linux-kernel - Re: [RFC PATCH v2 3/6] x86/pti: add a per-cpu variable pti

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: <20180110085016.GE14066@1wt.eu>
Date:   Wed, 10 Jan 2018 09:50:16 +0100
From:   Willy Tarreau <w@....eu>
To:     Ingo Molnar <mingo@...nel.org>
Cc:     linux-kernel@...r.kernel.org, x86@...nel.org,
        Andy Lutomirski <luto@...nel.org>,
        Borislav Petkov <bp@...en8.de>,
        Brian Gerst <brgerst@...il.com>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Josh Poimboeuf <jpoimboe@...hat.com>,
        "H. Peter Anvin" <hpa@...or.com>, Kees Cook <keescook@...omium.org>
Subject: Re: [RFC PATCH v2 3/6] x86/pti: add a per-cpu variable pti_disable

On Wed, Jan 10, 2018 at 09:01:02AM +0100, Ingo Molnar wrote:
> 
> * Willy Tarreau <w@....eu> wrote:
> 
> > [...] If we had "pit_enabled", something like this could be confusing because 
> > it's not obvious whether this pti_enabled *enforces* PTI or if its absence 
> > disables it :
> > 
> > 	cmpb $0, PER_CPU_VAR(pti_enabled)
> > 	jz .Lend\@
> 
> The natural sequence would be:
> 
> 	cmpb $1, PER_CPU_VAR(pti_enabled)
> 	jne .Lend\@
> 
> which is not confusing to me at all.

In fact I think I know now why it still poses me a problem : this
pti_enabled flag alone is not sufficient to enable PTI, it's just part
of the condition, as another part comes from the X86_FEATURE_PTI flag.
However, pti_disabled is sufficient to disable PTI so actually its
effect matches its name (note BTW that I called it "pti_disable" as a
verb indicating an action -- "I want to disable pti", and not as a past
form "pti is disabled").

Willy