powerpc.c 31.2 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
/*
 * 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>
A
Alexander Graf 已提交
25
#include <linux/hrtimer.h>
26
#include <linux/fs.h>
27
#include <linux/slab.h>
S
Scott Wood 已提交
28
#include <linux/file.h>
29
#include <linux/module.h>
30 31 32
#include <asm/cputable.h>
#include <asm/uaccess.h>
#include <asm/kvm_ppc.h>
33
#include <asm/tlbflush.h>
34
#include <asm/cputhreads.h>
35
#include <asm/irqflags.h>
36
#include "timing.h"
37
#include "irq.h"
P
Paul Mackerras 已提交
38
#include "../mm/mmu_decl.h"
39

40 41 42
#define CREATE_TRACE_POINTS
#include "trace.h"

43 44 45 46 47
struct kvmppc_ops *kvmppc_hv_ops;
EXPORT_SYMBOL_GPL(kvmppc_hv_ops);
struct kvmppc_ops *kvmppc_pr_ops;
EXPORT_SYMBOL_GPL(kvmppc_pr_ops);

48

49 50
int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
{
51
	return !!(v->arch.pending_exceptions) ||
52
	       v->requests;
53 54
}

55 56 57 58 59
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu)
{
	return 1;
}

60 61 62 63
/*
 * Common checks before entering the guest world.  Call with interrupts
 * disabled.
 *
64 65 66 67
 * returns:
 *
 * == 1 if we're ready to go into guest state
 * <= 0 if we need to go back to the host with return value
68 69 70
 */
int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)
{
S
Scott Wood 已提交
71 72 73 74
	int r;

	WARN_ON(irqs_disabled());
	hard_irq_disable();
75 76 77 78 79

	while (true) {
		if (need_resched()) {
			local_irq_enable();
			cond_resched();
S
Scott Wood 已提交
80
			hard_irq_disable();
81 82 83 84
			continue;
		}

		if (signal_pending(current)) {
85 86 87
			kvmppc_account_exit(vcpu, SIGNAL_EXITS);
			vcpu->run->exit_reason = KVM_EXIT_INTR;
			r = -EINTR;
88 89 90
			break;
		}

91 92 93 94 95 96 97 98
		vcpu->mode = IN_GUEST_MODE;

		/*
		 * Reading vcpu->requests must happen after setting vcpu->mode,
		 * so we don't miss a request because the requester sees
		 * OUTSIDE_GUEST_MODE and assumes we'll be checking requests
		 * before next entering the guest (and thus doesn't IPI).
		 */
99
		smp_mb();
100

101 102 103 104
		if (vcpu->requests) {
			/* Make sure we process requests preemptable */
			local_irq_enable();
			trace_kvm_check_requests(vcpu);
105
			r = kvmppc_core_check_requests(vcpu);
S
Scott Wood 已提交
106
			hard_irq_disable();
107 108 109
			if (r > 0)
				continue;
			break;
110 111 112 113 114 115 116 117
		}

		if (kvmppc_core_prepare_to_enter(vcpu)) {
			/* interrupts got enabled in between, so we
			   are back at square 1 */
			continue;
		}

118
		__kvm_guest_enter();
S
Scott Wood 已提交
119
		return 1;
120 121
	}

S
Scott Wood 已提交
122 123
	/* return to host */
	local_irq_enable();
124 125
	return r;
}
126
EXPORT_SYMBOL_GPL(kvmppc_prepare_to_enter);
127

128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
#if defined(CONFIG_PPC_BOOK3S_64) && defined(CONFIG_KVM_BOOK3S_PR_POSSIBLE)
static void kvmppc_swab_shared(struct kvm_vcpu *vcpu)
{
	struct kvm_vcpu_arch_shared *shared = vcpu->arch.shared;
	int i;

	shared->sprg0 = swab64(shared->sprg0);
	shared->sprg1 = swab64(shared->sprg1);
	shared->sprg2 = swab64(shared->sprg2);
	shared->sprg3 = swab64(shared->sprg3);
	shared->srr0 = swab64(shared->srr0);
	shared->srr1 = swab64(shared->srr1);
	shared->dar = swab64(shared->dar);
	shared->msr = swab64(shared->msr);
	shared->dsisr = swab32(shared->dsisr);
	shared->int_pending = swab32(shared->int_pending);
	for (i = 0; i < ARRAY_SIZE(shared->sr); i++)
		shared->sr[i] = swab32(shared->sr[i]);
}
#endif

149 150 151 152 153 154 155 156 157 158
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;

159
	if (!(kvmppc_get_msr(vcpu) & MSR_SF)) {
160 161 162 163 164 165 166 167
		/* 32 bit mode */
		param1 &= 0xffffffff;
		param2 &= 0xffffffff;
		param3 &= 0xffffffff;
		param4 &= 0xffffffff;
	}

	switch (nr) {
168
	case KVM_HCALL_TOKEN(KVM_HC_PPC_MAP_MAGIC_PAGE):
169
	{
170 171 172 173 174 175 176 177 178 179
#if defined(CONFIG_PPC_BOOK3S_64) && defined(CONFIG_KVM_BOOK3S_PR_POSSIBLE)
		/* Book3S can be little endian, find it out here */
		int shared_big_endian = true;
		if (vcpu->arch.intr_msr & MSR_LE)
			shared_big_endian = false;
		if (shared_big_endian != vcpu->arch.shared_big_endian)
			kvmppc_swab_shared(vcpu);
		vcpu->arch.shared_big_endian = shared_big_endian;
#endif

180 181 182 183 184 185 186 187 188 189 190 191
		if (!(param2 & MAGIC_PAGE_FLAG_NOT_MAPPED_NX)) {
			/*
			 * Older versions of the Linux magic page code had
			 * a bug where they would map their trampoline code
			 * NX. If that's the case, remove !PR NX capability.
			 */
			vcpu->arch.disable_kernel_nx = true;
			kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu);
		}

		vcpu->arch.magic_page_pa = param1 & ~0xfffULL;
		vcpu->arch.magic_page_ea = param2 & ~0xfffULL;
192

193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
#ifdef CONFIG_PPC_64K_PAGES
		/*
		 * Make sure our 4k magic page is in the same window of a 64k
		 * page within the guest and within the host's page.
		 */
		if ((vcpu->arch.magic_page_pa & 0xf000) !=
		    ((ulong)vcpu->arch.shared & 0xf000)) {
			void *old_shared = vcpu->arch.shared;
			ulong shared = (ulong)vcpu->arch.shared;
			void *new_shared;

			shared &= PAGE_MASK;
			shared |= vcpu->arch.magic_page_pa & 0xf000;
			new_shared = (void*)shared;
			memcpy(new_shared, old_shared, 0x1000);
			vcpu->arch.shared = new_shared;
		}
#endif

212
		r2 = KVM_MAGIC_FEAT_SR | KVM_MAGIC_FEAT_MAS0_TO_SPRG7;
213

214
		r = EV_SUCCESS;
215 216
		break;
	}
217 218
	case KVM_HCALL_TOKEN(KVM_HC_FEATURES):
		r = EV_SUCCESS;
219
#if defined(CONFIG_PPC_BOOK3S) || defined(CONFIG_KVM_E500V2)
220 221
		r2 |= (1 << KVM_FEATURE_MAGIC_PAGE);
#endif
222 223 224

		/* Second return value is in r4 */
		break;
225 226 227 228 229
	case EV_HCALL_TOKEN(EV_IDLE):
		r = EV_SUCCESS;
		kvm_vcpu_block(vcpu);
		clear_bit(KVM_REQ_UNHALT, &vcpu->requests);
		break;
230
	default:
231
		r = EV_UNIMPLEMENTED;
232 233 234
		break;
	}

235 236
	kvmppc_set_gpr(vcpu, 4, r2);

237 238
	return r;
}
239
EXPORT_SYMBOL_GPL(kvmppc_kvm_pv);
240

241 242 243 244 245 246 247 248 249 250 251 252 253
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;

	/* HV KVM can only do PAPR mode for now */
254
	if (!vcpu->arch.papr_enabled && is_kvmppc_hv_enabled(vcpu->kvm))
255 256
		goto out;

257 258 259 260 261
#ifdef CONFIG_KVM_BOOKE_HV
	if (!cpu_has_feature(CPU_FTR_EMB_HV))
		goto out;
#endif

262 263 264 265 266 267
	r = true;

out:
	vcpu->arch.sane = r;
	return r ? 0 : -EINVAL;
}
268
EXPORT_SYMBOL_GPL(kvmppc_sanity_check);
269

270 271 272 273 274
int kvmppc_emulate_mmio(struct kvm_run *run, struct kvm_vcpu *vcpu)
{
	enum emulation_result er;
	int r;

275
	er = kvmppc_emulate_loadstore(vcpu);
276 277 278 279 280 281
	switch (er) {
	case EMULATE_DONE:
		/* Future optimization: only reload non-volatiles if they were
		 * actually modified. */
		r = RESUME_GUEST_NV;
		break;
282 283 284
	case EMULATE_AGAIN:
		r = RESUME_GUEST;
		break;
285 286 287 288 289 290 291 292 293
	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:
294 295 296
	{
		u32 last_inst;

297
		kvmppc_get_last_inst(vcpu, INST_GENERIC, &last_inst);
298
		/* XXX Deliver Program interrupt to guest. */
299
		pr_emerg("%s: emulation failed (%08x)\n", __func__, last_inst);
300 301
		r = RESUME_HOST;
		break;
302
	}
303
	default:
304 305
		WARN_ON(1);
		r = RESUME_GUEST;
306 307 308 309
	}

	return r;
}
310
EXPORT_SYMBOL_GPL(kvmppc_emulate_mmio);
311

312 313 314
int kvmppc_st(struct kvm_vcpu *vcpu, ulong *eaddr, int size, void *ptr,
	      bool data)
{
315
	ulong mp_pa = vcpu->arch.magic_page_pa & KVM_PAM & PAGE_MASK;
316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
	struct kvmppc_pte pte;
	int r;

	vcpu->stat.st++;

	r = kvmppc_xlate(vcpu, *eaddr, data ? XLATE_DATA : XLATE_INST,
			 XLATE_WRITE, &pte);
	if (r < 0)
		return r;

	*eaddr = pte.raddr;

	if (!pte.may_write)
		return -EPERM;

331 332 333 334 335 336 337 338 339 340
	/* Magic page override */
	if (kvmppc_supports_magic_page(vcpu) && mp_pa &&
	    ((pte.raddr & KVM_PAM & PAGE_MASK) == mp_pa) &&
	    !(kvmppc_get_msr(vcpu) & MSR_PR)) {
		void *magic = vcpu->arch.shared;
		magic += pte.eaddr & 0xfff;
		memcpy(magic, ptr, size);
		return EMULATE_DONE;
	}

