linux-kernel - RE: [PATCH 3/7] microblaze: Define SMP safe bit operations

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:   Thu, 13 Feb 2020 09:24:46 +0000
From:   Stefan Asserhall <stefana@...inx.com>
To:     Peter Zijlstra <peterz@...radead.org>,
        Michal Simek <michals@...inx.com>
CC:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "monstr@...str.eu" <monstr@...str.eu>, git <git@...inx.com>,
        "arnd@...db.de" <arnd@...db.de>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Masahiro Yamada <yamada.masahiro@...ionext.com>,
        Will Deacon <will@...nel.org>
Subject: RE: [PATCH 3/7] microblaze: Define SMP safe bit operations

> On Thu, Feb 13, 2020 at 09:01:21AM +0000, Stefan Asserhall wrote:
> > The comment in the generic bitops.h says "You should recode these in
> > the native assembly language, if at all possible". I don't think using
> > the generic implementation will be as efficient as the current arch specific one.
> 
> That is a very crusty old recommendation. Please look at the compiler generated
> code.
> 
> We've extended the atomic_t operations the past few years and Will wrote the
> generic atomic bitops for Arm64, we're looking to convert more LL/SC archs to
> them.
> 
> There is currently one known issue with it, but Will has a patch-set pending to
> solve that (IIRC that only matters if you have stack protector on).
> 
> Also see this thread:
> 
>   https://lkml.kernel.org/r/875zimp0ay.fsf@mpe.ellerman.id.au
> 
> And these patches:
> 
>   https://lkml.kernel.org/r/20200123153341.19947-1-will@kernel.org

Thanks for the links. Sure, I agree that it is better to use the generic 
implementation if it is as efficient as the arch specific one, but I don't
think we should assume that it is.

Michal, would it be possible to replace the arch specific code and check
what we get?

Stefan