priv.c 27.9 KB
Newer Older
1
/*
2
 * handling privileged instructions
3
 *
4
 * Copyright IBM Corp. 2008, 2013
5 6 7 8 9 10 11 12 13 14
 *
 * 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>
 */

#include <linux/kvm.h>
15
#include <linux/gfp.h>
16
#include <linux/errno.h>
17
#include <linux/compat.h>
18
#include <asm/asm-offsets.h>
19
#include <asm/facility.h>
20 21 22 23
#include <asm/current.h>
#include <asm/debug.h>
#include <asm/ebcdic.h>
#include <asm/sysinfo.h>
24 25 26
#include <asm/pgtable.h>
#include <asm/pgalloc.h>
#include <asm/io.h>
27 28
#include <asm/ptrace.h>
#include <asm/compat.h>
29 30
#include "gaccess.h"
#include "kvm-s390.h"
31
#include "trace.h"
32

33 34 35 36 37
/* Handle SCK (SET CLOCK) interception */
static int handle_set_clock(struct kvm_vcpu *vcpu)
{
	struct kvm_vcpu *cpup;
	s64 hostclk, val;
38
	int i, rc;
39
	ar_t ar;
40 41 42 43 44
	u64 op2;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

45
	op2 = kvm_s390_get_base_disp_s(vcpu, &ar);
46 47
	if (op2 & 7)	/* Operand must be on a doubleword boundary */
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
48
	rc = read_guest(vcpu, op2, ar, &val, sizeof(val));
49 50
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

	if (store_tod_clock(&hostclk)) {
		kvm_s390_set_psw_cc(vcpu, 3);
		return 0;
	}
	val = (val - hostclk) & ~0x3fUL;

	mutex_lock(&vcpu->kvm->lock);
	kvm_for_each_vcpu(i, cpup, vcpu->kvm)
		cpup->arch.sie_block->epoch = val;
	mutex_unlock(&vcpu->kvm->lock);

	kvm_s390_set_psw_cc(vcpu, 0);
	return 0;
}

67 68 69
static int handle_set_prefix(struct kvm_vcpu *vcpu)
{
	u64 operand2;
70 71
	u32 address;
	int rc;
72
	ar_t ar;
73 74 75

	vcpu->stat.instruction_spx++;

76 77 78
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

79
	operand2 = kvm_s390_get_base_disp_s(vcpu, &ar);
80 81

	/* must be word boundary */
82 83
	if (operand2 & 3)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
84 85

	/* get the value */
86
	rc = read_guest(vcpu, operand2, ar, &address, sizeof(address));
87 88 89 90 91 92 93 94 95 96 97
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);

	address &= 0x7fffe000u;

	/*
	 * Make sure the new value is valid memory. We only need to check the
	 * first page, since address is 8k aligned and memory pieces are always
	 * at least 1MB aligned and have at least a size of 1MB.
	 */
	if (kvm_is_error_gpa(vcpu->kvm, address))
98
		return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
99

100
	kvm_s390_set_prefix(vcpu, address);
101
	trace_kvm_s390_handle_prefix(vcpu, 1, address);
102 103 104 105 106 107 108
	return 0;
}

static int handle_store_prefix(struct kvm_vcpu *vcpu)
{
	u64 operand2;
	u32 address;
109
	int rc;
110
	ar_t ar;
111 112

	vcpu->stat.instruction_stpx++;
113

114 115 116
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

117
	operand2 = kvm_s390_get_base_disp_s(vcpu, &ar);
118 119

	/* must be word boundary */
120 121
	if (operand2 & 3)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
122

123
	address = kvm_s390_get_prefix(vcpu);
124 125

	/* get the value */
126
	rc = write_guest(vcpu, operand2, ar, &address, sizeof(address));
127 128
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
129 130

	VCPU_EVENT(vcpu, 5, "storing prefix to %x", address);
131
	trace_kvm_s390_handle_prefix(vcpu, 0, address);
132 133 134 135 136
	return 0;
}

static int handle_store_cpu_address(struct kvm_vcpu *vcpu)
{
137 138 139
	u16 vcpu_id = vcpu->vcpu_id;
	u64 ga;
	int rc;
140
	ar_t ar;
141 142

	vcpu->stat.instruction_stap++;
143

144 145 146
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

147
	ga = kvm_s390_get_base_disp_s(vcpu, &ar);
148

149
	if (ga & 1)
150
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
151

152
	rc = write_guest(vcpu, ga, ar, &vcpu_id, sizeof(vcpu_id));
153 154
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
155

156 157
	VCPU_EVENT(vcpu, 5, "storing cpu address to %llx", ga);
	trace_kvm_s390_handle_stap(vcpu, ga);
158 159 160
	return 0;
}

161
static int __skey_check_enable(struct kvm_vcpu *vcpu)
162
{
163
	int rc = 0;
164
	if (!(vcpu->arch.sie_block->ictl & (ICTL_ISKE | ICTL_SSKE | ICTL_RRBE)))
165
		return rc;
166

