All of lore.kernel.org
 help / color / mirror / Atom feed
From: bugzilla-daemon@kernel.org
To: kvm@vger.kernel.org
Subject: [Bug 216033] New: KVM VMX nested virtualization: VMXON does not check guest CR0 against IA32_VMX_CR0_FIXED0
Date: Thu, 26 May 2022 03:54:23 +0000	[thread overview]
Message-ID: <bug-216033-28872@https.bugzilla.kernel.org/> (raw)

https://bugzilla.kernel.org/show_bug.cgi?id=216033

            Bug ID: 216033
           Summary: KVM VMX nested virtualization: VMXON does not check
                    guest CR0 against IA32_VMX_CR0_FIXED0
           Product: Virtualization
           Version: unspecified
    Kernel Version: 5.17.8-200.fc35.x86_64
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: kvm
          Assignee: virtualization_kvm@kernel-bugs.osdl.org
          Reporter: ercli@ucdavis.edu
        Regression: No

Created attachment 301050
  --> https://bugzilla.kernel.org/attachment.cgi?id=301050&action=edit
Guest hypervisor to reproduce this bug (xz compressed)

CPU model I am running: Intel(R) Core(TM) i7-4510U CPU @ 2.00GHz
Host kernel version: 5.17.8-200.fc35.x86_64
Host kernel arch: x86_64
Guest: a hypervisor I wrote myself, 32-bits, compressed and attached as
c.img.xz.
QEMU command line: qemu-system-x86_64 -m 512M -smp 1 -cpu Haswell,vmx=yes
-enable-kvm -serial stdio -drive media=disk,file=c.img,index=1
The problem does not go away if using -machine kernel_irqchip=off
Since the guest is a hypervisor, -accel tcg cannot be used (TCG does not
support nested virtualization)

Actual behavior: serial port shows:

...
CR0        = 0x0000000080000015
CR0 fixed0 = 0x0000000080000021
CR0 fixed1 = 0x00000000ffffffff
VMXON succeeds

Expected behaivor: serial port shows:

...
CR0        = 0x0000000080000015
CR0 fixed0 = 0x0000000080000021
CR0 fixed1 = 0x00000000ffffffff
[00]: unhandled exception 13 (0xd), halting!
[00]: error code: 0x00000000
[00]: state dump follows...
[00] CS:EIP ...
...

Explanation:

When the guest hypervisor starts VMX using the VMXON instruction, the guest
hypervisor's CR0 is not legal. IA32_VMX_CR0_FIXED0 = 0x0000000080000021. The
0x20 bit in this MSR is 1, which indicates that the 0x20 bit in CR0 must be 1
when executing VMXON. However, my hypervisor uses CR0 = 0x80000015 (the 0x20
bit is 0).

According to SDM 29.3, if "the values of CR0 and CR4 are not supported in VMX
operation", then a general protection exception (#GP(0)) should be raised. This
happens on real hardware, but not on KVM.

The relevant code in my hypervisor is:

https://github.com/lxylxy123456/uberxmhf/blob/770bdaa7afce560b9f46348bee5a05e2c680de06/xmhf/src/xmhf-core/xmhf-runtime/xmhf-startup/lhv-vmx.c#L250

The pseudo code is

print the hypervisor's CR0 to serial port
print MSR value IA32_VMX_CR0_FIXED0 to serial port
print MSR value IA32_VMX_CR0_FIXED1 to serial port
sleep for 3 seconds
run VMXON instruction
If succeed, write "VMXON succeeds" to serial port.
If VMXON receives an exception, write exception details to serial port ("[00]:
unhandled exception...")

To fix this bug, handle_vmon() in arch/x86/kvm/vmx/nested.c needs to be
updated. The check to CR0 and CR4 against IA32_VMX_CR0_FIXED0 etc need to be
added.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

             reply	other threads:[~2022-05-26  3:54 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-26  3:54 bugzilla-daemon [this message]
2022-05-26 16:18 ` [Bug 216033] KVM VMX nested virtualization: VMXON does not check guest CR0 against IA32_VMX_CR0_FIXED0 bugzilla-daemon
2022-09-02 18:39 ` bugzilla-daemon
2022-09-02 19:00 ` bugzilla-daemon
2022-09-02 19:45 ` bugzilla-daemon
2022-09-02 20:57 ` bugzilla-daemon

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=bug-216033-28872@https.bugzilla.kernel.org/ \
    --to=bugzilla-daemon@kernel.org \
    --cc=kvm@vger.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.