[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/vmalloc.h>
#include <linux/hrtimer.h>
#include <linux/fs.h>
#include <linux/slab.h>
#include <asm/cputable.h>
#include <asm/uaccess.h>
#include <asm/kvm_ppc.h>
#include <asm/tlbflush.h>
#include <asm/cputhreads.h>
#include "timing.h"
#include "../mm/mmu_decl.h"

#define CREATE_TRACE_POINTS
#include "trace.h"

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

int kvmppc_kvm_pv(struct kvm_vcpu *vcpu)
{
	int nr = kvmppc_get_gpr(vcpu, 11);
	int r;
	unsigned long __maybe_unused param1 = kvmppc_get_gpr(vcpu, 3);
	unsigned long __maybe_unused param2 = kvmppc_get_gpr(vcpu, 4);
	unsigned long __maybe_unused param3 = kvmppc_get_gpr(vcpu, 5);
	unsigned long __maybe_unused param4 = kvmppc_get_gpr(vcpu, 6);
	unsigned long r2 = 0;

	if (!(vcpu->arch.shared->msr & MSR_SF)) {
		/* 32 bit mode */
		param1 &= 0xffffffff;
		param2 &= 0xffffffff;
		param3 &= 0xffffffff;
		param4 &= 0xffffffff;
	}

	switch (nr) {
	case HC_VENDOR_KVM | KVM_HC_PPC_MAP_MAGIC_PAGE:
	{
		vcpu->arch.magic_page_pa = param1;
		vcpu->arch.magic_page_ea = param2;

		r2 = KVM_MAGIC_FEAT_SR | KVM_MAGIC_FEAT_MAS0_TO_SPRG7;

		r = HC_EV_SUCCESS;
		break;
	}
	case HC_VENDOR_KVM | KVM_HC_FEATURES:
		r = HC_EV_SUCCESS;
#if defined(CONFIG_PPC_BOOK3S) || defined(CONFIG_KVM_E500)
		/* XXX Missing magic page on 44x */
		r2 |= (1 << KVM_FEATURE_MAGIC_PAGE);
#endif

		/* Second return value is in r4 */
		break;
	default:
		r = HC_EV_UNIMPLEMENTED;
		break;
	}

	kvmppc_set_gpr(vcpu, 4, r2);

	return r;
}

int kvmppc_sanity_check(struct kvm_vcpu *vcpu)
{
	int r = false;

	/* We have to know what CPU to virtualize */
	if (!vcpu->arch.pvr)
		goto out;

	/* PAPR only works with book3s_64 */
	if ((vcpu->arch.cpu_type != KVM_CPU_3S_64) && vcpu->arch.papr_enabled)
		goto out;

#ifdef CONFIG_KVM_BOOK3S_64_HV
	/* HV KVM can only do PAPR mode for now */
	if (!vcpu->arch.papr_enabled)
		goto out;
#endif

	r = true;

out:
	vcpu->arch.sane = r;
	return r ? 0 : -EINVAL;
}

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__,
		       kvmppc_get_last_inst(vcpu));
		r = RESUME_HOST;
		break;
	default:
		BUG();
	}

	return r;
}

int kvm_arch_hardware_enable(void *garbage)
{
	return 0;
}

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 *)rtn = kvmppc_core_check_processor_compat();
}

int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
{
	if (type)
		return -EINVAL;

	return kvmppc_core_init_vm(kvm);
}

void kvm_arch_destroy_vm(struct kvm *kvm)
{
	unsigned int i;
	struct kvm_vcpu *vcpu;

	kvm_for_each_vcpu(i, vcpu, kvm)
		kvm_arch_vcpu_free(vcpu);

	mutex_lock(&kvm->lock);
	for (i = 0; i < atomic_read(&kvm->online_vcpus); i++)
		kvm->vcpus[i] = NULL;

	atomic_set(&kvm->online_vcpus, 0);

	kvmppc_core_destroy_vm(kvm);

	mutex_unlock(&kvm->lock);
}

void kvm_arch_sync_events(struct kvm *kvm)
{
}

int kvm_dev_ioctl_check_extension(long ext)
{
	int r;

	switch (ext) {
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_SREGS:
#else
	case KVM_CAP_PPC_SEGSTATE:
	case KVM_CAP_PPC_PAPR:
#endif
	case KVM_CAP_PPC_UNSET_IRQ:
	case KVM_CAP_PPC_IRQ_LEVEL:
	case KVM_CAP_ENABLE_CAP:
		r = 1;
		break;
#ifndef CONFIG_KVM_BOOK3S_64_HV
	case KVM_CAP_PPC_PAIRED_SINGLES:
	case KVM_CAP_PPC_OSI:
	case KVM_CAP_PPC_GET_PVINFO:
#ifdef CONFIG_KVM_E500
	case KVM_CAP_SW_TLB:
#endif
		r = 1;
		break;
	case KVM_CAP_COALESCED_MMIO:
		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
		break;
#endif
#ifdef CONFIG_KVM_BOOK3S_64_HV
	case KVM_CAP_SPAPR_TCE:
		r = 1;
		break;
	case KVM_CAP_PPC_SMT:
		r = threads_per_core;
		break;
	case KVM_CAP_PPC_RMA:
		r = 1;
		/* PPC970 requires an RMA */
		if (cpu_has_feature(CPU_FTR_ARCH_201))
			r = 2;
		break;
	case KVM_CAP_SYNC_MMU:
		r = cpu_has_feature(CPU_FTR_ARCH_206) ? 1 : 0;
		break;
#endif
	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_prepare_memory_region(struct kvm *kvm,
                                   struct kvm_memory_slot *memslot,
                                   struct kvm_memory_slot old,
                                   struct kvm_userspace_memory_region *mem,
                                   int user_alloc)
{
	return kvmppc_core_prepare_memory_region(kvm, mem);
}

void kvm_arch_commit_memory_region(struct kvm *kvm,
               struct kvm_userspace_memory_region *mem,
               struct kvm_memory_slot old,
               int user_alloc)
{
	kvmppc_core_commit_memory_region(kvm, mem);
}


