Hi, I would like to ask about the diagnostic function of SRAM ECC:
(1) If a single BIT ECC error is found at a certain address, will the SECDED module rewrite the corrected data into SRAM? Do I mean to correct the original SRAM data, or just correct the read data?
(2) In SRAM diagnosis, the hardware provides Hard error cache and livelock functions. Is the function of Hard error cache to store the address of 1bit ECC failure?
(3) If a 1-bit ECC failure occurs at a certain address, and then the SECDED module corrects it, then the Hard error cache will also record the address where the error occurred. If this address has a 1-bit ECC failure again, then the CPU By checking the Hard error cache, it is found that this address has failed, and no error alarm will be generated. Will the CPU continue to correct the wrong BIT bit on this address?
(4) If a 1-bit ECC failure occurs in multiple addresses, will these addresses be recorded in the Hard error cache?