From: Vitaly Kuznetsov <vkuznets@redhat.com>
To: Sean Christopherson <sean.j.christopherson@intel.com>
Cc: kvm@vger.kernel.org, Paolo Bonzini <pbonzini@redhat.com>,
Wanpeng Li <wanpengli@tencent.com>,
Jim Mattson <jmattson@google.com>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH] KVM: nVMX: properly pad struct kvm_vmx_nested_state_hdr
Date: Mon, 13 Jul 2020 17:54:52 +0200 [thread overview]
Message-ID: <878sfntnoz.fsf@vitty.brq.redhat.com> (raw)
In-Reply-To: <20200713151750.GA29901@linux.intel.com>
Sean Christopherson <sean.j.christopherson@intel.com> writes:
> On Mon, Jul 13, 2020 at 10:28:24AM +0200, Vitaly Kuznetsov wrote:
>> Holes in structs which are userspace ABI are undesireable.
>>
>> Fixes: 83d31e5271ac ("KVM: nVMX: fixes for preemption timer migration")
>> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
>> ---
>> Documentation/virt/kvm/api.rst | 2 +-
>> arch/x86/include/uapi/asm/kvm.h | 2 +-
>> 2 files changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst
>> index 320788f81a05..7beccda11978 100644
>> --- a/Documentation/virt/kvm/api.rst
>> +++ b/Documentation/virt/kvm/api.rst
>> @@ -4345,7 +4345,7 @@ Errors:
>> struct {
>> __u16 flags;
>> } smm;
>> -
>> + __u16 pad;
>
> I don't think this is sufficient. Before 83d31e5271ac, the struct was:
>
Before 850448f35aaf. Thanks, I was too lazy to check that.
> struct kvm_vmx_nested_state_hdr {
> __u64 vmxon_pa;
> __u64 vmcs12_pa;
>
> struct {
> __u16 flags;
> } smm;
> };
>
> which most/all compilers will pad out to 24 bytes on a 64-bit system. And
> although smm.flags is padded to 8 bytes, it's initialized as a 2 byte value.
>
> 714 struct kvm_vmx_nested_state_hdr boo;
> 715 u64 val;
> 716
> 717 BUILD_BUG_ON(sizeof(boo) != 3*8);
> 718 boo.smm.flags = 0;
> 0xffffffff810148a9 <+41>: xor %eax,%eax
> 0xffffffff810148ab <+43>: mov %ax,0x18(%rsp)
>
> 719
> 720 val = *((volatile u64 *)(&boo.smm.flags));
> 0xffffffff810148b0 <+48>: mov 0x18(%rsp),%rax
>
>
> Which means that userspace built for the old kernel will potentially send in
> garbage for the new 'flags' field due to it being uninitialized stack data,
> even with the layout after this patch.
>
> struct kvm_vmx_nested_state_hdr {
> __u64 vmxon_pa;
> __u64 vmcs12_pa;
>
> struct {
> __u16 flags;
> } smm;
> __u16 pad;
> __u32 flags;
> __u64 preemption_timer_deadline;
> };
>
> So to be backwards compatible I believe we need to add a __u32 pad as well,
> and to not cause internal padding issues, either make the new 'flags' a
> __u64 or pad that as well (or add and check a reserved __32). Making flags
> a __64 seems like the least wasteful approach, e.g.
>
> struct kvm_vmx_nested_state_hdr {
> __u64 vmxon_pa;
> __u64 vmcs12_pa;
>
> struct {
> __u16 flags;
> } smm;
> __u16 pad16;
> __u32 pad32;
> __u64 flags;
> __u64 preemption_timer_deadline;
> };
I see and I agree but the fix like that needs to get into 5.8 or an ABI
breakage is guaranteed. I'll send v2 immediately, hope Paolo will take a
look.
>
>
>> __u32 flags;
>> __u64 preemption_timer_deadline;
>> };
>> diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h
>> index 0780f97c1850..aae3df1cbd01 100644
>> --- a/arch/x86/include/uapi/asm/kvm.h
>> +++ b/arch/x86/include/uapi/asm/kvm.h
>> @@ -414,7 +414,7 @@ struct kvm_vmx_nested_state_hdr {
>> struct {
>> __u16 flags;
>> } smm;
>> -
>> + __u16 pad;
>> __u32 flags;
>> __u64 preemption_timer_deadline;
>> };
>> --
>> 2.25.4
>>
>
--
Vitaly
next prev parent reply other threads:[~2020-07-13 15:55 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-13 8:28 [PATCH] KVM: nVMX: properly pad struct kvm_vmx_nested_state_hdr Vitaly Kuznetsov
2020-07-13 15:17 ` Sean Christopherson
2020-07-13 15:54 ` Vitaly Kuznetsov [this message]
2020-07-27 11:43 ` Paolo Bonzini
2020-07-27 15:46 ` Sean Christopherson
2020-07-27 16:16 ` Paolo Bonzini
2020-07-28 15:59 ` 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=878sfntnoz.fsf@vitty.brq.redhat.com \
--to=vkuznets@redhat.com \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=sean.j.christopherson@intel.com \
--cc=wanpengli@tencent.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).