[linux-2.6-block.git] / arch / powerpc / kvm / book3s.c

/*
 * Copyright (C) 2009. SUSE Linux Products GmbH. All rights reserved.
 *
 * Authors:
 *    Alexander Graf <agraf@suse.de>
 *    Kevin Wolf <mail@kevin-wolf.de>
 *
 * Description:
 * This file is derived from arch/powerpc/kvm/44x.c,
 * by Hollis Blanchard <hollisb@us.ibm.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.
 */

#include <linux/kvm_host.h>
#include <linux/err.h>
#include <linux/export.h>
#include <linux/slab.h>

#include <asm/reg.h>
#include <asm/cputable.h>
#include <asm/cacheflush.h>
#include <asm/tlbflush.h>
#include <asm/uaccess.h>
#include <asm/io.h>
#include <asm/kvm_ppc.h>
#include <asm/kvm_book3s.h>
#include <asm/mmu_context.h>
#include <asm/page.h>
#include <linux/gfp.h>
#include <linux/sched.h>
#include <linux/vmalloc.h>
#include <linux/highmem.h>

#include "trace.h"

#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU

/* #define EXIT_DEBUG */

struct kvm_stats_debugfs_item debugfs_entries[] = {
	{ "exits",       VCPU_STAT(sum_exits) },
	{ "mmio",        VCPU_STAT(mmio_exits) },
	{ "sig",         VCPU_STAT(signal_exits) },
	{ "sysc",        VCPU_STAT(syscall_exits) },
	{ "inst_emu",    VCPU_STAT(emulated_inst_exits) },
	{ "dec",         VCPU_STAT(dec_exits) },
	{ "ext_intr",    VCPU_STAT(ext_intr_exits) },
	{ "queue_intr",  VCPU_STAT(queue_intr) },
	{ "halt_wakeup", VCPU_STAT(halt_wakeup) },
	{ "pf_storage",  VCPU_STAT(pf_storage) },
	{ "sp_storage",  VCPU_STAT(sp_storage) },
	{ "pf_instruc",  VCPU_STAT(pf_instruc) },
	{ "sp_instruc",  VCPU_STAT(sp_instruc) },
	{ "ld",          VCPU_STAT(ld) },
	{ "ld_slow",     VCPU_STAT(ld_slow) },
	{ "st",          VCPU_STAT(st) },
	{ "st_slow",     VCPU_STAT(st_slow) },
	{ NULL }
};

void kvmppc_core_load_host_debugstate(struct kvm_vcpu *vcpu)
{
}

void kvmppc_core_load_guest_debugstate(struct kvm_vcpu *vcpu)
{
}

void kvmppc_inject_interrupt(struct kvm_vcpu *vcpu, int vec, u64 flags)
{
	vcpu->arch.shared->srr0 = kvmppc_get_pc(vcpu);
	vcpu->arch.shared->srr1 = vcpu->arch.shared->msr | flags;
	kvmppc_set_pc(vcpu, kvmppc_interrupt_offset(vcpu) + vec);
	vcpu->arch.mmu.reset_msr(vcpu);
}

static int kvmppc_book3s_vec2irqprio(unsigned int vec)
{
	unsigned int prio;

	switch (vec) {
	case 0x100: prio = BOOK3S_IRQPRIO_SYSTEM_RESET;		break;
	case 0x200: prio = BOOK3S_IRQPRIO_MACHINE_CHECK;	break;
	case 0x300: prio = BOOK3S_IRQPRIO_DATA_STORAGE;		break;
	case 0x380: prio = BOOK3S_IRQPRIO_DATA_SEGMENT;		break;
	case 0x400: prio = BOOK3S_IRQPRIO_INST_STORAGE;		break;
	case 0x480: prio = BOOK3S_IRQPRIO_INST_SEGMENT;		break;
	case 0x500: prio = BOOK3S_IRQPRIO_EXTERNAL;		break;
	case 0x501: prio = BOOK3S_IRQPRIO_EXTERNAL_LEVEL;	break;
	case 0x600: prio = BOOK3S_IRQPRIO_ALIGNMENT;		break;
	case 0x700: prio = BOOK3S_IRQPRIO_PROGRAM;		break;
	case 0x800: prio = BOOK3S_IRQPRIO_FP_UNAVAIL;		break;
	case 0x900: prio = BOOK3S_IRQPRIO_DECREMENTER;		break;
	case 0xc00: prio = BOOK3S_IRQPRIO_SYSCALL;		break;
	case 0xd00: prio = BOOK3S_IRQPRIO_DEBUG;		break;
	case 0xf20: prio = BOOK3S_IRQPRIO_ALTIVEC;		break;
	case 0xf40: prio = BOOK3S_IRQPRIO_VSX;			break;
	default:    prio = BOOK3S_IRQPRIO_MAX;			break;
	}

	return prio;
}

static void kvmppc_book3s_dequeue_irqprio(struct kvm_vcpu *vcpu,
					  unsigned int vec)
{
	unsigned long old_pending = vcpu->arch.pending_exceptions;

	clear_bit(kvmppc_book3s_vec2irqprio(vec),
		  &vcpu->arch.pending_exceptions);

	kvmppc_update_int_pending(vcpu, vcpu->arch.pending_exceptions,
				  old_pending);
}

void kvmppc_book3s_queue_irqprio(struct kvm_vcpu *vcpu, unsigned int vec)
{
	vcpu->stat.queue_intr++;

	set_bit(kvmppc_book3s_vec2irqprio(vec),
		&vcpu->arch.pending_exceptions);
#ifdef EXIT_DEBUG
	printk(KERN_INFO "Queueing interrupt %x\n", vec);
#endif
}


