此篇限定研究范圍為FLASH區域的訪問規則。

以下結論均為實測所得：

一、讀

1.單字節讀例程，單字節讀可以從任意地址開始讀

uint8_t STMFLASH_ReadByte(uint32_t faddr)
{
  return *(volatile uint8_t*)faddr; 
}

2.半字（2個字節）讀例程， 半字讀只能從偶數地址開始讀 ，從奇數地址讀的話單片機就死機了

uint16_t STMFLASH_ReadHalfWord(uint32_t faddr)
{
  return *(volatile uint16_t*)faddr; 
}

3.字（4個字節）讀例程， 按字讀只能從4的整數倍地址開始讀 ，否則單片機死機

uint32_t STMFLASH_ReadWord(uint32_t faddr)
{
  return *(volatile uint32_t*)faddr; 
}

二、寫

1.單次寫入字節數：

我看G0x1的技術參考手冊上說一次寫入是72bit（64bit加ecc），我實際測試是單次寫入只能是32bit，以實際測試為準

（1）單字節寫，寫不了，單片機死機

我在HAL庫里面把庫函數FLASH_Program_DoubleWord改成如下進行驗證的

``

static void FLASH_Program_DoubleWord(uint32_t Address, uint64_t Data)

{

/* Set PG bit */

SET_BIT(FLASH->CR, FLASH_CR_PG);

*(uint8_t *)Address = (uint8_t)(Data&0xff);__ISB();

*((uint8_t *)Address+1) = (uint8_t)((Data>>8U)&0xff);__ISB();

*((uint8_t *)Address+2) = (uint8_t)((Data>>16U)&0xff);__ISB();

*((uint8_t *)Address+3) = (uint8_t)((Data>>24U)&0xff);__ISB();

*((uint8_t *)Address+4) = (uint8_t)((Data>>32U)&0xff);__ISB();

*((uint8_t *)Address+5) = (uint8_t)((Data>>40U)&0xff);__ISB();

*((uint8_t *)Address+6) = (uint8_t)((Data>>48U)&0xff);__ISB();

*((uint8_t *)Address+7) = (uint8_t)((Data>>56U)&0xff);

}

``

（2）**半字寫，寫不了，單片機死機

我在HAL庫里面把庫函數FLASH_Program_DoubleWord改成如下進行驗證的

static void FLASH_Program_DoubleWord(uint32_t Address, uint64_t Data)
{
  /* Set PG bit */
  SET_BIT(FLASH->CR, FLASH_CR_PG);


  *(uint16_t *)Address = (uint16_t)(Data&0xffff);__ISB();
  *((uint16_t *)Address+2) = (uint16_t)((Data>>16U)&0xffff);__ISB();
  *((uint16_t *)Address+4) = (uint16_t)((Data>>32U)&0xffff);__ISB();
  *((uint16_t *)Address+6) = (uint16_t)((Data>>48U)&0xffff);

}

（3）按字寫(4個字節，32bit)，可以，HAL庫函數FLASH_Program_DoubleWord里面就是按字寫的

static void FLASH_Program_DoubleWord(uint32_t Address, uint64_t Data)
{
  /* Set PG bit */
  SET_BIT(FLASH->CR, FLASH_CR_PG);


  /* Program first word */
  *(uint32_t *)Address = (uint32_t)Data;


  /* Barrier to ensure programming is performed in 2 steps, in right order
    (independently of compiler optimization behavior) */
  __ISB();


  /* Program second word */
  *(uint32_t *)(Address + 4U) = (uint32_t)(Data >> 32U);

}

（4） 按64bit寫， 寫不了，單片機死機

我在HAL庫里面把庫函數FLASH_Program_DoubleWord改成如下進行驗證的

static void FLASH_Program_DoubleWord(uint32_t Address, uint64_t Data)
{
  /* Set PG bit */
  SET_BIT(FLASH->CR, FLASH_CR_PG);
  
  *(uint64_t *)Address = (uint64_t)Data;
  }

2.寫入的首地址規則

經過實際驗證得出：**寫入首地址必須為8的整數倍，不然就會發生錯誤：如果為4的整數倍但不是8的整數倍，單片機不會死機，但是寫入的值全部變成0x3F，其他既不是4的整數倍也不是8的整數倍的地址，單片機會死機。

按以下代碼改變寫入首地址即可進行驗證

STMFLASH_Write(FLASH_CALIBRATION_ADDR+8,String,sizeof(String));
HAL_Delay(10);
STMFLASH_Read(FLASH_CALIBRATION_ADDR,buffer,sizeof(String)/4+1);
HAL_Delay(10);
Uart_Send_Data(buffer, sizeof(String));