linux-kernel - Re: [PATCH] x86/csum: Add KASAN/KCSAN instrumentation

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] [day] [month] [year] [list]

Date:   Thu, 27 Jan 2022 15:44:24 +0100
From:   Dmitry Vyukov <dvyukov@...gle.com>
To:     Jann Horn <jannh@...gle.com>
Cc:     Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
        "H. Peter Anvin" <hpa@...or.com>, linux-kernel@...r.kernel.org,
        Andrey Ryabinin <ryabinin.a.a@...il.com>,
        Alexander Potapenko <glider@...gle.com>,
        Andrey Konovalov <andreyknvl@...il.com>,
        kasan-dev@...glegroups.com, Eric Dumazet <edumazet@...gle.com>,
        Christoph Hellwig <hch@....de>
Subject: Re: [PATCH] x86/csum: Add KASAN/KCSAN instrumentation

On Wed, 26 Jan 2022 at 18:48, Jann Horn <jannh@...gle.com> wrote:
> > > In the optimized X86 version of the copy-with-checksum helpers, use
> > > instrument_*() before accessing buffers from assembly code so that KASAN
> > > and KCSAN don't have blind spots there.
> [...]
> > Can these potentially be called with KERNEL_DS as in some compat
> > syscalls? If so it's better to use instrument_copy_to/from_user.
> > Or probably it's better to use them anyway b/c we also want to know
> > about user accesses for uaccess logging and maybe other things.
>
> Christoph Hellwig has basically eradicated KERNEL_DS. :)
>
> In particular, on X86, set_fs(KERNEL_DS) doesn't really do anything
> anymore since commit 47058bb54b579 ("x86: remove address space
> overrides using set_fs()").

That's good!

But still need this for uaccess logging.