void kvmppc_core_queue_program(struct kvm_vcpu *vcpu, ulong flags)
{
	/* might as well deliver this straight away */
	kvmppc_inject_interrupt(vcpu, BOOK3S_INTERRUPT_PROGRAM, flags);
}

void kvmppc_core_queue_dec(struct kvm_vcpu *vcpu)
{
	kvmppc_book3s_queue_irqprio(vcpu, BOOK3S_INTERRUPT_DECREMENTER);
}

int kvmppc_core_pending_dec(struct kvm_vcpu *vcpu)
{
	return test_bit(BOOK3S_IRQPRIO_DECREMENTER, &vcpu->arch.pending_exceptions);
}

void kvmppc_core_dequeue_dec(struct kvm_vcpu *vcpu)
{
	kvmppc_book3s_dequeue_irqprio(vcpu, BOOK3S_INTERRUPT_DECREMENTER);
}

void kvmppc_core_queue_external(struct kvm_vcpu *vcpu,
                                struct kvm_interrupt *irq)
{
	unsigned int vec = BOOK3S_INTERRUPT_EXTERNAL;

	if (irq->irq == KVM_INTERRUPT_SET_LEVEL)
		vec = BOOK3S_INTERRUPT_EXTERNAL_LEVEL;

	kvmppc_book3s_queue_irqprio(vcpu, vec);
}

void kvmppc_core_dequeue_external(struct kvm_vcpu *vcpu,
                                  struct kvm_interrupt *irq)
{
	kvmppc_book3s_dequeue_irqprio(vcpu, BOOK3S_INTERRUPT_EXTERNAL);
	kvmppc_book3s_dequeue_irqprio(vcpu, BOOK3S_INTERRUPT_EXTERNAL_LEVEL);
}

int kvmppc_book3s_irqprio_deliver(struct kvm_vcpu *vcpu, unsigned int priority)
{
	int deliver = 1;
	int vec = 0;
	bool crit = kvmppc_critical_section(vcpu);

	switch (priority) {
	case BOOK3S_IRQPRIO_DECREMENTER:
		deliver = (vcpu->arch.shared->msr & MSR_EE) && !crit;
		vec = BOOK3S_INTERRUPT_DECREMENTER;
		break;
	case BOOK3S_IRQPRIO_EXTERNAL:
	case BOOK3S_IRQPRIO_EXTERNAL_LEVEL:
		deliver = (vcpu->arch.shared->msr & MSR_EE) && !crit;
		vec = BOOK3S_INTERRUPT_EXTERNAL;
		break;
	case BOOK3S_IRQPRIO_SYSTEM_RESET:
		vec = BOOK3S_INTERRUPT_SYSTEM_RESET;
		break;
	case BOOK3S_IRQPRIO_MACHINE_CHECK:
		vec = BOOK3S_INTERRUPT_MACHINE_CHECK;
		break;
	case BOOK3S_IRQPRIO_DATA_STORAGE:
		vec = BOOK3S_INTERRUPT_DATA_STORAGE;
		break;
	case BOOK3S_IRQPRIO_INST_STORAGE:
		vec = BOOK3S_INTERRUPT_INST_STORAGE;
		break;
	case BOOK3S_IRQPRIO_DATA_SEGMENT:
		vec = BOOK3S_INTERRUPT_DATA_SEGMENT;
		break;
	case BOOK3S_IRQPRIO_INST_SEGMENT:
		vec = BOOK3S_INTERRUPT_INST_SEGMENT;
		break;
	case BOOK3S_IRQPRIO_ALIGNMENT:
		vec = BOOK3S_INTERRUPT_ALIGNMENT;
		break;
	case BOOK3S_IRQPRIO_PROGRAM:
		vec = BOOK3S_INTERRUPT_PROGRAM;
		break;
	case BOOK3S_IRQPRIO_VSX:
		vec = BOOK3S_INTERRUPT_VSX;
		break;
	case BOOK3S_IRQPRIO_ALTIVEC:
		vec = BOOK3S_INTERRUPT_ALTIVEC;
		break;
	case BOOK3S_IRQPRIO_FP_UNAVAIL:
		vec = BOOK3S_INTERRUPT_FP_UNAVAIL;
		break;
	case BOOK3S_IRQPRIO_SYSCALL:
		vec = BOOK3S_INTERRUPT_SYSCALL;
		break;
	case BOOK3S_IRQPRIO_DEBUG:
		vec = BOOK3S_INTERRUPT_TRACE;
		break;
	case BOOK3S_IRQPRIO_PERFORMANCE_MONITOR:
		vec = BOOK3S_INTERRUPT_PERFMON;
		break;
	default:
		deliver = 0;
		printk(KERN_ERR "KVM: Unknown interrupt: 0x%x\n", priority);
		break;
	}

#if 0
	printk(KERN_INFO "Deliver interrupt 0x%x? %x\n", vec, deliver);
#endif

	if (deliver)
		kvmppc_inject_interrupt(vcpu, vec, 0);

	return deliver;
}

/*
 * This function determines if an irqprio should be cleared once issued.
 */
static bool clear_irqprio(struct kvm_vcpu *vcpu, unsigned int priority)
{
	switch (priority) {
		case BOOK3S_IRQPRIO_DECREMENTER:
			/* DEC interrupts get cleared by mtdec */
			return false;
		case BOOK3S_IRQPRIO_EXTERNAL_LEVEL:
			/* External interrupts get cleared by userspace */
			return false;
	}

	return true;
}

void kvmppc_core_prepare_to_enter(struct kvm_vcpu *vcpu)
{
	unsigned long *pending = &vcpu->arch.pending_exceptions;
	unsigned long old_pending = vcpu->arch.pending_exceptions;
	unsigned int priority;

