Allow provision for user defined TIMx_IRQ_Handlers

[geosmall]

Is your feature request/improvement related to a problem? Please describe.
Attempts to make full use of timer features in applications (i.e brushless motor control) require direct access to Timer IRQ Handlers. Current design of HardwareTimer.cpp takes full control of Timer IRQ Handlers. No user level compile switch is available to turn off these features. Any attempts to define sketch level Timer IRQ Handlers generates multiple definition of `TIMx_IRQHandler' errors.

Describe the solution you'd like to see
Provide user level build_opt.h switches to turn off IRQ based HardwareTimer (and associated pin PWM) functionality for advanced user applications.

I've hacked a quick and dirty solution with a user flag in build_opt.h with #ifndef EXCLUDE_HARDWARE_TIMER_MODULE in HardwareTimer.cpp, analog.cpp and wiring_digital.c files to get rid of the multiple definition errors, but have not thoroughly tested yet. I basically replaced #ifdef HAL_TIM_MODULE_ENABLED with my #ifndef so that feature can be turned off while still leaving the HAL_TIM_MODULE enabled. From there I can implement TIMx_IRQ_Handler at the sketch level.

There may be undesirable side effects I am unaware of. I'm guessing there may be a better way.

STM32 Core Design Philosophy Proposal
STM32 peripherals have many powerful features. A design philosophy to be considered for this core would be that features that take over IRQ Handlers (TIM, I2C, etc.) to provide base functionality are on by default but can optionally be switched off by users looking to take full control of the peripheral at the sketch level to gain more advanced functionality. This should also include HAL_XXX_MspInit() functions used to initialize HAL low level hardware initialization (i.e. if MspInit functions are implemented in core library functions then they should also be switchable off by user defines).

[fpistm]

Hi @geosmall
Right, I've already think about that and maybe I have a solution, anyway this will need some extra study.
Any proposal are welcome.

[ABOSTM]

Hi @geosmall,
There is an alternative:
You can directly replace NVIC vector with an IRQ handler of your choice:
NVIC_SetVector(irq_event_n, handler);
Just be warned that it will break any usage of concerned NVIC vector.
This NVIC API is directly usable in arduino sketch.

[geosmall]

There is an alternative:
You can directly replace NVIC vector with an IRQ handler of your choice:
NVIC_SetVector(irq_event_n, handler);

Hi @ABOSTM,
This approach sounds promising. I'm having trouble finding documentation on NVIC_SetVector(). Is there a good example out there? I'm presently working STM32F0 which I understand is differs in not having VTOR register. Does interrupt vector table first need to be relocated to SRAM in order to use NVIC_SetVector()?

[ABOSTM]

Hi @geosmall,
NVIC_SetVector() is defined in CMSIS so it is not part of this core,
but it comes with CMSIS which is automatically installed with this core.
For information about this function: official ARM documentation:
https://arm-software.github.io/CMSIS_5/Core/html/group__NVIC__gr.html#gab43c1c59d5c081f1bc725237f4b1f916

You are right STM32F0 doesn't support it as it is based on Cortex-M0.
And you are also right saying you need first to relocated to RAM in order to use NVIC_SetVector().

[geosmall]

Apparently doable even on Cortex-M0 using relocation to RAM after carving out enough space for the vectors with modified linker script. It is all a good bit of work and feels a bit work-around.

M0 approach here:
https://community.st.com/s/question/0D50X00009XkfRgSAJ/stm32f0-no-interrupt-after-vectortable-relocation

I thinking user #defines would be a better more versatile approach long term and work across variants. Having to modify linker scripts feels like a non-starter.

[geosmall]

I was able to get NVIC_SetVector() working on STM32F051K6 with the following sketch and linker script. Seems to be working when I insert a breakpoint within the sketch level interrupt routine My_SysTick_IRQHandler().

So this approach could work on F0 targets but would require linker script changes and the use of 256 bytes of RAM for the vector table.

