Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull KVM updates from Paolo Bonzini:
 "Small release overall.

  x86:
   - miscellaneous fixes
   - AVIC support (local APIC virtualization, AMD version)

  s390:
   - polling for interrupts after a VCPU goes to halted state is now
     enabled for s390
   - use hardware provided information about facility bits that do not
     need any hypervisor activity, and other fixes for cpu models and
     facilities
   - improve perf output
   - floating interrupt controller improvements.

  MIPS:
   - miscellaneous fixes

  PPC:
   - bugfixes only

  ARM:
   - 16K page size support
   - generic firmware probing layer for timer and GIC

  Christoffer Dall (KVM-ARM maintainer) says:
    "There are a few changes in this pull request touching things
     outside KVM, but they should all carry the necessary acks and it
     made the merge process much easier to do it this way."

  though actually the irqchip maintainers' acks didn't make it into the
  patches.  Marc Zyngier, who is both irqchip and KVM-ARM maintainer,
  later acked at http://mid.gmane.org/573351D1.4060303@arm.com ('more
  formally and for documentation purposes')"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (82 commits)
  KVM: MTRR: remove MSR 0x2f8
  KVM: x86: make hwapic_isr_update and hwapic_irr_update look the same
  svm: Manage vcpu load/unload when enable AVIC
  svm: Do not intercept CR8 when enable AVIC
  svm: Do not expose x2APIC when enable AVIC
  KVM: x86: Introducing kvm_x86_ops.apicv_post_state_restore
  svm: Add VMEXIT handlers for AVIC
  svm: Add interrupt injection via AVIC
  KVM: x86: Detect and Initialize AVIC support
  svm: Introduce new AVIC VMCB registers
  KVM: split kvm_vcpu_wake_up from kvm_vcpu_kick
  KVM: x86: Introducing kvm_x86_ops VCPU blocking/unblocking hooks
  KVM: x86: Introducing kvm_x86_ops VM init/destroy hooks
  KVM: x86: Rename kvm_apic_get_reg to kvm_lapic_get_reg
  KVM: x86: Misc LAPIC changes to expose helper functions
  KVM: shrink halt polling even more for invalid wakeups
  KVM: s390: set halt polling to 80 microseconds
  KVM: halt_polling: provide a way to qualify wakeups during poll
  KVM: PPC: Book3S HV: Re-enable XICS fast path for irqfd-generated interrupts
  kvm: Conditionally register IRQ bypass consumer
  ...

10 files changed, 122 insertions, 35 deletions

diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index 6da41fa..37b9017 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -38,7 +38,7 @@
  */
 #define KVM_NR_IRQCHIPS 1
 #define KVM_IRQCHIP_NUM_PINS 4096
-#define KVM_HALT_POLL_NS_DEFAULT 0
+#define KVM_HALT_POLL_NS_DEFAULT 80000
 
 /* s390-specific vcpu->requests bit members */
 #define KVM_REQ_ENABLE_IBS         8
@@ -247,6 +247,7 @@ struct kvm_vcpu_stat {
 	u32 exit_instruction;
 	u32 halt_successful_poll;
 	u32 halt_attempted_poll;
+	u32 halt_poll_invalid;
 	u32 halt_wakeup;
 	u32 instruction_lctl;
 	u32 instruction_lctlg;
@@ -544,10 +545,6 @@ struct kvm_vcpu_arch {
 	struct kvm_s390_local_interrupt local_int;
 	struct hrtimer    ckc_timer;
 	struct kvm_s390_pgm_info pgm;
-	union  {
-		struct cpuid	cpu_id;
-		u64		stidp_data;
-	};
 	struct gmap *gmap;
 	struct kvm_guestdbg_info_arch guestdbg;
 	unsigned long pfault_token;
@@ -605,7 +602,7 @@ struct kvm_s390_cpu_model {
 	__u64 fac_mask[S390_ARCH_FAC_LIST_SIZE_U64];
 	/* facility list requested by guest (in dma page) */
 	__u64 *fac_list;
-	struct cpuid cpu_id;
+	u64 cpuid;
 	unsigned short ibc;
 };
 
@@ -700,4 +697,6 @@ static inline void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
 static inline void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu) {}
 static inline void kvm_arch_vcpu_unblocking(struct kvm_vcpu *vcpu) {}
 