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;
	vcpu = kvmppc_core_vcpu_create(kvm, id);
	vcpu->arch.wqp = &vcpu->wq;
	if (!IS_ERR(vcpu))
		kvmppc_create_vcpu_debugfs(vcpu, id);
	return vcpu;
}

void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
	/* Make sure we're not using the vcpu anymore */
	hrtimer_cancel(&vcpu->arch.dec_timer);
	tasklet_kill(&vcpu->arch.tasklet);

	kvmppc_remove_vcpu_debugfs(vcpu);
	kvmppc_core_vcpu_free(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)
{
	return kvmppc_core_pending_dec(vcpu);
}

/*
 * low level hrtimer wake routine. Because this runs in hardirq context
 * we schedule a tasklet to do the real work.
 */
enum hrtimer_restart kvmppc_decrementer_wakeup(struct hrtimer *timer)
{
	struct kvm_vcpu *vcpu;

	vcpu = container_of(timer, struct kvm_vcpu, arch.dec_timer);
	tasklet_schedule(&vcpu->arch.tasklet);

	return HRTIMER_NORESTART;
}

int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
	hrtimer_init(&vcpu->arch.dec_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
	tasklet_init(&vcpu->arch.tasklet, kvmppc_decrementer_func, (ulong)vcpu);
	vcpu->arch.dec_timer.function = kvmppc_decrementer_wakeup;
	vcpu->arch.dec_expires = ~(u64)0;

#ifdef CONFIG_KVM_EXIT_TIMING
	mutex_init(&vcpu->arch.exit_timing_lock);
#endif

	return 0;
}

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

void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
#ifdef CONFIG_BOOKE
	/*
	 * vrsave (formerly usprg0) isn't used by Linux, but may
	 * be used by the guest.
	 *
	 * On non-booke this is associated with Altivec and
	 * is handled by code in book3s.c.
	 */
	mtspr(SPRN_VRSAVE, vcpu->arch.vrsave);
#endif
	kvmppc_core_vcpu_load(vcpu, cpu);
	vcpu->cpu = smp_processor_id();
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
	kvmppc_core_vcpu_put(vcpu);
#ifdef CONFIG_BOOKE
	vcpu->arch.vrsave = mfspr(SPRN_VRSAVE);
#endif
	vcpu->cpu = -1;
}

int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
                                        struct kvm_guest_debug *dbg)
{
	return -EINVAL;
}

static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,
                                     struct kvm_run *run)
{
	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, run->dcr.data);
}

static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
                                      struct kvm_run *run)
{
	u64 uninitialized_var(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 8: gpr = *(u64 *)run->mmio.data; break;
		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;
		}
	}

	if (vcpu->arch.mmio_sign_extend) {
		switch (run->mmio.len) {
#ifdef CONFIG_PPC64
		case 4:
			gpr = (s64)(s32)gpr;
			break;
#endif
		case 2:
			gpr = (s64)(s16)gpr;
			break;
		case 1:
			gpr = (s64)(s8)gpr;
			break;
		}
	}

	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);

	switch (vcpu->arch.io_gpr & KVM_REG_EXT_MASK) {
	case KVM_REG_GPR:
		kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
		break;
	case KVM_REG_FPR:
		vcpu->arch.fpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
		break;
#ifdef CONFIG_PPC_BOOK3S
	case KVM_REG_QPR:
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
		break;
	case KVM_REG_FQPR:
		vcpu->arch.fpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
		break;
#endif
	default:
		BUG();
	}
}

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;
	vcpu->arch.mmio_sign_extend = 0;

	return EMULATE_DO_MMIO;
}

/* Same as above, but sign extends */
int kvmppc_handle_loads(struct kvm_run *run, struct kvm_vcpu *vcpu,
                        unsigned int rt, unsigned int bytes, int is_bigendian)
{
	int r;

	r = kvmppc_handle_load(run, vcpu, rt, bytes, is_bigendian);
	vcpu->arch.mmio_sign_extend = 1;

	return r;
}

int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
                        u64 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 8: *(u64 *)data = val; break;
		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;

	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;
	} else if (vcpu->arch.osi_needed) {
		u64 *gprs = run->osi.gprs;
		int i;

		for (i = 0; i < 32; i++)
			kvmppc_set_gpr(vcpu, i, gprs[i]);
		vcpu->arch.osi_needed = 0;
	} else if (vcpu->arch.hcall_needed) {
		int i;

		kvmppc_set_gpr(vcpu, 3, run->papr_hcall.ret);
		for (i = 0; i < 9; ++i)
			kvmppc_set_gpr(vcpu, 4 + i, run->papr_hcall.args[i]);
		vcpu->arch.hcall_needed = 0;
	}

	r = kvmppc_vcpu_run(run, vcpu);