Test sketch:

/*
  DEMO to:
  Remap Wraith32 STM32F0051K6 interrupt table to RAM,
  then use NVIC_SetVector() to insert our own sketch based
  SysTick interrupt handler.  
*/

#define APPLICATION_ADDRESS     (uint32_t)0x08000000

__IO uint32_t VectorTable[48] __attribute__((section(".RAMVectorTable")));


extern "C" void My_SysTick_IRQHandler(void) {

  HAL_IncTick();

}


// the setup function runs once when you press reset or power the board
void setup() {

  // Relocate by software the vector table to the internal SRAM at 0x20000000
  // Copy the vector table from the Flash (mapped at the base of the application
  // load address 0x08000000) to the base address of the SRAM at 0x20000000
  for(uint32_t i = 0; i < 48; i++) {
    VectorTable[i] = *(__IO uint32_t*)(APPLICATION_ADDRESS + (i<<2));
  }

  /* Enable the SYSCFG peripheral clock*/
  __HAL_RCC_SYSCFG_CLK_ENABLE();

  // Remap SRAM at 0x00000000
  __disable_irq();
  __HAL_SYSCFG_REMAPMEMORY_SRAM();
  __enable_irq();

  NVIC_SetVector(SysTick_IRQn, (uint32_t)My_SysTick_IRQHandler);

  // initialize digital pin LED_BUILTIN as an output.
  pinMode(LED_BUILTIN, OUTPUT);

}


// the loop function runs over and over again forever
void loop() {

  digitalWrite(LED_BUILTIN, HIGH);   // turn the LED on (HIGH is the voltage level)
  delay(1000);                       // wait for a second
  digitalWrite(LED_BUILTIN, LOW);    // turn the LED off by making the voltage LOW
  delay(1000);                       // wait for a second

}

Linker script fro my custom variant:

/*
******************************************************************************
**
**  File        : LinkerScript.ld
**
**  Author      : Auto-generated by STM32CubeMx, modified by G. Small
**
**  Abstract    : Linker script for STM32F051K6Tx series
**                32Kbytes FLASH and 256 bytes SRAM, 7936 bytes RAM
**
**                Set heap size, stack size and stack location according
**                to application requirements.
**
**                Set memory bank area and size if external memory is used.
**
**  Target      : STMicroelectronics STM32
**
**  Distribution: The file is distributed “as is,” without any warranty
**                of any kind.
**
*****************************************************************************
*/

/* Entry Point */
ENTRY(Reset_Handler)

/* Highest address of the user mode stack */
_estack = 0x20002000;    /* end of RAM */
/* Generate a link error if heap and stack don't fit into RAM */
_Min_Heap_Size = 0x200;      /* required amount of heap  */
_Min_Stack_Size = 0x400; /* required amount of stack */

/* Specify the memory areas */
MEMORY
{
  SRAM  (xrw) : ORIGIN = 0x20000000,  LENGTH = 256
  RAM (xrw)   : ORIGIN = 0x20000100,  LENGTH = 7936
  FLASH (rx)  : ORIGIN = 0x8000000,   LENGTH = 32K
}

/* Define output sections */
SECTIONS
{
  /* The startup code goes first into FLASH */
  .isr_vector :
  {
    . = ALIGN(4);
    KEEP(*(.isr_vector)) /* Startup code */
    . = ALIGN(4);
  } >FLASH

  /* The program code and other data goes into FLASH */
  .text :
  {
    . = ALIGN(4);
    *(.text)           /* .text sections (code) */
    *(.text*)          /* .text* sections (code) */
    *(.glue_7)         /* glue arm to thumb code */
    *(.glue_7t)        /* glue thumb to arm code */
    *(.eh_frame)

    KEEP (*(.init))
    KEEP (*(.fini))

    . = ALIGN(4);
    _etext = .;        /* define a global symbols at end of code */
  } >FLASH

  /* Constant data goes into FLASH */
  .rodata :
  {
    . = ALIGN(4);
    *(.rodata)         /* .rodata sections (constants, strings, etc.) */
    *(.rodata*)        /* .rodata* sections (constants, strings, etc.) */
    . = ALIGN(4);
  } >FLASH

