linux-kernel - Re: [PATCH] x86_64: test that userspace stack is in fact NX

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:   Tue, 3 Oct 2023 13:46:20 -0700
From:   Dave Hansen <dave.hansen@...el.com>
To:     Ingo Molnar <mingo@...nel.org>
Cc:     Alexey Dobriyan <adobriyan@...il.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        linux-kernel@...r.kernel.org, "H. Peter Anvin" <hpa@...or.com>
Subject: Re: [PATCH] x86_64: test that userspace stack is in fact NX

On 10/3/23 12:30, Ingo Molnar wrote:
> * Ingo Molnar <mingo@...nel.org> wrote:
>> Because not having NX in 2023 on any system that is threatened is a
>> big security vulnerability in itself, and whether the vendor or owner
>> intentionally did that or not doesn't really matter, and a failing
>> kernel testcase will be the least of their problems.
> BTW., it's also questionable whether the owner is *aware* of the fact that 
> NX is not available: what if some kernel debug option cleared the NX flag, 
> unintended, or there's some serious firmware bug?
> 
> However unlikely those situations might be, I think unconditionally warning 
> about NX not available is a very 2023 thing to do.

100% agree for x86_64.  Any sane x86_64 system has NX and the rest are
noise that can live with the error message, unless someone shows up with
a compelling reason why not.

For 32-bit, the situation is reversed.  The majority of 32-bit-only CPUs
never had NX.  The only reason to even *do* this check on 32-bit is that
we think folks are running i386 kernels on x86_64 hardware _or_ we just
don't care about 32-bit in the first place.

In the end, I think if we're going to do this test on i386, we should
_also_ do the 5-lines-of-code CPUID check.  But I honestly don't care
that much.  I wouldn't NAK (or not merge) this patch over it.