Hi,
I have created an exception handling in MPC5746R for monitoring RAM fault check in following situations
1. Peripheral RAM
2. System RAM
For testing peripheral RAM I'm using EIM(error injection module) to induce error and I am able to see the exception handling is hitting for peripheral RAM un-correctable error.
I could't find any methods to test system RAM by inducing errors.So my concern is whether the exception handling function will work for system RAM errors (or) is there any method to induce system RAM errors.
Regards,
RR
Hi, by MPC5744P's EIM module you can inject ECC fault only to transfer control descriptor of eDMA.
Apparently, you could use core registers for this purpose:
If you enable E2EECSR0[INVC]=1 error injection and setup certain mask to E2EECSR0[CHKINVT] bits, subsequent write to SRAM creates error and following read of this area causes bus error.
Hi,
Thanks for your reply,We tried it in MPC5746R as follows.
/* asm static void Generate_system_RAM_fault(void)
{
!
mfdcr r22, 511
addis r22, 0, 0x0000
addi r22, 0, 0x1000
mtdcr 511, r22
}
And this set register E2EECSR0[INVC] as 1.But unable to see any update in MEMU registers corresponding to system RAM errors.
we tried E2EECSR0[CHKINVT] bits with certian mask and the result is same.
Whether the method we applied is correct or not...?
Please provide your suggestion.
Hi, I have just created example code based on approach I have already described in the link mentioned by you:
Example MPC5744P 1b+2b_RAM_ECC_error_injection GHS714
This approach could be possible to use for RAM error injection on MPC5746R.