linux-kernel - Re: [PATCH v4 1/3] dt-bindings: w1: UART 1-Wire bus

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <ZaYr7jdxoKMqjnO1@cjw-notebook>
Date: Tue, 16 Jan 2024 08:10:38 +0100
From: Christoph Winklhofer <cj.winklhofer@...il.com>
To: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
Cc: Rob Herring <robh@...nel.org>, Conor Dooley <conor+dt@...nel.org>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Jiri Slaby <jirislaby@...nel.org>,
	Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
	Jonathan Corbet <corbet@....net>, devicetree@...r.kernel.org,
	linux-kernel@...r.kernel.org, linux-serial@...r.kernel.org,
	linux-doc@...r.kernel.org
Subject: Re: [PATCH v4 1/3] dt-bindings: w1: UART 1-Wire bus

On Mon, Jan 15, 2024 at 07:02:55PM +0100, Krzysztof Kozlowski wrote:
> On 15/01/2024 18:36, Christoph Winklhofer wrote:
> > On Sun, Jan 14, 2024 at 04:55:42PM +0100, Krzysztof Kozlowski wrote:
> >> On 14/01/2024 15:47, Christoph Winklhofer wrote:
> >>> On Sun, Jan 14, 2024 at 11:54:48AM +0100, Krzysztof Kozlowski wrote:
> >>>> On 13/01/2024 19:04, Christoph Winklhofer wrote:
> >>>>> On Fri, Jan 12, 2024 at 07:39:17PM -0600, Rob Herring wrote:
> >>>>>> On Sat, Jan 06, 2024 at 05:02:24PM +0100, Christoph Winklhofer wrote:
> >>>>>>> Add device tree binding for UART 1-Wire bus.
> >>>>>>>
> >>>>>>> Signed-off-by: Christoph Winklhofer <cj.winklhofer@...il.com>
> >>>>>>> ---
> >>>>>>>  Documentation/devicetree/bindings/w1/w1-uart.yaml | 62 +++++++++++++++++++++++
> >>>>>>>  1 file changed, 62 insertions(+)
> >>>>>>>
> >>>>>>> diff --git a/Documentation/devicetree/bindings/w1/w1-uart.yaml b/Documentation/devicetree/bindings/w1/w1-uart.yaml
> >>>>>>> new file mode 100644
> >>>>>>> index 000000000000..6b90693b2ca0
> >>>>>>> --- /dev/null
> >>>>>>> +++ b/Documentation/devicetree/bindings/w1/w1-uart.yaml
> >>>>>>> @@ -0,0 +1,62 @@
> >>>>>>> +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
> >>>>>>> +%YAML 1.2
> >>>>>>> +---
> >>>>>>> +$id: http://devicetree.org/schemas/w1/w1-uart.yaml#
> >>>>>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> >>>>> ...
> >>>>>>> +properties:
> >>>>>>> +  compatible:
> >>>>>>> +    const: w1-uart
> >>>>>>> +
> >>>>>>> +  reset-speed:
> >>>>>>> +    $ref: /schemas/types.yaml#/definitions/uint32
> >>>>>>> +    default: 9600
> >>>>>>> +    description: |
> >>>>>>
> >>>>>> Don't need '|' if no formatting
> >>>>>>
> >>>>>
> >>>>> Ok.
> >>>>>
> >>>>>>> +      The baud rate for the 1-Wire reset and presence detect.
> >>>>>>> +
> >>>>>>> +  touch_0-speed:
> >>>>>>
> >>>>>> Don't use '_' in property names.
> >>>>>>
> >>>>>> I'm somewhat familar with 1-wire, but I don't get what 'touch' means 
> >>>>>> here. I assume these are low and high times which are a function of the 
> >>>>>> baudrate.
> >>>>>>
> >>>>>
> >>>>> I change the name to 'write-0-speed' and 'write-1-speed'. The function
> >>>>> in the w1-framework is named 'touch_bit' - therefore the previous
> >>>>> naming. 
> >>>>>
> >>>>> It is the baud-rate used in the 1-Wire cycle to write a 0-Bit
> >>>>> (write-0-speed) and to perform a 1-Wire cycle that writes a 1-Bit and
> >>>>> reads a 0-Bit or 1-Bit (write-1-speed).
> >>>>
> >>>>
> >>>> Then probably -bps:
> >>>> https://github.com/devicetree-org/dt-schema/blob/main/dtschema/schemas/property-units.yaml
> >>>
> >>> The serial.yaml uses prefix -speed for the baud rate but I can change it
> >>> to -bps.
> >>
> >> Do you reference serial.yaml?
> >>
> > 
> > No, serial.yaml is not referenced but 'onewire' will be a child-node of
> > a serial-device which already defines baud rate related properties
> > with -speed (e.g. max-speed although not used in w1-uart). Hence, I
> > thought -speed is typically used for baud rates.
> 
> Ah, it defines max-speed for childre, so for onewire. Re-using that
> property would make sense, but since you are defining completely new
> properties, let's use proper (-bps) naming.
> 

Ok.

> I still wonder, why would you use different baud rates for these three
> different operations?

The timing for the pulse on the 1-Wire bus, to perform a 1-Wire reset or
read/write is too different to use the same baud rate for them. The
pulse can only be generated with the transmission of a single byte (to
initiate the operation and to evaluate the response), since every UART
packet starts low and ends high, only a limited set of baud rates are
possible.

Read-0/1 and Write-1 have the same low-time and therefore share the same
baud rate setting, actually this baud could be higher. Otherwise, for
Write-0 the maximum baud-rate is somewhere near 115200.

The table lists the low-time range for the 1-Wire standard speed
(overdrive is currently not implemented by the driver):

  Operation | Low-Time (usec) | Baud rate (default)
  ----------+-----------------+--------------------
  Reset     | 480 - 640       | 9600
  Read-0/1  |   5 -  15       | 115200
  Write-1   |   5 -  15       | 115200
  Write-0   |  60 - 120       | 115200

I thought it makes sense to have the option to overwrite the default
baud rates. Alternatively, the driver could probe a fixed set of
different baud rates that are able to generate the timing patterns.
Initially, the driver always used 9600 and 115200 - however some UART
devices may not support these bauds. Now the driver requests a baud rate
and checks if it can create the timing pattern with the returned one.

Kind regards,
Christoph