Other Parts Discussed in Thread: TMDX570LC43HDK
Hi,
I need some advice on a couple of problems that I have encountered while using the CRC module on the Hercules TMS570LC4357.
I have implemented the CRC as follows:
- I have followed the guidance in SPNA235.
- I have implemented the CRC using Full-CPU mode.
- I have defined a linker generated CRC table for the .intvecs, .text, .const and .cinit.
The first problem that I encountered was that the call to crcSignGen(), when calculating the CRC for .const, causes ESM error 2.3 to be raised.
This is presumably due to an ECC problem.
Question 1. Do you know what circumstance might cause the calculation of the CRC to raise ESM error 2.3?
Note that my HL_sys_link.cmd already includes linker calculated ECC.
Originally, I was using vfill for VECTORS, FLASH0 and FLASH1.
i.e.
VECTORS (X) : origin=0x00000000 length=0x00000020 vfill=0xffffffff
FLASH0 (RX) : origin=0x00000020 length=0x001FFFE0 vfill=0xffffffff
FLASH1 (RX) : origin=0x00200000 length=0x00200000 vfill=0xffffffff
So, I tried changing vfill to fill and what I found was that the ESM error 2.3 was no longer raised.
Question 2. Do you know of any reason why changing vfill to fill would prevent ESM error 2.3?
The next problem that I encountered was that now the linker calculated CRC for .const is not the same as the CPU calculated CRC for .const.
The linker calculated CRC for .intvecs, .text and .cint are the same as the CPU calculated CRC for .intvecs, .text and .cint.
It is just the linker calculated CRC for .const that is not the same as the CPU calculated CRC for .const.
Question 3. Do you know of any reason why one CRC would be wrong while all of the other CRCs are correct?
Note that I have run this software on two different development boards (TMDX570LC43HDK) and I have obtained exactly the same result on both boards.
So, the cause of these problems cannot be due to the hardware.
Thank you.