+void kvm_arch_vcpu_block_finish(struct kvm_vcpu *vcpu);
+
 #endif
diff --git a/arch/s390/include/asm/sclp.h b/arch/s390/include/asm/sclp.h
index bd7893d..e4f6f73 100644
--- a/arch/s390/include/asm/sclp.h
+++ b/arch/s390/include/asm/sclp.h
@@ -69,6 +69,7 @@ struct sclp_info {
 	unsigned int max_cores;
 	unsigned long hsa_size;
 	unsigned long facilities;
+	unsigned int hmfai;
 };
 extern struct sclp_info sclp;
 
diff --git a/arch/s390/include/asm/sigp.h b/arch/s390/include/asm/sigp.h
index ec60cf7..1c8f33f 100644
--- a/arch/s390/include/asm/sigp.h
+++ b/arch/s390/include/asm/sigp.h
@@ -27,6 +27,7 @@
 
 /* SIGP cpu status bits */
 
+#define SIGP_STATUS_INVALID_ORDER	0x00000002UL
 #define SIGP_STATUS_CHECK_STOP		0x00000010UL
 #define SIGP_STATUS_STOPPED		0x00000040UL
 #define SIGP_STATUS_EXT_CALL_PENDING	0x00000080UL
diff --git a/arch/s390/include/uapi/asm/kvm.h b/arch/s390/include/uapi/asm/kvm.h
index 347fe5a..3b8e99e 100644
--- a/arch/s390/include/uapi/asm/kvm.h
+++ b/arch/s390/include/uapi/asm/kvm.h
@@ -25,6 +25,7 @@
 #define KVM_DEV_FLIC_APF_DISABLE_WAIT	5
 #define KVM_DEV_FLIC_ADAPTER_REGISTER	6
 #define KVM_DEV_FLIC_ADAPTER_MODIFY	7
+#define KVM_DEV_FLIC_CLEAR_IO_IRQ	8
 /*
  * We can have up to 4*64k pending subchannels + 8 adapter interrupts,
  * as well as up  to ASYNC_PF_PER_VCPU*KVM_MAX_VCPUS pfault done interrupts.
diff --git a/arch/s390/include/uapi/asm/sie.h b/arch/s390/include/uapi/asm/sie.h
index 5dbaa72..8fb5d4a 100644
--- a/arch/s390/include/uapi/asm/sie.h
+++ b/arch/s390/include/uapi/asm/sie.h
@@ -16,14 +16,19 @@
 	{ 0x01, "SIGP sense" },					\
 	{ 0x02, "SIGP external call" },				\
 	{ 0x03, "SIGP emergency signal" },			\
+	{ 0x04, "SIGP start" },					\
 	{ 0x05, "SIGP stop" },					\
 	{ 0x06, "SIGP restart" },				\
 	{ 0x09, "SIGP stop and store status" },			\
 	{ 0x0b, "SIGP initial cpu reset" },			\
+	{ 0x0c, "SIGP cpu reset" },				\
 	{ 0x0d, "SIGP set prefix" },				\
 	{ 0x0e, "SIGP store status at address" },		\
 	{ 0x12, "SIGP set architecture" },			\
-	{ 0x15, "SIGP sense running" }
+	{ 0x13, "SIGP conditional emergency signal" },		\
+	{ 0x15, "SIGP sense running" },				\
+	{ 0x16, "SIGP set multithreading"},			\
+	{ 0x17, "SIGP store additional status ait address"}
 
 #define icpt_prog_codes						\
 	{ 0x0001, "Prog Operation" },				\
diff --git a/arch/s390/kvm/Kconfig b/arch/s390/kvm/Kconfig
index 5ea5af3..b190023 100644
--- a/arch/s390/kvm/Kconfig
+++ b/arch/s390/kvm/Kconfig
@@ -28,6 +28,7 @@ config KVM
 	select HAVE_KVM_IRQCHIP
 	select HAVE_KVM_IRQFD
 	select HAVE_KVM_IRQ_ROUTING
+	select HAVE_KVM_INVALID_WAKEUPS
 	select SRCU
 	select KVM_VFIO
 	---help---
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 84efc2b..5a80af7 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -977,6 +977,11 @@ no_timer:
 
 void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu)
 {
+	/*
+	 * We cannot move this into the if, as the CPU might be already
+	 * in kvm_vcpu_block without having the waitqueue set (polling)
+	 */
+	vcpu->valid_wakeup = true;
 	if (swait_active(&vcpu->wq)) {
 		/*
 		 * The vcpu gave up the cpu voluntarily, mark it as a good
@@ -2034,6 +2039,27 @@ static int modify_io_adapter(struct kvm_device *dev,
 	return ret;
 }
 
