[deliverable/linux.git] / arch / powerpc / kvm / powerpc.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, write to the Free Software
 * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 *
 * Copyright IBM Corp. 2007
 *
 * Authors: Hollis Blanchard <hollisb@us.ibm.com>
 *          Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
 */

#include <linux/errno.h>
#include <linux/err.h>
#include <linux/kvm_host.h>
#include <linux/module.h>
#include <linux/vmalloc.h>
#include <linux/fs.h>
#include <asm/cputable.h>
#include <asm/uaccess.h>
#include <asm/kvm_ppc.h>
#include <asm/tlbflush.h>


gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
{
	return gfn;
}

int kvm_cpu_has_interrupt(struct kvm_vcpu *v)
{
	return !!(v->arch.pending_exceptions);
}

int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
{
	return !(v->arch.msr & MSR_WE);
}


int kvmppc_emulate_mmio(struct kvm_run *run, struct kvm_vcpu *vcpu)
{
	enum emulation_result er;
	int r;

	er = kvmppc_emulate_instruction(run, vcpu);
	switch (er) {
	case EMULATE_DONE:
		/* Future optimization: only reload non-volatiles if they were
		 * actually modified. */
		r = RESUME_GUEST_NV;
		break;
	case EMULATE_DO_MMIO:
		run->exit_reason = KVM_EXIT_MMIO;
		/* We must reload nonvolatiles because "update" load/store
		 * instructions modify register state. */
		/* Future optimization: only reload non-volatiles if they were
		 * actually modified. */
		r = RESUME_HOST_NV;
		break;
	case EMULATE_FAIL:
		/* XXX Deliver Program interrupt to guest. */
		printk(KERN_EMERG "%s: emulation failed (%08x)\n", __func__,
		       vcpu->arch.last_inst);
		r = RESUME_HOST;
		break;
	default:
		BUG();
	}

	return r;
}

void kvm_arch_hardware_enable(void *garbage)
{
}

void kvm_arch_hardware_disable(void *garbage)
{
}

int kvm_arch_hardware_setup(void)
{
	return 0;
}

void kvm_arch_hardware_unsetup(void)
{
}

void kvm_arch_check_processor_compat(void *rtn)
{
	int r;

	if (strcmp(cur_cpu_spec->platform, "ppc440") == 0)
		r = 0;
	else
		r = -ENOTSUPP;

	*(int *)rtn = r;
}

struct kvm *kvm_arch_create_vm(void)
{
	struct kvm *kvm;

	kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL);
	if (!kvm)
		return ERR_PTR(-ENOMEM);

	return kvm;
}

static void kvmppc_free_vcpus(struct kvm *kvm)
{
	unsigned int i;

	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
		if (kvm->vcpus[i]) {
			kvm_arch_vcpu_free(kvm->vcpus[i]);
			kvm->vcpus[i] = NULL;
		}
	}
}

void kvm_arch_destroy_vm(struct kvm *kvm)
{
	kvmppc_free_vcpus(kvm);
	kvm_free_physmem(kvm);
	kfree(kvm);
}

int kvm_dev_ioctl_check_extension(long ext)
{
	int r;

	switch (ext) {
	case KVM_CAP_USER_MEMORY:
		r = 1;
		break;
	case KVM_CAP_COALESCED_MMIO:
		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
		break;
	default:
		r = 0;
		break;
	}
	return r;

}

long kvm_arch_dev_ioctl(struct file *filp,
                        unsigned int ioctl, unsigned long arg)
{
	return -EINVAL;
}

int kvm_arch_set_memory_region(struct kvm *kvm,
                               struct kvm_userspace_memory_region *mem,
                               struct kvm_memory_slot old,
                               int user_alloc)
{
	return 0;
}

void kvm_arch_flush_shadow(struct kvm *kvm)
{
}

struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
{
	struct kvm_vcpu *vcpu;
	int err;

	vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
	if (!vcpu) {
		err = -ENOMEM;
		goto out;
	}

	err = kvm_vcpu_init(vcpu, kvm, id);
	if (err)
		goto free_vcpu;

	return vcpu;

free_vcpu:
	kmem_cache_free(kvm_vcpu_cache, vcpu);
out:
	return ERR_PTR(err);
}

void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
	kvm_vcpu_uninit(vcpu);
	kmem_cache_free(kvm_vcpu_cache, vcpu);
}

void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
{
	kvm_arch_vcpu_free(vcpu);
}

int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
{
	unsigned int priority = exception_priority[BOOKE_INTERRUPT_DECREMENTER];

	return test_bit(priority, &vcpu->arch.pending_exceptions);
}

static void kvmppc_decrementer_func(unsigned long data)
{
	struct kvm_vcpu *vcpu = (struct kvm_vcpu *)data;

