From: Xiaoyao Li <xiaoyao.li@intel.com>
To: Lisa Wang <wyihan@google.com>,
Andrew Jones <ajones@ventanamicro.com>,
Ackerley Tng <ackerleytng@google.com>,
Binbin Wu <binbin.wu@linux.intel.com>,
Chao Gao <chao.gao@intel.com>,
Chenyi Qiang <chenyi.qiang@intel.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
Erdem Aktas <erdemaktas@google.com>,
Ira Weiny <ira.weiny@intel.com>,
Isaku Yamahata <isaku.yamahata@intel.com>,
Kiryl Shutsemau <kas@kernel.org>,
linux-kselftest@vger.kernel.org,
Paolo Bonzini <pbonzini@redhat.com>,
"Pratik R. Sampat" <pratikrajesh.sampat@amd.com>,
Reinette Chatre <reinette.chatre@intel.com>,
Rick Edgecombe <rick.p.edgecombe@intel.com>,
Roger Wang <runanwang@google.com>,
Ryan Afranji <afranji@google.com>, Sagi Shahar <sagis@google.com>,
Sean Christopherson <seanjc@google.com>,
Shuah Khan <shuah@kernel.org>, Oliver Upton <oupton@kernel.org>
Cc: Jeremiah McReynolds <jmcrey@google.com>,
kvm@vger.kernel.org, linux-coco@lists.linux.dev,
linux-kernel@vger.kernel.org, x86@kernel.org
Subject: Re: [PATCH v14 15/22] KVM: selftests: Call KVM_TDX_INIT_VCPU when creating a new TDX vcpu
Date: Fri, 14 Aug 2026 16:32:59 +0800 [thread overview]
Message-ID: <31ddbd64-eeb4-4008-87fb-ff57de93635d@intel.com> (raw)
In-Reply-To: <20260722-tdx-selftests-v14-15-15ad654a50db@google.com>
On 7/23/2026 7:13 AM, Lisa Wang wrote:
> From: Sagi Shahar <sagis@google.com>
>
> TDX VMs need to issue the KVM_TDX_INIT_VCPU ioctl for each vcpu after
> vcpu creation.
>
> Since the cpuids for TD are managed by the TDX module, read the values
Please use CPUIDs instead of cpuids.
> virtualized for the TD using KVM_TDX_GET_CPUID and set them in kvm using
> KVM_SET_CPUID2 so that kvm has an accurate view of the VM cpuid values.
This does two things, and KVM_SET_CPUID2 isn't a hard requirement for
KVM_TDX_INIT_VCPU. we'd better split it into two patches.
> Signed-off-by: Sagi Shahar <sagis@google.com>
> Signed-off-by: Lisa Wang <wyihan@google.com>
> ---
> .../selftests/kvm/include/x86/tdx/tdx_util.h | 21 ++++++++++++++
> tools/testing/selftests/kvm/lib/x86/processor.c | 33 ++++++++++++++++------
> 2 files changed, 46 insertions(+), 8 deletions(-)
>
> diff --git a/tools/testing/selftests/kvm/include/x86/tdx/tdx_util.h b/tools/testing/selftests/kvm/include/x86/tdx/tdx_util.h
> index f5003cbaa106..07e7f5c90329 100644
> --- a/tools/testing/selftests/kvm/include/x86/tdx/tdx_util.h
> +++ b/tools/testing/selftests/kvm/include/x86/tdx/tdx_util.h
> @@ -44,6 +44,27 @@ static inline bool is_tdx_vm(struct kvm_vm *vm)
> } \
> })
>
> +#define __tdx_vcpu_ioctl(vcpu, cmd, _flags, arg) \
> +({ \
> + union { \
> + struct kvm_tdx_cmd c; \
> + unsigned long raw; \
> + } tdx_cmd = { .c = { \
> + .id = (cmd), \
> + .flags = (u32)(_flags), \
> + .data = (u64)(arg), \
> + } }; \
> + \
> + __vcpu_ioctl(vcpu, KVM_MEMORY_ENCRYPT_OP, &tdx_cmd.raw); \
> +})
> +
> +#define tdx_vcpu_ioctl(vcpu, cmd, flags, arg) \
> +({ \
> + int ret = __tdx_vcpu_ioctl(vcpu, cmd, flags, arg); \
> + TEST_ASSERT(!ret, "%s failed, errno: %d (%s)", \
> + #cmd, errno, strerror(errno)); \
> +})
It doesn't handle the tdx_cmd.c.hw_error for vcpu ioctl, but handle it
for vm ioctl, which looks inconsistent. This at least deserves a
justification in the changelog.