+static int clear_io_irq(struct kvm *kvm, struct kvm_device_attr *attr)
+
+{
+	const u64 isc_mask = 0xffUL << 24; /* all iscs set */
+	u32 schid;
+
+	if (attr->flags)
+		return -EINVAL;
+	if (attr->attr != sizeof(schid))
+		return -EINVAL;
+	if (copy_from_user(&schid, (void __user *) attr->addr, sizeof(schid)))
+		return -EFAULT;
+	kfree(kvm_s390_get_io_int(kvm, isc_mask, schid));
+	/*
+	 * If userspace is conforming to the architecture, we can have at most
+	 * one pending I/O interrupt per subchannel, so this is effectively a
+	 * clear all.
+	 */
+	return 0;
+}
+
 static int flic_set_attr(struct kvm_device *dev, struct kvm_device_attr *attr)
 {
 	int r = 0;
@@ -2067,6 +2093,9 @@ static int flic_set_attr(struct kvm_device *dev, struct kvm_device_attr *attr)
 	case KVM_DEV_FLIC_ADAPTER_MODIFY:
 		r = modify_io_adapter(dev, attr);
 		break;
+	case KVM_DEV_FLIC_CLEAR_IO_IRQ:
+		r = clear_io_irq(dev->kvm, attr);
+		break;
 	default:
 		r = -EINVAL;
 	}
@@ -2074,6 +2103,23 @@ static int flic_set_attr(struct kvm_device *dev, struct kvm_device_attr *attr)
 	return r;
 }
 
+static int flic_has_attr(struct kvm_device *dev,
+			     struct kvm_device_attr *attr)
+{
+	switch (attr->group) {
+	case KVM_DEV_FLIC_GET_ALL_IRQS:
+	case KVM_DEV_FLIC_ENQUEUE:
+	case KVM_DEV_FLIC_CLEAR_IRQS:
+	case KVM_DEV_FLIC_APF_ENABLE:
+	case KVM_DEV_FLIC_APF_DISABLE_WAIT:
+	case KVM_DEV_FLIC_ADAPTER_REGISTER:
+	case KVM_DEV_FLIC_ADAPTER_MODIFY:
+	case KVM_DEV_FLIC_CLEAR_IO_IRQ:
+		return 0;
+	}
+	return -ENXIO;
+}
+
 static int flic_create(struct kvm_device *dev, u32 type)
 {
 	if (!dev)
@@ -2095,6 +2141,7 @@ struct kvm_device_ops kvm_flic_ops = {
 	.name = "kvm-flic",
 	.get_attr = flic_get_attr,
 	.set_attr = flic_set_attr,
+	.has_attr = flic_has_attr,
 	.create = flic_create,
 	.destroy = flic_destroy,
 };
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 668c087..6d8ec3a 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -65,6 +65,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ "exit_instr_and_program_int", VCPU_STAT(exit_instr_and_program) },
 	{ "halt_successful_poll", VCPU_STAT(halt_successful_poll) },
 	{ "halt_attempted_poll", VCPU_STAT(halt_attempted_poll) },
+	{ "halt_poll_invalid", VCPU_STAT(halt_poll_invalid) },
 	{ "halt_wakeup", VCPU_STAT(halt_wakeup) },
 	{ "instruction_lctlg", VCPU_STAT(instruction_lctlg) },
 	{ "instruction_lctl", VCPU_STAT(instruction_lctl) },
@@ -118,9 +119,9 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 };
 
 /* upper facilities limit for kvm */
-unsigned long kvm_s390_fac_list_mask[] = {
-	0xffe6fffbfcfdfc40UL,
-	0x005e800000000000UL,
+unsigned long kvm_s390_fac_list_mask[16] = {
+	0xffe6000000000000UL,
+	0x005e000000000000UL,
 };
 
 unsigned long kvm_s390_fac_list_mask_size(void)
@@ -638,6 +639,7 @@ static int kvm_s390_get_tod(struct kvm *kvm, struct kvm_device_attr *attr)
 static int kvm_s390_set_processor(struct kvm *kvm, struct kvm_device_attr *attr)
 {
 	struct kvm_s390_vm_cpu_processor *proc;
+	u16 lowest_ibc, unblocked_ibc;
 	int ret = 0;
 
