book3s_hv.c 115.7 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
 * Copyright 2011 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com>
 * Copyright (C) 2009. SUSE Linux Products GmbH. All rights reserved.
 *
 * Authors:
 *    Paul Mackerras <paulus@au1.ibm.com>
 *    Alexander Graf <agraf@suse.de>
 *    Kevin Wolf <mail@kevin-wolf.de>
 *
 * Description: KVM functions specific to running on Book 3S
 * processors in hypervisor mode (specifically POWER7 and later).
 *
 * This file is derived from arch/powerpc/kvm/book3s.c,
 * by Alexander Graf <agraf@suse.de>.
 *
 * 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.
 */

#include <linux/kvm_host.h>
22
#include <linux/kernel.h>
23 24 25
#include <linux/err.h>
#include <linux/slab.h>
#include <linux/preempt.h>
26
#include <linux/sched/signal.h>
27
#include <linux/sched/stat.h>
28
#include <linux/delay.h>
29
#include <linux/export.h>
30 31
#include <linux/fs.h>
#include <linux/anon_inodes.h>
32
#include <linux/cpu.h>
33
#include <linux/cpumask.h>
34 35
#include <linux/spinlock.h>
#include <linux/page-flags.h>
36
#include <linux/srcu.h>
37
#include <linux/miscdevice.h>
38
#include <linux/debugfs.h>
39 40 41 42 43 44 45 46 47
#include <linux/gfp.h>
#include <linux/vmalloc.h>
#include <linux/highmem.h>
#include <linux/hugetlb.h>
#include <linux/kvm_irqfd.h>
#include <linux/irqbypass.h>
#include <linux/module.h>
#include <linux/compiler.h>
#include <linux/of.h>
48 49

#include <asm/reg.h>
50
#include <asm/ppc-opcode.h>
51
#include <asm/asm-prototypes.h>
52
#include <asm/debug.h>
53
#include <asm/disassemble.h>
54 55 56
#include <asm/cputable.h>
#include <asm/cacheflush.h>
#include <asm/tlbflush.h>
57
#include <linux/uaccess.h>
58 59 60 61 62 63
#include <asm/io.h>
#include <asm/kvm_ppc.h>
#include <asm/kvm_book3s.h>
#include <asm/mmu_context.h>
#include <asm/lppaca.h>
#include <asm/processor.h>
64
#include <asm/cputhreads.h>
65
#include <asm/page.h>
66
#include <asm/hvcall.h>
67
#include <asm/switch_to.h>
68
#include <asm/smp.h>
69
#include <asm/dbell.h>
70
#include <asm/hmi.h>
71
#include <asm/pnv-pci.h>
72
#include <asm/mmu.h>
73 74
#include <asm/opal.h>
#include <asm/xics.h>
75
#include <asm/xive.h>
76

77 78
#include "book3s.h"

79 80 81
#define CREATE_TRACE_POINTS
#include "trace_hv.h"

82 83 84 85
/* #define EXIT_DEBUG */
/* #define EXIT_DEBUG_SIMPLE */
/* #define EXIT_DEBUG_INT */

86 87
/* Used to indicate that a guest page fault needs to be handled */
#define RESUME_PAGE_FAULT	(RESUME_GUEST | RESUME_FLAG_ARCH1)
88 89
/* Used to indicate that a guest passthrough interrupt needs to be handled */
#define RESUME_PASSTHROUGH	(RESUME_GUEST | RESUME_FLAG_ARCH2)
90

91 92 93
/* Used as a "null" value for timebase values */
#define TB_NIL	(~(u64)0)

94 95
static DECLARE_BITMAP(default_enabled_hcalls, MAX_HCALL_OPCODE/4 + 1);

96
static int dynamic_mt_modes = 6;
97
module_param(dynamic_mt_modes, int, 0644);
98
MODULE_PARM_DESC(dynamic_mt_modes, "Set of allowed dynamic micro-threading modes: 0 (= none), 2, 4, or 6 (= 2 or 4)");
99
static int target_smt_mode;
100
module_param(target_smt_mode, int, 0644);
101
MODULE_PARM_DESC(target_smt_mode, "Target threads per core (0 = max)");
102

103 104 105 106
static bool indep_threads_mode = true;
module_param(indep_threads_mode, bool, S_IRUGO | S_IWUSR);
MODULE_PARM_DESC(indep_threads_mode, "Independent-threads mode (only on POWER9)");

107 108 109 110 111 112
#ifdef CONFIG_KVM_XICS
static struct kernel_param_ops module_param_ops = {
	.set = param_set_int,
	.get = param_get_int,
};

113
module_param_cb(kvm_irq_bypass, &module_param_ops, &kvm_irq_bypass, 0644);
114 115
MODULE_PARM_DESC(kvm_irq_bypass, "Bypass passthrough interrupt optimization");

116
module_param_cb(h_ipi_redirect, &module_param_ops, &h_ipi_redirect, 0644);
117 118 119
MODULE_PARM_DESC(h_ipi_redirect, "Redirect H_IPI wakeup to a free host core");
#endif

120 121 122
/* If set, the threads on each CPU core have to be in the same MMU mode */
static bool no_mixing_hpt_and_radix;

123
static void kvmppc_end_cede(struct kvm_vcpu *vcpu);
124
static int kvmppc_hv_setup_htab_rma(struct kvm_vcpu *vcpu);
125

126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
static inline struct kvm_vcpu *next_runnable_thread(struct kvmppc_vcore *vc,
		int *ip)
{
	int i = *ip;
	struct kvm_vcpu *vcpu;

	while (++i < MAX_SMT_THREADS) {
		vcpu = READ_ONCE(vc->runnable_threads[i]);
		if (vcpu) {
			*ip = i;
			return vcpu;
		}
	}
	return NULL;
}

/* Used to traverse the list of runnable threads for a given vcore */
#define for_each_runnable_thread(i, vcpu, vc) \
	for (i = -1; (vcpu = next_runnable_thread(vc, &i)); )

146 147
static bool kvmppc_ipi_thread(int cpu)
{
148 149 150 151 152 153 154 155 156 157
	unsigned long msg = PPC_DBELL_TYPE(PPC_DBELL_SERVER);

	/* On POWER9 we can use msgsnd to IPI any cpu */
	if (cpu_has_feature(CPU_FTR_ARCH_300)) {
		msg |= get_hard_smp_processor_id(cpu);
		smp_mb();
		__asm__ __volatile__ (PPC_MSGSND(%0) : : "r" (msg));
		return true;
	}

158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
	/* On POWER8 for IPIs to threads in the same core, use msgsnd */
	if (cpu_has_feature(CPU_FTR_ARCH_207S)) {
		preempt_disable();
		if (cpu_first_thread_sibling(cpu) ==
		    cpu_first_thread_sibling(smp_processor_id())) {
			msg |= cpu_thread_in_core(cpu);
			smp_mb();
			__asm__ __volatile__ (PPC_MSGSND(%0) : : "r" (msg));
			preempt_enable();
			return true;
		}
		preempt_enable();
	}

#if defined(CONFIG_PPC_ICP_NATIVE) && defined(CONFIG_SMP)
173
	if (cpu >= 0 && cpu < nr_cpu_ids) {
174
		if (paca_ptrs[cpu]->kvm_hstate.xics_phys) {
175 176 177 178
			xics_wake_cpu(cpu);
			return true;
		}
		opal_int_set_mfrr(get_hard_smp_processor_id(cpu), IPI_PRIORITY);
179 180 181 182 183 184 185
		return true;
	}
#endif

	return false;
}

186
static void kvmppc_fast_vcpu_kick_hv(struct kvm_vcpu *vcpu)
187
{
188
	int cpu;
189
	struct swait_queue_head *wqp;
190 191

	wqp = kvm_arch_vcpu_wq(vcpu);
192
	if (swq_has_sleeper(wqp)) {
193
		swake_up(wqp);
194 195 196
		++vcpu->stat.halt_wakeup;
	}

197 198
	cpu = READ_ONCE(vcpu->arch.thread_cpu);
	if (cpu >= 0 && kvmppc_ipi_thread(cpu))
199
		return;
200 201

	/* CPU points to the first thread of the core */
202
	cpu = vcpu->cpu;
203 204
	if (cpu >= 0 && cpu < nr_cpu_ids && cpu_online(cpu))
		smp_send_reschedule(cpu);
205 206
}

207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233
/*
 * We use the vcpu_load/put functions to measure stolen time.
 * Stolen time is counted as time when either the vcpu is able to
 * run as part of a virtual core, but the task running the vcore
 * is preempted or sleeping, or when the vcpu needs something done
 * in the kernel by the task running the vcpu, but that task is
 * preempted or sleeping.  Those two things have to be counted
 * separately, since one of the vcpu tasks will take on the job
 * of running the core, and the other vcpu tasks in the vcore will
 * sleep waiting for it to do that, but that sleep shouldn't count
 * as stolen time.
 *
 * Hence we accumulate stolen time when the vcpu can run as part of
 * a vcore using vc->stolen_tb, and the stolen time when the vcpu
 * needs its task to do other things in the kernel (for example,
 * service a page fault) in busy_stolen.  We don't accumulate
 * stolen time for a vcore when it is inactive, or for a vcpu
 * when it is in state RUNNING or NOTREADY.  NOTREADY is a bit of
 * a misnomer; it means that the vcpu task is not executing in
 * the KVM_VCPU_RUN ioctl, i.e. it is in userspace or elsewhere in
 * the kernel.  We don't have any way of dividing up that time
 * between time that the vcpu is genuinely stopped, time that
 * the task is actively working on behalf of the vcpu, and time
 * that the task is preempted, so we don't count any of it as
 * stolen.
 *
 * Updates to busy_stolen are protected by arch.tbacct_lock;
234 235 236 237
 * updates to vc->stolen_tb are protected by the vcore->stoltb_lock
 * lock.  The stolen times are measured in units of timebase ticks.
 * (Note that the != TB_NIL checks below are purely defensive;
 * they should never fail.)
238 239
 */

240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
static void kvmppc_core_start_stolen(struct kvmppc_vcore *vc)
{
	unsigned long flags;

	spin_lock_irqsave(&vc->stoltb_lock, flags);
	vc->preempt_tb = mftb();
	spin_unlock_irqrestore(&vc->stoltb_lock, flags);
}

static void kvmppc_core_end_stolen(struct kvmppc_vcore *vc)
{
	unsigned long flags;

	spin_lock_irqsave(&vc->stoltb_lock, flags);
	if (vc->preempt_tb != TB_NIL) {
		vc->stolen_tb += mftb() - vc->preempt_tb;
		vc->preempt_tb = TB_NIL;
	}
	spin_unlock_irqrestore(&vc->stoltb_lock, flags);
}

261
static void kvmppc_core_vcpu_load_hv(struct kvm_vcpu *vcpu, int cpu)
262
{
263
	struct kvmppc_vcore *vc = vcpu->arch.vcore;
264
	unsigned long flags;
265

266 267 268 269 270 271
	/*
	 * We can test vc->runner without taking the vcore lock,
	 * because only this task ever sets vc->runner to this
	 * vcpu, and once it is set to this vcpu, only this task
	 * ever sets it to NULL.
	 */
272 273 274
	if (vc->runner == vcpu && vc->vcore_state >= VCORE_SLEEPING)
		kvmppc_core_end_stolen(vc);

275
	spin_lock_irqsave(&vcpu->arch.tbacct_lock, flags);
276 277 278 279 280
	if (vcpu->arch.state == KVMPPC_VCPU_BUSY_IN_HOST &&
	    vcpu->arch.busy_preempt != TB_NIL) {
		vcpu->arch.busy_stolen += mftb() - vcpu->arch.busy_preempt;
		vcpu->arch.busy_preempt = TB_NIL;
	}
281
	spin_unlock_irqrestore(&vcpu->arch.tbacct_lock, flags);
282 283
}

284
static void kvmppc_core_vcpu_put_hv(struct kvm_vcpu *vcpu)
285
{
286
	struct kvmppc_vcore *vc = vcpu->arch.vcore;
287
	unsigned long flags;
288

289 290 291
	if (vc->runner == vcpu && vc->vcore_state >= VCORE_SLEEPING)
		kvmppc_core_start_stolen(vc);

292
	spin_lock_irqsave(&vcpu->arch.tbacct_lock, flags);
293 294
	if (vcpu->arch.state == KVMPPC_VCPU_BUSY_IN_HOST)
		vcpu->arch.busy_preempt = mftb();
295
	spin_unlock_irqrestore(&vcpu->arch.tbacct_lock, flags);
296 297
}

298
static void kvmppc_set_msr_hv(struct kvm_vcpu *vcpu, u64 msr)
299
{
300 301 302 303 304 305
	/*
	 * Check for illegal transactional state bit combination
	 * and if we find it, force the TS field to a safe state.
	 */
	if ((msr & MSR_TS_MASK) == MSR_TS_MASK)
		msr &= ~MSR_TS_MASK;
306
	vcpu->arch.shregs.msr = msr;
307
	kvmppc_end_cede(vcpu);
308 309
}

T
Thomas Huth 已提交
310
static void kvmppc_set_pvr_hv(struct kvm_vcpu *vcpu, u32 pvr)
311 312 313 314
{
	vcpu->arch.pvr = pvr;
}

315 316 317
/* Dummy value used in computing PCR value below */
#define PCR_ARCH_300	(PCR_ARCH_207 << 1)

T
Thomas Huth 已提交
318
static int kvmppc_set_arch_compat(struct kvm_vcpu *vcpu, u32 arch_compat)
319
{
320
	unsigned long host_pcr_bit = 0, guest_pcr_bit = 0;
321 322
	struct kvmppc_vcore *vc = vcpu->arch.vcore;

323 324 325 326 327 328 329 330 331 332 333 334
	/* We can (emulate) our own architecture version and anything older */
	if (cpu_has_feature(CPU_FTR_ARCH_300))
		host_pcr_bit = PCR_ARCH_300;
	else if (cpu_has_feature(CPU_FTR_ARCH_207S))
		host_pcr_bit = PCR_ARCH_207;
	else if (cpu_has_feature(CPU_FTR_ARCH_206))
		host_pcr_bit = PCR_ARCH_206;
	else
		host_pcr_bit = PCR_ARCH_205;

	/* Determine lowest PCR bit needed to run guest in given PVR level */
	guest_pcr_bit = host_pcr_bit;
335 336 337
	if (arch_compat) {
		switch (arch_compat) {
		case PVR_ARCH_205:
338
			guest_pcr_bit = PCR_ARCH_205;
339 340 341
			break;
		case PVR_ARCH_206:
		case PVR_ARCH_206p:
342
			guest_pcr_bit = PCR_ARCH_206;
343 344
			break;
		case PVR_ARCH_207:
345 346 347 348
			guest_pcr_bit = PCR_ARCH_207;
			break;
		case PVR_ARCH_300:
			guest_pcr_bit = PCR_ARCH_300;
349 350 351 352 353 354
			break;
		default:
			return -EINVAL;
		}
	}

355 356 357 358
	/* Check requested PCR bits don't exceed our capabilities */
	if (guest_pcr_bit > host_pcr_bit)
		return -EINVAL;

359 360
	spin_lock(&vc->lock);
	vc->arch_compat = arch_compat;
361 362
	/* Set all PCR bits for which guest_pcr_bit <= bit < host_pcr_bit */
	vc->pcr = host_pcr_bit - guest_pcr_bit;
363 364 365 366 367
	spin_unlock(&vc->lock);

	return 0;
}

T
Thomas Huth 已提交
368
static void kvmppc_dump_regs(struct kvm_vcpu *vcpu)
369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396
{
	int r;

	pr_err("vcpu %p (%d):\n", vcpu, vcpu->vcpu_id);
	pr_err("pc  = %.16lx  msr = %.16llx  trap = %x\n",
	       vcpu->arch.pc, vcpu->arch.shregs.msr, vcpu->arch.trap);
	for (r = 0; r < 16; ++r)
		pr_err("r%2d = %.16lx  r%d = %.16lx\n",
		       r, kvmppc_get_gpr(vcpu, r),
		       r+16, kvmppc_get_gpr(vcpu, r+16));
	pr_err("ctr = %.16lx  lr  = %.16lx\n",
	       vcpu->arch.ctr, vcpu->arch.lr);
	pr_err("srr0 = %.16llx srr1 = %.16llx\n",
	       vcpu->arch.shregs.srr0, vcpu->arch.shregs.srr1);
	pr_err("sprg0 = %.16llx sprg1 = %.16llx\n",
	       vcpu->arch.shregs.sprg0, vcpu->arch.shregs.sprg1);
	pr_err("sprg2 = %.16llx sprg3 = %.16llx\n",
	       vcpu->arch.shregs.sprg2, vcpu->arch.shregs.sprg3);
	pr_err("cr = %.8x  xer = %.16lx  dsisr = %.8x\n",
	       vcpu->arch.cr, vcpu->arch.xer, vcpu->arch.shregs.dsisr);
	pr_err("dar = %.16llx\n", vcpu->arch.shregs.dar);
	pr_err("fault dar = %.16lx dsisr = %.8x\n",
	       vcpu->arch.fault_dar, vcpu->arch.fault_dsisr);
	pr_err("SLB (%d entries):\n", vcpu->arch.slb_max);
	for (r = 0; r < vcpu->arch.slb_max; ++r)
		pr_err("  ESID = %.16llx VSID = %.16llx\n",
		       vcpu->arch.slb[r].orige, vcpu->arch.slb[r].origv);
	pr_err("lpcr = %.16lx sdr1 = %.16lx last_inst = %.8x\n",
397
	       vcpu->arch.vcore->lpcr, vcpu->kvm->arch.sdr1,
398 399 400
	       vcpu->arch.last_inst);
}

T
Thomas Huth 已提交
401
static struct kvm_vcpu *kvmppc_find_vcpu(struct kvm *kvm, int id)
402
{
403
	struct kvm_vcpu *ret;
404 405

	mutex_lock(&kvm->lock);
406
	ret = kvm_get_vcpu_by_id(kvm, id);
407 408 409 410 411 412
	mutex_unlock(&kvm->lock);
	return ret;
}

static void init_vpa(struct kvm_vcpu *vcpu, struct lppaca *vpa)
{
413
	vpa->__old_status |= LPPACA_OLD_SHARED_PROC;
414
	vpa->yield_count = cpu_to_be32(1);
415 416
}

417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
static int set_vpa(struct kvm_vcpu *vcpu, struct kvmppc_vpa *v,
		   unsigned long addr, unsigned long len)
{
	/* check address is cacheline aligned */
	if (addr & (L1_CACHE_BYTES - 1))
		return -EINVAL;
	spin_lock(&vcpu->arch.vpa_update_lock);
	if (v->next_gpa != addr || v->len != len) {
		v->next_gpa = addr;
		v->len = addr ? len : 0;
		v->update_pending = 1;
	}
	spin_unlock(&vcpu->arch.vpa_update_lock);
	return 0;
}

433 434 435 436
/* Length for a per-processor buffer is passed in at offset 4 in the buffer */
struct reg_vpa {
	u32 dummy;
	union {
437 438
		__be16 hword;
		__be32 word;
439 440 441 442 443 444 445 446 447 448
	} length;
};

static int vpa_is_registered(struct kvmppc_vpa *vpap)
{
	if (vpap->update_pending)
		return vpap->next_gpa != 0;
	return vpap->pinned_addr != NULL;
}

449 450 451 452 453
static unsigned long do_h_register_vpa(struct kvm_vcpu *vcpu,
				       unsigned long flags,
				       unsigned long vcpuid, unsigned long vpa)
{
	struct kvm *kvm = vcpu->kvm;
454
	unsigned long len, nb;
455 456
	void *va;
	struct kvm_vcpu *tvcpu;
457 458 459
	int err;
	int subfunc;
	struct kvmppc_vpa *vpap;
460 461 462 463 464

	tvcpu = kvmppc_find_vcpu(kvm, vcpuid);
	if (!tvcpu)
		return H_PARAMETER;

465 466 467 468 469
	subfunc = (flags >> H_VPA_FUNC_SHIFT) & H_VPA_FUNC_MASK;
	if (subfunc == H_VPA_REG_VPA || subfunc == H_VPA_REG_DTL ||
	    subfunc == H_VPA_REG_SLB) {
		/* Registering new area - address must be cache-line aligned */
		if ((vpa & (L1_CACHE_BYTES - 1)) || !vpa)
470
			return H_PARAMETER;
471 472

		/* convert logical addr to kernel addr and read length */
473 474
		va = kvmppc_pin_guest_page(kvm, vpa, &nb);
		if (va == NULL)
475
			return H_PARAMETER;
476
		if (subfunc == H_VPA_REG_VPA)
477
			len = be16_to_cpu(((struct reg_vpa *)va)->length.hword);
478
		else
479
			len = be32_to_cpu(((struct reg_vpa *)va)->length.word);
480
		kvmppc_unpin_guest_page(kvm, va, vpa, false);
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495

		/* Check length */
		if (len > nb || len < sizeof(struct reg_vpa))
			return H_PARAMETER;
	} else {
		vpa = 0;
		len = 0;
	}

	err = H_PARAMETER;
	vpap = NULL;
	spin_lock(&tvcpu->arch.vpa_update_lock);

	switch (subfunc) {
	case H_VPA_REG_VPA:		/* register VPA */
496 497 498 499 500 501
		/*
		 * The size of our lppaca is 1kB because of the way we align
		 * it for the guest to avoid crossing a 4kB boundary. We only
		 * use 640 bytes of the structure though, so we should accept
		 * clients that set a size of 640.
		 */
502 503
		BUILD_BUG_ON(sizeof(struct lppaca) != 640);
		if (len < sizeof(struct lppaca))
504
			break;
505 506 507 508 509 510
		vpap = &tvcpu->arch.vpa;
		err = 0;
		break;

	case H_VPA_REG_DTL:		/* register DTL */
		if (len < sizeof(struct dtl_entry))
511
			break;
512 513 514 515 516
		len -= len % sizeof(struct dtl_entry);

		/* Check that they have previously registered a VPA */
		err = H_RESOURCE;
		if (!vpa_is_registered(&tvcpu->arch.vpa))
517
			break;
518 519 520 521 522 523 524 525 526

		vpap = &tvcpu->arch.dtl;
		err = 0;
		break;

	case H_VPA_REG_SLB:		/* register SLB shadow buffer */
		/* Check that they have previously registered a VPA */
		err = H_RESOURCE;
		if (!vpa_is_registered(&tvcpu->arch.vpa))
527
			break;
528 529 530 531 532 533 534 535 536 537

		vpap = &tvcpu->arch.slb_shadow;
		err = 0;
		break;

	case H_VPA_DEREG_VPA:		/* deregister VPA */
		/* Check they don't still have a DTL or SLB buf registered */
		err = H_RESOURCE;
		if (vpa_is_registered(&tvcpu->arch.dtl) ||
		    vpa_is_registered(&tvcpu->arch.slb_shadow))
538
			break;
539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558

		vpap = &tvcpu->arch.vpa;
		err = 0;
		break;

	case H_VPA_DEREG_DTL:		/* deregister DTL */
		vpap = &tvcpu->arch.dtl;
		err = 0;
		break;

	case H_VPA_DEREG_SLB:		/* deregister SLB shadow buffer */
		vpap = &tvcpu->arch.slb_shadow;
		err = 0;
		break;
	}

	if (vpap) {
		vpap->next_gpa = vpa;
		vpap->len = len;
		vpap->update_pending = 1;
559
	}
560

