On Mon, Feb 20, 2017 at 2:40 PM, Andreas F?rber [off-list ref] wrote:

Am 16.02.2017 um 14:41 schrieb Arnd Bergmann:

quoted

On Wednesday, February 15, 2017 5:55:23 PM CET Andreas F?rber wrote:

quoted

diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
index 6117ac8..9d4213c 100644
--- a/drivers/tty/serial/Kconfig
+++ b/drivers/tty/serial/Kconfig

@@ -1677,6 +1677,25 @@ config SERIAL_MVEBU_CONSOLE
          and warnings and which allows logins in single user mode)
          Otherwise, say 'N'.

+config SERIAL_OWL
+       bool "Actions Semi Owl serial port support"
+       depends on ARCH_OWL || COMPILE_TEST
+       select SERIAL_CORE
+       help
+         This driver is for Actions Semi S500/S900 SoC's UART.
+         Say 'Y' here if you wish to use the on-board serial port.
+         Otherwise, say 'N'.
+

This could maybe use a word of information about how there isn't
actually a uart driver in this series, only an earlycon driver.

Yeah, I have added a sentence to the other one. Maybe also add a select?

I generally only use 'select' to avoid build failures, not to save the user
the time to make a choice.

quoted

Do you plan to add the actual tty driver soon? If not, maybe
only add one Kconfig symbol for the earlycon support.

So, I completed the driver the weekend and can see the full console
output, but there are still output glitches, possible locking issues up
to RCU stalls and soft lockups, and only stubbed out baudrate support.

Therefore I would rather keep the full driver as follow-up RFC in v2,
unless I get it fixed during Hackweek. In the earlycon I am confident,
and the only Kconfig change then is to switch from bool to tristate and
to drop the newly added sentence again.

Ok, sounds good.

     Arnd