linux-kernel - Re: [PATCH v2 02/10] netfilter: Avoid assigning 'const' pointer to non-const pointer

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:   Fri, 24 Jan 2020 09:36:54 -0800
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     Peter Zijlstra <peterz@...radead.org>
Cc:     Nick Desaulniers <ndesaulniers@...gle.com>,
        Will Deacon <will@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>,
        linux-arch <linux-arch@...r.kernel.org>,
        kernel-team <kernel-team@...roid.com>,
        Michael Ellerman <mpe@...erman.id.au>,
        Segher Boessenkool <segher@...nel.crashing.org>,
        Christian Borntraeger <borntraeger@...ibm.com>,
        Luc Van Oostenryck <luc.vanoostenryck@...il.com>,
        Arnd Bergmann <arnd@...db.de>,
        Peter Oberparleiter <oberpar@...ux.ibm.com>,
        Masahiro Yamada <masahiroy@...nel.org>,
        Pablo Neira Ayuso <pablo@...filter.org>,
        Jozsef Kadlecsik <kadlec@...filter.org>,
        Florian Westphal <fw@...len.de>,
        "David S. Miller" <davem@...emloft.net>
Subject: Re: [PATCH v2 02/10] netfilter: Avoid assigning 'const' pointer to
 non-const pointer

On Fri, Jan 24, 2020 at 12:25 AM Peter Zijlstra <peterz@...radead.org> wrote:
>
> Just for curiosity's sake. What does clang actually do in that case?

This shouldn't necessarily be clang-specific. If the variable itself
is 'const', it might go into a read-only section. So trying to modify
it will quite possibly hit a SIGSEGV in user space (and in kernel
space cause an oops).

Note that that is different from a const pointer to something that
wasn't originally const. That's just a "error out at compile time if
somebody tries to write through it", but the const'ness can be cast
away, because all the 'const' really said was that the object can't be
modified through _that_ pointer, not in general.

(That also means that the compiler can't necessarily even optimize
multiple accesses through a const pointer away, because the object
might be modified through another pointer that aliases the const one -
you'd need to also mark it "restrict" to tell the compiler that no
other pointer will alias).

                 Linus