167
	rc = s390_enable_skey();
168 169
	trace_kvm_s390_skey_related_inst(vcpu);
	vcpu->arch.sie_block->ictl &= ~(ICTL_ISKE | ICTL_SSKE | ICTL_RRBE);
170
	return rc;
171 172 173
}


174 175
static int handle_skey(struct kvm_vcpu *vcpu)
{
176
	int rc = __skey_check_enable(vcpu);
177

178 179
	if (rc)
		return rc;
180
	vcpu->stat.instruction_storage_key++;
181 182 183 184

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

185
	kvm_s390_rewind_psw(vcpu, 4);
186 187 188 189
	VCPU_EVENT(vcpu, 4, "%s", "retrying storage key operation");
	return 0;
}

190 191 192
static int handle_ipte_interlock(struct kvm_vcpu *vcpu)
{
	vcpu->stat.instruction_ipte_interlock++;
193
	if (psw_bits(vcpu->arch.sie_block->gpsw).p)
194 195
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);
	wait_event(vcpu->kvm->arch.ipte_wq, !ipte_lock_held(vcpu));
196
	kvm_s390_rewind_psw(vcpu, 4);
197 198 199 200
	VCPU_EVENT(vcpu, 4, "%s", "retrying ipte interlock operation");
	return 0;
}

T
Thomas Huth 已提交
201 202 203 204 205 206 207 208 209 210
static int handle_test_block(struct kvm_vcpu *vcpu)
{
	gpa_t addr;
	int reg2;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

	kvm_s390_get_regs_rre(vcpu, NULL, &reg2);
	addr = vcpu->run->s.regs.gprs[reg2] & PAGE_MASK;
211
	addr = kvm_s390_logical_to_effective(vcpu, addr);
212
	if (kvm_s390_check_low_addr_prot_real(vcpu, addr))
213
		return kvm_s390_inject_prog_irq(vcpu, &vcpu->arch.pgm);
T
Thomas Huth 已提交
214 215
	addr = kvm_s390_real_to_abs(vcpu, addr);

216
	if (kvm_is_error_gpa(vcpu->kvm, addr))
T
Thomas Huth 已提交
217 218 219 220 221
		return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
	/*
	 * We don't expect errors on modern systems, and do not care
	 * about storage keys (yet), so let's just clear the page.
	 */
222
	if (kvm_clear_guest(vcpu->kvm, addr, PAGE_SIZE))
T
Thomas Huth 已提交
223 224 225 226 227 228
		return -EFAULT;
	kvm_s390_set_psw_cc(vcpu, 0);
	vcpu->run->s.regs.gprs[0] = 0;
	return 0;
}

229
static int handle_tpi(struct kvm_vcpu *vcpu)
230
{
231
	struct kvm_s390_interrupt_info *inti;
H
Heiko Carstens 已提交
232 233
	unsigned long len;
	u32 tpi_data[3];
234
	int rc;
235
	u64 addr;
236
	ar_t ar;
237

238
	addr = kvm_s390_get_base_disp_s(vcpu, &ar);
239 240
	if (addr & 3)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
241

242
	inti = kvm_s390_get_io_int(vcpu->kvm, vcpu->arch.sie_block->gcr[6], 0);
243 244 245 246 247
	if (!inti) {
		kvm_s390_set_psw_cc(vcpu, 0);
		return 0;
	}

H
Heiko Carstens 已提交
248 249 250
	tpi_data[0] = inti->io.subchannel_id << 16 | inti->io.subchannel_nr;
	tpi_data[1] = inti->io.io_int_parm;
	tpi_data[2] = inti->io.io_int_word;
251 252 253 254 255
	if (addr) {
		/*
		 * Store the two-word I/O interruption code into the
		 * provided area.
		 */
H
Heiko Carstens 已提交
256
		len = sizeof(tpi_data) - 4;
257
		rc = write_guest(vcpu, addr, ar, &tpi_data, len);
258 259 260 261
		if (rc) {
			rc = kvm_s390_inject_prog_cond(vcpu, rc);
			goto reinject_interrupt;
		}
262 263 264 265 266
	} else {
		/*
		 * Store the three-word I/O interruption code into
		 * the appropriate lowcore area.
		 */
H
Heiko Carstens 已提交
267
		len = sizeof(tpi_data);
268 269
		if (write_guest_lc(vcpu, __LC_SUBCHANNEL_ID, &tpi_data, len)) {
			/* failed writes to the low core are not recoverable */
H
Heiko Carstens 已提交
270
			rc = -EFAULT;
271 272
			goto reinject_interrupt;
		}
273
	}
274 275 276 277 278 279

	/* irq was successfully handed to the guest */
	kfree(inti);
	kvm_s390_set_psw_cc(vcpu, 1);
	return 0;
reinject_interrupt:
280 281 282 283 284
	/*
	 * If we encounter a problem storing the interruption code, the
	 * instruction is suppressed from the guest's view: reinject the
	 * interrupt.
	 */
285 286 287 288
	if (kvm_s390_reinject_io_int(vcpu->kvm, inti)) {
		kfree(inti);
		rc = -EFAULT;
	}
289
	/* don't set the cc, a pgm irq was injected or we drop to user space */
H
Heiko Carstens 已提交
290
	return rc ? -EFAULT : 0;
291 292
}

