Merge tag 'kvm-arm-for-4-7-take2' of...

Merge tag 'kvm-arm-for-4-7-take2' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into kvm-next

KVM/ARM Changes for v4.7 take 2

"The GIC is dead; Long live the GIC"

This set of changes include the new vgic, which is a reimplementation of
our horribly broken legacy vgic implementation.  The two implementations
will live side-by-side (with the new being the configured default) for
one kernel release and then we'll remove it.

Also fixes a non-critical issue with virtual abort injection to guests.

arch/arm/include/asm/kvm_host.h

@@ -41,6 +41,8 @@
 
 #define KVM_MAX_VCPUS VGIC_V2_MAX_CPUS
 
+#define KVM_REQ_VCPU_EXIT	8
+
 u32 *kvm_vcpu_reg(struct kvm_vcpu *vcpu, u8 reg_num, u32 mode);
 int __attribute_const__ kvm_target_cpu(void);
 int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
@@ -226,6 +228,10 @@ static inline void kvm_arch_mmu_notifier_invalidate_page(struct kvm *kvm,
 
 struct kvm_vcpu *kvm_arm_get_running_vcpu(void);
 struct kvm_vcpu __percpu **kvm_get_running_vcpus(void);
+void kvm_arm_halt_guest(struct kvm *kvm);
+void kvm_arm_resume_guest(struct kvm *kvm);
+void kvm_arm_halt_vcpu(struct kvm_vcpu *vcpu);
+void kvm_arm_resume_vcpu(struct kvm_vcpu *vcpu);
 
 int kvm_arm_copy_coproc_indices(struct kvm_vcpu *vcpu, u64 __user *uindices);
 unsigned long kvm_arm_num_coproc_regs(struct kvm_vcpu *vcpu);

arch/arm/include/asm/kvm_mmio.h

@@ -28,6 +28,9 @@ struct kvm_decode {
 	bool sign_extend;
 };
 
+void kvm_mmio_write_buf(void *buf, unsigned int len, unsigned long data);
+unsigned long kvm_mmio_read_buf(const void *buf, unsigned int len);
+
 int kvm_handle_mmio_return(struct kvm_vcpu *vcpu, struct kvm_run *run);
 int io_mem_abort(struct kvm_vcpu *vcpu, struct kvm_run *run,
 		 phys_addr_t fault_ipa);

arch/arm/kvm/Kconfig

@@ -46,6 +46,13 @@ 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

@@ -21,7 +21,18 @@ obj-$(CONFIG_KVM_ARM_HOST) += hyp/
 obj-y += kvm-arm.o init.o interrupts.o
 obj-y += arm.o handle_exit.o guest.o mmu.o emulate.o reset.o
 obj-y += coproc.o coproc_a15.o coproc_a7.o mmio.o psci.o perf.o
+
+ifeq ($(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
+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

@@ -459,7 +459,7 @@ static void update_vttbr(struct kvm *kvm)
 static int kvm_vcpu_first_run_init(struct kvm_vcpu *vcpu)
 {
 	struct kvm *kvm = vcpu->kvm;
-	int ret;
+	int ret = 0;
 
 	if (likely(vcpu->arch.has_run_once))
 		return 0;
@@ -482,9 +482,9 @@ static int kvm_vcpu_first_run_init(struct kvm_vcpu *vcpu)
 	 * interrupts from the virtual timer with a userspace gic.
 	 */
 	if (irqchip_in_kernel(kvm) && vgic_initialized(kvm))
-		kvm_timer_enable(kvm);
+		ret = kvm_timer_enable(vcpu);
 
-	return 0;
+	return ret;
 }
 
 bool kvm_arch_intc_initialized(struct kvm *kvm)
@@ -492,30 +492,37 @@ bool kvm_arch_intc_initialized(struct kvm *kvm)
 	return vgic_initialized(kvm);
 }
 
-static void kvm_arm_halt_guest(struct kvm *kvm) __maybe_unused;
-static void kvm_arm_resume_guest(struct kvm *kvm) __maybe_unused;
-
-static void kvm_arm_halt_guest(struct kvm *kvm)
+void kvm_arm_halt_guest(struct kvm *kvm)
 {
 	int i;
 	struct kvm_vcpu *vcpu;
 
 	kvm_for_each_vcpu(i, vcpu, kvm)
 		vcpu->arch.pause = true;
-	force_vm_exit(cpu_all_mask);
+	kvm_make_all_cpus_request(kvm, KVM_REQ_VCPU_EXIT);
+}
+
+void kvm_arm_halt_vcpu(struct kvm_vcpu *vcpu)
+{
+	vcpu->arch.pause = true;
+	kvm_vcpu_kick(vcpu);
 }
 
-static void kvm_arm_resume_guest(struct kvm *kvm)
+void kvm_arm_resume_vcpu(struct kvm_vcpu *vcpu)
+{
+	struct swait_queue_head *wq = kvm_arch_vcpu_wq(vcpu);
+
+	vcpu->arch.pause = false;
+	swake_up(wq);
+}
+
+void kvm_arm_resume_guest(struct kvm *kvm)
 {
 	int i;
 	struct kvm_vcpu *vcpu;
 
-	kvm_for_each_vcpu(i, vcpu, kvm) {
-		struct swait_queue_head *wq = kvm_arch_vcpu_wq(vcpu);
-
-		vcpu->arch.pause = false;
-		swake_up(wq);
-	}
+	kvm_for_each_vcpu(i, vcpu, kvm)
+		kvm_arm_resume_vcpu(vcpu);
 }
 
 static void vcpu_sleep(struct kvm_vcpu *vcpu)

arch/arm/kvm/mmio.c

@@ -23,7 +23,7 @@
 
 #include "trace.h"
 
-static void mmio_write_buf(char *buf, unsigned int len, unsigned long data)
+void kvm_mmio_write_buf(void *buf, unsigned int len, unsigned long data)
 {
 	void *datap = NULL;
 	union {
@@ -55,7 +55,7 @@ static void mmio_write_buf(char *buf, unsigned int len, unsigned long data)
 	memcpy(buf, datap, len);
 }
 
-static unsigned long mmio_read_buf(char *buf, unsigned int len)
+unsigned long kvm_mmio_read_buf(const void *buf, unsigned int len)
 {
 	unsigned long data = 0;
 	union {
@@ -66,7 +66,7 @@ static unsigned long mmio_read_buf(char *buf, unsigned int len)
 
 	switch (len) {
 	case 1:
-		data = buf[0];
+		data = *(u8 *)buf;
 		break;
 	case 2:
 		memcpy(&tmp.hword, buf, len);
@@ -87,11 +87,10 @@ static unsigned long mmio_read_buf(char *buf, unsigned int len)
 
 /**
  * kvm_handle_mmio_return -- Handle MMIO loads after user space emulation
+ *			     or in-kernel IO emulation
+ *
  * @vcpu: The VCPU pointer
  * @run:  The VCPU run struct containing the mmio data
- *
- * This should only be called after returning from userspace for MMIO load
- * emulation.
  */
 int kvm_handle_mmio_return(struct kvm_vcpu *vcpu, struct kvm_run *run)
 {
@@ -104,7 +103,7 @@ int kvm_handle_mmio_return(struct kvm_vcpu *vcpu, struct kvm_run *run)
 		if (len > sizeof(unsigned long))
 			return -EINVAL;
 
