kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Edgecombe, Rick P" <rick.p.edgecombe@intel.com>
To: "seanjc@google.com" <seanjc@google.com>,
	"Gao, Chao" <chao.gao@intel.com>
Cc: "Yang, Weijiang" <weijiang.yang@intel.com>,
	"mingo@redhat.com" <mingo@redhat.com>,
	"x86@kernel.org" <x86@kernel.org>, "bp@alien8.de" <bp@alien8.de>,
	"dave.hansen@linux.intel.com" <dave.hansen@linux.intel.com>,
	"hpa@zytor.com" <hpa@zytor.com>,
	"john.allen@amd.com" <john.allen@amd.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"minipli@grsecurity.net" <minipli@grsecurity.net>,
	"tglx@linutronix.de" <tglx@linutronix.de>,
	"pbonzini@redhat.com" <pbonzini@redhat.com>,
	"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"xin@zytor.com" <xin@zytor.com>,
	"mlevitsk@redhat.com" <mlevitsk@redhat.com>
Subject: Re: [PATCH v12 15/24] KVM: VMX: Emulate read and write to CET MSRs
Date: Tue, 19 Aug 2025 17:19:42 +0000	[thread overview]
Message-ID: <a06cef50bff3ac618ec4feaa501d416f9841c7a1.camel@intel.com> (raw)
In-Reply-To: <aKShs0btGwLtYlVc@google.com>

On Tue, 2025-08-19 at 09:09 -0700, Sean Christopherson wrote:
> This emulation is wrong (in no small part because the architecture sucks).  From
> the SDM:
> 
>   If the processor does not support Intel 64 architecture, these fields have only
>   32 bits; bits 63:32 of the MSRs are reserved.
> 
>   On processors that support Intel 64 architecture this value cannot represent a
>   non-canonical address.
> 
>   In protected mode, only 31:0 are loaded.
> 
> That means KVM needs to drop bits 63:32 if the vCPU doesn't have LM or if the vCPU
> isn't in 64-bit mode.  The last one is especially frustrating, because software
> can still get a 64-bit value into the MSRs while running in protected, e.g. by
> switching to 64-bit mode, doing WRMSRs, then switching back to 32-bit mode.
> 
> But, there's probably no point in actually trying to correctly emulate/virtualize
> the Protected Mode behavior, because the MSRs can be written via XRSTOR, and to
> close that hole KVM would need to trap-and-emulate XRSTOR.  No thanks.
> 
> Unless someone has a better idea, I'm inclined to take an erratum for this, i.e.
> just sweep it under the rug.

Sounds ok to me. All I could think would be something like use the CR/EFER
interceptions and just exit to userspace if (CR0.PE && !EFER.LM && CR4.CET). But
this would require some rototilling and then likely remain un-exercised. Not
sure it's worth it.

  reply	other threads:[~2025-08-19 17:20 UTC|newest]