  .ARM.extab   : { *(.ARM.extab* .gnu.linkonce.armextab.*) } >FLASH
  .ARM : {
    __exidx_start = .;
    *(.ARM.exidx*)
    __exidx_end = .;
  } >FLASH

  .preinit_array     :
  {
    PROVIDE_HIDDEN (__preinit_array_start = .);
    KEEP (*(.preinit_array*))
    PROVIDE_HIDDEN (__preinit_array_end = .);
  } >FLASH
  .init_array :
  {
    PROVIDE_HIDDEN (__init_array_start = .);
    KEEP (*(SORT(.init_array.*)))
    KEEP (*(.init_array*))
    PROVIDE_HIDDEN (__init_array_end = .);
  } >FLASH
  .fini_array :
  {
    PROVIDE_HIDDEN (__fini_array_start = .);
    KEEP (*(SORT(.fini_array.*)))
    KEEP (*(.fini_array*))
    PROVIDE_HIDDEN (__fini_array_end = .);
  } >FLASH

  /* used by the startup to initialize data */
  _sidata = LOADADDR(.data);

  /* Initialized data sections goes into RAM, load LMA copy after code */
  .data : 
  {
    . = ALIGN(4);
    _sdata = .;        /* create a global symbol at data start */
    *(.data)           /* .data sections */
    *(.data*)          /* .data* sections */

    . = ALIGN(4);
    _edata = .;        /* define a global symbol at data end */
  } >RAM AT> FLASH

  
  /* Uninitialized data section */
  . = ALIGN(4);
  .bss :
  {
    /* This is used by the startup in order to initialize the .bss secion */
    _sbss = .;         /* define a global symbol at bss start */
    __bss_start__ = _sbss;
    *(.bss)
    *(.bss*)
    *(COMMON)

    . = ALIGN(4);
    _ebss = .;         /* define a global symbol at bss end */
    __bss_end__ = _ebss;
  } >RAM

  /* User_heap_stack section, used to check that there is enough RAM left */
  ._user_heap_stack :
  {
    . = ALIGN(8);
    PROVIDE ( end = . );
    PROVIDE ( _end = . );
    . = . + _Min_Heap_Size;
    . = . + _Min_Stack_Size;
    . = ALIGN(8);
  } >RAM

  

  /* Remove information from the standard libraries */
  /DISCARD/ :
  {
    libc.a ( * )
    libm.a ( * )
    libgcc.a ( * )
  }

  .ARM.attributes 0 : { *(.ARM.attributes) }
  .RAMVectorTable : { *(.RAMVectorTable) } >SRAM AT> FLASH

}

[geosmall]

Assuming use of NVIC_SetVector() per above to manage IRQ_Handlers at the sketch level, I run into the next issue. The application I am trying to port originated as an STM32CubeMx project. As such it makes use of HAL_XXX_MspInit() functions. As these functions are claimed within STM32 core I get a series of multiple definition linker errors.

The above will make porting of STM32CubeMx generated code or examples to sketches an onerous task. It is possible to work around this by writing local replacement HAL_XXX_Init() functions that then allow one to call their own local HAL_XXX_MspInit() functions, but this is all feels like unnecessary work.

I believe that the correct design philosophy is to provide user level build_opt.h switches to turn off TIM / ADC / COMP / Etc... functionality for advanced user applications.

Link errors generated:

Linking everything together...
"C:\Users\geoma\AppData\Local\Arduino15\packages\STM32\tools\arm-none-eabi-gcc\8.2.1-1.7/bin/arm-none-eabi-gcc" -mcpu=cortex-m0  -mthumb -g -Og --specs=nano.specs -Wl,--defsym=LD_FLASH_OFFSET=0 -Wl,--defsym=LD_MAX_SIZE=32768 -Wl,--defsym=LD_MAX_DATA_SIZE=8192 -Wl,--cref -Wl,--check-sections -Wl,--gc-sections -Wl,--entry=Reset_Handler -Wl,--unresolved-symbols=report-all -Wl,--warn-common "-TC:\Users\geoma\OneDrive\Arduino\hardware\STM32\stm32\variants\WRAITH32_F051K6/ldscript.ld" "-Wl,-Map,C:\Users\geoma\AppData\Local\Temp\arduino_build_500636/f051bldc.ino.map"   "-LC:\Users\geoma\AppData\Local\Arduino15\packages\STM32\tools\CMSIS\5.5.1/CMSIS/DSP/Lib/GCC/" -larm_cortexM0l_math -o "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636/f051bldc.ino.elf" "-LC:\Users\geoma\AppData\Local\Temp\arduino_build_500636" -Wl,--start-group "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\eeprom.c.o" "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\main.c.o" "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\stm32f0xx_hal_msp.c.o" "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\f051bldc.ino.cpp.o" "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\PeripheralPins.c.o" "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\variant.cpp.o" -Wl,--whole-archive "C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\core.a" -Wl,--no-whole-archive -lc -Wl,--end-group -lm -lgcc -lstdc++
c:/users/geoma/appdata/local/arduino15/packages/stm32/tools/arm-none-eabi-gcc/8.2.1-1.7/bin/../lib/gcc/arm-none-eabi/8.2.1/../../../../arm-none-eabi/bin/ld.exe: C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\core.a(timer.c.o): in function `HAL_TIM_Base_MspInit':

C:\Users\geoma\OneDrive\Arduino\hardware\STM32\stm32\cores\arduino\stm32/timer.c:30: multiple definition of `HAL_TIM_Base_MspInit'; C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\stm32f0xx_hal_msp.c.o:C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch/stm32f0xx_hal_msp.c:219: first defined here

c:/users/geoma/appdata/local/arduino15/packages/stm32/tools/arm-none-eabi-gcc/8.2.1-1.7/bin/../lib/gcc/arm-none-eabi/8.2.1/../../../../arm-none-eabi/bin/ld.exe: C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\core.a(timer.c.o): in function `HAL_TIM_Base_MspDeInit':

C:\Users\geoma\OneDrive\Arduino\hardware\STM32\stm32\cores\arduino\stm32/timer.c:50: multiple definition of `HAL_TIM_Base_MspDeInit'; C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\stm32f0xx_hal_msp.c.o:C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch/stm32f0xx_hal_msp.c:344: first defined here

c:/users/geoma/appdata/local/arduino15/packages/stm32/tools/arm-none-eabi-gcc/8.2.1-1.7/bin/../lib/gcc/arm-none-eabi/8.2.1/../../../../arm-none-eabi/bin/ld.exe: C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\core.a(analog.cpp.o): in function `HAL_ADC_MspInit':

C:\Users\geoma\OneDrive\Arduino\hardware\STM32\stm32\cores\arduino\stm32/analog.cpp:572: multiple definition of `HAL_ADC_MspInit'; C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\stm32f0xx_hal_msp.c.o:C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch/stm32f0xx_hal_msp.c:75: first defined here

c:/users/geoma/appdata/local/arduino15/packages/stm32/tools/arm-none-eabi-gcc/8.2.1-1.7/bin/../lib/gcc/arm-none-eabi/8.2.1/../../../../arm-none-eabi/bin/ld.exe: C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\core\core.a(analog.cpp.o): in function `HAL_ADC_MspDeInit':

C:\Users\geoma\OneDrive\Arduino\hardware\STM32\stm32\cores\arduino\stm32/analog.cpp:653: multiple definition of `HAL_ADC_MspDeInit'; C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch\stm32f0xx_hal_msp.c.o:C:\Users\geoma\AppData\Local\Temp\arduino_build_500636\sketch/stm32f0xx_hal_msp.c:123: first defined here

collect2.exe: error: ld returned 1 exit status
exit status 1
Error compiling for board Wraith32 speed controller.

[fpistm]

@geosmall

I believe that the correct design philosophy is to provide user level build_opt.h switches to turn off TIM / ADC / COMP / Etc... functionality for advanced user applications.

I guess this is the simplest way.

Anyway why use Arduino if you generate you code thanks CubeMX why not use directly CubeFw with CubeIDE or other?