341 342 343 344 345 346 347 348 349 350
	if (kvm_write_guest(vcpu->kvm, pte.raddr, ptr, size))
		return EMULATE_DO_MMIO;

	return EMULATE_DONE;
}
EXPORT_SYMBOL_GPL(kvmppc_st);

int kvmppc_ld(struct kvm_vcpu *vcpu, ulong *eaddr, int size, void *ptr,
		      bool data)
{
351
	ulong mp_pa = vcpu->arch.magic_page_pa & KVM_PAM & PAGE_MASK;
352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369
	struct kvmppc_pte pte;
	int rc;

	vcpu->stat.ld++;

	rc = kvmppc_xlate(vcpu, *eaddr, data ? XLATE_DATA : XLATE_INST,
			  XLATE_READ, &pte);
	if (rc)
		return rc;

	*eaddr = pte.raddr;

	if (!pte.may_read)
		return -EPERM;

	if (!data && !pte.may_execute)
		return -ENOEXEC;

370 371 372 373 374 375 376 377 378 379
	/* Magic page override */
	if (kvmppc_supports_magic_page(vcpu) && mp_pa &&
	    ((pte.raddr & KVM_PAM & PAGE_MASK) == mp_pa) &&
	    !(kvmppc_get_msr(vcpu) & MSR_PR)) {
		void *magic = vcpu->arch.shared;
		magic += pte.eaddr & 0xfff;
		memcpy(ptr, magic, size);
		return EMULATE_DONE;
	}

380 381
	if (kvm_read_guest(vcpu->kvm, pte.raddr, ptr, size))
		return EMULATE_DO_MMIO;
382 383 384 385 386

	return EMULATE_DONE;
}
EXPORT_SYMBOL_GPL(kvmppc_ld);

387
int kvm_arch_hardware_enable(void)
388
{
389
	return 0;
390 391 392 393 394 395 396 397 398
}

int kvm_arch_hardware_setup(void)
{
	return 0;
}

void kvm_arch_check_processor_compat(void *rtn)
{
399
	*(int *)rtn = kvmppc_core_check_processor_compat();
400 401
}

402
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
403
{
404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
	struct kvmppc_ops *kvm_ops = NULL;
	/*
	 * if we have both HV and PR enabled, default is HV
	 */
	if (type == 0) {
		if (kvmppc_hv_ops)
			kvm_ops = kvmppc_hv_ops;
		else
			kvm_ops = kvmppc_pr_ops;
		if (!kvm_ops)
			goto err_out;
	} else	if (type == KVM_VM_PPC_HV) {
		if (!kvmppc_hv_ops)
			goto err_out;
		kvm_ops = kvmppc_hv_ops;
	} else if (type == KVM_VM_PPC_PR) {
		if (!kvmppc_pr_ops)
			goto err_out;
		kvm_ops = kvmppc_pr_ops;
	} else
		goto err_out;

	if (kvm_ops->owner && !try_module_get(kvm_ops->owner))
		return -ENOENT;

	kvm->arch.kvm_ops = kvm_ops;
430
	return kvmppc_core_init_vm(kvm);
431 432
err_out:
	return -EINVAL;
433 434
}

435
void kvm_arch_destroy_vm(struct kvm *kvm)
436 437
{
	unsigned int i;
438
	struct kvm_vcpu *vcpu;
439

440 441 442 443 444 445 446 447 448 449
#ifdef CONFIG_KVM_XICS
	/*
	 * We call kick_all_cpus_sync() to ensure that all
	 * CPUs have executed any pending IPIs before we
	 * continue and free VCPUs structures below.
	 */
	if (is_kvmppc_hv_enabled(kvm))
		kick_all_cpus_sync();
#endif

450 451 452 453 454 455 456 457
	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);
458 459 460

	kvmppc_core_destroy_vm(kvm);

461
	mutex_unlock(&kvm->lock);
462 463 464

	/* drop the module reference */
	module_put(kvm->arch.kvm_ops->owner);
465 466
}

467
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
468 469
{
	int r;
470
	/* Assume we're using HV mode when the HV module is loaded */
471
	int hv_enabled = kvmppc_hv_ops ? 1 : 0;
472

473 474 475 476 477 478 479 480
	if (kvm) {
		/*
		 * Hooray - we know which VM type we're running on. Depend on
		 * that rather than the guess above.
		 */
		hv_enabled = is_kvmppc_hv_enabled(kvm);
	}

481
	switch (ext) {
S
Scott Wood 已提交
482 483
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_SREGS:
484
	case KVM_CAP_PPC_BOOKE_WATCHDOG:
485
	case KVM_CAP_PPC_EPR:
S
Scott Wood 已提交
486
#else
487
	case KVM_CAP_PPC_SEGSTATE:
488
	case KVM_CAP_PPC_HIOR:
489
	case KVM_CAP_PPC_PAPR:
S
Scott Wood 已提交
490
#endif
491
	case KVM_CAP_PPC_UNSET_IRQ:
492
	case KVM_CAP_PPC_IRQ_LEVEL:
493
	case KVM_CAP_ENABLE_CAP:
494
	case KVM_CAP_ENABLE_CAP_VM:
495
	case KVM_CAP_ONE_REG:
A
Alexander Graf 已提交
496
	case KVM_CAP_IOEVENTFD:
497
	case KVM_CAP_DEVICE_CTRL:
498 499 500
		r = 1;
		break;
	case KVM_CAP_PPC_PAIRED_SINGLES:
501
	case KVM_CAP_PPC_OSI:
502
	case KVM_CAP_PPC_GET_PVINFO:
503
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
504
	case KVM_CAP_SW_TLB:
S
Scott Wood 已提交
505
#endif
506
		/* We support this only for PR */
507
		r = !hv_enabled;
508
		break;
509
#ifdef CONFIG_KVM_MMIO
510 511 512
	case KVM_CAP_COALESCED_MMIO:
		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
		break;
513
#endif
514 515 516 517 518 519
#ifdef CONFIG_KVM_MPIC
	case KVM_CAP_IRQ_MPIC:
		r = 1;
		break;
#endif

