[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4468f388-b9e0-4407-9be3-260d2204e9d9@solid-run.com>
Date: Tue, 16 Jul 2024 08:16:00 +0000
From: Josua Mayer <josua@...id-run.com>
To: Andrew Lunn <andrew@...n.ch>
CC: Vinod Koul <vkoul@...nel.org>, Kishon Vijay Abraham I <kishon@...nel.org>,
Gregory Clement <gregory.clement@...tlin.com>, Sebastian Hesselbarth
<sebastian.hesselbarth@...il.com>, Rob Herring <robh@...nel.org>, Krzysztof
Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>, Yazan
Shhady <yazan.shhady@...id-run.com>, "linux-phy@...ts.infradead.org"
<linux-phy@...ts.infradead.org>, "linux-kernel@...r.kernel.org"
<linux-kernel@...r.kernel.org>, "linux-arm-kernel@...ts.infradead.org"
<linux-arm-kernel@...ts.infradead.org>, "devicetree@...r.kernel.org"
<devicetree@...r.kernel.org>
Subject: Re: [PATCH RFC 2/2] arm: dts: marvell: armada-38x: add description
for usb phys
Am 15.07.24 um 20:12 schrieb Andrew Lunn:
> On Mon, Jul 15, 2024 at 07:47:30PM +0200, Josua Mayer wrote:
>> Armada 38x has 3x USB-2.0 utmi phys. They are almost identical to the 2x
>> utmi phys on armada 8k.
>>
>> Add descriptions for all 3 phy ports.
>>
>> Also add a syscon node covering just the usb configuration registers.
>> Armada 8K have a syscon node covering configuration registers for
>> various functions including pinmux, woith dirvers using syscon framework
> woith -> with
>
>> for register access.
>>
>> Armada 388 has various drivers directly claiming some of those
>> configuration registers. Hence a similar syscon node would compete for
>> resources with these drivers.
> Do these drivers make exclusive use of these registers? Or are
> multiple drivers using the same registers and you need something to
> mediate in order to have atomic access?
On Armada 38x these drivers exclusively use their own registers.
I suspect armada 8k was trying to avoid declaring lots of tiny ranges,
by using a syscon node.
The syscon registers accessed by cp110-utmi driver are exclusively
for usb and not accessed elsewhere.
>
> In the old days, a register space could be mapped into multiple
> drivers, so long as all the drivers agreed to it. There are probably
> orion5x, kirkwood era mvebu drivers doing this.
>
> Andrew
Powered by blists - more mailing lists