[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180423174850.zjpx2mcs6dtfkgl3@suse.de>
Date: Mon, 23 Apr 2018 19:48:50 +0200
From: Joerg Roedel <jroedel@...e.de>
To: Kees Cook <keescook@...gle.com>
Cc: Joerg Roedel <joro@...tes.org>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...nel.org>,
"H . Peter Anvin" <hpa@...or.com>, X86 ML <x86@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
Linux-MM <linux-mm@...ck.org>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Andy Lutomirski <luto@...nel.org>,
Dave Hansen <dave.hansen@...el.com>,
Josh Poimboeuf <jpoimboe@...hat.com>,
Juergen Gross <jgross@...e.com>,
Peter Zijlstra <peterz@...radead.org>,
Borislav Petkov <bp@...en8.de>, Jiri Kosina <jkosina@...e.cz>,
Boris Ostrovsky <boris.ostrovsky@...cle.com>,
Brian Gerst <brgerst@...il.com>,
David Laight <David.Laight@...lab.com>,
Denys Vlasenko <dvlasenk@...hat.com>,
Eduardo Valentin <eduval@...zon.com>,
Greg KH <gregkh@...uxfoundation.org>,
Will Deacon <will.deacon@....com>,
Anthony Liguori <aliguori@...zon.com>,
Daniel Gruss <daniel.gruss@...k.tugraz.at>,
Hugh Dickins <hughd@...gle.com>,
Andrea Arcangeli <aarcange@...hat.com>,
Waiman Long <llong@...hat.com>, Pavel Machek <pavel@....cz>,
"David H . Gutteridge" <dhgutteridge@...patico.ca>
Subject: Re: [PATCH 28/37] x86/mm/pti: Map kernel-text to user-space on 32
bit kernels
On Mon, Apr 23, 2018 at 10:09:19AM -0700, Kees Cook wrote:
> On Mon, Apr 23, 2018 at 8:47 AM, Joerg Roedel <joro@...tes.org> wrote:
> > From: Joerg Roedel <jroedel@...e.de>
> >
> > Keeping the kernel text mapped with G bit set keeps its
> > entries in the TLB across kernel entry/exit and improved the
> > performance. The 64 bit x86 kernels already do this when
> > there is no PCID, so do this in 32 bit as well since PCID is
> > not even supported there.
>
> I think this should keep at least part of the logic as 64-bit since
> there are other reasons to turn off the Global flag:
>
> https://lkml.kernel.org/r/20180420222026.D0B4AAC9@viggo.jf.intel.com
That patch you linked is for function pti_kernel_image_global_ok() which
is used on 32 bit too. So any logic implemented for 64 bit to turn off
the global bit will automatically be used on 32 bit.
Regards,
Joerg
Powered by blists - more mailing lists