kvm-s390.c 26.8 KB
Newer Older
1
/*
2
 * hosting zSeries kernel virtual machines
3
 *
4
 * Copyright IBM Corp. 2008, 2009
5 6 7 8 9 10 11 12
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License (version 2 only)
 * as published by the Free Software Foundation.
 *
 *    Author(s): Carsten Otte <cotte@de.ibm.com>
 *               Christian Borntraeger <borntraeger@de.ibm.com>
 *               Heiko Carstens <heiko.carstens@de.ibm.com>
13
 *               Christian Ehrhardt <ehrhardt@de.ibm.com>
14 15 16 17 18
 */

#include <linux/compiler.h>
#include <linux/err.h>
#include <linux/fs.h>
19
#include <linux/hrtimer.h>
20 21 22 23 24
#include <linux/init.h>
#include <linux/kvm.h>
#include <linux/kvm_host.h>
#include <linux/module.h>
#include <linux/slab.h>
25
#include <linux/timer.h>
26
#include <asm/asm-offsets.h>
27 28
#include <asm/lowcore.h>
#include <asm/pgtable.h>
29
#include <asm/nmi.h>
30
#include <asm/switch_to.h>
31
#include <asm/sclp.h>
32
#include "kvm-s390.h"
33 34
#include "gaccess.h"

35 36
#define CREATE_TRACE_POINTS
#include "trace.h"
37
#include "trace-s390.h"
38

39 40 41 42
#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU

struct kvm_stats_debugfs_item debugfs_entries[] = {
	{ "userspace_handled", VCPU_STAT(exit_userspace) },
43
	{ "exit_null", VCPU_STAT(exit_null) },
44 45 46 47
	{ "exit_validity", VCPU_STAT(exit_validity) },
	{ "exit_stop_request", VCPU_STAT(exit_stop_request) },
	{ "exit_external_request", VCPU_STAT(exit_external_request) },
	{ "exit_external_interrupt", VCPU_STAT(exit_external_interrupt) },
48 49 50
	{ "exit_instruction", VCPU_STAT(exit_instruction) },
	{ "exit_program_interruption", VCPU_STAT(exit_program_interruption) },
	{ "exit_instr_and_program_int", VCPU_STAT(exit_instr_and_program) },
51
	{ "instruction_lctlg", VCPU_STAT(instruction_lctlg) },
52 53
	{ "instruction_lctl", VCPU_STAT(instruction_lctl) },
	{ "deliver_emergency_signal", VCPU_STAT(deliver_emergency_signal) },
54
	{ "deliver_external_call", VCPU_STAT(deliver_external_call) },
55 56 57 58 59 60 61
	{ "deliver_service_signal", VCPU_STAT(deliver_service_signal) },
	{ "deliver_virtio_interrupt", VCPU_STAT(deliver_virtio_interrupt) },
	{ "deliver_stop_signal", VCPU_STAT(deliver_stop_signal) },
	{ "deliver_prefix_signal", VCPU_STAT(deliver_prefix_signal) },
	{ "deliver_restart_signal", VCPU_STAT(deliver_restart_signal) },
	{ "deliver_program_interruption", VCPU_STAT(deliver_program_int) },
	{ "exit_wait_state", VCPU_STAT(exit_wait_state) },
62 63 64 65 66 67 68 69 70
	{ "instruction_stidp", VCPU_STAT(instruction_stidp) },
	{ "instruction_spx", VCPU_STAT(instruction_spx) },
	{ "instruction_stpx", VCPU_STAT(instruction_stpx) },
	{ "instruction_stap", VCPU_STAT(instruction_stap) },
	{ "instruction_storage_key", VCPU_STAT(instruction_storage_key) },
	{ "instruction_stsch", VCPU_STAT(instruction_stsch) },
	{ "instruction_chsc", VCPU_STAT(instruction_chsc) },
	{ "instruction_stsi", VCPU_STAT(instruction_stsi) },
	{ "instruction_stfl", VCPU_STAT(instruction_stfl) },
71
	{ "instruction_tprot", VCPU_STAT(instruction_tprot) },
72
	{ "instruction_sigp_sense", VCPU_STAT(instruction_sigp_sense) },
73
	{ "instruction_sigp_sense_running", VCPU_STAT(instruction_sigp_sense_running) },
74
	{ "instruction_sigp_external_call", VCPU_STAT(instruction_sigp_external_call) },
75 76 77 78 79
	{ "instruction_sigp_emergency", VCPU_STAT(instruction_sigp_emergency) },
	{ "instruction_sigp_stop", VCPU_STAT(instruction_sigp_stop) },
	{ "instruction_sigp_set_arch", VCPU_STAT(instruction_sigp_arch) },
	{ "instruction_sigp_set_prefix", VCPU_STAT(instruction_sigp_prefix) },
	{ "instruction_sigp_restart", VCPU_STAT(instruction_sigp_restart) },
80
	{ "diagnose_10", VCPU_STAT(diagnose_10) },
81
	{ "diagnose_44", VCPU_STAT(diagnose_44) },
82
	{ "diagnose_9c", VCPU_STAT(diagnose_9c) },
83 84 85
	{ NULL }
};

