Subject: MPC5605 memory-read issue at 0x00100010

[kasirajan]

Hello,

We have an issue with an ECU using MPC5605.

During an EOL memory-dump operation, our toolbox reads memory using a direct CPU byte access:

*((uint8_t *)address)

From the CAN log:

```text
Read 8 bytes from 0x00100008 -> successful
Read 8 bytes from 0x00100010 -> ECU stops responding and resets
```

We checked the MPC5606BK Reference Manual, `16234_MPC5606BRM`, Rev. 2:

- Table 3-1, page 49, shows `0x00100000 - 0x001FFFFF` as **Reserved**.
- Pages 871-872 describe the flash-memory map and flash-emulation mapping.

Could you please clarify:

1. Is `0x00100010` a valid/readable address for MPC5605?
2. If it is a reserved address, what happens when software reads it directly? Can it cause an exception or reset?
3. Which reference manual section should we use for the exact MPC5605 memory map?

We are also collecting the full MPC5605 part number and silicon revision from the ECU hardware.

Best regards,
kasirajan C

[petervlna · Accepted solution]

Hello,

Address 0x00100010 belongs to the range 0x00100000-0x001FFFFF, which is marked as Reserved in the MPC560xB memory map. Reserved regions must not be accessed by application software. Accessing such locations may generate a bus-related exception or other device-specific fault response, which can result in loss of communication or a reset if the exception is not handled.

Chapter 3
Memory Map

Best regards,

Peter

[kasirajan]

Thanks for your reply