[PATCH RFC 0/2] Introduce PRU UART driver

Judith Mendez posted 2 patches 9 months, 1 week ago
There is a newer version of this series
.../bindings/serial/ti,pruss-uart.yaml        |  54 +++++
drivers/tty/serial/8250/8250_pruss.c          | 213 ++++++++++++++++++
drivers/tty/serial/8250/Kconfig               |  10 +
drivers/tty/serial/8250/Makefile              |   1 +
4 files changed, 278 insertions(+)
create mode 100644 Documentation/devicetree/bindings/serial/ti,pruss-uart.yaml
create mode 100644 drivers/tty/serial/8250/8250_pruss.c
[PATCH RFC 0/2] Introduce PRU UART driver
Posted by Judith Mendez 9 months, 1 week ago
This patch series is sent as an RFC to get some initial comments
on the PRU UART driver.

The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
SoCs have a UART sub-module. This patch series introduces the driver and the
corresponding binding documentation for this sub-module.

The DTS patches for adding PRU nodes and enabling PRU UART will be added
in a later v1 version of the series if accepted.

This driver has been previously tested on the following boards:
am64x SK, am62x SK, and am335x SK boards.

Bin Liu (2):
  dt-bindings: serial: add binding documentation for TI PRUSS UART
  serial: 8250: Add PRUSS UART driver

 .../bindings/serial/ti,pruss-uart.yaml        |  54 +++++
 drivers/tty/serial/8250/8250_pruss.c          | 213 ++++++++++++++++++
 drivers/tty/serial/8250/Kconfig               |  10 +
 drivers/tty/serial/8250/Makefile              |   1 +
 4 files changed, 278 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/serial/ti,pruss-uart.yaml
 create mode 100644 drivers/tty/serial/8250/8250_pruss.c

-- 
2.49.0
Re: [PATCH RFC 0/2] Introduce PRU UART driver
Posted by Andy Shevchenko 9 months, 1 week ago
On Wed, Apr 30, 2025 at 07:31:11PM -0500, Judith Mendez wrote:
> This patch series is sent as an RFC to get some initial comments
> on the PRU UART driver.
> 
> The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
> SoCs have a UART sub-module. This patch series introduces the driver and the
> corresponding binding documentation for this sub-module.
> 
> The DTS patches for adding PRU nodes and enabling PRU UART will be added
> in a later v1 version of the series if accepted.
> 
> This driver has been previously tested on the following boards:
> am64x SK, am62x SK, and am335x SK boards.

The first and main question here: Have you checked the existing zillion of
drivers and become with an idea that none of them not even close to this one
(based on RTL)?

-- 
With Best Regards,
Andy Shevchenko
Re: [PATCH RFC 0/2] Introduce PRU UART driver
Posted by Judith Mendez 9 months ago
Hi Andy,

On 5/2/25 4:38 AM, Andy Shevchenko wrote:
> On Wed, Apr 30, 2025 at 07:31:11PM -0500, Judith Mendez wrote:
>> This patch series is sent as an RFC to get some initial comments
>> on the PRU UART driver.
>>
>> The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
>> SoCs have a UART sub-module. This patch series introduces the driver and the
>> corresponding binding documentation for this sub-module.
>>
>> The DTS patches for adding PRU nodes and enabling PRU UART will be added
>> in a later v1 version of the series if accepted.
>>
>> This driver has been previously tested on the following boards:
>> am64x SK, am62x SK, and am335x SK boards.
> 
> The first and main question here: Have you checked the existing zillion of
> drivers and become with an idea that none of them not even close to this one
> (based on RTL)?
> 

Thanks for reviewing.

I have looked through the drivers in tty/serial/8250 and I believe there
are no drivers we could leverage, especially since this driver has a
dependency on the PRU drivers for clock and power. This PORT_16550A UART
is in PRU module which is specific to TI, so besides simplifying the
driver a bit further, I don't think we can use another driver.

~ Judith
Re: [PATCH RFC 0/2] Introduce PRU UART driver
Posted by Greg Kroah-Hartman 9 months, 1 week ago
On Wed, Apr 30, 2025 at 07:31:11PM -0500, Judith Mendez wrote:
> This patch series is sent as an RFC to get some initial comments
> on the PRU UART driver.
> 
> The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
> SoCs have a UART sub-module. This patch series introduces the driver and the
> corresponding binding documentation for this sub-module.
> 
> The DTS patches for adding PRU nodes and enabling PRU UART will be added
> in a later v1 version of the series if accepted.
> 
> This driver has been previously tested on the following boards:
> am64x SK, am62x SK, and am335x SK boards.

