From: Claudio Imbrenda Date: Wed, 27 May 2026 14:43:55 +0000 (+0200) Subject: KVM: s390: Implement KVM_PRE_FAULT_MEMORY X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=bea77e83c0b9e1d98482a0c24f9ee9eb8dd61edc;p=thirdparty%2Fkernel%2Flinux.git KVM: s390: Implement KVM_PRE_FAULT_MEMORY Implement and enable the KVM_PRE_FAULT_MEMORY ioctl for s390. Faulted-in pages will be marked as accessed, unlike x86, otherwise they will trigger a minor fault when accessed. Avoiding such faults is one of the points of KVM_PRE_FAULT_MEMORY. Reviewed-by: Steffen Eiden Signed-off-by: Claudio Imbrenda Message-ID: <20260527144358.186359-3-imbrenda@linux.ibm.com> --- diff --git a/arch/s390/kvm/Kconfig b/arch/s390/kvm/Kconfig index 5b835bc6a194..8d3ee17a1bcb 100644 --- a/arch/s390/kvm/Kconfig +++ b/arch/s390/kvm/Kconfig @@ -30,6 +30,7 @@ config KVM select KVM_VFIO select VIRT_XFER_TO_GUEST_WORK select KVM_MMU_LOCKLESS_AGING + select KVM_GENERIC_PRE_FAULT_MEMORY help Support hosting paravirtualized guest machines using the SIE virtualization capability on the mainframe. This should work diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index e09960c2e6ed..f6521f16532a 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -630,6 +630,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) case KVM_CAP_S390_USER_OPEREXEC: case KVM_CAP_S390_KEYOP: case KVM_CAP_S390_VSIE_ESAMODE: + case KVM_CAP_PRE_FAULT_MEMORY: r = 1; break; case KVM_CAP_SET_GUEST_DEBUG2: @@ -5736,6 +5737,50 @@ out: return; } +/** + * kvm_arch_vcpu_pre_fault_memory() -- pre-fault and link gmap dat tables + * @vcpu: the vcpu that shall appear to have generated the fault-in. + * @range: the range that needs to be faulted in. + * + * The first page of the given range is faulted in and the corresponding gmap + * page tables are created, as if the given vCPU had performed a read + * operation. + * If the range starts outside any memslots, an error is returned. An error is + * also returned for UCONTROL VMs, which should instead use the + * KVM_S390_VCPU_FAULT ioctl. + * + * Return: + * * %-ENOENT if the range lies outside of a memslot. + * * %-EINVAL in case of invalid state (for example if the VM is UCONTROL). + * * %-EIO if errors happen while faulting-in the page (will trigger a warning + * in the caller). + * * other error codes < 0 in case of other errors. + * * otherwise a number > 0 of bytes that have been faulted in successfully. + */ +long kvm_arch_vcpu_pre_fault_memory(struct kvm_vcpu *vcpu, struct kvm_pre_fault_memory *range) +{ + struct guest_fault f = { .gfn = gpa_to_gfn(range->gpa), }; + gpa_t end; + int rc; + + if (kvm_is_ucontrol(vcpu->kvm)) + return -EINVAL; + + rc = kvm_s390_faultin_gfn(vcpu, NULL, &f); + if (rc == PGM_ADDRESSING) + return -ENOENT; + if (rc > 0) + return -EIO; + if (rc < 0) + return rc; + + if (f.ptep) + return PAGE_SIZE; + + end = ALIGN(range->gpa + PAGE_SIZE, f.crste_region3 ? _REGION3_SIZE : HPAGE_SIZE); + return min(range->size, end - range->gpa); +} + /** * kvm_test_age_gfn() - test young * @kvm: the kvm instance