boards/nucleo: share board.h and board.c#8065
Conversation
3a836fb to
5a3233c
Compare
|
rebased, as #8058 was merged. |
5a3233c to
74bd1b5
Compare
|
Looks good. I have no time to test it today. Let's give Murdock a try. |
|
please also consider #8189 to keep the doc consistent |
|
yes, will adapt the doc |
great! Merging #8189 (before this one), would be nice, too - but it needs approval, first 😉 |
|
will take a look later today |
|
#8189 is merged, can you rebase @haukepetersen ? |
74bd1b5 to
bee0f13
Compare
|
sorry for the delay: adapted doxygen groups to the changes in #8189 and rebased. |
|
Looks good for me. I'll test on a subset of boards next week. |
aabadie
left a comment
There was a problem hiding this comment.
Tested examples/default on a set of nucleo boards: works for nucleo32-l031, nucleo-l152 (when disabling LSE), nucleo-l476, nucleo32-l432, nucleo-f072, nucleo144-f207, nucleo32-f303, nucleo-f446 and nucleo144-f746. I don't have a nucleo-f103 (I lent it) but except this one I tested all nucleo cpu families.
Otherwise, I think I found an issue with doxygen grouping and there's the problem with the AUTO_INIT_LED0 that doesn't apply to nucleo144 boards.
boards/common/nucleo/board.c
Outdated
| */ | ||
|
|
||
| /** | ||
| * @ingroup boards_nucleo |
There was a problem hiding this comment.
If I get it right, it should be boards_common_nucleo
|
|
||
| /* initialization of on-board LEDs | ||
| * NOTE: LED0 must be explicitly enabled as it is also used for SPI_DEV(0) */ | ||
| #ifdef AUTO_INIT_LED0 |
There was a problem hiding this comment.
This is only the case for nucleo32 and nucleo64 boards. nucleo144 boards are using different pins for SPI and on-board leds.
|
Please rebase @haukepetersen |
bee0f13 to
99c847f
Compare
|
rebased and added a fix to always initialize LED0 for Nucleo144 boards. |
|
|
||
| # select OpenOCD configuration depending on CPU type | ||
| ifeq (,$(OPENOCD_CONFIG)) | ||
| ifeq ($(CPU),stm32f0) |
There was a problem hiding this comment.
how about
export OPENOCD_CONFIG := $(RIOTBOARD)/common/nucleo/dist/$(patsubst stm32%,openocd-%.cfg,$(CPU)).
And please add an error case, e.g., err out on if $(,$(filter stm32%,$(CPU))
There was a problem hiding this comment.
sounds good. Ok if I do it in a separate PR?
e8cc937 to
4f1ce14
Compare
|
adapted to #8405 |
|
@kaspar030 @aabadie: you think we could this in soonish? I get tired of maintaining this PR and all of your previous comments should be addressed :-) |
aabadie
left a comment
There was a problem hiding this comment.
Un-retested ACK. The change of auto init led0 should be ok with nucleo-144.
Thanks for rebasing again btw
|
Tested on nucleo-f401. ACK. |
|
finally, thanks a lot for reviewing and testing! |
Based on #8058and cuts out part from #8044As #8044 is getting out of hand, I think it makes sense to split its aspects into separate PRs. This PR does two things:
common/nucleoandcommon/nucleo64board.h(by type, one for 32, 64, and 144) and a single sharedboard.cfileEspecially the diff of the last commit (
175 additions and 2,293 deletions) seems to be a good start :-)