G - Physics – 06 – F
Patent
G - Physics
06
F
354/166, 354/225
G06F 11/22 (2006.01) G06F 11/16 (2006.01) G06F 11/27 (2006.01) G06F 7/00 (2006.01) G06F 7/48 (2006.01)
Patent
CA 1218748
83-318 Abstract of the Disclosure A mechanism for continually testing a floating point accelerator processor (FPAP) element or other processor element in a suitable multiprocessor system. At least two processors, such as an instruction execution processor (EU) and a FPAP, are connected to a common input bus to concurrently receive the same information (opcodes and operands). Both the EU and the FPAP decode the opcodes. When the FPAP decodes an opcode for an operation to be performed by the EU, the FPAP, instead of remaining idle while the EU operates, executes a diagnostic operation. The FPAP selects the particular diagnostic operation to perform in each instance from among a multiplicity of available diagnostic operations. The selection of a diagnostic operation is dependent on the instruction to be executed by the EU; in order to not slow down the overall execution rate of the system, a diagnostic operation is chosen whose execution time is matched to the execution time of the instruction being performed by the EU; that is, a diagnostic operation is selected such that the FPAP will finish the operation before the EU will finish executing its instruction. Operand data supplied to the EU on the input bus is used by the diagnostic operations, to add a degree of randomness to the test signals and permit detection of bits forced to a steady value of zero or one. For some diagnostic operations, one or more variables may be obtained from general purpose registers. The EU is not notified of the detection of an FPAP self-diagnostic error until the next operation which is to be performed by the FPAP, at which time such operation may be aborted or allowed to go to completion, depending on whether or not the source of the error will affect that operation. 83-318 23 Thus both the data paths and the control logic of the FPAP are checked during the execution of every non- floating point instruction. System response is not slowed, only an insignificant marginal cost is added and the diagnostic system is completely transparent to the programmer.
467090
Fossum Tryggve
Shively Milton L.
Digital Equipment Corporation
Smart & Biggar
LandOfFree
Method and apparatus for self-testing of floating point... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for self-testing of floating point..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for self-testing of floating point... will most certainly appreciate the feedback.
Profile ID: LFCA-PAI-O-1291808