powerpc.c 34.7 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
#include <linux/irqbypass.h>
#include <linux/kvm_irqfd.h>
32
#include <asm/cputable.h>
33
#include <linux/uaccess.h>
34
#include <asm/kvm_ppc.h>
35
#include <asm/tlbflush.h>
36
#include <asm/cputhreads.h>
37
#include <asm/irqflags.h>
38
#include <asm/iommu.h>
39
#include "timing.h"
40
#include "irq.h"
P
Paul Mackerras 已提交
41
#include "../mm/mmu_decl.h"
42

43 44 45
#define CREATE_TRACE_POINTS
#include "trace.h"

46 47 48 49 50
struct kvmppc_ops *kvmppc_hv_ops;
EXPORT_SYMBOL_GPL(kvmppc_hv_ops);
struct kvmppc_ops *kvmppc_pr_ops;
EXPORT_SYMBOL_GPL(kvmppc_pr_ops);

51

52 53
int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
{
54
	return !!(v->arch.pending_exceptions) ||
55
	       v->requests;
56 57
}

58 59 60 61 62
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu)
{
	return 1;
}

63 64 65 66
/*
 * Common checks before entering the guest world.  Call with interrupts
 * disabled.
 *
67 68 69 70
 * returns:
 *
 * == 1 if we're ready to go into guest state
 * <= 0 if we need to go back to the host with return value
71 72 73
 */
int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)
{
S
Scott Wood 已提交
74 75 76 77
	int r;

	WARN_ON(irqs_disabled());
	hard_irq_disable();
78 79 80 81 82

	while (true) {
		if (need_resched()) {
			local_irq_enable();
			cond_resched();
S
Scott Wood 已提交
83
			hard_irq_disable();
84 85 86 87
			continue;
		}

		if (signal_pending(current)) {
88 89 90
			kvmppc_account_exit(vcpu, SIGNAL_EXITS);
			vcpu->run->exit_reason = KVM_EXIT_INTR;
			r = -EINTR;
91 92 93
			break;
		}

94 95 96 97 98 99 100
		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).
101 102 103
		 * This also orders the write to mode from any reads
		 * to the page tables done while the VCPU is running.
		 * Please see the comment in kvm_flush_remote_tlbs.
104
		 */
105
		smp_mb();
106

107 108 109 110
		if (vcpu->requests) {
			/* Make sure we process requests preemptable */
			local_irq_enable();
			trace_kvm_check_requests(vcpu);
111
			r = kvmppc_core_check_requests(vcpu);
S
Scott Wood 已提交
112
			hard_irq_disable();
113 114 115
			if (r > 0)
				continue;
			break;
116 117 118 119 120 121 122 123
		}

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

124
		guest_enter_irqoff();
S
Scott Wood 已提交
125
		return 1;
126 127
	}

S
Scott Wood 已提交
128 129
	/* return to host */
	local_irq_enable();
130 131
	return r;
}
132
EXPORT_SYMBOL_GPL(kvmppc_prepare_to_enter);
133

134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
#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

155 156 157 158 159 160 161 162 163 164
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;

165
	if (!(kvmppc_get_msr(vcpu) & MSR_SF)) {
166 167 168 169 170 171 172 173
		/* 32 bit mode */
		param1 &= 0xffffffff;
		param2 &= 0xffffffff;
		param3 &= 0xffffffff;
		param4 &= 0xffffffff;
	}

	switch (nr) {
174
	case KVM_HCALL_TOKEN(KVM_HC_PPC_MAP_MAGIC_PAGE):
175
	{
176 177 178 179 180 181 182 183 184 185
#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

186 187 188 189 190 191 192 193 194 195 196 197
		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;
198

199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217
#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

218
		r2 = KVM_MAGIC_FEAT_SR | KVM_MAGIC_FEAT_MAS0_TO_SPRG7;
219

220
		r = EV_SUCCESS;
221 222
		break;
	}
223 224
	case KVM_HCALL_TOKEN(KVM_HC_FEATURES):
		r = EV_SUCCESS;
225
#if defined(CONFIG_PPC_BOOK3S) || defined(CONFIG_KVM_E500V2)
226 227
		r2 |= (1 << KVM_FEATURE_MAGIC_PAGE);
#endif
228 229 230

		/* Second return value is in r4 */
		break;
231 232 233 234 235
	case EV_HCALL_TOKEN(EV_IDLE):
		r = EV_SUCCESS;
		kvm_vcpu_block(vcpu);
		clear_bit(KVM_REQ_UNHALT, &vcpu->requests);
		break;
236
	default:
237
		r = EV_UNIMPLEMENTED;
238 239 240
		break;
	}

241 242
	kvmppc_set_gpr(vcpu, 4, r2);