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

	return r;
}

void kvm_vcpu_kick(struct kvm_vcpu *vcpu)
{
	int me;
	int cpu = vcpu->cpu;

	me = get_cpu();
	if (waitqueue_active(vcpu->arch.wqp)) {
		wake_up_interruptible(vcpu->arch.wqp);
		vcpu->stat.halt_wakeup++;
	} else if (cpu != me && cpu != -1) {
		smp_send_reschedule(vcpu->cpu);
	}
	put_cpu();
}

int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
{
	if (irq->irq == KVM_INTERRUPT_UNSET) {
		kvmppc_core_dequeue_external(vcpu, irq);
		return 0;
	}

	kvmppc_core_queue_external(vcpu, irq);
	kvm_vcpu_kick(vcpu);

	return 0;
}

static int kvm_vcpu_ioctl_enable_cap(struct kvm_vcpu *vcpu,
				     struct kvm_enable_cap *cap)
{
	int r;

	if (cap->flags)
		return -EINVAL;

	switch (cap->cap) {
	case KVM_CAP_PPC_OSI:
		r = 0;
		vcpu->arch.osi_enabled = true;
		break;
	case KVM_CAP_PPC_PAPR:
		r = 0;
		vcpu->arch.papr_enabled = true;
		break;
#ifdef CONFIG_KVM_E500
	case KVM_CAP_SW_TLB: {
		struct kvm_config_tlb cfg;
		void __user *user_ptr = (void __user *)(uintptr_t)cap->args[0];

		r = -EFAULT;
		if (copy_from_user(&cfg, user_ptr, sizeof(cfg)))
			break;

		r = kvm_vcpu_ioctl_config_tlb(vcpu, &cfg);
		break;
	}
#endif
	default:
		r = -EINVAL;
		break;
	}

	if (!r)
		r = kvmppc_sanity_check(vcpu);

	return r;
}

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);
		goto out;
	}

	case KVM_ENABLE_CAP:
	{
		struct kvm_enable_cap cap;
		r = -EFAULT;
		if (copy_from_user(&cap, argp, sizeof(cap)))
			goto out;
		r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
		break;
	}

#ifdef CONFIG_KVM_E500
	case KVM_DIRTY_TLB: {
		struct kvm_dirty_tlb dirty;
		r = -EFAULT;
		if (copy_from_user(&dirty, argp, sizeof(dirty)))
			goto out;
		r = kvm_vcpu_ioctl_dirty_tlb(vcpu, &dirty);
		break;
	}
#endif

	default:
		r = -EINVAL;
	}

out:
	return r;
}

int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf)
{
	return VM_FAULT_SIGBUS;
}

static int kvm_vm_ioctl_get_pvinfo(struct kvm_ppc_pvinfo *pvinfo)
{
	u32 inst_lis = 0x3c000000;
	u32 inst_ori = 0x60000000;
	u32 inst_nop = 0x60000000;
	u32 inst_sc = 0x44000002;
	u32 inst_imm_mask = 0xffff;

	/*
	 * The hypercall to get into KVM from within guest context is as
	 * follows:
	 *
	 *    lis r0, r0, KVM_SC_MAGIC_R0@h
	 *    ori r0, KVM_SC_MAGIC_R0@l
	 *    sc
	 *    nop
	 */
	pvinfo->hcall[0] = inst_lis | ((KVM_SC_MAGIC_R0 >> 16) & inst_imm_mask);
	pvinfo->hcall[1] = inst_ori | (KVM_SC_MAGIC_R0 & inst_imm_mask);
	pvinfo->hcall[2] = inst_sc;
	pvinfo->hcall[3] = inst_nop;

	return 0;
}

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

	switch (ioctl) {
	case KVM_PPC_GET_PVINFO: {
		struct kvm_ppc_pvinfo pvinfo;
		memset(&pvinfo, 0, sizeof(pvinfo));
		r = kvm_vm_ioctl_get_pvinfo(&pvinfo);
		if (copy_to_user(argp, &pvinfo, sizeof(pvinfo))) {
			r = -EFAULT;
			goto out;
		}

		break;
	}
#ifdef CONFIG_KVM_BOOK3S_64_HV
	case KVM_CREATE_SPAPR_TCE: {
		struct kvm_create_spapr_tce create_tce;
		struct kvm *kvm = filp->private_data;

		r = -EFAULT;
		if (copy_from_user(&create_tce, argp, sizeof(create_tce)))
			goto out;
		r = kvm_vm_ioctl_create_spapr_tce(kvm, &create_tce);
		goto out;
	}

	case KVM_ALLOCATE_RMA: {
		struct kvm *kvm = filp->private_data;
		struct kvm_allocate_rma rma;

		r = kvm_vm_ioctl_allocate_rma(kvm, &rma);
		if (r >= 0 && copy_to_user(argp, &rma, sizeof(rma)))
			r = -EFAULT;
		break;
	}
#endif /* CONFIG_KVM_BOOK3S_64_HV */