86
static unsigned long long *facilities;
87 88

/* Section: not file related */
89
int kvm_arch_hardware_enable(void *garbage)
90 91
{
	/* every s390 is virtualization enabled ;-) */
92
	return 0;
93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
}

void kvm_arch_hardware_disable(void *garbage)
{
}

int kvm_arch_hardware_setup(void)
{
	return 0;
}

void kvm_arch_hardware_unsetup(void)
{
}

void kvm_arch_check_processor_compat(void *rtn)
{
}

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

void kvm_arch_exit(void)
{
}

/* Section: device related */
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg)
{
	if (ioctl == KVM_S390_ENABLE_SIE)
		return s390_enable_sie();
	return -EINVAL;
}

int kvm_dev_ioctl_check_extension(long ext)
{
132 133
	int r;

134
	switch (ext) {
135
	case KVM_CAP_S390_PSW:
136
	case KVM_CAP_S390_GMAP:
137
	case KVM_CAP_SYNC_MMU:
138 139 140
#ifdef CONFIG_KVM_S390_UCONTROL
	case KVM_CAP_S390_UCONTROL:
#endif
141
	case KVM_CAP_SYNC_REGS:
142
	case KVM_CAP_ONE_REG:
143
	case KVM_CAP_ENABLE_CAP:
144
	case KVM_CAP_S390_CSS_SUPPORT:
C
Cornelia Huck 已提交
145
	case KVM_CAP_IOEVENTFD:
146 147
		r = 1;
		break;
148 149 150 151
	case KVM_CAP_NR_VCPUS:
	case KVM_CAP_MAX_VCPUS:
		r = KVM_MAX_VCPUS;
		break;
152
	case KVM_CAP_S390_COW:
153
		r = MACHINE_HAS_ESOP;
154
		break;
155
	default:
156
		r = 0;
157
	}
158
	return r;
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178
}

/* Section: vm related */
/*
 * Get (and clear) the dirty memory log for a memory slot.
 */
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
			       struct kvm_dirty_log *log)
{
	return 0;
}

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

	switch (ioctl) {
179 180 181 182 183 184 185 186 187
	case KVM_S390_INTERRUPT: {
		struct kvm_s390_interrupt s390int;

		r = -EFAULT;
		if (copy_from_user(&s390int, argp, sizeof(s390int)))
			break;
		r = kvm_s390_inject_vm(kvm, &s390int);
		break;
	}
188
	default:
189
		r = -ENOTTY;
190 191 192 193 194
	}

	return r;
}

195
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
196 197 198 199
{
	int rc;
	char debug_name[16];

200 201 202 203 204 205 206 207 208 209 210
	rc = -EINVAL;
#ifdef CONFIG_KVM_S390_UCONTROL
	if (type & ~KVM_VM_S390_UCONTROL)
		goto out_err;
	if ((type & KVM_VM_S390_UCONTROL) && (!capable(CAP_SYS_ADMIN)))
		goto out_err;
#else
	if (type)
		goto out_err;
#endif

211 212
	rc = s390_enable_sie();
	if (rc)
213
		goto out_err;
214

215 216
	rc = -ENOMEM;

217 218
	kvm->arch.sca = (struct sca_block *) get_zeroed_page(GFP_KERNEL);
	if (!kvm->arch.sca)
219
		goto out_err;
220 221 222 223 224 225 226

	sprintf(debug_name, "kvm-%u", current->pid);

	kvm->arch.dbf = debug_register(debug_name, 8, 2, 8 * sizeof(long));
	if (!kvm->arch.dbf)
		goto out_nodbf;

227 228 229
	spin_lock_init(&kvm->arch.float_int.lock);
	INIT_LIST_HEAD(&kvm->arch.float_int.list);

230 231 232
	debug_register_view(kvm->arch.dbf, &debug_sprintf_view);
	VM_EVENT(kvm, 3, "%s", "vm created");

233 234 235 236 237 238 239
	if (type & KVM_VM_S390_UCONTROL) {
		kvm->arch.gmap = NULL;
	} else {
		kvm->arch.gmap = gmap_alloc(current->mm);
		if (!kvm->arch.gmap)
			goto out_nogmap;
	}
240 241 242

	kvm->arch.css_support = 0;

243
	return 0;
244 245
out_nogmap:
	debug_unregister(kvm->arch.dbf);
246 247
out_nodbf:
	free_page((unsigned long)(kvm->arch.sca));
248 249
out_err:
	return rc;
250 251
}