-		data = mmio_read_buf(run->mmio.data, len);
+		data = kvm_mmio_read_buf(run->mmio.data, len);
 
 		if (vcpu->arch.mmio_decode.sign_extend &&
 		    len < sizeof(unsigned long)) {
@@ -190,7 +189,7 @@ int io_mem_abort(struct kvm_vcpu *vcpu, struct kvm_run *run,
 					       len);
 
 		trace_kvm_mmio(KVM_TRACE_MMIO_WRITE, len, fault_ipa, data);
-		mmio_write_buf(data_buf, len, data);
+		kvm_mmio_write_buf(data_buf, len, data);
 
 		ret = kvm_io_bus_write(vcpu, KVM_MMIO_BUS, fault_ipa, len,
 				       data_buf);
@@ -206,18 +205,19 @@ int io_mem_abort(struct kvm_vcpu *vcpu, struct kvm_run *run,
 	run->mmio.is_write	= is_write;
 	run->mmio.phys_addr	= fault_ipa;
 	run->mmio.len		= len;
-	if (is_write)
-		memcpy(run->mmio.data, data_buf, len);
 
 	if (!ret) {
 		/* We handled the access successfully in the kernel. */
+		if (!is_write)
+			memcpy(run->mmio.data, data_buf, len);
 		vcpu->stat.mmio_exit_kernel++;
 		kvm_handle_mmio_return(vcpu, run);
 		return 1;
-	} else {
-		vcpu->stat.mmio_exit_user++;
 	}
 
+	if (is_write)
+		memcpy(run->mmio.data, data_buf, len);
+	vcpu->stat.mmio_exit_user++;
 	run->exit_reason	= KVM_EXIT_MMIO;
 	return 0;
 }

arch/arm64/include/asm/kvm_host.h

@@ -43,6 +43,8 @@
 
 #define KVM_VCPU_MAX_FEATURES 4
 
+#define KVM_REQ_VCPU_EXIT	8
+
 int __attribute_const__ kvm_target_cpu(void);
 int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
 int kvm_arch_dev_ioctl_check_extension(long ext);
@@ -325,6 +327,10 @@ static inline void kvm_arch_mmu_notifier_invalidate_page(struct kvm *kvm,
 
 struct kvm_vcpu *kvm_arm_get_running_vcpu(void);
 struct kvm_vcpu * __percpu *kvm_get_running_vcpus(void);
+void kvm_arm_halt_guest(struct kvm *kvm);
+void kvm_arm_resume_guest(struct kvm *kvm);
+void kvm_arm_halt_vcpu(struct kvm_vcpu *vcpu);
+void kvm_arm_resume_vcpu(struct kvm_vcpu *vcpu);
 
 u64 __kvm_call_hyp(void *hypfn, ...);
 #define kvm_call_hyp(f, ...) __kvm_call_hyp(kvm_ksym_ref(f), ##__VA_ARGS__)

arch/arm64/include/asm/kvm_mmio.h

@@ -30,6 +30,9 @@ struct kvm_decode {
 	bool sign_extend;
 };
 
+void kvm_mmio_write_buf(void *buf, unsigned int len, unsigned long data);
+unsigned long kvm_mmio_read_buf(const void *buf, unsigned int len);
+
 int kvm_handle_mmio_return(struct kvm_vcpu *vcpu, struct kvm_run *run);
 int io_mem_abort(struct kvm_vcpu *vcpu, struct kvm_run *run,
 		 phys_addr_t fault_ipa);

arch/arm64/kvm/Kconfig

@@ -54,6 +54,13 @@ 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

@@ -20,10 +20,22 @@ 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) += $(KVM)/arm/vgic/vgic-irqfd.o
+kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-v2.o
+kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-v3.o
+kvm-$(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) += $(KVM)/arm/arch_timer.o
 kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o

arch/arm64/kvm/inject_fault.c

@@ -162,7 +162,7 @@ static void inject_abt64(struct kvm_vcpu *vcpu, bool is_iabt, unsigned long addr
 		esr |= (ESR_ELx_EC_IABT_CUR << ESR_ELx_EC_SHIFT);
 
 	if (!is_iabt)
-		esr |= ESR_ELx_EC_DABT_LOW;
+		esr |= ESR_ELx_EC_DABT_LOW << ESR_ELx_EC_SHIFT;
 
 	vcpu_sys_reg(vcpu, ESR_EL1) = esr | ESR_ELx_FSC_EXTABT;
 }

include/kvm/arm_arch_timer.h

@@ -24,9 +24,6 @@
 #include <linux/workqueue.h>
 
 struct arch_timer_kvm {
-	/* Is the timer enabled */
-	bool			enabled;
-
 	/* Virtual offset */
 	cycle_t			cntvoff;
 };
@@ -53,15 +50,15 @@ struct arch_timer_cpu {
 	/* Timer IRQ */
 	struct kvm_irq_level		irq;
 
-	/* VGIC mapping */
-	struct irq_phys_map		*map;
-
 	/* Active IRQ state caching */
 	bool				active_cleared_last;
+
+	/* Is the timer enabled */
+	bool			enabled;
 };
 
 int kvm_timer_hyp_init(void);
-void kvm_timer_enable(struct kvm *kvm);
+int kvm_timer_enable(struct kvm_vcpu *vcpu);
 void kvm_timer_init(struct kvm *kvm);
 int kvm_timer_vcpu_reset(struct kvm_vcpu *vcpu,
 			 const struct kvm_irq_level *irq);

include/kvm/arm_vgic.h

@@ -19,6 +19,10 @@
 #ifndef __ASM_ARM_KVM_VGIC_H
 #define __ASM_ARM_KVM_VGIC_H
 
+#ifdef CONFIG_KVM_NEW_VGIC
+#include <kvm/vgic/vgic.h>
+#else
+
 #include <linux/kernel.h>
 #include <linux/kvm.h>
 #include <linux/irqreturn.h>
@@ -158,7 +162,6 @@ struct vgic_io_device {
 struct irq_phys_map {
 	u32			virt_irq;
 	u32			phys_irq;
-	u32			irq;
 };
 
 struct irq_phys_map_entry {
@@ -305,9 +308,6 @@ struct vgic_cpu {
 	unsigned long   *active_shared;
 	unsigned long   *pend_act_shared;
 
-	/* Number of list registers on this CPU */
-	int		nr_lr;
-
 	/* CPU vif control registers for world switch */
 	union {
 		struct vgic_v2_cpu_if	vgic_v2;
@@ -342,17 +342,18 @@ void kvm_vgic_sync_hwstate(struct kvm_vcpu *vcpu);
 int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int irq_num,
 			bool level);
 int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid,
-			       struct irq_phys_map *map, bool level);
+			       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);
-struct irq_phys_map *kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu,
-					   int virt_irq, int irq);
-int kvm_vgic_unmap_phys_irq(struct kvm_vcpu *vcpu, struct irq_phys_map *map);
-bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, struct irq_phys_map *map);
+int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, int virt_irq, int 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);
 
 #define irqchip_in_kernel(k)	(!!((k)->arch.vgic.in_kernel))
 #define vgic_initialized(k)	(!!((k)->arch.vgic.nr_cpus))
 #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))
 
 int vgic_v2_probe(const struct gic_kvm_info *gic_kvm_info,
 		  const struct vgic_ops **ops,
@@ -370,4 +371,5 @@ static inline int vgic_v3_probe(const struct gic_kvm_info *gic_kvm_info,
 }
 #endif
 
