cpu: add stm32f2xx family support (based on #4497) by vincent-d · Pull Request #5397 · RIOT-OS/RIOT

vincent-d · 2016-04-26T13:37:03Z

Hi,

This pull request adds support for stm32f2xx family.
It's based on #4497, thanks @DipSwitch for the initial work.
The last commit adds the nucleo-f207 board based on stm32f207.

* Update headers from ST * Add linker scripts

The driver was imported from stm32f0 Fixes: - Fix 12/24h handling (CR flag badly used) - Fix interrupt flag clearance and interrupt name

Add missing DMA interrupts for UART Improve baudrate intialization: - Return error if baudrate is theorically unreachable - Implement oversampling by 8 method for high baudrates Add UART hardware flow control support Ensure uart tx thread safety with a mutex Allow setting of pins mode per UART

Remove unneeded workaround from stm32f1 Make pullup configurable Fix OAR1 register initialization improve i2c read functions with repeatead start conditions avoid to any loop to become infinite improve i2c driver error handling add missing coma in array initializers

DipSwitch · 2016-04-27T09:27:36Z

There is an updated version in de EWM... PR

toonst · 2016-04-28T14:09:30Z

@DipSwitch I had a look at the EWM PR, but it is a bit hard to see which parts are for the cpu and which are for the board. Also it seems to incorporate a lot of merges from other PRs. Can you indicate which parts are missing?

We tested this PR extensively on the nucleo-f207 board (and on our own HW) so I think it is ready for a merge.

vincent-d · 2016-05-10T10:38:19Z

What has to be done to have an implementation of the stm32f2 into RIOT ?
What is your feedback on this PR and what exactly prevents it from being merged ?

miri64 · 2016-05-10T11:14:03Z

Sorry, a significant portion of the RIOT community that usually looks into this kind of PRs is currently occupied with day job work. This should be finished by tonight.

Optional tip: I'm no expert on porting but usually it helps to split up a PR like this into smaller ones, so people aren't overwhelmed by the sheer number of code lines (34700 is quiet a lot).

vincent-d · 2016-05-10T11:53:59Z

@authmillenon thanks for the reply.

I think the high number of added lines comes mainly from the register header files update.

kaspar030 · 2016-05-10T12:41:57Z

boards/nucleo-f207/include/board.h

+#undef LED0_ON
+#undef LED0_OFF
+#undef LED0_TOGGLE
+#define LED0_PIN            GPIO_PIN(PORT_B, 0)


Could you please group LEDx_* together?

toonst · 2016-05-31T09:08:27Z

Hi guys, can we get an update on this?

thomaseichinger · 2016-06-29T13:27:10Z

cpu/stm32_common/include/periph_cpu_common.h

+enum {
+    AHB1,           /**< AHB1 bus */
+    AHB2,           /**< AHB2 bus */
+    AHB3            /**< AHB3 bus */


I'd move a change here to a separate PR and use it in all other stm32s too then.

You highlighted a good point. Actually, the number of AHB is not the same for all stm32 CPUs. This 'enum' should be part of the specific periph_cpu.h.

toonst · 2016-07-18T16:12:14Z

@haukepetersen Any idea when you will have time for reviewing this? We'd like to have it inside the next release.

kaspar030 · 2016-07-18T16:23:23Z

tests/periph_uart/main.c

 #include "msg.h"
 #include "ringbuffer.h"
 #include "periph/uart.h"
+#include "uart_stdio.h"


what's the reason for this?

This makes sure UART_STDIO_DEV is defined, so the init command fails when trying to use the shell UART as a parameter.
Maybe there's a better way to do this?
Edit: removed reference to local commit sha which was squashed

kaspar030 · 2016-07-18T16:27:17Z

@toonst @haukepetersen is in paternity leave until September.