243 244
	return r;
}
245
EXPORT_SYMBOL_GPL(kvmppc_kvm_pv);
246

247 248 249 250 251 252 253 254 255 256 257 258 259
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 */
260
	if (!vcpu->arch.papr_enabled && is_kvmppc_hv_enabled(vcpu->kvm))
261 262
		goto out;

263 264 265 266 267
#ifdef CONFIG_KVM_BOOKE_HV
	if (!cpu_has_feature(CPU_FTR_EMB_HV))
		goto out;
#endif

268 269 270 271 272 273
	r = true;

out:
	vcpu->arch.sane = r;
	return r ? 0 : -EINVAL;
}
274
EXPORT_SYMBOL_GPL(kvmppc_sanity_check);
275

276 277 278 279 280
int kvmppc_emulate_mmio(struct kvm_run *run, struct kvm_vcpu *vcpu)
{
	enum emulation_result er;
	int r;

281
	er = kvmppc_emulate_loadstore(vcpu);
282 283 284 285 286 287
	switch (er) {
	case EMULATE_DONE:
		/* Future optimization: only reload non-volatiles if they were
		 * actually modified. */
		r = RESUME_GUEST_NV;
		break;
288 289 290
	case EMULATE_AGAIN:
		r = RESUME_GUEST;
		break;
291 292 293 294 295 296 297 298 299
	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:
300 301 302
	{
		u32 last_inst;

303
		kvmppc_get_last_inst(vcpu, INST_GENERIC, &last_inst);
304
		/* XXX Deliver Program interrupt to guest. */
305
		pr_emerg("%s: emulation failed (%08x)\n", __func__, last_inst);
306 307
		r = RESUME_HOST;
		break;
308
	}
309
	default:
310 311
		WARN_ON(1);
		r = RESUME_GUEST;
312 313 314 315
	}

	return r;
}
316
EXPORT_SYMBOL_GPL(kvmppc_emulate_mmio);
317

318 319 320
int kvmppc_st(struct kvm_vcpu *vcpu, ulong *eaddr, int size, void *ptr,
	      bool data)
{
321
	ulong mp_pa = vcpu->arch.magic_page_pa & KVM_PAM & PAGE_MASK;
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
	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;

337 338 339 340 341 342 343 344 345 346
	/* 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;
	}

347 348 349 350 351 352 353 354 355 356
	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)
{
357
	ulong mp_pa = vcpu->arch.magic_page_pa & KVM_PAM & PAGE_MASK;
358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375
	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;

376 377 378 379 380 381 382 383 384 385
	/* 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;
	}

386 387
	if (kvm_read_guest(vcpu->kvm, pte.raddr, ptr, size))
		return EMULATE_DO_MMIO;
388 389 390 391 392

	return EMULATE_DONE;
}
EXPORT_SYMBOL_GPL(kvmppc_ld);

393
int kvm_arch_hardware_enable(void)
394
{
395
	return 0;
396 397 398 399 400 401 402 403 404
}

int kvm_arch_hardware_setup(void)
{
	return 0;
}

void kvm_arch_check_processor_compat(void *rtn)
{
405
	*(int *)rtn = kvmppc_core_check_processor_compat();
406 407
}

408
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
409
{
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
	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;
436
	return kvmppc_core_init_vm(kvm);
437 438
err_out:
	return -EINVAL;
439 440
}

441 442 443 444 445 446 447 448 449 450
bool kvm_arch_has_vcpu_debugfs(void)
{
	return false;
}

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

451
void kvm_arch_destroy_vm(struct kvm *kvm)
452 453
{
	unsigned int i;
454
	struct kvm_vcpu *vcpu;
455

456 457 458 459 460 461 462 463 464 465
#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

466 467 468 469 470 471 472 473
	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);
474 475 476

	kvmppc_core_destroy_vm(kvm);

477
	mutex_unlock(&kvm->lock);
478 479 480

	/* drop the module reference */
	module_put(kvm->arch.kvm_ops->owner);
481 482
}

483
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
484 485
{
	int r;
486
	/* Assume we're using HV mode when the HV module is loaded */
487
	int hv_enabled = kvmppc_hv_ops ? 1 : 0;
488

489 490 491 492 493 494 495 496
	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);
	}