520
#ifdef CONFIG_PPC_BOOK3S_64
521
	case KVM_CAP_SPAPR_TCE:
522
	case KVM_CAP_PPC_ALLOC_HTAB:
523
	case KVM_CAP_PPC_RTAS:
524
	case KVM_CAP_PPC_FIXUP_HCALL:
525
	case KVM_CAP_PPC_ENABLE_HCALL:
526 527 528
#ifdef CONFIG_KVM_XICS
	case KVM_CAP_IRQ_XICS:
#endif
529 530
		r = 1;
		break;
531
#endif /* CONFIG_PPC_BOOK3S_64 */
532
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
533
	case KVM_CAP_PPC_SMT:
534
		if (hv_enabled)
535
			r = threads_per_subcore;
536 537
		else
			r = 0;
538
		break;
539
	case KVM_CAP_PPC_RMA:
540
		r = 0;
541
		break;
542 543 544
	case KVM_CAP_PPC_HWRNG:
		r = kvmppc_hwrng_present();
		break;
545
#endif
546
	case KVM_CAP_SYNC_MMU:
547
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
548
		r = hv_enabled;
549 550 551 552
#elif defined(KVM_ARCH_WANT_MMU_NOTIFIER)
		r = 1;
#else
		r = 0;
553
#endif
554 555
		break;
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
556
	case KVM_CAP_PPC_HTAB_FD:
557
		r = hv_enabled;
558
		break;
559
#endif
560 561 562 563 564 565 566
	case KVM_CAP_NR_VCPUS:
		/*
		 * Recommending a number of CPUs is somewhat arbitrary; we
		 * return the number of present CPUs for -HV (since a host
		 * will have secondary threads "offline"), and for other KVM
		 * implementations just count online CPUs.
		 */
567
		if (hv_enabled)
568 569 570
			r = num_present_cpus();
		else
			r = num_online_cpus();
571
		break;
572 573 574
	case KVM_CAP_NR_MEMSLOTS:
		r = KVM_USER_MEM_SLOTS;
		break;
575 576 577
	case KVM_CAP_MAX_VCPUS:
		r = KVM_MAX_VCPUS;
		break;
578 579 580 581
#ifdef CONFIG_PPC_BOOK3S_64
	case KVM_CAP_PPC_GET_SMMU_INFO:
		r = 1;
		break;
582 583 584
	case KVM_CAP_SPAPR_MULTITCE:
		r = 1;
		break;
585
#endif
586 587 588 589 590 591 592 593 594 595 596 597 598 599
	default:
		r = 0;
		break;
	}
	return r;

}

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

600
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
601 602
			   struct kvm_memory_slot *dont)
{
603
	kvmppc_core_free_memslot(kvm, free, dont);
604 605
}

606 607
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages)
608
{
609
	return kvmppc_core_create_memslot(kvm, slot, npages);
610 611
}

612
int kvm_arch_prepare_memory_region(struct kvm *kvm,
613
				   struct kvm_memory_slot *memslot,
614
				   const struct kvm_userspace_memory_region *mem,
615
				   enum kvm_mr_change change)
616
{
617
	return kvmppc_core_prepare_memory_region(kvm, memslot, mem);
618 619
}

620
void kvm_arch_commit_memory_region(struct kvm *kvm,
621
				   const struct kvm_userspace_memory_region *mem,
622
				   const struct kvm_memory_slot *old,
623
				   const struct kvm_memory_slot *new,
624
				   enum kvm_mr_change change)
625
{
626
	kvmppc_core_commit_memory_region(kvm, mem, old, new);
627 628
}

629 630
void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
				   struct kvm_memory_slot *slot)
631
{
632
	kvmppc_core_flush_memslot(kvm, slot);
633 634
}

635 636
struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
{
637 638
	struct kvm_vcpu *vcpu;
	vcpu = kvmppc_core_vcpu_create(kvm, id);
639 640
	if (!IS_ERR(vcpu)) {
		vcpu->arch.wqp = &vcpu->wq;
641
		kvmppc_create_vcpu_debugfs(vcpu, id);
642
	}
643
	return vcpu;
644 645
}

646
void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
647 648 649
{
}

650 651
void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
652 653 654
	/* Make sure we're not using the vcpu anymore */
	hrtimer_cancel(&vcpu->arch.dec_timer);

655
	kvmppc_remove_vcpu_debugfs(vcpu);
S
Scott Wood 已提交
656 657 658 659 660

	switch (vcpu->arch.irq_type) {
	case KVMPPC_IRQ_MPIC:
		kvmppc_mpic_disconnect_vcpu(vcpu->arch.mpic, vcpu);
		break;
661 662 663
	case KVMPPC_IRQ_XICS:
		kvmppc_xics_free_icp(vcpu);
		break;
S
Scott Wood 已提交
664 665
	}

666
	kvmppc_core_vcpu_free(vcpu);
667 668 669 670 671 672 673 674 675
}

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

int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
{
676
	return kvmppc_core_pending_dec(vcpu);
677 678
}

