Abstract: A computing device is provided, including a processor configured to receive source code at a compiler. The source code may include at least one compound conditional having a plurality of conditions. For each condition of the plurality of conditions, the source code may further include a respective code block including instructions to evaluate the condition. For each ordering of a plurality of orderings of the plurality of conditions, the processor may determine that the ordering satisfies one or more legality constraints. For each ordering of the plurality of orderings that satisfy the one or more legality constraints, the processor may determine a respective estimated computational cost for that ordering. The processor may reorder the plurality of conditions to have an ordering that has a lowest estimated computational cost of the plurality of orderings that satisfy the one or more legality constraints.

Abstract: A computing device is provided, including a processor configured to receive source code at a compiler. The source code may include at least one compound conditional having a plurality of conditions. For each condition of the plurality of conditions, the source code may further include a respective code block including instructions to evaluate the condition. For each ordering of a plurality of orderings of the plurality of conditions, the processor may determine that the ordering satisfies one or more legality constraints. For each ordering of the plurality of orderings that satisfy the one or more legality constraints, the processor may determine a respective estimated computational cost for that ordering. The processor may reorder the plurality of conditions to have an ordering that has a lowest estimated computational cost of the plurality of orderings that satisfy the one or more legality constraints.

Abstract: One embodiment of the present invention sets forth a method for causing thread convergence. The method includes determining that a control flow graph representing a first section of a program includes at least two non-overlapping paths that extend from a first divergent node to a candidate node. The method also includes determining that the first divergent node is not a dominator of the candidate node or that the candidate node is not a post-dominator of the first divergent node. The method further includes identifying an external node and inserting a first instruction configured to cause a predicate variable to be set to true for a first set of threads that is to execute the external node. The method additionally includes inserting into the program a second divergent node configured to cause various threads to execute or not execute a first control flow path associated with the external node.