497
	switch (ext) {
S
Scott Wood 已提交
498 499
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_SREGS:
500
	case KVM_CAP_PPC_BOOKE_WATCHDOG:
501
	case KVM_CAP_PPC_EPR:
S
Scott Wood 已提交
502
#else
503
	case KVM_CAP_PPC_SEGSTATE:
504
	case KVM_CAP_PPC_HIOR:
505
	case KVM_CAP_PPC_PAPR:
S
Scott Wood 已提交
506
#endif
507
	case KVM_CAP_PPC_UNSET_IRQ:
508
	case KVM_CAP_PPC_IRQ_LEVEL:
509
	case KVM_CAP_ENABLE_CAP:
510
	case KVM_CAP_ENABLE_CAP_VM:
511
	case KVM_CAP_ONE_REG:
A
Alexander Graf 已提交
512
	case KVM_CAP_IOEVENTFD:
513
	case KVM_CAP_DEVICE_CTRL:
514 515 516
		r = 1;
		break;
	case KVM_CAP_PPC_PAIRED_SINGLES:
517
	case KVM_CAP_PPC_OSI:
518
	case KVM_CAP_PPC_GET_PVINFO:
519
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
520
	case KVM_CAP_SW_TLB:
S
Scott Wood 已提交
521
#endif
522
		/* We support this only for PR */
523
		r = !hv_enabled;
524
		break;
525
#ifdef CONFIG_KVM_MMIO
526 527 528
	case KVM_CAP_COALESCED_MMIO:
		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
		break;
529
#endif
530 531 532 533 534 535
#ifdef CONFIG_KVM_MPIC
	case KVM_CAP_IRQ_MPIC:
		r = 1;
		break;
#endif

536
#ifdef CONFIG_PPC_BOOK3S_64
537
	case KVM_CAP_SPAPR_TCE:
538
	case KVM_CAP_SPAPR_TCE_64:
539
	case KVM_CAP_PPC_RTAS:
540
	case KVM_CAP_PPC_FIXUP_HCALL:
541
	case KVM_CAP_PPC_ENABLE_HCALL:
542 543 544
#ifdef CONFIG_KVM_XICS
	case KVM_CAP_IRQ_XICS:
#endif
545 546
		r = 1;
		break;
547 548 549 550

	case KVM_CAP_PPC_ALLOC_HTAB:
		r = hv_enabled;
		break;
551
#endif /* CONFIG_PPC_BOOK3S_64 */
552
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
553
	case KVM_CAP_PPC_SMT:
554 555 556 557 558 559 560
		r = 0;
		if (hv_enabled) {
			if (cpu_has_feature(CPU_FTR_ARCH_300))
				r = 1;
			else
				r = threads_per_subcore;
		}
561
		break;
562
	case KVM_CAP_PPC_RMA:
563
		r = 0;
564
		break;
565 566 567
	case KVM_CAP_PPC_HWRNG:
		r = kvmppc_hwrng_present();
		break;
568
	case KVM_CAP_PPC_MMU_RADIX:
569
		r = !!(hv_enabled && radix_enabled());
570 571
		break;
	case KVM_CAP_PPC_MMU_HASH_V3:
572
		r = !!(hv_enabled && !radix_enabled() &&
573 574
		       cpu_has_feature(CPU_FTR_ARCH_300));
		break;
575
#endif
576
	case KVM_CAP_SYNC_MMU:
577
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
578
		r = hv_enabled;
579 580 581 582
#elif defined(KVM_ARCH_WANT_MMU_NOTIFIER)
		r = 1;
#else
		r = 0;
583
#endif
584 585
		break;
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
586
	case KVM_CAP_PPC_HTAB_FD:
587
		r = hv_enabled;
588
		break;
589
#endif
590 591 592 593 594 595 596
	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.
		 */
597
		if (hv_enabled)
598 599 600
			r = num_present_cpus();
		else
			r = num_online_cpus();
601
		break;
602 603 604
	case KVM_CAP_NR_MEMSLOTS:
		r = KVM_USER_MEM_SLOTS;
		break;
605 606 607
	case KVM_CAP_MAX_VCPUS:
		r = KVM_MAX_VCPUS;
		break;
608 609 610 611
#ifdef CONFIG_PPC_BOOK3S_64
	case KVM_CAP_PPC_GET_SMMU_INFO:
		r = 1;
		break;
612 613 614
	case KVM_CAP_SPAPR_MULTITCE:
		r = 1;
		break;
615
#endif
616 617 618 619
	case KVM_CAP_PPC_HTM:
		r = cpu_has_feature(CPU_FTR_TM_COMP) &&
		    is_kvmppc_hv_enabled(kvm);
		break;
620 621 622 623 624 625 626 627 628 629 630 631 632 633
	default:
		r = 0;
		break;
	}
	return r;

}

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

634
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
635 636
			   struct kvm_memory_slot *dont)
{
637
	kvmppc_core_free_memslot(kvm, free, dont);
638 639
}

640 641
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages)
642
{
643
	return kvmppc_core_create_memslot(kvm, slot, npages);
644 645
}

