| Message ID | 819322b8f25971f2b9933bfa4506e618508ad782.1712785629.git.isaku.yamahata@intel.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | KVM: Guest Memory Pre-Population API | expand |
On Wed, 2024-04-10 at 15:07 -0700, isaku.yamahata@intel.com wrote: > From: Isaku Yamahata <isaku.yamahata@intel.com> > > Add a new ioctl KVM_MAP_MEMORY in the KVM common code. It iterates on the > memory range and calls the arch-specific function. Add stub arch function > as a weak symbol. > > Suggested-by: Sean Christopherson <seanjc@google.com> > Signed-off-by: Isaku Yamahata <isaku.yamahata@intel.com> Reviewed-by: Rick Edgecombe <rick.p.edgecombe@intel.com> But one comment below. > --- > v2: > - Drop need_resched(). (David, Sean, Kai) > - Move cond_resched() at the end of loop. (Kai) > - Drop added check. (David) > - Use EINTR instead of ERESTART. (David, Sean) > - Fix srcu lock leak. (Kai, Sean) > - Add comment above copy_to_user(). > - Drop pointless comment. (Sean) > - Drop kvm_arch_vcpu_pre_map_memory(). (Sean) > - Don't overwrite error code. (Sean, David) > - Make the parameter in bytes, not pages. (Michael) > - Drop source member in struct kvm_memory_mapping. (Sean, Michael) > --- > include/linux/kvm_host.h | 3 +++ > include/uapi/linux/kvm.h | 9 +++++++ > virt/kvm/kvm_main.c | 54 ++++++++++++++++++++++++++++++++++++++++ > 3 files changed, 66 insertions(+) > > diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h > index 48f31dcd318a..e56a0c7e5b42 100644 > --- a/include/linux/kvm_host.h > +++ b/include/linux/kvm_host.h > @@ -2445,4 +2445,7 @@ static inline int kvm_gmem_get_pfn(struct kvm *kvm, > } > #endif /* CONFIG_KVM_PRIVATE_MEM */ > > +int kvm_arch_vcpu_map_memory(struct kvm_vcpu *vcpu, > + struct kvm_memory_mapping *mapping); > + > #endif > diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h > index 2190adbe3002..972aa9e054d3 100644 > --- a/include/uapi/linux/kvm.h > +++ b/include/uapi/linux/kvm.h > @@ -917,6 +917,7 @@ struct kvm_enable_cap { > #define KVM_CAP_MEMORY_ATTRIBUTES 233 > #define KVM_CAP_GUEST_MEMFD 234 > #define KVM_CAP_VM_TYPES 235 > +#define KVM_CAP_MAP_MEMORY 236 This can go in a later patch.
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 48f31dcd318a..e56a0c7e5b42 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -2445,4 +2445,7 @@ static inline int kvm_gmem_get_pfn(struct kvm *kvm, } #endif /* CONFIG_KVM_PRIVATE_MEM */ +int kvm_arch_vcpu_map_memory(struct kvm_vcpu *vcpu, + struct kvm_memory_mapping *mapping); + #endif diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h index 2190adbe3002..972aa9e054d3 100644 --- a/include/uapi/linux/kvm.h +++ b/include/uapi/linux/kvm.h @@ -917,6 +917,7 @@ struct kvm_enable_cap { #define KVM_CAP_MEMORY_ATTRIBUTES 233 #define KVM_CAP_GUEST_MEMFD 234 #define KVM_CAP_VM_TYPES 235 +#define KVM_CAP_MAP_MEMORY 236 struct kvm_irq_routing_irqchip { __u32 irqchip; @@ -1548,4 +1549,12 @@ struct kvm_create_guest_memfd { __u64 reserved[6]; }; +#define KVM_MAP_MEMORY _IOWR(KVMIO, 0xd5, struct kvm_memory_mapping) + +struct kvm_memory_mapping { + __u64 base_address; + __u64 size; + __u64 flags; +}; + #endif /* __LINUX_KVM_H */ diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index fb49c2a60200..f2ad9e106cdb 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4415,6 +4415,48 @@ static int kvm_vcpu_ioctl_get_stats_fd(struct kvm_vcpu *vcpu) return fd; } +__weak int kvm_arch_vcpu_map_memory(struct kvm_vcpu *vcpu, + struct kvm_memory_mapping *mapping) +{ + return -EOPNOTSUPP; +} + +static int kvm_vcpu_map_memory(struct kvm_vcpu *vcpu, + struct kvm_memory_mapping *mapping) +{ + int idx, r; + + if (mapping->flags) + return -EINVAL; + + if (!PAGE_ALIGNED(mapping->base_address) || + !PAGE_ALIGNED(mapping->size) || + mapping->base_address + mapping->size <= mapping->base_address) + return -EINVAL; + + vcpu_load(vcpu); + idx = srcu_read_lock(&vcpu->kvm->srcu); + + r = 0; + while (mapping->size) { + if (signal_pending(current)) { + r = -EINTR; + break; + } + + r = kvm_arch_vcpu_map_memory(vcpu, mapping); + if (r) + break; + + cond_resched(); + } + + srcu_read_unlock(&vcpu->kvm->srcu, idx); + vcpu_put(vcpu); + + return r; +} + static long kvm_vcpu_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg) { @@ -4616,6 +4658,18 @@ static long kvm_vcpu_ioctl(struct file *filp, r = kvm_vcpu_ioctl_get_stats_fd(vcpu); break; } + case KVM_MAP_MEMORY: { + struct kvm_memory_mapping mapping; + + r = -EFAULT; + if (copy_from_user(&mapping, argp, sizeof(mapping))) + break; + r = kvm_vcpu_map_memory(vcpu, &mapping); + /* Don't check error to tell the processed range. */ + if (copy_to_user(argp, &mapping, sizeof(mapping))) + r = -EFAULT; + break; + } default: r = kvm_arch_vcpu_ioctl(filp, ioctl, arg); }