 	mutex_lock(&kvm->lock);
@@ -652,9 +654,17 @@ static int kvm_s390_set_processor(struct kvm *kvm, struct kvm_device_attr *attr)
 	}
 	if (!copy_from_user(proc, (void __user *)attr->addr,
 			    sizeof(*proc))) {
-		memcpy(&kvm->arch.model.cpu_id, &proc->cpuid,
-		       sizeof(struct cpuid));
-		kvm->arch.model.ibc = proc->ibc;
+		kvm->arch.model.cpuid = proc->cpuid;
+		lowest_ibc = sclp.ibc >> 16 & 0xfff;
+		unblocked_ibc = sclp.ibc & 0xfff;
+		if (lowest_ibc) {
+			if (proc->ibc > unblocked_ibc)
+				kvm->arch.model.ibc = unblocked_ibc;
+			else if (proc->ibc < lowest_ibc)
+				kvm->arch.model.ibc = lowest_ibc;
+			else
+				kvm->arch.model.ibc = proc->ibc;
+		}
 		memcpy(kvm->arch.model.fac_list, proc->fac_list,
 		       S390_ARCH_FAC_LIST_SIZE_BYTE);
 	} else
@@ -687,7 +697,7 @@ static int kvm_s390_get_processor(struct kvm *kvm, struct kvm_device_attr *attr)
 		ret = -ENOMEM;
 		goto out;
 	}
-	memcpy(&proc->cpuid, &kvm->arch.model.cpu_id, sizeof(struct cpuid));
+	proc->cpuid = kvm->arch.model.cpuid;
 	proc->ibc = kvm->arch.model.ibc;
 	memcpy(&proc->fac_list, kvm->arch.model.fac_list,
 	       S390_ARCH_FAC_LIST_SIZE_BYTE);
@@ -1081,10 +1091,13 @@ static void kvm_s390_set_crycb_format(struct kvm *kvm)
 		kvm->arch.crypto.crycbd |= CRYCB_FORMAT1;
 }
 
-static void kvm_s390_get_cpu_id(struct cpuid *cpu_id)
+static u64 kvm_s390_get_initial_cpuid(void)
 {
-	get_cpu_id(cpu_id);
-	cpu_id->version = 0xff;
+	struct cpuid cpuid;
+
+	get_cpu_id(&cpuid);
+	cpuid.version = 0xff;
+	return *((u64 *) &cpuid);
 }
 
 static void kvm_s390_crypto_init(struct kvm *kvm)
@@ -1175,7 +1188,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	memcpy(kvm->arch.model.fac_list, kvm->arch.model.fac_mask,
 	       S390_ARCH_FAC_LIST_SIZE_BYTE);
 
-	kvm_s390_get_cpu_id(&kvm->arch.model.cpu_id);
+	kvm->arch.model.cpuid = kvm_s390_get_initial_cpuid();
 	kvm->arch.model.ibc = sclp.ibc & 0x0fff;
 
 	kvm_s390_crypto_init(kvm);
@@ -1624,7 +1637,6 @@ static void kvm_s390_vcpu_setup_model(struct kvm_vcpu *vcpu)
 {
 	struct kvm_s390_cpu_model *model = &vcpu->kvm->arch.model;
 
-	vcpu->arch.cpu_id = model->cpu_id;
 	vcpu->arch.sie_block->ibc = model->ibc;
 	if (test_kvm_facility(vcpu->kvm, 7))
 		vcpu->arch.sie_block->fac = (u32)(u64) model->fac_list;
@@ -1645,11 +1657,14 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 
 	kvm_s390_vcpu_setup_model(vcpu);
 
-	vcpu->arch.sie_block->ecb   = 6;
+	vcpu->arch.sie_block->ecb = 0x02;
+	if (test_kvm_facility(vcpu->kvm, 9))
+		vcpu->arch.sie_block->ecb |= 0x04;
 	if (test_kvm_facility(vcpu->kvm, 50) && test_kvm_facility(vcpu->kvm, 73))
 		vcpu->arch.sie_block->ecb |= 0x10;
 
-	vcpu->arch.sie_block->ecb2  = 8;
+	if (test_kvm_facility(vcpu->kvm, 8))
+		vcpu->arch.sie_block->ecb2 |= 0x08;
 	vcpu->arch.sie_block->eca   = 0xC1002000U;
 	if (sclp.has_siif)
 		vcpu->arch.sie_block->eca |= 1;
@@ -2971,13 +2986,31 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
 	return;
 }
 
