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

Pull kvm fixes from Paolo Bonzini:
 "As everyone back came back from conferences, here are the pending
  patches for Linux 6.0.

  ARM:

   - Fix for kmemleak with pKVM

  s390:

   - Fixes for VFIO with zPCI

   - smatch fix

  x86:

   - Ensure XSAVE-capable hosts always allow FP and SSE state to be
     saved and restored via KVM_{GET,SET}_XSAVE

   - Fix broken max_mmu_rmap_size stat

   - Fix compile error with old glibc that doesn't have gettid()"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: x86: Inject #UD on emulated XSETBV if XSAVES isn't enabled
  KVM: x86: Always enable legacy FP/SSE in allowed user XFEATURES
  KVM: x86: Reinstate kvm_vcpu_arch.guest_supported_xcr0
  KVM: x86/mmu: add missing update to max_mmu_rmap_size
  selftests: kvm: Fix a compile error in selftests/kvm/rseq_test.c
  KVM: s390: pci: register pci hooks without interpretation
  KVM: s390: pci: fix GAIT physical vs virtual pointers usage
  KVM: s390: Pass initialized arg even if unused
  KVM: s390: pci: fix plain integer as NULL pointer warnings
  KVM: arm64: Use kmemleak_free_part_phys() to unregister hyp_mem_base

12 files changed, 52 insertions, 27 deletions

diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index 2ff0ef62abad..917086be5c6b 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -2114,7 +2114,7 @@ static int finalize_hyp_mode(void)
 	 * at, which would end badly once inaccessible.
 	 */
 	kmemleak_free_part(__hyp_bss_start, __hyp_bss_end - __hyp_bss_start);
-	kmemleak_free_part(__va(hyp_mem_base), hyp_mem_size);
+	kmemleak_free_part_phys(hyp_mem_base, hyp_mem_size);
 	return pkvm_drop_host_privileges();
 }
 
diff --git a/arch/s390/kvm/gaccess.c b/arch/s390/kvm/gaccess.c
index 082ec5f2c3a5..0243b6e38d36 100644
--- a/arch/s390/kvm/gaccess.c
+++ b/arch/s390/kvm/gaccess.c
@@ -489,6 +489,8 @@ enum prot_type {
 	PROT_TYPE_ALC  = 2,
 	PROT_TYPE_DAT  = 3,
 	PROT_TYPE_IEP  = 4,
+	/* Dummy value for passing an initialized value when code != PGM_PROTECTION */
+	PROT_NONE,
 };
 
 static int trans_exc_ending(struct kvm_vcpu *vcpu, int code, unsigned long gva, u8 ar,
@@ -504,6 +506,10 @@ static int trans_exc_ending(struct kvm_vcpu *vcpu, int code, unsigned long gva,
 	switch (code) {
 	case PGM_PROTECTION:
 		switch (prot) {
+		case PROT_NONE:
+			/* We should never get here, acts like termination */
+			WARN_ON_ONCE(1);
+			break;
 		case PROT_TYPE_IEP:
 			tec->b61 = 1;
 			fallthrough;
@@ -968,8 +974,10 @@ static int guest_range_to_gpas(struct kvm_vcpu *vcpu, unsigned long ga, u8 ar,
 				return rc;
 		} else {
 			gpa = kvm_s390_real_to_abs(vcpu, ga);
-			if (kvm_is_error_gpa(vcpu->kvm, gpa))
+			if (kvm_is_error_gpa(vcpu->kvm, gpa)) {
 				rc = PGM_ADDRESSING;
+				prot = PROT_NONE;
+			}
 		}
 		if (rc)
 			return trans_exc(vcpu, rc, ga, ar, mode, prot);
@@ -1112,8 +1120,6 @@ int access_guest_with_key(struct kvm_vcpu *vcpu, unsigned long ga, u8 ar,
 		if (rc == PGM_PROTECTION && try_storage_prot_override)
 			rc = access_guest_page_with_key(vcpu->kvm, mode, gpas[idx],
 							data, fragment_len, PAGE_SPO_ACC);
-		if (rc == PGM_PROTECTION)
-			prot = PROT_TYPE_KEYC;
 		if (rc)
 			break;
 		len -= fragment_len;
@@ -1123,6 +1129,10 @@ int access_guest_with_key(struct kvm_vcpu *vcpu, unsigned long ga, u8 ar,
 	if (rc > 0) {
 		bool terminate = (mode == GACC_STORE) && (idx > 0);
 