	default:
		r = -ENOTTY;
	}

out:
	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>
bbf45ba5	24	#include <linux/vmalloc.h>
544c6761	25	#include <linux/hrtimer.h>
bbf45ba5	26	#include <linux/fs.h>
5a0e3ad6	27	#include <linux/slab.h>
bbf45ba5 HB	28	#include <asm/cputable.h>
	29	#include <asm/uaccess.h>
	30	#include <asm/kvm_ppc.h>
83aae4a8	31	#include <asm/tlbflush.h>
371fefd6	32	#include <asm/cputhreads.h>
73e75b41	33	#include "timing.h"
fad7b9b5	34	#include "../mm/mmu_decl.h"
bbf45ba5	35
46f43c6e MT	36	#define CREATE_TRACE_POINTS
	37	#include "trace.h"
	38
bbf45ba5 HB	39	int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
bbf45ba5 HB	40	{
666e7252	41	return !(v->arch.shared->msr & MSR_WE) \|\|
dfd4d47e SW	42	!!(v->arch.pending_exceptions) \|\|
dfd4d47e SW	43	v->requests;
bbf45ba5 HB	44	}
bbf45ba5 HB	45
2a342ed5 AG	46	int kvmppc_kvm_pv(struct kvm_vcpu *vcpu)
	47	{
	48	int nr = kvmppc_get_gpr(vcpu, 11);
	49	int r;
	50	unsigned long __maybe_unused param1 = kvmppc_get_gpr(vcpu, 3);
	51	unsigned long __maybe_unused param2 = kvmppc_get_gpr(vcpu, 4);
	52	unsigned long __maybe_unused param3 = kvmppc_get_gpr(vcpu, 5);
	53	unsigned long __maybe_unused param4 = kvmppc_get_gpr(vcpu, 6);
	54	unsigned long r2 = 0;
	55
	56	if (!(vcpu->arch.shared->msr & MSR_SF)) {
	57	/* 32 bit mode */
	58	param1 &= 0xffffffff;
	59	param2 &= 0xffffffff;
	60	param3 &= 0xffffffff;
	61	param4 &= 0xffffffff;
	62	}
	63
	64	switch (nr) {
5fc87407 AG	65	case HC_VENDOR_KVM \| KVM_HC_PPC_MAP_MAGIC_PAGE:
	66	{
	67	vcpu->arch.magic_page_pa = param1;
	68	vcpu->arch.magic_page_ea = param2;
	69
b5904972	70	r2 = KVM_MAGIC_FEAT_SR \| KVM_MAGIC_FEAT_MAS0_TO_SPRG7;
7508e16c	71
5fc87407 AG	72	r = HC_EV_SUCCESS;
	73	break;
	74	}
2a342ed5 AG	75	case HC_VENDOR_KVM \| KVM_HC_FEATURES:
2a342ed5 AG	76	r = HC_EV_SUCCESS;
a4cd8b23 SW	77	#if defined(CONFIG_PPC_BOOK3S) \|\| defined(CONFIG_KVM_E500)
a4cd8b23 SW	78	/* XXX Missing magic page on 44x */
5fc87407 AG	79	r2 \|= (1 << KVM_FEATURE_MAGIC_PAGE);
5fc87407 AG	80	#endif
2a342ed5 AG	81
2a342ed5 AG	82	/* Second return value is in r4 */
2a342ed5 AG	83	break;
	84	default:
	85	r = HC_EV_UNIMPLEMENTED;
	86	break;
	87	}
	88
7508e16c AG	89	kvmppc_set_gpr(vcpu, 4, r2);
7508e16c AG	90
2a342ed5 AG	91	return r;
2a342ed5 AG	92	}
bbf45ba5	93
af8f38b3 AG	94	int kvmppc_sanity_check(struct kvm_vcpu *vcpu)
	95	{
	96	int r = false;
	97
	98	/* We have to know what CPU to virtualize */
	99	if (!vcpu->arch.pvr)
	100	goto out;
	101
	102	/* PAPR only works with book3s_64 */
	103	if ((vcpu->arch.cpu_type != KVM_CPU_3S_64) && vcpu->arch.papr_enabled)
	104	goto out;
	105
	106	#ifdef CONFIG_KVM_BOOK3S_64_HV
	107	/* HV KVM can only do PAPR mode for now */
	108	if (!vcpu->arch.papr_enabled)
	109	goto out;
	110	#endif
	111
	112	r = true;
	113
	114	out:
	115	vcpu->arch.sane = r;
	116	return r ? 0 : -EINVAL;
	117	}
	118
bbf45ba5 HB	119	int kvmppc_emulate_mmio(struct kvm_run run, struct kvm_vcpu vcpu)
	120	{
	121	enum emulation_result er;
	122	int r;
	123
	124	er = kvmppc_emulate_instruction(run, vcpu);
	125	switch (er) {
	126	case EMULATE_DONE:
	127	/* Future optimization: only reload non-volatiles if they were
	128	* actually modified. */
	129	r = RESUME_GUEST_NV;
	130	break;
	131	case EMULATE_DO_MMIO:
	132	run->exit_reason = KVM_EXIT_MMIO;
	133	/* We must reload nonvolatiles because "update" load/store
	134	* instructions modify register state. */
	135	/* Future optimization: only reload non-volatiles if they were
	136	* actually modified. */
	137	r = RESUME_HOST_NV;
	138	break;
	139	case EMULATE_FAIL:
	140	/* XXX Deliver Program interrupt to guest. */
	141	printk(KERN_EMERG "%s: emulation failed (%08x)\n", __func__,
c7f38f46	142	kvmppc_get_last_inst(vcpu));
bbf45ba5 HB	143	r = RESUME_HOST;
	144	break;
	145	default:
	146	BUG();
	147	}
	148
	149	return r;
	150	}
	151
10474ae8	152	int kvm_arch_hardware_enable(void *garbage)
bbf45ba5	153	{
10474ae8	154	return 0;
bbf45ba5 HB	155	}
	156
	157	void kvm_arch_hardware_disable(void *garbage)
	158	{
	159	}
	160
	161	int kvm_arch_hardware_setup(void)
	162	{
	163	return 0;
	164	}
	165
	166	void kvm_arch_hardware_unsetup(void)
	167	{
	168	}
	169
	170	void kvm_arch_check_processor_compat(void *rtn)
	171	{
9dd921cf	172	(int )rtn = kvmppc_core_check_processor_compat();
