An error handling and reporting mechanism is capable of taking advantage of sophisticated error analysis performed after clocks have been stopped in response to an error detected in a controller. The controller provides services in a data processing system in response to requests for controller services from a plurality of requestors. The controller includes a plurality of ports for storing requests for controller services. A plurality of servers is coupled to the plurality of ports, and perform separate services associated with the requests for controller services stored in the plurality of ports. An error reporting mechanism is included which is responsive to a detected error in a particular server associated with a request in a particular port, for posting error status in the particular port and causing clock stoppage within a clock stop latency period. An error analysis mechanism analyzes the detected errors during the clock stoppage. Error handling logic is coupled with the error analysis mechanism, and is responsive to the posted error status in the ports, for notifying a requestor of an error status posted with a request in the particular port. The error handling logic includes a stall counter, which stalls the error handling mechanism in response to the posted error status for at least the clock latency period so that the clock stoppage occurs and the error analysis mechanism completes error analysis before the requestor is notified.

A CPU connected to a memory storing module programs has an instruction cache for reading a module header heading a module program and module instructions following the module header out of the memory, and storing and transferring them. An operand cache reads module data included in the module program of interest out of the memory, and stores and transfers them. A processing section performs arithmetic and logical operations in pipelining with the module header of the module program transferred from the instruction cache and with the module data transferred from the operand cache. A pipelined instruction control controls other circuitry of the CPU by decoding the module instructions of the module program transferred from the instuction cache. The CPU transfers the module header to the processing section via the instruction cache while writing the whole block of module header and module instructions to the instruction cache. Alternatively, the CPU may transfer the module header to the processing section via the operand cache while writing the whole block of module header and successive module instructions to the instruction cache.

A cache controller sits in parallel with a microprocessor bus and includes a tag RAM for associatively searching a directory for cache data-array addresses. Two normal address latches are provided to capture a cycle address in case the current cycle is extended by a pending tag RAM access. At any time, except when the next cycle has started, but during which the current cycle is in progress, one latch is open to an input buffer such that the input address is latched by that latch. The other latch holds the current cycle address until the cycle ends. The current cycle can be extended with snoops. The current cycle address has to be maintained as long as the cycle is still in progress. In the meantime, the external cycle might have ended and a next cycle started. The second address latch is used to capture the address corresponding to this new cycle. As signal selects which of the two latches will supply the address via a MUX to the tag RAM. Logic switches the two input address latches so that the tag RAM sees a constant address for the duration of a cycle while at the same time a new cycle address can be received.

The decoding of certain instructions cause an instruction unit of a production line data processing system to stall. Instructions still in the production line are executed, but no new instructions are sent into the production line until the instruction that caused the stall condition is executed. The execution of the instruction that caused the stall is completed by an execution unit taking over control of an address unit.

A pipelined data processing system comprises a series of pipeline stages through which instructions pass sequentially. If an error associated with a particular instruction (e.g. an illegal instruction) is detected, a Mayday signal is produced. This does not become active immediately. Instead it is passed down the pipeline along with the instruction that caused it, and becomes active only if that instruction attempts to terminate at the end of the pipeline. When the Mayday becomes active, it causes an interrupt to a diagnostic processor. If, on the other hand, the instruction that caused the Mayday is aborted, or does not attempt to terminate, then the Mayday is discarded and does not become active. This prevents unnecessary diagnostic actions.