Hi Simon,
On Wed, Feb 19, 2020 at 10:21 PM Simon Glass sjg@chromium.org wrote:
Hi Angelo,
On Wed, 19 Feb 2020 at 07:36, Angelo Dureghello angelo.dureghello@timesys.com wrote:
Hi Simon,
after this commit:
commit 29f7d05a347ab7a42577c67fdfb787ef91537302 Author: Simon Glass sjg@chromium.org Date: Sun Dec 29 21:19:17 2019 -0700
dm: core: Move ofdata_to_platdata() call earlier This method is supposed to extract platform data from the device tree. It should be done before the device itself is probed. Move it earlier in the device_probe() function. Signed-off-by: Simon Glass <sjg@chromium.org>stmark2 (mcf5441x based board) hangs silently.
I am debugging into this, since seems to be something wrong in my arch files, or related dm drivers, but if you have any hint about where to look, welcome.
Good to have a JTAG debugger or at least a working DEBUG_UART.
It could be that you have one device relying on another's ofdata, but the reading of it is in the probe() method instead of ofdata_to_platdata().
Also make sure you have proper error checking in these methods since silent failure (e.g. to read an address) can cause invalid data to be used by the driver.
You can also enable CONFIG_LOG_ERROR_RETURN if you are using log_ret() or log_msg_ret() in your drivers.
I can't actually even find the serial driver for m68k so I don't think I will be much help...
Regards, Simon
thanks, so, just for y i. Looks like the issue is that my driver (mcfuart.c) uses "dev->seq" to deduce uart port is 0. And dev->seq after your commit above is now -1 (it was 0).
So, looks like dev->seq can't be used now from _ofdata_to_platdata, where i was assigning
plat->port = dev->seq;
Will fix this in the driver.
Regards, Angelo