293 294
static int handle_tsch(struct kvm_vcpu *vcpu)
{
295 296
	struct kvm_s390_interrupt_info *inti = NULL;
	const u64 isc_mask = 0xffUL << 24; /* all iscs set */
297

298 299 300 301
	/* a valid schid has at least one bit set */
	if (vcpu->run->s.regs.gprs[1])
		inti = kvm_s390_get_io_int(vcpu->kvm, isc_mask,
					   vcpu->run->s.regs.gprs[1]);
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327

	/*
	 * Prepare exit to userspace.
	 * We indicate whether we dequeued a pending I/O interrupt
	 * so that userspace can re-inject it if the instruction gets
	 * a program check. While this may re-order the pending I/O
	 * interrupts, this is no problem since the priority is kept
	 * intact.
	 */
	vcpu->run->exit_reason = KVM_EXIT_S390_TSCH;
	vcpu->run->s390_tsch.dequeued = !!inti;
	if (inti) {
		vcpu->run->s390_tsch.subchannel_id = inti->io.subchannel_id;
		vcpu->run->s390_tsch.subchannel_nr = inti->io.subchannel_nr;
		vcpu->run->s390_tsch.io_int_parm = inti->io.io_int_parm;
		vcpu->run->s390_tsch.io_int_word = inti->io.io_int_word;
	}
	vcpu->run->s390_tsch.ipb = vcpu->arch.sie_block->ipb;
	kfree(inti);
	return -EREMOTE;
}

static int handle_io_inst(struct kvm_vcpu *vcpu)
{
	VCPU_EVENT(vcpu, 4, "%s", "I/O instruction");

328 329 330
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

331 332 333 334 335 336 337 338 339 340 341 342 343
	if (vcpu->kvm->arch.css_support) {
		/*
		 * Most I/O instructions will be handled by userspace.
		 * Exceptions are tpi and the interrupt portion of tsch.
		 */
		if (vcpu->arch.sie_block->ipa == 0xb236)
			return handle_tpi(vcpu);
		if (vcpu->arch.sie_block->ipa == 0xb235)
			return handle_tsch(vcpu);
		/* Handle in userspace. */
		return -EOPNOTSUPP;
	} else {
		/*
344
		 * Set condition code 3 to stop the guest from issuing channel
345 346
		 * I/O instructions.
		 */
347
		kvm_s390_set_psw_cc(vcpu, 3);
348 349 350 351
		return 0;
	}
}

352 353 354
static int handle_stfl(struct kvm_vcpu *vcpu)
{
	int rc;
355
	unsigned int fac;
356 357

	vcpu->stat.instruction_stfl++;
358 359 360 361

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

362 363 364 365
	/*
	 * We need to shift the lower 32 facility bits (bit 0-31) from a u64
	 * into a u32 memory representation. They will remain bits 0-31.
	 */
366
	fac = *vcpu->kvm->arch.model.fac->list >> 32;
367
	rc = write_guest_lc(vcpu, offsetof(struct _lowcore, stfl_fac_list),
368
			    &fac, sizeof(fac));
369
	if (rc)
370
		return rc;
371 372
	VCPU_EVENT(vcpu, 5, "store facility list value %x", fac);
	trace_kvm_s390_handle_stfl(vcpu, fac);
373 374 375
	return 0;
}

376 377
#define PSW_MASK_ADDR_MODE (PSW_MASK_EA | PSW_MASK_BA)
#define PSW_MASK_UNASSIGNED 0xb80800fe7fffffffUL
378
#define PSW_ADDR_24 0x0000000000ffffffUL
379 380
#define PSW_ADDR_31 0x000000007fffffffUL

T
Thomas Huth 已提交
381 382
int is_valid_psw(psw_t *psw)
{
383 384 385 386 387 388 389 390 391 392
	if (psw->mask & PSW_MASK_UNASSIGNED)
		return 0;
	if ((psw->mask & PSW_MASK_ADDR_MODE) == PSW_MASK_BA) {
		if (psw->addr & ~PSW_ADDR_31)
			return 0;
	}
	if (!(psw->mask & PSW_MASK_ADDR_MODE) && (psw->addr & ~PSW_ADDR_24))
		return 0;
	if ((psw->mask & PSW_MASK_ADDR_MODE) ==  PSW_MASK_EA)
		return 0;
T
Thomas Huth 已提交
393 394
	if (psw->addr & 1)
		return 0;
395 396 397
	return 1;
}

398 399
int kvm_s390_handle_lpsw(struct kvm_vcpu *vcpu)
{
400
	psw_t *gpsw = &vcpu->arch.sie_block->gpsw;
401
	psw_compat_t new_psw;
402
	u64 addr;
403
	int rc;
404
	ar_t ar;
405

406
	if (gpsw->mask & PSW_MASK_PSTATE)
407 408
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