@emmanuelsearch Could you find out if anyone at FU has this board available?

kaspar030 · 2016-07-18T16:28:04Z

I get the following when compiling examples/default:

"make" -C /home/kaspar/src/riot.tmp/sys/shell/commands
dac.c: In function 'dac_set':
dac.c:76:19: error: array subscript is above array bounds [-Werror=array-bounds]
     if (dac_config[line].chan) {
         ~~~~~~~~~~^~~~~~
dac.c: In function 'dac_poweron':
dac.c:91:38: error: array subscript is above array bounds [-Werror=array-bounds]
     DAC->CR |= (1 << (16 * dac_config[line].chan));
                            ~~~~~~~~~~^~~~~~
dac.c: In function 'dac_poweroff':
dac.c:96:39: error: array subscript is above array bounds [-Werror=array-bounds]
     DAC->CR &= ~(1 << (16 * dac_config[line].chan));
                             ~~~~~~~~~~^~~~~~
cc1: all warnings being treated as errors

gcc version 6.1.1 20160526 (Arch Repository)

kaspar030 · 2016-07-18T16:30:35Z

if anyone at FU has this board available?

If no maintainer has one available, I'd say, once everything compiles, let's merge anyways. Code-wise the PR looks very nice.

this is copied from DipSwitch's pr

toonst · 2016-07-19T12:44:35Z

adressed in 378a7a5. Thanks for reviewing!

toonst · 2016-07-19T14:22:28Z

I cleaned the history with rebase, should be ready for a merge now.

kYc0o · 2016-07-22T14:10:40Z

The commits seem a bit messy... is it possible to squash them and order them by type? for instance all the uart commits in a single one, same for i2c, spi, etc...

changes taken from DipSwitch's board: EMW3162 RIOT-OS#4498 PR. - Adds a functional implementation of the ADC - Implements low power mode

This board is based on a 144-pin stm32f207 cortex-m3

toonst · 2016-07-25T09:03:51Z

@kYc0o addressed. Also rebased onto latest master.

kYc0o · 2016-07-25T12:20:17Z

Thanks @toonst ! But sadly it will go into the next release... But as it's ready to merge it will go fast ;)

vincent-d · 2016-08-22T15:20:37Z

Hi guys, is there any chance this PR could be merged soon ?

kaspar030 · 2016-08-22T15:30:08Z

Compiles fine, looks OK. How do we proceed?

kaspar030 · 2016-08-22T15:35:40Z

I think we should merge this, if @vincent-d promises to help with basic maintenance of the port.

(merging code that maintainers cannot test is a more general discussion, I'll try to find consensus at devel@)

vincent-d · 2016-08-22T15:48:06Z

@kaspar030 of course I can help with the maintenance of the port :)

DipSwitch · 2016-08-22T15:50:25Z

I could also help when I have more spare time. :)

kaspar030 · 2016-08-24T10:06:34Z

I think we should merge this

Noone seems to disagree. ACK when Murdock is fine.

kaspar030 · 2016-08-24T17:16:25Z

Murdock complains about extra whitespaces in the vendor headers. Could you add those files to the doxygen excludes?

vincent-d · 2016-08-25T12:04:30Z

Vendor headers excluded from doxygen

kaspar030 · 2016-08-31T19:14:34Z

Well, it's the static tests that complain. But we don't care, so go!

DipSwitch and others added 7 commits December 13, 2015 11:00

stm32f2: add initial support for stm32f2

4064858

stm32f2: Update with last api changes

15b6814

* Update headers from ST * Add linker scripts

stm32f2: Add RTC driver

ab8f9f1

The driver was imported from stm32f0 Fixes: - Fix 12/24h handling (CR flag badly used) - Fix interrupt flag clearance and interrupt name

