小猫爪：S32K3学习笔记19-S32K3之如何集成SPD进自己的工程

1 前言

  前面已经将S32K3的与功能安全相关的硬件外设大致都梳理了一遍，前面在Safety简介中提到了NXP针对S32K3不同的安全机制共发布了免费的RTD和SPD，以及收费的SAF和SCST，如果恰逢你手头拮据，只能用免费的软件，而RTD和SPD又是两个独立的软件包，那么怎样将二者集成一下呢？

  首先下载SPD包并安装，有如下内容：

  我们需要用到的是SafetyBase， eMcem， Bist这三个模块。其中SafetyBase没啥东西，就是一些头文件，Bist模块实现了STCU2 BIST的功能，eMcem模块实现了FCCU, ERM, EIM, XBIC, DCM的功能。

2 EB的配置

  首先将SafetyBase， eMcem， Bist这三个模块对应的文件夹复制到EB的安装目录plugins下，让EB可以找到相关的模块。然后就是在自己的EB工程中添加这三个模块，如下：

  这样EB这边的配置就完成了。

3 S32DS的配置

  对于S32DS这边的配置，就更简单了，把EB生成的代码以及RTD,SPD的源代码加入到自己的工程里面去，然后还需要修改一下链接文件，具体修改的地方如下：

4 代码的集成

  接下来就是怎样在代码中集成SPD了，如果需要用SPD来实现一个基本的程序设计，那么在理想情况下，软件设计的基本流程如下：

  下面贴出SPD初始化的代码：

void Spd_Init(void)
{
	Mcu_ResetType  ResetReason = MCU_RESET_UNDEFINED;
	ResetReason = Mcu_GetResetReason();
	switch(ResetReason)
	{
		case MCU_POWER_ON_RESET:
			break;
	    //.....
		default:
			break;
	}
	/****************************************************************************
	 * if BIST(STCU) is not run yet, run LBIST and MBIST,
	 * this will trigger a ST_DONE functional reset.
	 * BIST must run with PLL enabled and CPU run in full-performance mode.
	 * BIST must run with FCCU faults detection disabled.
	 ***************************************************************************/
//#define DEMO_ENABLE_BIST
#ifdef DEMO_ENABLE_BIST
	Bist_StatusType bistStatus;
	uint32 stcuStatus;
	Std_ReturnType retStatus;
	Bist_LBistRDListType Bist_LBistRDList = 0U;
	Bist_MBistRDListType Bist_MBistRDList = 0U;

	bistStatus = Bist_GetExecStatus(BIST_SAFETYBOOT_CFG);

	if( bistStatus != BIST_OK )
	{
		if(bistStatus == BIST_ERROR)
		{
			/* Reads STCU ERR_STAT register to identify what HW error occured */
			stcuStatus = Bist_GetRawErrorStatus();
			(void) stcuStatus;
		}
		else if(bistStatus == BIST_FAILED)
		{
			/* Analyze which reset domain is failing */
			retStatus = Bist_GetFailRDs ( &Bist_LBistRDList, &Bist_MBistRDList );
			(void) retStatus;
		}
		else if(bistStatus == BIST_BUSY)
		{
			while( 1U );   /* Handle BIST HW busy state */
		}
		else if(bistStatus == BIST_NORUN)
		{
			/* Bist execution was not successfully performed due to HW hazard state. */
			Bist_Run(BIST_SAFETYBOOT_CFG);   /* Re-start execution */
		}
		else if(bistStatus == BIST_INTEGRITY_FAIL)
		{
			while( 1U );   /* Handle the integrity fail state */
		}
	}
#endif

	/* enable MemManageFault, BusFault, UsageFault in SCB to use individual exception handler. Otherwise, the HardFault handler is used */
	//S32_SCB->SHCSR |= (S32_SCB_SHCSR_MEMFAULTENA_MASK|S32_SCB_SHCSR_BUSFAULTENA_MASK|S32_SCB_SHCSR_USGFAULTENA_MASK);

	/**************************************************/
	/* eMcem example code                             */
	/**************************************************/
	/* User can get FCCU fault first if previous reset is FCCU FUNC Reset which is caused by certain FCCU fault */
	eMcem_GetErrors( &faultContainer );
	if(faultContainer.au32Faults[0] | faultContainer.au32Faults[1] | faultContainer.au32Faults[2])
	{
		printf("faultContainer = 0x%08X, 0x%08X, 0x%08X. \r\n", (unsigned int)faultContainer.au32Faults[0], (unsigned int)faultContainer.au32Faults[1], (unsigned int)faultContainer.au32Faults[2]);
	}

	/* Initialize eMcem */
    if( eMcem_Init( &eMcem_Config_0 ) != (Std_ReturnType)E_OK )
    {
		/* Handle initialization failure */
		printf("FCCU init failure.\r\n");
		printf("DCMROD3=0x%08X, DCMROD4=0x%08X, DCMROD5=0x%08X.\r\n", (unsigned int)IP_DCM_GPR->DCMROD3, (unsigned int)IP_DCM_GPR->DCMROD4, (unsigned int)IP_DCM_GPR->DCMROD5);
		printf("Trigger software destructive reset.\r\n");
		while( 1U );
	}
	printf("FCCU init completed.\r\n");
}

END