Verify exti test runs for stm32c092xx/stm32c091xx

[KozhinovAlexander]

Is your feature request related to a problem? Please describe.

Verify exti test runs for stm32c092xx/stm32c091xx

Describe the solution you'd like

EXTI test case shall run for mentioned SoCs

Describe alternatives you've considered

No response

Reason fr this item

Without the changes from the PR #93731 the c092 sample does not run.

[KozhinovAlexander]

Q1: Still open question: Why latest merged branch didn't catched mentioned twister fail?

1. Stm32 exti rework PR STM32 EXTI Rework #85508 was merged at about 10 am on 25.7.25. Last twister run before merge on 24.7.25.
2. Stm32c092 series support PR STM32: Add STM32C092 SoC support and NUCLEO-C092RC #93263 had last twister run before merge on 23.7.25 ref

2 was running in CI successfully before merge 2 days before 1 was merged. But 2 was merged after 1. Thus CI missed to catch the bug at test-runtime.

Q2: Does exti test builds in CI?

I. Running west twister -T tests/drivers/interrupt_controller/intc_exti_stm32 --vendor st on developer machine (Ubuntu 24.04 LTS OS) gives proper results:

INFO    - Total complete:  136/ 136  100%  built (not run):  133, filtered:    7, failed:    0, error:    0
INFO    - 1 test scenarios (140 configurations) selected, 7 configurations filtered (4 by static filter, 3 at runtime).
INFO    - 0 of 133 executed test configurations passed (0.00%), 133 built (not run), 0 failed, 0 errored, with no warnings in 445.62 seconds.
INFO    - 0 of 0 executed test cases passed (0.00%) on 0 out of total 1124 platforms (0.00%).
INFO    - 399 selected test cases not executed: 399 not run (built only).
INFO    - 0 test configurations executed on platforms, 133 test configurations were only built.
INFO    - Saving reports...

Result:

• This item was opened to answer Q1 - it was answered.
• Q2 ist still open and shall be referred to CI team.

[JarmouniA]

drivers.interrupt_controller.intc_exti_stm32 seems not to run in CI

I guess not everything runs in Zephyr Project CI or even builds, but it will certainly run in ST's Zephyr CI, which is enough I think (they actually have the hardware).

[KozhinovAlexander]

drivers.interrupt_controller.intc_exti_stm32 seems not to run in CI

I guess not everything runs in Zephyr Project CI or even builds, but it will certainly run in ST's Zephyr CI, which is enough I think (they actually have the hardware).

You're way to fast 😉 I am still investigating all the logs 😄

[KozhinovAlexander]

drivers.interrupt_controller.intc_exti_stm32 seems not to run in CI

I guess not everything runs in Zephyr Project CI or even builds, but it will certainly run in ST's Zephyr CI, which is enough I think (they actually have the hardware).

It looks like you're right. I can't find any nucleo_ pattern in twister build logs for example. But there are definitely some other boards based on stm32 mentioned

[KozhinovAlexander]

@erwango I think this is the reason, why CI didn't catched stm32c092xx exti bug.
Since bug itself was fixed - I can close this item.

Future WARNING: I do think, from time to time some stm32 will fail in CI in stm32 exti test. The first obvious fix of the problem shall be an update to SoC .dtsi file with number of exti lines appropriate for particular SoC according to RM.

[JarmouniA]

I looked a little bit deeper, and my guess is, this is what prevented the testcase from being selected: the dt_compat_enabled("test-st-stm32-exti") in

zephyr/tests/drivers/interrupt_controller/intc_exti_stm32/testcase.yaml

Line 18 in 2d4a119

filter: dt_compat_enabled("st,stm32-exti") and dt_compat_enabled("test-st-stm32-exti")

The EXTI test compat binding

zephyr/tests/drivers/interrupt_controller/intc_exti_stm32/dts/bindings/test-st-stm32-exti.yaml

Line 11 in 2d4a119

compatible: "test-st-stm32-exti"

does not include the base binding, hence the status property is not defined,

zephyr/dts/bindings/base/base.yaml

Line 6 in 5c7c398

status:

and dt_compat_enabled("test-st-stm32-exti") always gives false.

zephyr/scripts/kconfig/kconfigfunctions.py

Line 765 in 5c7c398

This function takes a 'compat' and returns "y" if we find a status "okay"

But this theory might be completely wrong, haven't actually tested it.

You could try opening a draft PR with dt_has_compat("test-st-stm32-exti") instead, and see what happens in CI.

[KozhinovAlexander]

@JarmouniA Seems to be low hanging fruits. Thank you. I will give it a try.