stm32f2: Update hwrng driver

d01da27

stm32f2: update timer and pwm drivers

70c8bff

PeterKietzmann assigned haukepetersen Apr 26, 2016

kaspar030 reviewed May 10, 2016
View reviewed changes

toonst mentioned this pull request Jun 29, 2016

stm32f2: Add DMA driver #5584

Closed

thomaseichinger reviewed Jun 29, 2016
View reviewed changes

kaspar030 reviewed Jul 18, 2016
View reviewed changes

kaspar030 self-assigned this Jul 18, 2016

kaspar030 added this to the Release 2016.07 milestone Jul 18, 2016

stm32f2: fix dac compilation

7a0e819

this is copied from DipSwitch's pr

toonst force-pushed the pr/stm32f2xx branch from 378a7a5 to 94123d0 Compare July 19, 2016 14:19

toonst mentioned this pull request Jul 25, 2016

tests/periph_uart: include uart_stdio.h #5686

Merged

Toon Stegen and others added 2 commits July 25, 2016 10:34

stm32f2: add adc and lpm

77c9ee0

changes taken from DipSwitch's board: EMW3162 RIOT-OS#4498 PR. - Adds a functional implementation of the ADC - Implements low power mode

board: Add nucleo-f207 board

146d462

This board is based on a 144-pin stm32f207 cortex-m3

toonst force-pushed the pr/stm32f2xx branch 3 times, most recently from e7f67da to 146d462 Compare July 25, 2016 09:02

kYc0o modified the milestones: Release 2016.10, Release 2016.07 Jul 25, 2016

kYc0o mentioned this pull request Jul 25, 2016

cpu: Initial import of the STM32F2xx family #4497

Closed

kaspar030 added the CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR label Aug 24, 2016

stm32f2: exlude vendor headers from doxygen

537d1d0

kaspar030 merged commit 26e4004 into RIOT-OS:master Aug 31, 2016

PeterKietzmann mentioned this pull request Jan 13, 2017

board: EMW3162 #4498

Closed

toonst deleted the pr/stm32f2xx branch August 21, 2018 15:32

Conversation

vincent-d commented Apr 26, 2016

Uh oh!

DipSwitch commented Apr 27, 2016 via email

Uh oh!

toonst commented Apr 28, 2016

Uh oh!

vincent-d commented May 10, 2016

Uh oh!

miri64 commented May 10, 2016

Uh oh!

vincent-d commented May 10, 2016

Uh oh!

kaspar030 May 10, 2016

Choose a reason for hiding this comment

Uh oh!

toonst May 11, 2016

Choose a reason for hiding this comment

Uh oh!

toonst commented May 31, 2016

Uh oh!

thomaseichinger Jun 29, 2016

Choose a reason for hiding this comment

Uh oh!

AurelienGONCE Jul 5, 2016

Choose a reason for hiding this comment

Uh oh!

toonst commented Jul 18, 2016

Uh oh!

kaspar030 Jul 18, 2016

Choose a reason for hiding this comment

Uh oh!

toonst Jul 18, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kaspar030 commented Jul 18, 2016

Uh oh!

kaspar030 commented Jul 18, 2016

Uh oh!

kaspar030 commented Jul 18, 2016

Uh oh!

toonst commented Jul 19, 2016

Uh oh!

toonst commented Jul 19, 2016

Uh oh!

kYc0o commented Jul 22, 2016

Uh oh!

toonst commented Jul 25, 2016

Uh oh!

kYc0o commented Jul 25, 2016

Uh oh!

vincent-d commented Aug 22, 2016

Uh oh!

kaspar030 commented Aug 22, 2016

Uh oh!

kaspar030 commented Aug 22, 2016

Uh oh!

vincent-d commented Aug 22, 2016

Uh oh!

DipSwitch commented Aug 22, 2016 via email

Uh oh!

kaspar030 commented Aug 24, 2016

Uh oh!

kaspar030 commented Aug 24, 2016

Uh oh!

vincent-d commented Aug 25, 2016

Uh oh!

kaspar030 commented Aug 31, 2016

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

11 participants

toonst Jul 18, 2016 •

edited

Loading