Arm:

- Only adjust the ID registers when no irqchip has been created once
   per VM run, instead of doing it once per vcpu, as this otherwise
   triggers a pretty bad conbsistency check failure in the sysreg code.
 
 - Make sure the per-vcpu Fine Grain Traps are computed before we load
   the system registers on the HW, as we otherwise start running without
   anything set until the first preemption of the vcpu.
 
 x86:
 
 - Fix selftests failure on AMD, checking for an optimization that was not
   happening anymore.
 -----BEGIN PGP SIGNATURE-----
 
 iQFIBAABCgAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAmkcpToUHHBib256aW5p
 QHJlZGhhdC5jb20ACgkQv/vSX3jHroMG4AgAl68nLWOdc6iRq/JRiiTOEx/urykc
 /nKqUz7YlefEbOB64F/ZADvmluIOTs85/bdutCTO/VzJ8lkimyhISygZzA/y6Gav
 XgAIekZ/QhriIqqfcrMGET+ug3EnOxCAo/M8kWmBtra7EPTrejUJhKtensBd4TXv
 GTcU+yxZJF7jLE84a8CuWVbHdSfyiLYP5V6cDeMtuqvZiR5cxqyzuL+KFri4jZ72
 2jxovTRpxjOh4n759Oe+eqEhl6tgWBEfOVvDMOP7hkcFGlFpsIkKGL5zR0b+xFNr
 jAm5AGbJMY8n8qSM2s0OS0+Md/3kiyhvL121XmV0iGRRc1Ceq3SWFB5pRg==
 =5CNi
 -----END PGP SIGNATURE-----

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

Pull kvm fixes from Paolo Bonzini:
 "Arm:

   - Only adjust the ID registers when no irqchip has been created once
     per VM run, instead of doing it once per vcpu, as this otherwise
     triggers a pretty bad conbsistency check failure in the sysreg code

   - Make sure the per-vcpu Fine Grain Traps are computed before we load
     the system registers on the HW, as we otherwise start running
     without anything set until the first preemption of the vcpu

  x86:

   - Fix selftests failure on AMD, checking for an optimization that was
     not happening anymore"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: SVM: Fix redundant updates of LBR MSR intercepts
  KVM: arm64: VHE: Compute fgt traps before activating them
  KVM: arm64: Finalize ID registers only once per VM

arch/arm64/include/asm/page.h

@@ -33,8 +33,8 @@ struct folio *vma_alloc_zeroed_movable_folio(struct vm_area_struct *vma,
 						unsigned long vaddr);
 #define vma_alloc_zeroed_movable_folio vma_alloc_zeroed_movable_folio
 
-void tag_clear_highpage(struct page *to);
+bool tag_clear_highpages(struct page *to, int numpages);
-#define __HAVE_ARCH_TAG_CLEAR_HIGHPAGE
+#define __HAVE_ARCH_TAG_CLEAR_HIGHPAGES
 
 #define clear_user_page(page, vaddr, pg)	clear_page(page)
 #define copy_user_page(to, from, vaddr, pg)	copy_page(to, from)

arch/arm64/kvm/arm.c

@@ -624,6 +624,7 @@ nommu:
 	kvm_timer_vcpu_load(vcpu);
 	kvm_vgic_load(vcpu);
 	kvm_vcpu_load_debug(vcpu);
+	kvm_vcpu_load_fgt(vcpu);
 	if (has_vhe())
 		kvm_vcpu_load_vhe(vcpu);
 	kvm_arch_vcpu_load_fp(vcpu);
@@ -642,7 +643,6 @@ nommu:
 		vcpu->arch.hcr_el2 |= HCR_TWI;
 
 	vcpu_set_pauth_traps(vcpu);
-	kvm_vcpu_load_fgt(vcpu);
 
 	if (is_protected_kvm_enabled()) {
 		kvm_call_hyp_nvhe(__pkvm_vcpu_load,

arch/arm64/kvm/sys_regs.c

@@ -5609,7 +5609,11 @@ int kvm_finalize_sys_regs(struct kvm_vcpu *vcpu)
 
 	guard(mutex)(&kvm->arch.config_lock);
 
-	if (!irqchip_in_kernel(kvm)) {
+	/*
+	 * This hacks into the ID registers, so only perform it when the
+	 * first vcpu runs, or the kvm_set_vm_id_reg() helper will scream.
+	 */
+	if (!irqchip_in_kernel(kvm) && !kvm_vm_has_ran_once(kvm)) {
 		u64 val;
 
 		val = kvm_read_vm_id_reg(kvm, SYS_ID_AA64PFR0_EL1) & ~ID_AA64PFR0_EL1_GIC;

arch/arm64/mm/fault.c

@@ -967,20 +967,21 @@ struct folio *vma_alloc_zeroed_movable_folio(struct vm_area_struct *vma,
 	return vma_alloc_folio(flags, 0, vma, vaddr);
 }
 
