Overall Objectives
Research Program
Application Domains
Highlights of the Year
New Software and Platforms
New Results
Bilateral Contracts and Grants with Industry
Partnerships and Cooperations
XML PDF e-pub
PDF e-Pub

Section: New Results

Studying Optimal Spilling in the Light of SSA

Participants : Florian Brandner [ENSTA ParisTech, previously Compsys] , Quentin Colombet [Apple, previously Compsys] , Alain Darte.

Recent developments in register allocation, mostly linked to static single assignment (SSA) form, have shown the benefits of decoupling the problem in two phases: a first spilling phase places load and store instructions so that the register pressure at all program points is small enough, and a second assignment and coalescing phase maps the variables to physical registers and reduces the number of move instructions among registers. We focused on the first phase, for which many open questions remain: in particular, we studied the notion of optimal spilling (what can be expressed?) and the impact of SSA form (does it help?).

To identify the important features for optimal spilling on load-store architectures, we developed a new integer linear programming formulation, more accurate and expressive than previous approaches. Among other features, we can express SSA φ-functions, memory-to-memory copies, and the fact that a value can be stored simultaneously in a register and in memory. Based on this formulation, we presented a thorough analysis of the results obtained for the SPECINT 2000 and EEMBC 1.1 benchmarks, from which we have drawn, among others, the following conclusions: (1) rematerialization is extremely important; (2) SSA complicates the formulation of optimal spilling, especially because of memory coalescing when the code is not in conventional SSA (CSSA); (3) micro-architectural features are significant and thus have to be accounted for; and (4) significant savings can be obtained in terms of static spill costs, cache miss rates, and dynamic instruction counts.

Parts of this work were published at CASES 2011 [18] . The journal publication [1] contains more detailed discussions, more examples illustrating new concepts and existing approaches, and additional experiments covering the observed worst-case behavior, a new post-latency heuristic, and empiric evidence showing why static spill costs are a poor metric. Three configurations were added: Appel and George under SSA, Koes and Goldstein, and the heuristic of Braun and Hack.