> void tdx_init_vm(struct kvm_vm *vm, u64 attributes);
> void tdx_vm_setup_boot_code_region(struct kvm_vm *vm);
> void tdx_vm_setup_boot_parameters_region(struct kvm_vm *vm, u32 nr_runnable_vcpus);
> diff --git a/tools/testing/selftests/kvm/lib/x86/processor.c b/tools/testing/selftests/kvm/lib/x86/processor.c
> index 8f8bb90fec35..f6a2aa6b86a4 100644
> --- a/tools/testing/selftests/kvm/lib/x86/processor.c
> +++ b/tools/testing/selftests/kvm/lib/x86/processor.c
> @@ -838,6 +838,17 @@ gva_t kvm_allocate_vcpu_stack(struct kvm_vm *vm)
> return stack_gva;
> }
>
> +static void tdx_vcpu_init(struct kvm_vm *vm, struct kvm_vcpu *vcpu)
> +{
> + struct kvm_cpuid2 *cpuid;
> +
> + cpuid = allocate_kvm_cpuid2(MAX_NR_CPUID_ENTRIES);
> + tdx_vcpu_ioctl(vcpu, KVM_TDX_GET_CPUID, 0, cpuid);
If I remember correctly, though KVM_TDX_GET_CPUID is a vcpu ioctl, the
data returned by KVM is retrived from TDX module and TDX module only
maintains the TD scope CPUID. So the CPUID returned here is TD scope,
for per-vcpu CPUIDs, e.g., x2apicid, we need to update them accordingly.
This seems not a functional gap. I'm OK to leave it to the future, but
please leave a TODO comment for it.
> + vcpu_init_cpuid(vcpu, cpuid);
> + free(cpuid);
> + tdx_vcpu_ioctl(vcpu, KVM_TDX_INIT_VCPU, 0, NULL);
> +}
> +
> struct kvm_vcpu *vm_arch_vcpu_add(struct kvm_vm *vm, u32 vcpu_id)
> {
> struct kvm_mp_state mp_state;
> @@ -845,15 +856,21 @@ struct kvm_vcpu *vm_arch_vcpu_add(struct kvm_vm *vm, u32 vcpu_id)
> struct kvm_regs regs;
>
> vcpu = __vm_vcpu_add(vm, vcpu_id);
> - vcpu_init_cpuid(vcpu, kvm_get_supported_cpuid());
> - vcpu_init_sregs(vm, vcpu);
> - vcpu_init_xcrs(vm, vcpu);
>
> - /* Setup guest general purpose registers */
> - vcpu_regs_get(vcpu, ®s);
> - regs.rflags = regs.rflags | 0x2;
> - regs.rsp = kvm_allocate_vcpu_stack(vm);
> - vcpu_regs_set(vcpu, ®s);
> + if (is_tdx_vm(vm)) {
> + tdx_vcpu_init(vm, vcpu);
> + } else {
> + vcpu_init_cpuid(vcpu, kvm_get_supported_cpuid());
> +
> + vcpu_init_sregs(vm, vcpu);
> + vcpu_init_xcrs(vm, vcpu);
> +
> + /* Setup guest general purpose registers */
> + vcpu_regs_get(vcpu, ®s);
> + regs.rflags = regs.rflags | 0x2;
> + regs.rsp = kvm_allocate_vcpu_stack(vm);
> + vcpu_regs_set(vcpu, ®s);
> + }
>
> /* Setup the MP state */
> mp_state.mp_state = 0;
>
next prev parent reply other threads:[~2026-08-14 8:33 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-22 23:13 [PATCH v14 00/22] TDX KVM selftests Lisa Wang
2026-07-22 23:13 ` [PATCH v14 01/22] KVM: selftests: Add macros to simplify creating VM shapes for non-default types Lisa Wang
2026-07-22 23:13 ` [PATCH v14 02/22] KVM: selftests: Update kvm_init_vm_address_properties() for TDX Lisa Wang
2026-08-13 23:17 ` Edgecombe, Rick P
2026-07-22 23:13 ` [PATCH v14 03/22] KVM: selftests: Initialize the TDX VM Lisa Wang
2026-07-23 8:44 ` Xiaoyao Li
2026-08-13 23:41 ` Edgecombe, Rick P
2026-08-13 23:41 ` Edgecombe, Rick P
2026-07-22 23:13 ` [PATCH v14 04/22] KVM: selftests: TDX: Use KVM_TDX_CAPABILITIES to validate TDs' attribute configuration Lisa Wang
2026-08-13 23:45 ` Edgecombe, Rick P
2026-07-22 23:13 ` [PATCH v14 05/22] KVM: selftests: Expose segment definitions to assembly files Lisa Wang