	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_DECREMENTER);

	if (waitqueue_active(&vcpu->wq)) {
		wake_up_interruptible(&vcpu->wq);
		vcpu->stat.halt_wakeup++;
	}
}

int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
	setup_timer(&vcpu->arch.dec_timer, kvmppc_decrementer_func,
	            (unsigned long)vcpu);

	return 0;
}

void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
	kvmppc_core_destroy_mmu(vcpu);
}

/* Note: clearing MSR[DE] just means that the debug interrupt will not be
 * delivered *immediately*. Instead, it simply sets the appropriate DBSR bits.
 * If those DBSR bits are still set when MSR[DE] is re-enabled, the interrupt
 * will be delivered as an "imprecise debug event" (which is indicated by
 * DBSR[IDE].
 */
static void kvmppc_disable_debug_interrupts(void)
{
	mtmsr(mfmsr() & ~MSR_DE);
}

static void kvmppc_restore_host_debug_state(struct kvm_vcpu *vcpu)
{
	kvmppc_disable_debug_interrupts();

	mtspr(SPRN_IAC1, vcpu->arch.host_iac[0]);
	mtspr(SPRN_IAC2, vcpu->arch.host_iac[1]);
	mtspr(SPRN_IAC3, vcpu->arch.host_iac[2]);
	mtspr(SPRN_IAC4, vcpu->arch.host_iac[3]);
	mtspr(SPRN_DBCR1, vcpu->arch.host_dbcr1);
	mtspr(SPRN_DBCR2, vcpu->arch.host_dbcr2);
	mtspr(SPRN_DBCR0, vcpu->arch.host_dbcr0);
	mtmsr(vcpu->arch.host_msr);
}

static void kvmppc_load_guest_debug_registers(struct kvm_vcpu *vcpu)
{
	struct kvm_guest_debug *dbg = &vcpu->guest_debug;
	u32 dbcr0 = 0;

	vcpu->arch.host_msr = mfmsr();
	kvmppc_disable_debug_interrupts();

	/* Save host debug register state. */
	vcpu->arch.host_iac[0] = mfspr(SPRN_IAC1);
	vcpu->arch.host_iac[1] = mfspr(SPRN_IAC2);
	vcpu->arch.host_iac[2] = mfspr(SPRN_IAC3);
	vcpu->arch.host_iac[3] = mfspr(SPRN_IAC4);
	vcpu->arch.host_dbcr0 = mfspr(SPRN_DBCR0);
	vcpu->arch.host_dbcr1 = mfspr(SPRN_DBCR1);
	vcpu->arch.host_dbcr2 = mfspr(SPRN_DBCR2);

	/* set registers up for guest */

	if (dbg->bp[0]) {
		mtspr(SPRN_IAC1, dbg->bp[0]);
		dbcr0 |= DBCR0_IAC1 | DBCR0_IDM;
	}
	if (dbg->bp[1]) {
		mtspr(SPRN_IAC2, dbg->bp[1]);
		dbcr0 |= DBCR0_IAC2 | DBCR0_IDM;
	}
	if (dbg->bp[2]) {
		mtspr(SPRN_IAC3, dbg->bp[2]);
		dbcr0 |= DBCR0_IAC3 | DBCR0_IDM;
	}
	if (dbg->bp[3]) {
		mtspr(SPRN_IAC4, dbg->bp[3]);
		dbcr0 |= DBCR0_IAC4 | DBCR0_IDM;
	}

	mtspr(SPRN_DBCR0, dbcr0);
	mtspr(SPRN_DBCR1, 0);
	mtspr(SPRN_DBCR2, 0);
}

void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
	int i;

	if (vcpu->guest_debug.enabled)
		kvmppc_load_guest_debug_registers(vcpu);