T
Thomas Huth 已提交
679
static enum hrtimer_restart kvmppc_decrementer_wakeup(struct hrtimer *timer)
A
Alexander Graf 已提交
680 681 682 683
{
	struct kvm_vcpu *vcpu;

	vcpu = container_of(timer, struct kvm_vcpu, arch.dec_timer);
684
	kvmppc_decrementer_func(vcpu);
A
Alexander Graf 已提交
685 686 687 688

	return HRTIMER_NORESTART;
}

689 690
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
691 692
	int ret;

A
Alexander Graf 已提交
693 694
	hrtimer_init(&vcpu->arch.dec_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
	vcpu->arch.dec_timer.function = kvmppc_decrementer_wakeup;
695
	vcpu->arch.dec_expires = ~(u64)0;
696

697 698 699
#ifdef CONFIG_KVM_EXIT_TIMING
	mutex_init(&vcpu->arch.exit_timing_lock);
#endif
700 701
	ret = kvmppc_subarch_vcpu_init(vcpu);
	return ret;
702 703 704 705
}

void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
706
	kvmppc_mmu_destroy(vcpu);
707
	kvmppc_subarch_vcpu_uninit(vcpu);
708 709 710 711
}

void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
712 713 714 715 716 717 718 719 720 721
#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
722
	kvmppc_core_vcpu_load(vcpu, cpu);
723 724 725 726
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
727
	kvmppc_core_vcpu_put(vcpu);
728 729 730
#ifdef CONFIG_BOOKE
	vcpu->arch.vrsave = mfspr(SPRN_VRSAVE);
#endif
731 732 733 734 735
}

static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
                                      struct kvm_run *run)
{
736
	u64 uninitialized_var(gpr);
737

738
	if (run->mmio.len > sizeof(gpr)) {
739 740 741 742
		printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
		return;
	}

743
	if (!vcpu->arch.mmio_host_swabbed) {
744
		switch (run->mmio.len) {
745
		case 8: gpr = *(u64 *)run->mmio.data; break;
746 747 748
		case 4: gpr = *(u32 *)run->mmio.data; break;
		case 2: gpr = *(u16 *)run->mmio.data; break;
		case 1: gpr = *(u8 *)run->mmio.data; break;
749 750 751
		}
	} else {
		switch (run->mmio.len) {
752 753 754
		case 8: gpr = swab64(*(u64 *)run->mmio.data); break;
		case 4: gpr = swab32(*(u32 *)run->mmio.data); break;
		case 2: gpr = swab16(*(u16 *)run->mmio.data); break;
755
		case 1: gpr = *(u8 *)run->mmio.data; break;
756 757
		}
	}
758

A
Alexander Graf 已提交
759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774
	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;
		}
	}

775
	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
776

777 778
	switch (vcpu->arch.io_gpr & KVM_MMIO_REG_EXT_MASK) {
	case KVM_MMIO_REG_GPR:
779 780
		kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
		break;
781
	case KVM_MMIO_REG_FPR:
782
		VCPU_FPR(vcpu, vcpu->arch.io_gpr & KVM_MMIO_REG_MASK) = gpr;
783
		break;
784
#ifdef CONFIG_PPC_BOOK3S
785 786
	case KVM_MMIO_REG_QPR:
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
787
		break;
788
	case KVM_MMIO_REG_FQPR:
789
		VCPU_FPR(vcpu, vcpu->arch.io_gpr & KVM_MMIO_REG_MASK) = gpr;
790
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
791
		break;
792
#endif
793 794 795
	default:
		BUG();
	}
796 797 798
}

int kvmppc_handle_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
799 800
		       unsigned int rt, unsigned int bytes,
		       int is_default_endian)
801
{
802
	int idx, ret;
803
	bool host_swabbed;
804

805
	/* Pity C doesn't have a logical XOR operator */
806
	if (kvmppc_need_byteswap(vcpu)) {
807
		host_swabbed = is_default_endian;
808
	} else {
809
		host_swabbed = !is_default_endian;
810
	}
811

812 813 814 815 816 817 818 819 820 821
	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;
822
	vcpu->arch.mmio_host_swabbed = host_swabbed;
823 824
	vcpu->mmio_needed = 1;
	vcpu->mmio_is_write = 0;
A
Alexander Graf 已提交
825
	vcpu->arch.mmio_sign_extend = 0;
826

827 828
	idx = srcu_read_lock(&vcpu->kvm->srcu);

829
	ret = kvm_io_bus_read(vcpu, KVM_MMIO_BUS, run->mmio.phys_addr,
830 831 832 833 834
			      bytes, &run->mmio.data);

	srcu_read_unlock(&vcpu->kvm->srcu, idx);

	if (!ret) {
A
Alexander Graf 已提交
835 836 837 838 839
		kvmppc_complete_mmio_load(vcpu, run);
		vcpu->mmio_needed = 0;
		return EMULATE_DONE;
	}

840 841
	return EMULATE_DO_MMIO;
}
842
EXPORT_SYMBOL_GPL(kvmppc_handle_load);
843

A
Alexander Graf 已提交
844 845
/* Same as above, but sign extends */
int kvmppc_handle_loads(struct kvm_run *run, struct kvm_vcpu *vcpu,
846 847
			unsigned int rt, unsigned int bytes,
			int is_default_endian)
A
Alexander Graf 已提交
848 849 850 851
{
	int r;

	vcpu->arch.mmio_sign_extend = 1;
852
	r = kvmppc_handle_load(run, vcpu, rt, bytes, is_default_endian);
A
Alexander Graf 已提交
853 854 855 856

	return r;
}

857
int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
858
			u64 val, unsigned int bytes, int is_default_endian)
859 860
{
	void *data = run->mmio.data;
861
	int idx, ret;
862
	bool host_swabbed;
863

