powerpc.c 24.1 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 30 31
#include <asm/cputable.h>
#include <asm/uaccess.h>
#include <asm/kvm_ppc.h>
32
#include <asm/tlbflush.h>
33
#include <asm/cputhreads.h>
34
#include <asm/irqflags.h>
35
#include "timing.h"
36
#include "irq.h"
P
Paul Mackerras 已提交
37
#include "../mm/mmu_decl.h"
38

39 40 41
#define CREATE_TRACE_POINTS
#include "trace.h"

42 43
struct kvmppc_ops *kvmppc_ops;

44 45
int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
{
46
	return !!(v->arch.pending_exceptions) ||
47
	       v->requests;
48 49
}

50 51 52 53 54
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu)
{
	return 1;
}

55 56 57 58
/*
 * Common checks before entering the guest world.  Call with interrupts
 * disabled.
 *
59 60 61 62
 * returns:
 *
 * == 1 if we're ready to go into guest state
 * <= 0 if we need to go back to the host with return value
63 64 65
 */
int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)
{
66
	int r = 1;
67 68 69 70 71 72 73 74 75 76 77

	WARN_ON_ONCE(!irqs_disabled());
	while (true) {
		if (need_resched()) {
			local_irq_enable();
			cond_resched();
			local_irq_disable();
			continue;
		}

		if (signal_pending(current)) {
78 79 80
			kvmppc_account_exit(vcpu, SIGNAL_EXITS);
			vcpu->run->exit_reason = KVM_EXIT_INTR;
			r = -EINTR;
81 82 83
			break;
		}

84 85 86 87 88 89 90 91
		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).
		 */
92
		smp_mb();
93

94 95 96 97
		if (vcpu->requests) {
			/* Make sure we process requests preemptable */
			local_irq_enable();
			trace_kvm_check_requests(vcpu);
98
			r = kvmppc_core_check_requests(vcpu);
99
			local_irq_disable();
100 101 102
			if (r > 0)
				continue;
			break;
103 104 105 106 107 108 109 110
		}

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

111 112 113 114 115 116 117
#ifdef CONFIG_PPC64
		/* lazy EE magic */
		hard_irq_disable();
		if (lazy_irq_pending()) {
			/* Got an interrupt in between, try again */
			local_irq_enable();
			local_irq_disable();
118
			kvm_guest_exit();
119 120 121 122
			continue;
		}
#endif

123
		kvm_guest_enter();
124 125 126 127 128
		break;
	}

	return r;
}
129
EXPORT_SYMBOL_GPL(kvmppc_prepare_to_enter);
130

131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
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) {
150
	case KVM_HCALL_TOKEN(KVM_HC_PPC_MAP_MAGIC_PAGE):
151 152 153 154
	{
		vcpu->arch.magic_page_pa = param1;
		vcpu->arch.magic_page_ea = param2;

155
		r2 = KVM_MAGIC_FEAT_SR | KVM_MAGIC_FEAT_MAS0_TO_SPRG7;
156

157
		r = EV_SUCCESS;
158 159
		break;
	}
160 161
	case KVM_HCALL_TOKEN(KVM_HC_FEATURES):
		r = EV_SUCCESS;
162
#if defined(CONFIG_PPC_BOOK3S) || defined(CONFIG_KVM_E500V2)
S
Scott Wood 已提交
163
		/* XXX Missing magic page on 44x */
164 165
		r2 |= (1 << KVM_FEATURE_MAGIC_PAGE);
#endif
166 167 168

		/* Second return value is in r4 */
		break;
169 170 171 172 173
	case EV_HCALL_TOKEN(EV_IDLE):
		r = EV_SUCCESS;
		kvm_vcpu_block(vcpu);
		clear_bit(KVM_REQ_UNHALT, &vcpu->requests);
		break;
174
	default:
175
		r = EV_UNIMPLEMENTED;
176 177 178
		break;
	}

179 180
	kvmppc_set_gpr(vcpu, 4, r2);

181 182
	return r;
}
183
EXPORT_SYMBOL_GPL(kvmppc_kvm_pv);
184

185 186 187 188 189 190 191 192 193 194 195 196 197
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 */
198
	if (!vcpu->arch.papr_enabled && kvmppc_ops->is_hv_enabled)
199 200
		goto out;

