linux-kernel - Re: [PATCH 0/2] ALSA: nm256: Fine-tuning for three function implementations

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: <s5h7euq9hqt.wl-tiwai@suse.de>
Date:   Thu, 16 Nov 2017 19:54:34 +0100
From:   Takashi Iwai <tiwai@...e.de>
To:     SF Markus Elfring <elfring@...rs.sourceforge.net>
Cc:     alsa-devel@...a-project.org,
        Arvind Yadav <arvind.yadav.cs@...il.com>,
        Jaroslav Kysela <perex@...ex.cz>,
        Takashi Sakamoto <o-takashi@...amocchi.jp>,
        kernel-janitors@...r.kernel.org,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 0/2] ALSA: nm256: Fine-tuning for three function implementations

On Thu, 16 Nov 2017 18:48:43 +0100,
SF Markus Elfring wrote:
> 
> >> Two update suggestions were taken into account
> >> from static source code analysis.
> > 
> > Markus, I'd apply this kind of patches only when they are really
> > tested on the hardware,
> 
> I can not test all software and hardware combinations (so far)
> for which I dare to show change possibilities.
> 
> 
> > or they were converted systematically by a script like spatch.
> 
> There is a general source code transformation pattern involved.
> So I find that it is systematic.
> 
> But I did not dare to develop a script variant for the semantic patch
> language (Coccinelle software) which can handle all special use cases
> as a few of them are already demonstrated in this tiny patch series.

Then you're doing everything by hands, and can be wrong -- that's the
heart of the problem.  The risk is bigger than the merit by applying
the patch.

So, just prove that your patch doesn't break anything.  Doesn't matter
whether it's a test with real hardware or with systematic checks.
Once when it's confirmed, we can apply it.  A very simple rule, and
this will be valid for most of other subsystems, too.


thanks,

Takashi

> 
> 
> > The reason is that you might break something
> 
> There are the usual software development risks.
> 
> 
> > (and you already broke things in the past).
> 
> I presented also some improvable update suggestions.
> 
> Another look on the corresponding circumstances might be interesting
> for further clarification.
> 
> 
> > The merit by such a patch is negligible in comparison of the risk of breakage.
> 
> I imagine that you might like a small object code reduction also for
> this software module.
> 
> 
> > These codes aren't too bad without fixing, after all;
> > everyone can read it pretty well as is.
> 
> The script "checkpatch.pl" points implementation details out for
> further considerations.
> 
> 
> > If these patches were tested on a real hardware,
> 
> I assume that this aspect can become a big challenge.
> 
> 
> > or at least on VM, so that you can show that they don't break anything,
> 
> Which test results would you trust (from me)?
> 
> 
> > I'll happily apply them for the next (4.16) kernel.
> 
> Thanks for your general offer.
> 
> 
> > Or, if you're really working on other real changes
> 
> I would find a bit more efficient exception handling useful.
> 
> 
> > (no cosmetic coding style fixes nor the code shuffling,
> 
> I propose to apply also corresponding checkpatch cosmetic.
> 
> 
> > but fixing a real bug)
> 
> I am trying to adjust the software situation a bit more for better
> run time characteristics.
> 
> 
> > *and* such a cleanup is mandatory as preliminary, it can be accepted, too.
> 
> There are change combinations needed for the proposed software
> design direction.
> Can you see positive effects here?
> 
> Regards,
> Markus
>