864
	/* Pity C doesn't have a logical XOR operator */
865
	if (kvmppc_need_byteswap(vcpu)) {
866
		host_swabbed = is_default_endian;
867
	} else {
868
		host_swabbed = !is_default_endian;
869
	}
870 871 872 873 874 875 876 877 878 879 880 881 882

	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'. */
883
	if (!host_swabbed) {
884
		switch (bytes) {
885
		case 8: *(u64 *)data = val; break;
886 887 888 889 890 891
		case 4: *(u32 *)data = val; break;
		case 2: *(u16 *)data = val; break;
		case 1: *(u8  *)data = val; break;
		}
	} else {
		switch (bytes) {
892 893 894 895
		case 8: *(u64 *)data = swab64(val); break;
		case 4: *(u32 *)data = swab32(val); break;
		case 2: *(u16 *)data = swab16(val); break;
		case 1: *(u8  *)data = val; break;
896 897 898
		}
	}

899 900
	idx = srcu_read_lock(&vcpu->kvm->srcu);

901
	ret = kvm_io_bus_write(vcpu, KVM_MMIO_BUS, run->mmio.phys_addr,
902 903 904 905 906
			       bytes, &run->mmio.data);

	srcu_read_unlock(&vcpu->kvm->srcu, idx);

	if (!ret) {
A
Alexander Graf 已提交
907 908 909 910
		vcpu->mmio_needed = 0;
		return EMULATE_DONE;
	}

911 912
	return EMULATE_DO_MMIO;
}
913
EXPORT_SYMBOL_GPL(kvmppc_handle_store);
914

915 916 917 918 919 920 921 922 923 924 925 926 927 928
int kvm_vcpu_ioctl_get_one_reg(struct kvm_vcpu *vcpu, struct kvm_one_reg *reg)
{
	int r = 0;
	union kvmppc_one_reg val;
	int size;

	size = one_reg_size(reg->id);
	if (size > sizeof(val))
		return -EINVAL;

	r = kvmppc_get_one_reg(vcpu, reg->id, &val);
	if (r == -EINVAL) {
		r = 0;
		switch (reg->id) {
929 930 931 932 933 934
#ifdef CONFIG_ALTIVEC
		case KVM_REG_PPC_VR0 ... KVM_REG_PPC_VR31:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
935
			val.vval = vcpu->arch.vr.vr[reg->id - KVM_REG_PPC_VR0];
936 937 938 939 940 941
			break;
		case KVM_REG_PPC_VSCR:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
942
			val = get_reg_val(reg->id, vcpu->arch.vr.vscr.u[3]);
943 944
			break;
		case KVM_REG_PPC_VRSAVE:
945
			val = get_reg_val(reg->id, vcpu->arch.vrsave);
946 947
			break;
#endif /* CONFIG_ALTIVEC */
948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979
		default:
			r = -EINVAL;
			break;
		}
	}

	if (r)
		return r;

	if (copy_to_user((char __user *)(unsigned long)reg->addr, &val, size))
		r = -EFAULT;

	return r;
}

int kvm_vcpu_ioctl_set_one_reg(struct kvm_vcpu *vcpu, struct kvm_one_reg *reg)
{
	int r;
	union kvmppc_one_reg val;
	int size;

	size = one_reg_size(reg->id);
	if (size > sizeof(val))
		return -EINVAL;

	if (copy_from_user(&val, (char __user *)(unsigned long)reg->addr, size))
		return -EFAULT;

	r = kvmppc_set_one_reg(vcpu, reg->id, &val);
	if (r == -EINVAL) {
		r = 0;
		switch (reg->id) {
980 981 982 983 984 985
#ifdef CONFIG_ALTIVEC
		case KVM_REG_PPC_VR0 ... KVM_REG_PPC_VR31:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
986
			vcpu->arch.vr.vr[reg->id - KVM_REG_PPC_VR0] = val.vval;
987 988 989 990 991 992
			break;
		case KVM_REG_PPC_VSCR:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
993
			vcpu->arch.vr.vscr.u[3] = set_reg_val(reg->id, val);
994 995
			break;
		case KVM_REG_PPC_VRSAVE:
996 997 998 999 1000
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
			vcpu->arch.vrsave = set_reg_val(reg->id, val);
1001 1002
			break;
#endif /* CONFIG_ALTIVEC */
1003 1004 1005 1006 1007 1008 1009 1010 1011
		default:
			r = -EINVAL;
			break;
		}
	}

	return r;
}

1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023
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;
1024 1025 1026 1027 1028 1029 1030
	} 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;
1031 1032 1033 1034 1035 1036 1037
	} 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;
1038 1039 1040 1041 1042
#ifdef CONFIG_BOOKE
	} else if (vcpu->arch.epr_needed) {
		kvmppc_set_epr(vcpu, run->epr.epr);
		vcpu->arch.epr_needed = 0;
#endif
1043 1044
	}

1045
	r = kvmppc_vcpu_run(run, vcpu);
1046 1047 1048 1049 1050 1051 1052 1053 1054

