Smart recompilation of performing matchup/difference after code generation

• US 5,367,683 A
• Filed: 06/26/1992
• Issued: 11/22/1994
• Est. Priority Date: 06/26/1992
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Digital Equipment Corporation (HP Inc.)
• Inventors: Brett, Bevin R.
• Primary Examiner(s): Shaw, Gareth D.
• Assistant Examiner(s): BUTLER, DENNIS
• Application Number:
  US07/904,816
• Time in Patent Office: 879 Days
• Field of Search: 395/425, 395/700, 364/254.3, 364/262.5, 364/280.4, 364/280.5
• US Class Current: 717/136
• CPC Class Codes: G06F 8/48 Incremental compilation software reuse G06F8/36

Abstract

A method and system for compiling a source program using smart recompilation. The invention allows fragments to contain "invocation specific" information, which is generated during a code generation phase of compilation. A hint generator attempts to preserve values of the invocation specific information between successive invocations of the compiler.

Claims

1. In a computer system, including a memory, a method of translating a source program comprising the steps of:

translating a first source program and producing a first global context table, said first global context table containing global information describing definitions from said first source program which can be referenced by source programs other than said first source program, said translating of said first source program including a first code generation step;

dividing, after said first code generation step, the information contained in said first global context table into a first group of fragments;

translating a second source program and producing a second global context table containing global information describing definitions from said second source program which can be referenced by source programs other than said second source program, said translating of said second program including a second code generation step;

dividing, after said second code generation step, the information contained in said second global context table into a second group of fragments;

determining whether a fragment in said second group has an equivalent fragment in said first group; and

translating source programs which are dependent on fragments from said second group determined not to have an equivalent fragment in said first group.

2. The method of claim 1, wherein the step of dividing the information contained in said first global context table includes assigning a version identifier to each fragment in said first group of fragments;

and

wherein, the method further includes assigning a new version identifier to fragments in said second group that do not have an equivalent fragment in said first group.

3. The method of claim 1, wherein the step of dividing the information contained in said first global context table further includes assigning a version identifier to each fragment in said first group of fragments;

and

wherein the method further includes maintaining that version number for a fragment in said second group having an equivalent fragment in said first group.

4. The method of claim 1, wherein the first global context table includes semantic information comprising at least one of a semantic nature identifier, a type identifier, a number of parameters, and a type of parameters.

5. The method of claim 1, wherein the first global context table includes invocation specific information comprising at least one of a global symbol name, programming variable information including size and representation information, and register offset information.

6. The method of claim 1, wherein said translating the first and second source programs respectively produces first and second object files, and wherein the first and second object files each contain a global symbol table.

7. The method of claim 6, wherein the global symbol table further includes a program section contribution table.

8. The method of claim 6, wherein the global symbol table further includes a global symbol name, a program section name, and a program section offset.

9. The method of claim 7, wherein the program section contribution table further includes a program section name and program section contents.

10. The method of claim 1 further including the step of matching fragments in said first group with corresponding fragments in the second group.

11. The method of claim 1 further comprising storing said first global context table in a library.

12. The method of claim 1 wherein said second source program is a subsequent version of the first source program.