409
	addr = kvm_s390_get_base_disp_s(vcpu, &ar);
410 411
	if (addr & 7)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
412

413
	rc = read_guest(vcpu, addr, ar, &new_psw, sizeof(new_psw));
414 415
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
416 417
	if (!(new_psw.mask & PSW32_MASK_BASE))
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
418 419 420 421
	gpsw->mask = (new_psw.mask & ~PSW32_MASK_BASE) << 32;
	gpsw->mask |= new_psw.addr & PSW32_ADDR_AMODE;
	gpsw->addr = new_psw.addr & ~PSW32_ADDR_AMODE;
	if (!is_valid_psw(gpsw))
422
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
423 424 425 426 427 428
	return 0;
}

static int handle_lpswe(struct kvm_vcpu *vcpu)
{
	psw_t new_psw;
429
	u64 addr;
430
	int rc;
431
	ar_t ar;
432

433 434 435
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

436
	addr = kvm_s390_get_base_disp_s(vcpu, &ar);
437 438
	if (addr & 7)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
439
	rc = read_guest(vcpu, addr, ar, &new_psw, sizeof(new_psw));
440 441
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
442 443
	vcpu->arch.sie_block->gpsw = new_psw;
	if (!is_valid_psw(&vcpu->arch.sie_block->gpsw))
444
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
445 446 447
	return 0;
}

448 449
static int handle_stidp(struct kvm_vcpu *vcpu)
{
450
	u64 stidp_data = vcpu->arch.stidp_data;
451
	u64 operand2;
452
	int rc;
453
	ar_t ar;
454 455

	vcpu->stat.instruction_stidp++;
456

457 458 459
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

460
	operand2 = kvm_s390_get_base_disp_s(vcpu, &ar);
461

462 463
	if (operand2 & 7)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
464

465
	rc = write_guest(vcpu, operand2, ar, &stidp_data, sizeof(stidp_data));
466 467
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
468 469 470 471 472 473 474 475 476 477

	VCPU_EVENT(vcpu, 5, "%s", "store cpu id");
	return 0;
}

static void handle_stsi_3_2_2(struct kvm_vcpu *vcpu, struct sysinfo_3_2_2 *mem)
{
	int cpus = 0;
	int n;

478
	cpus = atomic_read(&vcpu->kvm->online_vcpus);
479 480

	/* deal with other level 3 hypervisors */
481
	if (stsi(mem, 3, 2, 2))
482 483 484 485 486 487
		mem->count = 0;
	if (mem->count < 8)
		mem->count++;
	for (n = mem->count - 1; n > 0 ; n--)
		memcpy(&mem->vm[n], &mem->vm[n - 1], sizeof(mem->vm[0]));

488
	memset(&mem->vm[0], 0, sizeof(mem->vm[0]));
489 490 491 492 493 494 495 496 497 498 499
	mem->vm[0].cpus_total = cpus;
	mem->vm[0].cpus_configured = cpus;
	mem->vm[0].cpus_standby = 0;
	mem->vm[0].cpus_reserved = 0;
	mem->vm[0].caf = 1000;
	memcpy(mem->vm[0].name, "KVMguest", 8);
	ASCEBC(mem->vm[0].name, 8);
	memcpy(mem->vm[0].cpi, "KVM/Linux       ", 16);
	ASCEBC(mem->vm[0].cpi, 16);
}

500 501 502 503 504 505 506 507 508 509 510
static void insert_stsi_usr_data(struct kvm_vcpu *vcpu, u64 addr, ar_t ar,
				 u8 fc, u8 sel1, u16 sel2)
{
	vcpu->run->exit_reason = KVM_EXIT_S390_STSI;
	vcpu->run->s390_stsi.addr = addr;
	vcpu->run->s390_stsi.ar = ar;
	vcpu->run->s390_stsi.fc = fc;
	vcpu->run->s390_stsi.sel1 = sel1;
	vcpu->run->s390_stsi.sel2 = sel2;
}

511 512
static int handle_stsi(struct kvm_vcpu *vcpu)
{
513 514 515
	int fc = (vcpu->run->s.regs.gprs[0] & 0xf0000000) >> 28;
	int sel1 = vcpu->run->s.regs.gprs[0] & 0xff;
	int sel2 = vcpu->run->s.regs.gprs[1] & 0xffff;
516
	unsigned long mem = 0;
517
	u64 operand2;
518
	int rc = 0;
519
	ar_t ar;
520 521 522 523

	vcpu->stat.instruction_stsi++;
	VCPU_EVENT(vcpu, 4, "stsi: fc: %x sel1: %x sel2: %x", fc, sel1, sel2);

524 525 526
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

527
	if (fc > 3) {
528
		kvm_s390_set_psw_cc(vcpu, 3);
529 530
		return 0;
	}
531

532 533
	if (vcpu->run->s.regs.gprs[0] & 0x0fffff00
	    || vcpu->run->s.regs.gprs[1] & 0xffff0000)
534 535
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

536
	if (fc == 0) {
537
		vcpu->run->s.regs.gprs[0] = 3 << 28;
538
		kvm_s390_set_psw_cc(vcpu, 0);
539
		return 0;
540 541
	}