#ifdef EXIT_DEBUG
	if (vcpu->arch.pending_exceptions)
		printk(KERN_EMERG "KVM: Check pending: %lx\n", vcpu->arch.pending_exceptions);
#endif
	priority = __ffs(*pending);
	while (priority < BOOK3S_IRQPRIO_MAX) {
		if (kvmppc_book3s_irqprio_deliver(vcpu, priority) &&
		    clear_irqprio(vcpu, priority)) {
			clear_bit(priority, &vcpu->arch.pending_exceptions);
			break;
		}

		priority = find_next_bit(pending,
					 BITS_PER_BYTE * sizeof(*pending),
					 priority + 1);
	}

	/* Tell the guest about our interrupt status */
	kvmppc_update_int_pending(vcpu, *pending, old_pending);
}

pfn_t kvmppc_gfn_to_pfn(struct kvm_vcpu *vcpu, gfn_t gfn)
{
	ulong mp_pa = vcpu->arch.magic_page_pa;

	/* Magic page override */
	if (unlikely(mp_pa) &&
	    unlikely(((gfn << PAGE_SHIFT) & KVM_PAM) ==
		     ((mp_pa & PAGE_MASK) & KVM_PAM))) {
		ulong shared_page = ((ulong)vcpu->arch.shared) & PAGE_MASK;
		pfn_t pfn;

		pfn = (pfn_t)virt_to_phys((void*)shared_page) >> PAGE_SHIFT;
		get_page(pfn_to_page(pfn));
		return pfn;
	}

	return gfn_to_pfn(vcpu->kvm, gfn);
}

static int kvmppc_xlate(struct kvm_vcpu *vcpu, ulong eaddr, bool data,
			 struct kvmppc_pte *pte)
{
	int relocated = (vcpu->arch.shared->msr & (data ? MSR_DR : MSR_IR));
	int r;

	if (relocated) {
		r = vcpu->arch.mmu.xlate(vcpu, eaddr, pte, data);
	} else {
		pte->eaddr = eaddr;
		pte->raddr = eaddr & KVM_PAM;
		pte->vpage = VSID_REAL | eaddr >> 12;
		pte->may_read = true;
		pte->may_write = true;
		pte->may_execute = true;
		r = 0;
	}

	return r;
}

static hva_t kvmppc_bad_hva(void)
{
	return PAGE_OFFSET;
}

static hva_t kvmppc_pte_to_hva(struct kvm_vcpu *vcpu, struct kvmppc_pte *pte,
			       bool read)
{
	hva_t hpage;

	if (read && !pte->may_read)
		goto err;

	if (!read && !pte->may_write)
		goto err;

	hpage = gfn_to_hva(vcpu->kvm, pte->raddr >> PAGE_SHIFT);
	if (kvm_is_error_hva(hpage))
		goto err;

	return hpage | (pte->raddr & ~PAGE_MASK);
err:
	return kvmppc_bad_hva();
}

int kvmppc_st(struct kvm_vcpu *vcpu, ulong *eaddr, int size, void *ptr,
	      bool data)
{
	struct kvmppc_pte pte;

	vcpu->stat.st++;

	if (kvmppc_xlate(vcpu, *eaddr, data, &pte))
		return -ENOENT;

	*eaddr = pte.raddr;

	if (!pte.may_write)
		return -EPERM;

	if (kvm_write_guest(vcpu->kvm, pte.raddr, ptr, size))
		return EMULATE_DO_MMIO;

	return EMULATE_DONE;
}

int kvmppc_ld(struct kvm_vcpu *vcpu, ulong *eaddr, int size, void *ptr,
		      bool data)
{
	struct kvmppc_pte pte;
	hva_t hva = *eaddr;

	vcpu->stat.ld++;

	if (kvmppc_xlate(vcpu, *eaddr, data, &pte))
		goto nopte;

	*eaddr = pte.raddr;

	hva = kvmppc_pte_to_hva(vcpu, &pte, true);
	if (kvm_is_error_hva(hva))
		goto mmio;

	if (copy_from_user(ptr, (void __user *)hva, size)) {
		printk(KERN_INFO "kvmppc_ld at 0x%lx failed\n", hva);
		goto mmio;
	}

	return EMULATE_DONE;

nopte:
	return -ENOENT;
mmio:
	return EMULATE_DO_MMIO;
}

int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
{
	return 0;
}

int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
	int i;

	regs->pc = kvmppc_get_pc(vcpu);
	regs->cr = kvmppc_get_cr(vcpu);
	regs->ctr = kvmppc_get_ctr(vcpu);
	regs->lr = kvmppc_get_lr(vcpu);
	regs->xer = kvmppc_get_xer(vcpu);
	regs->msr = vcpu->arch.shared->msr;
	regs->srr0 = vcpu->arch.shared->srr0;
	regs->srr1 = vcpu->arch.shared->srr1;
	regs->pid = vcpu->arch.pid;
	regs->sprg0 = vcpu->arch.shared->sprg0;
	regs->sprg1 = vcpu->arch.shared->sprg1;
	regs->sprg2 = vcpu->arch.shared->sprg2;
	regs->sprg3 = vcpu->arch.shared->sprg3;
	regs->sprg4 = vcpu->arch.shared->sprg4;
	regs->sprg5 = vcpu->arch.shared->sprg5;
	regs->sprg6 = vcpu->arch.shared->sprg6;
	regs->sprg7 = vcpu->arch.shared->sprg7;