Why is this "RFC"?  What needs to be done to make it something that you
actually feel works properly and should be merged?

thanks,

greg k-h
Re: [PATCH RFC 0/2] Introduce PRU UART driver
Posted by Judith Mendez 9 months, 1 week ago
Hi Greg,

On 5/1/25 12:18 AM, Greg Kroah-Hartman wrote:
> On Wed, Apr 30, 2025 at 07:31:11PM -0500, Judith Mendez wrote:
>> This patch series is sent as an RFC to get some initial comments
>> on the PRU UART driver.
>>
>> The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
>> SoCs have a UART sub-module. This patch series introduces the driver and the
>> corresponding binding documentation for this sub-module.
>>
>> The DTS patches for adding PRU nodes and enabling PRU UART will be added
>> in a later v1 version of the series if accepted.
>>
>> This driver has been previously tested on the following boards:
>> am64x SK, am62x SK, and am335x SK boards.
> 
> Why is this "RFC"?  What needs to be done to make it something that you
> actually feel works properly and should be merged?

Nothing needs to be done IMO, the only reason it was sent as an RFC is
to get initial thoughts/issues that anyone might have with the driver
before sending v1.

If none, I will go ahead and send v1. Thanks for your attention Greg.

~ Judith
Re: [PATCH RFC 0/2] Introduce PRU UART driver
Posted by Andy Shevchenko 9 months, 1 week ago
On Thu, May 01, 2025 at 09:47:34AM -0500, Judith Mendez wrote:
> On 5/1/25 12:18 AM, Greg Kroah-Hartman wrote:
> > On Wed, Apr 30, 2025 at 07:31:11PM -0500, Judith Mendez wrote:
> > > This patch series is sent as an RFC to get some initial comments
> > > on the PRU UART driver.
> > > 
> > > The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
> > > SoCs have a UART sub-module. This patch series introduces the driver and the
> > > corresponding binding documentation for this sub-module.
> > > 
> > > The DTS patches for adding PRU nodes and enabling PRU UART will be added
> > > in a later v1 version of the series if accepted.
> > > 
> > > This driver has been previously tested on the following boards:
> > > am64x SK, am62x SK, and am335x SK boards.
> > 
> > Why is this "RFC"?  What needs to be done to make it something that you
> > actually feel works properly and should be merged?
> 
> Nothing needs to be done IMO, the only reason it was sent as an RFC is
> to get initial thoughts/issues that anyone might have with the driver
> before sending v1.
> 
> If none, I will go ahead and send v1. Thanks for your attention Greg.

I have tons of comments, please read my replies before sending a v1.

-- 
With Best Regards,
Andy Shevchenko
Re: [PATCH RFC 0/2] Introduce PRU UART driver
Posted by Judith Mendez 9 months, 1 week ago
Hi Andy,

On 5/2/25 4:51 AM, Andy Shevchenko wrote:
> On Thu, May 01, 2025 at 09:47:34AM -0500, Judith Mendez wrote:
>> On 5/1/25 12:18 AM, Greg Kroah-Hartman wrote:
>>> On Wed, Apr 30, 2025 at 07:31:11PM -0500, Judith Mendez wrote:
>>>> This patch series is sent as an RFC to get some initial comments
>>>> on the PRU UART driver.
>>>>
>>>> The ICSSM modules on am64x SoC and the PRUSS module on am62 SoC or am335x
>>>> SoCs have a UART sub-module. This patch series introduces the driver and the
>>>> corresponding binding documentation for this sub-module.
>>>>
>>>> The DTS patches for adding PRU nodes and enabling PRU UART will be added
>>>> in a later v1 version of the series if accepted.
>>>>
>>>> This driver has been previously tested on the following boards:
>>>> am64x SK, am62x SK, and am335x SK boards.
>>>
>>> Why is this "RFC"?  What needs to be done to make it something that you
>>> actually feel works properly and should be merged?
>>
>> Nothing needs to be done IMO, the only reason it was sent as an RFC is
>> to get initial thoughts/issues that anyone might have with the driver
>> before sending v1.
>>
>> If none, I will go ahead and send v1. Thanks for your attention Greg.
> 
> I have tons of comments, please read my replies before sending a v1.

Thanks for your comments, I need a day or so to review and then respond
to your comments and Andrews.

Judith

>