201 202 203 204 205
#ifdef CONFIG_KVM_BOOKE_HV
	if (!cpu_has_feature(CPU_FTR_EMB_HV))
		goto out;
#endif

206 207 208 209 210 211
	r = true;

out:
	vcpu->arch.sane = r;
	return r ? 0 : -EINVAL;
}
212
EXPORT_SYMBOL_GPL(kvmppc_sanity_check);
213

214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
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__,
237
		       kvmppc_get_last_inst(vcpu));
238 239 240
		r = RESUME_HOST;
		break;
	default:
241 242
		WARN_ON(1);
		r = RESUME_GUEST;
243 244 245 246
	}

	return r;
}
247
EXPORT_SYMBOL_GPL(kvmppc_emulate_mmio);
248

249
int kvm_arch_hardware_enable(void *garbage)
250
{
251
	return 0;
252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268
}

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)
{
269
	*(int *)rtn = kvmppc_core_check_processor_compat();
270 271
}

272
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
273
{
274 275 276
	if (type)
		return -EINVAL;

277
	return kvmppc_core_init_vm(kvm);
278 279
}

280
void kvm_arch_destroy_vm(struct kvm *kvm)
281 282
{
	unsigned int i;
283
	struct kvm_vcpu *vcpu;
284

285 286 287 288 289 290 291 292
	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);
293 294 295

	kvmppc_core_destroy_vm(kvm);

296
	mutex_unlock(&kvm->lock);
297 298
}

299 300 301 302
void kvm_arch_sync_events(struct kvm *kvm)
{
}

303 304 305 306 307
int kvm_dev_ioctl_check_extension(long ext)
{
	int r;

	switch (ext) {
S
Scott Wood 已提交
308 309
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_SREGS:
310
	case KVM_CAP_PPC_BOOKE_WATCHDOG:
311
	case KVM_CAP_PPC_EPR:
S
Scott Wood 已提交
312
#else
313
	case KVM_CAP_PPC_SEGSTATE:
314
	case KVM_CAP_PPC_HIOR:
315
	case KVM_CAP_PPC_PAPR:
S
Scott Wood 已提交
316
#endif
317
	case KVM_CAP_PPC_UNSET_IRQ:
318
	case KVM_CAP_PPC_IRQ_LEVEL:
319
	case KVM_CAP_ENABLE_CAP:
320
	case KVM_CAP_ONE_REG:
A
Alexander Graf 已提交
321
	case KVM_CAP_IOEVENTFD:
322
	case KVM_CAP_DEVICE_CTRL:
323 324 325
		r = 1;
		break;
	case KVM_CAP_PPC_PAIRED_SINGLES:
326
	case KVM_CAP_PPC_OSI:
327
	case KVM_CAP_PPC_GET_PVINFO:
328
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
329
	case KVM_CAP_SW_TLB:
S
Scott Wood 已提交
330
#endif
331 332
		/* We support this only for PR */
		r = !kvmppc_ops->is_hv_enabled;
333
		break;
334
#ifdef CONFIG_KVM_MMIO
335 336 337
	case KVM_CAP_COALESCED_MMIO:
		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
		break;
338
#endif
339 340 341 342 343 344
#ifdef CONFIG_KVM_MPIC
	case KVM_CAP_IRQ_MPIC:
		r = 1;
		break;
#endif

345
#ifdef CONFIG_PPC_BOOK3S_64
346
	case KVM_CAP_SPAPR_TCE:
347
	case KVM_CAP_PPC_ALLOC_HTAB:
348
	case KVM_CAP_PPC_RTAS:
349 350 351
#ifdef CONFIG_KVM_XICS
	case KVM_CAP_IRQ_XICS:
#endif
352 353
		r = 1;
		break;
354
#endif /* CONFIG_PPC_BOOK3S_64 */
355
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
356
	case KVM_CAP_PPC_SMT:
357 358 359 360
		if (kvmppc_ops->is_hv_enabled)
			r = threads_per_core;
		else
			r = 0;
361
		break;
362
	case KVM_CAP_PPC_RMA:
363
		r = kvmppc_ops->is_hv_enabled;
364
		/* PPC970 requires an RMA */
365
		if (r && cpu_has_feature(CPU_FTR_ARCH_201))
366
			r = 2;
367
		break;
368
#endif
369
	case KVM_CAP_SYNC_MMU:
370 371 372 373 374
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
		if (kvmppc_ops->is_hv_enabled)
			r = cpu_has_feature(CPU_FTR_ARCH_206) ? 1 : 0;
		else
			r = 0;
375 376 377 378
#elif defined(KVM_ARCH_WANT_MMU_NOTIFIER)
		r = 1;
#else
		r = 0;
379
#endif
380 381
		break;
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
382
	case KVM_CAP_PPC_HTAB_FD:
383
		r = kvmppc_ops->is_hv_enabled;
384
		break;
385
#endif
386 387 388 389 390 391 392
	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.
		 */
393 394 395 396
		if (kvmppc_ops->is_hv_enabled)
			r = num_present_cpus();
		else
			r = num_online_cpus();
397 398 399 400
		break;
	case KVM_CAP_MAX_VCPUS:
		r = KVM_MAX_VCPUS;
		break;
401 402 403 404 405
#ifdef CONFIG_PPC_BOOK3S_64
	case KVM_CAP_PPC_GET_SMMU_INFO:
		r = 1;
		break;
#endif
406 407 408 409 410 411 412 413 414 415 416 417 418 419
	default:
		r = 0;
		break;
	}
	return r;