	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);

	return 0;
}

int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
	int i;

	kvmppc_set_pc(vcpu, regs->pc);
	kvmppc_set_cr(vcpu, regs->cr);
	kvmppc_set_ctr(vcpu, regs->ctr);
	kvmppc_set_lr(vcpu, regs->lr);
	kvmppc_set_xer(vcpu, regs->xer);
	kvmppc_set_msr(vcpu, regs->msr);
	vcpu->arch.shared->srr0 = regs->srr0;
	vcpu->arch.shared->srr1 = regs->srr1;
	vcpu->arch.shared->sprg0 = regs->sprg0;
	vcpu->arch.shared->sprg1 = regs->sprg1;
	vcpu->arch.shared->sprg2 = regs->sprg2;
	vcpu->arch.shared->sprg3 = regs->sprg3;
	vcpu->arch.shared->sprg4 = regs->sprg4;
	vcpu->arch.shared->sprg5 = regs->sprg5;
	vcpu->arch.shared->sprg6 = regs->sprg6;
	vcpu->arch.shared->sprg7 = regs->sprg7;

	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);

	return 0;
}

int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
	return -ENOTSUPP;
}

int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
	return -ENOTSUPP;
}

int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
                                  struct kvm_translation *tr)
{
	return 0;
}

/*
 * Get (and clear) the dirty memory log for a memory slot.
 */
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				      struct kvm_dirty_log *log)
{
	struct kvm_memory_slot *memslot;
	struct kvm_vcpu *vcpu;
	ulong ga, ga_end;
	int is_dirty = 0;
	int r;
	unsigned long n;

	mutex_lock(&kvm->slots_lock);

	r = kvm_get_dirty_log(kvm, log, &is_dirty);
	if (r)
		goto out;

	/* If nothing is dirty, don't bother messing with page tables. */
	if (is_dirty) {
		memslot = id_to_memslot(kvm->memslots, log->slot);

		ga = memslot->base_gfn << PAGE_SHIFT;
		ga_end = ga + (memslot->npages << PAGE_SHIFT);

		kvm_for_each_vcpu(n, vcpu, kvm)
			kvmppc_mmu_pte_pflush(vcpu, ga, ga_end);

