From: Liran Alon <liran.alon@oracle.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: "Singh, Brijesh" <brijesh.singh@amd.com>,
Sean Christopherson <sean.j.christopherson@intel.com>,
"rkrcmar@redhat.com" <rkrcmar@redhat.com>,
"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
Boris Ostrovsky <boris.ostrovsky@oracle.com>
Subject: Re: [PATCH 1/2] KVM: SVM: Fix workaround for AMD Errata 1096
Date: Tue, 16 Jul 2019 22:47:46 +0300 [thread overview]
Message-ID: <E73AA180-9D11-4F74-83B2-1EF021F0951A@oracle.com> (raw)
In-Reply-To: <3cdd12c4-c3fa-5157-1a91-69e333750152@redhat.com>
> On 16 Jul 2019, at 22:39, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> On 16/07/19 21:34, Liran Alon wrote:
>>> When this errata is hit, the CPU will be at CPL3. From hardware
>>> point-of-view the below sequence happens:
>>>
>>> 1. CPL3 guest hits reserved bit NPT fault (MMIO access)
>> Why CPU needs to be at CPL3?
>> The requirement for SMAP should be that this page is user-accessible in guest page-tables.
>> Think on a case where guest have CR4.SMAP=1 and CR4.SMEP=0.
>>
>
> If you are not at CPL3, you'd get a SMAP NPF, not a RSVD NPF.
If CR4.SMEP=0, guest vCPU can execute a user-accessible page in guest page-tables with CPL<3.
This instruction will successfully execute and can cause by the data it references any type of #NPF. Including RSVD #NPF.
When hardware DecodeAssist microcode will attempt to read guest RIP though, it will get a SMAP violation because
data read is done by microcode with CPL<3 and is accessing user-accessible page.
Therefore, I still don’t think that guest vCPU CPL matters at all. Only whether code page is mapped in guest page-tables as user-accessible or not.
-Liran
>
> Paolo
next prev parent reply other threads:[~2019-07-16 19:48 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-15 20:30 KVM: SVM: Fix workaround for AMD Errata 1096 Liran Alon
2019-07-15 20:30 ` [PATCH 1/2] " Liran Alon
2019-07-16 15:48 ` Singh, Brijesh
2019-07-16 15:56 ` Liran Alon
2019-07-16 16:07 ` Liran Alon
2019-07-16 16:10 ` Singh, Brijesh
2019-07-16 16:20 ` Liran Alon
2019-07-16 16:41 ` Sean Christopherson
2019-07-16 16:56 ` Liran Alon
2019-07-16 17:27 ` Sean Christopherson
2019-07-16 17:27 ` Paolo Bonzini
2019-07-16 17:35 ` Liran Alon
2019-07-16 19:28 ` Singh, Brijesh
2019-07-16 19:34 ` Liran Alon
2019-07-16 19:39 ` Paolo Bonzini
2019-07-16 19:45 ` Sean Christopherson
2019-07-16 19:50 ` Liran Alon
2019-07-16 19:47 ` Liran Alon [this message]
2019-07-16 19:41 ` Sean Christopherson
2019-07-16 19:52 ` Liran Alon
2019-07-16 20:02 ` Singh, Brijesh
2019-07-16 20:07 ` Sean Christopherson
2019-07-16 20:13 ` Paolo Bonzini
2019-07-16 20:09 ` Liran Alon
2019-07-16 20:27 ` Singh, Brijesh
2019-07-16 20:54 ` Sean Christopherson
2019-07-16 21:53 ` Liran Alon
2019-07-16 18:05 ` Singh, Brijesh
2019-07-16 18:06 ` Singh, Brijesh
2019-07-15 20:30 ` [PATCH 2/2] KVM: x86: Rename need_emulation_on_page_fault() to handle_no_insn_on_page_fault() Liran Alon
2019-07-16 15:48 ` Sean Christopherson
2019-07-16 16:01 ` Liran Alon
2019-07-16 16:10 ` Sean Christopherson
2019-07-16 19:33 ` Singh, Brijesh
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=E73AA180-9D11-4F74-83B2-1EF021F0951A@oracle.com \
--to=liran.alon@oracle.com \
--cc=boris.ostrovsky@oracle.com \
--cc=brijesh.singh@amd.com \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=rkrcmar@redhat.com \
--cc=sean.j.christopherson@intel.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).