| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20230419093400.00000fc2.zhi.wang.linux@gmail.com>
Date: Wed, 19 Apr 2023 09:34:00 +0300
From: Zhi Wang <zhi.wang.linux@...il.com>
To: Sagi Shahar <sagis@...gle.com>
Cc: kvm@...r.kernel.org, linux-kernel@...r.kernel.org, x86@...nel.org,
Sean Christopherson <seanjc@...gle.com>,
Paolo Bonzini <pbonzini@...hat.com>,
Isaku Yamahata <isaku.yamahata@...el.com>,
Erdem Aktas <erdemaktas@...gle.com>,
David Matlack <dmatlack@...gle.com>,
Kai Huang <kai.huang@...el.com>,
Chao Peng <chao.p.peng@...ux.intel.com>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
Dave Hansen <dave.hansen@...ux.intel.com>
Subject: Re: [RFC PATCH 3/5] KVM: TDX: Add base implementation for
tdx_vm_move_enc_context_from
On Tue, 18 Apr 2023 10:47:44 -0700
Sagi Shahar <sagis@...gle.com> wrote:
> On Mon, Apr 17, 2023 at 11:28 PM Zhi Wang <zhi.wang.linux@...il.com> wrote:
> >
> > On Fri, 7 Apr 2023 20:19:19 +0000
> > Sagi Shahar <sagis@...gle.com> wrote:
> >
> > Is there any reaon that TDX doesn't need .vm_copy_enc_context_from? Or it is
> > going to be deprecated? The patch comments needs to be refined according to
> > Sean's KVM x86 maintainer book.
>
> To clarify, there are 2 types of migrations. live migration (between
> different hosts) and intra-host (between kvm instances in the same
> host) migration. This patchset deals with intra-host migration and
> doesn't add support for live migration.
>
> vm_copy_enc_context_from is currently used for setting up the
> migration helper for SEV live migration and therefore it is currently
> not needed in this patcheset.
Out of curiosity, Is this the migration helper you mentioned here also
a SEV VM?
>
> >
> > > This should mostly match the logic in sev_vm_move_enc_context_from.
> > >
> > > Signed-off-by: Sagi Shahar <sagis@...gle.com>
> > > ---
> > > arch/x86/kvm/vmx/main.c | 10 +++++++
> > > arch/x86/kvm/vmx/tdx.c | 56 ++++++++++++++++++++++++++++++++++++++
> > > arch/x86/kvm/vmx/tdx.h | 2 ++
> > > arch/x86/kvm/vmx/x86_ops.h | 5 ++++
> > > 4 files changed, 73 insertions(+)
> > >
> > > diff --git a/arch/x86/kvm/vmx/main.c b/arch/x86/kvm/vmx/main.c
> > > index 5b64fe5404958..9d5d0ac465bf6 100644
> > > --- a/arch/x86/kvm/vmx/main.c
> > > +++ b/arch/x86/kvm/vmx/main.c
> > > @@ -979,6 +979,14 @@ static int vt_vcpu_mem_enc_ioctl(struct kvm_vcpu *vcpu, void __user *argp)
> > > return tdx_vcpu_ioctl(vcpu, argp);
> > > }
> > >
> > > +static int vt_move_enc_context_from(struct kvm *kvm, unsigned int source_fd)
> > > +{
> > > + if (!is_td(kvm))
> > > + return -ENOTTY;
> > > +
> > > + return tdx_vm_move_enc_context_from(kvm, source_fd);
> > > +}
> > > +
> > > #define VMX_REQUIRED_APICV_INHIBITS \
> > > ( \
> > > BIT(APICV_INHIBIT_REASON_DISABLE)| \
> > > @@ -1141,6 +1149,8 @@ struct kvm_x86_ops vt_x86_ops __initdata = {
> > > .dev_mem_enc_ioctl = tdx_dev_ioctl,
> > > .mem_enc_ioctl = vt_mem_enc_ioctl,
> > > .vcpu_mem_enc_ioctl = vt_vcpu_mem_enc_ioctl,
> > > +
> > > + .vm_move_enc_context_from = vt_move_enc_context_from,
> > > };
> > >
> > > struct kvm_x86_init_ops vt_init_ops __initdata = {
> > > diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
> > > index 8af7e4e81c860..0999a6d827c99 100644
> > > --- a/arch/x86/kvm/vmx/tdx.c
> > > +++ b/arch/x86/kvm/vmx/tdx.c
> > > @@ -2826,3 +2826,59 @@ int __init tdx_init(void)