bbf45ba5 HB	173	}
bbf45ba5 HB	174
e08b9637	175	int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
bbf45ba5	176	{
e08b9637 CO	177	if (type)
	178	return -EINVAL;
	179
f9e0554d	180	return kvmppc_core_init_vm(kvm);
bbf45ba5 HB	181	}
bbf45ba5 HB	182
d89f5eff	183	void kvm_arch_destroy_vm(struct kvm *kvm)
bbf45ba5 HB	184	{
bbf45ba5 HB	185	unsigned int i;
988a2cae	186	struct kvm_vcpu *vcpu;
bbf45ba5	187
988a2cae GN	188	kvm_for_each_vcpu(i, vcpu, kvm)
	189	kvm_arch_vcpu_free(vcpu);
	190
	191	mutex_lock(&kvm->lock);
	192	for (i = 0; i < atomic_read(&kvm->online_vcpus); i++)
	193	kvm->vcpus[i] = NULL;
	194
	195	atomic_set(&kvm->online_vcpus, 0);
f9e0554d PM	196
	197	kvmppc_core_destroy_vm(kvm);
	198
988a2cae	199	mutex_unlock(&kvm->lock);
bbf45ba5 HB	200	}
bbf45ba5 HB	201
ad8ba2cd SY	202	void kvm_arch_sync_events(struct kvm *kvm)
	203	{
	204	}
	205
bbf45ba5 HB	206	int kvm_dev_ioctl_check_extension(long ext)
	207	{
	208	int r;
	209
	210	switch (ext) {
5ce941ee SW	211	#ifdef CONFIG_BOOKE
	212	case KVM_CAP_PPC_BOOKE_SREGS:
	213	#else
e15a1137	214	case KVM_CAP_PPC_SEGSTATE:
930b412a	215	case KVM_CAP_PPC_PAPR:
5ce941ee	216	#endif
18978768	217	case KVM_CAP_PPC_UNSET_IRQ:
7b4203e8	218	case KVM_CAP_PPC_IRQ_LEVEL:
71fbfd5f	219	case KVM_CAP_ENABLE_CAP:
de56a948 PM	220	r = 1;
	221	break;
	222	#ifndef CONFIG_KVM_BOOK3S_64_HV
	223	case KVM_CAP_PPC_PAIRED_SINGLES:
ad0a048b	224	case KVM_CAP_PPC_OSI:
15711e9c	225	case KVM_CAP_PPC_GET_PVINFO:
dc83b8bc SW	226	#ifdef CONFIG_KVM_E500
	227	case KVM_CAP_SW_TLB:
	228	#endif
e15a1137 AG	229	r = 1;
e15a1137 AG	230	break;
588968b6 LV	231	case KVM_CAP_COALESCED_MMIO:
	232	r = KVM_COALESCED_MMIO_PAGE_OFFSET;
	233	break;
54738c09 DG	234	#endif
	235	#ifdef CONFIG_KVM_BOOK3S_64_HV
	236	case KVM_CAP_SPAPR_TCE:
	237	r = 1;
	238	break;
371fefd6 PM	239	case KVM_CAP_PPC_SMT:
	240	r = threads_per_core;
	241	break;
aa04b4cc PM	242	case KVM_CAP_PPC_RMA:
aa04b4cc PM	243	r = 1;
9e368f29 PM	244	/* PPC970 requires an RMA */
	245	if (cpu_has_feature(CPU_FTR_ARCH_201))
	246	r = 2;
aa04b4cc	247	break;
342d3db7 PM	248	case KVM_CAP_SYNC_MMU:
	249	r = cpu_has_feature(CPU_FTR_ARCH_206) ? 1 : 0;
	250	break;
de56a948	251	#endif
bbf45ba5 HB	252	default:
	253	r = 0;
	254	break;
	255	}
	256	return r;
	257
	258	}
	259
	260	long kvm_arch_dev_ioctl(struct file *filp,
	261	unsigned int ioctl, unsigned long arg)
	262	{
	263	return -EINVAL;
	264	}
	265
f7784b8e MT	266	int kvm_arch_prepare_memory_region(struct kvm *kvm,
	267	struct kvm_memory_slot *memslot,
	268	struct kvm_memory_slot old,
	269	struct kvm_userspace_memory_region *mem,
	270	int user_alloc)
bbf45ba5	271	{
f9e0554d	272	return kvmppc_core_prepare_memory_region(kvm, mem);
bbf45ba5 HB	273	}
bbf45ba5 HB	274
f7784b8e MT	275	void kvm_arch_commit_memory_region(struct kvm *kvm,
	276	struct kvm_userspace_memory_region *mem,
	277	struct kvm_memory_slot old,
	278	int user_alloc)
	279	{
f9e0554d	280	kvmppc_core_commit_memory_region(kvm, mem);
f7784b8e MT	281	}
	282
	283
34d4cb8f MT	284	void kvm_arch_flush_shadow(struct kvm *kvm)
	285	{
	286	}
	287
bbf45ba5 HB	288	struct kvm_vcpu kvm_arch_vcpu_create(struct kvm kvm, unsigned int id)
bbf45ba5 HB	289	{
73e75b41 HB	290	struct kvm_vcpu *vcpu;
73e75b41 HB	291	vcpu = kvmppc_core_vcpu_create(kvm, id);
19ccb76a	292	vcpu->arch.wqp = &vcpu->wq;
06056bfb WY	293	if (!IS_ERR(vcpu))
06056bfb WY	294	kvmppc_create_vcpu_debugfs(vcpu, id);
73e75b41	295	return vcpu;
bbf45ba5 HB	296	}
	297
	298	void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
	299	{
a595405d AG	300	/* Make sure we're not using the vcpu anymore */
	301	hrtimer_cancel(&vcpu->arch.dec_timer);
	302	tasklet_kill(&vcpu->arch.tasklet);
	303
73e75b41	304	kvmppc_remove_vcpu_debugfs(vcpu);
db93f574	305	kvmppc_core_vcpu_free(vcpu);
bbf45ba5 HB	306	}
	307
	308	void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
	309	{
	310	kvm_arch_vcpu_free(vcpu);
	311	}
	312
	313	int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
	314	{
9dd921cf	315	return kvmppc_core_pending_dec(vcpu);
bbf45ba5 HB	316	}
bbf45ba5 HB	317
544c6761 AG	318	/*
	319	* low level hrtimer wake routine. Because this runs in hardirq context
	320	* we schedule a tasklet to do the real work.
	321	*/
	322	enum hrtimer_restart kvmppc_decrementer_wakeup(struct hrtimer *timer)
