eDMA Lockstep Implementation

取消
显示结果 
显示  仅  | 搜索替代 
您的意思是: 
已解决

eDMA Lockstep Implementation

跳至解决方案
1,167 次查看
jtillman
Contributor IV

I want to be sure we are interpreting the S32V234's RAM FMEDA cell `eDMA Lockstep Enabled` correctly

The safety manual states, "As eDMA is a replicated module, no software action is needed to detect faults inside the module."

Does this mean lockstep is enabled for every transaction by default? i.e. We (software engineers) don't need to synchronize 2 identical DMA transfer requests in software to implement this requirement?

I see NCF[10] is dedicated to eDMA Lockstep errors so if the above is true, it would seem our only action is to make sure we set the appropriate reaction to handling when that error is asserted

Thanks

标签 (1)
标记 (3)
0 项奖励
1 解答
1,150 次查看
Daniel_Wax
NXP Employee
NXP Employee

it is confirmed that the DMA lockstep is always enabled in S32V234. The customer needs to set the appropriate reaction for lockstep error which is signaled to FCCU at NCF[10].

在原帖中查看解决方案

1 回复
1,151 次查看
Daniel_Wax
NXP Employee
NXP Employee

it is confirmed that the DMA lockstep is always enabled in S32V234. The customer needs to set the appropriate reaction for lockstep error which is signaled to FCCU at NCF[10].