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

	return r;
}

int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
{
1055
	if (irq->irq == KVM_INTERRUPT_UNSET) {
1056
		kvmppc_core_dequeue_external(vcpu);
1057 1058 1059 1060
		return 0;
	}

	kvmppc_core_queue_external(vcpu, irq);
1061

1062
	kvm_vcpu_kick(vcpu);
1063

1064 1065 1066
	return 0;
}

1067 1068 1069 1070 1071 1072 1073 1074 1075
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) {
1076 1077 1078 1079
	case KVM_CAP_PPC_OSI:
		r = 0;
		vcpu->arch.osi_enabled = true;
		break;
1080 1081 1082 1083
	case KVM_CAP_PPC_PAPR:
		r = 0;
		vcpu->arch.papr_enabled = true;
		break;
1084 1085
	case KVM_CAP_PPC_EPR:
		r = 0;
1086 1087 1088 1089
		if (cap->args[0])
			vcpu->arch.epr_flags |= KVMPPC_EPR_USER;
		else
			vcpu->arch.epr_flags &= ~KVMPPC_EPR_USER;
1090
		break;
1091 1092 1093 1094 1095 1096
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_WATCHDOG:
		r = 0;
		vcpu->arch.watchdog_enabled = true;
		break;
#endif
1097
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
	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;
S
Scott Wood 已提交
1108 1109 1110 1111
	}
#endif
#ifdef CONFIG_KVM_MPIC
	case KVM_CAP_IRQ_MPIC: {
A
Al Viro 已提交
1112
		struct fd f;
S
Scott Wood 已提交
1113 1114 1115
		struct kvm_device *dev;

		r = -EBADF;
A
Al Viro 已提交
1116 1117
		f = fdget(cap->args[0]);
		if (!f.file)
S
Scott Wood 已提交
1118 1119 1120
			break;

		r = -EPERM;
A
Al Viro 已提交
1121
		dev = kvm_device_from_filp(f.file);
S
Scott Wood 已提交
1122 1123 1124
		if (dev)
			r = kvmppc_mpic_connect_vcpu(dev, vcpu, cap->args[1]);

A
Al Viro 已提交
1125
		fdput(f);
S
Scott Wood 已提交
1126
		break;
S
Scott Wood 已提交
1127 1128
	}
#endif
1129 1130
#ifdef CONFIG_KVM_XICS
	case KVM_CAP_IRQ_XICS: {
A
Al Viro 已提交
1131
		struct fd f;
1132 1133 1134
		struct kvm_device *dev;

		r = -EBADF;
A
Al Viro 已提交
1135 1136
		f = fdget(cap->args[0]);
		if (!f.file)
1137 1138 1139
			break;

		r = -EPERM;
A
Al Viro 已提交
1140
		dev = kvm_device_from_filp(f.file);
1141 1142 1143
		if (dev)
			r = kvmppc_xics_connect_vcpu(dev, vcpu, cap->args[1]);

A
Al Viro 已提交
1144
		fdput(f);
1145 1146 1147
		break;
	}
#endif /* CONFIG_KVM_XICS */
1148 1149 1150 1151 1152
	default:
		r = -EINVAL;
		break;
	}

1153 1154 1155
	if (!r)
		r = kvmppc_sanity_check(vcpu);

1156 1157 1158
	return r;
}

1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177
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;

1178 1179
	switch (ioctl) {
	case KVM_INTERRUPT: {
1180 1181 1182
		struct kvm_interrupt irq;
		r = -EFAULT;
		if (copy_from_user(&irq, argp, sizeof(irq)))
1183
			goto out;
1184
		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
1185
		goto out;
1186
	}
1187

1188 1189 1190 1191 1192 1193 1194 1195 1196
	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;
	}
S
Scott Wood 已提交
1197

1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211
	case KVM_SET_ONE_REG:
	case KVM_GET_ONE_REG:
	{
		struct kvm_one_reg reg;
		r = -EFAULT;
		if (copy_from_user(&reg, argp, sizeof(reg)))
			goto out;
		if (ioctl == KVM_SET_ONE_REG)
			r = kvm_vcpu_ioctl_set_one_reg(vcpu, &reg);
		else
			r = kvm_vcpu_ioctl_get_one_reg(vcpu, &reg);
		break;
	}

1212
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
1213 1214 1215 1216 1217 1218 1219 1220 1221
	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
1222 1223 1224 1225 1226 1227 1228 1229
	default:
		r = -EINVAL;
	}

out:
	return r;
}

1230 1231 1232 1233 1234
int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf)
{
	return VM_FAULT_SIGBUS;
}

1235 1236
static int kvm_vm_ioctl_get_pvinfo(struct kvm_ppc_pvinfo *pvinfo)
{
1237 1238 1239
	u32 inst_nop = 0x60000000;
#ifdef CONFIG_KVM_BOOKE_HV
	u32 inst_sc1 = 0x44000022;
1240 1241 1242 1243
	pvinfo->hcall[0] = cpu_to_be32(inst_sc1);
	pvinfo->hcall[1] = cpu_to_be32(inst_nop);
	pvinfo->hcall[2] = cpu_to_be32(inst_nop);
	pvinfo->hcall[3] = cpu_to_be32(inst_nop);
1244
#else
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258
	u32 inst_lis = 0x3c000000;
	u32 inst_ori = 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
	 */
1259 1260 1261 1262
	pvinfo->hcall[0] = cpu_to_be32(inst_lis | ((KVM_SC_MAGIC_R0 >> 16) & inst_imm_mask));
	pvinfo->hcall[1] = cpu_to_be32(inst_ori | (KVM_SC_MAGIC_R0 & inst_imm_mask));
	pvinfo->hcall[2] = cpu_to_be32(inst_sc);
	pvinfo->hcall[3] = cpu_to_be32(inst_nop);
1263
#endif
1264

