Re: [PATCH v2] ARM: dts: omap4-embt2ws: Add Bluetooth

From: Andreas Kemnade <andreas@kemnade.info>
To: Tony Lindgren <tony@atomide.com>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>,
	bcousson@baylibre.com, robh+dt@kernel.org,
	krzysztof.kozlowski+dt@linaro.org, conor+dt@kernel.org,
	linux-omap@vger.kernel.org, devicetree@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] ARM: dts: omap4-embt2ws: Add Bluetooth
Date: Sat, 7 Oct 2023 10:29:54 +0200	[thread overview]
Message-ID: <20231007102954.2eeb5edc@aktux> (raw)
In-Reply-To: <20231007072442.GW34982@atomide.com>

Hi Tony,

On Sat, 7 Oct 2023 10:24:42 +0300
Tony Lindgren <tony@atomide.com> wrote:

> * Tony Lindgren <tony@atomide.com> [231007 07:00]:
> > * Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> [231005 08:47]:  
> > > On 04/10/2023 09:03, Andreas Kemnade wrote:  
> > > > Since the required clock is now available, add bluetooth.
> > > > 
> > > > Note: Firmware (bts file) from device vendor reroutes tx for some time
> > > > during initialisation and later put it back, producing timeouts in
> > > > bluetooth initialisation but ignoring that command leads to proper
> > > > initialisation.
> > > > 
> > > > Signed-off-by: Andreas Kemnade <andreas@kemnade.info>
> > > > ---  
> > > 
> > > 
> > > Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>  
> > 
> > Applying into omap-for-v6.7/dt thanks.  
> 
> Looks like this causes the following #clock-cells warning so dropping
> this patch.
> 
> Regards,
> 
> Tony
> 
> arch/arm/boot/dts/ti/omap/omap4-epson-embt2ws.dts:457462.4: Warning (clocks_property):
> /ocp/interconnect@48000000/segment@0/target-module@6c000/serial@0/bluetooth-gnss:
> Missing property '#clock-cells' in node /ocp/interconnect@48000000/segment@0/targe
> 
> 
well, it depends on the wlan dts patch. I think we are better of with having
both applied now. It is a better description of the hardware, and that is what
the devicetree is about. Having wlan randomly working without undocumented
description is IMHO worse than with a clear description.
The reason why this did not come to light earlier is because I had tested it on top
of a bunch of regression fixes for the twl (which were accepted at that point) 
and I did miss the fact that I had also a had a local hack amoung them to enable
the clock by a simple register write in _probe it twl-core.c.

I think as Stephen has little comments about the last version and just not
collected anything else than fixes for -rc, I would think it simply will be
applied. I think he just has a late start.

Regards,
Andreas