ChibiOS Free Embedded RTOS

I experimented with ARMv8-M-ML-ALT and it would not bootup on a STM32U385.

My options to reproduce is: PORT_USE_SYSCALL=FALSE, CORTEX_USE_FPU=TRUE, and CH_DBG_ENABLE_STACK_CHECK=TRUE

Here is a patch to fix the problem:

diff --git a/os/common/ports/ARMv8-M-ML-ALT/compilers/GCC/chcoreasm.S b ...

Here is a quick patch to enable FDCAN support in STM32U3.


diff --git a/os/hal/ports/STM32/STM32U3xx/platform.mk b/os/hal/ports/STM32/STM32U3xx/platform.mk
index 249aa94ed..71b596f10 100644
--- a/os/hal/ports/STM32/STM32U3xx/platform.mk
+++ b/os/hal/ports/STM32/STM32U3xx ...

A few minor tweaks for STM32U385 port compilation using trunk baseline.

First:

diff --git a/os/hal/ports/STM32/LLD/EXTIv1/stm32_exti.h b/os/hal/ports/STM32/LLD/EXTIv1/stm32_exti.h
index 4b9dcd93a..5d1d06526 100644
--- a/os/hal/ports/STM32/LLD/EXTIv1/stm32_exti.h
+++ b/os/hal ...

I tried the fix with a STM32F417 (silicon "Z") and the calendar freezes. The sub-seconds keep on incrementing normally.

I think the errata is misleading. I found this statement in the reference manual to be important:

To ensure consistency between the 3 values, reading either RTC_SSR or RTC_TR ...

I recently upgraded a working project from 20.3.x to 21.6.x. I have custom code that uses chRegFirstThread() followed by chRegNextThread() to walk all threads. I noticed that chRegNextThread() always returns NULL for my application. This results in only seeing the first thread.

The following patch ...

I did not see the note. I think it is important to maintain consistency across threads with respect to the CH_DBG_FILL_THREADS option. The idle thread, from a thread perpective, is not special or different.

Can we instead fix is with the patch below?

Thanks!


diff --git a/os/rt/src/chinstances ...

I recently upgraded a project from 20.3.x to 21.6.x. The semaphore enqueuing test (7.2) consistently crashed when run with other code in my project. When I turned off CH_CFG_USE_SEMAPHORES_PRIORITY the problem went away. Looking at the code, I suspect this patch is the problem. With this patch, the ...

I recently upgraded a project from 20.3.x to 21.6.x. I have custom software that monitors the stack usage. I noticed in 21.6.x that the idle stack was not being filled with known values like 20.3.x. All other stacks worked normally.

Here is a patch that restores the functionality:

diff --git a/os ...

It worked! Thank you!

I am hoping that you can accept this patch for the STM32F4 platform that adds TIM9 interrupt support back. I have tested it and it works.

Here are the three files that need to change:


diff --git a/os/hal/ports/STM32/LLD/TIMv1/stm32_tim1_9_10_11.inc b/os/hal/ports/STM32 ...

Thanks. I think I got confused on the second item.