Fault when operating on double/float on STM32F4 - c

Question

I am working with STM32F407 microcontroller.
void mya_usart_init(USART_Type* USARTx, uint32_t baudRate, uint8_t UsartClockSpeedMHz) {
double a = (double)72/7;
double b = (double)UsartClockSpeedMHz;
// usart_div = (double)UsartClockSpeedMHz; // * 1000000 / (8 * (2) * baudRate);
//
// USARTx->BRR |= ((uint32_t)usart_div << 4) | (uint32_t) round(((usart_div - (uint32_t)usart_div) * 16)) ; // seperate integer part and decimal part
//
// // USARTx->CR1 &= ~USART_CR1_M; // this is reset value already
//
// USARTx->CR1 |= USART_CR1_RE | USART_CR1_TE | USART_CR1_UE; // Rx, Tx, and USART enable
return;
}
I have been dealing with USART, then I figured out that I cannot make this conversion from uint8_t to double. As seen from below, variable a is successfully gets its value, however, the variable b cannot be written with the desired value. I have added the assembly code, everything goes well but at this stage;
08000ac4: bl 0x800046c <__floatunsidf>
The MCU enters a fault with the following error in IDE(Eclipse based STM32CubeIDE).
No source available for "__floatunsidf() at 0x800046c"
I also noticed that, this assembly line above does not appear for (double)72/7 but appears for (double)UsartClockSpeedMHz. Why is that so?
So, to me, while it can convert 72/7 to double, it cannot, for some reason, convert UsartClockSpeedMHz, which is 84, given as an argument to the function, to double. What is the reason for this?
Thanks.
6 void mya_usart_init(USART_Type* USARTx, uint32_t baudRate, uint8_t UsartClockSpeedMHz) {
mya_usart_init:
08000aa8: push {r4, r7, lr}
08000aaa: sub sp, #36 ; 0x24
08000aac: add r7, sp, #0
08000aae: str r0, [r7, #12]
08000ab0: str r1, [r7, #8]
08000ab2: mov r3, r2
08000ab4: strb r3, [r7, #7]
10 double a = (double)72/7;
08000ab6: add r4, pc, #32 ; (adr r4, 0x8000ad8 <mya_usart_init+48>)
08000ab8: ldrd r3, r4, [r4]
08000abc: strd r3, r4, [r7, #24]
11 double b = (double)UsartClockSpeedMHz;
08000ac0: ldrb r3, [r7, #7]
08000ac2: mov r0, r3
08000ac4: bl 0x800046c <__floatunsidf>
08000ac8: mov r3, r0
08000aca: mov r4, r1
08000acc: strd r3, r4, [r7, #16]
21 return;
08000ad0: nop
22 }
update:
The build command is:
18:14:21 **** Incremental Build of configuration Debug for project MYA-STM32F4-DRIVERS ****
make -j8 all
arm-none-eabi-gcc "../Src/mya_usart.c" -mcpu=cortex-m4 -std=gnu11 -g3 -DSTM32 -DSTM32F407G_DISC1 -DSTM32F4 -DSTM32F407VGTx -DDEBUG -c -I../Inc -O0 -ffunction-sections -fdata-sections -Wall -fstack-usage -MMD -MP -MF"Src/mya_usart.d" -MT"Src/mya_usart.o" --specs=nano.specs -mfpu=fpv4-sp-d16 -mfloat-abi=hard -mthumb -o "Src/mya_usart.o"
../Src/mya_usart.c: In function 'mya_usart_init':
../Src/mya_usart.c:11:12: warning: unused variable 'b' [-Wunused-variable]
double b = (float)UsartClockSpeedMHz;
^
../Src/mya_usart.c:10:12: warning: unused variable 'a' [-Wunused-variable]
double a = (float)72/7;
^
arm-none-eabi-gcc -o "MYA-STM32F4-DRIVERS.elf" #"objects.list" -mcpu=cortex-m4 -T"/home/muyustan/Documents/STM32CubeIDE/workspace_1.3.0/MYA-STM32F4-DRIVERS/STM32F407VGTX_FLASH.ld" --specs=nosys.specs -Wl,-Map="MYA-STM32F4-DRIVERS.map" -Wl,--gc-sections -static --specs=nano.specs -mfpu=fpv4-sp-d16 -mfloat-abi=hard -mthumb -Wl,--start-group -lc -lm -Wl,--end-group
Finished building target: MYA-STM32F4-DRIVERS.elf
arm-none-eabi-objdump -h -S MYA-STM32F4-DRIVERS.elf > "MYA-STM32F4-DRIVERS.list"
arm-none-eabi-objcopy -O binary MYA-STM32F4-DRIVERS.elf "MYA-STM32F4-DRIVERS.bin"
arm-none-eabi-size MYA-STM32F4-DRIVERS.elf
text data bss dec hex filename
2804 8 1568 4380 111c MYA-STM32F4-DRIVERS.elf
Finished building: default.size.stdout
Finished building: MYA-STM32F4-DRIVERS.bin
Finished building: MYA-STM32F4-DRIVERS.list
I cannot even perform this in main(),
for (i=0 ; i<15; i++){
f*=10.01;
}
I get No source available for "__extendsfdf2() at 0x80004b0" for the above one.
If I change it to, (notice the 10.01f)
for (i=0 ; i<15; i++){
f*=10.01f;
}
This time, assembly code changes:
83 f*=10.01f;
080002fc: vldr s15, [r7]
08000300: vldr s14, [pc, #44] ; 0x8000330 <main+288>
08000304: vmul.f32 s15, s15, s14
08000308: vstr s15, [r7]
0800030c: ldr r3, [r7, #4]
0800030e: adds r3, #1
08000310: str r3, [r7, #4]
08000312: ldr r3, [r7, #4]
08000314: cmp r3, #14
08000316: ble.n 0x80002fc <main+236>
Executing the very first step, vldr s15, [r7] routes me to the following screen: