A hardware mechanism for dynamic extraction and relayout of program hot spots

Abstract

This paper presents a new mechanism for collecting and deploying runtime optimized code. The code-collecting component resides in the instruction retirement stage and lays out hot execution paths to improve instruction fetch rate as well as enable further code optimization. The code deployment component uses an extension to the Branch Target Buffer to migrate execution into the new code without modifying the original code. No significant delay is added to the total execution of the program due to these components. The code collection scheme enables safe runtime optimization along paths that span function boundaries. This technique provides a better platform for runtime optimization than trace caches, because the traces are longer and persist in main memory across context switches. Additionally, these traces are not as susceptible to transient behavior because they are restricted to frequently executed code. Empirical results show that on average this mechanism can achieve better instruction fetch rates using only 12KB of hardware than a trace cache requiring 15KB of hardware, while producing long, persistent traces more suited to optimization.

Selecting long atomic traces for high coverage

“…In particular, trace construction may be enhanced with optimization functionality that can increase performance [10][14] [25][7] [36] and reduce power consumption [35] [30]. This optimization potential can be significantly facilitated by (a) selecting longer traces, and (b) by treating each trace as a single atomic block although it may contain control transfer instructions [22] [26].…”

Probability Analysis to Improve the Confidence in Profiling Accuracy

“…e task granularity profiling is useful in code profiling and hot execution path detection [28]. Identifying program hot spots can support runtime optimization [29,30].…”

Selecting long atomic traces for high coverage

“…As far as we know, the global schemes considered here have not been considered elsewhere for trace selection. Analogous concepts exist for the selection of hot paths that can be stored in memory [22].…”

“…In particular, trace construction may be enhanced with optimization functionality that can increase performance [10][14] [25][7] [36] and reduce power consumption [35] [30]. This optimization potential can be significantly facilitated by (a) selecting longer traces, and (b) by treating each trace as a single atomic block although it may contain control transfer instructions [22] [26].…”