561 562
	spin_unlock(&tvcpu->arch.vpa_update_lock);

563
	return err;
564 565
}

566
static void kvmppc_update_vpa(struct kvm_vcpu *vcpu, struct kvmppc_vpa *vpap)
567
{
568
	struct kvm *kvm = vcpu->kvm;
569 570
	void *va;
	unsigned long nb;
571
	unsigned long gpa;
572

573 574 575 576 577 578 579 580 581 582 583 584 585 586
	/*
	 * We need to pin the page pointed to by vpap->next_gpa,
	 * but we can't call kvmppc_pin_guest_page under the lock
	 * as it does get_user_pages() and down_read().  So we
	 * have to drop the lock, pin the page, then get the lock
	 * again and check that a new area didn't get registered
	 * in the meantime.
	 */
	for (;;) {
		gpa = vpap->next_gpa;
		spin_unlock(&vcpu->arch.vpa_update_lock);
		va = NULL;
		nb = 0;
		if (gpa)
587
			va = kvmppc_pin_guest_page(kvm, gpa, &nb);
588 589 590 591 592
		spin_lock(&vcpu->arch.vpa_update_lock);
		if (gpa == vpap->next_gpa)
			break;
		/* sigh... unpin that one and try again */
		if (va)
593
			kvmppc_unpin_guest_page(kvm, va, gpa, false);
594 595 596 597 598 599 600 601 602
	}

	vpap->update_pending = 0;
	if (va && nb < vpap->len) {
		/*
		 * If it's now too short, it must be that userspace
		 * has changed the mappings underlying guest memory,
		 * so unregister the region.
		 */
603
		kvmppc_unpin_guest_page(kvm, va, gpa, false);
604
		va = NULL;
605 606
	}
	if (vpap->pinned_addr)
607 608 609
		kvmppc_unpin_guest_page(kvm, vpap->pinned_addr, vpap->gpa,
					vpap->dirty);
	vpap->gpa = gpa;
610
	vpap->pinned_addr = va;
611
	vpap->dirty = false;
612 613 614 615 616 617
	if (va)
		vpap->pinned_end = va + vpap->len;
}

static void kvmppc_update_vpas(struct kvm_vcpu *vcpu)
{
618 619 620 621 622
	if (!(vcpu->arch.vpa.update_pending ||
	      vcpu->arch.slb_shadow.update_pending ||
	      vcpu->arch.dtl.update_pending))
		return;

623 624
	spin_lock(&vcpu->arch.vpa_update_lock);
	if (vcpu->arch.vpa.update_pending) {
625
		kvmppc_update_vpa(vcpu, &vcpu->arch.vpa);
626 627
		if (vcpu->arch.vpa.pinned_addr)
			init_vpa(vcpu, vcpu->arch.vpa.pinned_addr);
628 629
	}
	if (vcpu->arch.dtl.update_pending) {
630
		kvmppc_update_vpa(vcpu, &vcpu->arch.dtl);
631 632 633 634
		vcpu->arch.dtl_ptr = vcpu->arch.dtl.pinned_addr;
		vcpu->arch.dtl_index = 0;
	}
	if (vcpu->arch.slb_shadow.update_pending)
635
		kvmppc_update_vpa(vcpu, &vcpu->arch.slb_shadow);
636 637 638
	spin_unlock(&vcpu->arch.vpa_update_lock);
}

639 640 641 642 643 644 645
/*
 * Return the accumulated stolen time for the vcore up until `now'.
 * The caller should hold the vcore lock.
 */
static u64 vcore_stolen_time(struct kvmppc_vcore *vc, u64 now)
{
	u64 p;
646
	unsigned long flags;
647

648 649
	spin_lock_irqsave(&vc->stoltb_lock, flags);
	p = vc->stolen_tb;
650
	if (vc->vcore_state != VCORE_INACTIVE &&
651 652 653
	    vc->preempt_tb != TB_NIL)
		p += now - vc->preempt_tb;
	spin_unlock_irqrestore(&vc->stoltb_lock, flags);
654 655 656
	return p;
}

657 658 659 660 661
static void kvmppc_create_dtl_entry(struct kvm_vcpu *vcpu,
				    struct kvmppc_vcore *vc)
{
	struct dtl_entry *dt;
	struct lppaca *vpa;
662 663 664
	unsigned long stolen;
	unsigned long core_stolen;
	u64 now;
665
	unsigned long flags;
666 667 668

	dt = vcpu->arch.dtl_ptr;
	vpa = vcpu->arch.vpa.pinned_addr;
669 670 671 672
	now = mftb();
	core_stolen = vcore_stolen_time(vc, now);
	stolen = core_stolen - vcpu->arch.stolen_logged;
	vcpu->arch.stolen_logged = core_stolen;
673
	spin_lock_irqsave(&vcpu->arch.tbacct_lock, flags);
674 675
	stolen += vcpu->arch.busy_stolen;
	vcpu->arch.busy_stolen = 0;
676
	spin_unlock_irqrestore(&vcpu->arch.tbacct_lock, flags);
677 678 679 680
	if (!dt || !vpa)
		return;
	memset(dt, 0, sizeof(struct dtl_entry));
	dt->dispatch_reason = 7;
681 682 683 684 685
	dt->processor_id = cpu_to_be16(vc->pcpu + vcpu->arch.ptid);
	dt->timebase = cpu_to_be64(now + vc->tb_offset);
	dt->enqueue_to_dispatch_time = cpu_to_be32(stolen);
	dt->srr0 = cpu_to_be64(kvmppc_get_pc(vcpu));
	dt->srr1 = cpu_to_be64(vcpu->arch.shregs.msr);
686 687 688 689 690 691
	++dt;
	if (dt == vcpu->arch.dtl.pinned_end)
		dt = vcpu->arch.dtl.pinned_addr;
	vcpu->arch.dtl_ptr = dt;
	/* order writing *dt vs. writing vpa->dtl_idx */
	smp_wmb();
692
	vpa->dtl_idx = cpu_to_be64(++vcpu->arch.dtl_index);
693
	vcpu->arch.dtl.dirty = true;
694 695
}

696 697 698 699 700 701
/* See if there is a doorbell interrupt pending for a vcpu */
static bool kvmppc_doorbell_pending(struct kvm_vcpu *vcpu)
{
	int thr;
	struct kvmppc_vcore *vc;

702 703 704 705 706 707 708 709 710
	if (vcpu->arch.doorbell_request)
		return true;
	/*
	 * Ensure that the read of vcore->dpdes comes after the read
	 * of vcpu->doorbell_request.  This barrier matches the
	 * lwsync in book3s_hv_rmhandlers.S just before the
	 * fast_guest_return label.
	 */
	smp_rmb();
711 712 713 714 715
	vc = vcpu->arch.vcore;
	thr = vcpu->vcpu_id - vc->first_vcpuid;
	return !!(vc->dpdes & (1 << thr));
}

716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745
static bool kvmppc_power8_compatible(struct kvm_vcpu *vcpu)
{
	if (vcpu->arch.vcore->arch_compat >= PVR_ARCH_207)
		return true;
	if ((!vcpu->arch.vcore->arch_compat) &&
	    cpu_has_feature(CPU_FTR_ARCH_207S))
		return true;
	return false;
}

static int kvmppc_h_set_mode(struct kvm_vcpu *vcpu, unsigned long mflags,
			     unsigned long resource, unsigned long value1,
			     unsigned long value2)
{
	switch (resource) {
	case H_SET_MODE_RESOURCE_SET_CIABR:
		if (!kvmppc_power8_compatible(vcpu))
			return H_P2;
		if (value2)
			return H_P4;
		if (mflags)
			return H_UNSUPPORTED_FLAG_START;
		/* Guests can't breakpoint the hypervisor */
		if ((value1 & CIABR_PRIV) == CIABR_PRIV_HYPER)
			return H_P3;
		vcpu->arch.ciabr  = value1;
		return H_SUCCESS;
	case H_SET_MODE_RESOURCE_SET_DAWR:
		if (!kvmppc_power8_compatible(vcpu))
			return H_P2;
746 747
		if (!ppc_breakpoint_available())
			return H_P2;
748 749 750 751 752 753 754 755 756 757 758 759
		if (mflags)
			return H_UNSUPPORTED_FLAG_START;
		if (value2 & DABRX_HYP)
			return H_P4;
		vcpu->arch.dawr  = value1;
		vcpu->arch.dawrx = value2;
		return H_SUCCESS;
	default:
		return H_TOO_HARD;
	}
}

760 761 762 763 764 765 766 767 768 769 770 771 772 773
static int kvm_arch_vcpu_yield_to(struct kvm_vcpu *target)
{
	struct kvmppc_vcore *vcore = target->arch.vcore;

	/*
	 * We expect to have been called by the real mode handler
	 * (kvmppc_rm_h_confer()) which would have directly returned
	 * H_SUCCESS if the source vcore wasn't idle (e.g. if it may
	 * have useful work to do and should not confer) so we don't
	 * recheck that here.
	 */

	spin_lock(&vcore->lock);
	if (target->arch.state == KVMPPC_VCPU_RUNNABLE &&
774 775
	    vcore->vcore_state != VCORE_INACTIVE &&
	    vcore->runner)
776 777 778 779 780 781 782 783 784 785 786 787 788 789
		target = vcore->runner;
	spin_unlock(&vcore->lock);

	return kvm_vcpu_yield_to(target);
}

static int kvmppc_get_yield_count(struct kvm_vcpu *vcpu)
{
	int yield_count = 0;
	struct lppaca *lppaca;

	spin_lock(&vcpu->arch.vpa_update_lock);
	lppaca = (struct lppaca *)vcpu->arch.vpa.pinned_addr;
	if (lppaca)
790
		yield_count = be32_to_cpu(lppaca->yield_count);
791 792 793 794
	spin_unlock(&vcpu->arch.vpa_update_lock);
	return yield_count;
}

795 796 797 798
int kvmppc_pseries_do_hcall(struct kvm_vcpu *vcpu)
{
	unsigned long req = kvmppc_get_gpr(vcpu, 3);
	unsigned long target, ret = H_SUCCESS;
799
	int yield_count;
800
	struct kvm_vcpu *tvcpu;
801
	int idx, rc;
802

803 804 805 806
	if (req <= MAX_HCALL_OPCODE &&
	    !test_bit(req/4, vcpu->kvm->arch.enabled_hcalls))
		return RESUME_HOST;

807 808 809 810 811 812 813 814 815 816 817 818
	switch (req) {
	case H_CEDE:
		break;
	case H_PROD:
		target = kvmppc_get_gpr(vcpu, 4);
		tvcpu = kvmppc_find_vcpu(vcpu->kvm, target);
		if (!tvcpu) {
			ret = H_PARAMETER;
			break;
		}
		tvcpu->arch.prodded = 1;
		smp_mb();
819 820
		if (tvcpu->arch.ceded)
			kvmppc_fast_vcpu_kick_hv(tvcpu);
821 822
		break;
	case H_CONFER:
823 824 825 826 827 828 829 830
		target = kvmppc_get_gpr(vcpu, 4);
		if (target == -1)
			break;
		tvcpu = kvmppc_find_vcpu(vcpu->kvm, target);
		if (!tvcpu) {
			ret = H_PARAMETER;
			break;
		}
831 832 833 834
		yield_count = kvmppc_get_gpr(vcpu, 5);
		if (kvmppc_get_yield_count(tvcpu) != yield_count)
			break;
		kvm_arch_vcpu_yield_to(tvcpu);
835 836 837 838 839 840
		break;
	case H_REGISTER_VPA:
		ret = do_h_register_vpa(vcpu, kvmppc_get_gpr(vcpu, 4),
					kvmppc_get_gpr(vcpu, 5),
					kvmppc_get_gpr(vcpu, 6));
		break;
841 842 843 844
	case H_RTAS:
		if (list_empty(&vcpu->kvm->arch.rtas_tokens))
			return RESUME_HOST;

845
		idx = srcu_read_lock(&vcpu->kvm->srcu);
846
		rc = kvmppc_rtas_hcall(vcpu);
847
		srcu_read_unlock(&vcpu->kvm->srcu, idx);
848 849 850 851 852 853 854 855

		if (rc == -ENOENT)
			return RESUME_HOST;
		else if (rc == 0)
			break;

		/* Send the error out to userspace via KVM_RUN */
		return rc;
856 857 858 859 860 861 862 863 864 865
	case H_LOGICAL_CI_LOAD:
		ret = kvmppc_h_logical_ci_load(vcpu);
		if (ret == H_TOO_HARD)
			return RESUME_HOST;
		break;
	case H_LOGICAL_CI_STORE:
		ret = kvmppc_h_logical_ci_store(vcpu);
		if (ret == H_TOO_HARD)
			return RESUME_HOST;
		break;
866 867 868 869 870 871 872 873
	case H_SET_MODE:
		ret = kvmppc_h_set_mode(vcpu, kvmppc_get_gpr(vcpu, 4),
					kvmppc_get_gpr(vcpu, 5),
					kvmppc_get_gpr(vcpu, 6),
					kvmppc_get_gpr(vcpu, 7));
		if (ret == H_TOO_HARD)
			return RESUME_HOST;
		break;
874 875 876 877
	case H_XIRR:
	case H_CPPR:
	case H_EOI:
	case H_IPI:
878 879
	case H_IPOLL:
	case H_XIRR_X:
880
		if (kvmppc_xics_enabled(vcpu)) {
881 882 883 884
			if (xive_enabled()) {
				ret = H_NOT_AVAILABLE;
				return RESUME_GUEST;
			}
885 886
			ret = kvmppc_xics_hcall(vcpu, req);
			break;
887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911
		}
		return RESUME_HOST;
	case H_PUT_TCE:
		ret = kvmppc_h_put_tce(vcpu, kvmppc_get_gpr(vcpu, 4),
						kvmppc_get_gpr(vcpu, 5),
						kvmppc_get_gpr(vcpu, 6));
		if (ret == H_TOO_HARD)
			return RESUME_HOST;
		break;
	case H_PUT_TCE_INDIRECT:
		ret = kvmppc_h_put_tce_indirect(vcpu, kvmppc_get_gpr(vcpu, 4),
						kvmppc_get_gpr(vcpu, 5),
						kvmppc_get_gpr(vcpu, 6),
						kvmppc_get_gpr(vcpu, 7));
		if (ret == H_TOO_HARD)
			return RESUME_HOST;
		break;
	case H_STUFF_TCE:
		ret = kvmppc_h_stuff_tce(vcpu, kvmppc_get_gpr(vcpu, 4),
						kvmppc_get_gpr(vcpu, 5),
						kvmppc_get_gpr(vcpu, 6),
						kvmppc_get_gpr(vcpu, 7));
		if (ret == H_TOO_HARD)
			return RESUME_HOST;
		break;
912 913 914 915 916 917 918 919
	default:
		return RESUME_HOST;
	}
	kvmppc_set_gpr(vcpu, 3, ret);
	vcpu->arch.hcall_needed = 0;
	return RESUME_GUEST;
}

920 921 922 923 924 925 926
static int kvmppc_hcall_impl_hv(unsigned long cmd)
{
	switch (cmd) {
	case H_CEDE:
	case H_PROD:
	case H_CONFER:
	case H_REGISTER_VPA:
927
	case H_SET_MODE:
928 929
	case H_LOGICAL_CI_LOAD:
	case H_LOGICAL_CI_STORE:
930 931 932 933 934 935 936 937 938 939 940 941 942 943 944
#ifdef CONFIG_KVM_XICS
	case H_XIRR:
	case H_CPPR:
	case H_EOI:
	case H_IPI:
	case H_IPOLL:
	case H_XIRR_X:
#endif
		return 1;
	}

	/* See if it's in the real-mode table */
	return kvmppc_hcall_impl_hv_realmode(cmd);
}

945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968
static int kvmppc_emulate_debug_inst(struct kvm_run *run,
					struct kvm_vcpu *vcpu)
{
	u32 last_inst;

	if (kvmppc_get_last_inst(vcpu, INST_GENERIC, &last_inst) !=
					EMULATE_DONE) {
		/*
		 * Fetch failed, so return to guest and
		 * try executing it again.
		 */
		return RESUME_GUEST;
	}

	if (last_inst == KVMPPC_INST_SW_BREAKPOINT) {
		run->exit_reason = KVM_EXIT_DEBUG;
		run->debug.arch.address = kvmppc_get_pc(vcpu);
		return RESUME_HOST;
	} else {
		kvmppc_core_queue_program(vcpu, SRR1_PROGILL);
		return RESUME_GUEST;
	}
}

969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
static void do_nothing(void *x)
{
}

static unsigned long kvmppc_read_dpdes(struct kvm_vcpu *vcpu)
{
	int thr, cpu, pcpu, nthreads;
	struct kvm_vcpu *v;
	unsigned long dpdes;

	nthreads = vcpu->kvm->arch.emul_smt_mode;
	dpdes = 0;
	cpu = vcpu->vcpu_id & ~(nthreads - 1);
	for (thr = 0; thr < nthreads; ++thr, ++cpu) {
		v = kvmppc_find_vcpu(vcpu->kvm, cpu);
		if (!v)
			continue;
		/*
		 * If the vcpu is currently running on a physical cpu thread,
		 * interrupt it in order to pull it out of the guest briefly,
		 * which will update its vcore->dpdes value.
		 */
		pcpu = READ_ONCE(v->cpu);
		if (pcpu >= 0)
			smp_call_function_single(pcpu, do_nothing, NULL, 1);
		if (kvmppc_doorbell_pending(v))
			dpdes |= 1 << thr;
	}
	return dpdes;
}

/*
 * On POWER9, emulate doorbell-related instructions in order to
 * give the guest the illusion of running on a multi-threaded core.
 * The instructions emulated are msgsndp, msgclrp, mfspr TIR,
 * and mfspr DPDES.
 */
static int kvmppc_emulate_doorbell_instr(struct kvm_vcpu *vcpu)
{
	u32 inst, rb, thr;
	unsigned long arg;
	struct kvm *kvm = vcpu->kvm;
	struct kvm_vcpu *tvcpu;

	if (kvmppc_get_last_inst(vcpu, INST_GENERIC, &inst) != EMULATE_DONE)
		return RESUME_GUEST;
	if (get_op(inst) != 31)
		return EMULATE_FAIL;
	rb = get_rb(inst);
	thr = vcpu->vcpu_id & (kvm->arch.emul_smt_mode - 1);
	switch (get_xop(inst)) {
	case OP_31_XOP_MSGSNDP:
		arg = kvmppc_get_gpr(vcpu, rb);
		if (((arg >> 27) & 0xf) != PPC_DBELL_SERVER)
			break;
		arg &= 0x3f;
		if (arg >= kvm->arch.emul_smt_mode)
			break;
		tvcpu = kvmppc_find_vcpu(kvm, vcpu->vcpu_id - thr + arg);
		if (!tvcpu)
			break;
		if (!tvcpu->arch.doorbell_request) {
			tvcpu->arch.doorbell_request = 1;
			kvmppc_fast_vcpu_kick_hv(tvcpu);
		}
		break;
	case OP_31_XOP_MSGCLRP:
		arg = kvmppc_get_gpr(vcpu, rb);
		if (((arg >> 27) & 0xf) != PPC_DBELL_SERVER)
			break;
		vcpu->arch.vcore->dpdes = 0;
		vcpu->arch.doorbell_request = 0;
		break;
	case OP_31_XOP_MFSPR:
		switch (get_sprn(inst)) {
		case SPRN_TIR:
			arg = thr;
			break;
		case SPRN_DPDES:
			arg = kvmppc_read_dpdes(vcpu);
			break;
		default:
			return EMULATE_FAIL;
		}
		kvmppc_set_gpr(vcpu, get_rt(inst), arg);
		break;
	default:
		return EMULATE_FAIL;
	}
	kvmppc_set_pc(vcpu, kvmppc_get_pc(vcpu) + 4);
	return RESUME_GUEST;
}

1062
/* Called with vcpu->arch.vcore->lock held */
1063 1064
static int kvmppc_handle_exit_hv(struct kvm_run *run, struct kvm_vcpu *vcpu,
				 struct task_struct *tsk)
1065 1066 1067 1068 1069
{
	int r = RESUME_HOST;

	vcpu->stat.sum_exits++;

1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087
	/*
	 * This can happen if an interrupt occurs in the last stages
	 * of guest entry or the first stages of guest exit (i.e. after
	 * setting paca->kvm_hstate.in_guest to KVM_GUEST_MODE_GUEST_HV
	 * and before setting it to KVM_GUEST_MODE_HOST_HV).
	 * That can happen due to a bug, or due to a machine check
	 * occurring at just the wrong time.
	 */
	if (vcpu->arch.shregs.msr & MSR_HV) {
		printk(KERN_EMERG "KVM trap in HV mode!\n");
		printk(KERN_EMERG "trap=0x%x | pc=0x%lx | msr=0x%llx\n",
			vcpu->arch.trap, kvmppc_get_pc(vcpu),
			vcpu->arch.shregs.msr);
		kvmppc_dump_regs(vcpu);
		run->exit_reason = KVM_EXIT_INTERNAL_ERROR;
		run->hw.hardware_exit_reason = vcpu->arch.trap;
		return RESUME_HOST;
	}
1088 1089 1090 1091 1092 1093 1094 1095 1096
	run->exit_reason = KVM_EXIT_UNKNOWN;
	run->ready_for_interrupt_injection = 1;
	switch (vcpu->arch.trap) {
	/* We're good on these - the host merely wanted to get our attention */
	case BOOK3S_INTERRUPT_HV_DECREMENTER:
		vcpu->stat.dec_exits++;
		r = RESUME_GUEST;
		break;
	case BOOK3S_INTERRUPT_EXTERNAL:
1097
	case BOOK3S_INTERRUPT_H_DOORBELL:
1098
	case BOOK3S_INTERRUPT_H_VIRT:
1099 1100 1101
		vcpu->stat.ext_intr_exits++;
		r = RESUME_GUEST;
		break;
1102
	/* SR/HMI/PMI are HV interrupts that host has handled. Resume guest.*/
1103
	case BOOK3S_INTERRUPT_HMI:
1104
	case BOOK3S_INTERRUPT_PERFMON:
1105
	case BOOK3S_INTERRUPT_SYSTEM_RESET:
1106 1107
		r = RESUME_GUEST;
		break;
1108
	case BOOK3S_INTERRUPT_MACHINE_CHECK:
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122
		/* Exit to guest with KVM_EXIT_NMI as exit reason */
		run->exit_reason = KVM_EXIT_NMI;
		run->hw.hardware_exit_reason = vcpu->arch.trap;
		/* Clear out the old NMI status from run->flags */
		run->flags &= ~KVM_RUN_PPC_NMI_DISP_MASK;
		/* Now set the NMI status */
		if (vcpu->arch.mce_evt.disposition == MCE_DISPOSITION_RECOVERED)
			run->flags |= KVM_RUN_PPC_NMI_DISP_FULLY_RECOV;
		else
			run->flags |= KVM_RUN_PPC_NMI_DISP_NOT_RECOV;

		r = RESUME_HOST;
		/* Print the MCE event to host console. */
		machine_check_print_event_info(&vcpu->arch.mce_evt, false);
1123
		break;
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142
	case BOOK3S_INTERRUPT_PROGRAM:
	{
		ulong flags;
		/*
		 * Normally program interrupts are delivered directly
		 * to the guest by the hardware, but we can get here
		 * as a result of a hypervisor emulation interrupt
		 * (e40) getting turned into a 700 by BML RTAS.
		 */
		flags = vcpu->arch.shregs.msr & 0x1f0000ull;
		kvmppc_core_queue_program(vcpu, flags);
		r = RESUME_GUEST;
		break;
	}
	case BOOK3S_INTERRUPT_SYSCALL:
	{
		/* hcall - punt to userspace */
		int i;