	323	{
	324	struct kvm_vcpu *vcpu;
	325
	326	vcpu = container_of(timer, struct kvm_vcpu, arch.dec_timer);
	327	tasklet_schedule(&vcpu->arch.tasklet);
	328
	329	return HRTIMER_NORESTART;
	330	}
	331
bbf45ba5 HB	332	int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
bbf45ba5 HB	333	{
544c6761 AG	334	hrtimer_init(&vcpu->arch.dec_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
	335	tasklet_init(&vcpu->arch.tasklet, kvmppc_decrementer_func, (ulong)vcpu);
	336	vcpu->arch.dec_timer.function = kvmppc_decrementer_wakeup;
de56a948	337	vcpu->arch.dec_expires = ~(u64)0;
bbf45ba5	338
09000adb BB	339	#ifdef CONFIG_KVM_EXIT_TIMING
	340	mutex_init(&vcpu->arch.exit_timing_lock);
	341	#endif
	342
bbf45ba5 HB	343	return 0;
	344	}
	345
	346	void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
	347	{
ecc0981f	348	kvmppc_mmu_destroy(vcpu);
bbf45ba5 HB	349	}
	350
	351	void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
	352	{
eab17672 SW	353	#ifdef CONFIG_BOOKE
	354	/*
	355	* vrsave (formerly usprg0) isn't used by Linux, but may
	356	* be used by the guest.
	357	*
	358	* On non-booke this is associated with Altivec and
	359	* is handled by code in book3s.c.
	360	*/
	361	mtspr(SPRN_VRSAVE, vcpu->arch.vrsave);
	362	#endif
9dd921cf	363	kvmppc_core_vcpu_load(vcpu, cpu);
de56a948	364	vcpu->cpu = smp_processor_id();
bbf45ba5 HB	365	}
	366
	367	void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
	368	{
9dd921cf	369	kvmppc_core_vcpu_put(vcpu);
eab17672 SW	370	#ifdef CONFIG_BOOKE
	371	vcpu->arch.vrsave = mfspr(SPRN_VRSAVE);
	372	#endif
de56a948	373	vcpu->cpu = -1;
bbf45ba5 HB	374	}
bbf45ba5 HB	375
d0bfb940	376	int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
f5d0906b	377	struct kvm_guest_debug *dbg)
bbf45ba5	378	{
f5d0906b	379	return -EINVAL;
bbf45ba5 HB	380	}
	381
	382	static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,
	383	struct kvm_run *run)
	384	{
8e5b26b5	385	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, run->dcr.data);
bbf45ba5 HB	386	}
	387
	388	static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
	389	struct kvm_run *run)
	390	{
69b61833	391	u64 uninitialized_var(gpr);
bbf45ba5	392
8e5b26b5	393	if (run->mmio.len > sizeof(gpr)) {
bbf45ba5 HB	394	printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
	395	return;
	396	}
	397
	398	if (vcpu->arch.mmio_is_bigendian) {
	399	switch (run->mmio.len) {
b104d066	400	case 8: gpr = (u64 )run->mmio.data; break;
8e5b26b5 AG	401	case 4: gpr = (u32 )run->mmio.data; break;
	402	case 2: gpr = (u16 )run->mmio.data; break;
	403	case 1: gpr = (u8 )run->mmio.data; break;
bbf45ba5 HB	404	}
	405	} else {
	406	/* Convert BE data from userland back to LE. */
	407	switch (run->mmio.len) {
8e5b26b5 AG	408	case 4: gpr = ld_le32((u32 *)run->mmio.data); break;
	409	case 2: gpr = ld_le16((u16 *)run->mmio.data); break;
	410	case 1: gpr = (u8 )run->mmio.data; break;
bbf45ba5 HB	411	}
bbf45ba5 HB	412	}
8e5b26b5	413
3587d534 AG	414	if (vcpu->arch.mmio_sign_extend) {
	415	switch (run->mmio.len) {
	416	#ifdef CONFIG_PPC64
	417	case 4:
	418	gpr = (s64)(s32)gpr;
	419	break;
	420	#endif
	421	case 2:
	422	gpr = (s64)(s16)gpr;
	423	break;
	424	case 1:
	425	gpr = (s64)(s8)gpr;
	426	break;
	427	}
	428	}
	429
8e5b26b5	430	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
b104d066 AG	431
	432	switch (vcpu->arch.io_gpr & KVM_REG_EXT_MASK) {
	433	case KVM_REG_GPR:
	434	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
	435	break;
	436	case KVM_REG_FPR:
	437	vcpu->arch.fpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
	438	break;
287d5611	439	#ifdef CONFIG_PPC_BOOK3S
b104d066 AG	440	case KVM_REG_QPR:
	441	vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
	442	break;
	443	case KVM_REG_FQPR:
	444	vcpu->arch.fpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
	445	vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_REG_MASK] = gpr;
	446	break;
287d5611	447	#endif
b104d066 AG	448	default:
	449	BUG();
	450	}
bbf45ba5 HB	451	}
	452
	453	int kvmppc_handle_load(struct kvm_run run, struct kvm_vcpu vcpu,
	454	unsigned int rt, unsigned int bytes, int is_bigendian)
	455	{
	456	if (bytes > sizeof(run->mmio.data)) {
	457	printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__,
