linux-kernel - Re: arch/arm/crypto/sha1_glue.c:34:8: warning: cast from 'void (*)(u32 *, const unsigned char *, unsigned int)' (aka 'void (*)(unsigned int *, const unsigned char *, unsigned int)') to 'sha1_block_fn *' (aka 'void (*)(struct sha1

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: <CAMj1kXGkdp_cwOWeX_49Y5xTCH8Y7X-LnM8pJufCH3yw7anBWQ@mail.gmail.com>
Date:   Thu, 8 Dec 2022 19:05:45 +0100
From:   Ard Biesheuvel <ardb@...nel.org>
To:     Herbert Xu <herbert@...dor.apana.org.au>
Cc:     kernel test robot <lkp@...el.com>,
        Robert Elliott <elliott@....com>, llvm@...ts.linux.dev,
        oe-kbuild-all@...ts.linux.dev, linux-kernel@...r.kernel.org,
        Linux Crypto Mailing List <linux-crypto@...r.kernel.org>
Subject: Re: arch/arm/crypto/sha1_glue.c:34:8: warning: cast from 'void
 (*)(u32 *, const unsigned char *, unsigned int)' (aka 'void (*)(unsigned int
 *, const unsigned char *, unsigned int)') to 'sha1_block_fn *' (aka 'void
 (*)(struct sha1_state *, const unsigned char ...

On Thu, 8 Dec 2022 at 05:59, Herbert Xu <herbert@...dor.apana.org.au> wrote:
>
> On Fri, Nov 04, 2022 at 07:33:43PM +0800, kernel test robot wrote:
> >
> > vim +34 arch/arm/crypto/sha1_glue.c
> >
> > f0be44f4fb1fae David McCullough 2012-09-07  23
> > 1f8673d31a999e Jussi Kivilinna  2014-07-29  24  asmlinkage void sha1_block_data_order(u32 *digest,
> > f0be44f4fb1fae David McCullough 2012-09-07  25                const unsigned char *data, unsigned int rounds);
> > f0be44f4fb1fae David McCullough 2012-09-07  26
> > 604682551aa511 Jussi Kivilinna  2014-07-29  27  int sha1_update_arm(struct shash_desc *desc, const u8 *data,
> > f0be44f4fb1fae David McCullough 2012-09-07  28                    unsigned int len)
> > f0be44f4fb1fae David McCullough 2012-09-07  29  {
> > 90451d6bdb787e Ard Biesheuvel   2015-04-09  30        /* make sure casting to sha1_block_fn() is safe */
> > 90451d6bdb787e Ard Biesheuvel   2015-04-09  31        BUILD_BUG_ON(offsetof(struct sha1_state, state) != 0);
> > f0be44f4fb1fae David McCullough 2012-09-07  32
> > 90451d6bdb787e Ard Biesheuvel   2015-04-09  33        return sha1_base_do_update(desc, data, len,
> > 90451d6bdb787e Ard Biesheuvel   2015-04-09 @34                                   (sha1_block_fn *)sha1_block_data_order);
> > f0be44f4fb1fae David McCullough 2012-09-07  35  }
> > 604682551aa511 Jussi Kivilinna  2014-07-29  36  EXPORT_SYMBOL_GPL(sha1_update_arm);
> > f0be44f4fb1fae David McCullough 2012-09-07  37
>
> So clan doesn't like the cast on the assembly function.
>
> Ard, why can't we just change the signature of the assembly
> function instead of casting?
>

We can, as the BUILD_BUG() will catch it if struct sha1_state gets
modified in an incompatible way.