}

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

420
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
421 422
			   struct kvm_memory_slot *dont)
{
423
	kvmppc_core_free_memslot(kvm, free, dont);
424 425
}

426 427
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages)
428
{
429
	return kvmppc_core_create_memslot(kvm, slot, npages);
430 431
}

432 433 434 435
void kvm_arch_memslots_updated(struct kvm *kvm)
{
}

436
int kvm_arch_prepare_memory_region(struct kvm *kvm,
437
				   struct kvm_memory_slot *memslot,
438 439
				   struct kvm_userspace_memory_region *mem,
				   enum kvm_mr_change change)
440
{
441
	return kvmppc_core_prepare_memory_region(kvm, memslot, mem);
442 443
}

444
void kvm_arch_commit_memory_region(struct kvm *kvm,
445
				   struct kvm_userspace_memory_region *mem,
446 447
				   const struct kvm_memory_slot *old,
				   enum kvm_mr_change change)
448
{
449
	kvmppc_core_commit_memory_region(kvm, mem, old);
450 451
}

452 453 454
void kvm_arch_flush_shadow_all(struct kvm *kvm)
{
}
455

456 457
void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
				   struct kvm_memory_slot *slot)
458
{
459
	kvmppc_core_flush_memslot(kvm, slot);
460 461
}

462 463
struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
{
464 465
	struct kvm_vcpu *vcpu;
	vcpu = kvmppc_core_vcpu_create(kvm, id);
466 467
	if (!IS_ERR(vcpu)) {
		vcpu->arch.wqp = &vcpu->wq;
468
		kvmppc_create_vcpu_debugfs(vcpu, id);
469
	}
470
	return vcpu;
471 472
}

473 474 475 476 477
int kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
{
	return 0;
}

478 479
void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
480 481 482 483
	/* Make sure we're not using the vcpu anymore */
	hrtimer_cancel(&vcpu->arch.dec_timer);
	tasklet_kill(&vcpu->arch.tasklet);

484
	kvmppc_remove_vcpu_debugfs(vcpu);
S
Scott Wood 已提交
485 486 487 488 489

	switch (vcpu->arch.irq_type) {
	case KVMPPC_IRQ_MPIC:
		kvmppc_mpic_disconnect_vcpu(vcpu->arch.mpic, vcpu);
		break;
490 491 492
	case KVMPPC_IRQ_XICS:
		kvmppc_xics_free_icp(vcpu);
		break;
S
Scott Wood 已提交
493 494
	}

495
	kvmppc_core_vcpu_free(vcpu);
496 497 498 499 500 501 502 503 504
}

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

int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
{
505
	return kvmppc_core_pending_dec(vcpu);
506 507
}

A
Alexander Graf 已提交
508 509 510 511 512 513 514 515 516 517 518 519 520 521
/*
 * 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;
}

522 523
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
524 525
	int ret;

A
Alexander Graf 已提交
526 527 528
	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;
529
	vcpu->arch.dec_expires = ~(u64)0;
530