		n = kvm_dirty_bitmap_bytes(memslot);
		memset(memslot->dirty_bitmap, 0, n);
	}

	r = 0;
out:
	mutex_unlock(&kvm->slots_lock);
	return r;
}
Commit	Line	Data
2f4cf5e4 AG	1	/*
	2	* Copyright (C) 2009. SUSE Linux Products GmbH. All rights reserved.
	3	*
	4	* Authors:
	5	* Alexander Graf <agraf@suse.de>
	6	* Kevin Wolf <mail@kevin-wolf.de>
	7	*
	8	* Description:
	9	* This file is derived from arch/powerpc/kvm/44x.c,
	10	* by Hollis Blanchard <hollisb@us.ibm.com>.
	11	*
	12	* This program is free software; you can redistribute it and/or modify
	13	* it under the terms of the GNU General Public License, version 2, as
	14	* published by the Free Software Foundation.
	15	*/
	16
	17	#include <linux/kvm_host.h>
	18	#include <linux/err.h>
66b15db6	19	#include <linux/export.h>
329d20ba	20	#include <linux/slab.h>
2f4cf5e4 AG	21
	22	#include <asm/reg.h>
	23	#include <asm/cputable.h>
	24	#include <asm/cacheflush.h>
	25	#include <asm/tlbflush.h>
	26	#include <asm/uaccess.h>
	27	#include <asm/io.h>
	28	#include <asm/kvm_ppc.h>
	29	#include <asm/kvm_book3s.h>
	30	#include <asm/mmu_context.h>
149dbdb1	31	#include <asm/page.h>
5a0e3ad6	32	#include <linux/gfp.h>
2f4cf5e4 AG	33	#include <linux/sched.h>
2f4cf5e4 AG	34	#include <linux/vmalloc.h>
9fb244a2	35	#include <linux/highmem.h>
2f4cf5e4	36
c4befc58 PM	37	#include "trace.h"
c4befc58 PM	38
2f4cf5e4 AG	39	#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU
	40
	41	/* #define EXIT_DEBUG */
07b0907d	42
2f4cf5e4 AG	43	struct kvm_stats_debugfs_item debugfs_entries[] = {
	44	{ "exits", VCPU_STAT(sum_exits) },
	45	{ "mmio", VCPU_STAT(mmio_exits) },
	46	{ "sig", VCPU_STAT(signal_exits) },
	47	{ "sysc", VCPU_STAT(syscall_exits) },
	48	{ "inst_emu", VCPU_STAT(emulated_inst_exits) },
	49	{ "dec", VCPU_STAT(dec_exits) },
	50	{ "ext_intr", VCPU_STAT(ext_intr_exits) },
	51	{ "queue_intr", VCPU_STAT(queue_intr) },
	52	{ "halt_wakeup", VCPU_STAT(halt_wakeup) },
	53	{ "pf_storage", VCPU_STAT(pf_storage) },
	54	{ "sp_storage", VCPU_STAT(sp_storage) },
	55	{ "pf_instruc", VCPU_STAT(pf_instruc) },
	56	{ "sp_instruc", VCPU_STAT(sp_instruc) },
	57	{ "ld", VCPU_STAT(ld) },
	58	{ "ld_slow", VCPU_STAT(ld_slow) },
	59	{ "st", VCPU_STAT(st) },
	60	{ "st_slow", VCPU_STAT(st_slow) },
	61	{ NULL }
	62	};
	63
	64	void kvmppc_core_load_host_debugstate(struct kvm_vcpu *vcpu)
	65	{
	66	}
	67
	68	void kvmppc_core_load_guest_debugstate(struct kvm_vcpu *vcpu)
	69	{
	70	}
	71
2f4cf5e4 AG	72	void kvmppc_inject_interrupt(struct kvm_vcpu *vcpu, int vec, u64 flags)
2f4cf5e4 AG	73	{
de7906c3 AG	74	vcpu->arch.shared->srr0 = kvmppc_get_pc(vcpu);
de7906c3 AG	75	vcpu->arch.shared->srr1 = vcpu->arch.shared->msr \| flags;
f05ed4d5	76	kvmppc_set_pc(vcpu, kvmppc_interrupt_offset(vcpu) + vec);
2f4cf5e4 AG	77	vcpu->arch.mmu.reset_msr(vcpu);
	78	}
	79
583617b7	80	static int kvmppc_book3s_vec2irqprio(unsigned int vec)
2f4cf5e4 AG	81	{
	82	unsigned int prio;
	83
2f4cf5e4 AG	84	switch (vec) {
	85	case 0x100: prio = BOOK3S_IRQPRIO_SYSTEM_RESET; break;
	86	case 0x200: prio = BOOK3S_IRQPRIO_MACHINE_CHECK; break;
	87	case 0x300: prio = BOOK3S_IRQPRIO_DATA_STORAGE; break;
	88	case 0x380: prio = BOOK3S_IRQPRIO_DATA_SEGMENT; break;
	89	case 0x400: prio = BOOK3S_IRQPRIO_INST_STORAGE; break;
	90	case 0x480: prio = BOOK3S_IRQPRIO_INST_SEGMENT; break;
	91	case 0x500: prio = BOOK3S_IRQPRIO_EXTERNAL; break;
17bd1580	92	case 0x501: prio = BOOK3S_IRQPRIO_EXTERNAL_LEVEL; break;
2f4cf5e4 AG	93	case 0x600: prio = BOOK3S_IRQPRIO_ALIGNMENT; break;
	94	case 0x700: prio = BOOK3S_IRQPRIO_PROGRAM; break;
	95	case 0x800: prio = BOOK3S_IRQPRIO_FP_UNAVAIL; break;
	96	case 0x900: prio = BOOK3S_IRQPRIO_DECREMENTER; break;
	97	case 0xc00: prio = BOOK3S_IRQPRIO_SYSCALL; break;
	98	case 0xd00: prio = BOOK3S_IRQPRIO_DEBUG; break;
	99	case 0xf20: prio = BOOK3S_IRQPRIO_ALTIVEC; break;
	100	case 0xf40: prio = BOOK3S_IRQPRIO_VSX; break;
	101	default: prio = BOOK3S_IRQPRIO_MAX; break;
	102	}
	103
583617b7 AG	104	return prio;
	105	}
	106
7706664d AG	107	static void kvmppc_book3s_dequeue_irqprio(struct kvm_vcpu *vcpu,
	108	unsigned int vec)
	109	{
f05ed4d5 PM	110	unsigned long old_pending = vcpu->arch.pending_exceptions;
f05ed4d5 PM	111
7706664d AG	112	clear_bit(kvmppc_book3s_vec2irqprio(vec),
