linux-kernel - Re: [PATCH v14 111/113] RFC: KVM: x86, TDX: Add check for setting CPUID

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 13 Jun 2023 10:31:18 -0700
From:   Isaku Yamahata <isaku.yamahata@...il.com>
To:     "Huang, Kai" <kai.huang@...el.com>
Cc:     "kvm@...r.kernel.org" <kvm@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "Yamahata, Isaku" <isaku.yamahata@...el.com>,
        "Christopherson,, Sean" <seanjc@...gle.com>,
        "Chen, Bo2" <chen.bo@...el.com>, "Shahar, Sagi" <sagis@...gle.com>,
        "isaku.yamahata@...il.com" <isaku.yamahata@...il.com>,
        "Aktas, Erdem" <erdemaktas@...gle.com>,
        "pbonzini@...hat.com" <pbonzini@...hat.com>,
        "dmatlack@...gle.com" <dmatlack@...gle.com>,
        "zhi.wang.linux@...il.com" <zhi.wang.linux@...il.com>
Subject: Re: [PATCH v14 111/113] RFC: KVM: x86, TDX: Add check for setting
 CPUID

On Tue, Jun 06, 2023 at 11:57:35PM +0000,
"Huang, Kai" <kai.huang@...el.com> wrote:

> On Sun, 2023-05-28 at 21:20 -0700, isaku.yamahata@...el.com wrote:
> > From: Isaku Yamahata <isaku.yamahata@...el.com>
> > 
> > Add a hook to setting CPUID for additional consistency check of
> > KVM_SET_CPUID2.
> > 
> > Because intel TDX or AMD SNP has restriction on the value of cpuid.  Some
> > value can't be changed after boot.  Some can be only set at the VM
> > creation time and can't be changed later.  Check if the new values are
> > consistent with the old values.
> 
> You might want to use some grammar tool to check against the changelog.
> 
> Also, personally I think it's better to briefly mention why we choose this
> design but not another (e.g., why we just don't make KVM remember all CPUIDs in
> TDH.MNG.INIT and simply ignores/rejects further KVM_SET_CPUID2).  It would be
> helpful for the reviewer, or some git blamer in the future.
> 

> And why this patch is placed at the very bottom of this series? This logic
> should belong to TD creation part which should be at a relative earlier position
> in this series.

Because this is nice-to-have, not necessary for this patch series to work.  


> > @@ -32,6 +32,13 @@ struct kvm_tdx {
> >  	atomic_t tdh_mem_track;
> >  
> >  	u64 tsc_offset;
> > +
> > +	/*
> > +	 * For KVM_SET_CPUID to check consistency. Remember the one passed to
> > +	 * TDH.MNG_INIT
> > +	 */
> > +	int cpuid_nent;
> > +	struct kvm_cpuid_entry2 *cpuid;
> 
> Since these CPUIDs may only be part of the vcpu's CPUIDs, you may want a more
> specific name, for instance, consistent_cpuid?
> 
> Also if you want this be common to AMD, then perhaps the comment should be
> improved too to be more generic.

Because I don't see apparent constraint on cpuid for SEV case in the spec, I
doubt they want it.  Please correct me if I'm wrong.

Given tdx_vcpu_check_cpuid() doesn't do much meaningful check, let's postpone
this patch for now.
-- 
Isaku Yamahata <isaku.yamahata@...il.com>