531 532 533
#ifdef CONFIG_KVM_EXIT_TIMING
	mutex_init(&vcpu->arch.exit_timing_lock);
#endif
534 535
	ret = kvmppc_subarch_vcpu_init(vcpu);
	return ret;
536 537 538 539
}

void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
540
	kvmppc_mmu_destroy(vcpu);
541
	kvmppc_subarch_vcpu_uninit(vcpu);
542 543 544 545
}

void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
546 547 548 549 550 551 552 553 554 555
#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
556
	kvmppc_core_vcpu_load(vcpu, cpu);
557 558 559 560
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
561
	kvmppc_core_vcpu_put(vcpu);
562 563 564
#ifdef CONFIG_BOOKE
	vcpu->arch.vrsave = mfspr(SPRN_VRSAVE);
#endif
565 566 567 568 569
}

static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,
                                     struct kvm_run *run)
{
570
	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, run->dcr.data);
571 572 573 574 575
}

static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
                                      struct kvm_run *run)
{
576
	u64 uninitialized_var(gpr);
577

578
	if (run->mmio.len > sizeof(gpr)) {
579 580 581 582 583 584
		printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
		return;
	}

	if (vcpu->arch.mmio_is_bigendian) {
		switch (run->mmio.len) {
585
		case 8: gpr = *(u64 *)run->mmio.data; break;
586 587 588
		case 4: gpr = *(u32 *)run->mmio.data; break;
		case 2: gpr = *(u16 *)run->mmio.data; break;
		case 1: gpr = *(u8 *)run->mmio.data; break;
589 590 591 592
		}
	} else {
		/* Convert BE data from userland back to LE. */
		switch (run->mmio.len) {
593 594 595
		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;
596 597
		}
	}
598

A
Alexander Graf 已提交
599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614
	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;
		}
	}

615
	kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
616

617 618
	switch (vcpu->arch.io_gpr & KVM_MMIO_REG_EXT_MASK) {
	case KVM_MMIO_REG_GPR:
619 620
		kvmppc_set_gpr(vcpu, vcpu->arch.io_gpr, gpr);
		break;
621 622
	case KVM_MMIO_REG_FPR:
		vcpu->arch.fpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
623
		break;
624
#ifdef CONFIG_PPC_BOOK3S
625 626
	case KVM_MMIO_REG_QPR:
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
627
		break;
628 629 630
	case KVM_MMIO_REG_FQPR:
		vcpu->arch.fpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
		vcpu->arch.qpr[vcpu->arch.io_gpr & KVM_MMIO_REG_MASK] = gpr;
631
		break;
632
#endif
633 634 635
	default:
		BUG();
	}
636 637 638 639 640
}

int kvmppc_handle_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
                       unsigned int rt, unsigned int bytes, int is_bigendian)
{
641 642
	int idx, ret;

643 644 645 646 647 648 649 650 651 652 653 654 655
	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;
A
Alexander Graf 已提交
656
	vcpu->arch.mmio_sign_extend = 0;
657

658 659 660 661 662 663 664 665
	idx = srcu_read_lock(&vcpu->kvm->srcu);

	ret = kvm_io_bus_read(vcpu->kvm, KVM_MMIO_BUS, run->mmio.phys_addr,
			      bytes, &run->mmio.data);

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

	if (!ret) {
A
Alexander Graf 已提交
666 667 668 669 670
		kvmppc_complete_mmio_load(vcpu, run);
		vcpu->mmio_needed = 0;
		return EMULATE_DONE;
	}

671 672
	return EMULATE_DO_MMIO;
}
673
EXPORT_SYMBOL_GPL(kvmppc_handle_load);
674

A
Alexander Graf 已提交
675 676 677 678 679 680 681
/* 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;

	vcpu->arch.mmio_sign_extend = 1;
A
Alexander Graf 已提交
682
	r = kvmppc_handle_load(run, vcpu, rt, bytes, is_bigendian);
A
Alexander Graf 已提交
683 684 685 686

	return r;
}

687
int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
688
                        u64 val, unsigned int bytes, int is_bigendian)
689 690
{
	void *data = run->mmio.data;
691
	int idx, ret;
692 693 694 695 696 697 698 699 700 701 702 703 704 705 706

	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) {
707
		case 8: *(u64 *)data = val; break;
708 709 710 711 712 713 714 715 716 717 718 719 720
		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;
		}
	}

721 722 723 724 725 726 727 728
	idx = srcu_read_lock(&vcpu->kvm->srcu);

	ret = kvm_io_bus_write(vcpu->kvm, KVM_MMIO_BUS, run->mmio.phys_addr,
			       bytes, &run->mmio.data);