252 253 254
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
{
	VCPU_EVENT(vcpu, 3, "%s", "free cpu");
255
	trace_kvm_s390_destroy_vcpu(vcpu->vcpu_id);
C
Carsten Otte 已提交
256 257 258 259 260 261 262
	if (!kvm_is_ucontrol(vcpu->kvm)) {
		clear_bit(63 - vcpu->vcpu_id,
			  (unsigned long *) &vcpu->kvm->arch.sca->mcn);
		if (vcpu->kvm->arch.sca->cpu[vcpu->vcpu_id].sda ==
		    (__u64) vcpu->arch.sie_block)
			vcpu->kvm->arch.sca->cpu[vcpu->vcpu_id].sda = 0;
	}
263
	smp_mb();
264 265 266 267

	if (kvm_is_ucontrol(vcpu->kvm))
		gmap_free(vcpu->arch.gmap);

268
	free_page((unsigned long)(vcpu->arch.sie_block));
269
	kvm_vcpu_uninit(vcpu);
270 271 272 273 274 275
	kfree(vcpu);
}

static void kvm_free_vcpus(struct kvm *kvm)
{
	unsigned int i;
276
	struct kvm_vcpu *vcpu;
277

278 279 280 281 282 283 284 285 286
	kvm_for_each_vcpu(i, vcpu, kvm)
		kvm_arch_vcpu_destroy(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);
	mutex_unlock(&kvm->lock);
287 288
}

289 290 291 292
void kvm_arch_sync_events(struct kvm *kvm)
{
}

293 294
void kvm_arch_destroy_vm(struct kvm *kvm)
{
295
	kvm_free_vcpus(kvm);
296
	free_page((unsigned long)(kvm->arch.sca));
297
	debug_unregister(kvm->arch.dbf);
298 299
	if (!kvm_is_ucontrol(kvm))
		gmap_free(kvm->arch.gmap);
300 301 302 303 304
}

/* Section: vcpu related */
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
305 306 307 308 309 310 311
	if (kvm_is_ucontrol(vcpu->kvm)) {
		vcpu->arch.gmap = gmap_alloc(current->mm);
		if (!vcpu->arch.gmap)
			return -ENOMEM;
		return 0;
	}

312
	vcpu->arch.gmap = vcpu->kvm->arch.gmap;
313 314
	vcpu->run->kvm_valid_regs = KVM_SYNC_PREFIX |
				    KVM_SYNC_GPRS |
315 316
				    KVM_SYNC_ACRS |
				    KVM_SYNC_CRS;
317 318 319 320 321
	return 0;
}

void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
322
	/* Nothing todo */
323 324 325 326 327 328 329 330
}

void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
	save_fp_regs(&vcpu->arch.host_fpregs);
	save_access_regs(vcpu->arch.host_acrs);
	vcpu->arch.guest_fpregs.fpc &= FPC_VALID_MASK;
	restore_fp_regs(&vcpu->arch.guest_fpregs);
331
	restore_access_regs(vcpu->run->s.regs.acrs);
332
	gmap_enable(vcpu->arch.gmap);
333
	atomic_set_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags);
334 335 336 337
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
338
	atomic_clear_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags);
339
	gmap_disable(vcpu->arch.gmap);
340
	save_fp_regs(&vcpu->arch.guest_fpregs);
341
	save_access_regs(vcpu->run->s.regs.acrs);
342 343 344 345 346 347 348 349 350
	restore_fp_regs(&vcpu->arch.host_fpregs);
	restore_access_regs(vcpu->arch.host_acrs);
}

static void kvm_s390_vcpu_initial_reset(struct kvm_vcpu *vcpu)
{
	/* this equals initial cpu reset in pop, but we don't switch to ESA */
	vcpu->arch.sie_block->gpsw.mask = 0UL;
	vcpu->arch.sie_block->gpsw.addr = 0UL;
351
	kvm_s390_set_prefix(vcpu, 0);
352 353 354 355 356 357 358 359 360
	vcpu->arch.sie_block->cputm     = 0UL;
	vcpu->arch.sie_block->ckc       = 0UL;
	vcpu->arch.sie_block->todpr     = 0;
	memset(vcpu->arch.sie_block->gcr, 0, 16 * sizeof(__u64));
	vcpu->arch.sie_block->gcr[0]  = 0xE0UL;
	vcpu->arch.sie_block->gcr[14] = 0xC2000000UL;
	vcpu->arch.guest_fpregs.fpc = 0;
	asm volatile("lfpc %0" : : "Q" (vcpu->arch.guest_fpregs.fpc));
	vcpu->arch.sie_block->gbea = 1;
361
	atomic_set_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags);
362 363
}

364 365 366 367 368
int kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
{
	return 0;
}

369 370
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
{
371 372 373
	atomic_set(&vcpu->arch.sie_block->cpuflags, CPUSTAT_ZARCH |
						    CPUSTAT_SM |
						    CPUSTAT_STOPPED);
374
	vcpu->arch.sie_block->ecb   = 6;
375
	vcpu->arch.sie_block->eca   = 0xC1002001U;
376
	vcpu->arch.sie_block->fac   = (int) (long) facilities;
377 378 379 380
	hrtimer_init(&vcpu->arch.ckc_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
	tasklet_init(&vcpu->arch.tasklet, kvm_s390_tasklet,
		     (unsigned long) vcpu);
	vcpu->arch.ckc_timer.function = kvm_s390_idle_wakeup;
381
	get_cpu_id(&vcpu->arch.cpu_id);
382
	vcpu->arch.cpu_id.version = 0xff;
383 384 385 386 387 388
	return 0;
}

struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
				      unsigned int id)
{
389 390 391 392 393 394 395
	struct kvm_vcpu *vcpu;
	int rc = -EINVAL;

	if (id >= KVM_MAX_VCPUS)
		goto out;

	rc = -ENOMEM;
396

397
	vcpu = kzalloc(sizeof(struct kvm_vcpu), GFP_KERNEL);
398
	if (!vcpu)
399
		goto out;
400