> > > INIT_LIST_HEAD(&per_cpu(associated_tdvcpus, cpu));
> > > return 0;
> > > }
> > > +
> > > +static __always_inline bool tdx_guest(struct kvm *kvm)
> > > +{
> > > + struct kvm_tdx *tdx_kvm = to_kvm_tdx(kvm);
> > > +
> > > + return tdx_kvm->finalized;
> > > +}
> > > +
> > > +static int tdx_migrate_from(struct kvm *dst, struct kvm *src)
> > > +{
> > > + return -EINVAL;
> > > +}
> > > +
> > > +int tdx_vm_move_enc_context_from(struct kvm *kvm, unsigned int source_fd)
> > > +{
> > > + struct kvm_tdx *dst_tdx = to_kvm_tdx(kvm);
> > > + struct file *src_kvm_file;
> > > + struct kvm_tdx *src_tdx;
> > > + struct kvm *src_kvm;
> > > + int ret;
> > > +
> > > + src_kvm_file = fget(source_fd);
> > > + if (!file_is_kvm(src_kvm_file)) {
> > > + ret = -EBADF;
> > > + goto out_fput;
> > > + }
> > > + src_kvm = src_kvm_file->private_data;
> > > + src_tdx = to_kvm_tdx(src_kvm);
> > > +
> > > + ret = pre_move_enc_context_from(kvm, src_kvm,
> > > + &dst_tdx->migration_in_progress,
> > > + &src_tdx->migration_in_progress);
> > > + if (ret)
> > > + goto out_fput;
> > > +
> > > + if (tdx_guest(kvm) || !tdx_guest(src_kvm)) {
> > > + ret = -EINVAL;
> > > + goto out_post;
> > > + }
> > > +
> > > + ret = tdx_migrate_from(kvm, src_kvm);
> > > + if (ret)
> > > + goto out_post;
> > > +
> > > + kvm_vm_dead(src_kvm);
> > > + ret = 0;
> > > +
> > > +out_post:
> > > + post_move_enc_context_from(kvm, src_kvm,
> > > + &dst_tdx->migration_in_progress,
> > > + &src_tdx->migration_in_progress);
> > > +out_fput:
> > > + if (src_kvm_file)
> > > + fput(src_kvm_file);
> > > + return ret;
> > > +}
> > > diff --git a/arch/x86/kvm/vmx/tdx.h b/arch/x86/kvm/vmx/tdx.h
> > > index 71818c5001862..21b7e710be1fd 100644
> > > --- a/arch/x86/kvm/vmx/tdx.h
> > > +++ b/arch/x86/kvm/vmx/tdx.h
> > > @@ -24,6 +24,8 @@ struct kvm_tdx {
> > > atomic_t tdh_mem_track;
> > >
> > > u64 tsc_offset;
> > > +
> > > + atomic_t migration_in_progress;
> > > };
> > >
> > > union tdx_exit_reason {
> > > diff --git a/arch/x86/kvm/vmx/x86_ops.h b/arch/x86/kvm/vmx/x86_ops.h
> > > index d049e0c72ed0c..275f5d75e9bf1 100644
> > > --- a/arch/x86/kvm/vmx/x86_ops.h
> > > +++ b/arch/x86/kvm/vmx/x86_ops.h
> > > @@ -187,6 +187,8 @@ int tdx_vcpu_ioctl(struct kvm_vcpu *vcpu, void __user *argp);
> > > void tdx_flush_tlb(struct kvm_vcpu *vcpu);
> > > int tdx_sept_tlb_remote_flush(struct kvm *kvm);
> > > void tdx_load_mmu_pgd(struct kvm_vcpu *vcpu, hpa_t root_hpa, int root_level);
> > > +
> > > +int tdx_vm_move_enc_context_from(struct kvm *kvm, unsigned int source_fd);
> > > #else
> > > static inline int tdx_init(void) { return 0; };
> > > static inline int tdx_hardware_setup(struct kvm_x86_ops *x86_ops) { return -ENOSYS; }
> > > @@ -241,6 +243,9 @@ static inline int tdx_vcpu_ioctl(struct kvm_vcpu *vcpu, void __user *argp) { ret
> > > static inline void tdx_flush_tlb(struct kvm_vcpu *vcpu) {}
> > > static inline int tdx_sept_tlb_remote_flush(struct kvm *kvm) { return 0; }
> > > static inline void tdx_load_mmu_pgd(struct kvm_vcpu *vcpu, hpa_t root_hpa, int root_level) {}
> > > +
> > > +static inline int tdx_vm_move_enc_context_from(struct kvm *kvm, u
> > > + nsigned int source_fd) { return -EOPNOTSUPP; }
> > > #endif
> > >
> > > #if defined(CONFIG_INTEL_TDX_HOST) && defined(CONFIG_KVM_SMM)
> >
Powered by blists - more mailing lists