Abstract
Compile-time code transformations which expose instruction-level parallelism (ILP) typically take into account the constraints imposed by all execution scenarios in the program. However, there are additional opportunities to increase ILP along some execution sequences if the constraints from alternative execution sequences can be ignored. Traditionally, profile information has been used to identify important execution sequences for aggressive compiler optimization and scheduling. This paper presents a set of static program analysis heuristics used in the IMPACT compiler to identify execution sequences for aggressive optimization. We show that the static program analysis heuristics identify execution sequences without hazardous conditions that tend to prohibit compiler optimizations. As a result, the static program analysis approach often achieves optimization results comparable to profile information in spite of its inferior branch prediction accuracies. This observation makes a strong case for using static program analysis with or without profile information to facilitate aggressive compiler optimization and scheduling.
| Original language | English (US) |
|---|---|
| Title of host publication | Proceedings of the Annual International Symposium on Microarchitecture |
| Editors | Anon |
| Publisher | Publ by IEEE |
| Pages | 247-255 |
| Number of pages | 9 |
| ISBN (Print) | 0818652802 |
| State | Published - Jan 1 1994 |
| Event | Proceedings of the 26th Annual International Symposium on Microarchitecture - Austin, TX, USA Duration: Dec 1 1993 → Dec 3 1993 |
Publication series
| Name | Proceedings of the Annual International Symposium on Microarchitecture |
|---|---|
| ISSN (Print) | 1072-4451 |
Other
| Other | Proceedings of the 26th Annual International Symposium on Microarchitecture |
|---|---|
| City | Austin, TX, USA |
| Period | 12/1/93 → 12/3/93 |
Fingerprint
ASJC Scopus subject areas
- Hardware and Architecture
- Software
Cite this
Superblock formation using static program analysis. / Hank, Richard E.; Mahlke, Scott A.; Bringmann, Roger A.; Gyllenhaal, John C.; Hwu, Wen-Mei W.
Proceedings of the Annual International Symposium on Microarchitecture. ed. / Anon. Publ by IEEE, 1994. p. 247-255 (Proceedings of the Annual International Symposium on Microarchitecture).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Superblock formation using static program analysis
AU - Hank, Richard E.
AU - Mahlke, Scott A.
AU - Bringmann, Roger A.
AU - Gyllenhaal, John C.
AU - Hwu, Wen-Mei W
PY - 1994/1/1
Y1 - 1994/1/1
N2 - Compile-time code transformations which expose instruction-level parallelism (ILP) typically take into account the constraints imposed by all execution scenarios in the program. However, there are additional opportunities to increase ILP along some execution sequences if the constraints from alternative execution sequences can be ignored. Traditionally, profile information has been used to identify important execution sequences for aggressive compiler optimization and scheduling. This paper presents a set of static program analysis heuristics used in the IMPACT compiler to identify execution sequences for aggressive optimization. We show that the static program analysis heuristics identify execution sequences without hazardous conditions that tend to prohibit compiler optimizations. As a result, the static program analysis approach often achieves optimization results comparable to profile information in spite of its inferior branch prediction accuracies. This observation makes a strong case for using static program analysis with or without profile information to facilitate aggressive compiler optimization and scheduling.
AB - Compile-time code transformations which expose instruction-level parallelism (ILP) typically take into account the constraints imposed by all execution scenarios in the program. However, there are additional opportunities to increase ILP along some execution sequences if the constraints from alternative execution sequences can be ignored. Traditionally, profile information has been used to identify important execution sequences for aggressive compiler optimization and scheduling. This paper presents a set of static program analysis heuristics used in the IMPACT compiler to identify execution sequences for aggressive optimization. We show that the static program analysis heuristics identify execution sequences without hazardous conditions that tend to prohibit compiler optimizations. As a result, the static program analysis approach often achieves optimization results comparable to profile information in spite of its inferior branch prediction accuracies. This observation makes a strong case for using static program analysis with or without profile information to facilitate aggressive compiler optimization and scheduling.
UR - http://www.scopus.com/inward/record.url?scp=0028015212&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0028015212&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:0028015212
SN - 0818652802
T3 - Proceedings of the Annual International Symposium on Microarchitecture
SP - 247
EP - 255
BT - Proceedings of the Annual International Symposium on Microarchitecture
A2 - Anon, null
PB - Publ by IEEE
ER -