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

	if (!ret) {
A
Alexander Graf 已提交
729 730 731 732
		vcpu->mmio_needed = 0;
		return EMULATE_DONE;
	}

733 734
	return EMULATE_DO_MMIO;
}
735
EXPORT_SYMBOL_GPL(kvmppc_handle_store);
736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752

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;
753 754 755 756 757 758 759
	} 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;
760 761 762 763 764 765 766
	} 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;
767 768 769 770 771
#ifdef CONFIG_BOOKE
	} else if (vcpu->arch.epr_needed) {
		kvmppc_set_epr(vcpu, run->epr.epr);
		vcpu->arch.epr_needed = 0;
#endif
772 773
	}

774
	r = kvmppc_vcpu_run(run, vcpu);
775 776 777 778 779 780 781 782 783

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

	return r;
}

int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
{
784
	if (irq->irq == KVM_INTERRUPT_UNSET) {
785
		kvmppc_core_dequeue_external(vcpu);
786 787 788 789
		return 0;
	}

	kvmppc_core_queue_external(vcpu, irq);
790

791
	kvm_vcpu_kick(vcpu);
792

793 794 795
	return 0;
}

796 797 798 799 800 801 802 803 804
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) {
805 806 807 808
	case KVM_CAP_PPC_OSI:
		r = 0;
		vcpu->arch.osi_enabled = true;
		break;
809 810 811 812
	case KVM_CAP_PPC_PAPR:
		r = 0;
		vcpu->arch.papr_enabled = true;
		break;
813 814
	case KVM_CAP_PPC_EPR:
		r = 0;
815 816 817 818
		if (cap->args[0])
			vcpu->arch.epr_flags |= KVMPPC_EPR_USER;
		else
			vcpu->arch.epr_flags &= ~KVMPPC_EPR_USER;
819
		break;
820 821 822 823 824 825
#ifdef CONFIG_BOOKE
	case KVM_CAP_PPC_BOOKE_WATCHDOG:
		r = 0;
		vcpu->arch.watchdog_enabled = true;
		break;
#endif
826
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
827 828 829 830 831 832 833 834 835 836
	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
837 838 839 840
	}
#endif
#ifdef CONFIG_KVM_MPIC
	case KVM_CAP_IRQ_MPIC: {
A
Al Viro 已提交
841
		struct fd f;
S
Scott Wood 已提交
842 843 844
		struct kvm_device *dev;

		r = -EBADF;
A
Al Viro 已提交
845 846
		f = fdget(cap->args[0]);
		if (!f.file)
S
Scott Wood 已提交
847 848 849
			break;

		r = -EPERM;
A
Al Viro 已提交
850
		dev = kvm_device_from_filp(f.file);
S
Scott Wood 已提交
851 852 853
		if (dev)
			r = kvmppc_mpic_connect_vcpu(dev, vcpu, cap->args[1]);

A
Al Viro 已提交
854
		fdput(f);
S
Scott Wood 已提交
855
		break;
S
Scott Wood 已提交
856 857
	}
#endif
858 859
#ifdef CONFIG_KVM_XICS
	case KVM_CAP_IRQ_XICS: {
A
Al Viro 已提交
860
		struct fd f;
861 862 863
		struct kvm_device *dev;

		r = -EBADF;
A
Al Viro 已提交
864 865
		f = fdget(cap->args[0]);
		if (!f.file)
866 867 868
			break;

		r = -EPERM;
A
Al Viro 已提交
869
		dev = kvm_device_from_filp(f.file);
870 871 872
		if (dev)
			r = kvmppc_xics_connect_vcpu(dev, vcpu, cap->args[1]);

A
Al Viro 已提交
873
		fdput(f);
874 875 876
		break;
	}
#endif /* CONFIG_KVM_XICS */
877 878 879 880 881
	default:
		r = -EINVAL;
		break;
	}

882 883 884
	if (!r)
		r = kvmppc_sanity_check(vcpu);

885 886 887
	return r;
}

888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906
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;

907 908
	switch (ioctl) {
	case KVM_INTERRUPT: {
909 910 911
		struct kvm_interrupt irq;
		r = -EFAULT;
		if (copy_from_user(&irq, argp, sizeof(irq)))
912
			goto out;
913
		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
914
		goto out;
915
	}
916

917 918 919 920 921 922 923 924 925
	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
926

