SPRUJ53B April 2024 – September 2024 TMS320F28P550SJ , TMS320F28P559SJ-Q1
For each correctable error, the count in the correctable error count register increments by one. When the value in this count register becomes equal to the value configured in the correctable error threshold register, an interrupt is generated to the CPU, if the interrupt is enabled in the correctable interrupt enable register. The user needs to configure the correctable error threshold register based on the system requirements. Also, the address for which the error occurred, gets latched into a register and a flag also gets set in a status register.
If there are uncorrectable errors, an NMI gets generated for the CPU. In this case also, the address for which the error occurred gets latched into a register, and a flag gets set in a status register.
Table 3-15 summarizes different error situations that can arise. These need to be handled appropriately in the software, using the status and interrupt indications provided.
Access Type | Error Found In | Error Type | Status Indication | Error Notification |
---|---|---|---|---|
Reads | Data read from memory | Uncorrectable Error (Single-bit error for Parity RAMs OR Double bit Error for ECC RAMs) |
Yes - CPU Read Error Address Register Data returned to CPU is incorrect | NMI for CPU access |
Reads | Data read from memory | Single-bit error for ECC RAMs | Yes - CPU Read Error Address Register Increment single error counter | Interrupt when error counter reaches the user programmable threshold for single errors |
Reads | Data read from PIE memory | Parity error | Yes - PIE Parity Error sets bit in MEM_CFG_REGS | Bit set in MEM_CFG_REGS |
Reads | Address | Address error | Yes - CPU Read Address Error Register Data returned to CPU is incorrect | NMI to CPU for CPU access |
During debug accesses, correctable as well as uncorrectable errors are masked.