2021-12-20 16:12:28

by Arnaud Pouliquen

[permalink] [raw]
Subject: [PATCH v2] ARM: multi_v7_defconfig: Enable CONFIG_RPMSG_TTY

The RPMsg TTY implements an inter-processor communication with a standard
TTY interface on top of the RPMsg framework.
This driver is a generic RPMsg client that can run on different platforms.

By enabling the RPMSG_TTY driver as module in multi_v7_defconfig, it makes
possible to automatically probe the rpmsg_tty driver by the RPMsg bus,
when the support of the RPMsg service is dynamically requested by the
co-processor firmware.

Signed-off-by: Arnaud Pouliquen <[email protected]>
---
delta vs V1:
add
---
arch/arm/configs/multi_v7_defconfig | 1 +
1 file changed, 1 insertion(+)

diff --git a/arch/arm/configs/multi_v7_defconfig b/arch/arm/configs/multi_v7_defconfig
index c951aeed2138..370ca4ae3981 100644
--- a/arch/arm/configs/multi_v7_defconfig
+++ b/arch/arm/configs/multi_v7_defconfig
@@ -372,6 +372,7 @@ CONFIG_SERIAL_ST_ASC_CONSOLE=y
CONFIG_SERIAL_STM32=y
CONFIG_SERIAL_STM32_CONSOLE=y
CONFIG_SERIAL_OWL=y
+CONFIG_RPMSG_TTY=m
CONFIG_SERIAL_DEV_BUS=y
CONFIG_VIRTIO_CONSOLE=y
CONFIG_ASPEED_KCS_IPMI_BMC=m
--
2.17.1



2021-12-20 16:58:32

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: [PATCH v2] ARM: multi_v7_defconfig: Enable CONFIG_RPMSG_TTY

On Mon, Dec 20, 2021 at 05:11:55PM +0100, Arnaud Pouliquen wrote:
> The RPMsg TTY implements an inter-processor communication with a standard
> TTY interface on top of the RPMsg framework.
> This driver is a generic RPMsg client that can run on different platforms.
>
> By enabling the RPMSG_TTY driver as module in multi_v7_defconfig, it makes
> possible to automatically probe the rpmsg_tty driver by the RPMsg bus,
> when the support of the RPMsg service is dynamically requested by the
> co-processor firmware.
>
> Signed-off-by: Arnaud Pouliquen <[email protected]>
> ---
> delta vs V1:
> add
> ---

"add"? What does that mean?

Do all v7 chips have this hardware?

thanks,

greg k-h

2021-12-21 09:06:33

by Arnaud Pouliquen

[permalink] [raw]
Subject: Re: [PATCH v2] ARM: multi_v7_defconfig: Enable CONFIG_RPMSG_TTY



On 12/20/21 5:58 PM, Greg Kroah-Hartman wrote:
> On Mon, Dec 20, 2021 at 05:11:55PM +0100, Arnaud Pouliquen wrote:
>> The RPMsg TTY implements an inter-processor communication with a standard
>> TTY interface on top of the RPMsg framework.
>> This driver is a generic RPMsg client that can run on different platforms.
>>
>> By enabling the RPMSG_TTY driver as module in multi_v7_defconfig, it makes
>> possible to automatically probe the rpmsg_tty driver by the RPMsg bus,
>> when the support of the RPMsg service is dynamically requested by the
>> co-processor firmware.
>>
>> Signed-off-by: Arnaud Pouliquen <[email protected]>
>> ---
>> delta vs V1:
>> add
>> ---
>
> "add"? What does that mean?

Oops, sorry, part of my sentence has disappeared
"Add more description in commit message to detail the objective ".

>
> Do all v7 chips have this hardware?

I would say yes and no

No - As mainly used today for hardware system integrating a main and a
coprocessor in a single chip.

Yes - as that depends on the RPMsg back-end

The RPMsg is divided in 3 layers
- the service layer that implement a service on top of the RPmsg
A parallel could be done between this layer and the Virtio devices such
as the "Virtio-Console".
=> the rpmsg _tty is part of this layer
=> rpmsg_char is another one ( already enabled in some configs)

- the transport layer in charge of the RPmsg protocol
- the back-end layer that is the hardware abstraction layer
- The main back-end used is the RPMsg VirtIO back-end which implements
the RPMSg over VirtIO.
- There are also some platform specific backends (e.g. QCOM, MediaTek)
- We have also some demos that implement the RPMsg protocol on top of a
serial link back-end allowing communication with an external
processor. Not up-streamed yet.

Thanks,
Arnaud

>
> thanks,
>
> greg k-h
>