1143 1144 1145 1146
		/* hypercall with MSR_PR has already been handled in rmode,
		 * and never reaches here.
		 */

1147 1148 1149 1150 1151 1152 1153 1154 1155
		run->papr_hcall.nr = kvmppc_get_gpr(vcpu, 3);
		for (i = 0; i < 9; ++i)
			run->papr_hcall.args[i] = kvmppc_get_gpr(vcpu, 4 + i);
		run->exit_reason = KVM_EXIT_PAPR_HCALL;
		vcpu->arch.hcall_needed = 1;
		r = RESUME_HOST;
		break;
	}
	/*
1156 1157 1158 1159 1160
	 * We get these next two if the guest accesses a page which it thinks
	 * it has mapped but which is not actually present, either because
	 * it is for an emulated I/O device or because the corresonding
	 * host page has been paged out.  Any other HDSI/HISI interrupts
	 * have been handled already.
1161 1162
	 */
	case BOOK3S_INTERRUPT_H_DATA_STORAGE:
1163
		r = RESUME_PAGE_FAULT;
1164 1165
		break;
	case BOOK3S_INTERRUPT_H_INST_STORAGE:
1166 1167 1168
		vcpu->arch.fault_dar = kvmppc_get_pc(vcpu);
		vcpu->arch.fault_dsisr = 0;
		r = RESUME_PAGE_FAULT;
1169 1170 1171
		break;
	/*
	 * This occurs if the guest executes an illegal instruction.
1172 1173 1174 1175
	 * If the guest debug is disabled, generate a program interrupt
	 * to the guest. If guest debug is enabled, we need to check
	 * whether the instruction is a software breakpoint instruction.
	 * Accordingly return to Guest or Host.
1176 1177
	 */
	case BOOK3S_INTERRUPT_H_EMUL_ASSIST:
1178 1179 1180 1181
		if (vcpu->arch.emul_inst != KVM_INST_FETCH_FAILED)
			vcpu->arch.last_inst = kvmppc_need_byteswap(vcpu) ?
				swab32(vcpu->arch.emul_inst) :
				vcpu->arch.emul_inst;
1182
		if (vcpu->guest_debug & KVM_GUESTDBG_USE_SW_BP) {
1183 1184
			/* Need vcore unlocked to call kvmppc_get_last_inst */
			spin_unlock(&vcpu->arch.vcore->lock);
1185
			r = kvmppc_emulate_debug_inst(run, vcpu);
1186
			spin_lock(&vcpu->arch.vcore->lock);
1187 1188 1189 1190
		} else {
			kvmppc_core_queue_program(vcpu, SRR1_PROGILL);
			r = RESUME_GUEST;
		}
1191 1192 1193
		break;
	/*
	 * This occurs if the guest (kernel or userspace), does something that
1194 1195 1196 1197
	 * is prohibited by HFSCR.
	 * On POWER9, this could be a doorbell instruction that we need
	 * to emulate.
	 * Otherwise, we just generate a program interrupt to the guest.
1198 1199
	 */
	case BOOK3S_INTERRUPT_H_FAC_UNAVAIL:
1200
		r = EMULATE_FAIL;
1201 1202 1203 1204
		if (((vcpu->arch.hfscr >> 56) == FSCR_MSGP_LG) &&
		    cpu_has_feature(CPU_FTR_ARCH_300)) {
			/* Need vcore unlocked to call kvmppc_get_last_inst */
			spin_unlock(&vcpu->arch.vcore->lock);
1205
			r = kvmppc_emulate_doorbell_instr(vcpu);
1206 1207
			spin_lock(&vcpu->arch.vcore->lock);
		}
1208 1209 1210 1211
		if (r == EMULATE_FAIL) {
			kvmppc_core_queue_program(vcpu, SRR1_PROGILL);
			r = RESUME_GUEST;
		}
1212
		break;
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225

#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
	case BOOK3S_INTERRUPT_HV_SOFTPATCH:
		/*
		 * This occurs for various TM-related instructions that
		 * we need to emulate on POWER9 DD2.2.  We have already
		 * handled the cases where the guest was in real-suspend
		 * mode and was transitioning to transactional state.
		 */
		r = kvmhv_p9_tm_emulation(vcpu);
		break;
#endif

1226 1227 1228
	case BOOK3S_INTERRUPT_HV_RM_HARD:
		r = RESUME_PASSTHROUGH;
		break;
1229 1230 1231 1232 1233
	default:
		kvmppc_dump_regs(vcpu);
		printk(KERN_EMERG "trap=0x%x | pc=0x%lx | msr=0x%llx\n",
			vcpu->arch.trap, kvmppc_get_pc(vcpu),
			vcpu->arch.shregs.msr);
1234
		run->hw.hardware_exit_reason = vcpu->arch.trap;
1235 1236 1237 1238 1239 1240 1241
		r = RESUME_HOST;
		break;
	}

	return r;
}

1242 1243
static int kvm_arch_vcpu_ioctl_get_sregs_hv(struct kvm_vcpu *vcpu,
					    struct kvm_sregs *sregs)
1244 1245 1246 1247
{
	int i;

	memset(sregs, 0, sizeof(struct kvm_sregs));
1248
	sregs->pvr = vcpu->arch.pvr;
1249 1250 1251 1252 1253 1254 1255 1256
	for (i = 0; i < vcpu->arch.slb_max; i++) {
		sregs->u.s.ppc64.slb[i].slbe = vcpu->arch.slb[i].orige;
		sregs->u.s.ppc64.slb[i].slbv = vcpu->arch.slb[i].origv;
	}

	return 0;
}

1257 1258
static int kvm_arch_vcpu_ioctl_set_sregs_hv(struct kvm_vcpu *vcpu,
					    struct kvm_sregs *sregs)
1259 1260 1261
{
	int i, j;

1262 1263 1264
	/* Only accept the same PVR as the host's, since we can't spoof it */
	if (sregs->pvr != vcpu->arch.pvr)
		return -EINVAL;
1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278

	j = 0;
	for (i = 0; i < vcpu->arch.slb_nr; i++) {
		if (sregs->u.s.ppc64.slb[i].slbe & SLB_ESID_V) {
			vcpu->arch.slb[j].orige = sregs->u.s.ppc64.slb[i].slbe;
			vcpu->arch.slb[j].origv = sregs->u.s.ppc64.slb[i].slbv;
			++j;
		}
	}
	vcpu->arch.slb_max = j;

	return 0;
}

1279 1280
static void kvmppc_set_lpcr(struct kvm_vcpu *vcpu, u64 new_lpcr,
		bool preserve_top32)
1281
{
1282
	struct kvm *kvm = vcpu->kvm;
1283 1284 1285
	struct kvmppc_vcore *vc = vcpu->arch.vcore;
	u64 mask;

1286
	mutex_lock(&kvm->lock);
1287
	spin_lock(&vc->lock);
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
	/*
	 * If ILE (interrupt little-endian) has changed, update the
	 * MSR_LE bit in the intr_msr for each vcpu in this vcore.
	 */
	if ((new_lpcr & LPCR_ILE) != (vc->lpcr & LPCR_ILE)) {
		struct kvm_vcpu *vcpu;
		int i;

		kvm_for_each_vcpu(i, vcpu, kvm) {
			if (vcpu->arch.vcore != vc)
				continue;
			if (new_lpcr & LPCR_ILE)
				vcpu->arch.intr_msr |= MSR_LE;
			else
				vcpu->arch.intr_msr &= ~MSR_LE;
		}
	}

1306 1307 1308
	/*
	 * Userspace can only modify DPFD (default prefetch depth),
	 * ILE (interrupt little-endian) and TC (translation control).
1309
	 * On POWER8 and POWER9 userspace can also modify AIL (alt. interrupt loc.).
1310 1311
	 */
	mask = LPCR_DPFD | LPCR_ILE | LPCR_TC;
1312 1313
	if (cpu_has_feature(CPU_FTR_ARCH_207S))
		mask |= LPCR_AIL;
1314 1315 1316 1317 1318 1319
	/*
	 * On POWER9, allow userspace to enable large decrementer for the
	 * guest, whether or not the host has it enabled.
	 */
	if (cpu_has_feature(CPU_FTR_ARCH_300))
		mask |= LPCR_LD;
1320 1321 1322 1323

	/* Broken 32-bit version of LPCR must not clear top bits */
	if (preserve_top32)
		mask &= 0xFFFFFFFF;
1324 1325
	vc->lpcr = (vc->lpcr & ~mask) | (new_lpcr & mask);
	spin_unlock(&vc->lock);
1326
	mutex_unlock(&kvm->lock);
1327 1328
}

1329 1330
static int kvmppc_get_one_reg_hv(struct kvm_vcpu *vcpu, u64 id,
				 union kvmppc_one_reg *val)
1331
{
1332 1333
	int r = 0;
	long int i;
1334

1335
	switch (id) {
1336 1337 1338
	case KVM_REG_PPC_DEBUG_INST:
		*val = get_reg_val(id, KVMPPC_INST_SW_BREAKPOINT);
		break;
1339
	case KVM_REG_PPC_HIOR:
1340 1341 1342 1343 1344
		*val = get_reg_val(id, 0);
		break;
	case KVM_REG_PPC_DABR:
		*val = get_reg_val(id, vcpu->arch.dabr);
		break;
1345 1346 1347
	case KVM_REG_PPC_DABRX:
		*val = get_reg_val(id, vcpu->arch.dabrx);
		break;
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362
	case KVM_REG_PPC_DSCR:
		*val = get_reg_val(id, vcpu->arch.dscr);
		break;
	case KVM_REG_PPC_PURR:
		*val = get_reg_val(id, vcpu->arch.purr);
		break;
	case KVM_REG_PPC_SPURR:
		*val = get_reg_val(id, vcpu->arch.spurr);
		break;
	case KVM_REG_PPC_AMR:
		*val = get_reg_val(id, vcpu->arch.amr);
		break;
	case KVM_REG_PPC_UAMOR:
		*val = get_reg_val(id, vcpu->arch.uamor);
		break;
1363
	case KVM_REG_PPC_MMCR0 ... KVM_REG_PPC_MMCRS:
1364 1365 1366 1367 1368 1369
		i = id - KVM_REG_PPC_MMCR0;
		*val = get_reg_val(id, vcpu->arch.mmcr[i]);
		break;
	case KVM_REG_PPC_PMC1 ... KVM_REG_PPC_PMC8:
		i = id - KVM_REG_PPC_PMC1;
		*val = get_reg_val(id, vcpu->arch.pmc[i]);
1370
		break;
1371 1372 1373 1374
	case KVM_REG_PPC_SPMC1 ... KVM_REG_PPC_SPMC2:
		i = id - KVM_REG_PPC_SPMC1;
		*val = get_reg_val(id, vcpu->arch.spmc[i]);
		break;
1375 1376 1377 1378 1379 1380
	case KVM_REG_PPC_SIAR:
		*val = get_reg_val(id, vcpu->arch.siar);
		break;
	case KVM_REG_PPC_SDAR:
		*val = get_reg_val(id, vcpu->arch.sdar);
		break;
1381 1382
	case KVM_REG_PPC_SIER:
		*val = get_reg_val(id, vcpu->arch.sier);
1383
		break;
1384 1385 1386 1387 1388 1389 1390 1391 1392
	case KVM_REG_PPC_IAMR:
		*val = get_reg_val(id, vcpu->arch.iamr);
		break;
	case KVM_REG_PPC_PSPB:
		*val = get_reg_val(id, vcpu->arch.pspb);
		break;
	case KVM_REG_PPC_DPDES:
		*val = get_reg_val(id, vcpu->arch.vcore->dpdes);
		break;
1393 1394 1395
	case KVM_REG_PPC_VTB:
		*val = get_reg_val(id, vcpu->arch.vcore->vtb);
		break;
1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421
	case KVM_REG_PPC_DAWR:
		*val = get_reg_val(id, vcpu->arch.dawr);
		break;
	case KVM_REG_PPC_DAWRX:
		*val = get_reg_val(id, vcpu->arch.dawrx);
		break;
	case KVM_REG_PPC_CIABR:
		*val = get_reg_val(id, vcpu->arch.ciabr);
		break;
	case KVM_REG_PPC_CSIGR:
		*val = get_reg_val(id, vcpu->arch.csigr);
		break;
	case KVM_REG_PPC_TACR:
		*val = get_reg_val(id, vcpu->arch.tacr);
		break;
	case KVM_REG_PPC_TCSCR:
		*val = get_reg_val(id, vcpu->arch.tcscr);
		break;
	case KVM_REG_PPC_PID:
		*val = get_reg_val(id, vcpu->arch.pid);
		break;
	case KVM_REG_PPC_ACOP:
		*val = get_reg_val(id, vcpu->arch.acop);
		break;
	case KVM_REG_PPC_WORT:
		*val = get_reg_val(id, vcpu->arch.wort);
1422
		break;
1423 1424 1425 1426 1427 1428
	case KVM_REG_PPC_TIDR:
		*val = get_reg_val(id, vcpu->arch.tid);
		break;
	case KVM_REG_PPC_PSSCR:
		*val = get_reg_val(id, vcpu->arch.psscr);
		break;
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445
	case KVM_REG_PPC_VPA_ADDR:
		spin_lock(&vcpu->arch.vpa_update_lock);
		*val = get_reg_val(id, vcpu->arch.vpa.next_gpa);
		spin_unlock(&vcpu->arch.vpa_update_lock);
		break;
	case KVM_REG_PPC_VPA_SLB:
		spin_lock(&vcpu->arch.vpa_update_lock);
		val->vpaval.addr = vcpu->arch.slb_shadow.next_gpa;
		val->vpaval.length = vcpu->arch.slb_shadow.len;
		spin_unlock(&vcpu->arch.vpa_update_lock);
		break;
	case KVM_REG_PPC_VPA_DTL:
		spin_lock(&vcpu->arch.vpa_update_lock);
		val->vpaval.addr = vcpu->arch.dtl.next_gpa;
		val->vpaval.length = vcpu->arch.dtl.len;
		spin_unlock(&vcpu->arch.vpa_update_lock);
		break;
1446 1447 1448
	case KVM_REG_PPC_TB_OFFSET:
		*val = get_reg_val(id, vcpu->arch.vcore->tb_offset);
		break;
1449
	case KVM_REG_PPC_LPCR:
1450
	case KVM_REG_PPC_LPCR_64:
1451 1452
		*val = get_reg_val(id, vcpu->arch.vcore->lpcr);
		break;
1453 1454 1455
	case KVM_REG_PPC_PPR:
		*val = get_reg_val(id, vcpu->arch.ppr);
		break;
1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487
#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
	case KVM_REG_PPC_TFHAR:
		*val = get_reg_val(id, vcpu->arch.tfhar);
		break;
	case KVM_REG_PPC_TFIAR:
		*val = get_reg_val(id, vcpu->arch.tfiar);
		break;
	case KVM_REG_PPC_TEXASR:
		*val = get_reg_val(id, vcpu->arch.texasr);
		break;
	case KVM_REG_PPC_TM_GPR0 ... KVM_REG_PPC_TM_GPR31:
		i = id - KVM_REG_PPC_TM_GPR0;
		*val = get_reg_val(id, vcpu->arch.gpr_tm[i]);
		break;
	case KVM_REG_PPC_TM_VSR0 ... KVM_REG_PPC_TM_VSR63:
	{
		int j;
		i = id - KVM_REG_PPC_TM_VSR0;
		if (i < 32)
			for (j = 0; j < TS_FPRWIDTH; j++)
				val->vsxval[j] = vcpu->arch.fp_tm.fpr[i][j];
		else {
			if (cpu_has_feature(CPU_FTR_ALTIVEC))
				val->vval = vcpu->arch.vr_tm.vr[i-32];
			else
				r = -ENXIO;
		}
		break;
	}
	case KVM_REG_PPC_TM_CR:
		*val = get_reg_val(id, vcpu->arch.cr_tm);
		break;
1488 1489 1490
	case KVM_REG_PPC_TM_XER:
		*val = get_reg_val(id, vcpu->arch.xer_tm);
		break;
1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521
	case KVM_REG_PPC_TM_LR:
		*val = get_reg_val(id, vcpu->arch.lr_tm);
		break;
	case KVM_REG_PPC_TM_CTR:
		*val = get_reg_val(id, vcpu->arch.ctr_tm);
		break;
	case KVM_REG_PPC_TM_FPSCR:
		*val = get_reg_val(id, vcpu->arch.fp_tm.fpscr);
		break;
	case KVM_REG_PPC_TM_AMR:
		*val = get_reg_val(id, vcpu->arch.amr_tm);
		break;
	case KVM_REG_PPC_TM_PPR:
		*val = get_reg_val(id, vcpu->arch.ppr_tm);
		break;
	case KVM_REG_PPC_TM_VRSAVE:
		*val = get_reg_val(id, vcpu->arch.vrsave_tm);
		break;
	case KVM_REG_PPC_TM_VSCR:
		if (cpu_has_feature(CPU_FTR_ALTIVEC))
			*val = get_reg_val(id, vcpu->arch.vr_tm.vscr.u[3]);
		else
			r = -ENXIO;
		break;
	case KVM_REG_PPC_TM_DSCR:
		*val = get_reg_val(id, vcpu->arch.dscr_tm);
		break;
	case KVM_REG_PPC_TM_TAR:
		*val = get_reg_val(id, vcpu->arch.tar_tm);
		break;
#endif
1522 1523 1524
	case KVM_REG_PPC_ARCH_COMPAT:
		*val = get_reg_val(id, vcpu->arch.vcore->arch_compat);
		break;
1525 1526 1527 1528
	case KVM_REG_PPC_DEC_EXPIRY:
		*val = get_reg_val(id, vcpu->arch.dec_expires +
				   vcpu->arch.vcore->tb_offset);
		break;
1529
	default:
1530
		r = -EINVAL;
1531 1532 1533 1534 1535 1536
		break;
	}

	return r;
}

1537 1538
static int kvmppc_set_one_reg_hv(struct kvm_vcpu *vcpu, u64 id,
				 union kvmppc_one_reg *val)
1539
{
1540 1541
	int r = 0;
	long int i;
1542
	unsigned long addr, len;
1543