646
int kvm_arch_prepare_memory_region(struct kvm *kvm,
647
				   struct kvm_memory_slot *memslot,
648
				   const struct kvm_userspace_memory_region *mem,
649
				   enum kvm_mr_change change)
650
{
651
	return kvmppc_core_prepare_memory_region(kvm, memslot, mem);
652 653
}

654
void kvm_arch_commit_memory_region(struct kvm *kvm,
655
				   const struct kvm_userspace_memory_region *mem,
656
				   const struct kvm_memory_slot *old,
657
				   const struct kvm_memory_slot *new,
658
				   enum kvm_mr_change change)
659
{
660
	kvmppc_core_commit_memory_region(kvm, mem, old, new);
661 662
}

663 664
void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
				   struct kvm_memory_slot *slot)
665
{
666
	kvmppc_core_flush_memslot(kvm, slot);
667 668
}

669 670
struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
{
671 672
	struct kvm_vcpu *vcpu;
	vcpu = kvmppc_core_vcpu_create(kvm, id);
673 674
	if (!IS_ERR(vcpu)) {
		vcpu->arch.wqp = &vcpu->wq;
675
		kvmppc_create_vcpu_debugfs(vcpu, id);
676
	}
677
	return vcpu;
678 679
}

680
void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
681 682 683
{
}

684 685
void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
686 687 688
	/* Make sure we're not using the vcpu anymore */
	hrtimer_cancel(&vcpu->arch.dec_timer);

689
	kvmppc_remove_vcpu_debugfs(vcpu);
S
Scott Wood 已提交
690 691 692 693 694

	switch (vcpu->arch.irq_type) {
	case KVMPPC_IRQ_MPIC:
		kvmppc_mpic_disconnect_vcpu(vcpu->arch.mpic, vcpu);
		break;
695 696 697
	case KVMPPC_IRQ_XICS:
		kvmppc_xics_free_icp(vcpu);
		break;
S
Scott Wood 已提交
698 699
	}

700
	kvmppc_core_vcpu_free(vcpu);
701 702 703 704 705 706 707 708 709
}

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

int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
{
710
	return kvmppc_core_pending_dec(vcpu);
711 712
}

T
Thomas Huth 已提交
713
static enum hrtimer_restart kvmppc_decrementer_wakeup(struct hrtimer *timer)
A
Alexander Graf 已提交
714 715 716 717
{
	struct kvm_vcpu *vcpu;

	vcpu = container_of(timer, struct kvm_vcpu, arch.dec_timer);
718
	kvmppc_decrementer_func(vcpu);
A
Alexander Graf 已提交
719 720 721 722

	return HRTIMER_NORESTART;
}

723 724
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
725 726
	int ret;

A
Alexander Graf 已提交
727 728
	hrtimer_init(&vcpu->arch.dec_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
	vcpu->arch.dec_timer.function = kvmppc_decrementer_wakeup;
729
	vcpu->arch.dec_expires = ~(u64)0;
730

731 732 733
#ifdef CONFIG_KVM_EXIT_TIMING
	mutex_init(&vcpu->arch.exit_timing_lock);
#endif
734 735
	ret = kvmppc_subarch_vcpu_init(vcpu);
	return ret;
736 737 738 739
}

void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
740
	kvmppc_mmu_destroy(vcpu);
741
	kvmppc_subarch_vcpu_uninit(vcpu);
742 743 744 745
}

void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
746 747 748 749 750 751 752 753 754 755
#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
756
	kvmppc_core_vcpu_load(vcpu, cpu);
757 758 759 760
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
761
	kvmppc_core_vcpu_put(vcpu);
762 763 764
#ifdef CONFIG_BOOKE
	vcpu->arch.vrsave = mfspr(SPRN_VRSAVE);
#endif
765 766
}

767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802
/*
 * irq_bypass_add_producer and irq_bypass_del_producer are only
 * useful if the architecture supports PCI passthrough.
 * irq_bypass_stop and irq_bypass_start are not needed and so
 * kvm_ops are not defined for them.
 */
bool kvm_arch_has_irq_bypass(void)
{
	return ((kvmppc_hv_ops && kvmppc_hv_ops->irq_bypass_add_producer) ||
		(kvmppc_pr_ops && kvmppc_pr_ops->irq_bypass_add_producer));
}

int kvm_arch_irq_bypass_add_producer(struct irq_bypass_consumer *cons,
				     struct irq_bypass_producer *prod)
{
	struct kvm_kernel_irqfd *irqfd =
		container_of(cons, struct kvm_kernel_irqfd, consumer);
	struct kvm *kvm = irqfd->kvm;

	if (kvm->arch.kvm_ops->irq_bypass_add_producer)
		return kvm->arch.kvm_ops->irq_bypass_add_producer(cons, prod);

	return 0;
}

void kvm_arch_irq_bypass_del_producer(struct irq_bypass_consumer *cons,
				      struct irq_bypass_producer *prod)
{
	struct kvm_kernel_irqfd *irqfd =
		container_of(cons, struct kvm_kernel_irqfd, consumer);
	struct kvm *kvm = irqfd->kvm;

	if (kvm->arch.kvm_ops->irq_bypass_del_producer)
		kvm->arch.kvm_ops->irq_bypass_del_producer(cons, prod);
}