	/* Mark every guest entry in the shadow TLB entry modified, so that they
	 * will all be reloaded on the next vcpu run (instead of being
	 * demand-faulted). */
	for (i = 0; i <= tlb_44x_hwater; i++)
		kvmppc_tlbe_set_modified(vcpu, i);
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
	if (vcpu->guest_debug.enabled)
		kvmppc_restore_host_debug_state(vcpu);

	/* Don't leave guest TLB entries resident when being de-scheduled. */
	/* XXX It would be nice to differentiate between heavyweight exit and
	 * sched_out here, since we could avoid the TLB flush for heavyweight
	 * exits. */
	_tlbia();
}

int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
                                    struct kvm_debug_guest *dbg)
{
	int i;

	vcpu->guest_debug.enabled = dbg->enabled;
	if (vcpu->guest_debug.enabled) {
		for (i=0; i < ARRAY_SIZE(vcpu->guest_debug.bp); i++) {
			if (dbg->breakpoints[i].enabled)
				vcpu->guest_debug.bp[i] = dbg->breakpoints[i].address;
			else
				vcpu->guest_debug.bp[i] = 0;
		}
	}

	return 0;
}

static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,
                                     struct kvm_run *run)
{
	u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
	*gpr = run->dcr.data;
}

static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
                                      struct kvm_run *run)
{
	u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];

	if (run->mmio.len > sizeof(*gpr)) {
		printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
		return;
	}

	if (vcpu->arch.mmio_is_bigendian) {
		switch (run->mmio.len) {
		case 4: *gpr = *(u32 *)run->mmio.data; break;
		case 2: *gpr = *(u16 *)run->mmio.data; break;
		case 1: *gpr = *(u8 *)run->mmio.data; break;
		}
	} else {
		/* Convert BE data from userland back to LE. */
		switch (run->mmio.len) {
		case 4: *gpr = ld_le32((u32 *)run->mmio.data); break;
		case 2: *gpr = ld_le16((u16 *)run->mmio.data); break;
		case 1: *gpr = *(u8 *)run->mmio.data; break;
		}
	}
}

int kvmppc_handle_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
                       unsigned int rt, unsigned int bytes, int is_bigendian)
{
	if (bytes > sizeof(run->mmio.data)) {
		printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__,
		       run->mmio.len);
	}

	run->mmio.phys_addr = vcpu->arch.paddr_accessed;
	run->mmio.len = bytes;
	run->mmio.is_write = 0;

	vcpu->arch.io_gpr = rt;
	vcpu->arch.mmio_is_bigendian = is_bigendian;
	vcpu->mmio_needed = 1;
	vcpu->mmio_is_write = 0;

	return EMULATE_DO_MMIO;
}

int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
                        u32 val, unsigned int bytes, int is_bigendian)
{
	void *data = run->mmio.data;

	if (bytes > sizeof(run->mmio.data)) {
		printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__,
		       run->mmio.len);
	}

	run->mmio.phys_addr = vcpu->arch.paddr_accessed;
	run->mmio.len = bytes;
	run->mmio.is_write = 1;
	vcpu->mmio_needed = 1;
	vcpu->mmio_is_write = 1;

	/* Store the value at the lowest bytes in 'data'. */
	if (is_bigendian) {
		switch (bytes) {
		case 4: *(u32 *)data = val; break;
		case 2: *(u16 *)data = val; break;
		case 1: *(u8  *)data = val; break;
		}
	} else {
		/* Store LE value into 'data'. */
		switch (bytes) {
		case 4: st_le32(data, val); break;
		case 2: st_le16(data, val); break;
		case 1: *(u8 *)data = val; break;
		}
	}

	return EMULATE_DO_MMIO;
}

int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
{
	int r;
	sigset_t sigsaved;

	vcpu_load(vcpu);

	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);

	if (vcpu->mmio_needed) {
		if (!vcpu->mmio_is_write)
			kvmppc_complete_mmio_load(vcpu, run);
		vcpu->mmio_needed = 0;
	} else if (vcpu->arch.dcr_needed) {
		if (!vcpu->arch.dcr_is_write)
			kvmppc_complete_dcr_load(vcpu, run);
		vcpu->arch.dcr_needed = 0;
	}

	kvmppc_check_and_deliver_interrupts(vcpu);

	local_irq_disable();
	kvm_guest_enter();
	r = __kvmppc_vcpu_run(run, vcpu);
	kvm_guest_exit();
	local_irq_enable();