+#endif	/* old VGIC include */
 #endif

include/kvm/vgic/vgic.h

+/*
+ * 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

@@ -275,6 +275,12 @@
 #define ICH_LR_ACTIVE_BIT		(1ULL << 63)
 #define ICH_LR_PHYS_ID_SHIFT		32
 #define ICH_LR_PHYS_ID_MASK		(0x3ffULL << ICH_LR_PHYS_ID_SHIFT)
+#define ICH_LR_PRIORITY_SHIFT		48
+
+/* These are for GICv2 emulation only */
+#define GICH_LR_VIRTUALID		(0x3ffUL << 0)
+#define GICH_LR_PHYSID_CPUID_SHIFT	(10)
+#define GICH_LR_PHYSID_CPUID		(7UL << GICH_LR_PHYSID_CPUID_SHIFT)
 
 #define ICH_MISR_EOI			(1 << 0)
 #define ICH_MISR_U			(1 << 1)

include/linux/irqchip/arm-gic.h

@@ -33,6 +33,7 @@
 
 #define GIC_DIST_CTRL			0x000
 #define GIC_DIST_CTR			0x004
+#define GIC_DIST_IIDR			0x008
 #define GIC_DIST_IGROUP			0x080
 #define GIC_DIST_ENABLE_SET		0x100
 #define GIC_DIST_ENABLE_CLEAR		0x180
@@ -76,6 +77,7 @@
 #define GICH_LR_VIRTUALID		(0x3ff << 0)
 #define GICH_LR_PHYSID_CPUID_SHIFT	(10)
 #define GICH_LR_PHYSID_CPUID		(0x3ff << GICH_LR_PHYSID_CPUID_SHIFT)
+#define GICH_LR_PRIORITY_SHIFT		23
 #define GICH_LR_STATE			(3 << 28)
 #define GICH_LR_PENDING_BIT		(1 << 28)
 #define GICH_LR_ACTIVE_BIT		(1 << 29)

virt/kvm/arm/arch_timer.c

@@ -20,6 +20,7 @@
 #include <linux/kvm.h>
 #include <linux/kvm_host.h>
 #include <linux/interrupt.h>
+#include <linux/irq.h>
 
 #include <clocksource/arm_arch_timer.h>
 #include <asm/arch_timer.h>
@@ -174,10 +175,10 @@ static void kvm_timer_update_irq(struct kvm_vcpu *vcpu, bool new_level)
 
 	timer->active_cleared_last = false;
 	timer->irq.level = new_level;
-	trace_kvm_timer_update_irq(vcpu->vcpu_id, timer->map->virt_irq,
+	trace_kvm_timer_update_irq(vcpu->vcpu_id, timer->irq.irq,
 				   timer->irq.level);
 	ret = kvm_vgic_inject_mapped_irq(vcpu->kvm, vcpu->vcpu_id,
-					 timer->map,
+					 timer->irq.irq,
 					 timer->irq.level);
 	WARN_ON(ret);
 }
@@ -196,7 +197,7 @@ static int kvm_timer_update_state(struct kvm_vcpu *vcpu)
 	 * because the guest would never see the interrupt.  Instead wait
 	 * until we call this function from kvm_timer_flush_hwstate.
 	 */
-	if (!vgic_initialized(vcpu->kvm))
+	if (!vgic_initialized(vcpu->kvm) || !timer->enabled)
 		return -ENODEV;
 
 	if (kvm_timer_should_fire(vcpu) != timer->irq.level)
@@ -274,10 +275,8 @@ void kvm_timer_flush_hwstate(struct kvm_vcpu *vcpu)
 	* to ensure that hardware interrupts from the timer triggers a guest
 	* exit.
 	*/
-	if (timer->irq.level || kvm_vgic_map_is_active(vcpu, timer->map))
-		phys_active = true;
-	else
-		phys_active = false;
+	phys_active = timer->irq.level ||
+			kvm_vgic_map_is_active(vcpu, timer->irq.irq);
 
 	/*
 	 * We want to avoid hitting the (re)distributor as much as
@@ -302,7 +301,7 @@ void kvm_timer_flush_hwstate(struct kvm_vcpu *vcpu)
 	if (timer->active_cleared_last && !phys_active)
 		return;
 
-	ret = irq_set_irqchip_state(timer->map->irq,
+	ret = irq_set_irqchip_state(host_vtimer_irq,
 				    IRQCHIP_STATE_ACTIVE,
 				    phys_active);
 	WARN_ON(ret);
@@ -334,7 +333,6 @@ int kvm_timer_vcpu_reset(struct kvm_vcpu *vcpu,
 			 const struct kvm_irq_level *irq)
 {
 	struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
-	struct irq_phys_map *map;
 
 	/*
 	 * The vcpu timer irq number cannot be determined in
@@ -353,15 +351,6 @@ int kvm_timer_vcpu_reset(struct kvm_vcpu *vcpu,
 	timer->cntv_ctl = 0;
 	kvm_timer_update_state(vcpu);
 
-	/*
-	 * Tell the VGIC that the virtual interrupt is tied to a
-	 * physical interrupt. We do that once per VCPU.
-	 */
-	map = kvm_vgic_map_phys_irq(vcpu, irq->irq, host_vtimer_irq);
-	if (WARN_ON(IS_ERR(map)))
-		return PTR_ERR(map);
-
-	timer->map = map;
 	return 0;
 }
 
@@ -487,14 +476,43 @@ void kvm_timer_vcpu_terminate(struct kvm_vcpu *vcpu)
 	struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
 
 	timer_disarm(timer);
-	if (timer->map)
-		kvm_vgic_unmap_phys_irq(vcpu, timer->map);
+	kvm_vgic_unmap_phys_irq(vcpu, timer->irq.irq);
 }
 
-void kvm_timer_enable(struct kvm *kvm)
+int kvm_timer_enable(struct kvm_vcpu *vcpu)
 {
-	if (kvm->arch.timer.enabled)
-		return;
+	struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
+	struct irq_desc *desc;
+	struct irq_data *data;
+	int phys_irq;
+	int ret;
+
+	if (timer->enabled)
+		return 0;
+
+	/*
+	 * Find the physical IRQ number corresponding to the host_vtimer_irq
+	 */
+	desc = irq_to_desc(host_vtimer_irq);
+	if (!desc) {
+		kvm_err("%s: no interrupt descriptor\n", __func__);
+		return -EINVAL;
+	}
+
+	data = irq_desc_get_irq_data(desc);
+	while (data->parent_data)
+		data = data->parent_data;
+
+	phys_irq = data->hwirq;
+
+	/*
+	 * Tell the VGIC that the virtual interrupt is tied to a
+	 * physical interrupt. We do that once per VCPU.
+	 */
+	ret = kvm_vgic_map_phys_irq(vcpu, timer->irq.irq, phys_irq);
+	if (ret)
+		return ret;
+
 
 	/*
 	 * There is a potential race here between VCPUs starting for the first
@@ -505,7 +523,9 @@ void kvm_timer_enable(struct kvm *kvm)
 	 * the arch timers are enabled.
 	 */
 	if (timecounter && wqueue)
-		kvm->arch.timer.enabled = 1;
+		timer->enabled = 1;
+
+	return 0;
 }
 
 void kvm_timer_init(struct kvm *kvm)

virt/kvm/arm/hyp/timer-sr.c

@@ -24,11 +24,10 @@
 /* vcpu is already in the HYP VA space */
 void __hyp_text __timer_save_state(struct kvm_vcpu *vcpu)
 {
-	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
 	struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
 	u64 val;
 
