前言

有童鞋說在gcc下，vsnprintf來替代rt_vsnprintf來打印浮點會引發死機

經過實際驗證，沒有發現死機

但并不建議直接用vsnprintf來替代rt_vsnprintf，打印浮點多半用于調試，可以獨立成一個專門的打印函數

vsnprintf只是用于把浮點格式化到字符串，只要字符串長度不溢出，理論上不會造成死機

如果發現死機，需要確認打印的環境，如中斷中，尤其在串口輸出的情況下

測試例程

RT-Thread Studio arm-none-eabi-gcc 環境：未發現死機現象

Keil MDK5 環境：未發現死機現象

cygwin gcc 環境：未發現死機現象

可以正常的用于浮點數的調試打印

/*

Copyright (c) 2006-2018, RT-Thread Development Team

SPDX-License-Identifier: Apache-2.0

Change Logs:

Date Author Notes

2018-11-06 SummerGift change to new framework
/
#include
#include
#include
#include "drv_gpio.h"
/ defined the LED0 pin: PA5 /
#define LED0_PIN GET_PIN(A, 5)
#include
#define DBG_BUFF_MAX_LEN 256
/ debug print : support float double */
int dbg_printf(const char fmt, ...)
{
va_list args;
static char rt_log_buf[DBG_BUFF_MAX_LEN] = { 0 };
va_start(args, fmt);
int length = vsnprintf(rt_log_buf, sizeof(rt_log_buf) - 1, fmt, args);
rt_kputs(rt_log_buf);
return length;
}
int main(void)
{
double a = 334455.00123345;
double b = 667788.14234234;
dbg_printf("double a = %lfrn", a);
int count = 1;
/ set LED0 pin mode to output */
rt_pin_mode(LED0_PIN, PIN_MODE_OUTPUT);
while (count++)
{
rt_pin_write(LED0_PIN, PIN_HIGH);
dbg_printf("double a + b = %lfrn", a+b+count);
rt_thread_mdelay(1000);
rt_pin_write(LED0_PIN, PIN_LOW);
rt_thread_mdelay(2000);
}
return RT_EOK;
}
cygwin的測試

/ vsnprintf example /
include
include
define DBG_BUFF_MAX_LEN 256
/ debug print : support float double /
int dbg_printf(const char *fmt, …)
{
va_list args;
static char rt_log_buf[DBG_BUFF_MAX_LEN] = { 0 };

va_start(args, fmt);
int length = vsnprintf(rt_log_buf, sizeof(rt_log_buf) - 1, fmt, args);
printf("%srn", rt_log_buf);
return length;
}

int main ()
{
double a = 123453.0001444;
dbg_printf(“%lf”, a);
return 0;
}

## 輸出效果

```c
msh >
  | /
- RT -     Thread Operating System
 / |      4.1.0 build Feb  3 2022 21:11:28
 2006 - 2022 Copyright by RT-Thread team
double a = 334455.001233
double a + b = 1002245.143576
msh >double a + b = 1002246.143576
double a + b = 1002247.143576
double a + b = 1002248.143576
double a + b = 1002249.143576
double a + b = 1002250.143576
double a + b = 1002251.143576
double a + b = 1002252.143576
double a + b = 1002253.143576
double a + b = 1002254.143576
double a + b = 1002255.143576
double a + b = 1002256.143576
double a + b = 1002257.143576
double a + b = 1002258.143576
double a + b = 1002259.143576
double a + b = 1002260.143576
double a + b = 1002261.143576
cygwin的打印：
$ ./a.exe
123453.000144

小結

rt_kprintf，采用不依賴標準C庫的方式，主要目的是為了減少ROM（Flash）的占用

如果實際中需要浮點打印（如打印到串口、打印到文件），可以基于vsnprintf編寫一個簡單的全功能格式打印函數

RT-Thread ulog可以開啟浮點支持，開啟后使用：vsnprintf