7706664d AG	113	&vcpu->arch.pending_exceptions);
9ee18b1e	114
f05ed4d5 PM	115	kvmppc_update_int_pending(vcpu, vcpu->arch.pending_exceptions,
f05ed4d5 PM	116	old_pending);
7706664d AG	117	}
7706664d AG	118
583617b7 AG	119	void kvmppc_book3s_queue_irqprio(struct kvm_vcpu *vcpu, unsigned int vec)
	120	{
	121	vcpu->stat.queue_intr++;
	122
	123	set_bit(kvmppc_book3s_vec2irqprio(vec),
	124	&vcpu->arch.pending_exceptions);
2f4cf5e4 AG	125	#ifdef EXIT_DEBUG
	126	printk(KERN_INFO "Queueing interrupt %x\n", vec);
	127	#endif
	128	}
	129
	130
25a8a02d	131	void kvmppc_core_queue_program(struct kvm_vcpu *vcpu, ulong flags)
2f4cf5e4	132	{
3cf658b6 PM	133	/* might as well deliver this straight away */
3cf658b6 PM	134	kvmppc_inject_interrupt(vcpu, BOOK3S_INTERRUPT_PROGRAM, flags);
2f4cf5e4 AG	135	}
	136
	137	void kvmppc_core_queue_dec(struct kvm_vcpu *vcpu)
	138	{
	139	kvmppc_book3s_queue_irqprio(vcpu, BOOK3S_INTERRUPT_DECREMENTER);
	140	}
	141
	142	int kvmppc_core_pending_dec(struct kvm_vcpu *vcpu)
	143	{
44075d95	144	return test_bit(BOOK3S_IRQPRIO_DECREMENTER, &vcpu->arch.pending_exceptions);
2f4cf5e4 AG	145	}
2f4cf5e4 AG	146
7706664d AG	147	void kvmppc_core_dequeue_dec(struct kvm_vcpu *vcpu)
	148	{
	149	kvmppc_book3s_dequeue_irqprio(vcpu, BOOK3S_INTERRUPT_DECREMENTER);
	150	}
	151
2f4cf5e4 AG	152	void kvmppc_core_queue_external(struct kvm_vcpu *vcpu,
	153	struct kvm_interrupt *irq)
	154	{
17bd1580 AG	155	unsigned int vec = BOOK3S_INTERRUPT_EXTERNAL;
	156
	157	if (irq->irq == KVM_INTERRUPT_SET_LEVEL)
	158	vec = BOOK3S_INTERRUPT_EXTERNAL_LEVEL;
	159
	160	kvmppc_book3s_queue_irqprio(vcpu, vec);
2f4cf5e4 AG	161	}
2f4cf5e4 AG	162
18978768 AG	163	void kvmppc_core_dequeue_external(struct kvm_vcpu *vcpu,
	164	struct kvm_interrupt *irq)
	165	{
	166	kvmppc_book3s_dequeue_irqprio(vcpu, BOOK3S_INTERRUPT_EXTERNAL);
17bd1580	167	kvmppc_book3s_dequeue_irqprio(vcpu, BOOK3S_INTERRUPT_EXTERNAL_LEVEL);
18978768 AG	168	}
18978768 AG	169
2f4cf5e4 AG	170	int kvmppc_book3s_irqprio_deliver(struct kvm_vcpu *vcpu, unsigned int priority)
	171	{
	172	int deliver = 1;
	173	int vec = 0;
f05ed4d5	174	bool crit = kvmppc_critical_section(vcpu);
2f4cf5e4 AG	175
	176	switch (priority) {
	177	case BOOK3S_IRQPRIO_DECREMENTER:
5c6cedf4	178	deliver = (vcpu->arch.shared->msr & MSR_EE) && !crit;
2f4cf5e4 AG	179	vec = BOOK3S_INTERRUPT_DECREMENTER;
	180	break;
	181	case BOOK3S_IRQPRIO_EXTERNAL:
17bd1580	182	case BOOK3S_IRQPRIO_EXTERNAL_LEVEL:
5c6cedf4	183	deliver = (vcpu->arch.shared->msr & MSR_EE) && !crit;
2f4cf5e4 AG	184	vec = BOOK3S_INTERRUPT_EXTERNAL;
	185	break;
	186	case BOOK3S_IRQPRIO_SYSTEM_RESET:
	187	vec = BOOK3S_INTERRUPT_SYSTEM_RESET;
	188	break;
	189	case BOOK3S_IRQPRIO_MACHINE_CHECK:
	190	vec = BOOK3S_INTERRUPT_MACHINE_CHECK;
	191	break;
	192	case BOOK3S_IRQPRIO_DATA_STORAGE:
	193	vec = BOOK3S_INTERRUPT_DATA_STORAGE;
	194	break;
	195	case BOOK3S_IRQPRIO_INST_STORAGE:
	196	vec = BOOK3S_INTERRUPT_INST_STORAGE;
	197	break;
	198	case BOOK3S_IRQPRIO_DATA_SEGMENT:
	199	vec = BOOK3S_INTERRUPT_DATA_SEGMENT;
	200	break;
	201	case BOOK3S_IRQPRIO_INST_SEGMENT:
	202	vec = BOOK3S_INTERRUPT_INST_SEGMENT;
	203	break;
	204	case BOOK3S_IRQPRIO_ALIGNMENT:
	205	vec = BOOK3S_INTERRUPT_ALIGNMENT;
	206	break;
	207	case BOOK3S_IRQPRIO_PROGRAM:
	208	vec = BOOK3S_INTERRUPT_PROGRAM;
	209	break;
	210	case BOOK3S_IRQPRIO_VSX:
	211	vec = BOOK3S_INTERRUPT_VSX;
	212	break;
	213	case BOOK3S_IRQPRIO_ALTIVEC:
	214	vec = BOOK3S_INTERRUPT_ALTIVEC;
	215	break;
	216	case BOOK3S_IRQPRIO_FP_UNAVAIL:
	217	vec = BOOK3S_INTERRUPT_FP_UNAVAIL;
	218	break;
	219	case BOOK3S_IRQPRIO_SYSCALL:
	220	vec = BOOK3S_INTERRUPT_SYSCALL;
	221	break;
	222	case BOOK3S_IRQPRIO_DEBUG:
	223	vec = BOOK3S_INTERRUPT_TRACE;
	224	break;
	225	case BOOK3S_IRQPRIO_PERFORMANCE_MONITOR:
	226	vec = BOOK3S_INTERRUPT_PERFMON;
	227	break;
	228	default:
	229	deliver = 0;
	230	printk(KERN_ERR "KVM: Unknown interrupt: 0x%x\n", priority);
	231	break;
	232	}
	233
	234	#if 0
	235	printk(KERN_INFO "Deliver interrupt 0x%x? %x\n", vec, deliver);
	236	#endif
	237
	238	if (deliver)
3cf658b6	239	kvmppc_inject_interrupt(vcpu, vec, 0);
2f4cf5e4 AG	240
	241	return deliver;
	242	}
	243
17bd1580 AG	244	/*
