MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs) #4435

bespsm · 2025-07-21T13:18:04Z

lib: add i2c mod to lib
lib: add bind_interrupts macros for async workflow
config: add I2C SDA,SCA pin traits code generation
config: add clock source for the I2C
config: add clock divider for the I2C
config: add I2C BusSpeed configuration
I2C: add blocking API: blocking_read, blocking_write, blocking_write_read
I2C: add async API: async_write, async_read, async_write_read
I2C: add embedded-hal (v0.2) API for blocking & async impl.
I2C: add embedded-hal (v1.0) API for blocking & async impl.
I2C-tests: checks for timer_period & check_clock_rate fn's
mspm0l1306 examples: add I2C blocking & async examples
mspm0l1306 examples: add -O2 optimization due to Flash limitations
mspm0g3507 examples: add I2C blocking & async examples

Looking forward for a feedback: @i509VCB

i509VCB

Still going through this.

embassy-mspm0/src/i2c.rs

embassy-mspm0/src/lib.rs

examples/mspm0g3507/src/bin/i2c.rs

bespsm · 2025-07-24T11:36:03Z

After switching to I2C standard mode, driver start working incorrectly (write-read test of 17 bytes test failed). I will do the research

Iooon · 2025-07-25T04:55:04Z

After switching to I2C standard mode, driver start working incorrectly (write-read test of 17 bytes test failed). I will do the research

Seems very likely that this is because of the wrong timer period. When executing the tests, a specific case with FastMode seems to work but all other cases fail. Probably the new calculation is a little bit wrong. Make sure you really work periods instead of frequencies (or account for that). That was my error at the beginning.

bespsm · 2025-07-28T07:58:14Z

After switching to I2C standard mode, driver start working incorrectly (write-read test of 17 bytes test failed). I will do the research

Seems very likely that this is because of the wrong timer period. When executing the tests, a specific case with FastMode seems to work but all other cases fail. Probably the new calculation is a little bit wrong. Make sure you really work periods instead of frequencies (or account for that). That was my error at the beginning.

This bug is reproducible even with old timer period calculation, I will dig into it more.

Added: this is expected behavior according TI. Blocking API is capable to transfer max 8 bytes in one transaction. It's explicitly mentioned in the same TI example and it shows the same faulty behavior. Async (with interrupts) API works well with different i2c speeds

i509VCB · 2025-08-02T01:15:25Z

Checking in, I assume the driver is still being worked on? (and whether it is working?)

bespsm · 2025-08-02T05:37:47Z

Checking in, I assume the driver is still being worked on? (and whether it is working?)

yep, still working on.

Added: @i509VCB I guess I addressed all issues

- lib: add i2c mod to lib - lib: add `bind_interrupts` mod for async workflow - lib: set SYSOSCBASE as system oscillator - config: add I2C SDA,SCA pin traits code generation - config: add clock source for the I2C - config: add clock divider for the I2C - config: add I2C BusSpeed configuration - I2C: add blocking API: blocking_read, blocking_write, blocking_write_read - I2C: add async API: async_write, async_read, async_write_read - I2C: add embedded-hal (v0.2) API for blocking & async impl. - I2C: add embedded-hal (v1.0) API for blocking & async impl. - I2C-tests: checks for timer_period & check_clock_rate fn's

- mspm0l1306 examples: add I2C blocking & async examples - mspm0l1306 examples: add -O2 optimization due to Flash limitations - mspm0g3507 examples: add I2C blocking & async examples

- i2c-config: automatically defines clock source based on input I2C rate - i2c: proper config functions naming - i2c-examples: adapt to changed API - i2c: save initialization pf cctr register

- blocking API for transfering max 8 bytes - async API has no such limitations

i509VCB · 2025-08-06T05:52:44Z

Implementation looks fine. I'll give the errata sheet a look just in case there is something possibly sneaky there we need to know about.

i509VCB · 2025-08-07T03:40:52Z

embassy-mspm0/src/i2c.rs

+            for byte in chunk {
+                *byte = self.info.regs.controller(0).crxdata().read().value();
+            }


We should verify this does not cause problems with I2C_ERR_08 errata.

@i509VCB I couldn't add errata to the project due to compiling issues. could you point, what exactly should I verify here?

https://www.ti.com/lit/pdf/slaz742, I2C_ERR_08 states:

FIFO Read directly after RXDONE interrupt will cause erroneous data to be read

The workaround listed is:

Wait 2 I2C CLK cycles for the FIFO to guarantee to have the latest data. I2C CLK will
based on the CLKSEL register in the I2C registers.

