boards/nucleo-l1: use LSI as default RTC clock by haukepetersen · Pull Request #6504 · RIOT-OS/RIOT

haukepetersen · 2017-01-27T13:51:40Z

fixes #6502
~~rebased on #6499 (needed for testing)~~

seems like not all nucleo-l1 boards have an external low-speed oscillator (LSE) attached. So I propose to make it configurable in the board's periph_conf and choose the LSI (internal low-speed oscillator) per default.

vincent-d · 2017-01-27T14:07:17Z

The problem of the LSI is its accuracy. But it works if application does not require accuracy for long periods I guess.

However, I took a new look to the RTC application note. It seems that the stm32l1 LSI is a 37kHz oscillator, then the prescaler register needs to be adapted. (AN3371, part 1.1.2, page 9)

PeterKietzmann · 2017-01-27T15:59:41Z

I'm on current master (includes #6499) without this PR running tests/periph_rtc. The output looks as expected:

2017-01-27 16:53:49,806 - INFO # RIOT RTC low-level driver test
2017-01-27 16:53:49,811 - INFO # This test will display 'Alarm' in 10 seconds
2017-01-27 16:53:49,811 - INFO # 
2017-01-27 16:53:49,812 - INFO # Initializing the RTC driver
2017-01-27 16:53:49,817 - INFO # Setting clock to 2011-12-13 14:15:15
2017-01-27 16:53:49,823 - INFO # Clock set to 2011-12-13 14:15:15
2017-01-27 16:53:49,824 - INFO # Setting alarm to 2011-12-13 14:15:25
2017-01-27 16:53:49,829 - INFO # Alarm set to 2011-12-13 14:15:25
2017-01-27 16:53:49,830 - INFO # The alarm should trigger every 10 seconds for 4 times.
2017-01-27 16:53:59,825 - INFO # Alarm!
2017-01-27 16:54:09,821 - INFO # Alarm!
2017-01-27 16:54:19,823 - INFO # Alarm!
2017-01-27 16:54:29,825 - INFO # Alarm!

So does with this PR.. Do we actually need it?

pwillem · 2017-01-27T17:40:40Z

cpu/stm32_common/periph/rtc.c

+#else
+    /* Enable the LSI clock */
+    RCC->CSR = RCC_CSR_LSION;
+    while (!(RCC->CSR & RCC_CSR_LSIRDY)) {}


For a F0 this didn't work. The LSI is off after each reboot but (RCC->BDCR & RCC_BDCR_RTCEN) == 1 so rtc_poweron() isn't called. Adding these 2 lines to clk_init() in cpu.c solved it for me

will look into this.

haukepetersen · 2017-01-28T13:26:01Z

@PeterKietzmann: probably your board is equipped with an external low-speed oscillator, so it works on master. But mine isn't, so it doesnt work on master -> ergo: this PR is needed.

vincent-d · 2017-01-30T16:45:28Z

I proposed some more fixes in haukepetersen#37

prescaler values updated
LSI for stm32fx

But I don't really like the LSI as default...

kYc0o · 2017-01-31T22:02:45Z

Confirmed that it works for nucleo-l152, but it breaks on nucleo-f091. Should this bug be postponed?

PeterKietzmann · 2017-01-31T22:32:02Z

I'd say yes. The release should go out tomorrow. Testing has mostly passed and I already added this issue to known issues in the preliminary (still invisible) release notes. Will take care of all 2017.01 labelled PRs tomorrow.

vincent-d · 2017-02-01T14:57:49Z

@kYc0o weird, I tested on a nucleo-f091 here and it worked :/

Anyway, I agree it should be postponed.

kYc0o · 2017-02-02T11:44:13Z

Well, mine gets stuck in setting the clock:

2017-02-02 12:42:52,413 - INFO # main(): This is RIOT! (Version: 2017.04-devel-8-gd1df5-snake.lan-pr/stm32l_lsi)
2017-02-02 12:42:52,413 - INFO # 
2017-02-02 12:42:52,415 - INFO # RIOT RTC low-level driver test
2017-02-02 12:42:52,420 - INFO # This test will display 'Alarm' in 10 seconds
2017-02-02 12:42:52,420 - INFO # 
2017-02-02 12:42:52,421 - INFO # Initializing the RTC driver
2017-02-02 12:42:52,425 - INFO # Setting clock to 2011-12-13 14:15:15

Same question: Which revision of nucleo-f091 do you have? does it have an external oscillator?

vincent-d · 2017-02-02T13:41:10Z

My nucleo-f091 has an external oscillator. I've just retested on it (with #define CLOCK_HAS_LSE (0) ) and I got the same issue, it gets stuck.

On a nucleo-f042 without LSE, it works fine.

vincent-d · 2017-02-02T13:48:27Z

This is not a RTC issue, this is a timer configuration issue. It get stuck in xtimer_usleep.

vincent-d · 2017-02-02T13:49:34Z

Seems related to #6494

kYc0o · 2017-02-02T14:07:45Z

Oh yeah! We need to merge first #6494 and then this will work for both boards. I'll add the Waiting for other PR flag.

I rebased this one with #6494 and everything works for both nucleo-l1 and nucleo-f091 boards.

kYc0o · 2017-02-07T13:47:36Z

@haukepetersen can you squash and optionally rebase to include #6494 ? Then I think this is ready.

PeterKietzmann · 2017-02-22T09:25:16Z

Ping @haukepetersen

aabadie · 2017-02-27T17:29:31Z

any update on this one @haukepetersen ?

haukepetersen · 2017-03-31T08:57:17Z

Yes, that makes sense. One more reason to handle the clock LS(I/E) clock only at one place in the code.

haukepetersen · 2017-03-31T08:58:13Z

I am looking into refining the RTT/RTC interfaces right now, and will in the same run look into a way to handle this for all STM32 CPUs

kYc0o · 2017-04-25T17:28:03Z

Can we merge this one for this release? I think that as long as the bug is solved, we can wait for the rework on this interface and merge it for the next release.

AFAIK this PR is a working state and can be merged asap (besides the need of squash).

haukepetersen · 2017-04-26T07:22:18Z

nope, won't merge this in the current state. Have a far better solution to the problem now (based on some clock initialization restructuring as done in #6907), so I will adapt this PR accordingly.

kYc0o · 2017-04-26T10:17:52Z

And you think we can get it for the release? That's only my point, that is broken and it will stay broken for this release...

haukepetersen · 2017-06-01T16:25:51Z

the solution as in this PR is sub-optimal, found something better by now: the rtc module should utilize stmclk_enable_lfclk() from newly introduced stm32_common/include/stmclk.h interface. The problem: this is not implement for many STM family members yet -> so once this is done, everything in this PR falls into place. Opened #7123 for tracking the implementation progress.

haukepetersen · 2017-06-01T17:35:54Z

pushed an update to the rtc driver using the stmclk interface -> much easier to the eye... But as said: all WIP as of now...

haukepetersen · 2017-08-23T13:37:51Z

closing in favor of #7504

kYc0o · 2018-01-11T17:22:52Z

#8024 led me here... So, I propose to enable the internal oscillator only for the nucleo-l152 board, since anyways nobody will use it for accurate RTC and all stm32l152 have LSI. Should I open a PR for that small change? I tested it on a board with external oscillator and works correctly with LSI set.

haukepetersen added the Type: bug The issue reports a bug / The PR fixes a bug (including spelling errors) label Jan 27, 2017

haukepetersen added this to the Release 2017.01 milestone Jan 27, 2017

haukepetersen assigned vincent-d, aabadie and PeterKietzmann Jan 27, 2017

haukepetersen mentioned this pull request Jan 27, 2017

cpu/stm32l1: RTC broken #6502

Closed

pwillem reviewed Jan 27, 2017

View reviewed changes

PeterKietzmann added the CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR label Jan 31, 2017

miri64 added the CI: needs squashing Commits in this PR need to be squashed; If set, CI systems will mark this PR as unmergable label Jan 31, 2017

