將float型別資料寫入暫存器中
在FreeBSD中針對暫存器在特殊情況下可能需要寫入float型別的值，具體實現步驟如下：
//變數
typedef struct{
float under_limit;
float over_limit;
}VFEPIO_SetRawLimit_t;
typedef struct{
VFEPIO_ReqHeader_t header;
union{
VFEPIO_SetRawLimit_t raw_limit;
}param;
}VFEPIO_Request_t;
VFEPIO_Request_t VFEPIO_Request;
VFEPIO_SetRawLimit_t raw_limit;
float under_limit;
float over_limit;
uint32_t *piTemp1;
uint32_t *piTemp2;
//程式碼的實現
raw_limit = VFEPIO_Request->param.raw_limit;
under_limit = raw_limit.under_limit ;
over_limit = raw_limit.over_limit;
piTemp1 = (uint32_t *)(&under_limit);
piTemp2 = (uint32_t *)(&over_limit);
pci_write_bar1(sc, UNDERFLOW_LIMIT_ADDR, *piTemp1);
pci_write_bar1(sc, OVERFLOW_LIMIT_ADDR, *piTemp2);
◆在C語言中也是可以證明這點的，具體實現如下
#include <stdio.h>
#include <stdlib.h>
int main()
{
float a = 1.23;
float b = 0.00;
float *d = 0;
int c = 0x00;
int *ptr ;
ptr = (int *)&a;
c = *ptr;
// printf("c = %x!\r\n", c);
d = (float *)&c;
b = *d;
printf("b = %0.2f!\r\n", b);
return 0;
}
結果如下：
-
在C#中可以編輯為如下
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace Float2Byte
{
class Program
{
static void Main(string[] args)
{
float f = -0.6f;
byte[] b = BitConverter.GetBytes(f);
Console.WriteLine(BitConverter.ToString(b));
float f3 = (float)(0.6);
byte[] b4 = BitConverter.GetBytes(f3);
Console.WriteLine(BitConverter.ToString(b4));
float f2 = 120.5f;
byte[] b2 = BitConverter.GetBytes(f2);
Console.WriteLine(BitConverter.ToString(b2));
float ff2 = 10.0f;
byte[] b3 = BitConverter.GetBytes(ff2);
Console.WriteLine(BitConverter.ToString(b3));
byte[] bb = { 0x00, 0x00, 0x20, 0x41 };
Console.WriteLine("\n" + BitConverter.ToSingle(bb, 0));
}
}
}
-
如果上下層需要通過byte陣列進行float資料的互動
（1）如下是底層傳到上層的處理，在底層的程式碼如下：
float highlimitvalue;
int32_t highlimitvalueTemp;
if (exmodthresholdlimitack.highlimit[1] == 0x80)
{
highlimitvalue = 0.0 - (exmodthresholdlimitack.highlimit[0] / 100.0);
}
else
{
highlimitvalue = (exmodthresholdlimitack.highlimit[0] / 100.0);
}
IIC_Status_Ret = 0x00;
IIC_Status_Ret = 0x00 | DVMOD_Status | exmodthresholdlimitack.err1; //DvmodExmodErr | ExmodScontErr
highlimitvalueTemp = *((int32_t *)(&highlimitvalue));
output->Data[0] = (uint8_t)( ( highlimitvalueTemp >> 0) & 0xff );
output->Data[1] = (uint8_t)( ( highlimitvalueTemp >> 8) & 0xff );
output->Data[2] = (uint8_t)( ( highlimitvalueTemp >> 16) & 0xff );
output->Data[3] = (uint8_t)( ( highlimitvalueTemp >> 24) & 0xff );
在上層的程式碼如下：
highlimit = BitConverter.ToSingle(getthresholdlimitack.highthreshold, 0);
highlimit = (short)(highlimit * 100) / 100.0;
（2）如下是上層傳到底層的處理，在上層的程式碼如下：
byte[] ByteThresholdlow = new byte[4];
byte[] ByteThresholdup = new byte[4];
ByteThresholdlow = BitConverter.GetBytes((float)Thresholdlow[exmodid, scontid]);
ByteThresholdup = BitConverter.GetBytes((float)Thresholdup[exmodid, scontid]);
在底層接收的操作處理程式碼如下：
uint8_t HighThresholdLimit[4];
uint8_t *pucTemp1;
float HighThresholdLimitValue;
HighThresholdLimit[0] = input->RcvData[7];
HighThresholdLimit[1] = input->RcvData[8];
HighThresholdLimit[2] = input->RcvData[9];
HighThresholdLimit[3] = input->RcvData[10];
pucTemp1 = (uint8_t*)&HighThresholdLimitValue;
pucTemp1[0] = HighThresholdLimit[0];
pucTemp1[1] = HighThresholdLimit[1];
pucTemp1[2] = HighThresholdLimit[2];
pucTemp1[3] = HighThresholdLimit[3];
if (HighThresholdLimitValue < 0)
{
exmodsetthresholdcmd.highlimit[1] = 0x80;
}
else
{
exmodsetthresholdcmd.highlimit[1] = 0x00;
}
exmodsetthresholdcmd.highlimit[0] = (uint8_t)(fabs(HighThresholdLimitValue) * 100);