Bull Hn Information Systems, Inc.

Honeywell Information Systems Inc.

Honeywell Bull, Inc.

George J Barlow

James W Keeley

Chester M Nibby, Jr

John W Conway

Richard C Zelley

Donald L Smith

Richard A Lemay

Arthur Peters

Kenneth E Bruce

David B O'Keefe

Frank V Cassarino, Jr

Elmer W Carroll

John J Bradley

Richard P Brown

Ralph M Lombardo, Jr

Michelle A Pence

Victor M Morganti

Raymond D Bowden, Iii

Myrl Kennedy Bailey, Jr

Thomas F Joyce

William E Woods

Jian-Kuo Shen

Philip E Stanley

Robert B Johnson

Richard P Kelly

Robert C Miller

Bruce H Tarbox

Jeffrey S Somers

James B Geyer

Edward F Getson, Jr

John W Bradley

Patrick E Prange

Douglas L Riikonen

Marc E Sanfacon

Amy E Gilfeather

George J Bekampis

Wallace A Martland

Robert V Ledoux, Deceased

In order to provide communication between two processors in a data processing system, a target processor includes apparatus that can store data signal groups from a source processor. Having stored a data signal group from the source processor, the target processor notifies the source processor of the receipt of the data signal group. In response to the presence of the stored data signal group, the target processor executes a interprocessor command or instruction identified by the transferred data signal group. The source processor at a preselected time, executes an instruction to determine if the command designated by the data signal group stored in the target processor has been executed. The commands specified by the transferred data signal groups can be executed under hardware control by the target processor in a relatively short time immediately following completion of the instruction in execution in the target processor at the time of the transfer of the data signal group. The interprocessor communication has only a minor impact on the performance of the source processor and the target processor because of the command implementation by apparatus.

Apparatus and method for interprocessor communication

A multiprocessor computer system includes a number of processors, each processor having an interrupt mechanism and connecting in common to a system bus over which interrupt requests are communicated. When a processor accepts an interrupt request from another processor, it generates an acknowledge response on the system bus. If such processor contains a previous and pending interrupt request of an equal or higher priority level, it generates a not acknowledge response on the system bus and refuses the interrupt request. At the completion of servicing an interrupt request, each processor places on the system bus, an interrupt completed command including an address identifying such processor, a code designating a priority level to which it has switched and a code indicating that the processor has completing servicing an interrupt request. Each interrupt mechanism further includes interrupt retry means containing a refused interrupt register means for storing the channel address of the processor generating the not acknowledge in response and the priority level code contained in the interrupt request, and a level monitor logic unit connected to the system bus. The level monitor logic unit detects interrupt completed commands and compares the channel address and priority level code in the interrupt completed command to channel address and level code stored in the refused interrupt register means. When the processor channel addresses match and the level code is less than the level stored in the refused interrupt register means, the monitor logic unit generates a retry interrupt output. The processor retries the corresponding previously refused interrupt request upon receipt of the retry interrupt output.

Apparatus and method for providing more effective reiterations of

A bus interface priority network provides access to a system bus by a plurality of different types of requestors as a function of the types of transactions they are required to process. The network includes programmable circuit for identifying the type of requestor and selecting a delay for accessing the system bus on the basis of requestor type thereby eliminating the need to adjust timing to the slowest requestor.

Programmable system bus priority network

A multiprocessor computer system having a first processor having a first interrupt mechanism for generating interrupt requests, a second processor having a second interrupt mechanism, and a system bus for communicating interrupt requests from the first processor to the second processor. The second interrupt mechanism is responsive to an interrupt request by generating an acknowledge response on the system bus when the second processor accepts the interrupt request and generating a not acknowledge response on the system bus when the second processor contains a previous and pending interrupt request of higher level and refuses the interrupt request. The second interrupt mechanism is responsive to the completion of servicing of an interrupt request by the second processor by placing on the system bus an interrupt completed command, which includes an address identifying the second processor and a code indicating that the second processor has completing servicing an interrupt request. The first processor includes an interrupt retry means, which includes a refused interrupt register means responsive to a not acknowledge response from the second processor in response to an interrupt requested from the first processor for storing the channel number of the second processor, and level monitor logic connected from the system bus.