803 804 805
static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
                                      struct kvm_run *run)
{
806
	u64 uninitialized_var(gpr);
807

808
	if (run->mmio.len > sizeof(gpr)) {
809 810 811 812
		printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
		return;
	}

813
	if (!vcpu->arch.mmio_host_swabbed) {
814
		switch (run->mmio.len) {
815
		case 8: gpr = *(u64 *)run->mmio.data; break;
816 817 818
		case 4: gpr = *(u32 *)run->mmio.data; break;
		case 2: gpr = *(u16 *)run->mmio.data; break;
		case 1: gpr = *(u8 *)run->mmio.data; break;
819 820 821
		}
	} else {
		switch (run->mmio.len) {
822 823 824
		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;
825
		case 1: gpr = *(u8 *)run->mmio.data; break;
826 827
		}
	}
828

A
Alexander Graf 已提交
829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844
	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;
		}
	}

845
	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
846

847 848
	switch (vcpu->arch.io_gpr & KVM_MMIO_REG_EXT_MASK) {
	case KVM_MMIO_REG_GPR:
849 850
		kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
		break;
851
	case KVM_MMIO_REG_FPR:
852
		VCPU_FPR(vcpu, vcpu->arch.io_gpr & KVM_MMIO_REG_MASK) = gpr;
853
		break;
854
#ifdef CONFIG_PPC_BOOK3S
855 856
	case KVM_MMIO_REG_QPR:
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
857
		break;
858
	case KVM_MMIO_REG_FQPR:
859
		VCPU_FPR(vcpu, vcpu->arch.io_gpr & KVM_MMIO_REG_MASK) = gpr;
860
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
861
		break;
862
#endif
863 864 865
	default:
		BUG();
	}
866 867
}

868 869 870
static int __kvmppc_handle_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
				unsigned int rt, unsigned int bytes,
				int is_default_endian, int sign_extend)
871
{
872
	int idx, ret;
873
	bool host_swabbed;
874

875
	/* Pity C doesn't have a logical XOR operator */
876
	if (kvmppc_need_byteswap(vcpu)) {
877
		host_swabbed = is_default_endian;
878
	} else {
879
		host_swabbed = !is_default_endian;
880
	}
881

882 883 884 885 886 887 888 889 890 891
	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;
892
	vcpu->arch.mmio_host_swabbed = host_swabbed;
893 894
	vcpu->mmio_needed = 1;
	vcpu->mmio_is_write = 0;
895
	vcpu->arch.mmio_sign_extend = sign_extend;
896

897 898
	idx = srcu_read_lock(&vcpu->kvm->srcu);

899
	ret = kvm_io_bus_read(vcpu, KVM_MMIO_BUS, run->mmio.phys_addr,
900 901 902 903 904
			      bytes, &run->mmio.data);

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

	if (!ret) {
A
Alexander Graf 已提交
905 906 907 908 909
		kvmppc_complete_mmio_load(vcpu, run);
		vcpu->mmio_needed = 0;
		return EMULATE_DONE;
	}

910 911
	return EMULATE_DO_MMIO;
}
912 913 914 915 916 917 918

int kvmppc_handle_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
		       unsigned int rt, unsigned int bytes,
		       int is_default_endian)
{
	return __kvmppc_handle_load(run, vcpu, rt, bytes, is_default_endian, 0);
}
919
EXPORT_SYMBOL_GPL(kvmppc_handle_load);
920

A
Alexander Graf 已提交
921 922
/* Same as above, but sign extends */
int kvmppc_handle_loads(struct kvm_run *run, struct kvm_vcpu *vcpu,
923 924
			unsigned int rt, unsigned int bytes,
			int is_default_endian)
A
Alexander Graf 已提交
925
{
926
	return __kvmppc_handle_load(run, vcpu, rt, bytes, is_default_endian, 1);
A
Alexander Graf 已提交
927 928
}

929
int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
930
			u64 val, unsigned int bytes, int is_default_endian)
931 932
{
	void *data = run->mmio.data;
933
	int idx, ret;
934
	bool host_swabbed;
935