1544
	switch (id) {
1545 1546
	case KVM_REG_PPC_HIOR:
		/* Only allow this to be set to zero */
1547
		if (set_reg_val(id, *val))
1548 1549
			r = -EINVAL;
		break;
1550 1551 1552
	case KVM_REG_PPC_DABR:
		vcpu->arch.dabr = set_reg_val(id, *val);
		break;
1553 1554 1555
	case KVM_REG_PPC_DABRX:
		vcpu->arch.dabrx = set_reg_val(id, *val) & ~DABRX_HYP;
		break;
1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570
	case KVM_REG_PPC_DSCR:
		vcpu->arch.dscr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_PURR:
		vcpu->arch.purr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_SPURR:
		vcpu->arch.spurr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_AMR:
		vcpu->arch.amr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_UAMOR:
		vcpu->arch.uamor = set_reg_val(id, *val);
		break;
1571
	case KVM_REG_PPC_MMCR0 ... KVM_REG_PPC_MMCRS:
1572 1573 1574 1575 1576 1577 1578
		i = id - KVM_REG_PPC_MMCR0;
		vcpu->arch.mmcr[i] = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_PMC1 ... KVM_REG_PPC_PMC8:
		i = id - KVM_REG_PPC_PMC1;
		vcpu->arch.pmc[i] = set_reg_val(id, *val);
		break;
1579 1580 1581 1582
	case KVM_REG_PPC_SPMC1 ... KVM_REG_PPC_SPMC2:
		i = id - KVM_REG_PPC_SPMC1;
		vcpu->arch.spmc[i] = set_reg_val(id, *val);
		break;
1583 1584 1585 1586 1587 1588
	case KVM_REG_PPC_SIAR:
		vcpu->arch.siar = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_SDAR:
		vcpu->arch.sdar = set_reg_val(id, *val);
		break;
1589 1590
	case KVM_REG_PPC_SIER:
		vcpu->arch.sier = set_reg_val(id, *val);
1591
		break;
1592 1593 1594 1595 1596 1597 1598 1599 1600
	case KVM_REG_PPC_IAMR:
		vcpu->arch.iamr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_PSPB:
		vcpu->arch.pspb = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_DPDES:
		vcpu->arch.vcore->dpdes = set_reg_val(id, *val);
		break;
1601 1602 1603
	case KVM_REG_PPC_VTB:
		vcpu->arch.vcore->vtb = set_reg_val(id, *val);
		break;
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632
	case KVM_REG_PPC_DAWR:
		vcpu->arch.dawr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_DAWRX:
		vcpu->arch.dawrx = set_reg_val(id, *val) & ~DAWRX_HYP;
		break;
	case KVM_REG_PPC_CIABR:
		vcpu->arch.ciabr = set_reg_val(id, *val);
		/* Don't allow setting breakpoints in hypervisor code */
		if ((vcpu->arch.ciabr & CIABR_PRIV) == CIABR_PRIV_HYPER)
			vcpu->arch.ciabr &= ~CIABR_PRIV;	/* disable */
		break;
	case KVM_REG_PPC_CSIGR:
		vcpu->arch.csigr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TACR:
		vcpu->arch.tacr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TCSCR:
		vcpu->arch.tcscr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_PID:
		vcpu->arch.pid = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_ACOP:
		vcpu->arch.acop = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_WORT:
		vcpu->arch.wort = set_reg_val(id, *val);
1633
		break;
1634 1635 1636 1637 1638 1639
	case KVM_REG_PPC_TIDR:
		vcpu->arch.tid = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_PSSCR:
		vcpu->arch.psscr = set_reg_val(id, *val) & PSSCR_GUEST_VIS;
		break;
1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
	case KVM_REG_PPC_VPA_ADDR:
		addr = set_reg_val(id, *val);
		r = -EINVAL;
		if (!addr && (vcpu->arch.slb_shadow.next_gpa ||
			      vcpu->arch.dtl.next_gpa))
			break;
		r = set_vpa(vcpu, &vcpu->arch.vpa, addr, sizeof(struct lppaca));
		break;
	case KVM_REG_PPC_VPA_SLB:
		addr = val->vpaval.addr;
		len = val->vpaval.length;
		r = -EINVAL;
		if (addr && !vcpu->arch.vpa.next_gpa)
			break;
		r = set_vpa(vcpu, &vcpu->arch.slb_shadow, addr, len);
		break;
	case KVM_REG_PPC_VPA_DTL:
		addr = val->vpaval.addr;
		len = val->vpaval.length;
		r = -EINVAL;
1660 1661
		if (addr && (len < sizeof(struct dtl_entry) ||
			     !vcpu->arch.vpa.next_gpa))
1662 1663 1664 1665
			break;
		len -= len % sizeof(struct dtl_entry);
		r = set_vpa(vcpu, &vcpu->arch.dtl, addr, len);
		break;
1666
	case KVM_REG_PPC_TB_OFFSET:
1667 1668 1669 1670 1671 1672 1673 1674
		/*
		 * POWER9 DD1 has an erratum where writing TBU40 causes
		 * the timebase to lose ticks.  So we don't let the
		 * timebase offset be changed on P9 DD1.  (It is
		 * initialized to zero.)
		 */
		if (cpu_has_feature(CPU_FTR_POWER9_DD1))
			break;
1675 1676 1677 1678
		/* round up to multiple of 2^24 */
		vcpu->arch.vcore->tb_offset =
			ALIGN(set_reg_val(id, *val), 1UL << 24);
		break;
1679
	case KVM_REG_PPC_LPCR:
1680 1681 1682 1683
		kvmppc_set_lpcr(vcpu, set_reg_val(id, *val), true);
		break;
	case KVM_REG_PPC_LPCR_64:
		kvmppc_set_lpcr(vcpu, set_reg_val(id, *val), false);
1684
		break;
1685 1686 1687
	case KVM_REG_PPC_PPR:
		vcpu->arch.ppr = set_reg_val(id, *val);
		break;
1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718
#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
	case KVM_REG_PPC_TFHAR:
		vcpu->arch.tfhar = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TFIAR:
		vcpu->arch.tfiar = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TEXASR:
		vcpu->arch.texasr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_GPR0 ... KVM_REG_PPC_TM_GPR31:
		i = id - KVM_REG_PPC_TM_GPR0;
		vcpu->arch.gpr_tm[i] = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_VSR0 ... KVM_REG_PPC_TM_VSR63:
	{
		int j;
		i = id - KVM_REG_PPC_TM_VSR0;
		if (i < 32)
			for (j = 0; j < TS_FPRWIDTH; j++)
				vcpu->arch.fp_tm.fpr[i][j] = val->vsxval[j];
		else
			if (cpu_has_feature(CPU_FTR_ALTIVEC))
				vcpu->arch.vr_tm.vr[i-32] = val->vval;
			else
				r = -ENXIO;
		break;
	}
	case KVM_REG_PPC_TM_CR:
		vcpu->arch.cr_tm = set_reg_val(id, *val);
		break;
1719 1720 1721
	case KVM_REG_PPC_TM_XER:
		vcpu->arch.xer_tm = set_reg_val(id, *val);
		break;
1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752
	case KVM_REG_PPC_TM_LR:
		vcpu->arch.lr_tm = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_CTR:
		vcpu->arch.ctr_tm = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_FPSCR:
		vcpu->arch.fp_tm.fpscr = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_AMR:
		vcpu->arch.amr_tm = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_PPR:
		vcpu->arch.ppr_tm = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_VRSAVE:
		vcpu->arch.vrsave_tm = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_VSCR:
		if (cpu_has_feature(CPU_FTR_ALTIVEC))
			vcpu->arch.vr.vscr.u[3] = set_reg_val(id, *val);
		else
			r = - ENXIO;
		break;
	case KVM_REG_PPC_TM_DSCR:
		vcpu->arch.dscr_tm = set_reg_val(id, *val);
		break;
	case KVM_REG_PPC_TM_TAR:
		vcpu->arch.tar_tm = set_reg_val(id, *val);
		break;
#endif
1753 1754 1755
	case KVM_REG_PPC_ARCH_COMPAT:
		r = kvmppc_set_arch_compat(vcpu, set_reg_val(id, *val));
		break;
1756 1757 1758 1759
	case KVM_REG_PPC_DEC_EXPIRY:
		vcpu->arch.dec_expires = set_reg_val(id, *val) -
			vcpu->arch.vcore->tb_offset;
		break;
1760
	default:
1761
		r = -EINVAL;
1762 1763 1764 1765 1766 1767
		break;
	}

	return r;
}

1768 1769 1770 1771 1772 1773 1774
/*
 * On POWER9, threads are independent and can be in different partitions.
 * Therefore we consider each thread to be a subcore.
 * There is a restriction that all threads have to be in the same
 * MMU mode (radix or HPT), unfortunately, but since we only support
 * HPT guests on a HPT host so far, that isn't an impediment yet.
 */
1775
static int threads_per_vcore(struct kvm *kvm)
1776
{
1777
	if (kvm->arch.threads_indep)
1778 1779 1780 1781
		return 1;
	return threads_per_subcore;
}

1782 1783 1784 1785 1786 1787 1788 1789 1790 1791
static struct kvmppc_vcore *kvmppc_vcore_create(struct kvm *kvm, int core)
{
	struct kvmppc_vcore *vcore;

	vcore = kzalloc(sizeof(struct kvmppc_vcore), GFP_KERNEL);

	if (vcore == NULL)
		return NULL;

	spin_lock_init(&vcore->lock);
1792
	spin_lock_init(&vcore->stoltb_lock);
1793
	init_swait_queue_head(&vcore->wq);
1794 1795
	vcore->preempt_tb = TB_NIL;
	vcore->lpcr = kvm->arch.lpcr;
1796
	vcore->first_vcpuid = core * kvm->arch.smt_mode;
1797
	vcore->kvm = kvm;
1798
	INIT_LIST_HEAD(&vcore->preempt_list);
1799 1800 1801 1802

	return vcore;
}

1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814
#ifdef CONFIG_KVM_BOOK3S_HV_EXIT_TIMING
static struct debugfs_timings_element {
	const char *name;
	size_t offset;
} timings[] = {
	{"rm_entry",	offsetof(struct kvm_vcpu, arch.rm_entry)},
	{"rm_intr",	offsetof(struct kvm_vcpu, arch.rm_intr)},
	{"rm_exit",	offsetof(struct kvm_vcpu, arch.rm_exit)},
	{"guest",	offsetof(struct kvm_vcpu, arch.guest_time)},
	{"cede",	offsetof(struct kvm_vcpu, arch.cede_time)},
};

1815
#define N_TIMINGS	(ARRAY_SIZE(timings))
1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950

struct debugfs_timings_state {
	struct kvm_vcpu	*vcpu;
	unsigned int	buflen;
	char		buf[N_TIMINGS * 100];
};

static int debugfs_timings_open(struct inode *inode, struct file *file)
{
	struct kvm_vcpu *vcpu = inode->i_private;
	struct debugfs_timings_state *p;

	p = kzalloc(sizeof(*p), GFP_KERNEL);
	if (!p)
		return -ENOMEM;

	kvm_get_kvm(vcpu->kvm);
	p->vcpu = vcpu;
	file->private_data = p;

	return nonseekable_open(inode, file);
}

static int debugfs_timings_release(struct inode *inode, struct file *file)
{
	struct debugfs_timings_state *p = file->private_data;

	kvm_put_kvm(p->vcpu->kvm);
	kfree(p);
	return 0;
}

static ssize_t debugfs_timings_read(struct file *file, char __user *buf,
				    size_t len, loff_t *ppos)
{
	struct debugfs_timings_state *p = file->private_data;
	struct kvm_vcpu *vcpu = p->vcpu;
	char *s, *buf_end;
	struct kvmhv_tb_accumulator tb;
	u64 count;
	loff_t pos;
	ssize_t n;
	int i, loops;
	bool ok;

	if (!p->buflen) {
		s = p->buf;
		buf_end = s + sizeof(p->buf);
		for (i = 0; i < N_TIMINGS; ++i) {
			struct kvmhv_tb_accumulator *acc;

			acc = (struct kvmhv_tb_accumulator *)
				((unsigned long)vcpu + timings[i].offset);
			ok = false;
			for (loops = 0; loops < 1000; ++loops) {
				count = acc->seqcount;
				if (!(count & 1)) {
					smp_rmb();
					tb = *acc;
					smp_rmb();
					if (count == acc->seqcount) {
						ok = true;
						break;
					}
				}
				udelay(1);
			}
			if (!ok)
				snprintf(s, buf_end - s, "%s: stuck\n",
					timings[i].name);
			else
				snprintf(s, buf_end - s,
					"%s: %llu %llu %llu %llu\n",
					timings[i].name, count / 2,
					tb_to_ns(tb.tb_total),
					tb_to_ns(tb.tb_min),
					tb_to_ns(tb.tb_max));
			s += strlen(s);
		}
		p->buflen = s - p->buf;
	}

	pos = *ppos;
	if (pos >= p->buflen)
		return 0;
	if (len > p->buflen - pos)
		len = p->buflen - pos;
	n = copy_to_user(buf, p->buf + pos, len);
	if (n) {
		if (n == len)
			return -EFAULT;
		len -= n;
	}
	*ppos = pos + len;
	return len;
}

static ssize_t debugfs_timings_write(struct file *file, const char __user *buf,
				     size_t len, loff_t *ppos)
{
	return -EACCES;
}

static const struct file_operations debugfs_timings_ops = {
	.owner	 = THIS_MODULE,
	.open	 = debugfs_timings_open,
	.release = debugfs_timings_release,
	.read	 = debugfs_timings_read,
	.write	 = debugfs_timings_write,
	.llseek	 = generic_file_llseek,
};

/* Create a debugfs directory for the vcpu */
static void debugfs_vcpu_init(struct kvm_vcpu *vcpu, unsigned int id)
{
	char buf[16];
	struct kvm *kvm = vcpu->kvm;

	snprintf(buf, sizeof(buf), "vcpu%u", id);
	if (IS_ERR_OR_NULL(kvm->arch.debugfs_dir))
		return;
	vcpu->arch.debugfs_dir = debugfs_create_dir(buf, kvm->arch.debugfs_dir);
	if (IS_ERR_OR_NULL(vcpu->arch.debugfs_dir))
		return;
	vcpu->arch.debugfs_timings =
		debugfs_create_file("timings", 0444, vcpu->arch.debugfs_dir,
				    vcpu, &debugfs_timings_ops);
}

#else /* CONFIG_KVM_BOOK3S_HV_EXIT_TIMING */
static void debugfs_vcpu_init(struct kvm_vcpu *vcpu, unsigned int id)
{
}
#endif /* CONFIG_KVM_BOOK3S_HV_EXIT_TIMING */

1951 1952
static struct kvm_vcpu *kvmppc_core_vcpu_create_hv(struct kvm *kvm,
						   unsigned int id)
1953 1954
{
	struct kvm_vcpu *vcpu;
1955
	int err;
1956 1957
	int core;
	struct kvmppc_vcore *vcore;
1958

1959
	err = -ENOMEM;
1960
	vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
1961 1962 1963 1964 1965 1966 1967 1968
	if (!vcpu)
		goto out;

	err = kvm_vcpu_init(vcpu, kvm, id);
	if (err)
		goto free_vcpu;

	vcpu->arch.shared = &vcpu->arch.shregs;
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979
#ifdef CONFIG_KVM_BOOK3S_PR_POSSIBLE
	/*
	 * The shared struct is never shared on HV,
	 * so we can always use host endianness
	 */
#ifdef __BIG_ENDIAN__
	vcpu->arch.shared_big_endian = true;
#else
	vcpu->arch.shared_big_endian = false;
#endif
#endif
1980 1981 1982
	vcpu->arch.mmcr[0] = MMCR0_FC;
	vcpu->arch.ctrl = CTRL_RUNLATCH;
	/* default to host PVR, since we can't spoof it */
1983
	kvmppc_set_pvr_hv(vcpu, mfspr(SPRN_PVR));
1984
	spin_lock_init(&vcpu->arch.vpa_update_lock);
1985 1986
	spin_lock_init(&vcpu->arch.tbacct_lock);
	vcpu->arch.busy_preempt = TB_NIL;
1987
	vcpu->arch.intr_msr = MSR_SF | MSR_ME;
1988

1989 1990 1991 1992 1993
	/*
	 * Set the default HFSCR for the guest from the host value.
	 * This value is only used on POWER9.
	 * On POWER9 DD1, TM doesn't work, so we make sure to
	 * prevent the guest from using it.
1994 1995
	 * On POWER9, we want to virtualize the doorbell facility, so we
	 * turn off the HFSCR bit, which causes those instructions to trap.
1996 1997
	 */
	vcpu->arch.hfscr = mfspr(SPRN_HFSCR);
1998 1999 2000
	if (cpu_has_feature(CPU_FTR_P9_TM_HV_ASSIST))
		vcpu->arch.hfscr |= HFSCR_TM;
	else if (!cpu_has_feature(CPU_FTR_TM_COMP))
2001
		vcpu->arch.hfscr &= ~HFSCR_TM;
2002 2003
	if (cpu_has_feature(CPU_FTR_ARCH_300))
		vcpu->arch.hfscr &= ~HFSCR_MSGP;
2004

2005 2006
	kvmppc_mmu_book3s_hv_init(vcpu);

2007
	vcpu->arch.state = KVMPPC_VCPU_NOTREADY;
2008 2009 2010 2011

	init_waitqueue_head(&vcpu->arch.cpu_run);

	mutex_lock(&kvm->lock);
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
	vcore = NULL;
	err = -EINVAL;
	core = id / kvm->arch.smt_mode;
	if (core < KVM_MAX_VCORES) {
		vcore = kvm->arch.vcores[core];
		if (!vcore) {
			err = -ENOMEM;
			vcore = kvmppc_vcore_create(kvm, core);
			kvm->arch.vcores[core] = vcore;
			kvm->arch.online_vcores++;
		}
2023 2024 2025 2026 2027 2028 2029 2030 2031 2032
	}
	mutex_unlock(&kvm->lock);

	if (!vcore)
		goto free_vcpu;

	spin_lock(&vcore->lock);
	++vcore->num_threads;
	spin_unlock(&vcore->lock);
	vcpu->arch.vcore = vcore;
2033
	vcpu->arch.ptid = vcpu->vcpu_id - vcore->first_vcpuid;
2034
	vcpu->arch.thread_cpu = -1;
2035
	vcpu->arch.prev_cpu = -1;
2036

2037 2038 2039
	vcpu->arch.cpu_type = KVM_CPU_3S_64;
	kvmppc_sanity_check(vcpu);

2040 2041
	debugfs_vcpu_init(vcpu, id);

2042 2043 2044
	return vcpu;

free_vcpu:
2045
	kmem_cache_free(kvm_vcpu_cache, vcpu);
2046 2047 2048 2049
out:
	return ERR_PTR(err);
}

2050 2051 2052 2053
static int kvmhv_set_smt_mode(struct kvm *kvm, unsigned long smt_mode,
			      unsigned long flags)
{
	int err;
2054
	int esmt = 0;
2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071

	if (flags)
		return -EINVAL;
	if (smt_mode > MAX_SMT_THREADS || !is_power_of_2(smt_mode))
		return -EINVAL;
	if (!cpu_has_feature(CPU_FTR_ARCH_300)) {
		/*
		 * On POWER8 (or POWER7), the threading mode is "strict",
		 * so we pack smt_mode vcpus per vcore.
		 */
		if (smt_mode > threads_per_subcore)
			return -EINVAL;
	} else {
		/*
		 * On POWER9, the threading mode is "loose",
		 * so each vcpu gets its own vcore.
		 */
2072
		esmt = smt_mode;
2073 2074 2075 2076 2077 2078
		smt_mode = 1;
	}
	mutex_lock(&kvm->lock);
	err = -EBUSY;
	if (!kvm->arch.online_vcores) {
		kvm->arch.smt_mode = smt_mode;
2079
		kvm->arch.emul_smt_mode = esmt;
2080 2081 2082 2083 2084 2085 2086
		err = 0;
	}
	mutex_unlock(&kvm->lock);

	return err;
}

2087 2088 2089 2090 2091 2092 2093
static void unpin_vpa(struct kvm *kvm, struct kvmppc_vpa *vpa)
{
	if (vpa->pinned_addr)
		kvmppc_unpin_guest_page(kvm, vpa->pinned_addr, vpa->gpa,
					vpa->dirty);
}

2094
static void kvmppc_core_vcpu_free_hv(struct kvm_vcpu *vcpu)
2095
{
2096
	spin_lock(&vcpu->arch.vpa_update_lock);
2097 2098 2099
	unpin_vpa(vcpu->kvm, &vcpu->arch.dtl);
	unpin_vpa(vcpu->kvm, &vcpu->arch.slb_shadow);
	unpin_vpa(vcpu->kvm, &vcpu->arch.vpa);
2100
	spin_unlock(&vcpu->arch.vpa_update_lock);
2101
	kvm_vcpu_uninit(vcpu);
2102
	kmem_cache_free(kvm_vcpu_cache, vcpu);
2103 2104
}

2105 2106 2107 2108 2109 2110
static int kvmppc_core_check_requests_hv(struct kvm_vcpu *vcpu)
{
	/* Indicate we want to get back into the guest */
	return 1;
}

2111
static void kvmppc_set_timer(struct kvm_vcpu *vcpu)
2112
{
2113
	unsigned long dec_nsec, now;
2114

2115 2116 2117 2118
	now = get_tb();
	if (now > vcpu->arch.dec_expires) {
		/* decrementer has already gone negative */
		kvmppc_core_queue_dec(vcpu);
2119
		kvmppc_core_prepare_to_enter(vcpu);
2120
		return;
2121
	}
2122 2123
	dec_nsec = (vcpu->arch.dec_expires - now) * NSEC_PER_SEC
		   / tb_ticks_per_sec;
T
Thomas Gleixner 已提交
2124
	hrtimer_start(&vcpu->arch.dec_timer, dec_nsec, HRTIMER_MODE_REL);
2125
	vcpu->arch.timer_running = 1;
2126 2127
}

2128
static void kvmppc_end_cede(struct kvm_vcpu *vcpu)
2129
{
2130 2131 2132 2133 2134
	vcpu->arch.ceded = 0;
	if (vcpu->arch.timer_running) {
		hrtimer_try_to_cancel(&vcpu->arch.dec_timer);
		vcpu->arch.timer_running = 0;
	}
2135 2136
}

2137
extern int __kvmppc_vcore_entry(void);
2138

2139 2140
static void kvmppc_remove_runnable(struct kvmppc_vcore *vc,
				   struct kvm_vcpu *vcpu)
2141
{
2142 2143
	u64 now;

2144 2145
	if (vcpu->arch.state != KVMPPC_VCPU_RUNNABLE)
		return;
2146
	spin_lock_irq(&vcpu->arch.tbacct_lock);
2147 2148 2149 2150 2151
	now = mftb();
	vcpu->arch.busy_stolen += vcore_stolen_time(vc, now) -
		vcpu->arch.stolen_logged;
	vcpu->arch.busy_preempt = now;
	vcpu->arch.state = KVMPPC_VCPU_BUSY_IN_HOST;
2152
	spin_unlock_irq(&vcpu->arch.tbacct_lock);
2153
	--vc->n_runnable;
2154
	WRITE_ONCE(vc->runnable_threads[vcpu->arch.ptid], NULL);
2155 2156
}

2157 2158 2159
static int kvmppc_grab_hwthread(int cpu)
{
	struct paca_struct *tpaca;
2160
	long timeout = 10000;
2161

2162
	tpaca = paca_ptrs[cpu];
2163 2164

	/* Ensure the thread won't go into the kernel if it wakes */
2165
	tpaca->kvm_hstate.kvm_vcpu = NULL;
2166
	tpaca->kvm_hstate.kvm_vcore = NULL;
2167 2168 2169
	tpaca->kvm_hstate.napping = 0;
	smp_wmb();
	tpaca->kvm_hstate.hwthread_req = 1;
2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194

	/*
	 * If the thread is already executing in the kernel (e.g. handling
	 * a stray interrupt), wait for it to get back to nap mode.
	 * The smp_mb() is to ensure that our setting of hwthread_req
	 * is visible before we look at hwthread_state, so if this
	 * races with the code at system_reset_pSeries and the thread
	 * misses our setting of hwthread_req, we are sure to see its
	 * setting of hwthread_state, and vice versa.
	 */
	smp_mb();
	while (tpaca->kvm_hstate.hwthread_state == KVM_HWTHREAD_IN_KERNEL) {
		if (--timeout <= 0) {
			pr_err("KVM: couldn't grab cpu %d\n", cpu);
			return -EBUSY;
		}
		udelay(1);
	}
	return 0;
}

static void kvmppc_release_hwthread(int cpu)
{
	struct paca_struct *tpaca;

2195
	tpaca = paca_ptrs[cpu];
2196
	tpaca->kvm_hstate.hwthread_req = 0;
2197
	tpaca->kvm_hstate.kvm_vcpu = NULL;
2198 2199
	tpaca->kvm_hstate.kvm_vcore = NULL;
	tpaca->kvm_hstate.kvm_split_mode = NULL;
2200 2201
}

2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218
static void radix_flush_cpu(struct kvm *kvm, int cpu, struct kvm_vcpu *vcpu)
{
	int i;

	cpu = cpu_first_thread_sibling(cpu);
	cpumask_set_cpu(cpu, &kvm->arch.need_tlb_flush);
	/*
	 * Make sure setting of bit in need_tlb_flush precedes
	 * testing of cpu_in_guest bits.  The matching barrier on
	 * the other side is the first smp_mb() in kvmppc_run_core().
	 */
	smp_mb();
	for (i = 0; i < threads_per_core; ++i)
		if (cpumask_test_cpu(cpu + i, &kvm->arch.cpu_in_guest))
			smp_call_function_single(cpu + i, do_nothing, NULL, 1);
}

2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243
static void kvmppc_prepare_radix_vcpu(struct kvm_vcpu *vcpu, int pcpu)
{
	struct kvm *kvm = vcpu->kvm;