Thread overview: 51+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-12  2:55 [PATCH v12 00/24] Enable CET Virtualization Chao Gao
2025-08-12  2:55 ` [PATCH v12 01/24] KVM: x86: Rename kvm_{g,s}et_msr()* to show that they emulate guest accesses Chao Gao
2025-09-01  7:07   ` Xiaoyao Li
2025-09-02 14:28     ` Sean Christopherson
2025-08-12  2:55 ` [PATCH v12 02/24] KVM: x86: Use double-underscore read/write MSR helpers as appropriate Chao Gao
2025-08-12  2:55 ` [PATCH v12 03/24] KVM: x86: Add kvm_msr_{read,write}() helpers Chao Gao
2025-08-12  2:55 ` [PATCH v12 04/24] KVM: x86: Manually clear MPX state only on INIT Chao Gao
2025-08-12  2:55 ` [PATCH v12 05/24] KVM: x86: Zero XSTATE components on INIT by iterating over supported features Chao Gao
2025-08-12  2:55 ` [PATCH v12 06/24] KVM: x86: Introduce KVM_{G,S}ET_ONE_REG uAPIs support Chao Gao
2025-08-19 17:37   ` Sean Christopherson
2025-08-20  8:28     ` Chao Gao
2025-08-21 13:20       ` Chao Gao
2025-08-12  2:55 ` [PATCH v12 07/24] KVM: x86: Report XSS as to-be-saved if there are supported features Chao Gao
2025-08-12  2:55 ` [PATCH v12 08/24] KVM: x86: Refresh CPUID on write to guest MSR_IA32_XSS Chao Gao
2025-08-12  2:55 ` [PATCH v12 09/24] KVM: x86: Initialize kvm_caps.supported_xss Chao Gao
2025-08-12  2:55 ` [PATCH v12 10/24] KVM: x86: Load guest FPU state when access XSAVE-managed MSRs Chao Gao
2025-08-12  2:55 ` [PATCH v12 11/24] KVM: x86: Add fault checks for guest CR4.CET setting Chao Gao
2025-08-12  2:55 ` [PATCH v12 12/24] KVM: x86: Report KVM supported CET MSRs as to-be-saved Chao Gao
2025-08-12  2:55 ` [PATCH v12 13/24] KVM: VMX: Introduce CET VMCS fields and control bits Chao Gao
2025-08-12  2:55 ` [PATCH v12 14/24] KVM: x86: Enable guest SSP read/write interface with new uAPIs Chao Gao
2025-08-12  2:55 ` [PATCH v12 15/24] KVM: VMX: Emulate read and write to CET MSRs Chao Gao
2025-08-19 16:09   ` Sean Christopherson
2025-08-19 17:19     ` Edgecombe, Rick P [this message]
2025-08-19 17:50       ` Sean Christopherson
2025-08-19 17:53     ` Xin Li
2025-08-19 18:35       ` Sean Christopherson
2025-08-20  2:32     ` Chao Gao
2025-08-20 14:12       ` Sean Christopherson
2025-08-12  2:55 ` [PATCH v12 16/24] KVM: x86: Save and reload SSP to/from SMRAM Chao Gao
2025-08-12  2:55 ` [PATCH v12 17/24] KVM: VMX: Set up interception for CET MSRs Chao Gao
2025-08-19 16:11   ` Sean Christopherson
2025-08-19 18:05     ` Xin Li
2025-08-19 18:45       ` Sean Christopherson
2025-08-20  2:10         ` Chao Gao
2025-08-12  2:55 ` [PATCH v12 18/24] KVM: VMX: Set host constant supervisor states to VMCS fields Chao Gao
2025-08-18  7:57   ` Chao Gao
2025-08-12  2:55 ` [PATCH v12 19/24] KVM: x86: Don't emulate instructions guarded by CET Chao Gao
2025-08-12  2:55 ` [PATCH v12 20/24] KVM: x86: Enable CET virtualization for VMX and advertise to userspace Chao Gao
2025-08-19 18:49   ` Sean Christopherson
2025-08-12  2:55 ` [PATCH v12 21/24] KVM: nVMX: Virtualize NO_HW_ERROR_CODE_CC for L1 event injection to L2 Chao Gao
2025-08-12  2:55 ` [PATCH v12 22/24] KVM: nVMX: Enable CET support for nested guest Chao Gao
2025-08-15 14:24   ` Chao Gao
2025-08-19 17:18   ` Sean Christopherson
2025-08-12  2:55 ` [PATCH v12 23/24] KVM: nVMX: Add consistency checks for CR0.WP and CR4.CET Chao Gao
2025-08-19 17:27   ` Sean Christopherson
2025-08-20  7:29     ` Chao Gao
2025-08-12  2:55 ` [PATCH v12 24/24] KVM: nVMX: Add consistency checks for CET states Chao Gao
2025-08-15 22:57 ` [PATCH v12 00/24] Enable CET Virtualization Edgecombe, Rick P
2025-08-18  7:53   ` Chao Gao
2025-08-18 22:25 ` Xin Li
2025-08-19 23:11 ` Sean Christopherson

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=a06cef50bff3ac618ec4feaa501d416f9841c7a1.camel@intel.com \
    --to=rick.p.edgecombe@intel.com \
    --cc=bp@alien8.de \
    --cc=chao.gao@intel.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=hpa@zytor.com \
    --cc=john.allen@amd.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=minipli@grsecurity.net \
    --cc=mlevitsk@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=seanjc@google.com \
    --cc=tglx@linutronix.de \
    --cc=weijiang.yang@intel.com \
    --cc=x86@kernel.org \
    --cc=xin@zytor.com \
    /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;
as well as URLs for NNTP newsgroup(s).