-	if (kvm->arch.timer.enabled) {
+	if (timer->enabled) {
 		timer->cntv_ctl = read_sysreg_el0(cntv_ctl);
 		timer->cntv_cval = read_sysreg_el0(cntv_cval);
 	}
@@ -60,7 +59,7 @@ void __hyp_text __timer_restore_state(struct kvm_vcpu *vcpu)
 	val |= CNTHCTL_EL1PCTEN;
 	write_sysreg(val, cnthctl_el2);
 
-	if (kvm->arch.timer.enabled) {
+	if (timer->enabled) {
 		write_sysreg(kvm->arch.timer.cntvoff, cntvoff_el2);
 		write_sysreg_el0(timer->cntv_cval, cntv_cval);
 		isb();

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

@@ -21,11 +21,18 @@
 
 #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 = vcpu->arch.vgic_cpu.nr_lr;
+	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
 	u32 eisr0, eisr1;
 	int i;
 	bool expect_mi;
@@ -67,7 +74,7 @@ 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 = vcpu->arch.vgic_cpu.nr_lr;
+	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
 	u32 elrsr0, elrsr1;
 
 	elrsr0 = readl_relaxed(base + GICH_ELRSR0);
@@ -86,7 +93,7 @@ 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 = vcpu->arch.vgic_cpu.nr_lr;
+	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
 	int i;
 
 	for (i = 0; i < nr_lr; i++) {
@@ -141,13 +148,13 @@ 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 i, nr_lr;
+	int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+	int i;
 	u64 live_lrs = 0;
 
 	if (!base)
 		return;
 
-	nr_lr = vcpu->arch.vgic_cpu.nr_lr;
 
 	for (i = 0; i < nr_lr; i++)
 		if (cpu_if->vgic_lr[i] & GICH_LR_STATE)

virt/kvm/arm/pmu.c

@@ -436,7 +436,14 @@ static int kvm_arm_pmu_v3_init(struct kvm_vcpu *vcpu)
 	return 0;
 }
 
-static bool irq_is_valid(struct kvm *kvm, int irq, bool is_ppi)
+#define irq_is_ppi(irq) ((irq) >= VGIC_NR_SGIS && (irq) < VGIC_NR_PRIVATE_IRQS)
+
+/*
+ * For one VM the interrupt type must be same for each vcpu.
+ * As a PPI, the interrupt number is the same for all vcpus,
+ * while as an SPI it must be a separate number per vcpu.
+ */
+static bool pmu_irq_is_valid(struct kvm *kvm, int irq)
 {
 	int i;
 	struct kvm_vcpu *vcpu;
@@ -445,7 +452,7 @@ static bool irq_is_valid(struct kvm *kvm, int irq, bool is_ppi)
 		if (!kvm_arm_pmu_irq_initialized(vcpu))
 			continue;
 
-		if (is_ppi) {
+		if (irq_is_ppi(irq)) {
 			if (vcpu->arch.pmu.irq_num != irq)
 				return false;
 		} else {
@@ -457,7 +464,6 @@ static bool irq_is_valid(struct kvm *kvm, int irq, bool is_ppi)
 	return true;
 }
 
-
 int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
 {
 	switch (attr->attr) {
@@ -471,14 +477,11 @@ int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
 		if (get_user(irq, uaddr))
 			return -EFAULT;
 
-		/*
-		 * The PMU overflow interrupt could be a PPI or SPI, but for one
-		 * VM the interrupt type must be same for each vcpu. As a PPI,
-		 * the interrupt number is the same for all vcpus, while as an
-		 * SPI it must be a separate number per vcpu.
-		 */
-		if (irq < VGIC_NR_SGIS || irq >= vcpu->kvm->arch.vgic.nr_irqs ||
-		    !irq_is_valid(vcpu->kvm, irq, irq < VGIC_NR_PRIVATE_IRQS))
+		/* The PMU overflow interrupt can be a PPI or a valid SPI. */
+		if (!(irq_is_ppi(irq) || vgic_valid_spi(vcpu->kvm, irq)))
+			return -EINVAL;
+
+		if (!pmu_irq_is_valid(vcpu->kvm, irq))
 			return -EINVAL;
 
 		if (kvm_arm_pmu_irq_initialized(vcpu))

virt/kvm/arm/vgic-v2.c

@@ -171,7 +171,7 @@ static const struct vgic_ops vgic_v2_ops = {
 	.enable			= vgic_v2_enable,
 };
 
-static struct vgic_params vgic_v2_params;
+struct vgic_params __section(.hyp.text) vgic_v2_params;
 
 static void vgic_cpu_init_lrs(void *params)
 {
@@ -201,6 +201,8 @@ int vgic_v2_probe(const struct gic_kvm_info *gic_kvm_info,
 	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;

virt/kvm/arm/vgic-v3.c

@@ -29,12 +29,6 @@
 #include <asm/kvm_asm.h>
 #include <asm/kvm_mmu.h>
 
-/* These are for GICv2 emulation only */
-#define GICH_LR_VIRTUALID		(0x3ffUL << 0)
-#define GICH_LR_PHYSID_CPUID_SHIFT	(10)
-#define GICH_LR_PHYSID_CPUID		(7UL << GICH_LR_PHYSID_CPUID_SHIFT)
-#define ICH_LR_VIRTUALID_MASK		(BIT_ULL(32) - 1)
-
 static u32 ich_vtr_el2;
 
 static struct vgic_lr vgic_v3_get_lr(const struct kvm_vcpu *vcpu, int lr)
@@ -43,7 +37,7 @@ static struct vgic_lr vgic_v3_get_lr(const struct kvm_vcpu *vcpu, int lr)
 	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_VIRTUALID_MASK;
+		lr_desc.irq = val & ICH_LR_VIRTUAL_ID_MASK;
 	else
 		lr_desc.irq = val & GICH_LR_VIRTUALID;
 

virt/kvm/arm/vgic.c

@@ -690,12 +690,11 @@ bool vgic_handle_cfg_reg(u32 *reg, struct kvm_exit_mmio *mmio,
  */
 void vgic_unqueue_irqs(struct kvm_vcpu *vcpu)
 {
-	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
 	u64 elrsr = vgic_get_elrsr(vcpu);
 	unsigned long *elrsr_ptr = u64_to_bitmask(&elrsr);
 	int i;
 
-	for_each_clear_bit(i, elrsr_ptr, vgic_cpu->nr_lr) {
+	for_each_clear_bit(i, elrsr_ptr, vgic->nr_lr) {
 		struct vgic_lr lr = vgic_get_lr(vcpu, i);
 
 		/*
@@ -820,7 +819,6 @@ static int vgic_handle_mmio_access(struct kvm_vcpu *vcpu,
 	struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
 	struct vgic_io_device *iodev = container_of(this,
 						    struct vgic_io_device, dev);
-	struct kvm_run *run = vcpu->run;
 	const struct vgic_io_range *range;
 	struct kvm_exit_mmio mmio;
 	bool updated_state;
@@ -849,12 +847,6 @@ static int vgic_handle_mmio_access(struct kvm_vcpu *vcpu,
 		updated_state = false;
 	}
 	spin_unlock(&dist->lock);
-	run->mmio.is_write	= is_write;
-	run->mmio.len		= len;
-	run->mmio.phys_addr	= addr;
-	memcpy(run->mmio.data, val, len);
-
-	kvm_handle_mmio_return(vcpu, run);
 
 	if (updated_state)
 		vgic_kick_vcpus(vcpu->kvm);
@@ -1102,18 +1094,18 @@ static bool dist_active_irq(struct kvm_vcpu *vcpu)
 	return test_bit(vcpu->vcpu_id, dist->irq_active_on_cpu);
 }
 
-bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, struct irq_phys_map *map)
+bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq)
 {
 	int i;
 
-	for (i = 0; i < vcpu->arch.vgic_cpu.nr_lr; i++) {
+	for (i = 0; i < vgic->nr_lr; i++) {
 		struct vgic_lr vlr = vgic_get_lr(vcpu, i);
 
-		if (vlr.irq == map->virt_irq && vlr.state & LR_STATE_ACTIVE)
+		if (vlr.irq == virt_irq && vlr.state & LR_STATE_ACTIVE)
 			return true;
 	}
 
-	return vgic_irq_is_active(vcpu, map->virt_irq);
+	return vgic_irq_is_active(vcpu, virt_irq);
 }
 
 /*
@@ -1521,7 +1513,6 @@ static int vgic_validate_injection(struct kvm_vcpu *vcpu, int irq, int level)
 }
 
 static int vgic_update_irq_pending(struct kvm *kvm, int cpuid,
-				   struct irq_phys_map *map,
 				   unsigned int irq_num, bool level)
 {
 	struct vgic_dist *dist = &kvm->arch.vgic;
@@ -1660,14 +1651,14 @@ int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int irq_num,
 	if (map)
 		return -EINVAL;
 
-	return vgic_update_irq_pending(kvm, cpuid, NULL, irq_num, level);
+	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
- * @map:     Pointer to a irq_phys_map structure describing the mapping
+ * @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
@@ -1678,7 +1669,7 @@ int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int irq_num,
  * being HIGH and 0 being LOW and all devices being active-HIGH.
  */
 int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid,
-			       struct irq_phys_map *map, bool level)
+			       unsigned int virt_irq, bool level)
 {
 	int ret;
 
@@ -1686,7 +1677,7 @@ int kvm_vgic_inject_mapped_irq(struct kvm *kvm, int cpuid,
 	if (ret)
 		return ret;
 
-	return vgic_update_irq_pending(kvm, cpuid, map, map->virt_irq, level);
+	return vgic_update_irq_pending(kvm, cpuid, virt_irq, level);
 }
 
 static irqreturn_t vgic_maintenance_handler(int irq, void *data)
@@ -1712,43 +1703,28 @@ static struct list_head *vgic_get_irq_phys_map_list(struct kvm_vcpu *vcpu,
 /**
  * kvm_vgic_map_phys_irq - map a virtual IRQ to a physical IRQ
  * @vcpu: The VCPU pointer
- * @virt_irq: The virtual irq number
- * @irq: The Linux IRQ number
+ * @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
- * Linux irq (@irq). On injection, @virt_irq will be associated with
- * the physical interrupt represented by @irq. This mapping can be
+ * 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 a valid pointer on success, and an error pointer otherwise
+ * Returns 0 on success or an error value otherwise.
  */
-struct irq_phys_map *kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu,
-					   int virt_irq, int irq)
+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;
-	struct irq_desc *desc;
-	struct irq_data *data;
-	int phys_irq;
-
-	desc = irq_to_desc(irq);
-	if (!desc) {
-		kvm_err("%s: no interrupt descriptor\n", __func__);
-		return ERR_PTR(-EINVAL);
-	}
-
-	data = irq_desc_get_irq_data(desc);
-	while (data->parent_data)
-		data = data->parent_data;
-
-	phys_irq = data->hwirq;
+	int ret = 0;
 
 	/* Create a new mapping */
 	entry = kzalloc(sizeof(*entry), GFP_KERNEL);
 	if (!entry)
-		return ERR_PTR(-ENOMEM);
+		return -ENOMEM;
 
 	spin_lock(&dist->irq_phys_map_lock);
 
@@ -1756,9 +1732,8 @@ struct irq_phys_map *kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu,
 	map = vgic_irq_map_search(vcpu, virt_irq);
 	if (map) {
 		/* Make sure this mapping matches */
-		if (map->phys_irq != phys_irq	||
-		    map->irq      != irq)
-			map = ERR_PTR(-EINVAL);
+		if (map->phys_irq != phys_irq)
+			ret = -EINVAL;
 
 		/* Found an existing, valid mapping */
 		goto out;
@@ -1767,7 +1742,6 @@ struct irq_phys_map *kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu,
 	map           = &entry->map;
 	map->virt_irq = virt_irq;
 	map->phys_irq = phys_irq;
-	map->irq      = irq;
 
 	list_add_tail_rcu(&entry->entry, root);
 
@@ -1775,9 +1749,9 @@ struct irq_phys_map *kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu,
 	spin_unlock(&dist->irq_phys_map_lock);
 	/* If we've found a hit in the existing list, free the useless
 	 * entry */
-	if (IS_ERR(map) || map != &entry->map)
+	if (ret || map != &entry->map)
 		kfree(entry);
-	return map;
+	return ret;
 }
 
 static struct irq_phys_map *vgic_irq_map_search(struct kvm_vcpu *vcpu,
@@ -1813,25 +1787,22 @@ static void vgic_free_phys_irq_map_rcu(struct rcu_head *rcu)
 /**
  * kvm_vgic_unmap_phys_irq - Remove a virtual to physical IRQ mapping
  * @vcpu: The VCPU pointer
- * @map: The pointer to a mapping obtained through kvm_vgic_map_phys_irq
+ * @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, struct irq_phys_map *map)
+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;
 
-	if (!map)
-		return -EINVAL;
-
-	root = vgic_get_irq_phys_map_list(vcpu, map->virt_irq);
+	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 == map) {
+		if (entry->map.virt_irq == virt_irq) {
 			list_del_rcu(&entry->entry);
 			call_rcu(&entry->rcu, vgic_free_phys_irq_map_rcu);
 			break;
@@ -1887,13 +1858,6 @@ static int vgic_vcpu_init_maps(struct kvm_vcpu *vcpu, int nr_irqs)
 		return -ENOMEM;
 	}
 
-	/*
-	 * Store the number of LRs per vcpu, so we don't have to go
-	 * all the way to the distributor structure to find out. Only
-	 * assembly code should use this one.
-	 */
-	vgic_cpu->nr_lr = vgic->nr_lr;
-
 	return 0;
 }
 

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

+/*
+ * 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/>.
+ */
+
+#include <linux/kvm.h>
+#include <linux/kvm_host.h>
+#include <trace/events/kvm.h>
+
+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 int irqchip,
+			 unsigned int 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/vgic-mmio.h

+/*
+ * 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 __KVM_ARM_VGIC_MMIO_H__
+#define __KVM_ARM_VGIC_MMIO_H__
+
+struct vgic_register_region {
+	unsigned int reg_offset;
+	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);
+};
+
+extern struct kvm_io_device_ops kvm_io_gic_ops;
+
+#define VGIC_ACCESS_8bit	1
+#define VGIC_ACCESS_32bit	2
+#define VGIC_ACCESS_64bit	4
+
+/*
+ * Generate a mask that covers the number of bytes required to address
+ * up to 1024 interrupts, each represented by <bits> bits. This assumes
+ * that <bits> is a power of two.
+ */
+#define VGIC_ADDR_IRQ_MASK(bits) (((bits) * 1024 / 8) - 1)
+
+/*
+ * (addr & mask) gives us the byte offset for the INT ID, so we want to
+ * divide this with 'bytes per irq' to get the INT ID, which is given
+ * by '(bits) / 8'.  But we do this with fixed-point-arithmetic and
+ * take advantage of the fact that division by a fraction equals
+ * multiplication with the inverted fraction, and scale up both the
+ * numerator and denominator with 8 to support at most 64 bits per IRQ:
+ */
+#define VGIC_ADDR_TO_INTID(addr, bits)  (((addr) & VGIC_ADDR_IRQ_MASK(bits)) * \
+					64 / (bits) / 8)
+
+/*
+ * Some VGIC registers store per-IRQ information, with a different number
+ * of bits per IRQ. For those registers this macro is used.
+ * The _WITH_LENGTH version instantiates registers with a fixed length
+ * and is mutually exclusive with the _PER_IRQ version.
+ */
+#define REGISTER_DESC_WITH_BITS_PER_IRQ(off, rd, wr, bpi, acc)		\
+	{								\
+		.reg_offset = off,					\
+		.bits_per_irq = bpi,					\
+		.len = bpi * 1024 / 8,					\
+		.access_flags = acc,					\
+		.read = rd,						\
+		.write = wr,						\
+	}
+
+#define REGISTER_DESC_WITH_LENGTH(off, rd, wr, length, acc)		\
+	{								\
+		.reg_offset = off,					\
+		.bits_per_irq = 0,					\
+		.len = length,						\
+		.access_flags = acc,					\
+		.read = rd,						\
+		.write = wr,						\
+	}
+
+int kvm_vgic_register_mmio_region(struct kvm *kvm, struct kvm_vcpu *vcpu,
+				  struct vgic_register_region *reg_desc,
+				  struct vgic_io_device *region,
+				  int nr_irqs, bool offset_private);
+
+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 vgic_mmio_read_raz(struct kvm_vcpu *vcpu,
+				 gpa_t addr, unsigned int len);
+
+unsigned long vgic_mmio_read_rao(struct kvm_vcpu *vcpu,
+				 gpa_t addr, unsigned int len);
+
+void vgic_mmio_write_wi(struct kvm_vcpu *vcpu, gpa_t addr,
+			unsigned int len, unsigned long val);
+
+unsigned long vgic_mmio_read_enable(struct kvm_vcpu *vcpu,
+				    gpa_t addr, unsigned int len);
+
+void vgic_mmio_write_senable(struct kvm_vcpu *vcpu,
+			     gpa_t addr, unsigned int len,
+			     unsigned long val);
+
+void vgic_mmio_write_cenable(struct kvm_vcpu *vcpu,
+			     gpa_t addr, unsigned int len,
+			     unsigned long val);
+
+unsigned long vgic_mmio_read_pending(struct kvm_vcpu *vcpu,
+				     gpa_t addr, unsigned int len);
+
+void vgic_mmio_write_spending(struct kvm_vcpu *vcpu,
+			      gpa_t addr, unsigned int len,
+			      unsigned long val);
+
+void vgic_mmio_write_cpending(struct kvm_vcpu *vcpu,
+			      gpa_t addr, unsigned int len,
+			      unsigned long val);
+
+unsigned long vgic_mmio_read_active(struct kvm_vcpu *vcpu,
+				    gpa_t addr, unsigned int len);
+
+void vgic_mmio_write_cactive(struct kvm_vcpu *vcpu,
+			     gpa_t addr, unsigned int len,
+			     unsigned long val);
+
+void vgic_mmio_write_sactive(struct kvm_vcpu *vcpu,
+			     gpa_t addr, unsigned int len,
+			     unsigned long val);
+
+unsigned long vgic_mmio_read_priority(struct kvm_vcpu *vcpu,
+				      gpa_t addr, unsigned int len);
+
+void vgic_mmio_write_priority(struct kvm_vcpu *vcpu,
+			      gpa_t addr, unsigned int len,
+			      unsigned long val);
+
+unsigned long vgic_mmio_read_config(struct kvm_vcpu *vcpu,
+				    gpa_t addr, unsigned int len);
+
+void vgic_mmio_write_config(struct kvm_vcpu *vcpu,
+			    gpa_t addr, unsigned int len,
+			    unsigned long val);
+
+unsigned int vgic_v2_init_dist_iodev(struct vgic_io_device *dev);
+
+unsigned int vgic_v3_init_dist_iodev(struct vgic_io_device *dev);
+
+#endif

