linux-kernel - RE: [PATCH RESEND v3] x86/fpu: use _Alignof to avoid UB in TYPE

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: <1f8cc34bc3af40468e9de6307ff4ad1a@AcuMS.aculab.com>
Date:   Wed, 2 Nov 2022 21:41:44 +0000
From:   David Laight <David.Laight@...LAB.COM>
To:     'Borislav Petkov' <bp@...en8.de>, YingChi Long <me@...lyc.cn>
CC:     "chang.seok.bae@...el.com" <chang.seok.bae@...el.com>,
        "dave.hansen@...ux.intel.com" <dave.hansen@...ux.intel.com>,
        "hpa@...or.com" <hpa@...or.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "mingo@...hat.com" <mingo@...hat.com>,
        "ndesaulniers@...gle.com" <ndesaulniers@...gle.com>,
        "pbonzini@...hat.com" <pbonzini@...hat.com>,
        "tglx@...utronix.de" <tglx@...utronix.de>,
        "x86@...nel.org" <x86@...nel.org>
Subject: RE: [PATCH RESEND v3] x86/fpu: use _Alignof to avoid UB in TYPE_ALIGN

From: Borislav Petkov
> Sent: 02 November 2022 16:56
> 
> On Sat, Oct 29, 2022 at 08:25:52PM +0800, YingChi Long wrote:
> > WG14 N2350 made very clear that it is an UB having type definitions
> > with in "offsetof".
> 
> Pls put the working group URL note here, not below in a Link tag.
> 
> Also, write out "UB" pls.

I'm also pretty sure it is only undefined because a compiler
is allowed to add padding between structure members.
So the type definition inside offsetof() could be laid out
differently from any other similar definition.

However the kernel requires that the compiler only adds the
minimum padding required to align members.
So in the kernel it is actually fine.

OTOH using Alignof() (etc) is cleaner.

This is all similar to (probably) why clang objects to arithmetic
on NULL (ie implementing offsetof by looking at the address of
a field when NULL is cast to the struct type).
This is only undefined because the NULL pointer might not
have the all-zero bit pattern.
But pretty much every C ABI uses the zero bit pattern for NULL.
If it used any other value then most of the memset() of structures
would need changing. So for portability they ought to generate
warnings as well.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)