2026-08-13 23:50 ` Edgecombe, Rick P
2026-07-22 23:13 ` [PATCH v14 06/22] tools: include: Add kbuild.h for assembly structure offsets Lisa Wang
2026-07-22 23:13 ` [PATCH v14 07/22] KVM: selftests: Introduce structures for TDX guest boot parameters Lisa Wang
2026-07-22 23:13 ` [PATCH v14 08/22] KVM: selftests: Add TDX boot code Lisa Wang
2026-07-22 23:13 ` [PATCH v14 09/22] KVM: selftests: Expose functions to get default sregs values Lisa Wang
2026-08-14 0:44 ` Edgecombe, Rick P
2026-08-14 2:36 ` Xiaoyao Li
2026-08-14 15:14 ` Edgecombe, Rick P
2026-07-22 23:13 ` [PATCH v14 10/22] KVM: selftests: Set up TDX boot code region Lisa Wang
2026-07-22 23:13 ` [PATCH v14 11/22] KVM: selftests: Set up TDX boot parameters region Lisa Wang
2026-08-11 6:32 ` Binbin Wu
2026-07-22 23:13 ` [PATCH v14 12/22] KVM: selftests: Require guest_memfd for TDX VMs Lisa Wang
2026-08-11 7:43 ` Binbin Wu
2026-08-14 7:42 ` Xiaoyao Li
2026-07-22 23:13 ` [PATCH v14 13/22] KVM: selftests: Support guest_memfd in-place conversion Lisa Wang
2026-07-22 23:13 ` [PATCH v14 14/22] KVM: selftests: Expose function to allocate vCPU stack Lisa Wang
2026-08-14 8:10 ` Xiaoyao Li
2026-07-22 23:13 ` [PATCH v14 15/22] KVM: selftests: Call KVM_TDX_INIT_VCPU when creating a new TDX vcpu Lisa Wang
2026-08-14 8:32 ` Xiaoyao Li [this message]
2026-07-22 23:13 ` [PATCH v14 16/22] KVM: selftests: Load per-vCPU guest stack in TDX boot parameters Lisa Wang
2026-08-14 8:39 ` Xiaoyao Li
2026-07-22 23:13 ` [PATCH v14 17/22] KVM: selftests: Set entry point for TDX guest code Lisa Wang
2026-08-14 8:43 ` Xiaoyao Li
2026-07-22 23:13 ` [PATCH v14 18/22] KVM: selftests: Add helpers to init TDX memory and finalize VM Lisa Wang
2026-07-22 23:13 ` [PATCH v14 19/22] KVM: selftests: Finalize TD memory as part of kvm_arch_vm_finalize_vcpus Lisa Wang
2026-07-22 23:13 ` [PATCH v14 20/22] KVM: selftests: Implement MMIO WRITE for the TDX VM Lisa Wang
2026-07-28 22:56 ` Ackerley Tng
2026-07-22 23:13 ` [PATCH v14 21/22] KVM: selftests: Add ucall support for TDX Lisa Wang
2026-07-22 23:13 ` [PATCH v14 22/22] KVM: selftests: Add TDX lifecycle test Lisa Wang
2026-08-13 22:47 ` [PATCH v14 00/22] TDX KVM selftests Edgecombe, Rick P
2026-08-13 23:05 ` Edgecombe, Rick P
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=31ddbd64-eeb4-4008-87fb-ff57de93635d@intel.com \
--to=xiaoyao.li@intel.com \
--cc=ackerleytng@google.com \
--cc=afranji@google.com \
--cc=ajones@ventanamicro.com \
--cc=binbin.wu@linux.intel.com \
--cc=chao.gao@intel.com \
--cc=chenyi.qiang@intel.com \
--cc=dave.hansen@linux.intel.com \
--cc=erdemaktas@google.com \
--cc=ira.weiny@intel.com \
--cc=isaku.yamahata@intel.com \
--cc=jmcrey@google.com \
--cc=kas@kernel.org \
--cc=kvm@vger.kernel.org \
--cc=linux-coco@lists.linux.dev \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=oupton@kernel.org \
--cc=pbonzini@redhat.com \
--cc=pratikrajesh.sampat@amd.com \
--cc=reinette.chatre@intel.com \
--cc=rick.p.edgecombe@intel.com \
--cc=runanwang@google.com \
--cc=sagis@google.com \
--cc=seanjc@google.com \
--cc=shuah@kernel.org \
--cc=wyihan@google.com \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox