Aggressive Inlining for VLIW

Aggressive Function Inlining: Preventing Loop Blockings in the Instruction Cache

Aggressive function inlining can lead to significant improvements in execution time. This potential is reduced by extensive instruction cache (Icache) misses caused by code expansion involved with aggressive inlining. It is very difficult to predict which inlinings will lead to Icache conflicts, as the exact location of code lines in the executable depends on completing the inlining first. In t...

Profile-driven Inlining for Erlang

Partial Inlining

Technical Summary 1 Motivation Subprogram inlining is an optimization that has been studied extensively in the By replacing a procedure call with a copy of the body of the called procedure, it ooers a number of beneets: the overhead associated with the procedure call is eliminated; the inlined code can be optimized for the particular call site that was expanded; and the caller's code can potent...

Type-Safe Method Inlining

In a typed language such as Java, inlining of virtual methods does not always preserve typability. The best known solution to this problem is to insert type casts, which may hurt performance. This paper presents a solution that never hurts performance. The solution is based on a transformation that modifies static type annotations and changes some virtual calls into static calls, which can then...

Inlining as staged computation

Inlining and specialization appear in various forms throughout the implementation of modern programming languages. From mere compiler optimizations to sophisticated techniques in partial evaluation, they are omnipresent, yet each application is treated differently. This paper is an attempt at uncovering the relations between inlining (as done in production compilers) and staged computation (as ...