On Mon, 17 Oct 2016, Phil Blundell wrote:
> On Mon, 2016-10-17 at 07:57 -0400, Robert P. J. Day wrote:
> so is it a fair statement that, pedantically, i might as well just
> > define everything in terms of SERIAL_CONSOLES and think that way?
> Yes. SERIAL_CONSOLE predates SERIAL_CONSOLES and remains supported
> in the configuration for backwards compatibility. New
> configurations and recipes should probably use SERIAL_CONSOLES and
> ignore SERIAL_CONSOLE entirely.
i'm good with that, thanks, and one more question related to setting
baud rate for the console port, if i may. what is the precedence order
for the baud rate that is actually used, given that there are a number
of places this can be specified?
i'm building a systemd-based system, and i'm looking at the
do_install() code in systemd-serialgetty.bb, and as i see it, if one
is using u-boot and a device tree, the baud rate can be specified in a
number of places.
1) apparently, a number of boards defined in u-boot can set the baud
rate in their ft_board_setup() routine.
2) u-boot can pass the baud rate on the kernel command line
3) the device tree blob for the target board might have the baud rate
hard-coded in the DTB
4) using SERIAL_CONSOLES
am i missing anything? if i follow the code, i'm sure i can puzzle
it out eventually.
Robert P. J. Day Ottawa, Ontario, CANADA
Openembedded-core mailing list