401 402
	vcpu->arch.sie_block = (struct kvm_s390_sie_block *)
					get_zeroed_page(GFP_KERNEL);
403 404 405 406 407

	if (!vcpu->arch.sie_block)
		goto out_free_cpu;

	vcpu->arch.sie_block->icpua = id;
C
Carsten Otte 已提交
408 409 410 411 412 413 414 415 416 417 418 419 420
	if (!kvm_is_ucontrol(kvm)) {
		if (!kvm->arch.sca) {
			WARN_ON_ONCE(1);
			goto out_free_cpu;
		}
		if (!kvm->arch.sca->cpu[id].sda)
			kvm->arch.sca->cpu[id].sda =
				(__u64) vcpu->arch.sie_block;
		vcpu->arch.sie_block->scaoh =
			(__u32)(((__u64)kvm->arch.sca) >> 32);
		vcpu->arch.sie_block->scaol = (__u32)(__u64)kvm->arch.sca;
		set_bit(63 - id, (unsigned long *) &kvm->arch.sca->mcn);
	}
421

422 423 424
	spin_lock_init(&vcpu->arch.local_int.lock);
	INIT_LIST_HEAD(&vcpu->arch.local_int.list);
	vcpu->arch.local_int.float_int = &kvm->arch.float_int;
425
	spin_lock(&kvm->arch.float_int.lock);
426 427
	kvm->arch.float_int.local_int[id] = &vcpu->arch.local_int;
	init_waitqueue_head(&vcpu->arch.local_int.wq);
428
	vcpu->arch.local_int.cpuflags = &vcpu->arch.sie_block->cpuflags;
429
	spin_unlock(&kvm->arch.float_int.lock);
430

431 432
	rc = kvm_vcpu_init(vcpu, kvm, id);
	if (rc)
433
		goto out_free_sie_block;
434 435
	VM_EVENT(kvm, 3, "create cpu %d at %p, sie block at %p", id, vcpu,
		 vcpu->arch.sie_block);
436
	trace_kvm_s390_create_vcpu(id, vcpu, vcpu->arch.sie_block);
437 438

	return vcpu;
439 440
out_free_sie_block:
	free_page((unsigned long)(vcpu->arch.sie_block));
441 442
out_free_cpu:
	kfree(vcpu);
443
out:
444 445 446 447 448 449 450 451 452 453
	return ERR_PTR(rc);
}

int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
{
	/* kvm common code refers to this, but never calls it */
	BUG();
	return 0;
}

454 455 456 457 458 459 460
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu)
{
	/* kvm common code refers to this, but never calls it */
	BUG();
	return 0;
}

461 462 463 464 465 466
static int kvm_arch_vcpu_ioctl_get_one_reg(struct kvm_vcpu *vcpu,
					   struct kvm_one_reg *reg)
{
	int r = -EINVAL;

	switch (reg->id) {
467 468 469 470 471 472 473 474
	case KVM_REG_S390_TODPR:
		r = put_user(vcpu->arch.sie_block->todpr,
			     (u32 __user *)reg->addr);
		break;
	case KVM_REG_S390_EPOCHDIFF:
		r = put_user(vcpu->arch.sie_block->epoch,
			     (u64 __user *)reg->addr);
		break;
475 476 477 478 479 480 481 482
	case KVM_REG_S390_CPU_TIMER:
		r = put_user(vcpu->arch.sie_block->cputm,
			     (u64 __user *)reg->addr);
		break;
	case KVM_REG_S390_CLOCK_COMP:
		r = put_user(vcpu->arch.sie_block->ckc,
			     (u64 __user *)reg->addr);
		break;
483 484 485 486 487 488 489 490 491 492 493 494 495
	default:
		break;
	}

	return r;
}

static int kvm_arch_vcpu_ioctl_set_one_reg(struct kvm_vcpu *vcpu,
					   struct kvm_one_reg *reg)
{
	int r = -EINVAL;

	switch (reg->id) {
496 497 498 499 500 501 502 503
	case KVM_REG_S390_TODPR:
		r = get_user(vcpu->arch.sie_block->todpr,
			     (u32 __user *)reg->addr);
		break;
	case KVM_REG_S390_EPOCHDIFF:
		r = get_user(vcpu->arch.sie_block->epoch,
			     (u64 __user *)reg->addr);
		break;
504 505 506 507 508 509 510 511
	case KVM_REG_S390_CPU_TIMER:
		r = get_user(vcpu->arch.sie_block->cputm,
			     (u64 __user *)reg->addr);
		break;
	case KVM_REG_S390_CLOCK_COMP:
		r = get_user(vcpu->arch.sie_block->ckc,
			     (u64 __user *)reg->addr);
		break;
512 513 514 515 516 517
	default:
		break;
	}