+static inline unsigned long nonhyp_mask(int i)
+{
+	unsigned int nonhyp_fai = (sclp.hmfai << i * 2) >> 30;
+
+	return 0x0000ffffffffffffUL >> (nonhyp_fai << 4);
+}
+
+void kvm_arch_vcpu_block_finish(struct kvm_vcpu *vcpu)
+{
+	vcpu->valid_wakeup = false;
+}
+
 static int __init kvm_s390_init(void)
 {
+	int i;
+
 	if (!sclp.has_sief2) {
 		pr_info("SIE not available\n");
 		return -ENODEV;
 	}
 
+	for (i = 0; i < 16; i++)
+		kvm_s390_fac_list_mask[i] |=
+			S390_lowcore.stfle_fac_list[i] & nonhyp_mask(i);
+
 	return kvm_init(NULL, sizeof(struct kvm_vcpu), 0, THIS_MODULE);
 }
 
diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c
index 0a1591d..95916fa 100644
--- a/arch/s390/kvm/priv.c
+++ b/arch/s390/kvm/priv.c
@@ -439,7 +439,7 @@ static int handle_lpswe(struct kvm_vcpu *vcpu)
 
 static int handle_stidp(struct kvm_vcpu *vcpu)
 {
-	u64 stidp_data = vcpu->arch.stidp_data;
+	u64 stidp_data = vcpu->kvm->arch.model.cpuid;
 	u64 operand2;
 	int rc;
 	ar_t ar;
@@ -670,8 +670,9 @@ static int handle_pfmf(struct kvm_vcpu *vcpu)
 	if (vcpu->run->s.regs.gprs[reg1] & PFMF_RESERVED)
 		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
 
-	/* Only provide non-quiescing support if the host supports it */
-	if (vcpu->run->s.regs.gprs[reg1] & PFMF_NQ && !test_facility(14))
+	/* Only provide non-quiescing support if enabled for the guest */
+	if (vcpu->run->s.regs.gprs[reg1] & PFMF_NQ &&
+	    !test_kvm_facility(vcpu->kvm, 14))
 		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
 
 	/* No support for conditional-SSKE */
@@ -744,7 +745,7 @@ static int handle_essa(struct kvm_vcpu *vcpu)
 {
 	/* entries expected to be 1FF */
 	int entries = (vcpu->arch.sie_block->cbrlo & ~PAGE_MASK) >> 3;
-	unsigned long *cbrlo, cbrle;
+	unsigned long *cbrlo;
 	struct gmap *gmap;
 	int i;
 
@@ -765,17 +766,9 @@ static int handle_essa(struct kvm_vcpu *vcpu)
 	vcpu->arch.sie_block->cbrlo &= PAGE_MASK;	/* reset nceo */
 	cbrlo = phys_to_virt(vcpu->arch.sie_block->cbrlo);
 	down_read(&gmap->mm->mmap_sem);
-	for (i = 0; i < entries; ++i) {
-		cbrle = cbrlo[i];
-		if (unlikely(cbrle & ~PAGE_MASK || cbrle < 2 * PAGE_SIZE))
-			/* invalid entry */
-			break;
-		/* try to free backing */
-		__gmap_zap(gmap, cbrle);
-	}
+	for (i = 0; i < entries; ++i)
+		__gmap_zap(gmap, cbrlo[i]);
 	up_read(&gmap->mm->mmap_sem);
-	if (i < entries)
-		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c
index 77c22d6..28ea0ca 100644
--- a/arch/s390/kvm/sigp.c
+++ b/arch/s390/kvm/sigp.c
@@ -240,6 +240,12 @@ static int __sigp_sense_running(struct kvm_vcpu *vcpu,
 	struct kvm_s390_local_interrupt *li;
 	int rc;
 
+	if (!test_kvm_facility(vcpu->kvm, 9)) {
+		*reg &= 0xffffffff00000000UL;
+		*reg |= SIGP_STATUS_INVALID_ORDER;
+		return SIGP_CC_STATUS_STORED;
+	}
+
 	li = &dst_vcpu->arch.local_int;
 	if (atomic_read(li->cpuflags) & CPUSTAT_RUNNING) {
 		/* running */