+		if (rc == PGM_PROTECTION)
+			prot = PROT_TYPE_KEYC;
+		else
+			prot = PROT_NONE;
 		rc = trans_exc_ending(vcpu, rc, ga, ar, mode, prot, terminate);
 	}
 out_unlock:
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index b9c944b262c7..ab569faf0df2 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -3324,7 +3324,7 @@ static void aen_host_forward(unsigned long si)
 	if (gaite->count == 0)
 		return;
 	if (gaite->aisb != 0)
-		set_bit_inv(gaite->aisbo, (unsigned long *)gaite->aisb);
+		set_bit_inv(gaite->aisbo, phys_to_virt(gaite->aisb));
 
 	kvm = kvm_s390_pci_si_to_kvm(aift, si);
 	if (!kvm)
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index edfd4bbd0cba..b7ef0b71014d 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -505,7 +505,7 @@ int kvm_arch_init(void *opaque)
 		goto out;
 	}
 
-	if (kvm_s390_pci_interp_allowed()) {
+	if (IS_ENABLED(CONFIG_VFIO_PCI_ZDEV_KVM)) {
 		rc = kvm_s390_pci_init();
 		if (rc) {
 			pr_err("Unable to allocate AIFT for PCI\n");
@@ -527,7 +527,7 @@ out:
 void kvm_arch_exit(void)
 {
 	kvm_s390_gib_destroy();
-	if (kvm_s390_pci_interp_allowed())
+	if (IS_ENABLED(CONFIG_VFIO_PCI_ZDEV_KVM))
 		kvm_s390_pci_exit();
 	debug_unregister(kvm_s390_dbf);
 	debug_unregister(kvm_s390_dbf_uv);
diff --git a/arch/s390/kvm/pci.c b/arch/s390/kvm/pci.c
index bb8c335d17b9..c50c1645c0ae 100644
--- a/arch/s390/kvm/pci.c
+++ b/arch/s390/kvm/pci.c
@@ -58,7 +58,7 @@ static int zpci_setup_aipb(u8 nisc)
 	if (!zpci_aipb)
 		return -ENOMEM;
 
-	aift->sbv = airq_iv_create(ZPCI_NR_DEVICES, AIRQ_IV_ALLOC, 0);
+	aift->sbv = airq_iv_create(ZPCI_NR_DEVICES, AIRQ_IV_ALLOC, NULL);
 	if (!aift->sbv) {
 		rc = -ENOMEM;
 		goto free_aipb;
@@ -71,7 +71,7 @@ static int zpci_setup_aipb(u8 nisc)
 		rc = -ENOMEM;
 		goto free_sbv;
 	}
-	aift->gait = (struct zpci_gaite *)page_to_phys(page);
+	aift->gait = (struct zpci_gaite *)page_to_virt(page);
 
 	zpci_aipb->aipb.faisb = virt_to_phys(aift->sbv->vector);
 	zpci_aipb->aipb.gait = virt_to_phys(aift->gait);
@@ -373,7 +373,7 @@ static int kvm_s390_pci_aif_disable(struct zpci_dev *zdev, bool force)
 		gaite->gisc = 0;
 		gaite->aisbo = 0;
 		gaite->gisa = 0;
-		aift->kzdev[zdev->aisb] = 0;
+		aift->kzdev[zdev->aisb] = NULL;
 		/* Clear zdev info */
 		airq_iv_free_bit(aift->sbv, zdev->aisb);
 		airq_iv_release(zdev->aibv);
@@ -672,23 +672,31 @@ out:
 