	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &sigsaved, NULL);

	vcpu_put(vcpu);

	return r;
}

int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
{
	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_EXTERNAL);

	if (waitqueue_active(&vcpu->wq)) {
		wake_up_interruptible(&vcpu->wq);
		vcpu->stat.halt_wakeup++;
	}

	return 0;
}

int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
                                    struct kvm_mp_state *mp_state)
{
	return -EINVAL;
}

int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
                                    struct kvm_mp_state *mp_state)
{
	return -EINVAL;
}

long kvm_arch_vcpu_ioctl(struct file *filp,
                         unsigned int ioctl, unsigned long arg)
{
	struct kvm_vcpu *vcpu = filp->private_data;
	void __user *argp = (void __user *)arg;
	long r;

	switch (ioctl) {
	case KVM_INTERRUPT: {
		struct kvm_interrupt irq;
		r = -EFAULT;
		if (copy_from_user(&irq, argp, sizeof(irq)))
			goto out;
		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
		break;
	}
	default:
		r = -EINVAL;
	}

out:
	return r;
}

int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm, struct kvm_dirty_log *log)
{
	return -ENOTSUPP;
}

long kvm_arch_vm_ioctl(struct file *filp,
                       unsigned int ioctl, unsigned long arg)
{
	long r;

	switch (ioctl) {
	default:
		r = -EINVAL;
	}

	return r;
}

int kvm_arch_init(void *opaque)
{
	return 0;
}

void kvm_arch_exit(void)
{
}
Commit	Line	Data
bbf45ba5 HB	1	/*
	2	* This program is free software; you can redistribute it and/or modify
	3	* it under the terms of the GNU General Public License, version 2, as
	4	* published by the Free Software Foundation.
	5	*
	6	* This program is distributed in the hope that it will be useful,
	7	* but WITHOUT ANY WARRANTY; without even the implied warranty of
	8	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	9	* GNU General Public License for more details.
	10	*
	11	* You should have received a copy of the GNU General Public License
	12	* along with this program; if not, write to the Free Software
	13	* Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
	14	*
	15	* Copyright IBM Corp. 2007
	16	*
	17	* Authors: Hollis Blanchard <hollisb@us.ibm.com>
	18	* Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
	19	*/
	20
	21	#include <linux/errno.h>
	22	#include <linux/err.h>
	23	#include <linux/kvm_host.h>
	24	#include <linux/module.h>
	25	#include <linux/vmalloc.h>
	26	#include <linux/fs.h>
	27	#include <asm/cputable.h>
	28	#include <asm/uaccess.h>
	29	#include <asm/kvm_ppc.h>
83aae4a8	30	#include <asm/tlbflush.h>
bbf45ba5 HB	31
	32
	33	gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
	34	{
	35	return gfn;
	36	}
	37
	38	int kvm_cpu_has_interrupt(struct kvm_vcpu *v)
	39	{
45c5eb67	40	return !!(v->arch.pending_exceptions);
bbf45ba5 HB	41	}
	42
	43	int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
	44	{
45c5eb67	45	return !(v->arch.msr & MSR_WE);
bbf45ba5 HB	46	}
	47
	48
	49	int kvmppc_emulate_mmio(struct kvm_run run, struct kvm_vcpu vcpu)
	50	{
	51	enum emulation_result er;
	52	int r;
	53
	54	er = kvmppc_emulate_instruction(run, vcpu);
	55	switch (er) {
	56	case EMULATE_DONE:
	57	/* Future optimization: only reload non-volatiles if they were
	58	* actually modified. */
	59	r = RESUME_GUEST_NV;
	60	break;
	61	case EMULATE_DO_MMIO:
	62	run->exit_reason = KVM_EXIT_MMIO;
	63	/* We must reload nonvolatiles because "update" load/store
	64	* instructions modify register state. */
	65	/* Future optimization: only reload non-volatiles if they were
	66	* actually modified. */
	67	r = RESUME_HOST_NV;
	68	break;
	69	case EMULATE_FAIL:
	70	/* XXX Deliver Program interrupt to guest. */
	71	printk(KERN_EMERG "%s: emulation failed (%08x)\n", __func__,
	72	vcpu->arch.last_inst);
	73	r = RESUME_HOST;
	74	break;
	75	default:
	76	BUG();
	77	}
	78
	79	return r;
	80	}
	81
	82	void kvm_arch_hardware_enable(void *garbage)
	83	{
	84	}
	85
	86	void kvm_arch_hardware_disable(void *garbage)
	87	{
	88	}
	89
	90	int kvm_arch_hardware_setup(void)
	91	{
	92	return 0;
	93	}
	94
	95	void kvm_arch_hardware_unsetup(void)
	96	{
	97	}
	98
	99	void kvm_arch_check_processor_compat(void *rtn)
	100	{
	101	int r;
	102
	103	if (strcmp(cur_cpu_spec->platform, "ppc440") == 0)
	104	r = 0;
	105	else
	106	r = -ENOTSUPP;
	107
	108	(int )rtn = r;
	109	}
110
111	struct kvm *kvm_arch_create_vm(void)
112	{
113	struct kvm *kvm;
114
115	kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL);
116	if (!kvm)
117	return ERR_PTR(-ENOMEM);
118
119	return kvm;
120	}
121
122	static void kvmppc_free_vcpus(struct kvm *kvm)
123	{
124	unsigned int i;
125
126	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
127	if (kvm->vcpus[i]) {
128	kvm_arch_vcpu_free(kvm->vcpus[i]);
129	kvm->vcpus[i] = NULL;
130	}
131	}
132	}
133
134	void kvm_arch_destroy_vm(struct kvm *kvm)
135	{
136	kvmppc_free_vcpus(kvm);
137	kvm_free_physmem(kvm);
138	kfree(kvm);
139	}
140
141	int kvm_dev_ioctl_check_extension(long ext)
142	{
143	int r;
144
145	switch (ext) {
146	case KVM_CAP_USER_MEMORY:
147	r = 1;
148	break;
588968b6 LV	149	case KVM_CAP_COALESCED_MMIO:
	150	r = KVM_COALESCED_MMIO_PAGE_OFFSET;
	151	break;
bbf45ba5 HB	152	default:
	153	r = 0;
	154	break;
	155	}
	156	return r;
	157
	158	}
	159
	160	long kvm_arch_dev_ioctl(struct file *filp,
	161	unsigned int ioctl, unsigned long arg)
	162	{
	163	return -EINVAL;
	164	}
	165
	166	int kvm_arch_set_memory_region(struct kvm *kvm,
	167	struct kvm_userspace_memory_region *mem,
	168	struct kvm_memory_slot old,
	169	int user_alloc)
	170	{
	171	return 0;
	172	}
	173
34d4cb8f MT	174	void kvm_arch_flush_shadow(struct kvm *kvm)
	175	{
	176	}
	177
bbf45ba5 HB	178	struct kvm_vcpu kvm_arch_vcpu_create(struct kvm kvm, unsigned int id)
	179	{
	180	struct kvm_vcpu *vcpu;
	181	int err;
	182
	183	vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
	184	if (!vcpu) {
	185	err = -ENOMEM;
	186	goto out;
	187	}
	188
	189	err = kvm_vcpu_init(vcpu, kvm, id);
	190	if (err)
	191	goto free_vcpu;
	192
	193	return vcpu;
	194
	195	free_vcpu:
	196	kmem_cache_free(kvm_vcpu_cache, vcpu);
	197	out:
	198	return ERR_PTR(err);
	199	}
	200
	201	void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
	202	{
	203	kvm_vcpu_uninit(vcpu);
	204	kmem_cache_free(kvm_vcpu_cache, vcpu);
	205	}
	206
	207	void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
	208	{
	209	kvm_arch_vcpu_free(vcpu);
	210	}
	211