	458	run->mmio.len);
	459	}
	460
	461	run->mmio.phys_addr = vcpu->arch.paddr_accessed;
	462	run->mmio.len = bytes;
	463	run->mmio.is_write = 0;
	464
	465	vcpu->arch.io_gpr = rt;
	466	vcpu->arch.mmio_is_bigendian = is_bigendian;
	467	vcpu->mmio_needed = 1;
	468	vcpu->mmio_is_write = 0;
3587d534	469	vcpu->arch.mmio_sign_extend = 0;
bbf45ba5 HB	470
	471	return EMULATE_DO_MMIO;
	472	}
	473
3587d534 AG	474	/* Same as above, but sign extends */
	475	int kvmppc_handle_loads(struct kvm_run run, struct kvm_vcpu vcpu,
	476	unsigned int rt, unsigned int bytes, int is_bigendian)
	477	{
	478	int r;
	479
	480	r = kvmppc_handle_load(run, vcpu, rt, bytes, is_bigendian);
	481	vcpu->arch.mmio_sign_extend = 1;
	482
	483	return r;
	484	}
	485
bbf45ba5	486	int kvmppc_handle_store(struct kvm_run run, struct kvm_vcpu vcpu,
b104d066	487	u64 val, unsigned int bytes, int is_bigendian)
bbf45ba5 HB	488	{
	489	void *data = run->mmio.data;
	490
	491	if (bytes > sizeof(run->mmio.data)) {
	492	printk(KERN_ERR "%s: bad MMIO length: %d\n", __func__,
	493	run->mmio.len);
	494	}
	495
	496	run->mmio.phys_addr = vcpu->arch.paddr_accessed;
	497	run->mmio.len = bytes;
	498	run->mmio.is_write = 1;
	499	vcpu->mmio_needed = 1;
	500	vcpu->mmio_is_write = 1;
	501
	502	/* Store the value at the lowest bytes in 'data'. */
	503	if (is_bigendian) {
	504	switch (bytes) {
b104d066	505	case 8: (u64 )data = val; break;
bbf45ba5 HB	506	case 4: (u32 )data = val; break;
	507	case 2: (u16 )data = val; break;
	508	case 1: (u8 )data = val; break;
	509	}
	510	} else {
	511	/* Store LE value into 'data'. */
	512	switch (bytes) {
	513	case 4: st_le32(data, val); break;
	514	case 2: st_le16(data, val); break;
	515	case 1: (u8 )data = val; break;
	516	}
	517	}
	518
	519	return EMULATE_DO_MMIO;
	520	}
	521
	522	int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu vcpu, struct kvm_run run)
	523	{
	524	int r;
	525	sigset_t sigsaved;
	526
	527	if (vcpu->sigset_active)
	528	sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
	529
	530	if (vcpu->mmio_needed) {
	531	if (!vcpu->mmio_is_write)
	532	kvmppc_complete_mmio_load(vcpu, run);
	533	vcpu->mmio_needed = 0;
	534	} else if (vcpu->arch.dcr_needed) {
	535	if (!vcpu->arch.dcr_is_write)
	536	kvmppc_complete_dcr_load(vcpu, run);
	537	vcpu->arch.dcr_needed = 0;
ad0a048b AG	538	} else if (vcpu->arch.osi_needed) {
	539	u64 *gprs = run->osi.gprs;
	540	int i;
	541
	542	for (i = 0; i < 32; i++)
	543	kvmppc_set_gpr(vcpu, i, gprs[i]);
	544	vcpu->arch.osi_needed = 0;
de56a948 PM	545	} else if (vcpu->arch.hcall_needed) {
	546	int i;
	547
	548	kvmppc_set_gpr(vcpu, 3, run->papr_hcall.ret);
	549	for (i = 0; i < 9; ++i)
	550	kvmppc_set_gpr(vcpu, 4 + i, run->papr_hcall.args[i]);
	551	vcpu->arch.hcall_needed = 0;
bbf45ba5 HB	552	}
bbf45ba5 HB	553
df6909e5	554	r = kvmppc_vcpu_run(run, vcpu);
bbf45ba5 HB	555
	556	if (vcpu->sigset_active)
	557	sigprocmask(SIG_SETMASK, &sigsaved, NULL);
	558
	559	return r;
	560	}
	561
dfd4d47e SW	562	void kvm_vcpu_kick(struct kvm_vcpu *vcpu)
dfd4d47e SW	563	{
ae21216b AG	564	int me;
	565	int cpu = vcpu->cpu;
	566
	567	me = get_cpu();
4e72dbe1	568	if (waitqueue_active(vcpu->arch.wqp)) {
dfd4d47e SW	569	wake_up_interruptible(vcpu->arch.wqp);
dfd4d47e SW	570	vcpu->stat.halt_wakeup++;
ae21216b	571	} else if (cpu != me && cpu != -1) {
dfd4d47e SW	572	smp_send_reschedule(vcpu->cpu);
dfd4d47e SW	573	}
ae21216b	574	put_cpu();
dfd4d47e SW	575	}
dfd4d47e SW	576
bbf45ba5 HB	577	int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu vcpu, struct kvm_interrupt irq)
bbf45ba5 HB	578	{
19ccb76a	579	if (irq->irq == KVM_INTERRUPT_UNSET) {
18978768	580	kvmppc_core_dequeue_external(vcpu, irq);
19ccb76a PM	581	return 0;
	582	}
	583
	584	kvmppc_core_queue_external(vcpu, irq);
dfd4d47e	585	kvm_vcpu_kick(vcpu);
45c5eb67	586
bbf45ba5 HB	587	return 0;
	588	}
	589
71fbfd5f AG	590	static int kvm_vcpu_ioctl_enable_cap(struct kvm_vcpu *vcpu,
	591	struct kvm_enable_cap *cap)
	592	{
	593	int r;
	594
	595	if (cap->flags)
	596	return -EINVAL;
	597
	598	switch (cap->cap) {
ad0a048b AG	599	case KVM_CAP_PPC_OSI:
	600	r = 0;
	601	vcpu->arch.osi_enabled = true;
	602	break;
930b412a AG	603	case KVM_CAP_PPC_PAPR:
	604	r = 0;
	605	vcpu->arch.papr_enabled = true;
	606	break;