936
	/* Pity C doesn't have a logical XOR operator */
937
	if (kvmppc_need_byteswap(vcpu)) {
938
		host_swabbed = is_default_endian;
939
	} else {
940
		host_swabbed = !is_default_endian;
941
	}
942 943 944 945 946 947 948 949 950 951 952 953 954

	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'. */
955
	if (!host_swabbed) {
956
		switch (bytes) {
957
		case 8: *(u64 *)data = val; break;
958 959 960 961 962 963
		case 4: *(u32 *)data = val; break;
		case 2: *(u16 *)data = val; break;
		case 1: *(u8  *)data = val; break;
		}
	} else {
		switch (bytes) {
964 965 966 967
		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;
968 969 970
		}
	}

971 972
	idx = srcu_read_lock(&vcpu->kvm->srcu);

973
	ret = kvm_io_bus_write(vcpu, KVM_MMIO_BUS, run->mmio.phys_addr,
974 975 976 977 978
			       bytes, &run->mmio.data);

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

	if (!ret) {
A
Alexander Graf 已提交
979 980 981 982
		vcpu->mmio_needed = 0;
		return EMULATE_DONE;
	}

983 984
	return EMULATE_DO_MMIO;
}
985
EXPORT_SYMBOL_GPL(kvmppc_handle_store);
986

987 988 989 990 991 992 993 994 995 996 997 998 999 1000
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) {
1001 1002 1003 1004 1005 1006
#ifdef CONFIG_ALTIVEC
		case KVM_REG_PPC_VR0 ... KVM_REG_PPC_VR31:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
1007
			val.vval = vcpu->arch.vr.vr[reg->id - KVM_REG_PPC_VR0];
1008 1009 1010 1011 1012 1013
			break;
		case KVM_REG_PPC_VSCR:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
1014
			val = get_reg_val(reg->id, vcpu->arch.vr.vscr.u[3]);
1015 1016
			break;
		case KVM_REG_PPC_VRSAVE:
1017
			val = get_reg_val(reg->id, vcpu->arch.vrsave);
1018 1019
			break;
#endif /* CONFIG_ALTIVEC */
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051
		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) {
1052 1053 1054 1055 1056 1057
#ifdef CONFIG_ALTIVEC
		case KVM_REG_PPC_VR0 ... KVM_REG_PPC_VR31:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
1058
			vcpu->arch.vr.vr[reg->id - KVM_REG_PPC_VR0] = val.vval;
1059 1060 1061 1062 1063 1064
			break;
		case KVM_REG_PPC_VSCR:
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
1065
			vcpu->arch.vr.vscr.u[3] = set_reg_val(reg->id, val);
1066 1067
			break;
		case KVM_REG_PPC_VRSAVE:
1068 1069 1070 1071 1072
			if (!cpu_has_feature(CPU_FTR_ALTIVEC)) {
				r = -ENXIO;
				break;
			}
			vcpu->arch.vrsave = set_reg_val(reg->id, val);
1073 1074
			break;
#endif /* CONFIG_ALTIVEC */
1075 1076 1077 1078 1079 1080 1081 1082 1083
		default:
			r = -EINVAL;
			break;
		}
	}

	return r;
}

1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095
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;
1096 1097 1098 1099 1100 1101 1102
	} 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;
1103 1104 1105 1106 1107 1108 1109
	} 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;
1110 1111 1112 1113 1114
#ifdef CONFIG_BOOKE
	} else if (vcpu->arch.epr_needed) {
		kvmppc_set_epr(vcpu, run->epr.epr);
		vcpu->arch.epr_needed = 0;
#endif
1115 1116
	}

1117
	r = kvmppc_vcpu_run(run, vcpu);
1118 1119 1120 1121 1122 1123 1124 1125 1126

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

	return r;
}

int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
{
1127
	if (irq->irq == KVM_INTERRUPT_UNSET) {
1128
		kvmppc_core_dequeue_external(vcpu);
1129 1130 1131 1132
		return 0;
	}

	kvmppc_core_queue_external(vcpu, irq);
1133

1134
	kvm_vcpu_kick(vcpu);
1135

1136 1137 1138
	return 0;
}

1139 1140 1141 1142 1143 1144 1145 1146 1147
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) {
1148 1149 1150 1151
	case KVM_CAP_PPC_OSI:
		r = 0;
		vcpu->arch.osi_enabled = true;
		break;
1152 1153 1154 1155
	case KVM_CAP_PPC_PAPR:
		r = 0;
		vcpu->arch.papr_enabled = true;
		break;
1156 1157
	case KVM_CAP_PPC_EPR:
		r = 0;
1158 1159 1160 1161
		if (cap->args[0])
			vcpu->arch.epr_flags |= KVMPPC_EPR_USER;
		else
			vcpu->arch.epr_flags &= ~KVMPPC_EPR_USER;
1162
		break;
1163 1164 1165 1166 1167 1168
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_WATCHDOG:
		r = 0;
		vcpu->arch.watchdog_enabled = true;
		break;
#endif
1169
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
	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
1180 1181 1182 1183
	}
#endif
#ifdef CONFIG_KVM_MPIC
	case KVM_CAP_IRQ_MPIC: {
A
Al Viro 已提交
1184
		struct fd f;
S
Scott Wood 已提交
1185 1186 1187
		struct kvm_device *dev;

		r = -EBADF;
A
Al Viro 已提交
1188 1189
		f = fdget(cap->args[0]);
		if (!f.file)
S
Scott Wood 已提交
1190 1191 1192
			break;

		r = -EPERM;
A
Al Viro 已提交
1193
		dev = kvm_device_from_filp(f.file);
S
Scott Wood 已提交
1194 1195 1196
		if (dev)
			r = kvmppc_mpic_connect_vcpu(dev, vcpu, cap->args[1]);

A
Al Viro 已提交
1197
		fdput(f);
S
Scott Wood 已提交
1198
		break;
S
Scott Wood 已提交
1199 1200
	}
#endif
1201 1202
#ifdef CONFIG_KVM_XICS
	case KVM_CAP_IRQ_XICS: {
A
Al Viro 已提交
1203
		struct fd f;
1204 1205 1206
		struct kvm_device *dev;

		r = -EBADF;
A
Al Viro 已提交
1207 1208
		f = fdget(cap->args[0]);
		if (!f.file)
1209 1210 1211
			break;

		r = -EPERM;
A
Al Viro 已提交
1212
		dev = kvm_device_from_filp(f.file);
1213 1214 1215
		if (dev)
			r = kvmppc_xics_connect_vcpu(dev, vcpu, cap->args[1]);

A
Al Viro 已提交
1216
		fdput(f);
1217 1218 1219
		break;
	}
#endif /* CONFIG_KVM_XICS */
1220 1221 1222 1223 1224
	default:
		r = -EINVAL;
		break;
	}

1225 1226 1227
	if (!r)
		r = kvmppc_sanity_check(vcpu);

1228 1229 1230
	return r;
}

1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
bool kvm_arch_intc_initialized(struct kvm *kvm)
{
#ifdef CONFIG_KVM_MPIC
	if (kvm->arch.mpic)
		return true;
#endif
#ifdef CONFIG_KVM_XICS
	if (kvm->arch.xics)
		return true;
#endif
	return false;
}

1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262
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;

1263 1264
	switch (ioctl) {
	case KVM_INTERRUPT: {
1265 1266 1267
		struct kvm_interrupt irq;
		r = -EFAULT;
		if (copy_from_user(&irq, argp, sizeof(irq)))
1268
			goto out;
1269
		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
1270
		goto out;
1271
	}
1272

1273 1274 1275 1276 1277 1278 1279 1280 1281
	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
1282

1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
	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;
	}

1297
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
1298 1299 1300 1301 1302 1303 1304 1305 1306
	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
1307 1308 1309 1310 1311 1312 1313 1314
	default:
		r = -EINVAL;
	}

out:
	return r;
}

1315 1316 1317 1318 1319
int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf)
{
	return VM_FAULT_SIGBUS;
}

1320 1321
static int kvm_vm_ioctl_get_pvinfo(struct kvm_ppc_pvinfo *pvinfo)
{
1322 1323 1324
	u32 inst_nop = 0x60000000;
#ifdef CONFIG_KVM_BOOKE_HV
	u32 inst_sc1 = 0x44000022;
1325 1326 1327 1328
	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);
1329
#else
1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
	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
	 */
1344 1345 1346 1347
	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);
1348
#endif
1349

1350 1351
	pvinfo->flags = KVM_PPC_PVINFO_FLAGS_EV_IDLE;

1352 1353 1354
	return 0;
}

1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
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;
}

1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384

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;
1385 1386
		if (!kvmppc_book3s_hcall_implemented(kvm, hcall))
			break;
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402
		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;
}