542
	operand2 = kvm_s390_get_base_disp_s(vcpu, &ar);
543 544 545 546 547

	if (operand2 & 0xfff)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	switch (fc) {
548 549 550 551
	case 1: /* same handling for 1 and 2 */
	case 2:
		mem = get_zeroed_page(GFP_KERNEL);
		if (!mem)
552
			goto out_no_data;
553
		if (stsi((void *) mem, fc, sel1, sel2))
554
			goto out_no_data;
555 556 557
		break;
	case 3:
		if (sel1 != 2 || sel2 != 2)
558
			goto out_no_data;
559 560
		mem = get_zeroed_page(GFP_KERNEL);
		if (!mem)
561
			goto out_no_data;
562 563 564 565
		handle_stsi_3_2_2(vcpu, (void *) mem);
		break;
	}

566
	rc = write_guest(vcpu, operand2, ar, (void *)mem, PAGE_SIZE);
567 568 569
	if (rc) {
		rc = kvm_s390_inject_prog_cond(vcpu, rc);
		goto out;
570
	}
571 572 573 574
	if (vcpu->kvm->arch.user_stsi) {
		insert_stsi_usr_data(vcpu, operand2, ar, fc, sel1, sel2);
		rc = -EREMOTE;
	}
575
	trace_kvm_s390_handle_stsi(vcpu, fc, sel1, sel2, operand2);
576
	free_page(mem);
577
	kvm_s390_set_psw_cc(vcpu, 0);
578
	vcpu->run->s.regs.gprs[0] = 0;
579
	return rc;
580
out_no_data:
581
	kvm_s390_set_psw_cc(vcpu, 3);
582
out:
583
	free_page(mem);
584
	return rc;
585 586
}

587
static const intercept_handler_t b2_handlers[256] = {
588
	[0x02] = handle_stidp,
589
	[0x04] = handle_set_clock,
590 591 592
	[0x10] = handle_set_prefix,
	[0x11] = handle_store_prefix,
	[0x12] = handle_store_cpu_address,
593
	[0x21] = handle_ipte_interlock,
594 595 596
	[0x29] = handle_skey,
	[0x2a] = handle_skey,
	[0x2b] = handle_skey,
T
Thomas Huth 已提交
597
	[0x2c] = handle_test_block,
598 599 600 601 602 603 604 605 606 607 608 609 610
	[0x30] = handle_io_inst,
	[0x31] = handle_io_inst,
	[0x32] = handle_io_inst,
	[0x33] = handle_io_inst,
	[0x34] = handle_io_inst,
	[0x35] = handle_io_inst,
	[0x36] = handle_io_inst,
	[0x37] = handle_io_inst,
	[0x38] = handle_io_inst,
	[0x39] = handle_io_inst,
	[0x3a] = handle_io_inst,
	[0x3b] = handle_io_inst,
	[0x3c] = handle_io_inst,
611
	[0x50] = handle_ipte_interlock,
612 613 614
	[0x5f] = handle_io_inst,
	[0x74] = handle_io_inst,
	[0x76] = handle_io_inst,
615 616
	[0x7d] = handle_stsi,
	[0xb1] = handle_stfl,
617
	[0xb2] = handle_lpswe,
618 619
};

620
int kvm_s390_handle_b2(struct kvm_vcpu *vcpu)
621 622 623
{
	intercept_handler_t handler;

624
	/*
625 626 627 628
	 * A lot of B2 instructions are priviledged. Here we check for
	 * the privileged ones, that we can handle in the kernel.
	 * Anything else goes to userspace.
	 */
629
	handler = b2_handlers[vcpu->arch.sie_block->ipa & 0x00ff];
630 631 632
	if (handler)
		return handler(vcpu);

633
	return -EOPNOTSUPP;
634
}
635

636 637 638 639
static int handle_epsw(struct kvm_vcpu *vcpu)
{
	int reg1, reg2;

640
	kvm_s390_get_regs_rre(vcpu, &reg1, &reg2);
641 642

	/* This basically extracts the mask half of the psw. */
643
	vcpu->run->s.regs.gprs[reg1] &= 0xffffffff00000000UL;
644 645
	vcpu->run->s.regs.gprs[reg1] |= vcpu->arch.sie_block->gpsw.mask >> 32;
	if (reg2) {
646
		vcpu->run->s.regs.gprs[reg2] &= 0xffffffff00000000UL;
647
		vcpu->run->s.regs.gprs[reg2] |=
648
			vcpu->arch.sie_block->gpsw.mask & 0x00000000ffffffffUL;
649 650 651 652
	}
	return 0;
}