	212	int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
	213	{
	214	unsigned int priority = exception_priority[BOOKE_INTERRUPT_DECREMENTER];
	215
	216	return test_bit(priority, &vcpu->arch.pending_exceptions);
	217	}
	218
	219	static void kvmppc_decrementer_func(unsigned long data)
	220	{
	221	struct kvm_vcpu vcpu = (struct kvm_vcpu )data;
	222
	223	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_DECREMENTER);
45c5eb67 HB	224
	225	if (waitqueue_active(&vcpu->wq)) {
	226	wake_up_interruptible(&vcpu->wq);
	227	vcpu->stat.halt_wakeup++;
	228	}
bbf45ba5 HB	229	}
	230
	231	int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
	232	{
	233	setup_timer(&vcpu->arch.dec_timer, kvmppc_decrementer_func,
	234	(unsigned long)vcpu);
	235
	236	return 0;
	237	}
	238
	239	void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
	240	{
c30f8a6c	241	kvmppc_core_destroy_mmu(vcpu);
bbf45ba5 HB	242	}
bbf45ba5 HB	243
6a0ab738 HB	244	/* Note: clearing MSR[DE] just means that the debug interrupt will not be
	245	* delivered immediately. Instead, it simply sets the appropriate DBSR bits.
	246	* If those DBSR bits are still set when MSR[DE] is re-enabled, the interrupt
	247	* will be delivered as an "imprecise debug event" (which is indicated by
	248	* DBSR[IDE].
	249	*/
	250	static void kvmppc_disable_debug_interrupts(void)
	251	{
	252	mtmsr(mfmsr() & ~MSR_DE);
	253	}
	254
	255	static void kvmppc_restore_host_debug_state(struct kvm_vcpu *vcpu)
	256	{
	257	kvmppc_disable_debug_interrupts();
	258
	259	mtspr(SPRN_IAC1, vcpu->arch.host_iac[0]);
	260	mtspr(SPRN_IAC2, vcpu->arch.host_iac[1]);
	261	mtspr(SPRN_IAC3, vcpu->arch.host_iac[2]);
	262	mtspr(SPRN_IAC4, vcpu->arch.host_iac[3]);
	263	mtspr(SPRN_DBCR1, vcpu->arch.host_dbcr1);
	264	mtspr(SPRN_DBCR2, vcpu->arch.host_dbcr2);
	265	mtspr(SPRN_DBCR0, vcpu->arch.host_dbcr0);
	266	mtmsr(vcpu->arch.host_msr);
	267	}
	268
	269	static void kvmppc_load_guest_debug_registers(struct kvm_vcpu *vcpu)
	270	{
	271	struct kvm_guest_debug *dbg = &vcpu->guest_debug;
	272	u32 dbcr0 = 0;
	273
	274	vcpu->arch.host_msr = mfmsr();
	275	kvmppc_disable_debug_interrupts();
	276
	277	/* Save host debug register state. */
	278	vcpu->arch.host_iac[0] = mfspr(SPRN_IAC1);
	279	vcpu->arch.host_iac[1] = mfspr(SPRN_IAC2);
	280	vcpu->arch.host_iac[2] = mfspr(SPRN_IAC3);
	281	vcpu->arch.host_iac[3] = mfspr(SPRN_IAC4);
	282	vcpu->arch.host_dbcr0 = mfspr(SPRN_DBCR0);
	283	vcpu->arch.host_dbcr1 = mfspr(SPRN_DBCR1);
	284	vcpu->arch.host_dbcr2 = mfspr(SPRN_DBCR2);
	285
	286	/* set registers up for guest */
	287
	288	if (dbg->bp[0]) {
	289	mtspr(SPRN_IAC1, dbg->bp[0]);
	290	dbcr0 \|= DBCR0_IAC1 \| DBCR0_IDM;
	291	}
	292	if (dbg->bp[1]) {
	293	mtspr(SPRN_IAC2, dbg->bp[1]);
	294	dbcr0 \|= DBCR0_IAC2 \| DBCR0_IDM;
	295	}
	296	if (dbg->bp[2]) {
	297	mtspr(SPRN_IAC3, dbg->bp[2]);
	298	dbcr0 \|= DBCR0_IAC3 \| DBCR0_IDM;
	299	}
	300	if (dbg->bp[3]) {
	301	mtspr(SPRN_IAC4, dbg->bp[3]);
	302	dbcr0 \|= DBCR0_IAC4 \| DBCR0_IDM;
	303	}
	304
	305	mtspr(SPRN_DBCR0, dbcr0);
	306	mtspr(SPRN_DBCR1, 0);
	307	mtspr(SPRN_DBCR2, 0);
308	}
309
bbf45ba5 HB	310	void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
bbf45ba5 HB	311	{
83aae4a8 HB	312	int i;
83aae4a8 HB	313
