boards: slwstk6000b: add support #8824

basilfx · 2018-03-22T21:21:23Z

Contribution description

This PR add supports for the Silicon Labs SLWSTK6000B development kit for wireless applications. Although the files are very similar to the other PRs I have submitted, this board is actually a base board that can host different CPU modules. All CPU modules have fixed pins for certain functions, including ones for the onboard LCD and Si7021 temperature sensor.

I came up with an idea similar to how I supported the EFM32 family, by using modules.txt as a 'database' for mapping the base board to a CPU module. I don't think duplicating this board for all the different modules would work out.

To compile, you actually have to override the BOARD_MODULE (defaults to SLWBR4162a) using something like this:

BOARD=slwstk6000b BOARD_MODULE=slwrb4162a make -j8

I currently have two CPU modules supported: SLWBR4150A and SLWBR4162A. However, there are 15+ more modules that fit this base board.

Issues/PRs references

#8585
#8520
#8630
#8569

basilfx · 2018-03-22T21:22:17Z

Asking @kaibeckmann to test this PR, since he has access to the boards.

kaibeckmann · 2018-03-23T16:17:35Z

Hey I found a couple of things.
For the two modules I tested, the cpu was wrong. Therefore the right header files were missing:

efr32mg1p233f256gm48.h
efr32mg12p332f1024gl125.h

I like the submodule system, but I think the new definition for the pins is not necessary. First the names are derived from the slwrb4162a, but the datesheets of the other modules have different names for the pins with the same GPIO. That is quite confusing. And than it's not necessary, because the GPIOs seem to be the same (at least for the 3 or 4 datasheets I checked).

The main difference of the modules are the cpu and additional periphery, like SPI-Flash, Touch-Slider, radio transceiver and if there are separate GPIOs to enable sensors, dispay etc.

basilfx · 2018-03-25T15:58:53Z

Thanks @kaibeckmann for testing.

The mapping of modules to CPU is based on the list on the page of the SLWSTK6000B itself.

However, looking at the reference manuals, I do see different MCU's being used. I'll ask Silicon Labs to update their list, somehow.

Regarding the pin numbers: it is needed. If you look at the reference manuals, you can see that some MCU pins map to different module pins. For instance, the brd4150a maps P0 to PA2, but it maps P0 to PD8 on the brd4162a. Then there are also a lot of pins that aren't connected on some boards.

The pin numbers are confusing at first, because they don't link to the reference manuals, where you can see the actual module pins. Here are some:

kaibeckmann · 2018-03-26T14:30:17Z

Hmm I see your point, but the module pin names are confusing. It makes the understanding and using of the boards for someone new more complex.

Maybe a more elaborate name schema would help? Something like MODULE_PIN_P0 ?

If the MCU pin for the basic peripherals really can change, than the multiplexing part should properly be moved to the module header as well (UART, I2C, SPI, ...).

basilfx · 2018-03-26T14:43:23Z

I can change it to MODULE_PIN_P0. That is a better name, and I will also put in the docs that it concerns the module pinout.

I haven't encountered a reason for moving the multiplexing part as well, but if it is necessary, I'll do it. I want to keep as much as possible in board.h and periph_conf.h.

basilfx · 2018-03-26T21:41:13Z

I have fixed the board name to module mapping, based on the reference manuals and renamed the pin names.

kaibeckmann

I think I should have pressed the "submit review" button ...
Some smaller issues / bugs.

kaibeckmann · 2018-03-23T15:45:41Z

boards/slwstk6000b/Makefile.include

+
+# setup JLink for flashing
+export JLINK_DEVICE := $(CPU_MODEL)
+include $(RIOTMAKE)/tools/jlink.inc.mk


Debugging does not work for me, as long as the DBG variable is not set to the gdb program.
I had added following lines between line 18 and 19 in my version:

include $(RIOTMAKE)/tools/gdb.inc.mk export DBG:=$(GDB)

I think this is a bug. I compared JLink with OpenOCD, and OpenOCD sets DBG=GDB, but JLink doesn't. Will look into this tonight.

#8853 should address this.

kaibeckmann · 2018-03-23T15:46:57Z

boards/slwstk6000b/include/periph_conf.h

+};
+
+#define I2C_NUMOF           PERIPH_NUMOF(i2c_config)
+#define I2C_0_ISR           isr_i2c1


that is the wrong isr handler, must be isr_i2c0

kaibeckmann · 2018-03-23T15:48:02Z

boards/slwstk6000b/include/periph_conf.h

+    {
+        .dev = USART0,
+        .rx_pin = F6,
+        .tx_pin = F7,


RX and TX are swaped.

kaibeckmann · 2018-03-23T16:22:53Z

boards/slwstk6000b/Makefile.include

+include $(RIOTMAKE)/tools/serial.inc.mk
+
+# setup JLink for flashing
+export JLINK_DEVICE := $(CPU_MODEL)


the $(CPU_MODEL) does not work for the EFR32MG1P family. For these the string EFR32MG1PxxxF256 is needed.

I've added additional data in modules.txt.

basilfx · 2018-04-03T09:54:26Z

@kaibeckmann do the latest adaptions work for you? The debug problem should now be addressed.

kaibeckmann · 2018-04-03T16:24:19Z

OK I tested it for the slwrb4150a and slwrb4162a with the default app and the tests-core. It works fine.

basilfx · 2018-04-04T19:02:56Z

Updated vendor headers per #8873.

kaibeckmann · 2018-04-05T17:09:16Z

OK I ran the tests again, everything is working as expected.

basilfx · 2018-04-05T17:12:18Z

Ping @kYc0o or @jia200x.

kYc0o · 2018-04-05T17:16:27Z

boards/slwstk6000b/modules/slwrb4162a/include/board_module.h

+#define MODULE_PIN_P43      GPIO_PIN(PF, 10)
+#define MODULE_PIN_P44      GPIO_PIN(PF, 11)
+#define MODULE_PIN_P45      GPIO_PIN(PF, 12)
+/** @} */


How different are these definitions to the ones for the other module?

So far as I can tell, different enough. There are modules that only define half of these pins.

I was just wondering if we could merge them in a single file (with #ifdefs), since I see that at least the beginning of the definitions are the same.

I propose to start this way, and change if it turns out to be easier to combine in the future.

I expect that there might be additional features or possibilities, looking at the list of supported modules (bottom three for example).

Ok, please squash.

kYc0o · 2018-04-05T17:17:30Z

cpu/efm32/families/efr32mg12p/vectors.c

@@ -0,0 +1,135 @@
+/*
+ * Copyright (C) 2015-2018 Freie Universität Berlin
+ *


Freie Universität?

Yes, file was copy-pasted from EZR32WG, by @haukepetersen.

basilfx · 2018-04-05T18:04:39Z

Squashed and rebased. Murdock is happy.

basilfx · 2018-04-06T11:01:25Z

@kYc0o @kaibeckmann Good to go?

kYc0o

Good for me, ACK.

basilfx · 2018-04-06T12:00:34Z

Thanks!

basilfx added Platform: ARM Platform: This PR/issue effects ARM-based platforms Area: boards Area: Board ports labels Mar 22, 2018

basilfx requested review from jia200x and kYc0o March 22, 2018 21:21

basilfx force-pushed the feature/efm32_slwstk6200b branch from ad5bfa8 to 5e0e83d Compare March 26, 2018 21:40

basilfx added CI: needs squashing Commits in this PR need to be squashed; If set, CI systems will mark this PR as unmergable CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR labels Mar 26, 2018

kaibeckmann suggested changes Mar 29, 2018

View reviewed changes

basilfx force-pushed the feature/efm32_slwstk6200b branch from 5e0e83d to ae3b910 Compare March 29, 2018 13:03