653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675
#define PFMF_RESERVED   0xfffc0101UL
#define PFMF_SK         0x00020000UL
#define PFMF_CF         0x00010000UL
#define PFMF_UI         0x00008000UL
#define PFMF_FSC        0x00007000UL
#define PFMF_NQ         0x00000800UL
#define PFMF_MR         0x00000400UL
#define PFMF_MC         0x00000200UL
#define PFMF_KEY        0x000000feUL

static int handle_pfmf(struct kvm_vcpu *vcpu)
{
	int reg1, reg2;
	unsigned long start, end;

	vcpu->stat.instruction_pfmf++;

	kvm_s390_get_regs_rre(vcpu, &reg1, &reg2);

	if (!MACHINE_HAS_PFMF)
		return kvm_s390_inject_program_int(vcpu, PGM_OPERATION);

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
676
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);
677 678 679 680 681

	if (vcpu->run->s.regs.gprs[reg1] & PFMF_RESERVED)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	/* Only provide non-quiescing support if the host supports it */
682
	if (vcpu->run->s.regs.gprs[reg1] & PFMF_NQ && !test_facility(14))
683 684 685 686 687 688 689
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	/* No support for conditional-SSKE */
	if (vcpu->run->s.regs.gprs[reg1] & (PFMF_MR | PFMF_MC))
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	start = vcpu->run->s.regs.gprs[reg2] & PAGE_MASK;
690
	start = kvm_s390_logical_to_effective(vcpu, start);
T
Thomas Huth 已提交
691

692 693 694 695 696 697 698 699
	switch (vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) {
	case 0x00000000:
		end = (start + (1UL << 12)) & ~((1UL << 12) - 1);
		break;
	case 0x00001000:
		end = (start + (1UL << 20)) & ~((1UL << 20) - 1);
		break;
	case 0x00002000:
700 701 702 703 704
		/* only support 2G frame size if EDAT2 is available and we are
		   not in 24-bit addressing mode */
		if (!test_kvm_facility(vcpu->kvm, 78) ||
		    psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_AMODE_24BIT)
			return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
705
		end = (start + (1UL << 31)) & ~((1UL << 31) - 1);
706
		break;
707 708 709
	default:
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
	}
710 711

	if (vcpu->run->s.regs.gprs[reg1] & PFMF_CF) {
712
		if (kvm_s390_check_low_addr_prot_real(vcpu, start))
713 714 715
			return kvm_s390_inject_prog_irq(vcpu, &vcpu->arch.pgm);
	}

716
	while (start < end) {
T
Thomas Huth 已提交
717 718 719 720 721 722 723 724 725
		unsigned long useraddr, abs_addr;

		/* Translate guest address to host address */
		if ((vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) == 0)
			abs_addr = kvm_s390_real_to_abs(vcpu, start);
		else
			abs_addr = start;
		useraddr = gfn_to_hva(vcpu->kvm, gpa_to_gfn(abs_addr));
		if (kvm_is_error_hva(useraddr))
726 727 728 729 730 731 732 733
			return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);

		if (vcpu->run->s.regs.gprs[reg1] & PFMF_CF) {
			if (clear_user((void __user *)useraddr, PAGE_SIZE))
				return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
		}

		if (vcpu->run->s.regs.gprs[reg1] & PFMF_SK) {
734 735 736 737
			int rc = __skey_check_enable(vcpu);

			if (rc)
				return rc;
738 739 740 741 742 743 744 745 746 747 748 749 750
			if (set_guest_storage_key(current->mm, useraddr,
					vcpu->run->s.regs.gprs[reg1] & PFMF_KEY,
					vcpu->run->s.regs.gprs[reg1] & PFMF_NQ))
				return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
		}

		start += PAGE_SIZE;
	}
	if (vcpu->run->s.regs.gprs[reg1] & PFMF_FSC)
		vcpu->run->s.regs.gprs[reg2] = end;
	return 0;
}

751 752 753 754 755 756 757 758 759 760 761
static int handle_essa(struct kvm_vcpu *vcpu)
{
	/* entries expected to be 1FF */
	int entries = (vcpu->arch.sie_block->cbrlo & ~PAGE_MASK) >> 3;
	unsigned long *cbrlo, cbrle;
	struct gmap *gmap;
	int i;

	VCPU_EVENT(vcpu, 5, "cmma release %d pages", entries);
	gmap = vcpu->arch.gmap;
	vcpu->stat.instruction_essa++;
762
	if (!vcpu->kvm->arch.use_cmma)
763 764 765 766 767 768 769 770 771
		return kvm_s390_inject_program_int(vcpu, PGM_OPERATION);

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

	if (((vcpu->arch.sie_block->ipb & 0xf0000000) >> 28) > 6)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	/* Rewind PSW to repeat the ESSA instruction */
772
	kvm_s390_rewind_psw(vcpu, 4);
773 774 775 776 777 778 779 780 781
	vcpu->arch.sie_block->cbrlo &= PAGE_MASK;	/* reset nceo */
	cbrlo = phys_to_virt(vcpu->arch.sie_block->cbrlo);
	down_read(&gmap->mm->mmap_sem);
	for (i = 0; i < entries; ++i) {
		cbrle = cbrlo[i];
		if (unlikely(cbrle & ~PAGE_MASK || cbrle < 2 * PAGE_SIZE))
			/* invalid entry */
			break;
		/* try to free backing */
782
		__gmap_zap(gmap, cbrle);
783 784 785 786 787 788 789
	}
	up_read(&gmap->mm->mmap_sem);
	if (i < entries)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
	return 0;
}