	/*
	 * With radix, the guest can do TLB invalidations itself,
	 * and it could choose to use the local form (tlbiel) if
	 * it is invalidating a translation that has only ever been
	 * used on one vcpu.  However, that doesn't mean it has
	 * only ever been used on one physical cpu, since vcpus
	 * can move around between pcpus.  To cope with this, when
	 * a vcpu moves from one pcpu to another, we need to tell
	 * any vcpus running on the same core as this vcpu previously
	 * ran to flush the TLB.  The TLB is shared between threads,
	 * so we use a single bit in .need_tlb_flush for all 4 threads.
	 */
	if (vcpu->arch.prev_cpu != pcpu) {
		if (vcpu->arch.prev_cpu >= 0 &&
		    cpu_first_thread_sibling(vcpu->arch.prev_cpu) !=
		    cpu_first_thread_sibling(pcpu))
			radix_flush_cpu(kvm, vcpu->arch.prev_cpu, vcpu);
		vcpu->arch.prev_cpu = pcpu;
	}
}

2244
static void kvmppc_start_thread(struct kvm_vcpu *vcpu, struct kvmppc_vcore *vc)
2245 2246 2247
{
	int cpu;
	struct paca_struct *tpaca;
2248
	struct kvm *kvm = vc->kvm;
2249

2250 2251 2252 2253 2254 2255 2256
	cpu = vc->pcpu;
	if (vcpu) {
		if (vcpu->arch.timer_running) {
			hrtimer_try_to_cancel(&vcpu->arch.dec_timer);
			vcpu->arch.timer_running = 0;
		}
		cpu += vcpu->arch.ptid;
2257
		vcpu->cpu = vc->pcpu;
2258
		vcpu->arch.thread_cpu = cpu;
2259
		cpumask_set_cpu(cpu, &kvm->arch.cpu_in_guest);
2260
	}
2261
	tpaca = paca_ptrs[cpu];
2262
	tpaca->kvm_hstate.kvm_vcpu = vcpu;
2263
	tpaca->kvm_hstate.ptid = cpu - vc->pcpu;
2264
	tpaca->kvm_hstate.fake_suspend = 0;
2265
	/* Order stores to hstate.kvm_vcpu etc. before store to kvm_vcore */
2266
	smp_wmb();
2267
	tpaca->kvm_hstate.kvm_vcore = vc;
2268
	if (cpu != smp_processor_id())
2269
		kvmppc_ipi_thread(cpu);
2270
}
2271

2272
static void kvmppc_wait_for_nap(int n_threads)
2273
{
2274 2275
	int cpu = smp_processor_id();
	int i, loops;
2276

2277 2278
	if (n_threads <= 1)
		return;
2279 2280 2281
	for (loops = 0; loops < 1000000; ++loops) {
		/*
		 * Check if all threads are finished.
2282
		 * We set the vcore pointer when starting a thread
2283
		 * and the thread clears it when finished, so we look
2284
		 * for any threads that still have a non-NULL vcore ptr.
2285
		 */
2286
		for (i = 1; i < n_threads; ++i)
2287
			if (paca_ptrs[cpu + i]->kvm_hstate.kvm_vcore)
2288
				break;
2289
		if (i == n_threads) {
2290 2291
			HMT_medium();
			return;
2292
		}
2293
		HMT_low();
2294 2295
	}
	HMT_medium();
2296
	for (i = 1; i < n_threads; ++i)
2297
		if (paca_ptrs[cpu + i]->kvm_hstate.kvm_vcore)
2298
			pr_err("KVM: CPU %d seems to be stuck\n", cpu + i);
2299 2300 2301 2302
}

/*
 * Check that we are on thread 0 and that any other threads in
2303 2304
 * this core are off-line.  Then grab the threads so they can't
 * enter the kernel.
2305 2306 2307 2308
 */
static int on_primary_thread(void)
{
	int cpu = smp_processor_id();
2309
	int thr;
2310

2311 2312
	/* Are we on a primary subcore? */
	if (cpu_thread_in_subcore(cpu))
2313
		return 0;
2314 2315 2316

	thr = 0;
	while (++thr < threads_per_subcore)
2317 2318
		if (cpu_online(cpu + thr))
			return 0;
2319 2320

	/* Grab all hw threads so they can't go into the kernel */
2321
	for (thr = 1; thr < threads_per_subcore; ++thr) {
2322 2323 2324 2325 2326 2327 2328 2329
		if (kvmppc_grab_hwthread(cpu + thr)) {
			/* Couldn't grab one; let the others go */
			do {
				kvmppc_release_hwthread(cpu + thr);
			} while (--thr > 0);
			return 0;
		}
	}
2330 2331 2332
	return 1;
}

2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361
/*
 * A list of virtual cores for each physical CPU.
 * These are vcores that could run but their runner VCPU tasks are
 * (or may be) preempted.
 */
struct preempted_vcore_list {
	struct list_head	list;
	spinlock_t		lock;
};

static DEFINE_PER_CPU(struct preempted_vcore_list, preempted_vcores);

static void init_vcore_lists(void)
{
	int cpu;

	for_each_possible_cpu(cpu) {
		struct preempted_vcore_list *lp = &per_cpu(preempted_vcores, cpu);
		spin_lock_init(&lp->lock);
		INIT_LIST_HEAD(&lp->list);
	}
}

static void kvmppc_vcore_preempt(struct kvmppc_vcore *vc)
{
	struct preempted_vcore_list *lp = this_cpu_ptr(&preempted_vcores);

	vc->vcore_state = VCORE_PREEMPT;
	vc->pcpu = smp_processor_id();
2362
	if (vc->num_threads < threads_per_vcore(vc->kvm)) {
2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373
		spin_lock(&lp->lock);
		list_add_tail(&vc->preempt_list, &lp->list);
		spin_unlock(&lp->lock);
	}

	/* Start accumulating stolen time */
	kvmppc_core_start_stolen(vc);
}

static void kvmppc_vcore_end_preempt(struct kvmppc_vcore *vc)
{
2374
	struct preempted_vcore_list *lp;
2375 2376 2377

	kvmppc_core_end_stolen(vc);
	if (!list_empty(&vc->preempt_list)) {
2378
		lp = &per_cpu(preempted_vcores, vc->pcpu);
2379 2380 2381 2382 2383 2384 2385
		spin_lock(&lp->lock);
		list_del_init(&vc->preempt_list);
		spin_unlock(&lp->lock);
	}
	vc->vcore_state = VCORE_INACTIVE;
}

2386 2387 2388 2389
/*
 * This stores information about the virtual cores currently
 * assigned to a physical core.
 */
2390
struct core_info {
2391 2392
	int		n_subcores;
	int		max_subcore_threads;
2393
	int		total_threads;
2394
	int		subcore_threads[MAX_SUBCORES];
2395
	struct kvmppc_vcore *vc[MAX_SUBCORES];
2396 2397
};

2398 2399
/*
 * This mapping means subcores 0 and 1 can use threads 0-3 and 4-7
2400
 * respectively in 2-way micro-threading (split-core) mode on POWER8.
2401 2402 2403
 */
static int subcore_thread_map[MAX_SUBCORES] = { 0, 4, 2, 6 };

2404 2405 2406
static void init_core_info(struct core_info *cip, struct kvmppc_vcore *vc)
{
	memset(cip, 0, sizeof(*cip));
2407 2408
	cip->n_subcores = 1;
	cip->max_subcore_threads = vc->num_threads;
2409
	cip->total_threads = vc->num_threads;
2410
	cip->subcore_threads[0] = vc->num_threads;
2411
	cip->vc[0] = vc;
2412 2413 2414 2415
}

static bool subcore_config_ok(int n_subcores, int n_threads)
{
2416
	/*
2417 2418
	 * POWER9 "SMT4" cores are permanently in what is effectively a 4-way
	 * split-core mode, with one thread per subcore.
2419 2420 2421 2422 2423
	 */
	if (cpu_has_feature(CPU_FTR_ARCH_300))
		return n_subcores <= 4 && n_threads == 1;

	/* On POWER8, can only dynamically split if unsplit to begin with */
2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435
	if (n_subcores > 1 && threads_per_subcore < MAX_SMT_THREADS)
		return false;
	if (n_subcores > MAX_SUBCORES)
		return false;
	if (n_subcores > 1) {
		if (!(dynamic_mt_modes & 2))
			n_subcores = 4;
		if (n_subcores > 2 && !(dynamic_mt_modes & 4))
			return false;
	}

	return n_subcores * roundup_pow_of_two(n_threads) <= MAX_SMT_THREADS;
2436 2437
}

2438
static void init_vcore_to_run(struct kvmppc_vcore *vc)
2439 2440 2441 2442 2443
{
	vc->entry_exit_map = 0;
	vc->in_guest = 0;
	vc->napping_threads = 0;
	vc->conferring_threads = 0;
2444
	vc->tb_offset_applied = 0;
2445 2446
}

2447 2448 2449 2450 2451 2452 2453 2454
static bool can_dynamic_split(struct kvmppc_vcore *vc, struct core_info *cip)
{
	int n_threads = vc->num_threads;
	int sub;

	if (!cpu_has_feature(CPU_FTR_ARCH_207S))
		return false;

2455 2456
	/* Some POWER9 chips require all threads to be in the same MMU mode */
	if (no_mixing_hpt_and_radix &&
2457 2458 2459
	    kvm_is_radix(vc->kvm) != kvm_is_radix(cip->vc[0]->kvm))
		return false;

2460 2461
	if (n_threads < cip->max_subcore_threads)
		n_threads = cip->max_subcore_threads;
2462
	if (!subcore_config_ok(cip->n_subcores + 1, n_threads))
2463
		return false;
2464
	cip->max_subcore_threads = n_threads;
2465 2466 2467 2468 2469

	sub = cip->n_subcores;
	++cip->n_subcores;
	cip->total_threads += vc->num_threads;
	cip->subcore_threads[sub] = vc->num_threads;
2470 2471 2472
	cip->vc[sub] = vc;
	init_vcore_to_run(vc);
	list_del_init(&vc->preempt_list);
2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486

	return true;
}

/*
 * Work out whether it is possible to piggyback the execution of
 * vcore *pvc onto the execution of the other vcores described in *cip.
 */
static bool can_piggyback(struct kvmppc_vcore *pvc, struct core_info *cip,
			  int target_threads)
{
	if (cip->total_threads + pvc->num_threads > target_threads)
		return false;

2487
	return can_dynamic_split(pvc, cip);
2488 2489
}

2490 2491
static void prepare_threads(struct kvmppc_vcore *vc)
{
2492 2493
	int i;
	struct kvm_vcpu *vcpu;
2494

2495
	for_each_runnable_thread(i, vcpu, vc) {
2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508
		if (signal_pending(vcpu->arch.run_task))
			vcpu->arch.ret = -EINTR;
		else if (vcpu->arch.vpa.update_pending ||
			 vcpu->arch.slb_shadow.update_pending ||
			 vcpu->arch.dtl.update_pending)
			vcpu->arch.ret = RESUME_GUEST;
		else
			continue;
		kvmppc_remove_runnable(vc, vcpu);
		wake_up(&vcpu->arch.cpu_run);
	}
}

2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539
static void collect_piggybacks(struct core_info *cip, int target_threads)
{
	struct preempted_vcore_list *lp = this_cpu_ptr(&preempted_vcores);
	struct kvmppc_vcore *pvc, *vcnext;

	spin_lock(&lp->lock);
	list_for_each_entry_safe(pvc, vcnext, &lp->list, preempt_list) {
		if (!spin_trylock(&pvc->lock))
			continue;
		prepare_threads(pvc);
		if (!pvc->n_runnable) {
			list_del_init(&pvc->preempt_list);
			if (pvc->runner == NULL) {
				pvc->vcore_state = VCORE_INACTIVE;
				kvmppc_core_end_stolen(pvc);
			}
			spin_unlock(&pvc->lock);
			continue;
		}
		if (!can_piggyback(pvc, cip, target_threads)) {
			spin_unlock(&pvc->lock);
			continue;
		}
		kvmppc_core_end_stolen(pvc);
		pvc->vcore_state = VCORE_PIGGYBACK;
		if (cip->total_threads >= target_threads)
			break;
	}
	spin_unlock(&lp->lock);
}

2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551
static bool recheck_signals(struct core_info *cip)
{
	int sub, i;
	struct kvm_vcpu *vcpu;

	for (sub = 0; sub < cip->n_subcores; ++sub)
		for_each_runnable_thread(i, vcpu, cip->vc[sub])
			if (signal_pending(vcpu->arch.run_task))
				return true;
	return false;
}

2552
static void post_guest_process(struct kvmppc_vcore *vc, bool is_master)
2553
{
2554
	int still_running = 0, i;
2555 2556
	u64 now;
	long ret;
2557
	struct kvm_vcpu *vcpu;
2558

2559
	spin_lock(&vc->lock);
2560
	now = get_tb();
2561
	for_each_runnable_thread(i, vcpu, vc) {
2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576
		/* cancel pending dec exception if dec is positive */
		if (now < vcpu->arch.dec_expires &&
		    kvmppc_core_pending_dec(vcpu))
			kvmppc_core_dequeue_dec(vcpu);

		trace_kvm_guest_exit(vcpu);

		ret = RESUME_GUEST;
		if (vcpu->arch.trap)
			ret = kvmppc_handle_exit_hv(vcpu->arch.kvm_run, vcpu,
						    vcpu->arch.run_task);

		vcpu->arch.ret = ret;
		vcpu->arch.trap = 0;

2577 2578 2579 2580
		if (is_kvmppc_resume_guest(vcpu->arch.ret)) {
			if (vcpu->arch.pending_exceptions)
				kvmppc_core_prepare_to_enter(vcpu);
			if (vcpu->arch.ceded)
2581
				kvmppc_set_timer(vcpu);
2582 2583 2584
			else
				++still_running;
		} else {
2585 2586 2587 2588
			kvmppc_remove_runnable(vc, vcpu);
			wake_up(&vcpu->arch.cpu_run);
		}
	}
2589
	if (!is_master) {
2590
		if (still_running > 0) {
2591
			kvmppc_vcore_preempt(vc);
2592 2593 2594 2595 2596 2597
		} else if (vc->runner) {
			vc->vcore_state = VCORE_PREEMPT;
			kvmppc_core_start_stolen(vc);
		} else {
			vc->vcore_state = VCORE_INACTIVE;
		}
2598 2599
		if (vc->n_runnable > 0 && vc->runner == NULL) {
			/* make sure there's a candidate runner awake */
2600 2601
			i = -1;
			vcpu = next_runnable_thread(vc, &i);
2602 2603 2604 2605
			wake_up(&vcpu->arch.cpu_run);
		}
	}
	spin_unlock(&vc->lock);
2606 2607
}

2608 2609 2610 2611 2612
/*
 * Clear core from the list of active host cores as we are about to
 * enter the guest. Only do this if it is the primary thread of the
 * core (not if a subcore) that is entering the guest.
 */
2613
static inline int kvmppc_clear_host_core(unsigned int cpu)
2614 2615 2616 2617
{
	int core;

	if (!kvmppc_host_rm_ops_hv || cpu_thread_in_core(cpu))
2618
		return 0;
2619 2620 2621 2622 2623 2624 2625
	/*
	 * Memory barrier can be omitted here as we will do a smp_wmb()
	 * later in kvmppc_start_thread and we need ensure that state is
	 * visible to other CPUs only after we enter guest.
	 */
	core = cpu >> threads_shift;
	kvmppc_host_rm_ops_hv->rm_core[core].rm_state.in_host = 0;
2626
	return 0;
2627 2628 2629 2630 2631 2632 2633
}

/*
 * Advertise this core as an active host core since we exited the guest
 * Only need to do this if it is the primary thread of the core that is
 * exiting.
 */
2634
static inline int kvmppc_set_host_core(unsigned int cpu)
2635 2636 2637 2638
{
	int core;

	if (!kvmppc_host_rm_ops_hv || cpu_thread_in_core(cpu))
2639
		return 0;
2640 2641 2642 2643 2644 2645 2646

	/*
	 * Memory barrier can be omitted here because we do a spin_unlock
	 * immediately after this which provides the memory barrier.
	 */
	core = cpu >> threads_shift;
	kvmppc_host_rm_ops_hv->rm_core[core].rm_state.in_host = 1;
2647
	return 0;
2648 2649
}

2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661
static void set_irq_happened(int trap)
{
	switch (trap) {
	case BOOK3S_INTERRUPT_EXTERNAL:
		local_paca->irq_happened |= PACA_IRQ_EE;
		break;
	case BOOK3S_INTERRUPT_H_DOORBELL:
		local_paca->irq_happened |= PACA_IRQ_DBELL;
		break;
	case BOOK3S_INTERRUPT_HMI:
		local_paca->irq_happened |= PACA_IRQ_HMI;
		break;
2662 2663 2664
	case BOOK3S_INTERRUPT_SYSTEM_RESET:
		replay_system_reset();
		break;
2665 2666 2667
	}
}

2668 2669 2670 2671
/*
 * Run a set of guest threads on a physical core.
 * Called with vc->lock held.
 */
2672
static noinline void kvmppc_run_core(struct kvmppc_vcore *vc)
2673
{
2674
	struct kvm_vcpu *vcpu;
2675
	int i;
2676
	int srcu_idx;
2677
	struct core_info core_info;
2678
	struct kvmppc_vcore *pvc;
2679 2680 2681 2682 2683
	struct kvm_split_mode split_info, *sip;
	int split, subcore_size, active;
	int sub;
	bool thr0_done;
	unsigned long cmd_bit, stat_bit;
2684 2685
	int pcpu, thr;
	int target_threads;
2686
	int controlled_threads;
2687
	int trap;
2688
	bool is_power8;
2689
	bool hpt_on_radix;
2690

2691 2692 2693 2694 2695 2696 2697 2698 2699
	/*
	 * Remove from the list any threads that have a signal pending
	 * or need a VPA update done
	 */
	prepare_threads(vc);

	/* if the runner is no longer runnable, let the caller pick a new one */
	if (vc->runner->arch.state != KVMPPC_VCPU_RUNNABLE)
		return;
2700 2701

	/*
2702
	 * Initialize *vc.
2703
	 */
2704
	init_vcore_to_run(vc);
2705
	vc->preempt_tb = TB_NIL;
2706

2707 2708 2709 2710 2711
	/*
	 * Number of threads that we will be controlling: the same as
	 * the number of threads per subcore, except on POWER9,
	 * where it's 1 because the threads are (mostly) independent.
	 */
2712
	controlled_threads = threads_per_vcore(vc->kvm);
2713

2714
	/*
2715 2716 2717
	 * Make sure we are running on primary threads, and that secondary
	 * threads are offline.  Also check if the number of threads in this
	 * guest are greater than the current system threads per guest.
2718
	 * On POWER9, we need to be not in independent-threads mode if
2719 2720
	 * this is a HPT guest on a radix host machine where the
	 * CPU threads may not be in different MMU modes.
2721
	 */
2722 2723
	hpt_on_radix = no_mixing_hpt_and_radix && radix_enabled() &&
		!kvm_is_radix(vc->kvm);
2724 2725 2726
	if (((controlled_threads > 1) &&
	     ((vc->num_threads > threads_per_subcore) || !on_primary_thread())) ||
	    (hpt_on_radix && vc->kvm->arch.threads_indep)) {
2727
		for_each_runnable_thread(i, vcpu, vc) {
2728
			vcpu->arch.ret = -EBUSY;
2729 2730 2731
			kvmppc_remove_runnable(vc, vcpu);
			wake_up(&vcpu->arch.cpu_run);
		}
2732 2733 2734
		goto out;
	}

2735 2736 2737 2738 2739 2740
	/*
	 * See if we could run any other vcores on the physical core
	 * along with this one.
	 */
	init_core_info(&core_info, vc);
	pcpu = smp_processor_id();
2741
	target_threads = controlled_threads;
2742 2743 2744 2745
	if (target_smt_mode && target_smt_mode < target_threads)
		target_threads = target_smt_mode;
	if (vc->num_threads < target_threads)
		collect_piggybacks(&core_info, target_threads);
2746

2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762
	/*
	 * On radix, arrange for TLB flushing if necessary.
	 * This has to be done before disabling interrupts since
	 * it uses smp_call_function().
	 */
	pcpu = smp_processor_id();
	if (kvm_is_radix(vc->kvm)) {
		for (sub = 0; sub < core_info.n_subcores; ++sub)
			for_each_runnable_thread(i, vcpu, core_info.vc[sub])
				kvmppc_prepare_radix_vcpu(vcpu, pcpu);
	}

	/*
	 * Hard-disable interrupts, and check resched flag and signals.
	 * If we need to reschedule or deliver a signal, clean up
	 * and return without going into the guest(s).
2763
	 * If the mmu_ready flag has been cleared, don't go into the
2764
	 * guest because that means a HPT resize operation is in progress.
2765 2766 2767 2768
	 */
	local_irq_disable();
	hard_irq_disable();
	if (lazy_irq_pending() || need_resched() ||
2769
	    recheck_signals(&core_info) || !vc->kvm->arch.mmu_ready) {
2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785
		local_irq_enable();
		vc->vcore_state = VCORE_INACTIVE;
		/* Unlock all except the primary vcore */
		for (sub = 1; sub < core_info.n_subcores; ++sub) {
			pvc = core_info.vc[sub];
			/* Put back on to the preempted vcores list */
			kvmppc_vcore_preempt(pvc);
			spin_unlock(&pvc->lock);
		}
		for (i = 0; i < controlled_threads; ++i)
			kvmppc_release_hwthread(pcpu + i);
		return;
	}

	kvmppc_clear_host_core(pcpu);

2786 2787 2788 2789 2790
	/* Decide on micro-threading (split-core) mode */
	subcore_size = threads_per_subcore;
	cmd_bit = stat_bit = 0;
	split = core_info.n_subcores;
	sip = NULL;
2791 2792 2793
	is_power8 = cpu_has_feature(CPU_FTR_ARCH_207S)
		&& !cpu_has_feature(CPU_FTR_ARCH_300);

2794
	if (split > 1 || hpt_on_radix) {
2795 2796 2797
		sip = &split_info;
		memset(&split_info, 0, sizeof(split_info));
		for (sub = 0; sub < core_info.n_subcores; ++sub)
2798
			split_info.vc[sub] = core_info.vc[sub];
2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815

		if (is_power8) {
			if (split == 2 && (dynamic_mt_modes & 2)) {
				cmd_bit = HID0_POWER8_1TO2LPAR;
				stat_bit = HID0_POWER8_2LPARMODE;
			} else {
				split = 4;
				cmd_bit = HID0_POWER8_1TO4LPAR;
				stat_bit = HID0_POWER8_4LPARMODE;
			}
			subcore_size = MAX_SMT_THREADS / split;
			split_info.rpr = mfspr(SPRN_RPR);
			split_info.pmmar = mfspr(SPRN_PMMAR);
			split_info.ldbar = mfspr(SPRN_LDBAR);
			split_info.subcore_size = subcore_size;
		} else {
			split_info.subcore_size = 1;
2816 2817 2818 2819 2820 2821 2822
			if (hpt_on_radix) {
				/* Use the split_info for LPCR/LPIDR changes */
				split_info.lpcr_req = vc->lpcr;
				split_info.lpidr_req = vc->kvm->arch.lpid;
				split_info.host_lpcr = vc->kvm->arch.host_lpcr;
				split_info.do_set = 1;
			}
2823 2824
		}