A data processing system includes a system management unit, a number of central subsystems, a main memory and a number of peripheral subsystems all coupled in common to a system bus. Any subsystem may generate a command which includes a first field specifying a destination subsystem, a second field specifying the operation the destination subsystem is to perform. If a response is required, the subsystem generating the initial command may specify a third subsystem for receiving the response command.

Data processing system having a bus command generated by one subsystem

A multiprocessor computer system includes first processors, second processors, a system management means for performing system management functions, including detecting pending power shut-downs and sending power shut-down messages addressed to each of the first processors warning of pending power shut-downs, and a system bus for communication between the first and second processors and the system management means, including the communication of pending power shut-down messages. The first processors include interrupt handling means responsive to pending power shut-down messages for executing power shut-down routines for placing the first processors into a known state before power termination, but the second processors inherently do not include a power shut-down capability. In each of the second processors, a power shut-down means is provided to place the second processors in a known state before a power termination, including a bus monitor connected from the system bus and responsive to any power shut-down message addressed to a first processor for generating an output indicating the occurrence of a power shut-down message to a first processor. The second processor also includes non-maskable interrupt logic connected from the power shut-down message output of the bus monitor and responsive to the power shut-down message output of the bus monitor for generating a non-maskable interrupt output to the second processor. The second processor is in turn responsive to a non-maskable interrupt output of the non-maskable logic for querying the non-maskable logic to determine the nature of the interrupt, and responsive to the indicated occurrence of a power shut-down message to any first processor for executing a power shut-down routine for placing the second processor in a known state before the termination of power.

Means for providing a graceful power shut-down capability in a

A data processing unit includes a number of tightly coupled central subsystems, a number of peripheral subsystems, a main memory and a system management facility all coupled in common to a system bus. The system management unit has top priority on the system bus and includes centralized resources which provide apparatus for indicating the status of power and temperature, booting the subsystems, testing the subsystems, timing central subsystem functions, and allowing local and remote maintenance access to the subsystems. The system management facility receives commands from the central subystem to read from and write into the timers as well as to read the status of the overall system. The system management facility generates special commands to the central subsystem to indicate when the timers have decremented to ZERO as well as special commands to aid in hardware and software debugging.

Multiprocessor system with centralized initialization, testing and

A multiprocessor system includes a plurality of central subsystem (CSS) units, a plurality of memory units and input/output units which connect in common to a system bus for transferring requests between a pair of units on a priority basis defined by a distributed bus priority network included as part of the system bus. A private bus (P bus) connects all of the CSS units and memory units in common for high speed block data transfers. Each CSS unit includes input circuits which couple to the priority network for detecting when the system bus is in an idle state. P bus logic circuits couple to the P bus and generate a transfer request in response to a request from its CSS unit only when the P bus is detected to be in an idle state. The idle signals from both buses are used to generate a system bus request for P bus access only when both buses are in an idle state so as to eliminate the need to contend for system bus use.

Minimum contention processor and system bus system

A multiprocessor system includes a plurality of identical central subsystem (CSS) units, a plurality of memory subsystem units and input/output units which connect in common to a system bus. Requests are transferred between a pair of units on a priority basis defined by a distributed bus priority network included as part of the system bus. Each CSS unit includes cycle stealer logic circuits which grant bus cycles on a round robin basis. The cycle stealer logic circuits are connected to receive high priority request signals from the network and refuse acceptance of a cycle granted to such CSS unit as a low priority requester thereby passing it along to a next lower priority CSS unit.

Method and apparatus for a high performance round robin distributed bus

Memory retry logic to improve the resilience of system memory operations with respect to system errors or faults which prevent a memory read operation from being completed on a first attempt by allowing the memory to retry the operation once. The memory retry logic detects the occurrence of an improper response from the system element requesting a memory read operation when attempting to initiate the system bus operation for reading the data from memory to the requesting element and, if an improper response indicating that the requesting element is not accepting the bus operation request is detected, stores the memory operation request and the requested data and retries the data transmission on the next available bus cycle. If the memory receives an improper response of a specified type during a bus operation of a memory burst, the memory will terminate the operation and proceed to the next requested operation.

