Data-handling system with operatively interconnected processors

• US 4,334,305 A
• Filed: 02/14/1980
• Issued: 06/08/1982
• Est. Priority Date: 02/15/1979
• Status: Expired due to Term

• Current Assignee: Cselt - Centro Studi E Laboratori Telecommunicazioni SPA
• Original Assignee: Cselt - Centro Studi E Laboratori Telecommunicazioni SPA
• Inventors: Girardi, Gerardo
• Primary Examiner(s): Olms, Douglas W.
• Application Number:
  US06/121,445
• Time in Patent Office: 845 Days
• Field of Search: 370/86, 370/89, 370/92, 364/200
• US Class Current: 370/359
• CPC Class Codes: G06F 15/17337 Direct connection machines, e.g. completely connected computers, point to point communication networks coupling between buses G06F13/4004
• CPC Class Codes: H04L 12/433 with asynchronous transmission, e.g. token ring, register insertion

Abstract

A number of data processors have access, via respective interfaces, to a common ring memory made up of cascaded shift-register sections. Each interface comprises a pair of shift registers or a pair of read/write memories one of which, at a given instant, is operatively connected to the ring memory while the other coacts with the processor, their roles being interchanged from time to time by a switchover signal from a logic network in response to an instruction from the associated processor. The logic networks of all interfaces receive clock pulses from a common time base and, optionally, may be able to inhibit temporarily the emission of such pulses pending completion of an exchange of information with the ring memory.

Claims

1. In a data-handling system, in combination:

a multiplicity of data processors operating independently of one another;

a ring memory including a multiplicity of cyclically interconnected shift registers;

a multiplicity of interfaces each connected to a respective processor and interposed between adjoining shift registers of said ring memory for enabling the exchange of data between the ring memory and the respective processor, each of said interfaces including a pair of data stores, circuit means for temporarily inserting one of said data stores in a series connection between said adjoining shift registers while simultaneously establishing a two-way connection between the other of said data stores and the respective processor, and switchover means for interchanging said connections in response to an instruction from the respective processor; and

a source of clock pulses common to all said interfaces for timing the circulation of data through said ring memory by way of a signal path including the series-connected data store of each of said interfaces.

2. The combination defined in claim 1 wherein each of said interfaces further comprises comparison means connected to the respective processor and to said signal path for actuating said switchover means only upon detecting on said signal path an identification code matching said instruction.

3. The combination defined in claim 1, further comprising a signal generator common to all said interfaces for informing same of the nature of oncoming data as a criterion for actuation of said switchover means in response to said instruction.

4. The combination defined in claim 1, 3 or 4 wherein said interfaces are provided with individual connections to said source for temporarily inhibiting the emission of said clock pulses during actuation of said switchover means.

5. The combination defined in claim 1, 3 or 4 wherein said data stores are a pair of additional shift registers.

6. The combination defined in claim 1, 3 or 4 wherein said data stores are a pair of read/write memories, each of said interfaces further comprising counting means stepped by said clock pulses for sequentially addressing respective cells of said read/write memories.