SPRUII0F May 2019 – June 2024 TMS320F28384D , TMS320F28384D-Q1 , TMS320F28384S , TMS320F28384S-Q1 , TMS320F28386D , TMS320F28386D-Q1 , TMS320F28386S , TMS320F28386S-Q1 , TMS320F28388D , TMS320F28388S
Error detection is done while reading the data from memory. The error detection is performed for data as well as address. For parity memory, only a single-bit error gets detected, whereas in the case of ECC memory, along with a single-bit error, a double-bit error also gets detected. These errors are called correctable errors and uncorrectable errors.
The following are characteristics of these errors:
Correctable errors get corrected by the memory controller module and then correct data is given back as read data to the master. It is also written back into the memory to prevent double-bit error due to another single-bit error at the same memory address.
Size of write access initiated by bus masters on CM subsystem can be byte or half-word or full word. Since ECC/Parity is calculated for 16-it word, in case of byte write access memory controller performs Read-Modify-Write operation (read 16-bit from RAM block, modify the specific byte with new data, calculate ECC/parity for new data and write the data and ECC/parity back into RAM block).