2825 2826 2827
		/* order writes to split_info before kvm_split_mode pointer */
		smp_wmb();
	}
2828 2829

	for (thr = 0; thr < controlled_threads; ++thr) {
2830 2831 2832 2833 2834
		struct paca_struct *paca = paca_ptrs[pcpu + thr];

		paca->kvm_hstate.tid = thr;
		paca->kvm_hstate.napping = 0;
		paca->kvm_hstate.kvm_split_mode = sip;
2835
	}
2836

2837
	/* Initiate micro-threading (split-core) on POWER8 if required */
2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849
	if (cmd_bit) {
		unsigned long hid0 = mfspr(SPRN_HID0);

		hid0 |= cmd_bit | HID0_POWER8_DYNLPARDIS;
		mb();
		mtspr(SPRN_HID0, hid0);
		isync();
		for (;;) {
			hid0 = mfspr(SPRN_HID0);
			if (hid0 & stat_bit)
				break;
			cpu_relax();
2850
		}
2851
	}
2852

2853 2854 2855
	/* Start all the threads */
	active = 0;
	for (sub = 0; sub < core_info.n_subcores; ++sub) {
2856
		thr = is_power8 ? subcore_thread_map[sub] : sub;
2857 2858
		thr0_done = false;
		active |= 1 << thr;
2859 2860 2861 2862 2863 2864 2865 2866 2867
		pvc = core_info.vc[sub];
		pvc->pcpu = pcpu + thr;
		for_each_runnable_thread(i, vcpu, pvc) {
			kvmppc_start_thread(vcpu, pvc);
			kvmppc_create_dtl_entry(vcpu, pvc);
			trace_kvm_guest_enter(vcpu);
			if (!vcpu->arch.ptid)
				thr0_done = true;
			active |= 1 << (thr + vcpu->arch.ptid);
2868
		}
2869 2870 2871 2872 2873 2874
		/*
		 * We need to start the first thread of each subcore
		 * even if it doesn't have a vcpu.
		 */
		if (!thr0_done)
			kvmppc_start_thread(NULL, pvc);
2875
	}
2876

2877 2878 2879 2880 2881 2882
	/*
	 * Ensure that split_info.do_nap is set after setting
	 * the vcore pointer in the PACA of the secondaries.
	 */
	smp_mb();

2883 2884 2885 2886
	/*
	 * When doing micro-threading, poke the inactive threads as well.
	 * This gets them to the nap instruction after kvm_do_nap,
	 * which reduces the time taken to unsplit later.
2887 2888
	 * For POWER9 HPT guest on radix host, we need all the secondary
	 * threads woken up so they can do the LPCR/LPIDR change.
2889
	 */
2890
	if (cmd_bit || hpt_on_radix) {
2891
		split_info.do_nap = 1;	/* ask secondaries to nap when done */
2892 2893 2894
		for (thr = 1; thr < threads_per_subcore; ++thr)
			if (!(active & (1 << thr)))
				kvmppc_ipi_thread(pcpu + thr);
2895
	}
2896

2897
	vc->vcore_state = VCORE_RUNNING;
2898
	preempt_disable();
2899 2900 2901

	trace_kvmppc_run_core(vc, 0);

2902
	for (sub = 0; sub < core_info.n_subcores; ++sub)
2903
		spin_unlock(&core_info.vc[sub]->lock);
2904

2905 2906 2907 2908 2909
	/*
	 * Interrupts will be enabled once we get into the guest,
	 * so tell lockdep that we're about to enable interrupts.
	 */
	trace_hardirqs_on();
2910

2911
	guest_enter_irqoff();
2912

2913
	srcu_idx = srcu_read_lock(&vc->kvm->srcu);
2914

2915 2916
	this_cpu_disable_ftrace();

2917
	trap = __kvmppc_vcore_entry();
2918

2919 2920
	this_cpu_enable_ftrace();

2921 2922
	srcu_read_unlock(&vc->kvm->srcu, srcu_idx);

2923 2924 2925
	trace_hardirqs_off();
	set_irq_happened(trap);

2926
	spin_lock(&vc->lock);
2927
	/* prevent other vcpu threads from doing kvmppc_start_thread() now */
2928
	vc->vcore_state = VCORE_EXITING;
2929

2930
	/* wait for secondary threads to finish writing their state to memory */
2931
	kvmppc_wait_for_nap(controlled_threads);
2932 2933

	/* Return to whole-core mode if we split the core earlier */
2934
	if (cmd_bit) {
2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949
		unsigned long hid0 = mfspr(SPRN_HID0);
		unsigned long loops = 0;

		hid0 &= ~HID0_POWER8_DYNLPARDIS;
		stat_bit = HID0_POWER8_2LPARMODE | HID0_POWER8_4LPARMODE;
		mb();
		mtspr(SPRN_HID0, hid0);
		isync();
		for (;;) {
			hid0 = mfspr(SPRN_HID0);
			if (!(hid0 & stat_bit))
				break;
			cpu_relax();
			++loops;
		}
2950 2951 2952
	} else if (hpt_on_radix) {
		/* Wait for all threads to have seen final sync */
		for (thr = 1; thr < controlled_threads; ++thr) {
2953 2954 2955
			struct paca_struct *paca = paca_ptrs[pcpu + thr];

			while (paca->kvm_hstate.kvm_split_mode) {
2956 2957 2958 2959 2960
				HMT_low();
				barrier();
			}
			HMT_medium();
		}
2961
	}
2962
	split_info.do_nap = 0;
2963

2964 2965 2966
	kvmppc_set_host_core(pcpu);

	local_irq_enable();
2967
	guest_exit();
2968

2969
	/* Let secondaries go back to the offline loop */
2970
	for (i = 0; i < controlled_threads; ++i) {
2971 2972 2973
		kvmppc_release_hwthread(pcpu + i);
		if (sip && sip->napped[i])
			kvmppc_ipi_thread(pcpu + i);
2974
		cpumask_clear_cpu(pcpu + i, &vc->kvm->arch.cpu_in_guest);
2975 2976
	}

2977
	spin_unlock(&vc->lock);
2978

2979 2980
	/* make sure updates to secondary vcpu structs are visible now */
	smp_mb();
2981

2982 2983
	preempt_enable();

2984 2985 2986 2987
	for (sub = 0; sub < core_info.n_subcores; ++sub) {
		pvc = core_info.vc[sub];
		post_guest_process(pvc, pvc == vc);
	}
2988

2989
	spin_lock(&vc->lock);
2990 2991

 out:
2992
	vc->vcore_state = VCORE_INACTIVE;
2993
	trace_kvmppc_run_core(vc, 1);
2994 2995
}

2996 2997 2998 2999
/*
 * Wait for some other vcpu thread to execute us, and
 * wake us up when we need to handle something in the host.
 */
3000 3001
static void kvmppc_wait_for_exec(struct kvmppc_vcore *vc,
				 struct kvm_vcpu *vcpu, int wait_state)
3002 3003 3004
{
	DEFINE_WAIT(wait);

3005
	prepare_to_wait(&vcpu->arch.cpu_run, &wait, wait_state);
3006 3007
	if (vcpu->arch.state == KVMPPC_VCPU_RUNNABLE) {
		spin_unlock(&vc->lock);
3008
		schedule();
3009 3010
		spin_lock(&vc->lock);
	}
3011 3012 3013
	finish_wait(&vcpu->arch.cpu_run, &wait);
}

3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030
static void grow_halt_poll_ns(struct kvmppc_vcore *vc)
{
	/* 10us base */
	if (vc->halt_poll_ns == 0 && halt_poll_ns_grow)
		vc->halt_poll_ns = 10000;
	else
		vc->halt_poll_ns *= halt_poll_ns_grow;
}

static void shrink_halt_poll_ns(struct kvmppc_vcore *vc)
{
	if (halt_poll_ns_shrink == 0)
		vc->halt_poll_ns = 0;
	else
		vc->halt_poll_ns /= halt_poll_ns_shrink;
}

3031 3032 3033 3034 3035
#ifdef CONFIG_KVM_XICS
static inline bool xive_interrupt_pending(struct kvm_vcpu *vcpu)
{
	if (!xive_enabled())
		return false;
3036
	return vcpu->arch.irq_pending || vcpu->arch.xive_saved_state.pipr <
3037 3038 3039 3040 3041 3042 3043 3044 3045
		vcpu->arch.xive_saved_state.cppr;
}
#else
static inline bool xive_interrupt_pending(struct kvm_vcpu *vcpu)
{
	return false;
}
#endif /* CONFIG_KVM_XICS */

3046 3047 3048
static bool kvmppc_vcpu_woken(struct kvm_vcpu *vcpu)
{
	if (vcpu->arch.pending_exceptions || vcpu->arch.prodded ||
3049
	    kvmppc_doorbell_pending(vcpu) || xive_interrupt_pending(vcpu))
3050 3051 3052 3053 3054
		return true;

	return false;
}

3055 3056
/*
 * Check to see if any of the runnable vcpus on the vcore have pending
3057 3058 3059 3060 3061 3062 3063 3064
 * exceptions or are no longer ceded
 */
static int kvmppc_vcore_check_block(struct kvmppc_vcore *vc)
{
	struct kvm_vcpu *vcpu;
	int i;

	for_each_runnable_thread(i, vcpu, vc) {
3065
		if (!vcpu->arch.ceded || kvmppc_vcpu_woken(vcpu))
3066 3067 3068 3069 3070 3071
			return 1;
	}

	return 0;
}

3072 3073 3074 3075 3076 3077
/*
 * All the vcpus in this vcore are idle, so wait for a decrementer
 * or external interrupt to one of the vcpus.  vc->lock is held.
 */
static void kvmppc_vcore_blocked(struct kvmppc_vcore *vc)
{
3078
	ktime_t cur, start_poll, start_wait;
3079 3080
	int do_sleep = 1;
	u64 block_ns;
3081
	DECLARE_SWAITQUEUE(wait);
3082

3083
	/* Poll for pending exceptions and ceded state */
3084
	cur = start_poll = ktime_get();
3085
	if (vc->halt_poll_ns) {
3086 3087
		ktime_t stop = ktime_add_ns(start_poll, vc->halt_poll_ns);
		++vc->runner->stat.halt_attempted_poll;
3088

3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102
		vc->vcore_state = VCORE_POLLING;
		spin_unlock(&vc->lock);

		do {
			if (kvmppc_vcore_check_block(vc)) {
				do_sleep = 0;
				break;
			}
			cur = ktime_get();
		} while (single_task_running() && ktime_before(cur, stop));

		spin_lock(&vc->lock);
		vc->vcore_state = VCORE_INACTIVE;

3103 3104
		if (!do_sleep) {
			++vc->runner->stat.halt_successful_poll;
3105
			goto out;
3106
		}
3107 3108
	}

3109 3110 3111
	prepare_to_swait(&vc->wq, &wait, TASK_INTERRUPTIBLE);

	if (kvmppc_vcore_check_block(vc)) {
3112
		finish_swait(&vc->wq, &wait);
3113
		do_sleep = 0;
3114 3115 3116
		/* If we polled, count this as a successful poll */
		if (vc->halt_poll_ns)
			++vc->runner->stat.halt_successful_poll;
3117
		goto out;
3118 3119
	}

3120 3121
	start_wait = ktime_get();

3122
	vc->vcore_state = VCORE_SLEEPING;
3123
	trace_kvmppc_vcore_blocked(vc, 0);
3124
	spin_unlock(&vc->lock);
3125
	schedule();
3126
	finish_swait(&vc->wq, &wait);
3127 3128
	spin_lock(&vc->lock);
	vc->vcore_state = VCORE_INACTIVE;
3129
	trace_kvmppc_vcore_blocked(vc, 1);
3130
	++vc->runner->stat.halt_successful_wait;
3131 3132 3133 3134

	cur = ktime_get();

out:
3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152
	block_ns = ktime_to_ns(cur) - ktime_to_ns(start_poll);

	/* Attribute wait time */
	if (do_sleep) {
		vc->runner->stat.halt_wait_ns +=
			ktime_to_ns(cur) - ktime_to_ns(start_wait);
		/* Attribute failed poll time */
		if (vc->halt_poll_ns)
			vc->runner->stat.halt_poll_fail_ns +=
				ktime_to_ns(start_wait) -
				ktime_to_ns(start_poll);
	} else {
		/* Attribute successful poll time */
		if (vc->halt_poll_ns)
			vc->runner->stat.halt_poll_success_ns +=
				ktime_to_ns(cur) -
				ktime_to_ns(start_poll);
	}
3153 3154

	/* Adjust poll time */
3155
	if (halt_poll_ns) {
3156 3157 3158
		if (block_ns <= vc->halt_poll_ns)
			;
		/* We slept and blocked for longer than the max halt time */
3159
		else if (vc->halt_poll_ns && block_ns > halt_poll_ns)
3160 3161
			shrink_halt_poll_ns(vc);
		/* We slept and our poll time is too small */
3162 3163
		else if (vc->halt_poll_ns < halt_poll_ns &&
				block_ns < halt_poll_ns)
3164
			grow_halt_poll_ns(vc);
3165 3166
		if (vc->halt_poll_ns > halt_poll_ns)
			vc->halt_poll_ns = halt_poll_ns;
3167 3168 3169 3170
	} else
		vc->halt_poll_ns = 0;

	trace_kvmppc_vcore_wakeup(do_sleep, block_ns);
3171
}
3172

3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191
static int kvmhv_setup_mmu(struct kvm_vcpu *vcpu)
{
	int r = 0;
	struct kvm *kvm = vcpu->kvm;

	mutex_lock(&kvm->lock);
	if (!kvm->arch.mmu_ready) {
		if (!kvm_is_radix(kvm))
			r = kvmppc_hv_setup_htab_rma(vcpu);
		if (!r) {
			if (cpu_has_feature(CPU_FTR_ARCH_300))
				kvmppc_setup_partition_table(kvm);
			kvm->arch.mmu_ready = 1;
		}
	}
	mutex_unlock(&kvm->lock);
	return r;
}

3192 3193
static int kvmppc_run_vcpu(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu)
{
3194
	int n_ceded, i, r;
3195
	struct kvmppc_vcore *vc;
3196
	struct kvm_vcpu *v;
3197

3198 3199
	trace_kvmppc_run_vcpu_enter(vcpu);

3200 3201 3202
	kvm_run->exit_reason = 0;
	vcpu->arch.ret = RESUME_GUEST;
	vcpu->arch.trap = 0;
3203
	kvmppc_update_vpas(vcpu);
3204 3205 3206 3207 3208 3209

	/*
	 * Synchronize with other threads in this virtual core
	 */
	vc = vcpu->arch.vcore;
	spin_lock(&vc->lock);
3210
	vcpu->arch.ceded = 0;
3211 3212
	vcpu->arch.run_task = current;
	vcpu->arch.kvm_run = kvm_run;
3213
	vcpu->arch.stolen_logged = vcore_stolen_time(vc, mftb());
3214
	vcpu->arch.state = KVMPPC_VCPU_RUNNABLE;
3215
	vcpu->arch.busy_preempt = TB_NIL;
3216
	WRITE_ONCE(vc->runnable_threads[vcpu->arch.ptid], vcpu);
3217 3218
	++vc->n_runnable;

3219 3220 3221 3222 3223
	/*
	 * This happens the first time this is called for a vcpu.
	 * If the vcore is already running, we may be able to start
	 * this thread straight away and have it join in.
	 */
3224
	if (!signal_pending(current)) {
3225 3226
		if ((vc->vcore_state == VCORE_PIGGYBACK ||
		     vc->vcore_state == VCORE_RUNNING) &&
3227
			   !VCORE_IS_EXITING(vc)) {
3228
			kvmppc_create_dtl_entry(vcpu, vc);
3229
			kvmppc_start_thread(vcpu, vc);
3230
			trace_kvm_guest_enter(vcpu);
3231
		} else if (vc->vcore_state == VCORE_SLEEPING) {
3232
			swake_up(&vc->wq);
3233 3234
		}

3235
	}
3236

3237 3238
	while (vcpu->arch.state == KVMPPC_VCPU_RUNNABLE &&
	       !signal_pending(current)) {
3239 3240
		/* See if the MMU is ready to go */
		if (!vcpu->kvm->arch.mmu_ready) {
3241
			spin_unlock(&vc->lock);
3242
			r = kvmhv_setup_mmu(vcpu);
3243 3244 3245
			spin_lock(&vc->lock);
			if (r) {
				kvm_run->exit_reason = KVM_EXIT_FAIL_ENTRY;
3246 3247
				kvm_run->fail_entry.
					hardware_entry_failure_reason = 0;
3248 3249 3250 3251 3252
				vcpu->arch.ret = r;
				break;
			}
		}

3253 3254 3255
		if (vc->vcore_state == VCORE_PREEMPT && vc->runner == NULL)
			kvmppc_vcore_end_preempt(vc);

3256
		if (vc->vcore_state != VCORE_INACTIVE) {
3257
			kvmppc_wait_for_exec(vc, vcpu, TASK_INTERRUPTIBLE);
3258 3259
			continue;
		}
3260
		for_each_runnable_thread(i, v, vc) {
3261
			kvmppc_core_prepare_to_enter(v);
3262 3263 3264 3265 3266 3267 3268 3269
			if (signal_pending(v->arch.run_task)) {
				kvmppc_remove_runnable(vc, v);
				v->stat.signal_exits++;
				v->arch.kvm_run->exit_reason = KVM_EXIT_INTR;
				v->arch.ret = -EINTR;
				wake_up(&v->arch.cpu_run);
			}
		}
3270 3271 3272
		if (!vc->n_runnable || vcpu->arch.state != KVMPPC_VCPU_RUNNABLE)
			break;
		n_ceded = 0;
3273
		for_each_runnable_thread(i, v, vc) {
3274
			if (!kvmppc_vcpu_woken(v))
3275
				n_ceded += v->arch.ceded;
3276 3277 3278
			else
				v->arch.ceded = 0;
		}
3279 3280
		vc->runner = vcpu;
		if (n_ceded == vc->n_runnable) {
3281
			kvmppc_vcore_blocked(vc);
3282
		} else if (need_resched()) {
3283
			kvmppc_vcore_preempt(vc);
3284 3285
			/* Let something else run */
			cond_resched_lock(&vc->lock);
3286 3287
			if (vc->vcore_state == VCORE_PREEMPT)
				kvmppc_vcore_end_preempt(vc);
3288
		} else {
3289
			kvmppc_run_core(vc);
3290
		}
3291
		vc->runner = NULL;
3292
	}
3293

3294 3295
	while (vcpu->arch.state == KVMPPC_VCPU_RUNNABLE &&
	       (vc->vcore_state == VCORE_RUNNING ||
3296 3297
		vc->vcore_state == VCORE_EXITING ||
		vc->vcore_state == VCORE_PIGGYBACK))
3298
		kvmppc_wait_for_exec(vc, vcpu, TASK_UNINTERRUPTIBLE);
3299

3300 3301 3302
	if (vc->vcore_state == VCORE_PREEMPT && vc->runner == NULL)
		kvmppc_vcore_end_preempt(vc);

3303 3304 3305 3306 3307 3308 3309 3310 3311
	if (vcpu->arch.state == KVMPPC_VCPU_RUNNABLE) {
		kvmppc_remove_runnable(vc, vcpu);
		vcpu->stat.signal_exits++;
		kvm_run->exit_reason = KVM_EXIT_INTR;
		vcpu->arch.ret = -EINTR;
	}

	if (vc->n_runnable && vc->vcore_state == VCORE_INACTIVE) {
		/* Wake up some vcpu to run the core */
3312 3313
		i = -1;
		v = next_runnable_thread(vc, &i);
3314
		wake_up(&v->arch.cpu_run);
3315 3316
	}

3317
	trace_kvmppc_run_vcpu_exit(vcpu, kvm_run);
3318 3319
	spin_unlock(&vc->lock);
	return vcpu->arch.ret;
3320 3321
}

3322
static int kvmppc_vcpu_run_hv(struct kvm_run *run, struct kvm_vcpu *vcpu)
3323 3324
{
	int r;
3325
	int srcu_idx;
3326
	unsigned long ebb_regs[3] = {};	/* shut up GCC */
3327 3328
	unsigned long user_tar = 0;
	unsigned int user_vrsave;
3329
	struct kvm *kvm;
3330

3331 3332 3333 3334 3335
	if (!vcpu->arch.sane) {
		run->exit_reason = KVM_EXIT_INTERNAL_ERROR;
		return -EINVAL;
	}

3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349
	/*
	 * Don't allow entry with a suspended transaction, because
	 * the guest entry/exit code will lose it.
	 * If the guest has TM enabled, save away their TM-related SPRs
	 * (they will get restored by the TM unavailable interrupt).
	 */
#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
	if (cpu_has_feature(CPU_FTR_TM) && current->thread.regs &&
	    (current->thread.regs->msr & MSR_TM)) {
		if (MSR_TM_ACTIVE(current->thread.regs->msr)) {
			run->exit_reason = KVM_EXIT_FAIL_ENTRY;
			run->fail_entry.hardware_entry_failure_reason = 0;
			return -EINVAL;
		}
3350 3351
		/* Enable TM so we can read the TM SPRs */
		mtmsr(mfmsr() | MSR_TM);
3352 3353 3354 3355 3356 3357 3358
		current->thread.tm_tfhar = mfspr(SPRN_TFHAR);
		current->thread.tm_tfiar = mfspr(SPRN_TFIAR);
		current->thread.tm_texasr = mfspr(SPRN_TEXASR);
		current->thread.regs->msr &= ~MSR_TM;
	}
#endif

3359 3360
	kvmppc_core_prepare_to_enter(vcpu);

3361 3362 3363 3364 3365 3366
	/* No need to go into the guest when all we'll do is come back out */
	if (signal_pending(current)) {
		run->exit_reason = KVM_EXIT_INTR;
		return -EINTR;
	}

3367 3368 3369
	kvm = vcpu->kvm;
	atomic_inc(&kvm->arch.vcpus_running);
	/* Order vcpus_running vs. mmu_ready, see kvmppc_alloc_reset_hpt */
3370 3371
	smp_mb();

3372 3373
	flush_all_to_thread(current);

