Computer Program Code Size Partitioning System for Multiple Memory Multi-Processing Systems

A computer program product for computer program code size partitioning for multiple memory multi-processor systems, the computer program product having a medium with a computer program embodied thereon, the computer program, when executed on a computing device, causes the computing device to:

• identify at least one system parameter of a computer system comprising one or more disparate processing nodes;
• receive computer program code comprising a program to be run on the computer system;
• generate a program representation based on received computer program code;
• identify at least one single-entry-single-exit (SESE) region based on the program representation;
• identify at least one store-size-specific SESE region based on identified SESE regions and the at least one system parameter, thereby forming a specified node;
• identify SESE regions proximate to the identified store-size-specific SESE region based on an ability of the specified node to process the identified proximate SESE regions, wherein the identified proximate SESE regions are grouped with the identified store-size-specific SESE regions and are configured to be compiled for the specified node;
• present the identified store-size-specific SESE regions and the identified proximate SESE regions to a user;
• receive user input from the user as to a grouping that should be used for the identified store-size-specific SESE regions and the identified proximate SESE regions;
• group each store-size-specific SESE region with at least one proximate SESE region into a node-specific subroutine based on the user input;
• modify the computer program code based on the node-specific subroutines;
• and
• compile the modified computer program code including each node-specific subroutine based on a specified node characteristic.

A compiler, comprising:

• an object file reader module configured to identify at least one system parameter of a computer system comprising one or more disparate processing nodes;
• a whole program analyzer and optimizer module coupled to the object file reader module and configured:
• to receive computer program code comprising a program to be run on the computer system,
• to generate a program representation based on received computer program code,
• to identify at least one single-entry-single-exit (SESE) region based on the program representation,
• to identify at least one store-size-specific SESE region based on identified SESE regions and the at least one system parameter, thereby forming a specified node,
• to identify SESE regions proximate to the identified store-size-specific SESE region based on an ability of the specified node to process the identified proximate SESE regions, wherein the identified proximate SESE regions are grouped with the identified store-size-specific SESE regions and are configured to be compiled for the specified node,
• to present the identified store-size-specific SESE regions and the identified proximate SESE regions to a user, and
• to receive user input from the user as to a grouping that should be used for the identified store-size-specific SESE regions and the identified proximate SESE regions, and
• a store size partitioning module coupled to the whole program analyzer and optimizer module and configured:
• to group each store-size-specific SESE region with at least one proximate SESE region into a node-specific subroutine based on the user input,
• to modify the computer program code based on the node-specific subroutines, and
• to compile the modified computer program code including each node-specific subroutine based on a specified node characteristic.