1403 1404 1405
long kvm_arch_vm_ioctl(struct file *filp,
                       unsigned int ioctl, unsigned long arg)
{
1406
	struct kvm *kvm __maybe_unused = filp->private_data;
1407
	void __user *argp = (void __user *)arg;
1408 1409 1410
	long r;

	switch (ioctl) {
1411 1412
	case KVM_PPC_GET_PVINFO: {
		struct kvm_ppc_pvinfo pvinfo;
1413
		memset(&pvinfo, 0, sizeof(pvinfo));
1414 1415 1416 1417 1418 1419 1420 1421
		r = kvm_vm_ioctl_get_pvinfo(&pvinfo);
		if (copy_to_user(argp, &pvinfo, sizeof(pvinfo))) {
			r = -EFAULT;
			goto out;
		}

		break;
	}
1422 1423 1424 1425 1426 1427 1428 1429 1430
	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;
	}
1431
#ifdef CONFIG_PPC_BOOK3S_64
1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444
	case KVM_CREATE_SPAPR_TCE_64: {
		struct kvm_create_spapr_tce_64 create_tce_64;

		r = -EFAULT;
		if (copy_from_user(&create_tce_64, argp, sizeof(create_tce_64)))
			goto out;
		if (create_tce_64.flags) {
			r = -EINVAL;
			goto out;
		}
		r = kvm_vm_ioctl_create_spapr_tce(kvm, &create_tce_64);
		goto out;
	}
1445 1446
	case KVM_CREATE_SPAPR_TCE: {
		struct kvm_create_spapr_tce create_tce;
1447
		struct kvm_create_spapr_tce_64 create_tce_64;
1448 1449 1450 1451

		r = -EFAULT;
		if (copy_from_user(&create_tce, argp, sizeof(create_tce)))
			goto out;
1452 1453 1454 1455 1456 1457 1458 1459

		create_tce_64.liobn = create_tce.liobn;
		create_tce_64.page_shift = IOMMU_PAGE_SHIFT_4K;
		create_tce_64.offset = 0;
		create_tce_64.size = create_tce.window_size >>
				IOMMU_PAGE_SHIFT_4K;
		create_tce_64.flags = 0;
		r = kvm_vm_ioctl_create_spapr_tce(kvm, &create_tce_64);
1460 1461
		goto out;
	}
1462 1463
	case KVM_PPC_GET_SMMU_INFO: {
		struct kvm_ppc_smmu_info info;
1464
		struct kvm *kvm = filp->private_data;
1465 1466

		memset(&info, 0, sizeof(info));
1467
		r = kvm->arch.kvm_ops->get_smmu_info(kvm, &info);
1468 1469 1470 1471
		if (r >= 0 && copy_to_user(argp, &info, sizeof(info)))
			r = -EFAULT;
		break;
	}
1472 1473 1474 1475 1476 1477
	case KVM_PPC_RTAS_DEFINE_TOKEN: {
		struct kvm *kvm = filp->private_data;

		r = kvm_vm_ioctl_rtas_define_token(kvm, argp);
		break;
	}
1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502
	case KVM_PPC_CONFIGURE_V3_MMU: {
		struct kvm *kvm = filp->private_data;
		struct kvm_ppc_mmuv3_cfg cfg;

		r = -EINVAL;
		if (!kvm->arch.kvm_ops->configure_mmu)
			goto out;
		r = -EFAULT;
		if (copy_from_user(&cfg, argp, sizeof(cfg)))
			goto out;
		r = kvm->arch.kvm_ops->configure_mmu(kvm, &cfg);
		break;
	}
	case KVM_PPC_GET_RMMU_INFO: {
		struct kvm *kvm = filp->private_data;
		struct kvm_ppc_rmmu_info info;

		r = -EINVAL;
		if (!kvm->arch.kvm_ops->get_rmmu_info)
			goto out;
		r = kvm->arch.kvm_ops->get_rmmu_info(kvm, &info);
		if (r >= 0 && copy_to_user(argp, &info, sizeof(info)))
			r = -EFAULT;
		break;
	}
1503 1504 1505 1506
	default: {
		struct kvm *kvm = filp->private_data;
		r = kvm->arch.kvm_ops->arch_vm_ioctl(filp, ioctl, arg);
	}
1507
#else /* CONFIG_PPC_BOOK3S_64 */
1508
	default:
1509
		r = -ENOTTY;
1510
#endif
1511
	}
1512
out:
1513 1514 1515
	return r;
}

1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532
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;
}
1533
EXPORT_SYMBOL_GPL(kvmppc_alloc_lpid);
1534 1535 1536 1537 1538

void kvmppc_claim_lpid(long lpid)
{
	set_bit(lpid, lpid_inuse);
}
1539
EXPORT_SYMBOL_GPL(kvmppc_claim_lpid);
1540 1541 1542 1543 1544

void kvmppc_free_lpid(long lpid)
{
	clear_bit(lpid, lpid_inuse);
}
1545
EXPORT_SYMBOL_GPL(kvmppc_free_lpid);
1546 1547 1548 1549 1550 1551

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));
}
1552
EXPORT_SYMBOL_GPL(kvmppc_init_lpid);
1553

1554 1555 1556 1557 1558
int kvm_arch_init(void *opaque)
{
	return 0;
}

1559
EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_ppc_instr);