 int kvm_s390_pci_init(void)
 {
+	zpci_kvm_hook.kvm_register = kvm_s390_pci_register_kvm;
+	zpci_kvm_hook.kvm_unregister = kvm_s390_pci_unregister_kvm;
+
+	if (!kvm_s390_pci_interp_allowed())
+		return 0;
+
 	aift = kzalloc(sizeof(struct zpci_aift), GFP_KERNEL);
 	if (!aift)
 		return -ENOMEM;
 
 	spin_lock_init(&aift->gait_lock);
 	mutex_init(&aift->aift_lock);
-	zpci_kvm_hook.kvm_register = kvm_s390_pci_register_kvm;
-	zpci_kvm_hook.kvm_unregister = kvm_s390_pci_unregister_kvm;
 
 	return 0;
 }
 
 void kvm_s390_pci_exit(void)
 {
-	mutex_destroy(&aift->aift_lock);
 	zpci_kvm_hook.kvm_register = NULL;
 	zpci_kvm_hook.kvm_unregister = NULL;
 
+	if (!kvm_s390_pci_interp_allowed())
+		return;
+
+	mutex_destroy(&aift->aift_lock);
+
 	kfree(aift);
 }
diff --git a/arch/s390/kvm/pci.h b/arch/s390/kvm/pci.h
index 3a3606c3a0fe..486d06ef563f 100644
--- a/arch/s390/kvm/pci.h
+++ b/arch/s390/kvm/pci.h
@@ -46,9 +46,9 @@ extern struct zpci_aift *aift;
 static inline struct kvm *kvm_s390_pci_si_to_kvm(struct zpci_aift *aift,
 						 unsigned long si)
 {
-	if (!IS_ENABLED(CONFIG_VFIO_PCI_ZDEV_KVM) || aift->kzdev == 0 ||
-	    aift->kzdev[si] == 0)
-		return 0;
+	if (!IS_ENABLED(CONFIG_VFIO_PCI_ZDEV_KVM) || !aift->kzdev ||
+	    !aift->kzdev[si])
+		return NULL;
 	return aift->kzdev[si]->kvm;
 };
 
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 2c96c43c313a..aa381ab69a19 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -729,6 +729,7 @@ struct kvm_vcpu_arch {
 	struct fpu_guest guest_fpu;
 
 	u64 xcr0;
+	u64 guest_supported_xcr0;
 
 	struct kvm_pio_request pio;
 	void *pio_data;
diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c
index 75dcf7a72605..4c1c2c06e96b 100644
--- a/arch/x86/kvm/cpuid.c
+++ b/arch/x86/kvm/cpuid.c
@@ -315,7 +315,6 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
 {
 	struct kvm_lapic *apic = vcpu->arch.apic;
 	struct kvm_cpuid_entry2 *best;
-	u64 guest_supported_xcr0;
 
 	best = kvm_find_cpuid_entry(vcpu, 1);
 	if (best && apic) {
@@ -327,10 +326,16 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
 		kvm_apic_set_version(vcpu);
 	}
 
-	guest_supported_xcr0 =
+	vcpu->arch.guest_supported_xcr0 =
 		cpuid_get_supported_xcr0(vcpu->arch.cpuid_entries, vcpu->arch.cpuid_nent);
 
-	vcpu->arch.guest_fpu.fpstate->user_xfeatures = guest_supported_xcr0;
+	/*
+	 * FP+SSE can always be saved/restored via KVM_{G,S}ET_XSAVE, even if
+	 * XSAVE/XCRO are not exposed to the guest, and even if XSAVE isn't
+	 * supported by the host.
+	 */
+	vcpu->arch.guest_fpu.fpstate->user_xfeatures = vcpu->arch.guest_supported_xcr0 |
+						       XFEATURE_MASK_FPSSE;
 
 	kvm_update_pv_runtime(vcpu);
 
diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
index d5ec3a2ed5a4..aacb28c83e43 100644
--- a/arch/x86/kvm/emulate.c
+++ b/arch/x86/kvm/emulate.c
@@ -4132,6 +4132,9 @@ static int em_xsetbv(struct x86_emulate_ctxt *ctxt)
 {
 	u32 eax, ecx, edx;
 