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

+/*
+ * 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/>.
+ */
+
+#include <linux/irqchip/arm-gic.h>
+#include <linux/kvm.h>
+#include <linux/kvm_host.h>
+#include <kvm/arm_vgic.h>
+#include <asm/kvm_mmu.h>
+
+#include "vgic.h"
+
+/*
+ * 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;
+}
+
+void vgic_v2_process_maintenance(struct kvm_vcpu *vcpu)
+{
+	struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2;
+
+	if (cpuif->vgic_misr & GICH_MISR_EOI) {
+		u64 eisr = cpuif->vgic_eisr;
+		unsigned long *eisr_bmap = u64_to_bitmask(&eisr);
+		int lr;
+
+		for_each_set_bit(lr, eisr_bmap, kvm_vgic_global_state.nr_lr) {
+			u32 intid = cpuif->vgic_lr[lr] & GICH_LR_VIRTUALID;
+
+			WARN_ON(cpuif->vgic_lr[lr] & GICH_LR_STATE);
+
+			kvm_notify_acked_irq(vcpu->kvm, 0,
+					     intid - VGIC_NR_PRIVATE_IRQS);
+		}
+	}
+
+	/* check and disable underflow maintenance IRQ */
+	cpuif->vgic_hcr &= ~GICH_HCR_UIE;
+
+	/*
+	 * 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.
+	 */
+	cpuif->vgic_eisr = 0;
+}
+
+void vgic_v2_set_underflow(struct kvm_vcpu *vcpu)
+{
+	struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2;
+
+	cpuif->vgic_hcr |= GICH_HCR_UIE;
+}
+
+/*
+ * transfer the content of the LRs back into the corresponding ap_list:
+ * - active bit is transferred as is
+ * - pending bit is
+ *   - transferred as is in case of edge sensitive IRQs
+ *   - set to the line-level (resample time) for level sensitive IRQs
+ */
+void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
+{
+	struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2;
+	int lr;
+
+	for (lr = 0; lr < vcpu->arch.vgic_cpu.used_lrs; lr++) {
+		u32 val = cpuif->vgic_lr[lr];
+		u32 intid = val & GICH_LR_VIRTUALID;
+		struct vgic_irq *irq;
+
+		irq = vgic_get_irq(vcpu->kvm, vcpu, intid);
+
+		spin_lock(&irq->irq_lock);
+
+		/* Always preserve the active bit */
+		irq->active = !!(val & GICH_LR_ACTIVE_BIT);
+
+		/* Edge is the only case where we preserve the pending bit */
+		if (irq->config == VGIC_CONFIG_EDGE &&
+		    (val & GICH_LR_PENDING_BIT)) {
+			irq->pending = true;
+
+			if (vgic_irq_is_sgi(intid)) {
+				u32 cpuid = val & GICH_LR_PHYSID_CPUID;
+
+				cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT;
+				irq->source |= (1 << cpuid);
+			}
+		}
+
+		/* Clear soft pending state when level IRQs have been acked */
+		if (irq->config == VGIC_CONFIG_LEVEL &&
+		    !(val & GICH_LR_PENDING_BIT)) {
+			irq->soft_pending = false;
+			irq->pending = irq->line_level;
+		}
+
+		spin_unlock(&irq->irq_lock);
+	}
+}
+
+/*
+ * Populates the particular LR with the state of a given IRQ:
+ * - for an edge sensitive IRQ the pending state is cleared in struct vgic_irq
+ * - for a level sensitive IRQ the pending state value is unchanged;
+ *   it is dictated directly by the input level
+ *
+ * If @irq describes an SGI with multiple sources, we choose the
+ * lowest-numbered source VCPU and clear that bit in the source bitmap.
+ *
+ * The irq_lock must be held by the caller.
+ */
+void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
+{
+	u32 val = irq->intid;
+
+	if (irq->pending) {
+		val |= GICH_LR_PENDING_BIT;
+
+		if (irq->config == VGIC_CONFIG_EDGE)
+			irq->pending = false;
+
+		if (vgic_irq_is_sgi(irq->intid)) {
+			u32 src = ffs(irq->source);
+
+			BUG_ON(!src);
+			val |= (src - 1) << GICH_LR_PHYSID_CPUID_SHIFT;
+			irq->source &= ~(1 << (src - 1));
+			if (irq->source)
+				irq->pending = true;
+		}
+	}
+
+	if (irq->active)
+		val |= GICH_LR_ACTIVE_BIT;
+
+	if (irq->hw) {
+		val |= GICH_LR_HW;
+		val |= irq->hwintid << GICH_LR_PHYSID_CPUID_SHIFT;
+	} else {
+		if (irq->config == VGIC_CONFIG_LEVEL)
+			val |= GICH_LR_EOI;
+	}
+
+	/* The GICv2 LR only holds five bits of priority. */
+	val |= (irq->priority >> 3) << GICH_LR_PRIORITY_SHIFT;
+
+	vcpu->arch.vgic_cpu.vgic_v2.vgic_lr[lr] = val;
+}
+
+void vgic_v2_clear_lr(struct kvm_vcpu *vcpu, int lr)
+{
+	vcpu->arch.vgic_cpu.vgic_v2.vgic_lr[lr] = 0;
+}
+
+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;
+}
+
+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;
+}
+
+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;
+}
+
+/* check for overlapping regions and for regions crossing the end of memory */
+static bool vgic_v2_check_base(gpa_t dist_base, gpa_t cpu_base)
+{
+	if (dist_base + KVM_VGIC_V2_DIST_SIZE < dist_base)
+		return false;
+	if (cpu_base + KVM_VGIC_V2_CPU_SIZE < cpu_base)
+		return false;
+
+	if (dist_base + KVM_VGIC_V2_DIST_SIZE <= cpu_base)
+		return true;
+	if (cpu_base + KVM_VGIC_V2_CPU_SIZE <= dist_base)
+		return true;
+
+	return false;
+}
+
+int vgic_v2_map_resources(struct kvm *kvm)
+{
+	struct vgic_dist *dist = &kvm->arch.vgic;
+	int ret = 0;
+
+	if (vgic_ready(kvm))
+		goto out;
+
+	if (IS_VGIC_ADDR_UNDEF(dist->vgic_dist_base) ||
+	    IS_VGIC_ADDR_UNDEF(dist->vgic_cpu_base)) {
+		kvm_err("Need to set vgic cpu and dist addresses first\n");
+		ret = -ENXIO;
+		goto out;
+	}
+
+	if (!vgic_v2_check_base(dist->vgic_dist_base, dist->vgic_cpu_base)) {
+		kvm_err("VGIC CPU and dist frames overlap\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/*
+	 * 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 initialize VGIC dynamic data structures\n");
+		goto out;
+	}
+
+	ret = vgic_register_dist_iodev(kvm, dist->vgic_dist_base, VGIC_V2);
+	if (ret) {
+		kvm_err("Unable to register VGIC MMIO regions\n");
+		goto out;
+	}
+
+	ret = kvm_phys_addr_ioremap(kvm, dist->vgic_cpu_base,
+				    kvm_vgic_global_state.vcpu_base,
+				    KVM_VGIC_V2_CPU_SIZE, true);
+	if (ret) {
+		kvm_err("Unable to remap VGIC CPU to VCPU\n");
+		goto out;
+	}
+
+	dist->ready = true;
+
+out:
+	if (ret)
+		kvm_vgic_destroy(kvm);
+	return ret;
+}
+
+/**
+ * vgic_v2_probe - probe for a GICv2 compatible interrupt controller in DT
+ * @node:	pointer to the DT node
+ *
+ * Returns 0 if a GICv2 has been found, returns an error code otherwise
+ */
+int vgic_v2_probe(const struct gic_kvm_info *info)
+{
+	int ret;
+	u32 vtr;
+
+	if (!info->vctrl.start) {
+		kvm_err("GICH not present in the firmware table\n");
+		return -ENXIO;
+	}
+
+	if (!PAGE_ALIGNED(info->vcpu.start)) {
+		kvm_err("GICV physical address 0x%llx not page aligned\n",
+			(unsigned long long)info->vcpu.start);
+		return -ENXIO;
+	}
+
+	if (!PAGE_ALIGNED(resource_size(&info->vcpu))) {
+		kvm_err("GICV size 0x%llx not a multiple of page size 0x%lx\n",
+			(unsigned long long)resource_size(&info->vcpu),
+			PAGE_SIZE);
+		return -ENXIO;
+	}
+
+	kvm_vgic_global_state.vctrl_base = ioremap(info->vctrl.start,
+						   resource_size(&info->vctrl));
+	if (!kvm_vgic_global_state.vctrl_base) {
+		kvm_err("Cannot ioremap GICH\n");
+		return -ENOMEM;
+	}
+
+	vtr = readl_relaxed(kvm_vgic_global_state.vctrl_base + GICH_VTR);
+	kvm_vgic_global_state.nr_lr = (vtr & 0x3f) + 1;
+
+	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");
+		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;
+
+	kvm_info("vgic-v2@%llx\n", info->vctrl.start);
+
+	return 0;
+}