6a0ab738 HB	314	if (vcpu->guest_debug.enabled)
6a0ab738 HB	315	kvmppc_load_guest_debug_registers(vcpu);
83aae4a8 HB	316
	317	/* Mark every guest entry in the shadow TLB entry modified, so that they
	318	* will all be reloaded on the next vcpu run (instead of being
	319	* demand-faulted). */
	320	for (i = 0; i <= tlb_44x_hwater; i++)
	321	kvmppc_tlbe_set_modified(vcpu, i);
bbf45ba5 HB	322	}
	323
	324	void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
	325	{
6a0ab738 HB	326	if (vcpu->guest_debug.enabled)
6a0ab738 HB	327	kvmppc_restore_host_debug_state(vcpu);
83aae4a8 HB	328
	329	/* Don't leave guest TLB entries resident when being de-scheduled. */
	330	/* XXX It would be nice to differentiate between heavyweight exit and
	331	* sched_out here, since we could avoid the TLB flush for heavyweight
	332	* exits. */
	333	_tlbia();
bbf45ba5 HB	334	}
bbf45ba5 HB	335
bbf45ba5 HB	336	int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
	337	struct kvm_debug_guest *dbg)
	338	{
6a0ab738 HB	339	int i;
	340
	341	vcpu->guest_debug.enabled = dbg->enabled;
	342	if (vcpu->guest_debug.enabled) {
	343	for (i=0; i < ARRAY_SIZE(vcpu->guest_debug.bp); i++) {
	344	if (dbg->breakpoints[i].enabled)
	345	vcpu->guest_debug.bp[i] = dbg->breakpoints[i].address;
	346	else
	347	vcpu->guest_debug.bp[i] = 0;
	348	}
	349	}
	350
	351	return 0;
bbf45ba5 HB	352	}
	353
	354	static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,
	355	struct kvm_run *run)
	356	{
	357	u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
	358	*gpr = run->dcr.data;
	359	}
	360
	361	static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
	362	struct kvm_run *run)
	363	{
	364	u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
	365
	366	if (run->mmio.len > sizeof(*gpr)) {
	367	printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
	368	return;
	369	}
	370
	371	if (vcpu->arch.mmio_is_bigendian) {
	372	switch (run->mmio.len) {
	373	case 4: gpr = (u32 *)run->mmio.data; break;
	374	case 2: gpr = (u16 *)run->mmio.data; break;
	375	case 1: gpr = (u8 *)run->mmio.data; break;
	376	}
	377	} else {
	378	/* Convert BE data from userland back to LE. */
	379	switch (run->mmio.len) {
	380	case 4: gpr = ld_le32((u32 )run->mmio.data); break;
	381	case 2: gpr = ld_le16((u16 )run->mmio.data); break;
	382	case 1: gpr = (u8 *)run->mmio.data; break;
	383	}
	384	}
	385	}
	386
	387	int kvmppc_handle_load(struct kvm_run run, struct kvm_vcpu vcpu,
	388	unsigned int rt, unsigned int bytes, int is_bigendian)
	389	{
	390	if (bytes > sizeof(run->mmio.data)) {
	391	printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__,
	392	run->mmio.len);
	393	}
	394
	395	run->mmio.phys_addr = vcpu->arch.paddr_accessed;
	396	run->mmio.len = bytes;
	397	run->mmio.is_write = 0;
	398
	399	vcpu->arch.io_gpr = rt;
	400	vcpu->arch.mmio_is_bigendian = is_bigendian;
	401	vcpu->mmio_needed = 1;
	402	vcpu->mmio_is_write = 0;
	403
	404	return EMULATE_DO_MMIO;
	405	}
	406
	407	int kvmppc_handle_store(struct kvm_run run, struct kvm_vcpu vcpu,
	408	u32 val, unsigned int bytes, int is_bigendian)
	409	{
	410	void *data = run->mmio.data;
	411
	412	if (bytes > sizeof(run->mmio.data)) {
	413	printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__,
