[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAGXu5jKkPJAHezzX1o0wQySaHLjYW-muGJ8W0XeDGvvsn0h3Zg@mail.gmail.com>
Date: Wed, 11 May 2016 19:11:57 -0700
From: Kees Cook <keescook@...omium.org>
To: Andi Kleen <ak@...ux.intel.com>
Cc: Hector Marco-Gisbert <hecmargi@....es>,
Andy Lutomirski <luto@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>,
"H. Peter Anvin" <hpa@...or.com>,
"x86@...nel.org" <x86@...nel.org>, Brian Gerst <brgerst@...il.com>,
Borislav Petkov <bp@...e.de>,
Huaitong Han <huaitong.han@...el.com>,
Ismael Ripoll Ripoll <iripoll@....es>
Subject: Re: [PATCH] x86_64: Disabling read-implies-exec when the stack is executable
On Wed, May 11, 2016 at 3:40 PM, Andi Kleen <ak@...ux.intel.com> wrote:
>> However, I would tend to agree: RIE should only be needed on 32-bit
>> since 64-bit started its life knowing about no-exec permissions.
>
> NX was not in the original AMD K8 chips. Was only added some time later.
So we should retain this behavior for all of 64-bit?
>> set_personality_64bit()'s (which is confusingly just an initializer
>> and not called during the personality() syscall) comment about this
>> makes no sense to me:
>>
>> /* TBD: overwrites user setup. Should have two bits.
>> But 64bit processes have always behaved this way,
>> so it's not too bad. The main problem is just that
>> 32bit childs are affected again. */
>> current->personality &= ~READ_IMPLIES_EXEC;
>
> What does not make sense?
I just don't have enough context to make sense of it. What two bits?
Always behaved what way?Affected by what?
-Kees
--
Kees Cook
Chrome OS & Brillo Security
Powered by blists - more mailing lists