927 928 929 930 931 932 933 934 935 936 937 938 939 940
	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;
	}

941
#if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC)
S
Scott Wood 已提交
942 943 944 945 946 947 948 949 950
	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
951 952 953 954 955 956 957 958
	default:
		r = -EINVAL;
	}

out:
	return r;
}

959 960 961 962 963
int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf)
{
	return VM_FAULT_SIGBUS;
}

964 965
static int kvm_vm_ioctl_get_pvinfo(struct kvm_ppc_pvinfo *pvinfo)
{
966 967 968 969 970 971 972 973
	u32 inst_nop = 0x60000000;
#ifdef CONFIG_KVM_BOOKE_HV
	u32 inst_sc1 = 0x44000022;
	pvinfo->hcall[0] = inst_sc1;
	pvinfo->hcall[1] = inst_nop;
	pvinfo->hcall[2] = inst_nop;
	pvinfo->hcall[3] = inst_nop;
#else
974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991
	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
	 */
	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;
992
#endif
993

994 995
	pvinfo->flags = KVM_PPC_PVINFO_FLAGS_EV_IDLE;

996 997 998
	return 0;
}

999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010
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;
}

1011 1012 1013
long kvm_arch_vm_ioctl(struct file *filp,
                       unsigned int ioctl, unsigned long arg)
{
1014
	struct kvm *kvm __maybe_unused = filp->private_data;
1015
	void __user *argp = (void __user *)arg;
1016 1017 1018
	long r;

	switch (ioctl) {
1019 1020
	case KVM_PPC_GET_PVINFO: {
		struct kvm_ppc_pvinfo pvinfo;
1021
		memset(&pvinfo, 0, sizeof(pvinfo));
1022 1023 1024 1025 1026 1027 1028 1029
		r = kvm_vm_ioctl_get_pvinfo(&pvinfo);
		if (copy_to_user(argp, &pvinfo, sizeof(pvinfo))) {
			r = -EFAULT;
			goto out;
		}

		break;
	}
1030
#ifdef CONFIG_PPC_BOOK3S_64
1031 1032 1033 1034 1035 1036 1037 1038 1039
	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;
	}
1040 1041 1042 1043
	case KVM_PPC_GET_SMMU_INFO: {
		struct kvm_ppc_smmu_info info;

		memset(&info, 0, sizeof(info));
1044
		r = kvmppc_ops->get_smmu_info(kvm, &info);
1045 1046 1047 1048
		if (r >= 0 && copy_to_user(argp, &info, sizeof(info)))
			r = -EFAULT;
		break;
	}
1049 1050 1051 1052 1053 1054
	case KVM_PPC_RTAS_DEFINE_TOKEN: {
		struct kvm *kvm = filp->private_data;

		r = kvm_vm_ioctl_rtas_define_token(kvm, argp);
		break;
	}
1055 1056 1057 1058
	default:
		r = kvmppc_ops->arch_vm_ioctl(filp, ioctl, arg);

#else /* CONFIG_PPC_BOOK3S_64 */
1059
	default:
1060
		r = -ENOTTY;
1061
#endif
1062
	}
1063
out:
1064 1065 1066
	return r;
}

1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083
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;
}
1084
EXPORT_SYMBOL_GPL(kvmppc_alloc_lpid);
1085 1086 1087 1088 1089

void kvmppc_claim_lpid(long lpid)
{
	set_bit(lpid, lpid_inuse);
}
1090
EXPORT_SYMBOL_GPL(kvmppc_claim_lpid);
1091 1092 1093 1094 1095

void kvmppc_free_lpid(long lpid)
{
	clear_bit(lpid, lpid_inuse);
}
1096
EXPORT_SYMBOL_GPL(kvmppc_free_lpid);
1097 1098 1099 1100 1101 1102

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));
}
1103
EXPORT_SYMBOL_GPL(kvmppc_init_lpid);
1104

1105 1106
int kvm_arch_init(void *opaque)
{
1107 1108 1109 1110 1111
	if (kvmppc_ops) {
		printk(KERN_ERR "kvm: already loaded the other module\n");
		return -EEXIST;
	}
	kvmppc_ops = (struct kvmppc_ops *)opaque;
1112 1113 1114 1115 1116
	return 0;
}

void kvm_arch_exit(void)
{
1117
	kvmppc_ops = NULL;
1118
}