PeterKietzmann modified the milestones: Release 2017.01, Release 2017.04 Feb 1, 2017

kYc0o added State: waiting for other PR State: The PR requires another PR to be merged first and removed State: waiting for other PR State: The PR requires another PR to be merged first labels Feb 2, 2017

kaspar030 modified the milestones: Release 2017.07, Release 2017.04 Apr 21, 2017

kYc0o modified the milestones: Release 2017.04, Release 2017.07 Apr 25, 2017

kYc0o mentioned this pull request Apr 25, 2017

RTC alarm triggered with 1s delay #6331

Closed

kaspar030 modified the milestones: Release 2017.07, Release 2017.04 May 8, 2017

haukepetersen and others added 3 commits June 1, 2017 18:16

cpu/stm32: add option to run RTC with LSI

fe52aba

boards/nucleo-l1: run RTC with LSI

ca4055a

cpu/stm32_common: add LSI option with RTC driver

492dc87

haukepetersen added State: WIP State: The PR is still work-in-progress and its code is not in its final presentable form yet and removed CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR labels Jun 1, 2017

haukepetersen mentioned this pull request Jun 1, 2017

cpu: implement the stmclk interface for all STM32 CPUs #7123

Closed

9 tasks

WIP

cf0da44

haukepetersen force-pushed the fix_stm32l1_rtc_lse branch from 3fa0951 to cf0da44 Compare June 1, 2017 17:35

kYc0o modified the milestones: Release 2017.10, Release 2017.07 Jun 30, 2017

haukepetersen mentioned this pull request Aug 23, 2017

cpu/stm32: reworked RTC implementation #7504

Merged

8 tasks

haukepetersen closed this Aug 23, 2017

haukepetersen deleted the fix_stm32l1_rtc_lse branch August 23, 2017 13:37

miri64 added the CI: needs squashing label Oct 5, 2018

Conversation

haukepetersen commented Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vincent-d commented Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

PeterKietzmann commented Jan 27, 2017

Uh oh!

pwillem Jan 27, 2017

Choose a reason for hiding this comment

Uh oh!

haukepetersen Jan 28, 2017

Choose a reason for hiding this comment

Uh oh!

haukepetersen commented Jan 28, 2017

Uh oh!

vincent-d commented Jan 30, 2017

Uh oh!

kYc0o commented Jan 31, 2017

Uh oh!

PeterKietzmann commented Jan 31, 2017

Uh oh!

vincent-d commented Feb 1, 2017

Uh oh!

kYc0o commented Feb 2, 2017

Uh oh!

vincent-d commented Feb 2, 2017

Uh oh!

vincent-d commented Feb 2, 2017

Uh oh!

vincent-d commented Feb 2, 2017

Uh oh!

kYc0o commented Feb 2, 2017

Uh oh!

kYc0o commented Feb 7, 2017

Uh oh!

PeterKietzmann commented Feb 22, 2017

Uh oh!

aabadie commented Feb 27, 2017

Uh oh!

haukepetersen commented Mar 31, 2017

Uh oh!

haukepetersen commented Mar 31, 2017

Uh oh!

kYc0o commented Apr 25, 2017

Uh oh!

haukepetersen commented Apr 26, 2017

Uh oh!

kYc0o commented Apr 26, 2017

Uh oh!

haukepetersen commented Jun 1, 2017

Uh oh!

haukepetersen commented Jun 1, 2017

Uh oh!

haukepetersen commented Aug 23, 2017

Uh oh!

kYc0o commented Jan 11, 2018

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

haukepetersen commented Jan 27, 2017 •

edited

Loading

vincent-d commented Jan 27, 2017 •

edited

Loading