linux-kernel - Re: [GIT PULL] pid: use flex array

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, 30 Jun 2023 09:01:54 -0700
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     David Laight <David.Laight@...lab.com>
Cc:     Christian Brauner <brauner@...nel.org>,
        Kees Cook <keescook@...omium.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [GIT PULL] pid: use flex array

On Fri, 30 Jun 2023 at 04:21, David Laight <David.Laight@...lab.com> wrote:
>
> The bit pattern used for a NULL pointer is implementation defined.

Oh, pretty much everything we do in the kernel is implementation defined.

And yes, we do indeed basically rely on the bit pattern for NULL being
all zeroes, and that example I gave with PCI_IOBASE is just one of
many.

And yes, compilers will sometimes complain about the things we do.

On s390, for example, the low memory range is special (and the kernel
mapping is special), so s390 uses all these pointers that are
*literally* small constants, and we've had to turn off some compiler
warnings because gcc would think that we're doing small offsets off
NULL.

But "implementation defined" is actually the good case.

The problem case is "undefined", when we sometimes want to do exactly
those kinds of things. We try to generally avoid it, but we sometimes
end up using compiler switches to say "turn off that part of your
logic" (strict overflow comes to mind).

             Linus