-void tag_clear_highpage(struct page *page)
+bool tag_clear_highpages(struct page *page, int numpages)
 {
 	/*
 	 * Check if MTE is supported and fall back to clear_highpage().
 	 * get_huge_zero_folio() unconditionally passes __GFP_ZEROTAGS and
-	 * post_alloc_hook() will invoke tag_clear_highpage().
+	 * post_alloc_hook() will invoke tag_clear_highpages().
 	 */
-	if (!system_supports_mte()) {
+	if (!system_supports_mte())
-		clear_highpage(page);
+		return false;
-		return;
-	}
 
-	/* Newly allocated page, shouldn't have been tagged yet */
+	/* Newly allocated pages, shouldn't have been tagged yet */
-	WARN_ON_ONCE(!try_page_mte_tagging(page));
+	for (int i = 0; i < numpages; i++, page++) {
-	mte_zero_clear_page_tags(page_address(page));
+		WARN_ON_ONCE(!try_page_mte_tagging(page));
-	set_page_mte_tagged(page);
+		mte_zero_clear_page_tags(page_address(page));
+		set_page_mte_tagged(page);
+	}
+	return true;
 }

arch/x86/kvm/svm/svm.c

@@ -705,7 +705,11 @@ void *svm_alloc_permissions_map(unsigned long size, gfp_t gfp_mask)
 
 static void svm_recalc_lbr_msr_intercepts(struct kvm_vcpu *vcpu)
 {
-	bool intercept = !(to_svm(vcpu)->vmcb->control.virt_ext & LBR_CTL_ENABLE_MASK);
+	struct vcpu_svm *svm = to_svm(vcpu);
+	bool intercept = !(svm->vmcb->control.virt_ext & LBR_CTL_ENABLE_MASK);
+
+	if (intercept == svm->lbr_msrs_intercepted)
+		return;
 
 	svm_set_intercept_for_msr(vcpu, MSR_IA32_LASTBRANCHFROMIP, MSR_TYPE_RW, intercept);
 	svm_set_intercept_for_msr(vcpu, MSR_IA32_LASTBRANCHTOIP, MSR_TYPE_RW, intercept);
@@ -714,6 +718,8 @@ static void svm_recalc_lbr_msr_intercepts(struct kvm_vcpu *vcpu)
 
 	if (sev_es_guest(vcpu->kvm))
 		svm_set_intercept_for_msr(vcpu, MSR_IA32_DEBUGCTLMSR, MSR_TYPE_RW, intercept);
+
+	svm->lbr_msrs_intercepted = intercept;
 }
 
 void svm_vcpu_free_msrpm(void *msrpm)
@@ -1221,6 +1227,7 @@ static int svm_vcpu_create(struct kvm_vcpu *vcpu)
 	}
 
 	svm->x2avic_msrs_intercepted = true;
+	svm->lbr_msrs_intercepted = true;
 
 	svm->vmcb01.ptr = page_address(vmcb01_page);
 	svm->vmcb01.pa = __sme_set(page_to_pfn(vmcb01_page) << PAGE_SHIFT);

arch/x86/kvm/svm/svm.h

@@ -336,6 +336,7 @@ struct vcpu_svm {
 	bool guest_state_loaded;
 
 	bool x2avic_msrs_intercepted;
+	bool lbr_msrs_intercepted;
 
 	/* Guest GIF value, used when vGIF is not enabled */
 	bool guest_gif;

include/linux/highmem.h

@@ -249,10 +249,12 @@ static inline void clear_highpage_kasan_tagged(struct page *page)
 	kunmap_local(kaddr);
 }
 
-#ifndef __HAVE_ARCH_TAG_CLEAR_HIGHPAGE
+#ifndef __HAVE_ARCH_TAG_CLEAR_HIGHPAGES
 
-static inline void tag_clear_highpage(struct page *page)
+/* Return false to let people know we did not initialize the pages */
+static inline bool tag_clear_highpages(struct page *page, int numpages)
 {
+	return false;
 }
 
 #endif

mm/page_alloc.c

@@ -1822,14 +1822,9 @@ inline void post_alloc_hook(struct page *page, unsigned int order,
 	 * If memory tags should be zeroed
 	 * (which happens only when memory should be initialized as well).
 	 */
-	if (zero_tags) {
+	if (zero_tags)
-		/* Initialize both memory and memory tags. */
+		init = !tag_clear_highpages(page, 1 << order);
-		for (i = 0; i != 1 << order; ++i)
-			tag_clear_highpage(page + i);
 
-		/* Take note that memory was initialized by the loop above. */
-		init = false;
-	}
 	if (!should_skip_kasan_unpoison(gfp_flags) &&
 	    kasan_unpoison_pages(page, order, init)) {
 		/* Take note that memory was initialized by KASAN. */