1265 1266
	pvinfo->flags = KVM_PPC_PVINFO_FLAGS_EV_IDLE;

1267 1268 1269
	return 0;
}

1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_event,
			  bool line_status)
{
	if (!irqchip_in_kernel(kvm))
		return -ENXIO;

	irq_event->status = kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID,
					irq_event->irq, irq_event->level,
					line_status);
	return 0;
}

1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299

static int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
				   struct kvm_enable_cap *cap)
{
	int r;

	if (cap->flags)
		return -EINVAL;

	switch (cap->cap) {
#ifdef CONFIG_KVM_BOOK3S_64_HANDLER
	case KVM_CAP_PPC_ENABLE_HCALL: {
		unsigned long hcall = cap->args[0];

		r = -EINVAL;
		if (hcall > MAX_HCALL_OPCODE || (hcall & 3) ||
		    cap->args[1] > 1)
			break;
1300 1301
		if (!kvmppc_book3s_hcall_implemented(kvm, hcall))
			break;
1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317
		if (cap->args[1])
			set_bit(hcall / 4, kvm->arch.enabled_hcalls);
		else
			clear_bit(hcall / 4, kvm->arch.enabled_hcalls);
		r = 0;
		break;
	}
#endif
	default:
		r = -EINVAL;
		break;
	}

	return r;
}

1318 1319 1320
long kvm_arch_vm_ioctl(struct file *filp,
                       unsigned int ioctl, unsigned long arg)
{
1321
	struct kvm *kvm __maybe_unused = filp->private_data;
1322
	void __user *argp = (void __user *)arg;
1323 1324 1325
	long r;

	switch (ioctl) {
1326 1327
	case KVM_PPC_GET_PVINFO: {
		struct kvm_ppc_pvinfo pvinfo;
1328
		memset(&pvinfo, 0, sizeof(pvinfo));
1329 1330 1331 1332 1333 1334 1335 1336
		r = kvm_vm_ioctl_get_pvinfo(&pvinfo);
		if (copy_to_user(argp, &pvinfo, sizeof(pvinfo))) {
			r = -EFAULT;
			goto out;
		}

		break;
	}
1337 1338 1339 1340 1341 1342 1343 1344 1345
	case KVM_ENABLE_CAP:
	{
		struct kvm_enable_cap cap;
		r = -EFAULT;
		if (copy_from_user(&cap, argp, sizeof(cap)))
			goto out;
		r = kvm_vm_ioctl_enable_cap(kvm, &cap);
		break;
	}
1346
#ifdef CONFIG_PPC_BOOK3S_64
1347 1348 1349 1350 1351 1352 1353 1354 1355
	case KVM_CREATE_SPAPR_TCE: {
		struct kvm_create_spapr_tce create_tce;

		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;
	}
1356 1357
	case KVM_PPC_GET_SMMU_INFO: {
		struct kvm_ppc_smmu_info info;
1358
		struct kvm *kvm = filp->private_data;
1359 1360

		memset(&info, 0, sizeof(info));
1361
		r = kvm->arch.kvm_ops->get_smmu_info(kvm, &info);
1362 1363 1364 1365
		if (r >= 0 && copy_to_user(argp, &info, sizeof(info)))
			r = -EFAULT;
		break;
	}
1366 1367 1368 1369 1370 1371
	case KVM_PPC_RTAS_DEFINE_TOKEN: {
		struct kvm *kvm = filp->private_data;

		r = kvm_vm_ioctl_rtas_define_token(kvm, argp);
		break;
	}
1372 1373 1374 1375
	default: {
		struct kvm *kvm = filp->private_data;
		r = kvm->arch.kvm_ops->arch_vm_ioctl(filp, ioctl, arg);
	}
1376
#else /* CONFIG_PPC_BOOK3S_64 */
1377
	default:
1378
		r = -ENOTTY;
1379
#endif
1380
	}
1381
out:
1382 1383 1384
	return r;
}

1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401
static unsigned long lpid_inuse[BITS_TO_LONGS(KVMPPC_NR_LPIDS)];
static unsigned long nr_lpids;

long kvmppc_alloc_lpid(void)
{
	long lpid;

	do {
		lpid = find_first_zero_bit(lpid_inuse, KVMPPC_NR_LPIDS);
		if (lpid >= nr_lpids) {
			pr_err("%s: No LPIDs free\n", __func__);
			return -ENOMEM;
		}
	} while (test_and_set_bit(lpid, lpid_inuse));

	return lpid;
}
1402
EXPORT_SYMBOL_GPL(kvmppc_alloc_lpid);
1403 1404 1405 1406 1407

void kvmppc_claim_lpid(long lpid)
{
	set_bit(lpid, lpid_inuse);
}
1408
EXPORT_SYMBOL_GPL(kvmppc_claim_lpid);
1409 1410 1411 1412 1413

void kvmppc_free_lpid(long lpid)
{
	clear_bit(lpid, lpid_inuse);
}
1414
EXPORT_SYMBOL_GPL(kvmppc_free_lpid);
1415 1416 1417 1418 1419 1420

void kvmppc_init_lpid(unsigned long nr_lpids_param)
{
	nr_lpids = min_t(unsigned long, KVMPPC_NR_LPIDS, nr_lpids_param);
	memset(lpid_inuse, 0, sizeof(lpid_inuse));
}
1421
EXPORT_SYMBOL_GPL(kvmppc_init_lpid);
1422

1423 1424 1425 1426 1427
int kvm_arch_init(void *opaque)
{
	return 0;
}

1428
EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_ppc_instr);