	return r;
}
518

519 520 521 522 523 524 525 526
static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
{
	kvm_s390_vcpu_initial_reset(vcpu);
	return 0;
}

int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
527
	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
528 529 530 531 532
	return 0;
}

int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
533
	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
534 535 536 537 538 539
	return 0;
}

int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs)
{
540
	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
541
	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
542
	restore_access_regs(vcpu->run->s.regs.acrs);
543 544 545 546 547 548
	return 0;
}

int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs)
{
549
	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
550 551 552 553 554 555 556
	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
	return 0;
}

int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
	memcpy(&vcpu->arch.guest_fpregs.fprs, &fpu->fprs, sizeof(fpu->fprs));
557
	vcpu->arch.guest_fpregs.fpc = fpu->fpc & FPC_VALID_MASK;
C
Carsten Otte 已提交
558
	restore_fp_regs(&vcpu->arch.guest_fpregs);
559 560 561 562 563 564 565 566 567 568 569 570 571 572
	return 0;
}

int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
	memcpy(&fpu->fprs, &vcpu->arch.guest_fpregs.fprs, sizeof(fpu->fprs));
	fpu->fpc = vcpu->arch.guest_fpregs.fpc;
	return 0;
}

static int kvm_arch_vcpu_ioctl_set_initial_psw(struct kvm_vcpu *vcpu, psw_t psw)
{
	int rc = 0;

573
	if (!(atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_STOPPED))
574
		rc = -EBUSY;
575 576 577 578
	else {
		vcpu->run->psw_mask = psw.mask;
		vcpu->run->psw_addr = psw.addr;
	}
579 580 581 582 583 584 585 586 587
	return rc;
}

int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				  struct kvm_translation *tr)
{
	return -EINVAL; /* not implemented yet */
}

J
Jan Kiszka 已提交
588 589
int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
					struct kvm_guest_debug *dbg)
590 591 592 593
{
	return -EINVAL; /* not implemented yet */
}

594 595 596 597 598 599 600 601 602 603 604 605
int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
				    struct kvm_mp_state *mp_state)
{
	return -EINVAL; /* not implemented yet */
}

int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
				    struct kvm_mp_state *mp_state)
{
	return -EINVAL; /* not implemented yet */
}

606
static int __vcpu_run(struct kvm_vcpu *vcpu)
607
{
608 609
	int rc;

610
	memcpy(&vcpu->arch.sie_block->gg14, &vcpu->run->s.regs.gprs[14], 16);
611 612 613 614

	if (need_resched())
		schedule();

615 616 617
	if (test_thread_flag(TIF_MCCK_PENDING))
		s390_handle_mcck();

618 619
	if (!kvm_is_ucontrol(vcpu->kvm))
		kvm_s390_deliver_pending_interrupts(vcpu);
C
Carsten Otte 已提交
620

621
	vcpu->arch.sie_block->icptcode = 0;
622
	preempt_disable();
623
	kvm_guest_enter();
624
	preempt_enable();
625 626
	VCPU_EVENT(vcpu, 6, "entering sie flags %x",
		   atomic_read(&vcpu->arch.sie_block->cpuflags));
627 628
	trace_kvm_s390_sie_enter(vcpu,
				 atomic_read(&vcpu->arch.sie_block->cpuflags));
629
	rc = sie64a(vcpu->arch.sie_block, vcpu->run->s.regs.gprs);
630 631 632 633 634
	if (rc) {
		if (kvm_is_ucontrol(vcpu->kvm)) {
			rc = SIE_INTERCEPT_UCONTROL;
		} else {
			VCPU_EVENT(vcpu, 3, "%s", "fault in sie instruction");
635
			trace_kvm_s390_sie_fault(vcpu);
636 637 638
			kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
			rc = 0;
		}
639
	}
640 641
	VCPU_EVENT(vcpu, 6, "exit sie icptcode %d",
		   vcpu->arch.sie_block->icptcode);
642
	trace_kvm_s390_sie_exit(vcpu, vcpu->arch.sie_block->icptcode);
643 644
	kvm_guest_exit();

645
	memcpy(&vcpu->run->s.regs.gprs[14], &vcpu->arch.sie_block->gg14, 16);
646
	return rc;
647 648 649 650
}

int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
{
651
	int rc;
652 653
	sigset_t sigsaved;

654
rerun_vcpu:
655 656 657
	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);

658
	atomic_clear_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags);
659

660 661
	BUG_ON(vcpu->kvm->arch.float_int.local_int[vcpu->vcpu_id] == NULL);

