Loop Rolling for Code Size Reduction

Abstract

Code size is critical for resource-constrained devices, where memory and storage are limited. Compilers, therefore, should offer optimizations aimed at code reduction. One such optimization is loop rerolling, which transforms a partially unrolled loop into a fully rolled one. However, existing techniques are limited and rarely applicable to real-world programs. They are incapable of handling partial rerolling or straight-line code.In this paper, we propose RoLAG, a novel code-size optimization that creates loops out of straight-line code. It identifies isomorphic code by aligning SSA graphs in a bottom-up fashion. The aligned code is later rolled into a loop. In addition, we propose several optimizations that increase the amount of aligned code by identifying specific patterns of code. Finally, an analysis is used to estimate the profitability of the rolled loop before deciding which version should be kept in the code.Our evaluation of RoLAG on full programs from MiBench and SPEC 2017 show absolute reductions of up to 88 KB while LLVM's technique is hardly distinguishable from the baseline with no rerolling. Finally, our results show that RoLAG is highly applicable to real-world code extracted from popular GitHub repositories. RoLAG is triggered several orders of magnitude more often than LLVM's rerolling, resulting in meaningful reductions on real-world functions.

HyBF: A Hybrid Branch Fusion Strategy for Code Size Reduction

“…Merging. Function merging is an important code size optimization, where recent work has shown significant reduction gains [26,28].…”

“…Loop rolling is another important code-size optimization that works by merging similar code into a loop [15,26]. The state-of-the-art technique, RoLAG [26], searches for isomorphic instructions within a single basic block and builds an alignment graph representing the groups of isomorphic instructions that can be rolled into a loop.…”

“…Loop rolling is another important code-size optimization that works by merging similar code into a loop [15,26]. The state-of-the-art technique, RoLAG [26], searches for isomorphic instructions within a single basic block and builds an alignment graph representing the groups of isomorphic instructions that can be rolled into a loop. While both loop rolling and branch fusion work within functions, they exploit different opportunities for merging similar code, since the former considers similarities within a basic block while branch fusion consider similarities across the regions that form an if-then-else construction.…”

“…Function outlining identifies equivalent basic blocks across all functions and extracts them into a function, replacing all the copies by a function call [4]. Loop rolling transforms equivalent instructions from a single block into a loop [27].…”