Steps:

IRQ handler wakes the tasking which is pending

poll_fn closure is run, when RXDONE is set then return from poll_fn.

Check a branch and maybe stop the controller transmission.

Read from the fifo

The time gap between step 1 and 4 is non-deterministic. To properly address the errata we would need to insert a wait for 2 I2C clock cycles. Although giving that a thought we would need to know the core clock speed to issue the correct number of cycles to delay. But that is going to depend on #4445.

I'll open an issue about that errata, as we will need to deal with it later.

i509VCB · 2025-08-07T03:41:28Z

Giving the errata a check:

I2C_ERR_01 - SMBUS related issue. We don't implement SMBUS mode yet so we can deal with that later.
I2C_ERR_02 - Quick command related issue. Only applicable in target mode, so ignore it.
I2C_ERR_03 - I2C peripheral mode cannot wake up device when sourced from MFCLK. Maybe something to keep in mind when we allow entering STOP/STANDBY. Ignore for now.
I2C_ERR_04 - Only relevant in target mode
I2C_ERR_05 - We don't touch active bit during transmission, not applicable.
I2C_ERR_06 - SMBus timeout again, so ignore.
I2C_ERR_07 - We do single CCTR writes. So this is worked around indirectly.
I2C_ERR_08 - Maybe an issue, see comment
I2C_ERR_09 - Target mode again
I2C_ERR_10 - Low power, doesn't need to be done yet.

i509VCB · 2025-08-08T23:37:26Z

Flaky tests... Lets hope another run works.

i509VCB · 2025-08-08T23:51:52Z

@Dirbaio when you can please merge. 5340dk tests are flaking.

bespsm · 2025-08-11T13:02:35Z

@Dirbaio kind reminder to check this one

@Dirbaio when you can please merge. 5340dk tests are flaking.

bespsm · 2025-08-13T07:57:08Z

@i509VCB @Dirbaio I see that more PR get merged recently, Could we try again?

Iooon

Sorry to throws this in now, but I noticed that the tests in the i2c.rs file don't work. If the implementation seems to be correct we should at least change the tests.
Failing tests are:
i2c::tests::ti_calculate_timer_period
i2c::tests::ti_calculate_timer_period_2
i2c::tests::ti_calculate_timer_period_3
i2c::tests::ti_calculate_timer_period_4

bespsm · 2025-08-14T08:40:04Z

Sorry to throws this in now, but I noticed that the tests in the i2c.rs file don't work. If the implementation seems to be correct we should at least change the tests. Failing tests are: i2c::tests::ti_calculate_timer_period i2c::tests::ti_calculate_timer_period_2 i2c::tests::ti_calculate_timer_period_3 i2c::tests::ti_calculate_timer_period_4

fixed and tested with oscillator

bespsm · 2025-08-14T08:43:07Z

@i509VCB @Dirbaio I see that more PR get merged recently, Could we try again?

kind reminder for another attempt 🥲

i509VCB · 2025-08-14T16:18:47Z

I'll open a pull request to disable the 5340 tests, as those seem to have broken. I can only use the merge queue myself.

bespsm force-pushed the mspm0-i2c branch from 61a87a1 to b3112c1 Compare July 21, 2025 13:49

i509VCB reviewed Jul 21, 2025

View reviewed changes

embassy-mspm0/src/lib.rs Outdated Show resolved Hide resolved

i509VCB reviewed Jul 21, 2025

View reviewed changes

examples/mspm0g3507/src/bin/i2c.rs Show resolved Hide resolved

bespsm force-pushed the mspm0-i2c branch 3 times, most recently from 669590c to f0eb466 Compare July 23, 2025 15:39

bespsm force-pushed the mspm0-i2c branch from fa81d3a to 659ccba Compare July 29, 2025 14:47

bespsm changed the title ~~MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs)~~ [WIP] MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs) Jul 30, 2025

bespsm mentioned this pull request Aug 1, 2025

peripheral: add fifo size field for the relevant peripherals mspm0-rs/mspm0-data#9