662 663 664
	switch (kvm_run->exit_reason) {
	case KVM_EXIT_S390_SIEIC:
	case KVM_EXIT_UNKNOWN:
665
	case KVM_EXIT_INTR:
666
	case KVM_EXIT_S390_RESET:
667
	case KVM_EXIT_S390_UCONTROL:
668
	case KVM_EXIT_S390_TSCH:
669 670 671 672 673
		break;
	default:
		BUG();
	}

674 675
	vcpu->arch.sie_block->gpsw.mask = kvm_run->psw_mask;
	vcpu->arch.sie_block->gpsw.addr = kvm_run->psw_addr;
676 677 678 679
	if (kvm_run->kvm_dirty_regs & KVM_SYNC_PREFIX) {
		kvm_run->kvm_dirty_regs &= ~KVM_SYNC_PREFIX;
		kvm_s390_set_prefix(vcpu, kvm_run->s.regs.prefix);
	}
680 681 682 683 684
	if (kvm_run->kvm_dirty_regs & KVM_SYNC_CRS) {
		kvm_run->kvm_dirty_regs &= ~KVM_SYNC_CRS;
		memcpy(&vcpu->arch.sie_block->gcr, &kvm_run->s.regs.crs, 128);
		kvm_s390_set_prefix(vcpu, kvm_run->s.regs.prefix);
	}
685

686
	might_fault();
687 688

	do {
689 690 691
		rc = __vcpu_run(vcpu);
		if (rc)
			break;
692 693 694 695
		if (kvm_is_ucontrol(vcpu->kvm))
			rc = -EOPNOTSUPP;
		else
			rc = kvm_handle_sie_intercept(vcpu);
696 697
	} while (!signal_pending(current) && !rc);

698 699 700
	if (rc == SIE_INTERCEPT_RERUNVCPU)
		goto rerun_vcpu;

701 702
	if (signal_pending(current) && !rc) {
		kvm_run->exit_reason = KVM_EXIT_INTR;
703
		rc = -EINTR;
704
	}
705

706 707 708 709 710 711 712 713 714 715
#ifdef CONFIG_KVM_S390_UCONTROL
	if (rc == SIE_INTERCEPT_UCONTROL) {
		kvm_run->exit_reason = KVM_EXIT_S390_UCONTROL;
		kvm_run->s390_ucontrol.trans_exc_code =
			current->thread.gmap_addr;
		kvm_run->s390_ucontrol.pgm_code = 0x10;
		rc = 0;
	}
#endif

716
	if (rc == -EOPNOTSUPP) {
717 718 719 720 721 722 723 724 725 726 727 728 729
		/* intercept cannot be handled in-kernel, prepare kvm-run */
		kvm_run->exit_reason         = KVM_EXIT_S390_SIEIC;
		kvm_run->s390_sieic.icptcode = vcpu->arch.sie_block->icptcode;
		kvm_run->s390_sieic.ipa      = vcpu->arch.sie_block->ipa;
		kvm_run->s390_sieic.ipb      = vcpu->arch.sie_block->ipb;
		rc = 0;
	}

	if (rc == -EREMOTE) {
		/* intercept was handled, but userspace support is needed
		 * kvm_run has been prepared by the handler */
		rc = 0;
	}
730

731 732
	kvm_run->psw_mask     = vcpu->arch.sie_block->gpsw.mask;
	kvm_run->psw_addr     = vcpu->arch.sie_block->gpsw.addr;
733
	kvm_run->s.regs.prefix = vcpu->arch.sie_block->prefix;
734
	memcpy(&kvm_run->s.regs.crs, &vcpu->arch.sie_block->gcr, 128);
735

736 737 738 739
	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &sigsaved, NULL);

	vcpu->stat.exit_userspace++;
740
	return rc;
741 742
}

743
static int __guestcopy(struct kvm_vcpu *vcpu, u64 guestdest, void *from,
744 745 746 747 748 749 750 751 752 753 754 755 756 757
		       unsigned long n, int prefix)
{
	if (prefix)
		return copy_to_guest(vcpu, guestdest, from, n);
	else
		return copy_to_guest_absolute(vcpu, guestdest, from, n);
}

/*
 * store status at address
 * we use have two special cases:
 * KVM_S390_STORE_STATUS_NOADDR: -> 0x1200 on 64 bit
 * KVM_S390_STORE_STATUS_PREFIXED: -> prefix
 */
758
int kvm_s390_vcpu_store_status(struct kvm_vcpu *vcpu, unsigned long addr)
759
{
760
	unsigned char archmode = 1;
761 762 763 764 765 766 767 768 769 770 771 772 773 774 775
	int prefix;

	if (addr == KVM_S390_STORE_STATUS_NOADDR) {
		if (copy_to_guest_absolute(vcpu, 163ul, &archmode, 1))
			return -EFAULT;
		addr = SAVE_AREA_BASE;
		prefix = 0;
	} else if (addr == KVM_S390_STORE_STATUS_PREFIXED) {
		if (copy_to_guest(vcpu, 163ul, &archmode, 1))
			return -EFAULT;
		addr = SAVE_AREA_BASE;
		prefix = 1;
	} else
		prefix = 0;