virt/kvm/arm/vgic/vgic-v3.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/>.
+ */
+
+#include <linux/irqchip/arm-gic-v3.h>
+#include <linux/kvm.h>
+#include <linux/kvm_host.h>
+#include <kvm/arm_vgic.h>
+#include <asm/kvm_mmu.h>
+#include <asm/kvm_asm.h>
+
+#include "vgic.h"
+
+void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu)
+{
+	struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3;
+	u32 model = vcpu->kvm->arch.vgic.vgic_model;
+
+	if (cpuif->vgic_misr & ICH_MISR_EOI) {
+		unsigned long eisr_bmap = cpuif->vgic_eisr;
+		int lr;
+
+		for_each_set_bit(lr, &eisr_bmap, kvm_vgic_global_state.nr_lr) {
+			u32 intid;
+			u64 val = cpuif->vgic_lr[lr];
+
+			if (model == KVM_DEV_TYPE_ARM_VGIC_V3)
+				intid = val & ICH_LR_VIRTUAL_ID_MASK;
+			else
+				intid = val & GICH_LR_VIRTUALID;
+
+			WARN_ON(cpuif->vgic_lr[lr] & ICH_LR_STATE);
+
+			kvm_notify_acked_irq(vcpu->kvm, 0,
+					     intid - VGIC_NR_PRIVATE_IRQS);
+		}
+
+		/*
+		 * 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.
+		 */
+		cpuif->vgic_eisr = 0;
+	}
+
+	cpuif->vgic_hcr &= ~ICH_HCR_UIE;
+}
+
+void vgic_v3_set_underflow(struct kvm_vcpu *vcpu)
+{
+	struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3;
+
+	cpuif->vgic_hcr |= ICH_HCR_UIE;
+}
+
+void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
+{
+	struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3;
+	u32 model = vcpu->kvm->arch.vgic.vgic_model;
+	int lr;
+
+	for (lr = 0; lr < vcpu->arch.vgic_cpu.used_lrs; lr++) {
+		u64 val = cpuif->vgic_lr[lr];
+		u32 intid;
+		struct vgic_irq *irq;
+
+		if (model == KVM_DEV_TYPE_ARM_VGIC_V3)
+			intid = val & ICH_LR_VIRTUAL_ID_MASK;
+		else
+			intid = val & GICH_LR_VIRTUALID;
+		irq = vgic_get_irq(vcpu->kvm, vcpu, intid);
+
+		spin_lock(&irq->irq_lock);
+
+		/* Always preserve the active bit */
+		irq->active = !!(val & ICH_LR_ACTIVE_BIT);
+
+		/* Edge is the only case where we preserve the pending bit */
+		if (irq->config == VGIC_CONFIG_EDGE &&
+		    (val & ICH_LR_PENDING_BIT)) {
+			irq->pending = true;
+
+			if (vgic_irq_is_sgi(intid) &&
+			    model == KVM_DEV_TYPE_ARM_VGIC_V2) {
+				u32 cpuid = val & GICH_LR_PHYSID_CPUID;
+
+				cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT;
+				irq->source |= (1 << cpuid);
+			}
+		}
+
+		/* Clear soft pending state when level irqs have been acked */
+		if (irq->config == VGIC_CONFIG_LEVEL &&
+		    !(val & ICH_LR_PENDING_BIT)) {
+			irq->soft_pending = false;
+			irq->pending = irq->line_level;
+		}
+
+		spin_unlock(&irq->irq_lock);
+	}
+}
+
+/* Requires the irq to be locked already */
+void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
+{
+	u32 model = vcpu->kvm->arch.vgic.vgic_model;
+	u64 val = irq->intid;
+
+	if (irq->pending) {
+		val |= ICH_LR_PENDING_BIT;
+
+		if (irq->config == VGIC_CONFIG_EDGE)
+			irq->pending = false;
+
+		if (vgic_irq_is_sgi(irq->intid) &&
+		    model == KVM_DEV_TYPE_ARM_VGIC_V2) {
+			u32 src = ffs(irq->source);
+
+			BUG_ON(!src);
+			val |= (src - 1) << GICH_LR_PHYSID_CPUID_SHIFT;
+			irq->source &= ~(1 << (src - 1));
+			if (irq->source)
+				irq->pending = true;
+		}
+	}
+
+	if (irq->active)
+		val |= ICH_LR_ACTIVE_BIT;
+
+	if (irq->hw) {
+		val |= ICH_LR_HW;
+		val |= ((u64)irq->hwintid) << ICH_LR_PHYS_ID_SHIFT;
+	} else {
+		if (irq->config == VGIC_CONFIG_LEVEL)
+			val |= ICH_LR_EOI;
+	}
+
+	/*
+	 * We currently only support Group1 interrupts, which is a
+	 * known defect. This needs to be addressed at some point.
+	 */
+	if (model == KVM_DEV_TYPE_ARM_VGIC_V3)
+		val |= ICH_LR_GROUP;
+
+	val |= (u64)irq->priority << ICH_LR_PRIORITY_SHIFT;
+
+	vcpu->arch.vgic_cpu.vgic_v3.vgic_lr[lr] = val;
+}
+
+void vgic_v3_clear_lr(struct kvm_vcpu *vcpu, int lr)
+{
+	vcpu->arch.vgic_cpu.vgic_v3.vgic_lr[lr] = 0;
+}
+
+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;
+}
+
+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;
+}
+
+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;
+}
+
+/* check for overlapping regions and for regions crossing the end of memory */
+static bool vgic_v3_check_base(struct kvm *kvm)
+{
+	struct vgic_dist *d = &kvm->arch.vgic;
+	gpa_t redist_size = KVM_VGIC_V3_REDIST_SIZE;
+
+	redist_size *= atomic_read(&kvm->online_vcpus);
+
+	if (d->vgic_dist_base + KVM_VGIC_V3_DIST_SIZE < d->vgic_dist_base)
+		return false;
+	if (d->vgic_redist_base + redist_size < d->vgic_redist_base)
+		return false;
+
+	if (d->vgic_dist_base + KVM_VGIC_V3_DIST_SIZE <= d->vgic_redist_base)
+		return true;
+	if (d->vgic_redist_base + redist_size <= d->vgic_dist_base)
+		return true;
+
+	return false;
+}
+
+int vgic_v3_map_resources(struct kvm *kvm)
+{
+	int ret = 0;
+	struct vgic_dist *dist = &kvm->arch.vgic;
+
+	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;
+	}
+
+	if (!vgic_v3_check_base(kvm)) {
+		kvm_err("VGIC redist and dist frames overlap\n");
+		ret = -EINVAL;
+		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_dist_iodev(kvm, dist->vgic_dist_base, VGIC_V3);
+	if (ret) {
+		kvm_err("Unable to register VGICv3 dist MMIO regions\n");
+		goto out;
+	}
+
+	ret = vgic_register_redist_iodevs(kvm, dist->vgic_redist_base);
+	if (ret) {
+		kvm_err("Unable to register VGICv3 redist MMIO regions\n");
+		goto out;
+	}
+
+	dist->ready = true;
+
+out:
+	if (ret)
+		kvm_vgic_destroy(kvm);
+	return ret;
+}
+
+/**
+ * vgic_v3_probe - probe for a GICv3 compatible interrupt controller in DT
+ * @node:	pointer to the DT node
+ *
+ * Returns 0 if a GICv3 has been found, returns an error code otherwise
+ */
+int vgic_v3_probe(const struct gic_kvm_info *info)
+{
+	u32 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...
+	 */
+	kvm_vgic_global_state.nr_lr = (ich_vtr_el2 & 0xf) + 1;
+	kvm_vgic_global_state.can_emulate_gicv2 = false;
+
+	if (!info->vcpu.start) {
+		kvm_info("GICv3: no GICV resource entry\n");
+		kvm_vgic_global_state.vcpu_base = 0;
+	} else if (!PAGE_ALIGNED(info->vcpu.start)) {
+		pr_warn("GICV physical address 0x%llx not page aligned\n",
+			(unsigned long long)info->vcpu.start);
+		kvm_vgic_global_state.vcpu_base = 0;
+	} else if (!PAGE_ALIGNED(resource_size(&info->vcpu))) {
+		pr_warn("GICV size 0x%llx not a multiple of page size 0x%lx\n",
+			(unsigned long long)resource_size(&info->vcpu),
+			PAGE_SIZE);
+		kvm_vgic_global_state.vcpu_base = 0;
+	} 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);
+		kvm_info("vgic-v2@%llx\n", info->vcpu.start);
+	}
+	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;
+	kvm_vgic_global_state.max_gic_vcpus = VGIC_V3_MAX_CPUS;
+
+	return 0;
+}