Re: Porting Zephyr to STM32F103C8T6 Minimum System Development Board
toggle quoted messageShow quoted text
Thanks for your speedy response, that was very helpful. After adding the pin mux settings, I was able to get the UART to work.
The part that I am worried about is,
diff --git a/ext/hal/st/stm32cube/stm32f1xx/soc/stm32f1xx.h b/ext/hal/st/stm32cube/stm32f1xx/soc/stm32f1xx.h
index 333095b..8dc37db 100644
@@ -145,7 +145,7 @@
+#elif defined(STM32F103xB) || defined(STM32F103x8)
Since, I couldn't find a stm32f103x8.h (ST's own sdk ships with an stm32f10xxx.h), I am just using stm32f103xb.h instead. Is this acceptable?. For now things look okay, but will have to validate this a little further to know for sure. Once I document the board, will send the PR.
It seems your porting miss the pinmux configuration driver file, cf:
Then, you'll have to create a new CONFIG value "SOC_STM32F103X8". We need to identify stm32 SoC variants with the last letter in order to keep the Flash size information.
You'll use this value to fill flash/ram in in dts/arm/st/mem.h (please rebase if you don't see it).
Regarding upstream; yes, we'll be please to have this board supported!
On 14 May 2017 at 23:20, Siddharth Chandrasekaran <siddharth@...>
I am trying to port Zephyr into one of the cheap, breadboard friendly, "STM32F103C8T6 Minimum System Development Boards" (link not attached due to overwhelming results in Google) and I am facing some difficulty.
So far, with the other ported boards as a reference, I was able to setup a new board and get the blinky sample application to work. However, I am unable to get console logs to show up in UART_1 as expected. I have tried the hello_world and shell applications without much luck.
Digging a little deeper, I realised that the SOC support had STM32103XB and there isn't a STM32103X8 or a STM32103XX. Now I am beginning now wondering if there is a reason why the SOC was called STM32103XB and not STM32103XX, as the data sheet for the two of them is unified. I also verified that the peripheral register address map for UART 1 is in fact at the the same location as in the STM32103XB, which means at least UART 1 should work even if the SOCs aren't compatible.
Here is a patch to my work so far http://embedjournal.com/board_stm32_min_dev.patch
. Do let me know If I am missing something. Also, is this a hardware that the upstream will be interested in? It's a pretty popular design although there are slight incistencies. I can raise a PR after successful validation if it means anything.
Zephyr-devel mailing list
Join firstname.lastname@example.org to automatically receive all group messages.