Lean programs, branch mispredictions, and sorting

TY - GEN

T1 - Lean programs, branch mispredictions, and sorting

AU - Elmasry, Amr Ahmed Abd Elmoneim

AU - Katajainen, Jyrki

N1 - Conference code: 6

PY - 2012

Y1 - 2012

N2 - According to a folk theorem, every program can be transformed into a program that produces the same output and only has one loop. We generalize this to a form where the resulting program has one loop and no other branches than the one associated with the loop control. For this branch, branch prediction is easy even for a static branch predictor. If the original program is of length κ, measured in the number of assembly-language instructions, and runs in t(n) time for an input of size n, the transformed program is of length O(κ) and runs in O(κt(n)) time. Normally sorting programs are short, but still κ may be too large for practical purposes. Therefore, we provide more efficient hand-tailored heapsort and mergesort programs. Our programs retain most features of the original programs-e.g. they perform the same number of element comparisons-and they induce O(1) branch mispredictions. On computers where branch mispredictions were expensive, some of our programs were, for integer data and small instances, faster than the counterparts in the GNU implementation of the C++ standard library.

AB - According to a folk theorem, every program can be transformed into a program that produces the same output and only has one loop. We generalize this to a form where the resulting program has one loop and no other branches than the one associated with the loop control. For this branch, branch prediction is easy even for a static branch predictor. If the original program is of length κ, measured in the number of assembly-language instructions, and runs in t(n) time for an input of size n, the transformed program is of length O(κ) and runs in O(κt(n)) time. Normally sorting programs are short, but still κ may be too large for practical purposes. Therefore, we provide more efficient hand-tailored heapsort and mergesort programs. Our programs retain most features of the original programs-e.g. they perform the same number of element comparisons-and they induce O(1) branch mispredictions. On computers where branch mispredictions were expensive, some of our programs were, for integer data and small instances, faster than the counterparts in the GNU implementation of the C++ standard library.

U2 - 10.1007/978-3-642-30347-0_14

DO - 10.1007/978-3-642-30347-0_14

M3 - Article in proceedings

SN - 978-3-642-30346-3

T3 - Lecture notes in computer science

SP - 119

EP - 130

BT - Fun with Algorithms

A2 - Kranakis, Evangelos

A2 - Krizanc, Danny

A2 - Luccio, Flaminia

PB - Springer

T2 - The Sixth International conference on Fun with Algorithms

Y2 - 4 June 2012 through 6 June 2012

ER -