From mboxrd@z Thu Jan  1 00:00:00 1970
From: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Subject: Re: [RFC 5/7] soc: qcom: Add Shared Memory Driver
Date: Thu, 30 Oct 2014 08:04:31 -0700
Message-ID: <20141030150429.GX28611@sonymobile.com>
References: <1412037291-16880-1-git-send-email-bjorn.andersson@sonymobile.com>
 <1412037291-16880-6-git-send-email-bjorn.andersson@sonymobile.com>
 <CADMYwHxLtAZZMUGX3jo0URAezqGuho5XB=hQJp8k+acBiENrhg@mail.gmail.com>
 <20141030003806.GU28611@sonymobile.com>
 <CAK=WgbbP8UQAbFauE80PnF5bj3FTAkB2KfF0kFE8Pt23jQrDhA@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <CAK=WgbbP8UQAbFauE80PnF5bj3FTAkB2KfF0kFE8Pt23jQrDhA@mail.gmail.com>
Sender: linux-kernel-owner@vger.kernel.org
To: Ohad Ben-Cohen <ohad@wizery.com>
Cc: Kumar Gala <galak@codeaurora.org>, Andy Gross <agross@codeaurora.org>, Arnd Bergmann <arnd@arndb.de>, "devicetree@vger.kernel.org" <devicetree@vger.kernel.org>, linux-arm-kernel <linux-arm-kernel@lists.infradead.org>, "linux-arm-msm@vger.kernel.org" <linux-arm-msm@vger.kernel.org>, Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, Grant Likely <grant.likely@linaro.org>, Ian Campbell <ijc+devicetree@hellion.org.uk>, Lee Jones <lee.jones@linaro.org>, Liam Girdwood <lgirdwood@gmail.com>, Mark Brown <broonie@kernel.org>, Mark Rutland <mark.rutland@arm.com>, Pawel Moll <pawel.moll@arm.com>, Rob Herring <robh+dt@kernel.org>, Samuel Ortiz <sameo@linux.intel.com>, Kevin Hilman <khilman@deeprootsystems.com>
List-Id: devicetree@vger.kernel.org

On Thu 30 Oct 06:34 PDT 2014, Ohad Ben-Cohen wrote:

> Hi Bjorn,
> 
> On Thu, Oct 30, 2014 at 2:38 AM, Bjorn Andersson
> <bjorn.andersson@sonymobile.com> wrote:
> > My concern is that merging SMD into rpmsg will still give us two different
> > solutions, except that they will be deeply tangled.
> 
> I agree.
> 
> It's very clear that SMD and rpmsg are different. But fundamentally,
> they do exactly the same: new messages are placed in a ring buffer and
> then an interrupt is triggered.
> 
> If Qualcomm wanted to adopt rpmsg I'm confident there would have been
> no technical barriers to do so.
> 

Their wish list seems long, but I think you're right :)

> But given that SMD is not going anywhere, I personally hope that you
> could at least avoid having two different implementations of it merged
> upstream.
> 

Looking at the lack of functionality in mach-msm it seems like it's quite a lot
of work getting to a point where anyone could actually test smd on those devices.

If anyone have any input on how I get a piece of user space that will run with
mainline and communicate with the modem to test this I'm happily listening.

Regards,
Bjorn