776 777 778 779 780 781 782 783
	/*
	 * The guest FPRS and ACRS are in the host FPRS/ACRS due to the lazy
	 * copying in vcpu load/put. Lets update our copies before we save
	 * it into the save area
	 */
	save_fp_regs(&vcpu->arch.guest_fpregs);
	save_access_regs(vcpu->run->s.regs.acrs);

784
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, fp_regs),
785 786 787
			vcpu->arch.guest_fpregs.fprs, 128, prefix))
		return -EFAULT;

788
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, gp_regs),
789
			vcpu->run->s.regs.gprs, 128, prefix))
790 791
		return -EFAULT;

792
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, psw),
793 794 795
			&vcpu->arch.sie_block->gpsw, 16, prefix))
		return -EFAULT;

796
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, pref_reg),
797 798 799 800
			&vcpu->arch.sie_block->prefix, 4, prefix))
		return -EFAULT;

	if (__guestcopy(vcpu,
801
			addr + offsetof(struct save_area, fp_ctrl_reg),
802 803 804
			&vcpu->arch.guest_fpregs.fpc, 4, prefix))
		return -EFAULT;

805
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, tod_reg),
806 807 808
			&vcpu->arch.sie_block->todpr, 4, prefix))
		return -EFAULT;

809
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, timer),
810 811 812
			&vcpu->arch.sie_block->cputm, 8, prefix))
		return -EFAULT;

813
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, clk_cmp),
814 815 816
			&vcpu->arch.sie_block->ckc, 8, prefix))
		return -EFAULT;

817
	if (__guestcopy(vcpu, addr + offsetof(struct save_area, acc_regs),
818
			&vcpu->run->s.regs.acrs, 64, prefix))
819 820 821
		return -EFAULT;

	if (__guestcopy(vcpu,
822
			addr + offsetof(struct save_area, ctrl_regs),
823 824 825 826 827
			&vcpu->arch.sie_block->gcr, 128, prefix))
		return -EFAULT;
	return 0;
}

828 829 830 831 832 833 834 835 836
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) {
837 838 839 840 841 842 843
	case KVM_CAP_S390_CSS_SUPPORT:
		if (!vcpu->kvm->arch.css_support) {
			vcpu->kvm->arch.css_support = 1;
			trace_kvm_s390_enable_css(vcpu->kvm);
		}
		r = 0;
		break;
844 845 846 847 848 849 850
	default:
		r = -EINVAL;
		break;
	}
	return r;
}

851 852 853 854 855
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;
856
	long r;
857

858 859
	switch (ioctl) {
	case KVM_S390_INTERRUPT: {
860 861
		struct kvm_s390_interrupt s390int;

862
		r = -EFAULT;
863
		if (copy_from_user(&s390int, argp, sizeof(s390int)))
864 865 866
			break;
		r = kvm_s390_inject_vcpu(vcpu, &s390int);
		break;
867
	}
868
	case KVM_S390_STORE_STATUS:
869 870
		r = kvm_s390_vcpu_store_status(vcpu, arg);
		break;
871 872 873
	case KVM_S390_SET_INITIAL_PSW: {
		psw_t psw;

874
		r = -EFAULT;
875
		if (copy_from_user(&psw, argp, sizeof(psw)))
876 877 878
			break;
		r = kvm_arch_vcpu_ioctl_set_initial_psw(vcpu, psw);
		break;
879 880
	}
	case KVM_S390_INITIAL_RESET:
881 882
		r = kvm_arch_vcpu_ioctl_initial_reset(vcpu);
		break;
883 884 885 886 887 888 889 890 891 892 893 894
	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)))
			break;
		if (ioctl == KVM_SET_ONE_REG)
			r = kvm_arch_vcpu_ioctl_set_one_reg(vcpu, &reg);
		else
			r = kvm_arch_vcpu_ioctl_get_one_reg(vcpu, &reg);
		break;
	}
895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930
#ifdef CONFIG_KVM_S390_UCONTROL
	case KVM_S390_UCAS_MAP: {
		struct kvm_s390_ucas_mapping ucasmap;

		if (copy_from_user(&ucasmap, argp, sizeof(ucasmap))) {
			r = -EFAULT;
			break;
		}

		if (!kvm_is_ucontrol(vcpu->kvm)) {
			r = -EINVAL;
			break;
		}

		r = gmap_map_segment(vcpu->arch.gmap, ucasmap.user_addr,
				     ucasmap.vcpu_addr, ucasmap.length);
		break;
	}
	case KVM_S390_UCAS_UNMAP: {
		struct kvm_s390_ucas_mapping ucasmap;

		if (copy_from_user(&ucasmap, argp, sizeof(ucasmap))) {
			r = -EFAULT;
			break;
		}

		if (!kvm_is_ucontrol(vcpu->kvm)) {
			r = -EINVAL;
			break;
		}

		r = gmap_unmap_segment(vcpu->arch.gmap, ucasmap.vcpu_addr,
			ucasmap.length);
		break;
	}
#endif
931 932 933 934 935 936
	case KVM_S390_VCPU_FAULT: {
		r = gmap_fault(arg, vcpu->arch.gmap);
		if (!IS_ERR_VALUE(r))
			r = 0;
		break;
	}
937 938 939 940 941 942 943 944 945
	case KVM_ENABLE_CAP:
	{
		struct kvm_enable_cap cap;
		r = -EFAULT;
		if (copy_from_user(&cap, argp, sizeof(cap)))
			break;
		r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
		break;
	}
946
	default:
947
		r = -ENOTTY;
948
	}
949
	return r;
950 951
}