	414	run->mmio.len);
	415	}
416
417	run->mmio.phys_addr = vcpu->arch.paddr_accessed;
418	run->mmio.len = bytes;
419	run->mmio.is_write = 1;
420	vcpu->mmio_needed = 1;
421	vcpu->mmio_is_write = 1;
422
423	/* Store the value at the lowest bytes in 'data'. */
424	if (is_bigendian) {
425	switch (bytes) {
426	case 4: (u32 )data = val; break;
427	case 2: (u16 )data = val; break;
428	case 1: (u8 )data = val; break;
429	}
430	} else {
431	/* Store LE value into 'data'. */
432	switch (bytes) {
433	case 4: st_le32(data, val); break;
434	case 2: st_le16(data, val); break;
435	case 1: (u8 )data = val; break;
436	}
437	}
438
439	return EMULATE_DO_MMIO;
440	}
441
442	int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu vcpu, struct kvm_run run)
443	{
444	int r;
445	sigset_t sigsaved;
446
45c5eb67 HB	447	vcpu_load(vcpu);
45c5eb67 HB	448
bbf45ba5 HB	449	if (vcpu->sigset_active)
	450	sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
	451
	452	if (vcpu->mmio_needed) {
	453	if (!vcpu->mmio_is_write)
	454	kvmppc_complete_mmio_load(vcpu, run);
	455	vcpu->mmio_needed = 0;
	456	} else if (vcpu->arch.dcr_needed) {
	457	if (!vcpu->arch.dcr_is_write)
	458	kvmppc_complete_dcr_load(vcpu, run);
	459	vcpu->arch.dcr_needed = 0;
	460	}
	461
	462	kvmppc_check_and_deliver_interrupts(vcpu);
	463
	464	local_irq_disable();
	465	kvm_guest_enter();
	466	r = __kvmppc_vcpu_run(run, vcpu);
	467	kvm_guest_exit();
	468	local_irq_enable();
	469
	470	if (vcpu->sigset_active)
	471	sigprocmask(SIG_SETMASK, &sigsaved, NULL);
	472
45c5eb67 HB	473	vcpu_put(vcpu);
45c5eb67 HB	474
bbf45ba5 HB	475	return r;
	476	}
	477
	478	int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu vcpu, struct kvm_interrupt irq)
	479	{
	480	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_EXTERNAL);
45c5eb67 HB	481
	482	if (waitqueue_active(&vcpu->wq)) {
	483	wake_up_interruptible(&vcpu->wq);
	484	vcpu->stat.halt_wakeup++;
	485	}
	486
bbf45ba5 HB	487	return 0;
	488	}
	489
	490	int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
	491	struct kvm_mp_state *mp_state)
	492	{
	493	return -EINVAL;
	494	}
	495
	496	int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
	497	struct kvm_mp_state *mp_state)
	498	{
	499	return -EINVAL;
	500	}
	501
	502	long kvm_arch_vcpu_ioctl(struct file *filp,
	503	unsigned int ioctl, unsigned long arg)
	504	{
	505	struct kvm_vcpu *vcpu = filp->private_data;
	506	void __user argp = (void __user )arg;
	507	long r;
	508
	509	switch (ioctl) {
	510	case KVM_INTERRUPT: {
	511	struct kvm_interrupt irq;
	512	r = -EFAULT;
	513	if (copy_from_user(&irq, argp, sizeof(irq)))
	514	goto out;
	515	r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
	516	break;
	517	}
	518	default:
	519	r = -EINVAL;
	520	}
	521
	522	out:
	523	return r;
	524	}
	525
	526	int kvm_vm_ioctl_get_dirty_log(struct kvm kvm, struct kvm_dirty_log log)
	527	{
	528	return -ENOTSUPP;
	529	}
	530
	531	long kvm_arch_vm_ioctl(struct file *filp,
	532	unsigned int ioctl, unsigned long arg)
	533	{
	534	long r;
	535
	536	switch (ioctl) {
	537	default:
	538	r = -EINVAL;
	539	}
	540
	541	return r;
	542	}
	543
	544	int kvm_arch_init(void *opaque)
	545	{
	546	return 0;
	547	}
	548
	549	void kvm_arch_exit(void)
	550	{
551	}