ifdef CONFIG_KVM_ARM_VGIC_V3
 void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu);
 void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu);
@@ -71,6 +84,10 @@ void vgic_v3_enable(struct kvm_vcpu *vcpu);
 int vgic_v3_probe(const struct gic_kvm_info *info);
 int vgic_v3_map_resources(struct kvm *kvm);
 int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t dist_base_address);
+bool vgic_has_its(struct kvm *kvm);
+int kvm_vgic_register_its_device(void);
+void vgic_enable_lpis(struct kvm_vcpu *vcpu);
+int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi);
 #else
 static inline void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu)
 {
@@ -122,9 +139,28 @@ static inline int vgic_register_redist_iodevs(struct kvm *kvm,
 {
 	return -ENODEV;
 }
+
+static inline bool vgic_has_its(struct kvm *kvm)
+{
+	return false;
+}
+
+static inline int kvm_vgic_register_its_device(void)
+{
+	return -ENODEV;
+}
+
+static inline void vgic_enable_lpis(struct kvm_vcpu *vcpu)
+{
+}
+
+static inline int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
+{
+	return -ENODEV;
+}
 #endif
 
-void kvm_register_vgic_device(unsigned long type);
+int kvm_register_vgic_device(unsigned long type);
 int vgic_lazy_init(struct kvm *kvm);
 int vgic_init(struct kvm *kvm);
 

virt/kvm/kvm_main.c

@@ -3545,6 +3545,30 @@ int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
 	return r;
 }
 
+struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
+					 gpa_t addr)
+{
+	struct kvm_io_bus *bus;
+	int dev_idx, srcu_idx;
+	struct kvm_io_device *iodev = NULL;
+
+	srcu_idx = srcu_read_lock(&kvm->srcu);
+
+	bus = srcu_dereference(kvm->buses[bus_idx], &kvm->srcu);
+
+	dev_idx = kvm_io_bus_get_first_dev(bus, addr, 1);
+	if (dev_idx < 0)
+		goto out_unlock;
+
+	iodev = bus->range[dev_idx].dev;
+
+out_unlock:
+	srcu_read_unlock(&kvm->srcu, srcu_idx);
+
+	return iodev;
+}
+EXPORT_SYMBOL_GPL(kvm_io_bus_get_dev);
+
 static struct notifier_block kvm_cpu_notifier = {
 	.notifier_call = kvm_cpu_hotplug,
 };