	245	* This function determines if an irqprio should be cleared once issued.
	246	*/
	247	static bool clear_irqprio(struct kvm_vcpu *vcpu, unsigned int priority)
	248	{
	249	switch (priority) {
	250	case BOOK3S_IRQPRIO_DECREMENTER:
	251	/* DEC interrupts get cleared by mtdec */
	252	return false;
	253	case BOOK3S_IRQPRIO_EXTERNAL_LEVEL:
	254	/* External interrupts get cleared by userspace */
	255	return false;
	256	}
	257
	258	return true;
	259	}
	260
7e28e60e	261	void kvmppc_core_prepare_to_enter(struct kvm_vcpu *vcpu)
2f4cf5e4 AG	262	{
2f4cf5e4 AG	263	unsigned long *pending = &vcpu->arch.pending_exceptions;
90bba358	264	unsigned long old_pending = vcpu->arch.pending_exceptions;
2f4cf5e4 AG	265	unsigned int priority;
2f4cf5e4 AG	266
2f4cf5e4 AG	267	#ifdef EXIT_DEBUG
	268	if (vcpu->arch.pending_exceptions)
	269	printk(KERN_EMERG "KVM: Check pending: %lx\n", vcpu->arch.pending_exceptions);
	270	#endif
	271	priority = __ffs(*pending);
ada7ba17	272	while (priority < BOOK3S_IRQPRIO_MAX) {
7706664d	273	if (kvmppc_book3s_irqprio_deliver(vcpu, priority) &&
17bd1580	274	clear_irqprio(vcpu, priority)) {
2f4cf5e4 AG	275	clear_bit(priority, &vcpu->arch.pending_exceptions);
	276	break;
	277	}
	278
	279	priority = find_next_bit(pending,
	280	BITS_PER_BYTE * sizeof(*pending),
	281	priority + 1);
	282	}
90bba358 AG	283
90bba358 AG	284	/* Tell the guest about our interrupt status */
f05ed4d5	285	kvmppc_update_int_pending(vcpu, *pending, old_pending);
2f4cf5e4 AG	286	}
2f4cf5e4 AG	287
e8508940 AG	288	pfn_t kvmppc_gfn_to_pfn(struct kvm_vcpu *vcpu, gfn_t gfn)
	289	{
	290	ulong mp_pa = vcpu->arch.magic_page_pa;
	291
	292	/* Magic page override */
	293	if (unlikely(mp_pa) &&
	294	unlikely(((gfn << PAGE_SHIFT) & KVM_PAM) ==
	295	((mp_pa & PAGE_MASK) & KVM_PAM))) {
	296	ulong shared_page = ((ulong)vcpu->arch.shared) & PAGE_MASK;
	297	pfn_t pfn;
	298
	299	pfn = (pfn_t)virt_to_phys((void*)shared_page) >> PAGE_SHIFT;
	300	get_page(pfn_to_page(pfn));
	301	return pfn;
	302	}
	303
	304	return gfn_to_pfn(vcpu->kvm, gfn);
	305	}
	306
2f4cf5e4 AG	307	static int kvmppc_xlate(struct kvm_vcpu *vcpu, ulong eaddr, bool data,
	308	struct kvmppc_pte *pte)
	309	{
666e7252	310	int relocated = (vcpu->arch.shared->msr & (data ? MSR_DR : MSR_IR));
2f4cf5e4 AG	311	int r;
	312
	313	if (relocated) {
	314	r = vcpu->arch.mmu.xlate(vcpu, eaddr, pte, data);
	315	} else {
	316	pte->eaddr = eaddr;
28e83b4f	317	pte->raddr = eaddr & KVM_PAM;
3eeafd7d	318	pte->vpage = VSID_REAL \| eaddr >> 12;
2f4cf5e4 AG	319	pte->may_read = true;
	320	pte->may_write = true;
	321	pte->may_execute = true;
	322	r = 0;
	323	}
	324
	325	return r;
	326	}
	327
	328	static hva_t kvmppc_bad_hva(void)
	329	{
	330	return PAGE_OFFSET;
	331	}
	332
	333	static hva_t kvmppc_pte_to_hva(struct kvm_vcpu vcpu, struct kvmppc_pte pte,
	334	bool read)
	335	{
	336	hva_t hpage;
	337
	338	if (read && !pte->may_read)
	339	goto err;
	340
	341	if (!read && !pte->may_write)
	342	goto err;
	343
	344	hpage = gfn_to_hva(vcpu->kvm, pte->raddr >> PAGE_SHIFT);
	345	if (kvm_is_error_hva(hpage))
	346	goto err;
	347
	348	return hpage \| (pte->raddr & ~PAGE_MASK);
	349	err:
	350	return kvmppc_bad_hva();
	351	}
	352
5467a97d AG	353	int kvmppc_st(struct kvm_vcpu vcpu, ulong eaddr, int size, void *ptr,
5467a97d AG	354	bool data)
2f4cf5e4 AG	355	{
2f4cf5e4 AG	356	struct kvmppc_pte pte;
2f4cf5e4 AG	357
	358	vcpu->stat.st++;
	359
5467a97d	360	if (kvmppc_xlate(vcpu, *eaddr, data, &pte))
9fb244a2	361	return -ENOENT;
5467a97d AG	362
5467a97d AG	363	*eaddr = pte.raddr;
2f4cf5e4	364
9fb244a2 AG	365	if (!pte.may_write)
9fb244a2 AG	366	return -EPERM;
2f4cf5e4	367
9fb244a2 AG	368	if (kvm_write_guest(vcpu->kvm, pte.raddr, ptr, size))
9fb244a2 AG	369	return EMULATE_DO_MMIO;
2f4cf5e4	370
5467a97d	371	return EMULATE_DONE;
2f4cf5e4 AG	372	}
2f4cf5e4 AG	373
5467a97d	374	int kvmppc_ld(struct kvm_vcpu vcpu, ulong eaddr, int size, void *ptr,
2f4cf5e4 AG	375	bool data)
	376	{
	377	struct kvmppc_pte pte;
5467a97d	378	hva_t hva = *eaddr;
2f4cf5e4 AG	379
	380	vcpu->stat.ld++;
	381
5467a97d AG	382	if (kvmppc_xlate(vcpu, *eaddr, data, &pte))
	383	goto nopte;
	384
	385	*eaddr = pte.raddr;
2f4cf5e4 AG	386
	387	hva = kvmppc_pte_to_hva(vcpu, &pte, true);
	388	if (kvm_is_error_hva(hva))
