linux-kernel - Re: [PATCH 2/3] spi: dt-bindings: Add num-cs property for mpfs-spi

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <1edb6c4c1a66c1a2009278b99f897e3a71b592c6.camel@microchip.com>
Date: Fri, 3 May 2024 12:54:19 +0000
From: <Prajna.Rajendrakumar@...rochip.com>
To: <broonie@...nel.org>, <krzk@...nel.org>
CC: <linux-spi@...r.kernel.org>, <linux-riscv@...ts.infradead.org>,
	<Conor.Dooley@...rochip.com>, <devicetree@...r.kernel.org>,
	<robh@...nel.org>, <linux-kernel@...r.kernel.org>, <krzk+dt@...nel.org>,
	<Valentina.FernandezAlanis@...rochip.com>, <Daire.McNamara@...rochip.com>
Subject: Re: [PATCH 2/3] spi: dt-bindings: Add num-cs property for mpfs-spi

On Thu, 2024-05-02 at 16:53 +0200, Krzysztof Kozlowski wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you
> know the content is safe
> 
> On 02/05/2024 16:34, Prajna Rajendra Kumar wrote:
> > The PolarFire SoC SPI controller supports multiple chip selects,but
> > in
> > the MSS, only one CS line is physically wired. To reflect this
> > hardware
> > limitation in the device tree, the binding enforces that the 'num-
> > cs'
> > property defaults to 1 and cannot exceed 1 unless additional
> > chip select lines are explicitly defined using GPIO descriptors.
> > 
> 

Hi,

> You marked it as Fix for bug, but I don't understand where the bug
> is.

The bug was that the PolarFire SoC SPI "hard" controller supports eight
chip selects, but only one chip select is connected and can be used.
This was not reflected in the device tree because default num-cs was
never set. Hence, it's marked as a fix.

> Do you describe above the issue or the solution?

It describes both the issue and the solution, but I will revise the
commit message as Conor Dooley suggested.
> 
> > +  - if:
> > +      properties:
> > +        compatible:
> > +          contains:
> > +            const: microchip,mpfs-spi
> > +      not:
> > +        required:
> > +          - cs-gpios
> 
> I don't understand what you are expressing here. Did you actually
> validate it that it achieves exactly what you want?

Since the controller supports only one chip select, the num-cs should
default to 1 and cannot exceed 1 unless GPIOs are used as chip selects.

> > +    then:
> > +      properties:
> > +        num-cs:
> > +          default: 1
> > +          maximum: 1
> > +
> >  unevaluatedProperties: false
> > 
> >  examples:
> 
> Best regards,
> Krzysztof
> 
Best regards,
Prajna