790
static const intercept_handler_t b9_handlers[256] = {
791
	[0x8a] = handle_ipte_interlock,
792
	[0x8d] = handle_epsw,
793 794
	[0x8e] = handle_ipte_interlock,
	[0x8f] = handle_ipte_interlock,
795
	[0xab] = handle_essa,
796
	[0xaf] = handle_pfmf,
797 798 799 800 801 802 803 804
};

int kvm_s390_handle_b9(struct kvm_vcpu *vcpu)
{
	intercept_handler_t handler;

	/* This is handled just as for the B2 instructions. */
	handler = b9_handlers[vcpu->arch.sie_block->ipa & 0x00ff];
805 806 807
	if (handler)
		return handler(vcpu);

808 809 810
	return -EOPNOTSUPP;
}

811 812 813 814
int kvm_s390_handle_lctl(struct kvm_vcpu *vcpu)
{
	int reg1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4;
	int reg3 = vcpu->arch.sie_block->ipa & 0x000f;
815 816
	int reg, rc, nr_regs;
	u32 ctl_array[16];
817
	u64 ga;
818
	ar_t ar;
819 820 821 822 823 824

	vcpu->stat.instruction_lctl++;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

825
	ga = kvm_s390_get_base_disp_rs(vcpu, &ar);
826

827
	if (ga & 3)
828 829
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

830 831
	VCPU_EVENT(vcpu, 5, "lctl r1:%x, r3:%x, addr:%llx", reg1, reg3, ga);
	trace_kvm_s390_handle_lctl(vcpu, 0, reg1, reg3, ga);
832

833
	nr_regs = ((reg3 - reg1) & 0xf) + 1;
834
	rc = read_guest(vcpu, ga, ar, ctl_array, nr_regs * sizeof(u32));
835 836
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
837
	reg = reg1;
838
	nr_regs = 0;
839 840
	do {
		vcpu->arch.sie_block->gcr[reg] &= 0xffffffff00000000ul;
841
		vcpu->arch.sie_block->gcr[reg] |= ctl_array[nr_regs++];
842 843 844 845
		if (reg == reg3)
			break;
		reg = (reg + 1) % 16;
	} while (1);
846
	kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu);
847 848 849
	return 0;
}

850 851 852 853
int kvm_s390_handle_stctl(struct kvm_vcpu *vcpu)
{
	int reg1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4;
	int reg3 = vcpu->arch.sie_block->ipa & 0x000f;
854 855
	int reg, rc, nr_regs;
	u32 ctl_array[16];
856
	u64 ga;
857
	ar_t ar;
858 859 860 861 862 863

	vcpu->stat.instruction_stctl++;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

864
	ga = kvm_s390_get_base_disp_rs(vcpu, &ar);
865 866 867 868 869 870 871 872

	if (ga & 3)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	VCPU_EVENT(vcpu, 5, "stctl r1:%x, r3:%x, addr:%llx", reg1, reg3, ga);
	trace_kvm_s390_handle_stctl(vcpu, 0, reg1, reg3, ga);

	reg = reg1;
873
	nr_regs = 0;
874
	do {
875
		ctl_array[nr_regs++] = vcpu->arch.sie_block->gcr[reg];
876 877 878 879
		if (reg == reg3)
			break;
		reg = (reg + 1) % 16;
	} while (1);
880
	rc = write_guest(vcpu, ga, ar, ctl_array, nr_regs * sizeof(u32));
881
	return rc ? kvm_s390_inject_prog_cond(vcpu, rc) : 0;
882 883
}

884 885 886 887
static int handle_lctlg(struct kvm_vcpu *vcpu)
{
	int reg1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4;
	int reg3 = vcpu->arch.sie_block->ipa & 0x000f;
888 889 890
	int reg, rc, nr_regs;
	u64 ctl_array[16];
	u64 ga;
891
	ar_t ar;
892 893 894 895 896 897

	vcpu->stat.instruction_lctlg++;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

898
	ga = kvm_s390_get_base_disp_rsy(vcpu, &ar);
899

900
	if (ga & 7)
901 902
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

903 904
	VCPU_EVENT(vcpu, 5, "lctlg r1:%x, r3:%x, addr:%llx", reg1, reg3, ga);
	trace_kvm_s390_handle_lctl(vcpu, 1, reg1, reg3, ga);
905

906
	nr_regs = ((reg3 - reg1) & 0xf) + 1;
907
	rc = read_guest(vcpu, ga, ar, ctl_array, nr_regs * sizeof(u64));
908 909 910 911
	if (rc)
		return kvm_s390_inject_prog_cond(vcpu, rc);
	reg = reg1;
	nr_regs = 0;
912
	do {
913
		vcpu->arch.sie_block->gcr[reg] = ctl_array[nr_regs++];
914 915 916 917
		if (reg == reg3)
			break;
		reg = (reg + 1) % 16;
	} while (1);
918
	kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu);
919 920 921
	return 0;
}