Merged

bespsm force-pushed the mspm0-i2c branch from 659ccba to 9275b43 Compare August 4, 2025 08:16

bespsm added 9 commits August 4, 2025 10:19

mspm0: Add I2C Controller examples for mspm0l1306, mspm0g3507 MCUs

f9753f3

- mspm0l1306 examples: add I2C blocking & async examples - mspm0l1306 examples: add -O2 optimization due to Flash limitations - mspm0g3507 examples: add I2C blocking & async examples

mspm0-I2C: add type for I2C clock rates + fixed comments

45852b8

mspm0-I2C: automate source clock definition

917a509

- i2c-config: automatically defines clock source based on input I2C rate - i2c: proper config functions naming - i2c-examples: adapt to changed API - i2c: save initialization pf cctr register

mspm0-I2C: remove SYSOSC init, make ClockDiv::Into private

dc52ead

mspm0-I2C: mention blocking API's restrictions

d6a87b4

- blocking API for transfering max 8 bytes - async API has no such limitations

mspm0-I2C: update mspm0-metapac revision

934e780

mspm0-I2C: forward fifo size to i2c controller

8091155

mspm0-I2C: remove type time:Herz usage

3b7b343

bespsm force-pushed the mspm0-i2c branch from 9275b43 to 3b7b343 Compare August 4, 2025 08:19

mspm0-I2C: replace examples for mspm0l1306 & mspm0g3507 with AD5171

e78959e

bespsm changed the title ~~[WIP] MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs)~~ MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs) Aug 4, 2025

i509VCB reviewed Aug 7, 2025

View reviewed changes

i509VCB approved these changes Aug 8, 2025

View reviewed changes

i509VCB added this pull request to the merge queue Aug 8, 2025

i509VCB mentioned this pull request Aug 8, 2025

mspm0: work around I2C_ERR_08 #4519

Open

github-merge-queue bot removed this pull request from the merge queue due to failed status checks Aug 8, 2025

i509VCB added this pull request to the merge queue Aug 8, 2025

github-merge-queue bot removed this pull request from the merge queue due to failed status checks Aug 8, 2025

i509VCB added this pull request to the merge queue Aug 11, 2025

github-merge-queue bot removed this pull request from the merge queue due to failed status checks Aug 11, 2025

Iooon suggested changes Aug 13, 2025

View reviewed changes

bespsm added 2 commits August 14, 2025 10:30

mspm0-I2C: fix calculate_timer_period function & tests

b66fa9a

embassy-mspm0: add changes to the CHANGELOG.md

f84eb9a

bespsm force-pushed the mspm0-i2c branch from 50fb8e9 to f84eb9a Compare August 14, 2025 08:31

bespsm requested a review from Iooon August 14, 2025 08:40

i509VCB mentioned this pull request Aug 14, 2025

ci: 5340dk is dead #4551

Merged

i509VCB added this pull request to the merge queue Aug 14, 2025

Merged via the queue into embassy-rs:main with commit 32f142d Aug 14, 2025
8 checks passed

MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs) #4435

MSPM0: Add I2C Controller (blocking & async) + examples for mspm0l1306, mspm0g3507 (tested MCUs) #4435

Uh oh!

Conversation

bespsm commented Jul 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

i509VCB left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bespsm commented Jul 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Iooon commented Jul 25, 2025

Uh oh!

bespsm commented Jul 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

i509VCB commented Aug 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bespsm commented Aug 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

i509VCB commented Aug 6, 2025

Uh oh!

i509VCB Aug 7, 2025

Choose a reason for hiding this comment

Uh oh!

bespsm Aug 7, 2025

Choose a reason for hiding this comment

Uh oh!

i509VCB Aug 8, 2025 • 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!

i509VCB commented Aug 7, 2025

Uh oh!

Uh oh!

i509VCB commented Aug 8, 2025

Uh oh!

Uh oh!

i509VCB commented Aug 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bespsm commented Aug 11, 2025

Uh oh!

Uh oh!

bespsm commented Aug 13, 2025

Uh oh!

Iooon left a comment

Choose a reason for hiding this comment

Uh oh!

bespsm commented Aug 14, 2025

Uh oh!

bespsm commented Aug 14, 2025

Uh oh!

i509VCB commented Aug 14, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

bespsm commented Jul 21, 2025 •

edited

Loading

bespsm commented Jul 24, 2025 •

edited

Loading

bespsm commented Jul 28, 2025 •

edited

Loading

i509VCB commented Aug 2, 2025 •

edited

Loading

bespsm commented Aug 2, 2025 •

edited

Loading

i509VCB Aug 8, 2025 •

edited

Loading

i509VCB commented Aug 8, 2025 •

edited

Loading