Re: [PATCH 1/2] ASoC: qcom: remove ipq806x Storm machine driver

[Dmitry Baryshkov]

On Thu, Jul 23, 2026 at 08:38:09PM +0200, Christian Marangi wrote:
> On Wed, Jul 22, 2026 at 01:48:30PM +0300, Dmitry Baryshkov wrote:
> > On Wed, 22 Jul 2026 at 13:38, Christian Marangi <ansuelsmth@xxxxxxxxx> wrote:
> > >
> > > On Wed, Jul 22, 2026 at 10:00:34AM +0200, Konrad Dybcio wrote:
> > > > On 7/21/26 8:12 PM, Dmitry Baryshkov wrote:
> > > > > The Storm board machine driver targets Google's IPQ806x-based Storm
> > > > > reference board (a MAX98357A DAC wired to the LPASS I2S bus). This board
> > > > > has no upstream device tree, no active users and has seen no functional
> > > > > changes in years, so the driver is dead weight.
> > > > >
> > > > > Remove the driver together with its Kconfig entry and Makefile rule.
> > > > >
> > > > > Assisted-by: Claude:claude-opus-4-8
> > > >
> > > > Using AI to `git rm` is definitely tokenmaxxing!
> > > >
> > > > > Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxxxxxxxx>
> > > > > ---
> > > >
> > > > Reviewed-by: Konrad Dybcio <konrad.dybcio@xxxxxxxxxxxxxxxx>
> > > >
> > >
> > > Honestly I'm pretty angry about the latest changes on ipq806x. First the
> > > dtsi just because there weren't device (it's ok and understandable) now we
> > > are starting to remove driver...
> > >
> > > There are godzillion situation where there is a driver but absolutely no
> > > dtsi. Lets remove the dwmac driver, usb phy, pcie and everything else at
> > > this point and treat this soc like it's a 80-90 code...
> > >
> > > I will post dtsi with actual device for the SoC but really I'm not
> > > following all this hurry in cleaning code for device that still gets used
> > > and actually some company even produced new one (even if it's EOL)
> > 
> > We have no way to check that the device is being used other than the
> > DTSI present in the kernel, being updated from time to time with
> > non-trivial changes.
> > 
> > For the patchset in question, do you (or the community) actually use
> > OnHub? If so, please consider upstreaming DT files (and also updating
> > schema FWIW).
> > If not, the files will eventually go away. Just having unused drivers
> > _adds_ a burden. I stumbled upon the driver while reviewing a patch
> > changing the driver (for the sake of changing the ASoC API).
> > 
> > And FWIW the series doesn't touch the IPQ806x LPASS driver (which has
> > a DT counterpart).
> > 
> 
> Yes on OpenWrt we support 3 OnHub google devices from Asus and Tplink and
> they also use the storm driver for the strange attached speaker.

Ack. Then pretty please, could the OpenWrt community please contribute
those to the mainline. It is a quest and a process, but it helps us to
understand actual hardware users.  Not to mention that getting platforms
support upstream would lower the burden of rebasing on the new upstream
kernels (assuming that you are using close-to-upstream kernel) and it
would also save from similar cleanup attempts.

-- 
With best wishes
Dmitry