linux-kernel - Re: [PATCHv7 12/19] x86/mm: Adjust virtual address space layout in early boot.

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:   Thu, 28 Sep 2017 15:38:38 +0200
From:   Ingo Molnar <mingo@...nel.org>
To:     "Kirill A. Shutemov" <kirill@...temov.name>
Cc:     "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
        Ingo Molnar <mingo@...hat.com>,
        Linus Torvalds <torvalds@...ux-foundation.org>, x86@...nel.org,
        Thomas Gleixner <tglx@...utronix.de>,
        "H. Peter Anvin" <hpa@...or.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Andy Lutomirski <luto@...capital.net>,
        Cyrill Gorcunov <gorcunov@...nvz.org>,
        Borislav Petkov <bp@...e.de>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCHv7 12/19] x86/mm: Adjust virtual address space layout in
 early boot.


* Kirill A. Shutemov <kirill@...temov.name> wrote:

> On Thu, Sep 28, 2017 at 10:31:55AM +0200, Ingo Molnar wrote:
> > 
> > * Kirill A. Shutemov <kirill.shutemov@...ux.intel.com> wrote:
> > 
> > > We need to adjust virtual address space to support switching between
> > > paging modes.
> > > 
> > > The adjustment happens in __startup_64().
> > 
> > > +#ifdef CONFIG_X86_5LEVEL
> > > +	if (__read_cr4() & X86_CR4_LA57) {
> > > +		pgtable_l5_enabled = 1;
> > > +		pgdir_shift = 48;
> > > +		ptrs_per_p4d = 512;
> > > +	}
> > > +#endif
> > 
> > So CR4 really sucks as a parameter passing interface - was it us who enabled LA57 
> > in the early boot code, right? Couldn't we add a flag which gets set there, or 
> > something?
> 
> It's not necessary that we enabled LA57. At least I tried to write code
> that doesn't assume this. We enable it if bootloader haven't done this
> already for us.
> 
> What is so awful about using CR4 as passing interface? It's one-time
> check, so performance shouldn't be an issue.

As a starter, this code is in generic x86 code [choose_random_location()], is this 
CR4 bit known to AMD as well and is it guaranteed to be sane across all x86 CPUs? 
I don't think so.

CR4 is a poor interface to pass CPU features through. Generaly we try to enumerate 
CPU features via CPUID, and/or enable synthetic CPU features in certain cases, and 
work from there.

Thanks,

	Ingo