dc83b8bc SW	607	#ifdef CONFIG_KVM_E500
	608	case KVM_CAP_SW_TLB: {
	609	struct kvm_config_tlb cfg;
	610	void __user user_ptr = (void __user )(uintptr_t)cap->args[0];
	611
	612	r = -EFAULT;
	613	if (copy_from_user(&cfg, user_ptr, sizeof(cfg)))
	614	break;
	615
	616	r = kvm_vcpu_ioctl_config_tlb(vcpu, &cfg);
	617	break;
	618	}
	619	#endif
71fbfd5f AG	620	default:
	621	r = -EINVAL;
	622	break;
	623	}
	624
af8f38b3 AG	625	if (!r)
	626	r = kvmppc_sanity_check(vcpu);
	627
71fbfd5f AG	628	return r;
	629	}
	630
bbf45ba5 HB	631	int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
	632	struct kvm_mp_state *mp_state)
	633	{
	634	return -EINVAL;
	635	}
	636
	637	int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
	638	struct kvm_mp_state *mp_state)
	639	{
	640	return -EINVAL;
	641	}
	642
	643	long kvm_arch_vcpu_ioctl(struct file *filp,
	644	unsigned int ioctl, unsigned long arg)
	645	{
	646	struct kvm_vcpu *vcpu = filp->private_data;
	647	void __user argp = (void __user )arg;
	648	long r;
	649
93736624 AK	650	switch (ioctl) {
93736624 AK	651	case KVM_INTERRUPT: {
bbf45ba5 HB	652	struct kvm_interrupt irq;
	653	r = -EFAULT;
	654	if (copy_from_user(&irq, argp, sizeof(irq)))
93736624	655	goto out;
bbf45ba5	656	r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
93736624	657	goto out;
bbf45ba5	658	}
19483d14	659
71fbfd5f AG	660	case KVM_ENABLE_CAP:
	661	{
	662	struct kvm_enable_cap cap;
	663	r = -EFAULT;
	664	if (copy_from_user(&cap, argp, sizeof(cap)))
	665	goto out;
	666	r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
	667	break;
	668	}
dc83b8bc SW	669
	670	#ifdef CONFIG_KVM_E500
	671	case KVM_DIRTY_TLB: {
	672	struct kvm_dirty_tlb dirty;
	673	r = -EFAULT;
	674	if (copy_from_user(&dirty, argp, sizeof(dirty)))
	675	goto out;
	676	r = kvm_vcpu_ioctl_dirty_tlb(vcpu, &dirty);
	677	break;
	678	}
	679	#endif
	680
bbf45ba5 HB	681	default:
	682	r = -EINVAL;
	683	}
	684
	685	out:
	686	return r;
	687	}
	688
5b1c1493 CO	689	int kvm_arch_vcpu_fault(struct kvm_vcpu vcpu, struct vm_fault vmf)
	690	{
	691	return VM_FAULT_SIGBUS;
	692	}
	693
15711e9c AG	694	static int kvm_vm_ioctl_get_pvinfo(struct kvm_ppc_pvinfo *pvinfo)
	695	{
	696	u32 inst_lis = 0x3c000000;
	697	u32 inst_ori = 0x60000000;
	698	u32 inst_nop = 0x60000000;
	699	u32 inst_sc = 0x44000002;
	700	u32 inst_imm_mask = 0xffff;
	701
	702	/*
	703	* The hypercall to get into KVM from within guest context is as
	704	* follows:
	705	*
	706	* lis r0, r0, KVM_SC_MAGIC_R0@h
	707	* ori r0, KVM_SC_MAGIC_R0@l
	708	* sc
	709	* nop
	710	*/
	711	pvinfo->hcall[0] = inst_lis \| ((KVM_SC_MAGIC_R0 >> 16) & inst_imm_mask);
	712	pvinfo->hcall[1] = inst_ori \| (KVM_SC_MAGIC_R0 & inst_imm_mask);
	713	pvinfo->hcall[2] = inst_sc;
	714	pvinfo->hcall[3] = inst_nop;
	715
	716	return 0;
	717	}
	718
bbf45ba5 HB	719	long kvm_arch_vm_ioctl(struct file *filp,
	720	unsigned int ioctl, unsigned long arg)
	721	{
15711e9c	722	void __user argp = (void __user )arg;
bbf45ba5 HB	723	long r;
	724
	725	switch (ioctl) {
15711e9c AG	726	case KVM_PPC_GET_PVINFO: {
15711e9c AG	727	struct kvm_ppc_pvinfo pvinfo;
d8cdddcd	728	memset(&pvinfo, 0, sizeof(pvinfo));
15711e9c AG	729	r = kvm_vm_ioctl_get_pvinfo(&pvinfo);
	730	if (copy_to_user(argp, &pvinfo, sizeof(pvinfo))) {
	731	r = -EFAULT;
	732	goto out;
	733	}
	734
	735	break;
	736	}
54738c09 DG	737	#ifdef CONFIG_KVM_BOOK3S_64_HV
	738	case KVM_CREATE_SPAPR_TCE: {
	739	struct kvm_create_spapr_tce create_tce;
	740	struct kvm *kvm = filp->private_data;
	741
	742	r = -EFAULT;
	743	if (copy_from_user(&create_tce, argp, sizeof(create_tce)))
	744	goto out;
	745	r = kvm_vm_ioctl_create_spapr_tce(kvm, &create_tce);
	746	goto out;
	747	}
aa04b4cc PM	748
	749	case KVM_ALLOCATE_RMA: {
	750	struct kvm *kvm = filp->private_data;
	751	struct kvm_allocate_rma rma;
	752
	753	r = kvm_vm_ioctl_allocate_rma(kvm, &rma);
	754	if (r >= 0 && copy_to_user(argp, &rma, sizeof(rma)))
	755	r = -EFAULT;
	756	break;
	757	}
54738c09 DG	758	#endif /* CONFIG_KVM_BOOK3S_64_HV */
54738c09 DG	759
bbf45ba5	760	default:
367e1319	761	r = -ENOTTY;
bbf45ba5 HB	762	}
bbf45ba5 HB	763
15711e9c	764	out:
bbf45ba5 HB	765	return r;
	766	}
	767
	768	int kvm_arch_init(void *opaque)
	769	{
	770	return 0;
	771	}
	772
	773	void kvm_arch_exit(void)
	774	{
	775	}