Method and system for automatic generation of processor datapaths
Systems and method for automatically generating a set of shared processor datapaths from the description of the behavior of one or more ISA operations is presented. The operations may include, for example, the standard operations of a processor necessary to support an application language such as C or C++ on the ISA. Such operations, for example, may represent a configurable processor ISA. The operations may also include one or more extension operations defined by one or more designers. Thus, a description of the behaviors of the various standard and/or extension operations that compose the ISA of an instance of a standard or configurable processor is used to automatically generate a set of shared processor datapaths that implement the behavior of those operations.
1. A computer-implemented method for automatically generating a plurality of shared datapaths, comprising:
receiving a behavior description for each of a set of operations;
automatically generating the plurality of shared datapaths from a plurality of seed datapaths to implement behavior of the set of operations; and
identifying common hardware resources of the plurality of shared datapaths, wherein certain datapaths of the plurality of shared datapaths are automatically merged based on the received behavior description to share the identified common hardware resources.
2. The method of claim 1 , wherein at least one of the shared common hardware resources contains parts of a multiplier.
3. The method of claim 1 , wherein the set of operations is of an instruction set architecture (ISA) and includes at least one base instruction set architecture operation and at least one designer-defined extension operation, and wherein at least one of the shared common hardware resources is used by the at least one base instruction set architecture operation and the at least one designer-defined extension operation.
4. The method of claim 1 , wherein the behavior of two or more operations of the set of operations is described with a single shared datapath of the plurality of shared datapaths.
5. The method of claim 1 , wherein the hardware comprising the plurality of shared datapaths includes a register transfer level (RTL) description corresponding to a hardware implementation of the plurality of shared datapaths in a processor.
6. The method of claim 1 , wherein automatically generating includes constraining timing requirements of input and output signals of the datapaths.
7. The method of claim 1 , wherein at least one of the generated datapaths is scheduled to span a plurality of pipeline stages.
8. The method of claim 7 , wherein automatically generating includes moving at least one of the common hardware resources between pipeline stages to increase sharing.
9. The method of claim 1 , wherein automatically generating includes generating the datapaths so a number of pipeline registers is reduced.
10. The method of claim 1 , wherein at least one generated datapath uses a same hardware resource in a least two different pipeline stages for a same operation of the set of operations.
11. The method of claim 1 , wherein automatically generating is performed such that a clock period of the resulting hardware does not exceed a designer-specified limit.
12. The method of claim 1 , wherein automatically generating is performed such that an area of the resulting hardware does not exceed a designer-specified limit.
13. The method of claim 1 , wherein at least one of the shared common hardware resources contains part of a shift unit.
14. The method of claim 1 , wherein automatically generating is performed in accordance with an estimate of a hardware area of each logic operator associated with the set of operators.
15. The method of claim 1 , wherein automatically generating is performed in accordance with an estimate of a hardware timing delay of each logic operator associated with the set of operators.
16. The method of claim 1 , wherein at least one operation of the set of operations is a designer-defined extension operation.
17. The method of claim 16 , wherein at least one operation of the set of operations is a base operation of the ISA.