5467a97d	389	goto mmio;
2f4cf5e4 AG	390
	391	if (copy_from_user(ptr, (void __user *)hva, size)) {
	392	printk(KERN_INFO "kvmppc_ld at 0x%lx failed\n", hva);
5467a97d	393	goto mmio;
2f4cf5e4 AG	394	}
2f4cf5e4 AG	395
5467a97d	396	return EMULATE_DONE;
2f4cf5e4	397
5467a97d	398	nopte:
2f4cf5e4	399	return -ENOENT;
5467a97d AG	400	mmio:
5467a97d AG	401	return EMULATE_DO_MMIO;
2f4cf5e4 AG	402	}
2f4cf5e4 AG	403
2f4cf5e4 AG	404	int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
	405	{
	406	return 0;
	407	}
	408
	409	int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu vcpu, struct kvm_regs regs)
	410	{
	411	int i;
	412
c7f38f46	413	regs->pc = kvmppc_get_pc(vcpu);
992b5b29	414	regs->cr = kvmppc_get_cr(vcpu);
c7f38f46 AG	415	regs->ctr = kvmppc_get_ctr(vcpu);
c7f38f46 AG	416	regs->lr = kvmppc_get_lr(vcpu);
992b5b29	417	regs->xer = kvmppc_get_xer(vcpu);
666e7252	418	regs->msr = vcpu->arch.shared->msr;
de7906c3 AG	419	regs->srr0 = vcpu->arch.shared->srr0;
de7906c3 AG	420	regs->srr1 = vcpu->arch.shared->srr1;
2f4cf5e4	421	regs->pid = vcpu->arch.pid;
a73a9599 AG	422	regs->sprg0 = vcpu->arch.shared->sprg0;
	423	regs->sprg1 = vcpu->arch.shared->sprg1;
	424	regs->sprg2 = vcpu->arch.shared->sprg2;
	425	regs->sprg3 = vcpu->arch.shared->sprg3;
b5904972 SW	426	regs->sprg4 = vcpu->arch.shared->sprg4;
	427	regs->sprg5 = vcpu->arch.shared->sprg5;
	428	regs->sprg6 = vcpu->arch.shared->sprg6;
	429	regs->sprg7 = vcpu->arch.shared->sprg7;
2f4cf5e4 AG	430
2f4cf5e4 AG	431	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
8e5b26b5	432	regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
2f4cf5e4 AG	433
	434	return 0;
	435	}
	436
	437	int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu vcpu, struct kvm_regs regs)
	438	{
	439	int i;
	440
c7f38f46	441	kvmppc_set_pc(vcpu, regs->pc);
992b5b29	442	kvmppc_set_cr(vcpu, regs->cr);
c7f38f46 AG	443	kvmppc_set_ctr(vcpu, regs->ctr);
c7f38f46 AG	444	kvmppc_set_lr(vcpu, regs->lr);
992b5b29	445	kvmppc_set_xer(vcpu, regs->xer);
2f4cf5e4	446	kvmppc_set_msr(vcpu, regs->msr);
de7906c3 AG	447	vcpu->arch.shared->srr0 = regs->srr0;
de7906c3 AG	448	vcpu->arch.shared->srr1 = regs->srr1;
a73a9599 AG	449	vcpu->arch.shared->sprg0 = regs->sprg0;
	450	vcpu->arch.shared->sprg1 = regs->sprg1;
	451	vcpu->arch.shared->sprg2 = regs->sprg2;
	452	vcpu->arch.shared->sprg3 = regs->sprg3;
b5904972 SW	453	vcpu->arch.shared->sprg4 = regs->sprg4;
	454	vcpu->arch.shared->sprg5 = regs->sprg5;
	455	vcpu->arch.shared->sprg6 = regs->sprg6;
	456	vcpu->arch.shared->sprg7 = regs->sprg7;
2f4cf5e4	457
8e5b26b5 AG	458	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
8e5b26b5 AG	459	kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
2f4cf5e4 AG	460
	461	return 0;
	462	}
	463
2f4cf5e4 AG	464	int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu vcpu, struct kvm_fpu fpu)
	465	{
	466	return -ENOTSUPP;
	467	}
	468
	469	int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu vcpu, struct kvm_fpu fpu)
	470	{
	471	return -ENOTSUPP;
	472	}
	473
	474	int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
	475	struct kvm_translation *tr)
	476	{
	477	return 0;
	478	}
	479
	480	/*
	481	* Get (and clear) the dirty memory log for a memory slot.
	482	*/
	483	int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
	484	struct kvm_dirty_log *log)
	485	{
	486	struct kvm_memory_slot *memslot;
	487	struct kvm_vcpu *vcpu;
	488	ulong ga, ga_end;
	489	int is_dirty = 0;
87bf6e7d TY	490	int r;
87bf6e7d TY	491	unsigned long n;
2f4cf5e4	492
79fac95e	493	mutex_lock(&kvm->slots_lock);
2f4cf5e4 AG	494
	495	r = kvm_get_dirty_log(kvm, log, &is_dirty);
	496	if (r)
	497	goto out;
	498
	499	/* If nothing is dirty, don't bother messing with page tables. */
	500	if (is_dirty) {
28a37544	501	memslot = id_to_memslot(kvm->memslots, log->slot);
2f4cf5e4 AG	502
	503	ga = memslot->base_gfn << PAGE_SHIFT;
	504	ga_end = ga + (memslot->npages << PAGE_SHIFT);
	505
	506	kvm_for_each_vcpu(n, vcpu, kvm)
	507	kvmppc_mmu_pte_pflush(vcpu, ga, ga_end);
	508
87bf6e7d	509	n = kvm_dirty_bitmap_bytes(memslot);
2f4cf5e4 AG	510	memset(memslot->dirty_bitmap, 0, n);
	511	}
	512
	513	r = 0;
	514	out:
79fac95e	515	mutex_unlock(&kvm->slots_lock);
2f4cf5e4 AG	516	return r;
2f4cf5e4 AG	517	}