952 953 954 955 956 957 958 959 960 961 962 963 964
int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf)
{
#ifdef CONFIG_KVM_S390_UCONTROL
	if ((vmf->pgoff == KVM_S390_SIE_PAGE_OFFSET)
		 && (kvm_is_ucontrol(vcpu->kvm))) {
		vmf->page = virt_to_page(vcpu->arch.sie_block);
		get_page(vmf->page);
		return 0;
	}
#endif
	return VM_FAULT_SIGBUS;
}

965 966 967 968 969 970 971 972 973 974
void kvm_arch_free_memslot(struct kvm_memory_slot *free,
			   struct kvm_memory_slot *dont)
{
}

int kvm_arch_create_memslot(struct kvm_memory_slot *slot, unsigned long npages)
{
	return 0;
}

975
/* Section: memory related */
976 977
int kvm_arch_prepare_memory_region(struct kvm *kvm,
				   struct kvm_memory_slot *memslot,
978 979
				   struct kvm_userspace_memory_region *mem,
				   enum kvm_mr_change change)
980 981 982 983 984 985 986 987
{
	/* A few sanity checks. We can have exactly one memory slot which has
	   to start at guest virtual zero and which has to be located at a
	   page boundary in userland and which has to end at a page boundary.
	   The memory in userland is ok to be fragmented into various different
	   vmas. It is okay to mmap() and munmap() stuff in this slot after
	   doing this call at any time */

988
	if (mem->slot)
989 990 991 992 993
		return -EINVAL;

	if (mem->guest_phys_addr)
		return -EINVAL;

994
	if (mem->userspace_addr & 0xffffful)
995 996
		return -EINVAL;

997
	if (mem->memory_size & 0xffffful)
998 999
		return -EINVAL;

1000 1001 1002 1003 1004
	return 0;
}

void kvm_arch_commit_memory_region(struct kvm *kvm,
				struct kvm_userspace_memory_region *mem,
1005 1006
				const struct kvm_memory_slot *old,
				enum kvm_mr_change change)
1007
{
1008
	int rc;
1009

1010 1011 1012 1013 1014 1015 1016 1017 1018 1019
	/* If the basics of the memslot do not change, we do not want
	 * to update the gmap. Every update causes several unnecessary
	 * segment translation exceptions. This is usually handled just
	 * fine by the normal fault handler + gmap, but it will also
	 * cause faults on the prefix page of running guest CPUs.
	 */
	if (old->userspace_addr == mem->userspace_addr &&
	    old->base_gfn * PAGE_SIZE == mem->guest_phys_addr &&
	    old->npages * PAGE_SIZE == mem->memory_size)
		return;
1020 1021 1022 1023

	rc = gmap_map_segment(kvm->arch.gmap, mem->userspace_addr,
		mem->guest_phys_addr, mem->memory_size);
	if (rc)
1024
		printk(KERN_WARNING "kvm-s390: failed to commit memory region\n");
1025
	return;
1026 1027
}

1028 1029 1030 1031 1032 1033
void kvm_arch_flush_shadow_all(struct kvm *kvm)
{
}

void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
				   struct kvm_memory_slot *slot)
1034 1035 1036
{
}

1037 1038
static int __init kvm_s390_init(void)
{
1039
	int ret;
1040
	ret = kvm_init(NULL, sizeof(struct kvm_vcpu), 0, THIS_MODULE);
1041 1042 1043 1044 1045
	if (ret)
		return ret;

	/*
	 * guests can ask for up to 255+1 double words, we need a full page
L
Lucas De Marchi 已提交
1046
	 * to hold the maximum amount of facilities. On the other hand, we
1047 1048
	 * only set facilities that are known to work in KVM.
	 */
1049
	facilities = (unsigned long long *) get_zeroed_page(GFP_KERNEL|GFP_DMA);
1050 1051 1052 1053
	if (!facilities) {
		kvm_exit();
		return -ENOMEM;
	}
1054
	memcpy(facilities, S390_lowcore.stfle_fac_list, 16);
1055
	facilities[0] &= 0xff00fff3f47c0000ULL;
1056
	facilities[1] &= 0x001c000000000000ULL;
1057
	return 0;
1058 1059 1060 1061
}

static void __exit kvm_s390_exit(void)
{
1062
	free_page((unsigned long) facilities);
1063 1064 1065 1066 1067
	kvm_exit();
}

module_init(kvm_s390_init);
module_exit(kvm_s390_exit);