linux-kernel - Re: [alsa-devel] [RFC PATCH] ALSA: core: Add DMA share buffer support

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: <20190125182515.GD6939@sirena.org.uk>
Date:   Fri, 25 Jan 2019 18:25:15 +0000
From:   Mark Brown <broonie@...nel.org>
To:     Takashi Iwai <tiwai@...e.de>
Cc:     Leo Yan <leo.yan@...aro.org>, alsa-devel@...a-project.org,
        bgoswami@...eaurora.org, gustavo@...eddedor.com,
        srinivas.kandagatla@...aro.org, mchehab+samsung@...nel.org,
        sr@...x.de, daniel.thompson@...aro.org, corbet@....net,
        philburk@...gle.com, willy@...radead.org, jmiller@...erware.com,
        keescook@...omium.org, arnd@...db.de, colyli@...e.de,
        ckeepax@...nsource.wolfsonmicro.com, anna-maria@...utronix.de,
        mathieu.poirier@...aro.org, Baolin Wang <baolin.wang@...aro.org>,
        sboyd@...nel.org, linux-kernel@...r.kernel.org, vkoul@...nel.org,
        joe@...ches.com
Subject: Re: [alsa-devel] [RFC PATCH] ALSA: core: Add DMA share buffer  support

On Fri, Jan 25, 2019 at 02:19:22PM +0100, Takashi Iwai wrote:
> Leo Yan wrote:

> > If we directly use the device node /dev/snd/ as file descriptor, even
> > though we specify flag O_EXCL when open it, but it still is not an
> > anon inode file descriptor.  Thus this is not safe enough and will be
> > blocked by SELinux.  On the other hand, this patch wants to use
> > dma-buf framework to provide file descriptor for the audio buffer, and
> > this audio buffer can be one of mutiple audio buffers in the system
> > and it can be shared to any audio client program.

> Hrm, it sounds like a workaround just to bypass SELinux check...

> The sound server can open another PCM stream with O_APPEND, and pass
> that fd to the client, too?

So long as we can teach SELinux that they're safe to export, yeah.

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)