linux-kernel - Re: [alsa-devel] [PATCH] ASoC: dwc: disallow building designware

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: <s5h37d2su8h.wl-tiwai@suse.de>
Date:   Thu, 20 Apr 2017 22:24:14 +0200
From:   Takashi Iwai <tiwai@...e.de>
To:     Mark Brown <broonie@...nel.org>
Cc:     Jose Abreu <Jose.Abreu@...opsys.com>,
        Lubomir Rintel <lkundrak@...sk>, alsa-devel@...a-project.org,
        linux-kernel@...r.kernel.org, Arnd Bergmann <arnd@...db.de>
Subject: Re: [alsa-devel] [PATCH] ASoC: dwc: disallow building  designware_pcm as a module

On Thu, 20 Apr 2017 21:46:46 +0200,
Mark Brown wrote:
> 
> On Wed, Apr 19, 2017 at 05:48:15PM +0100, Jose Abreu wrote:
> 
> > What do you think Mark? If you want to keep the PCM as a module
> > then we will need to abstract this more, by reducing the
> > dependencies.
> 
> I think forcing this to be built in to the kernel (which is what the
> commit message says the change is going to do) is an obviously bad
> idea.  Anything we add to the base kernel image needs to have a good
> reason to be there and it is hard to think what that reason might be for
> any audio driver, we need to be able to put this code into a module.

Well, I guess the original patch description caused a big confusion.
As far as I see, the intention of the patch is not about the module or
built-in kernel.  Instead it's rather to fold designware_pcm stuff
into the single designware_i2s driver.

The former is merely an extension of the latter driver, and the latter
invokes the former directly.  Thus there is little merit to keep them
separate.  I think the current code is even buggy, which allows to
leave CONFIG_SND_DESIGNWARE_I2S=y and CONFIG_SND_DESIGNWARE_PCM=m.

So, I think Lubomir's change is right.  But the patch subject and
description should be rephrased.

One thing I don't like is the rename of the file.  But in this
particular case, it's unavoidable unless we rename the module name.

BTW, we should drop the superfluous EXPORT_SYMBOL*(), too.

thanks,

Takashi