Re: [PATCH v2 1/2] dt-bindings: remoteproc: qcom: Add firmware bindings for Q6V5

From: Brian Norris <briannorris@chromium.org>
To: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Sibi Sankar <sibis@codeaurora.org>,
	david.brown@linaro.org, robh+dt@kernel.org, mark.rutland@arm.com,
	andy.gross@linaro.org, akdwived@codeaurora.org,
	clew@codeaurora.org, linux-kernel@vger.kernel.org,
	linux-arm-msm-owner@vger.kernel.org, ohad@wizery.com,
	linux-remoteproc@vger.kernel.org, devicetree@vger.kernel.org
Subject: Re: [PATCH v2 1/2] dt-bindings: remoteproc: qcom: Add firmware bindings for Q6V5
Date: Thu, 3 Jan 2019 16:11:59 -0800	[thread overview]
Message-ID: <20190104001158.GA200069@google.com> (raw)
In-Reply-To: <20190104000145.GJ31596@builder>

Hi Bjorn,

On Thu, Jan 03, 2019 at 04:01:45PM -0800, Bjorn Andersson wrote:
> On Thu 03 Jan 15:50 PST 2019, Brian Norris wrote:
> 
> > On Thu, Jan 03, 2019 at 03:30:14PM -0800, Brian Norris wrote:
> > > On Fri, Dec 28, 2018 at 10:18:18AM +0530, Sibi Sankar wrote:
> > > > +- firmware-name:
> > > > +	Usage: optional
> > > > +	Value type: <string>
> > > > +	Definition: must list the relative firmware image path for the
> > > > +		    Hexagon Core.
> > > 
> > > Relative to what? I still think it's a terrible idea that your driver
> > > looks for files at the top-level /lib/firmware/ directory, but now
> > > you're leaking this into the device tree. This should at a bare minimum
> > > be namespaced to something like the qcom/ sub-directory. But ideally,
> > > the driver would automatically be deriving a further sub-directory of
> > > qcom/ based on the chipset or something, and then the only thing you'd
> > > describe here is some kind of variant string -- something akin to
> > > ath10k's qcom,ath10k-calibration-variant (see
> > > Documentation/devicetree/bindings/net/wireless/qcom,ath10k.txt), which
> > > doesn't require a full path-name or any hierarchy.
> > 
> > Oh, I see Rob actually recommended this binding in v1, and it's (sort
> > of) in use by a few other drivers. Is it really expected that we put
> > arbitrary pathnames in device tree? None of the binding documentation
> > seems very specific to me, and their implementations *do* allow
> > arbitrary text. As it stands today, this is a great recipe for name
> > collision -- e.g., how the driver today suggests "modem.XYZ" names; is
> > Qualcomm really the only one out there making modems? :D
> > 
> > So my natural instinct is to avoid this. But if that's what everybody
> > wants...
> > 
> 
> I share your concern about this, but I came to suggest this as the
> driver cares about platforms but the firmware is (often?)
> device/product-specific.
> 
> E.g. we will serve the MTP and Pixel 3 with the qcom,sdm845-adsp-pas
> compatible, but they are unlikely to run the same adsp firmware. This
> allows the individual dtb to specify which firmware the driver should
> use.

I understand this, but that still doesn't mean we should be suggesting
each DTB to clutter the top-level firmware search path, especially since
lazy people will probably just use "modem.mdt" and similar. That means
you no longer can ship the same rootfs that supports both QCOM and
<other> modems, if <other> modem also uses the same lazy format.

It seems like a much better practice to at least enforce a particular
prefix to things. e.g., the driver could assume:

  qcom/sdm845-adsp-pas/ (or if you must, just qcom/)

and your DTB only gets to add .../<your-string-here> to that path.

In case it isn't clear: I think it's also severely misguided that the
existing driver gets away with lines like

	request_firmware(&fw, "modem.mdt", ...);

today ;)

Brian