Variadic functions (like prinf)

[diego-z.esse-ti]

Hi community,

I wrote the following function

UINT PrintOut(const char* format, ...)
{
	va_list args;
	va_start(args, format);

	char buffer[1024] = { 0 };
	vsprintf(buffer, format, args);

	va_end(args);

	return PrintString(buffer);
}

No problems at compiling time, but at runtime the execution falls in HardFault.

Any suggestion?

[Pavel A.]

The code looks OK. Debug as usual. HardFault where? The debugger of CubeIDE has convenient tool to analyze HardFaults.

[Andrew Neil]

So where, exactly, does the Hard Fault occur?

And what Fault is reported?

Some links here on debugging Cortex-M Hard Faults:

https://community.arm.com/support-forums/f/embedded-forum/3257/debugging-a-cortex-m0-hard-fault

See also CubeIDE Fault Analyser:

 

https://www.st.com/resource/en/user_manual/um2609-stm32cubeide-user-guide-stmicroelectronics.pdf 

[alister]

Line 6 is a waste of code. Isn't line 7 copying data over buffer? In embedded development especially, all code should be for a reason and no code should be without a reason.

Line 7 should use vsnprintf. Else what would happen if the format or its expansions overrun buffer?

Ways this code could fail are to numerous to list. Let it hard-fault in a debugger and then debug it. Configure your IDE or makefile to generate a link map file and check that for the addresses in the hard-fault information. Check ST's applicable programmer manual for your variety of STM32. Search community for other posts on hard faults.

[Andrew Neil]

---

@diego-z.esse-ti wrote:

I wrote the following function

---

Why?

Why not just use printf() - you just have to add a simple write() function to support output to whatever device you require; eg, UART.

The bulk of the code in printf() is  accounted for by vsprintf() anyhow - so you're not saving much (if anything).

https://shawnhymel.com/1873/how-to-use-printf-on-stm32/

 

[Tesla DeLorean]

Probably related to 1KB stack usage.

See if making buffer static changes failure or results in Linker flagging memory size issue.

[diego-z.esse-ti]

Thank you @Tesla DeLorean,

   you nailed the problem!

[Pavel A.]

Then you have a worse problem. No detection of stack growing out of limits.