3374
	/* Save userspace EBB and other register values */
3375 3376 3377 3378
	if (cpu_has_feature(CPU_FTR_ARCH_207S)) {
		ebb_regs[0] = mfspr(SPRN_EBBHR);
		ebb_regs[1] = mfspr(SPRN_EBBRR);
		ebb_regs[2] = mfspr(SPRN_BESCR);
3379
		user_tar = mfspr(SPRN_TAR);
3380
	}
3381
	user_vrsave = mfspr(SPRN_VRSAVE);
3382

3383
	vcpu->arch.wqp = &vcpu->arch.vcore->wq;
3384
	vcpu->arch.pgdir = current->mm->pgd;
3385
	vcpu->arch.state = KVMPPC_VCPU_BUSY_IN_HOST;
3386

3387 3388 3389 3390 3391
	do {
		r = kvmppc_run_vcpu(run, vcpu);

		if (run->exit_reason == KVM_EXIT_PAPR_HCALL &&
		    !(vcpu->arch.shregs.msr & MSR_PR)) {
3392
			trace_kvm_hcall_enter(vcpu);
3393
			r = kvmppc_pseries_do_hcall(vcpu);
3394
			trace_kvm_hcall_exit(vcpu, r);
3395
			kvmppc_core_prepare_to_enter(vcpu);
3396
		} else if (r == RESUME_PAGE_FAULT) {
3397
			srcu_idx = srcu_read_lock(&kvm->srcu);
3398 3399
			r = kvmppc_book3s_hv_page_fault(run, vcpu,
				vcpu->arch.fault_dar, vcpu->arch.fault_dsisr);
3400
			srcu_read_unlock(&kvm->srcu, srcu_idx);
3401 3402 3403 3404 3405 3406
		} else if (r == RESUME_PASSTHROUGH) {
			if (WARN_ON(xive_enabled()))
				r = H_SUCCESS;
			else
				r = kvmppc_xics_rm_complete(vcpu, 0);
		}
3407
	} while (is_kvmppc_resume_guest(r));
3408

3409
	/* Restore userspace EBB and other register values */
3410 3411 3412 3413
	if (cpu_has_feature(CPU_FTR_ARCH_207S)) {
		mtspr(SPRN_EBBHR, ebb_regs[0]);
		mtspr(SPRN_EBBRR, ebb_regs[1]);
		mtspr(SPRN_BESCR, ebb_regs[2]);
3414 3415
		mtspr(SPRN_TAR, user_tar);
		mtspr(SPRN_FSCR, current->thread.fscr);
3416
	}
3417
	mtspr(SPRN_VRSAVE, user_vrsave);
3418

3419
	vcpu->arch.state = KVMPPC_VCPU_NOTREADY;
3420
	atomic_dec(&kvm->arch.vcpus_running);
3421 3422 3423
	return r;
}

3424
static void kvmppc_add_seg_page_size(struct kvm_ppc_one_seg_page_size **sps,
3425
				     int shift, int sllp)
3426
{
3427 3428 3429 3430
	(*sps)->page_shift = shift;
	(*sps)->slb_enc = sllp;
	(*sps)->enc[0].page_shift = shift;
	(*sps)->enc[0].pte_enc = kvmppc_pgsize_lp_encoding(shift, shift);
3431
	/*
3432
	 * Add 16MB MPSS support (may get filtered out by userspace)
3433
	 */
3434 3435 3436 3437 3438 3439
	if (shift != 24) {
		int penc = kvmppc_pgsize_lp_encoding(shift, 24);
		if (penc != -1) {
			(*sps)->enc[1].page_shift = 24;
			(*sps)->enc[1].pte_enc = penc;
		}
3440
	}
3441 3442 3443
	(*sps)++;
}

3444 3445
static int kvm_vm_ioctl_get_smmu_info_hv(struct kvm *kvm,
					 struct kvm_ppc_smmu_info *info)
3446 3447 3448
{
	struct kvm_ppc_one_seg_page_size *sps;

3449 3450 3451 3452 3453 3454 3455 3456
	/*
	 * POWER7, POWER8 and POWER9 all support 32 storage keys for data.
	 * POWER7 doesn't support keys for instruction accesses,
	 * POWER8 and POWER9 do.
	 */
	info->data_keys = 32;
	info->instr_keys = cpu_has_feature(CPU_FTR_ARCH_207S) ? 32 : 0;

3457 3458 3459
	/* POWER7, 8 and 9 all have 1T segments and 32-entry SLB */
	info->flags = KVM_PPC_PAGE_SIZES_REAL | KVM_PPC_1T_SEGMENTS;
	info->slb_size = 32;
3460 3461 3462

	/* We only support these sizes for now, and no muti-size segments */
	sps = &info->sps[0];
3463 3464 3465
	kvmppc_add_seg_page_size(&sps, 12, 0);
	kvmppc_add_seg_page_size(&sps, 16, SLB_VSID_L | SLB_VSID_LP_01);
	kvmppc_add_seg_page_size(&sps, 24, SLB_VSID_L);
3466 3467 3468 3469

	return 0;
}

3470 3471 3472
/*
 * Get (and clear) the dirty memory log for a memory slot.
 */
3473 3474
static int kvm_vm_ioctl_get_dirty_log_hv(struct kvm *kvm,
					 struct kvm_dirty_log *log)
3475
{
3476
	struct kvm_memslots *slots;
3477
	struct kvm_memory_slot *memslot;
3478
	int i, r;
3479
	unsigned long n;
3480
	unsigned long *buf, *p;
3481
	struct kvm_vcpu *vcpu;
3482 3483 3484 3485

	mutex_lock(&kvm->slots_lock);

	r = -EINVAL;
3486
	if (log->slot >= KVM_USER_MEM_SLOTS)
3487 3488
		goto out;

3489 3490
	slots = kvm_memslots(kvm);
	memslot = id_to_memslot(slots, log->slot);
3491 3492 3493 3494
	r = -ENOENT;
	if (!memslot->dirty_bitmap)
		goto out;

3495
	/*
3496 3497
	 * Use second half of bitmap area because both HPT and radix
	 * accumulate bits in the first half.
3498
	 */
3499
	n = kvm_dirty_bitmap_bytes(memslot);
3500 3501
	buf = memslot->dirty_bitmap + n / sizeof(long);
	memset(buf, 0, n);
3502

3503 3504 3505 3506
	if (kvm_is_radix(kvm))
		r = kvmppc_hv_get_dirty_log_radix(kvm, memslot, buf);
	else
		r = kvmppc_hv_get_dirty_log_hpt(kvm, memslot, buf);
3507 3508 3509
	if (r)
		goto out;

3510 3511 3512 3513 3514 3515 3516 3517 3518 3519
	/*
	 * We accumulate dirty bits in the first half of the
	 * memslot's dirty_bitmap area, for when pages are paged
	 * out or modified by the host directly.  Pick up these
	 * bits and add them to the map.
	 */
	p = memslot->dirty_bitmap;
	for (i = 0; i < n / sizeof(long); ++i)
		buf[i] |= xchg(&p[i], 0);

3520 3521 3522 3523 3524 3525 3526 3527 3528
	/* Harvest dirty bits from VPA and DTL updates */
	/* Note: we never modify the SLB shadow buffer areas */
	kvm_for_each_vcpu(i, vcpu, kvm) {
		spin_lock(&vcpu->arch.vpa_update_lock);
		kvmppc_harvest_vpa_dirty(&vcpu->arch.vpa, memslot, buf);
		kvmppc_harvest_vpa_dirty(&vcpu->arch.dtl, memslot, buf);
		spin_unlock(&vcpu->arch.vpa_update_lock);
	}

3529
	r = -EFAULT;
3530
	if (copy_to_user(log->dirty_bitmap, buf, n))
3531 3532 3533 3534 3535 3536 3537 3538
		goto out;

	r = 0;
out:
	mutex_unlock(&kvm->slots_lock);
	return r;
}

3539 3540
static void kvmppc_core_free_memslot_hv(struct kvm_memory_slot *free,
					struct kvm_memory_slot *dont)
3541 3542 3543 3544
{
	if (!dont || free->arch.rmap != dont->arch.rmap) {
		vfree(free->arch.rmap);
		free->arch.rmap = NULL;
3545
	}
3546 3547
}

3548 3549
static int kvmppc_core_create_memslot_hv(struct kvm_memory_slot *slot,
					 unsigned long npages)
3550 3551 3552 3553
{
	slot->arch.rmap = vzalloc(npages * sizeof(*slot->arch.rmap));
	if (!slot->arch.rmap)
		return -ENOMEM;
3554

3555 3556
	return 0;
}
3557

3558 3559
static int kvmppc_core_prepare_memory_region_hv(struct kvm *kvm,
					struct kvm_memory_slot *memslot,
3560
					const struct kvm_userspace_memory_region *mem)
3561
{
3562
	return 0;
3563 3564
}

3565
static void kvmppc_core_commit_memory_region_hv(struct kvm *kvm,
3566
				const struct kvm_userspace_memory_region *mem,
3567 3568
				const struct kvm_memory_slot *old,
				const struct kvm_memory_slot *new)
3569
{
3570 3571
	unsigned long npages = mem->memory_size >> PAGE_SHIFT;

3572 3573 3574 3575 3576 3577 3578 3579
	/*
	 * If we are making a new memslot, it might make
	 * some address that was previously cached as emulated
	 * MMIO be no longer emulated MMIO, so invalidate
	 * all the caches of emulated MMIO translations.
	 */
	if (npages)
		atomic64_inc(&kvm->arch.mmio_update);
3580 3581
}

3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607
/*
 * Update LPCR values in kvm->arch and in vcores.
 * Caller must hold kvm->lock.
 */
void kvmppc_update_lpcr(struct kvm *kvm, unsigned long lpcr, unsigned long mask)
{
	long int i;
	u32 cores_done = 0;

	if ((kvm->arch.lpcr & mask) == lpcr)
		return;

	kvm->arch.lpcr = (kvm->arch.lpcr & ~mask) | lpcr;

	for (i = 0; i < KVM_MAX_VCORES; ++i) {
		struct kvmppc_vcore *vc = kvm->arch.vcores[i];
		if (!vc)
			continue;
		spin_lock(&vc->lock);
		vc->lpcr = (vc->lpcr & ~mask) | lpcr;
		spin_unlock(&vc->lock);
		if (++cores_done >= kvm->arch.online_vcores)
			break;
	}
}

3608 3609 3610 3611 3612
static void kvmppc_mmu_destroy_hv(struct kvm_vcpu *vcpu)
{
	return;
}

3613
void kvmppc_setup_partition_table(struct kvm *kvm)
3614 3615 3616
{
	unsigned long dw0, dw1;

3617 3618 3619 3620 3621 3622
	if (!kvm_is_radix(kvm)) {
		/* PS field - page size for VRMA */
		dw0 = ((kvm->arch.vrma_slb_v & SLB_VSID_L) >> 1) |
			((kvm->arch.vrma_slb_v & SLB_VSID_LP) << 1);
		/* HTABSIZE and HTABORG fields */
		dw0 |= kvm->arch.sdr1;
3623

3624 3625 3626 3627 3628 3629 3630
		/* Second dword as set by userspace */
		dw1 = kvm->arch.process_table;
	} else {
		dw0 = PATB_HR | radix__get_tree_size() |
			__pa(kvm->arch.pgtable) | RADIX_PGD_INDEX_SIZE;
		dw1 = PATB_GR | kvm->arch.process_table;
	}
3631 3632 3633 3634

	mmu_partition_table_set_entry(kvm->arch.lpid, dw0, dw1);
}

3635 3636 3637 3638
/*
 * Set up HPT (hashed page table) and RMA (real-mode area).
 * Must be called with kvm->lock held.
 */
3639
static int kvmppc_hv_setup_htab_rma(struct kvm_vcpu *vcpu)
3640 3641 3642 3643 3644 3645
{
	int err = 0;
	struct kvm *kvm = vcpu->kvm;
	unsigned long hva;
	struct kvm_memory_slot *memslot;
	struct vm_area_struct *vma;
3646
	unsigned long lpcr = 0, senc;
3647
	unsigned long psize, porder;
3648
	int srcu_idx;
3649

3650
	/* Allocate hashed page table (if not done already) and reset it */
3651
	if (!kvm->arch.hpt.virt) {
3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662
		int order = KVM_DEFAULT_HPT_ORDER;
		struct kvm_hpt_info info;

		err = kvmppc_allocate_hpt(&info, order);
		/* If we get here, it means userspace didn't specify a
		 * size explicitly.  So, try successively smaller
		 * sizes if the default failed. */
		while ((err == -ENOMEM) && --order >= PPC_MIN_HPT_ORDER)
			err  = kvmppc_allocate_hpt(&info, order);

		if (err < 0) {
3663 3664 3665
			pr_err("KVM: Couldn't alloc HPT\n");
			goto out;
		}
3666 3667

		kvmppc_set_hpt(kvm, &info);
3668 3669
	}

3670
	/* Look up the memslot for guest physical address 0 */
3671
	srcu_idx = srcu_read_lock(&kvm->srcu);
3672
	memslot = gfn_to_memslot(kvm, 0);
3673

3674 3675 3676
	/* We must have some memory at 0 by now */
	err = -EINVAL;
	if (!memslot || (memslot->flags & KVM_MEMSLOT_INVALID))
3677
		goto out_srcu;
3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689

	/* Look up the VMA for the start of this memory slot */
	hva = memslot->userspace_addr;
	down_read(&current->mm->mmap_sem);
	vma = find_vma(current->mm, hva);
	if (!vma || vma->vm_start > hva || (vma->vm_flags & VM_IO))
		goto up_out;

	psize = vma_kernel_pagesize(vma);

	up_read(&current->mm->mmap_sem);

3690
	/* We can handle 4k, 64k or 16M pages in the VRMA */
3691 3692 3693 3694 3695 3696 3697
	if (psize >= 0x1000000)
		psize = 0x1000000;
	else if (psize >= 0x10000)
		psize = 0x10000;
	else
		psize = 0x1000;
	porder = __ilog2(psize);
3698

3699 3700 3701 3702 3703
	senc = slb_pgsize_encoding(psize);
	kvm->arch.vrma_slb_v = senc | SLB_VSID_B_1T |
		(VRMA_VSID << SLB_VSID_SHIFT_1T);
	/* Create HPTEs in the hash page table for the VRMA */
	kvmppc_map_vrma(vcpu, memslot, porder);
3704

3705 3706 3707 3708 3709 3710
	/* Update VRMASD field in the LPCR */
	if (!cpu_has_feature(CPU_FTR_ARCH_300)) {
		/* the -4 is to account for senc values starting at 0x10 */
		lpcr = senc << (LPCR_VRMASD_SH - 4);
		kvmppc_update_lpcr(kvm, lpcr, LPCR_VRMASD);
	}
3711

3712
	/* Order updates to kvm->arch.lpcr etc. vs. mmu_ready */
3713 3714
	smp_wmb();
	err = 0;
3715 3716
 out_srcu:
	srcu_read_unlock(&kvm->srcu, srcu_idx);
3717 3718
 out:
	return err;
3719

3720 3721
 up_out:
	up_read(&current->mm->mmap_sem);
3722
	goto out_srcu;
3723 3724
}

3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752
/* Must be called with kvm->lock held and mmu_ready = 0 and no vcpus running */
int kvmppc_switch_mmu_to_hpt(struct kvm *kvm)
{
	kvmppc_free_radix(kvm);
	kvmppc_update_lpcr(kvm, LPCR_VPM1,
			   LPCR_VPM1 | LPCR_UPRT | LPCR_GTSE | LPCR_HR);
	kvmppc_rmap_reset(kvm);
	kvm->arch.radix = 0;
	kvm->arch.process_table = 0;
	return 0;
}

/* Must be called with kvm->lock held and mmu_ready = 0 and no vcpus running */
int kvmppc_switch_mmu_to_radix(struct kvm *kvm)
{
	int err;

	err = kvmppc_init_vm_radix(kvm);
	if (err)
		return err;

	kvmppc_free_hpt(&kvm->arch.hpt);
	kvmppc_update_lpcr(kvm, LPCR_UPRT | LPCR_GTSE | LPCR_HR,
			   LPCR_VPM1 | LPCR_UPRT | LPCR_GTSE | LPCR_HR);
	kvm->arch.radix = 1;
	return 0;
}

3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786
#ifdef CONFIG_KVM_XICS
/*
 * Allocate a per-core structure for managing state about which cores are
 * running in the host versus the guest and for exchanging data between
 * real mode KVM and CPU running in the host.
 * This is only done for the first VM.
 * The allocated structure stays even if all VMs have stopped.
 * It is only freed when the kvm-hv module is unloaded.
 * It's OK for this routine to fail, we just don't support host
 * core operations like redirecting H_IPI wakeups.
 */
void kvmppc_alloc_host_rm_ops(void)
{
	struct kvmppc_host_rm_ops *ops;
	unsigned long l_ops;
	int cpu, core;
	int size;

	/* Not the first time here ? */
	if (kvmppc_host_rm_ops_hv != NULL)
		return;

	ops = kzalloc(sizeof(struct kvmppc_host_rm_ops), GFP_KERNEL);
	if (!ops)
		return;

	size = cpu_nr_cores() * sizeof(struct kvmppc_host_rm_core);
	ops->rm_core = kzalloc(size, GFP_KERNEL);

	if (!ops->rm_core) {
		kfree(ops);
		return;
	}

3787
	cpus_read_lock();
3788

3789 3790 3791 3792 3793 3794 3795 3796
	for (cpu = 0; cpu < nr_cpu_ids; cpu += threads_per_core) {
		if (!cpu_online(cpu))
			continue;

		core = cpu >> threads_shift;
		ops->rm_core[core].rm_state.in_host = 1;
	}

3797 3798
	ops->vcpu_kick = kvmppc_fast_vcpu_kick_hv;

3799 3800 3801 3802 3803 3804 3805 3806 3807 3808
	/*
	 * Make the contents of the kvmppc_host_rm_ops structure visible
	 * to other CPUs before we assign it to the global variable.
	 * Do an atomic assignment (no locks used here), but if someone
	 * beats us to it, just free our copy and return.
	 */
	smp_wmb();
	l_ops = (unsigned long) ops;

	if (cmpxchg64((unsigned long *)&kvmppc_host_rm_ops_hv, 0, l_ops)) {
3809
		cpus_read_unlock();
3810 3811
		kfree(ops->rm_core);
		kfree(ops);
3812
		return;
3813
	}
3814

3815 3816 3817 3818 3819
	cpuhp_setup_state_nocalls_cpuslocked(CPUHP_KVM_PPC_BOOK3S_PREPARE,
					     "ppc/kvm_book3s:prepare",
					     kvmppc_set_host_core,
					     kvmppc_clear_host_core);
	cpus_read_unlock();
3820 3821 3822 3823 3824
}

void kvmppc_free_host_rm_ops(void)
{
	if (kvmppc_host_rm_ops_hv) {
3825
		cpuhp_remove_state_nocalls(CPUHP_KVM_PPC_BOOK3S_PREPARE);
3826 3827 3828 3829 3830 3831 3832
		kfree(kvmppc_host_rm_ops_hv->rm_core);
		kfree(kvmppc_host_rm_ops_hv);
		kvmppc_host_rm_ops_hv = NULL;
	}
}
#endif

3833
static int kvmppc_core_init_vm_hv(struct kvm *kvm)
3834
{
3835
	unsigned long lpcr, lpid;
3836
	char buf[32];
3837
	int ret;
3838

3839 3840 3841
	/* Allocate the guest's logical partition ID */

	lpid = kvmppc_alloc_lpid();
3842
	if ((long)lpid < 0)
3843 3844
		return -ENOMEM;
	kvm->arch.lpid = lpid;
3845

3846 3847
	kvmppc_alloc_host_rm_ops();

3848 3849 3850 3851
	/*
	 * Since we don't flush the TLB when tearing down a VM,
	 * and this lpid might have previously been used,
	 * make sure we flush on each core before running the new VM.
3852 3853
	 * On POWER9, the tlbie in mmu_partition_table_set_entry()
	 * does this flush for us.
3854
	 */
3855 3856
	if (!cpu_has_feature(CPU_FTR_ARCH_300))
		cpumask_setall(&kvm->arch.need_tlb_flush);
3857

3858 3859 3860 3861
	/* Start out with the default set of hcalls enabled */
	memcpy(kvm->arch.enabled_hcalls, default_enabled_hcalls,
	       sizeof(kvm->arch.enabled_hcalls));

3862 3863
	if (!cpu_has_feature(CPU_FTR_ARCH_300))
		kvm->arch.host_sdr1 = mfspr(SPRN_SDR1);
3864

3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875
	/* Init LPCR for virtual RMA mode */
	kvm->arch.host_lpid = mfspr(SPRN_LPID);
	kvm->arch.host_lpcr = lpcr = mfspr(SPRN_LPCR);
	lpcr &= LPCR_PECE | LPCR_LPES;
	lpcr |= (4UL << LPCR_DPFD_SH) | LPCR_HDICE |
		LPCR_VPM0 | LPCR_VPM1;
	kvm->arch.vrma_slb_v = SLB_VSID_B_1T |
		(VRMA_VSID << SLB_VSID_SHIFT_1T);
	/* On POWER8 turn on online bit to enable PURR/SPURR */
	if (cpu_has_feature(CPU_FTR_ARCH_207S))
		lpcr |= LPCR_ONL;
3876 3877 3878
	/*
	 * On POWER9, VPM0 bit is reserved (VPM0=1 behaviour is assumed)
	 * Set HVICE bit to enable hypervisor virtualization interrupts.
3879 3880 3881
	 * Set HEIC to prevent OS interrupts to go to hypervisor (should
	 * be unnecessary but better safe than sorry in case we re-enable
	 * EE in HV mode with this LPCR still set)
3882 3883
	 */
	if (cpu_has_feature(CPU_FTR_ARCH_300)) {
3884
		lpcr &= ~LPCR_VPM0;
3885 3886 3887 3888 3889 3890 3891 3892
		lpcr |= LPCR_HVICE | LPCR_HEIC;

		/*
		 * If xive is enabled, we route 0x500 interrupts directly
		 * to the guest.
		 */
		if (xive_enabled())
			lpcr |= LPCR_LPES;
3893 3894
	}

3895
	/*
3896
	 * If the host uses radix, the guest starts out as radix.
3897 3898 3899
	 */
	if (radix_enabled()) {
		kvm->arch.radix = 1;
3900
		kvm->arch.mmu_ready = 1;
3901 3902 3903 3904 3905 3906 3907 3908 3909 3910
		lpcr &= ~LPCR_VPM1;
		lpcr |= LPCR_UPRT | LPCR_GTSE | LPCR_HR;
		ret = kvmppc_init_vm_radix(kvm);
		if (ret) {
			kvmppc_free_lpid(kvm->arch.lpid);
			return ret;
		}
		kvmppc_setup_partition_table(kvm);
	}

3911
	kvm->arch.lpcr = lpcr;
3912

3913 3914 3915
	/* Initialization for future HPT resizes */
	kvm->arch.resize_hpt = NULL;

3916 3917 3918 3919
	/*
	 * Work out how many sets the TLB has, for the use of
	 * the TLB invalidation loop in book3s_hv_rmhandlers.S.
	 */
3920
	if (radix_enabled())
3921 3922
		kvm->arch.tlb_sets = POWER9_TLB_SETS_RADIX;	/* 128 */
	else if (cpu_has_feature(CPU_FTR_ARCH_300))
3923 3924 3925 3926 3927 3928
		kvm->arch.tlb_sets = POWER9_TLB_SETS_HASH;	/* 256 */
	else if (cpu_has_feature(CPU_FTR_ARCH_207S))
		kvm->arch.tlb_sets = POWER8_TLB_SETS;		/* 512 */
	else
		kvm->arch.tlb_sets = POWER7_TLB_SETS;		/* 128 */

3929
	/*
3930 3931
	 * Track that we now have a HV mode VM active. This blocks secondary
	 * CPU threads from coming online.
3932 3933
	 * On POWER9, we only need to do this if the "indep_threads_mode"
	 * module parameter has been set to N.
3934
	 */
3935 3936 3937
	if (cpu_has_feature(CPU_FTR_ARCH_300))
		kvm->arch.threads_indep = indep_threads_mode;
	if (!kvm->arch.threads_indep)
3938
		kvm_hv_vm_activated();
3939

3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950
	/*
	 * Initialize smt_mode depending on processor.
	 * POWER8 and earlier have to use "strict" threading, where
	 * all vCPUs in a vcore have to run on the same (sub)core,
	 * whereas on POWER9 the threads can each run a different
	 * guest.
	 */
	if (!cpu_has_feature(CPU_FTR_ARCH_300))
		kvm->arch.smt_mode = threads_per_subcore;
	else
		kvm->arch.smt_mode = 1;
3951
	kvm->arch.emul_smt_mode = 1;
3952

3953 3954 3955 3956 3957 3958 3959 3960
	/*
	 * Create a debugfs directory for the VM
	 */
	snprintf(buf, sizeof(buf), "vm%d", current->pid);
	kvm->arch.debugfs_dir = debugfs_create_dir(buf, kvm_debugfs_dir);
	if (!IS_ERR_OR_NULL(kvm->arch.debugfs_dir))
		kvmppc_mmu_debugfs_init(kvm);

3961
	return 0;
3962 3963
}

3964 3965 3966 3967
static void kvmppc_free_vcores(struct kvm *kvm)
{
	long int i;

3968
	for (i = 0; i < KVM_MAX_VCORES; ++i)
3969 3970 3971 3972
		kfree(kvm->arch.vcores[i]);
	kvm->arch.online_vcores = 0;
}

3973
static void kvmppc_core_destroy_vm_hv(struct kvm *kvm)
3974
{
3975 3976
	debugfs_remove_recursive(kvm->arch.debugfs_dir);

3977
	if (!kvm->arch.threads_indep)
3978
		kvm_hv_vm_deactivated();
3979

3980
	kvmppc_free_vcores(kvm);
3981

3982 3983
	kvmppc_free_lpid(kvm->arch.lpid);

3984 3985 3986
	if (kvm_is_radix(kvm))
		kvmppc_free_radix(kvm);
	else
3987
		kvmppc_free_hpt(&kvm->arch.hpt);
3988 3989

	kvmppc_free_pimap(kvm);
3990 3991
}

3992 3993 3994
/* We don't need to emulate any privileged instructions or dcbz */
static int kvmppc_core_emulate_op_hv(struct kvm_run *run, struct kvm_vcpu *vcpu,
				     unsigned int inst, int *advance)
3995
{
3996
	return EMULATE_FAIL;
3997 3998
}

3999 4000
static int kvmppc_core_emulate_mtspr_hv(struct kvm_vcpu *vcpu, int sprn,
					ulong spr_val)
4001 4002 4003 4004
{
	return EMULATE_FAIL;
}

4005 4006
static int kvmppc_core_emulate_mfspr_hv(struct kvm_vcpu *vcpu, int sprn,
					ulong *spr_val)
4007 4008 4009 4010
{
	return EMULATE_FAIL;
}

4011
static int kvmppc_core_check_processor_compat_hv(void)
4012
{
4013 4014
	if (!cpu_has_feature(CPU_FTR_HVMODE) ||
	    !cpu_has_feature(CPU_FTR_ARCH_206))
4015
		return -EIO;
4016

4017
	return 0;
4018 4019
}

4020 4021 4022 4023 4024 4025 4026
#ifdef CONFIG_KVM_XICS

void kvmppc_free_pimap(struct kvm *kvm)
{
	kfree(kvm->arch.pimap);
}

4027
static struct kvmppc_passthru_irqmap *kvmppc_alloc_pimap(void)
4028 4029 4030
{
	return kzalloc(sizeof(struct kvmppc_passthru_irqmap), GFP_KERNEL);
}
4031 4032 4033 4034 4035 4036 4037

static int kvmppc_set_passthru_irq(struct kvm *kvm, int host_irq, int guest_gsi)
{
	struct irq_desc *desc;
	struct kvmppc_irq_map *irq_map;
	struct kvmppc_passthru_irqmap *pimap;
	struct irq_chip *chip;
4038
	int i, rc = 0;
4039

4040 4041 4042
	if (!kvm_irq_bypass)
		return 1;

4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062
	desc = irq_to_desc(host_irq);
	if (!desc)
		return -EIO;

	mutex_lock(&kvm->lock);

	pimap = kvm->arch.pimap;
	if (pimap == NULL) {
		/* First call, allocate structure to hold IRQ map */
		pimap = kvmppc_alloc_pimap();
		if (pimap == NULL) {
			mutex_unlock(&kvm->lock);
			return -ENOMEM;
		}
		kvm->arch.pimap = pimap;
	}

	/*
	 * For now, we only support interrupts for which the EOI operation
	 * is an OPAL call followed by a write to XIRR, since that's
4063
	 * what our real-mode EOI code does, or a XIVE interrupt
4064 4065
	 */
	chip = irq_data_get_irq_chip(&desc->irq_data);
4066
	if (!chip || !(is_pnv_opal_msi(chip) || is_xive_irq(chip))) {
4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097
		pr_warn("kvmppc_set_passthru_irq_hv: Could not assign IRQ map for (%d,%d)\n",
			host_irq, guest_gsi);
		mutex_unlock(&kvm->lock);
		return -ENOENT;
	}

	/*
	 * See if we already have an entry for this guest IRQ number.
	 * If it's mapped to a hardware IRQ number, that's an error,
	 * otherwise re-use this entry.
	 */
	for (i = 0; i < pimap->n_mapped; i++) {
		if (guest_gsi == pimap->mapped[i].v_hwirq) {
			if (pimap->mapped[i].r_hwirq) {
				mutex_unlock(&kvm->lock);
				return -EINVAL;
			}
			break;
		}
	}

	if (i == KVMPPC_PIRQ_MAPPED) {
		mutex_unlock(&kvm->lock);
		return -EAGAIN;		/* table is full */
	}

	irq_map = &pimap->mapped[i];

	irq_map->v_hwirq = guest_gsi;
	irq_map->desc = desc;

4098 4099 4100 4101 4102 4103 4104
	/*
	 * Order the above two stores before the next to serialize with
	 * the KVM real mode handler.
	 */
	smp_wmb();
	irq_map->r_hwirq = desc->irq_data.hwirq;

4105 4106 4107
	if (i == pimap->n_mapped)
		pimap->n_mapped++;

4108 4109 4110 4111 4112 4113
	if (xive_enabled())
		rc = kvmppc_xive_set_mapped(kvm, guest_gsi, desc);
	else
		kvmppc_xics_set_mapped(kvm, guest_gsi, desc->irq_data.hwirq);
	if (rc)
		irq_map->r_hwirq = 0;
4114

4115 4116 4117 4118 4119 4120 4121 4122 4123
	mutex_unlock(&kvm->lock);

	return 0;
}

static int kvmppc_clr_passthru_irq(struct kvm *kvm, int host_irq, int guest_gsi)
{
	struct irq_desc *desc;
	struct kvmppc_passthru_irqmap *pimap;
4124
	int i, rc = 0;
4125

4126 4127 4128
	if (!kvm_irq_bypass)
		return 0;

4129 4130 4131 4132 4133
	desc = irq_to_desc(host_irq);
	if (!desc)
		return -EIO;

	mutex_lock(&kvm->lock);
4134 4135
	if (!kvm->arch.pimap)
		goto unlock;
4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148

	pimap = kvm->arch.pimap;

	for (i = 0; i < pimap->n_mapped; i++) {
		if (guest_gsi == pimap->mapped[i].v_hwirq)
			break;
	}

	if (i == pimap->n_mapped) {
		mutex_unlock(&kvm->lock);
		return -ENODEV;
	}

4149 4150 4151 4152
	if (xive_enabled())
		rc = kvmppc_xive_clr_mapped(kvm, guest_gsi, pimap->mapped[i].desc);
	else
		kvmppc_xics_clr_mapped(kvm, guest_gsi, pimap->mapped[i].r_hwirq);
4153

4154
	/* invalidate the entry (what do do on error from the above ?) */
4155 4156 4157 4158 4159 4160
	pimap->mapped[i].r_hwirq = 0;

	/*
	 * We don't free this structure even when the count goes to
	 * zero. The structure is freed when we destroy the VM.
	 */
4161
 unlock:
4162
	mutex_unlock(&kvm->lock);
4163
	return rc;
4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201
}

static int kvmppc_irq_bypass_add_producer_hv(struct irq_bypass_consumer *cons,
					     struct irq_bypass_producer *prod)
{
	int ret = 0;
	struct kvm_kernel_irqfd *irqfd =
		container_of(cons, struct kvm_kernel_irqfd, consumer);

	irqfd->producer = prod;

	ret = kvmppc_set_passthru_irq(irqfd->kvm, prod->irq, irqfd->gsi);
	if (ret)
		pr_info("kvmppc_set_passthru_irq (irq %d, gsi %d) fails: %d\n",
			prod->irq, irqfd->gsi, ret);

	return ret;
}

static void kvmppc_irq_bypass_del_producer_hv(struct irq_bypass_consumer *cons,
					      struct irq_bypass_producer *prod)
{
	int ret;
	struct kvm_kernel_irqfd *irqfd =
		container_of(cons, struct kvm_kernel_irqfd, consumer);

	irqfd->producer = NULL;

	/*
	 * When producer of consumer is unregistered, we change back to
	 * default external interrupt handling mode - KVM real mode
	 * will switch back to host.
	 */
	ret = kvmppc_clr_passthru_irq(irqfd->kvm, prod->irq, irqfd->gsi);
	if (ret)
		pr_warn("kvmppc_clr_passthru_irq (irq %d, gsi %d) fails: %d\n",
			prod->irq, irqfd->gsi, ret);
}
4202 4203
#endif

4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218
static long kvm_arch_vm_ioctl_hv(struct file *filp,
				 unsigned int ioctl, unsigned long arg)
{
	struct kvm *kvm __maybe_unused = filp->private_data;
	void __user *argp = (void __user *)arg;
	long r;

	switch (ioctl) {

	case KVM_PPC_ALLOCATE_HTAB: {
		u32 htab_order;

		r = -EFAULT;
		if (get_user(htab_order, (u32 __user *)argp))
			break;
4219
		r = kvmppc_alloc_reset_hpt(kvm, htab_order);
4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235
		if (r)
			break;
		r = 0;
		break;
	}

	case KVM_PPC_GET_HTAB_FD: {
		struct kvm_get_htab_fd ghf;

		r = -EFAULT;
		if (copy_from_user(&ghf, argp, sizeof(ghf)))
			break;
		r = kvm_vm_ioctl_get_htab_fd(kvm, &ghf);
		break;
	}

4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257
	case KVM_PPC_RESIZE_HPT_PREPARE: {
		struct kvm_ppc_resize_hpt rhpt;

		r = -EFAULT;
		if (copy_from_user(&rhpt, argp, sizeof(rhpt)))
			break;

		r = kvm_vm_ioctl_resize_hpt_prepare(kvm, &rhpt);
		break;
	}

	case KVM_PPC_RESIZE_HPT_COMMIT: {
		struct kvm_ppc_resize_hpt rhpt;

		r = -EFAULT;
		if (copy_from_user(&rhpt, argp, sizeof(rhpt)))
			break;

		r = kvm_vm_ioctl_resize_hpt_commit(kvm, &rhpt);
		break;
	}

4258 4259 4260 4261 4262 4263 4264
	default:
		r = -ENOTTY;
	}

	return r;
}

4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298
/*
 * List of hcall numbers to enable by default.
 * For compatibility with old userspace, we enable by default
 * all hcalls that were implemented before the hcall-enabling
 * facility was added.  Note this list should not include H_RTAS.
 */
static unsigned int default_hcall_list[] = {
	H_REMOVE,
	H_ENTER,
	H_READ,
	H_PROTECT,
	H_BULK_REMOVE,
	H_GET_TCE,
	H_PUT_TCE,
	H_SET_DABR,
	H_SET_XDABR,
	H_CEDE,
	H_PROD,
	H_CONFER,
	H_REGISTER_VPA,
#ifdef CONFIG_KVM_XICS
	H_EOI,
	H_CPPR,
	H_IPI,
	H_IPOLL,
	H_XIRR,
	H_XIRR_X,
#endif
	0
};

static void init_default_hcalls(void)
{
	int i;
4299
	unsigned int hcall;
4300

4301 4302 4303 4304 4305
	for (i = 0; default_hcall_list[i]; ++i) {
		hcall = default_hcall_list[i];
		WARN_ON(!kvmppc_hcall_impl_hv(hcall));
		__set_bit(hcall / 4, default_enabled_hcalls);
	}
4306 4307
}

4308 4309
static int kvmhv_configure_mmu(struct kvm *kvm, struct kvm_ppc_mmuv3_cfg *cfg)
{
4310
	unsigned long lpcr;
4311
	int radix;
4312
	int err;
4313 4314 4315 4316 4317 4318 4319 4320 4321 4322

	/* If not on a POWER9, reject it */
	if (!cpu_has_feature(CPU_FTR_ARCH_300))
		return -ENODEV;

	/* If any unknown flags set, reject it */
	if (cfg->flags & ~(KVM_PPC_MMUV3_RADIX | KVM_PPC_MMUV3_GTSE))
		return -EINVAL;

	/* GR (guest radix) bit in process_table field must match */
4323
	radix = !!(cfg->flags & KVM_PPC_MMUV3_RADIX);
4324
	if (!!(cfg->process_table & PATB_GR) != radix)
4325 4326 4327 4328 4329 4330
		return -EINVAL;

	/* Process table size field must be reasonable, i.e. <= 24 */
	if ((cfg->process_table & PRTS_MASK) > 24)
		return -EINVAL;

4331 4332 4333 4334
	/* We can change a guest to/from radix now, if the host is radix */
	if (radix && !radix_enabled())
		return -EINVAL;

4335
	mutex_lock(&kvm->lock);
4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354
	if (radix != kvm_is_radix(kvm)) {
		if (kvm->arch.mmu_ready) {
			kvm->arch.mmu_ready = 0;
			/* order mmu_ready vs. vcpus_running */
			smp_mb();
			if (atomic_read(&kvm->arch.vcpus_running)) {
				kvm->arch.mmu_ready = 1;
				err = -EBUSY;
				goto out_unlock;
			}
		}
		if (radix)
			err = kvmppc_switch_mmu_to_radix(kvm);
		else
			err = kvmppc_switch_mmu_to_hpt(kvm);
		if (err)
			goto out_unlock;
	}

4355 4356 4357 4358 4359
	kvm->arch.process_table = cfg->process_table;
	kvmppc_setup_partition_table(kvm);

	lpcr = (cfg->flags & KVM_PPC_MMUV3_GTSE) ? LPCR_GTSE : 0;
	kvmppc_update_lpcr(kvm, lpcr, LPCR_GTSE);
4360
	err = 0;
4361

4362 4363 4364
 out_unlock:
	mutex_unlock(&kvm->lock);
	return err;
4365 4366
}

4367
static struct kvmppc_ops kvm_ops_hv = {
4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397
	.get_sregs = kvm_arch_vcpu_ioctl_get_sregs_hv,
	.set_sregs = kvm_arch_vcpu_ioctl_set_sregs_hv,
	.get_one_reg = kvmppc_get_one_reg_hv,
	.set_one_reg = kvmppc_set_one_reg_hv,
	.vcpu_load   = kvmppc_core_vcpu_load_hv,
	.vcpu_put    = kvmppc_core_vcpu_put_hv,
	.set_msr     = kvmppc_set_msr_hv,
	.vcpu_run    = kvmppc_vcpu_run_hv,
	.vcpu_create = kvmppc_core_vcpu_create_hv,
	.vcpu_free   = kvmppc_core_vcpu_free_hv,
	.check_requests = kvmppc_core_check_requests_hv,
	.get_dirty_log  = kvm_vm_ioctl_get_dirty_log_hv,
	.flush_memslot  = kvmppc_core_flush_memslot_hv,
	.prepare_memory_region = kvmppc_core_prepare_memory_region_hv,
	.commit_memory_region  = kvmppc_core_commit_memory_region_hv,
	.unmap_hva_range = kvm_unmap_hva_range_hv,
	.age_hva  = kvm_age_hva_hv,
	.test_age_hva = kvm_test_age_hva_hv,
	.set_spte_hva = kvm_set_spte_hva_hv,
	.mmu_destroy  = kvmppc_mmu_destroy_hv,
	.free_memslot = kvmppc_core_free_memslot_hv,
	.create_memslot = kvmppc_core_create_memslot_hv,
	.init_vm =  kvmppc_core_init_vm_hv,
	.destroy_vm = kvmppc_core_destroy_vm_hv,
	.get_smmu_info = kvm_vm_ioctl_get_smmu_info_hv,
	.emulate_op = kvmppc_core_emulate_op_hv,
	.emulate_mtspr = kvmppc_core_emulate_mtspr_hv,
	.emulate_mfspr = kvmppc_core_emulate_mfspr_hv,
	.fast_vcpu_kick = kvmppc_fast_vcpu_kick_hv,
	.arch_vm_ioctl  = kvm_arch_vm_ioctl_hv,
4398
	.hcall_implemented = kvmppc_hcall_impl_hv,
4399 4400 4401 4402
#ifdef CONFIG_KVM_XICS
	.irq_bypass_add_producer = kvmppc_irq_bypass_add_producer_hv,
	.irq_bypass_del_producer = kvmppc_irq_bypass_del_producer_hv,
#endif
4403 4404
	.configure_mmu = kvmhv_configure_mmu,
	.get_rmmu_info = kvmhv_get_rmmu_info,
4405
	.set_smt_mode = kvmhv_set_smt_mode,
4406 4407
};

4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418
static int kvm_init_subcore_bitmap(void)
{
	int i, j;
	int nr_cores = cpu_nr_cores();
	struct sibling_subcore_state *sibling_subcore_state;

	for (i = 0; i < nr_cores; i++) {
		int first_cpu = i * threads_per_core;
		int node = cpu_to_node(first_cpu);

		/* Ignore if it is already allocated. */
4419
		if (paca_ptrs[first_cpu]->sibling_subcore_state)
4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433
			continue;

		sibling_subcore_state =
			kmalloc_node(sizeof(struct sibling_subcore_state),
							GFP_KERNEL, node);
		if (!sibling_subcore_state)
			return -ENOMEM;

		memset(sibling_subcore_state, 0,
				sizeof(struct sibling_subcore_state));

		for (j = 0; j < threads_per_core; j++) {
			int cpu = first_cpu + j;

4434 4435
			paca_ptrs[cpu]->sibling_subcore_state =
						sibling_subcore_state;
4436 4437 4438 4439 4440
		}
	}
	return 0;
}

4441 4442 4443 4444 4445
static int kvmppc_radix_possible(void)
{
	return cpu_has_feature(CPU_FTR_ARCH_300) && radix_enabled();
}

4446
static int kvmppc_book3s_init_hv(void)
4447 4448
{
	int r;
4449 4450 4451 4452 4453
	/*
	 * FIXME!! Do we need to check on all cpus ?
	 */
	r = kvmppc_core_check_processor_compat_hv();
	if (r < 0)
4454
		return -ENODEV;
4455

4456 4457 4458 4459
	r = kvm_init_subcore_bitmap();
	if (r)
		return r;

4460 4461
	/*
	 * We need a way of accessing the XICS interrupt controller,
4462
	 * either directly, via paca_ptrs[cpu]->kvm_hstate.xics_phys, or
4463 4464 4465
	 * indirectly, via OPAL.
	 */
#ifdef CONFIG_SMP
4466
	if (!xive_enabled() && !local_paca->kvm_hstate.xics_phys) {
4467 4468 4469 4470 4471 4472 4473 4474 4475 4476
		struct device_node *np;

		np = of_find_compatible_node(NULL, NULL, "ibm,opal-intc");
		if (!np) {
			pr_err("KVM-HV: Cannot determine method for accessing XICS\n");
			return -ENODEV;
		}
	}
#endif

4477 4478
	kvm_ops_hv.owner = THIS_MODULE;
	kvmppc_hv_ops = &kvm_ops_hv;
4479

4480 4481
	init_default_hcalls();

4482 4483
	init_vcore_lists();

4484
	r = kvmppc_mmu_hv_init();
4485 4486 4487 4488 4489
	if (r)
		return r;

	if (kvmppc_radix_possible())
		r = kvmppc_radix_init();
4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502

	/*
	 * POWER9 chips before version 2.02 can't have some threads in
	 * HPT mode and some in radix mode on the same core.
	 */
	if (cpu_has_feature(CPU_FTR_ARCH_300)) {
		unsigned int pvr = mfspr(SPRN_PVR);
		if ((pvr >> 16) == PVR_POWER9 &&
		    (((pvr & 0xe000) == 0 && (pvr & 0xfff) < 0x202) ||
		     ((pvr & 0xe000) == 0x2000 && (pvr & 0xfff) < 0x101)))
			no_mixing_hpt_and_radix = true;
	}

4503 4504 4505
	return r;
}

4506
static void kvmppc_book3s_exit_hv(void)
4507
{
4508
	kvmppc_free_host_rm_ops();
4509 4510
	if (kvmppc_radix_possible())
		kvmppc_radix_exit();
4511
	kvmppc_hv_ops = NULL;
4512 4513
}

4514 4515
module_init(kvmppc_book3s_init_hv);
module_exit(kvmppc_book3s_exit_hv);
4516
MODULE_LICENSE("GPL");
4517 4518
MODULE_ALIAS_MISCDEV(KVM_MINOR);
MODULE_ALIAS("devname:kvm");