+	if (!(ctxt->ops->get_cr(ctxt, 4) & X86_CR4_OSXSAVE))
+		return emulate_ud(ctxt);
+
 	eax = reg_read(ctxt, VCPU_REGS_RAX);
 	edx = reg_read(ctxt, VCPU_REGS_RDX);
 	ecx = reg_read(ctxt, VCPU_REGS_RCX);
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index e418ef3ecfcb..3552e6af3684 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -1596,6 +1596,8 @@ static void __rmap_add(struct kvm *kvm,
 	rmap_head = gfn_to_rmap(gfn, sp->role.level, slot);
 	rmap_count = pte_list_add(cache, spte, rmap_head);
 
+	if (rmap_count > kvm->stat.max_mmu_rmap_size)
+		kvm->stat.max_mmu_rmap_size = rmap_count;
 	if (rmap_count > RMAP_RECYCLE_THRESHOLD) {
 		kvm_zap_all_rmap_sptes(kvm, rmap_head);
 		kvm_flush_remote_tlbs_with_address(
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 43a6a7efc6ec..b0c47b41c264 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1011,15 +1011,10 @@ void kvm_load_host_xsave_state(struct kvm_vcpu *vcpu)
 }
 EXPORT_SYMBOL_GPL(kvm_load_host_xsave_state);
 
-static inline u64 kvm_guest_supported_xcr0(struct kvm_vcpu *vcpu)
-{
-	return vcpu->arch.guest_fpu.fpstate->user_xfeatures;
-}
-
 #ifdef CONFIG_X86_64
 static inline u64 kvm_guest_supported_xfd(struct kvm_vcpu *vcpu)
 {
-	return kvm_guest_supported_xcr0(vcpu) & XFEATURE_MASK_USER_DYNAMIC;
+	return vcpu->arch.guest_supported_xcr0 & XFEATURE_MASK_USER_DYNAMIC;
 }
 #endif
 
@@ -1042,7 +1037,7 @@ static int __kvm_set_xcr(struct kvm_vcpu *vcpu, u32 index, u64 xcr)
 	 * saving.  However, xcr0 bit 0 is always set, even if the
 	 * emulated CPU does not support XSAVE (see kvm_vcpu_reset()).
 	 */
-	valid_bits = kvm_guest_supported_xcr0(vcpu) | XFEATURE_MASK_FP;
+	valid_bits = vcpu->arch.guest_supported_xcr0 | XFEATURE_MASK_FP;
 	if (xcr0 & ~valid_bits)
 		return 1;
 
@@ -1070,6 +1065,7 @@ static int __kvm_set_xcr(struct kvm_vcpu *vcpu, u32 index, u64 xcr)
 
 int kvm_emulate_xsetbv(struct kvm_vcpu *vcpu)
 {
+	/* Note, #UD due to CR4.OSXSAVE=0 has priority over the intercept. */
 	if (static_call(kvm_x86_get_cpl)(vcpu) != 0 ||
 	    __kvm_set_xcr(vcpu, kvm_rcx_read(vcpu), kvm_read_edx_eax(vcpu))) {
 		kvm_inject_gp(vcpu, 0);
diff --git a/tools/testing/selftests/kvm/rseq_test.c b/tools/testing/selftests/kvm/rseq_test.c
index fac248a43666..6f88da7e60be 100644
--- a/tools/testing/selftests/kvm/rseq_test.c
+++ b/tools/testing/selftests/kvm/rseq_test.c
@@ -227,7 +227,7 @@ int main(int argc, char *argv[])
 	ucall_init(vm, NULL);
 
 	pthread_create(&migration_thread, NULL, migration_worker,
-		       (void *)(unsigned long)gettid());
+		       (void *)(unsigned long)syscall(SYS_gettid));
 
 	for (i = 0; !done; i++) {
 		vcpu_run(vcpu);