linux-kernel - Re: [PATCH 2/2] tools/x86/kcpuid: Add AMD leaf 0x8000001E

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]

Message-ID: <20210317003011.GD49151@shbuild999.sh.intel.com>
Date:   Wed, 17 Mar 2021 08:30:11 +0800
From:   Feng Tang <feng.tang@...el.com>
To:     Sean Christopherson <seanjc@...gle.com>
Cc:     Borislav Petkov <bp@...en8.de>, X86 ML <x86@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/2] tools/x86/kcpuid: Add AMD leaf 0x8000001E

Hi Boris and Sean,

On Tue, Mar 16, 2021 at 10:04:44AM -0700, Sean Christopherson wrote:
> On Tue, Mar 16, 2021, Borislav Petkov wrote:
> > On Tue, Mar 16, 2021 at 03:42:23PM +0800, Feng Tang wrote:
> > > Also I'm wondering for some basic leaf and extended leaf which
> > > may has different definition for different vendors, do we need
> > > to seprate the csv to a general one and vendor specific ones.
> > 
> > Do you know of such?

No. When I read the patch, I googled some doc for the registers definition
which I found different from Intel's manual.

> > 
> > Because AFAIK vendors own, more or less, each range. Like, Intel owns
> > the base range and AMD the extended so there should be no conflicts
> > actually...
> 
> There are no known conflicts, and all sorts of things would break horribly if
> any CPU vendor (or hypervsior) were careless enough to redefine a CPUID bit.

Great to know these sharing policy between vendors, which will save many
troubles for us :)

Also I just took a look at code of cpuid, which has some functions like
	print_leafX_vendorA
	print_leafX_vendorB
	print_leafX_vendorC
but as you mentioned, I didn't find obvious overlaps of specific bits.

Thanks,
Feng