Method and apparatus for memory retry

A means for ascertaining the health, or basic operational status, of a system unit. A 'health check' provides an indication of either 'yes', the system unit is operational, or 'no', the system unit is either inoperative or there is a question as to whether the system is operational. The test is performed by requesting that the system unit perform a high priority 'short' operation and noting the response provided to the request; the actual execution of the request is unimportant and it is the response of the unit under test to the receipt of the request for a bus operation that is the actual indicator of the status of the unit being tested. The requested operation is not directed at the unit whose operational status is to be determined, but instead at a bus interface unit which performs bus operations for the unit to be tested and whose responses to requests for bus operations are effected by the operational status of the unit that is to be tested. The operation is not directed at an actual element in the bus interface unit, but at a phantom, or nonexistent, element.

Method and apparatus for performing health tests of units of a data

A method and apparatus for resetting memory state when power is applied to the system. The memory has memory elements, a refresh clock and a refresh counter for counting refresh cycles and providing refresh signals to the memory elements, the memory elements and refresh means being connected from the power system and from a battery back-up means. A state detection means is connected from the refresh counter for detecting a change in state of the refresh counter to a state equivalent to the reset state of the refresh counter and asserting a state change signal. A means responsive to the state change signal and to the occurrence of the reset signal provides a memory controller reset signal, so that the memory controller reset signal occurs in synchronization with the change of state of the refresh counter to a state equivalent to the refresh counter reset state. The memory reset further includes a time-out counter means responsive to the assertion of the reset signal and to the refresh clock for counting refresh cycles in synchronization with the refresh counter. A time-out detection means is responsive to the time out counter means for providing a time out signal when the time-out counter has counted a refresh period plus one clock cycle and to the state change signal for providing the memory controller reset signal when the time-out counter has counted a refresh cycle plus one clock period and the state change signal has not been asserted.

Method and apparatus for resetting a memory upon power recovery

The pipelined central processing system (CSS) units of a multiprocessor system are tightly coupled to connect in common to a system bus for sharing main memory and input/output controllers/devices. The CSS includes several circuit boards for the different VLSI circuit chip pipelined stages and associated control circuits in addition to the bus interface unit (BIU) circuits. Each board includes one or more unusual event (UEV) detector circuits for signaling when the behavior of a stage is abnormal. The UEV fault signals from each board are collected by the BIU board. When a UEV fault is detected, the BIU board circuits prevent any further communications with the system bus and broadcasts the UEV fault signal to the other boards causing the different pipelined stages to emulate the completion of the instructions within the pipeline thereby flushing it. It is thereafter placed in a nonpipelined mode. Control circuits execute a UEV handler microcode routine which reads the contents of syndrome registers included on each board containing the UEV indicator states in addition to error signals into register file working locations. The UEV signal is then cleared enabling the BIU to resume communications with the sytem bus.

Recovery method and apparatus for a pipelined processing unit of a

A multiprocessor system includes a plurality of identical central subsystem (CSS) units, a plurality of memory and input/output (I/O) subsystem units which connect in common to an asynchronous bus system. Each CSS subsystem unit includes a cancel command mechanism for enabling each such unit to effectively withdraw from the asynchronous bus or switch the state of a resource such as a memory or I/O lock mechanism included in such subsystem without otherwise disturbing the state of such subsystems.

Cancel mechanism for resilient resource management and control

A multiprocessor system includes a system management facility (SMF) unit, a plurality of central subsystem (CSS) units, a plurality of memory subsystem units and first and second pluralities of input/output units which connect in common to a system bus. Requests are transferred between a pair of units on a distributed bus priority network included as part of the system bus on the basis of the unit's physical position on the bus relative to one end of the bus. The SMF unit positioned at the high priority end of the bus includes fast recovery bus request logic circuits which connect to the high priority request line of the priority network. Each of the CSS units positioned after the SMF unit on either side of the memory subsystems includes bus request logic circuits which connect only to the low priority request line. The memory subsystems each include bus request logic circuits which connect to both the high and low priority request lines for accepting and granting cycles from higher and lower priority units. This enables the positioning of the first and second pluralities of input/output units positioned below the memory subsystems to operate as high and low priority requestors.

Growing community of inventors

Tewksbury, MA, United States of America

George J Barlow