linux-kernel - Re: [PATCH v4 2/2] powerpc/uaccess: Implement unsafe_copy_to

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: <49YBKb2VVkz9sT6@ozlabs.org>
Date:   Fri, 29 May 2020 14:24:17 +1000 (AEST)
From:   Michael Ellerman <patch-notifications@...erman.id.au>
To:     Christophe Leroy <christophe.leroy@....fr>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Paul Mackerras <paulus@...ba.org>, npiggin@...il.com,
        segher@...nel.crashing.org
Cc:     linuxppc-dev@...ts.ozlabs.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v4 2/2] powerpc/uaccess: Implement unsafe_copy_to_user() as a simple loop

On Fri, 2020-04-17 at 17:08:52 UTC, Christophe Leroy wrote:
> At the time being, unsafe_copy_to_user() is based on
> raw_copy_to_user() which calls __copy_tofrom_user().
> 
> __copy_tofrom_user() is a big optimised function to copy big amount
> of data. It aligns destinations to cache line in order to use
> dcbz instruction.
> 
> Today unsafe_copy_to_user() is called only from filldir().
> It is used to mainly copy small amount of data like filenames,
> so __copy_tofrom_user() is not fit.
> 
> Also, unsafe_copy_to_user() is used within user_access_begin/end
> sections. In those section, it is preferable to not call functions.
> 
> Rewrite unsafe_copy_to_user() as a macro that uses __put_user_goto().
> We first perform a loop of long, then we finish with necessary
> complements.
> 
> unsafe_copy_to_user() might be used in the near future to copy
> fixed-size data, like pt_regs structs during signal processing.
> Having it as a macro allows GCC to optimise it for instead when
> it knows the size in advance, it can unloop loops, drop complements
> when the size is a multiple of longs, etc ...
> 
> Signed-off-by: Christophe Leroy <christophe.leroy@....fr>

Applied to powerpc topic/uaccess-ppc, thanks.

https://git.kernel.org/powerpc/c/17bc43367fc2a720400d21c745db641c654c1e6b

cheers