922 923 924 925
static int handle_stctg(struct kvm_vcpu *vcpu)
{
	int reg1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4;
	int reg3 = vcpu->arch.sie_block->ipa & 0x000f;
926 927 928
	int reg, rc, nr_regs;
	u64 ctl_array[16];
	u64 ga;
929
	ar_t ar;
930 931 932 933 934 935

	vcpu->stat.instruction_stctg++;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

936
	ga = kvm_s390_get_base_disp_rsy(vcpu, &ar);
937 938 939 940 941 942 943

	if (ga & 7)
		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);

	VCPU_EVENT(vcpu, 5, "stctg r1:%x, r3:%x, addr:%llx", reg1, reg3, ga);
	trace_kvm_s390_handle_stctl(vcpu, 1, reg1, reg3, ga);

944 945
	reg = reg1;
	nr_regs = 0;
946
	do {
947
		ctl_array[nr_regs++] = vcpu->arch.sie_block->gcr[reg];
948 949 950 951
		if (reg == reg3)
			break;
		reg = (reg + 1) % 16;
	} while (1);
952
	rc = write_guest(vcpu, ga, ar, ctl_array, nr_regs * sizeof(u64));
953
	return rc ? kvm_s390_inject_prog_cond(vcpu, rc) : 0;
954 955
}

956
static const intercept_handler_t eb_handlers[256] = {
957
	[0x2f] = handle_lctlg,
958
	[0x25] = handle_stctg,
959 960
};

961
int kvm_s390_handle_eb(struct kvm_vcpu *vcpu)
962 963 964 965 966 967 968 969 970
{
	intercept_handler_t handler;

	handler = eb_handlers[vcpu->arch.sie_block->ipb & 0xff];
	if (handler)
		return handler(vcpu);
	return -EOPNOTSUPP;
}

971 972
static int handle_tprot(struct kvm_vcpu *vcpu)
{
973
	u64 address1, address2;
974 975 976
	unsigned long hva, gpa;
	int ret = 0, cc = 0;
	bool writable;
977
	ar_t ar;
978 979 980

	vcpu->stat.instruction_tprot++;

981 982 983
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);

984
	kvm_s390_get_base_disp_sse(vcpu, &address1, &address2, &ar, NULL);
985

986 987 988 989 990 991
	/* we only handle the Linux memory detection case:
	 * access key == 0
	 * everything else goes to userspace. */
	if (address2 & 0xf0)
		return -EOPNOTSUPP;
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_DAT)
992
		ipte_lock(vcpu);
993
	ret = guest_translate_address(vcpu, address1, ar, &gpa, 1);
994 995 996
	if (ret == PGM_PROTECTION) {
		/* Write protected? Try again with read-only... */
		cc = 1;
997
		ret = guest_translate_address(vcpu, address1, ar, &gpa, 0);
998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008
	}
	if (ret) {
		if (ret == PGM_ADDRESSING || ret == PGM_TRANSLATION_SPEC) {
			ret = kvm_s390_inject_program_int(vcpu, ret);
		} else if (ret > 0) {
			/* Translation not available */
			kvm_s390_set_psw_cc(vcpu, 3);
			ret = 0;
		}
		goto out_unlock;
	}
1009

1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022
	hva = gfn_to_hva_prot(vcpu->kvm, gpa_to_gfn(gpa), &writable);
	if (kvm_is_error_hva(hva)) {
		ret = kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
	} else {
		if (!writable)
			cc = 1;		/* Write not permitted ==> read-only */
		kvm_s390_set_psw_cc(vcpu, cc);
		/* Note: CC2 only occurs for storage keys (not supported yet) */
	}
out_unlock:
	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_DAT)
		ipte_unlock(vcpu);
	return ret;
1023 1024 1025 1026 1027 1028 1029 1030 1031 1032
}

int kvm_s390_handle_e5(struct kvm_vcpu *vcpu)
{
	/* For e5xx... instructions we only handle TPROT */
	if ((vcpu->arch.sie_block->ipa & 0x00ff) == 0x01)
		return handle_tprot(vcpu);
	return -EOPNOTSUPP;
}

1033 1034 1035 1036 1037
static int handle_sckpf(struct kvm_vcpu *vcpu)
{
	u32 value;

	if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
1038
		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);
1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049

	if (vcpu->run->s.regs.gprs[0] & 0x00000000ffff0000)
		return kvm_s390_inject_program_int(vcpu,
						   PGM_SPECIFICATION);

	value = vcpu->run->s.regs.gprs[0] & 0x000000000000ffff;
	vcpu->arch.sie_block->todpr = value;

	return 0;
}

1050
static const intercept_handler_t x01_handlers[256] = {
1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
	[0x07] = handle_sckpf,
};

int kvm_s390_handle_01(struct kvm_vcpu *vcpu)
{
	intercept_handler_t handler;

	handler = x01_handlers[vcpu->arch.sie_block->ipa & 0x00ff];
	if (handler)
		return handler(vcpu);
	return -EOPNOTSUPP;
}