basilfx mentioned this pull request Mar 29, 2018

dist/tools: jlink: use GDB as default for DBG #8853

Merged

basilfx force-pushed the feature/efm32_slwstk6200b branch from ae3b910 to 1aa560e Compare April 2, 2018 18:09

basilfx mentioned this pull request Apr 3, 2018

pkg: gecko_sdk: update version #8873

Merged

basilfx force-pushed the feature/efm32_slwstk6200b branch from 1aa560e to e53967d Compare April 4, 2018 19:02

kYc0o reviewed Apr 5, 2018

View reviewed changes

basilfx added 4 commits April 5, 2018 19:40

cpu/efm32: efr32mg12p: add vendor headers

d1441be

cpu/efm32: efr32mg1p: extend vendor headers

9937f29

cpu/efm32: efr32mg12p: add support

b40cf30

boards: slwstk6000b: add support

2a8712a

basilfx force-pushed the feature/efm32_slwstk6200b branch from e53967d to 2a8712a Compare April 5, 2018 17:42

kYc0o approved these changes Apr 6, 2018

View reviewed changes

kYc0o merged commit ce43846 into RIOT-OS:master Apr 6, 2018

basilfx deleted the feature/efm32_slwstk6200b branch April 12, 2023 21:52

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

boards: slwstk6000b: add support #8824

boards: slwstk6000b: add support #8824

basilfx commented Mar 22, 2018 •

edited

Loading

basilfx commented Mar 22, 2018

kaibeckmann commented Mar 23, 2018

basilfx commented Mar 25, 2018 •

edited

Loading

kaibeckmann commented Mar 26, 2018

basilfx commented Mar 26, 2018

basilfx commented Mar 26, 2018

kaibeckmann left a comment

kaibeckmann Mar 23, 2018

basilfx Mar 29, 2018

basilfx Mar 29, 2018

kaibeckmann Mar 23, 2018

basilfx Mar 29, 2018

kaibeckmann Mar 23, 2018

basilfx Mar 29, 2018

kaibeckmann Mar 23, 2018

basilfx Mar 29, 2018

basilfx commented Apr 3, 2018

kaibeckmann commented Apr 3, 2018

basilfx commented Apr 4, 2018

kaibeckmann commented Apr 5, 2018 •

edited

Loading

basilfx commented Apr 5, 2018

kYc0o Apr 5, 2018

basilfx Apr 5, 2018

kYc0o Apr 5, 2018

basilfx Apr 5, 2018 •

edited

Loading

kYc0o Apr 5, 2018

kYc0o Apr 5, 2018

basilfx Apr 5, 2018

basilfx commented Apr 5, 2018

basilfx commented Apr 6, 2018

kYc0o left a comment

basilfx commented Apr 6, 2018

boards: slwstk6000b: add support #8824

boards: slwstk6000b: add support #8824

Conversation

basilfx commented Mar 22, 2018 • edited Loading

Contribution description

Issues/PRs references

basilfx commented Mar 22, 2018

kaibeckmann commented Mar 23, 2018

basilfx commented Mar 25, 2018 • edited Loading

kaibeckmann commented Mar 26, 2018

basilfx commented Mar 26, 2018

basilfx commented Mar 26, 2018

kaibeckmann left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

basilfx commented Apr 3, 2018

kaibeckmann commented Apr 3, 2018

basilfx commented Apr 4, 2018

kaibeckmann commented Apr 5, 2018 • edited Loading

basilfx commented Apr 5, 2018

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

basilfx Apr 5, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

basilfx commented Apr 5, 2018

basilfx commented Apr 6, 2018

kYc0o left a comment

Choose a reason for hiding this comment

basilfx commented Apr 6, 2018

basilfx commented Mar 22, 2018 •

edited

Loading

basilfx commented Mar 25, 2018 •

edited

Loading

kaibeckmann commented Apr 5, 2018 •

edited

Loading

basilfx Apr 5, 2018 •

edited

Loading