SPRUJ17I March 2022 – August 2025 AM2631 , AM2631-Q1 , AM2632 , AM2632-Q1 , AM2634 , AM2634-Q1
The configuration MMRs for each Error Group N are backed by 3 flops in order to protect against single or double-bit errors. When written, all 3 bits are set to the same value. When read (and for functioning of the internal state machines) the value is the OR of all 3 bits. Whenever any of the bits disagree, the Configuration Error interrupt is asserted (if enabled). The registers covered by this mechanism are below:
Config Error Interrupt Raw Status/Set Register (Base Address + 0x10)
Config Error Interrupt Enabled Status/Clear Register (Base Address + 0x14)
Config Error Interrupt Enabled Clear Register (Base Address + 0x1C)
Error Group N Event Raw Status/Set Register (Base Address + 0x400 + N*0x20 + 0x00)
Error Group N Interrupt Enabled Status/Clear Register (Base Address + 0x400 + N*0x20 + 0x04)
Error Group N Interrupt Enabled Set Register (Base Address + 0x400 + N*0x20 + 0x08)
Error Group N Interrupt Priority Register (Base Address + 0x400 + N*0x20 + 0x10)
Error Group N Error Pin Influence Set Register (Base Address + 0x400 + N*0x20 + 0x14)
Error Group N Error Pin Influence Clear Register (Base Address + 0x400 + N*0x20 + 0x18)
The Error Pin Control Register (Base Address + 0x40) contains a multi-bit field. The Error Pin Counter Pre-Load Register (Base Address + 0x4C) should also be read and checked periodically by software. The key value ensures normal operation on the error pin and that an error even will be generated if one occurs. Software should periodically read check the KEY bit field value and make sure it is 0x0. If the value is not 0x0, software must re-write it to this key value (unless in test mode forcing an error on the pin) to ensure the normal operation.
Table 13-322 lists the KEY values and their respective meaning.
| ESM_PIN_CTRL[3-0] KEY | Description |
|---|---|
| N/A | Registers are inaccessible. Device disables the I/O and pulls down internally. |
| 0x0 (Normal) | Normal operation mode - Error pin will activate when an enabled error event occurs. |
| 0xA (Force Error) | Force error mode - Forces the error pin active. To clear the error pin (return to the ESM_IDLE state) write this field back to normal mode (writing a CLEAR event will also work). Force error mode must be set only while in IDLE. Attempting force error while in another state will have no effect. |
| 0x5 (CLEAR) | CLEAR Event - generates a CLEAR event to the ESM state machine. KEY will return to normal mode (0x0) on the next cycle. |
| Other Values | All other values - Normal mode. Writing any of these values will have no effect. When reading any of these values indicates that one or more bits have experienced a single event upset, software should write the field back to 0x0. The ESM will continue to operate in normal mode. |