Representable Correcting Terms for Possibly Underflowing Floating Point Operations

Authors:
Sylvie Boldo;Marc Daumas
Affiliations:
-;-
Venue:
ARITH '03 Proceedings of the 16th IEEE Symposium on Computer Arithmetic (ARITH-16'03)
Year:
2003

Citing 0
Cited 5

Semantics of roundoff error propagation in finite precision calculations

Higher-Order and Symbolic Computation
Improved bound for stochastic formal correctness of numerical algorithms

Innovations in Systems and Software Engineering
An overview of semantics for the validation of numerical programs

VMCAI'05 Proceedings of the 6th international conference on Verification, Model Checking, and Abstract Interpretation
Pitfalls of a full floating-point proof: example on the formal proof of the veltkamp/dekker algorithms

IJCAR'06 Proceedings of the Third international joint conference on Automated Reasoning
Semantics-based transformation of arithmetic expressions

SAS'07 Proceedings of the 14th international conference on Static Analysis

Quantified Score

Hi-index	0.00

Visualization

Abstract

Studying floating point arithmetic, authors have shown that the implemented operations (addition, subtraction, multiplication, division and square root) can compute a result and an exact correcting term using the same format as the inputs. Following a path initiated in 1965, many authors supposed that neither underflow nor overflow occurred in the process. Overflow is not critical as this kind of exception creates persisting non numeric quantities. Underflow may be fatal to the process as it returns wrong numeric values with little warning. Our new conditions guarantee that the correcting term is exact when the result is a number. We have validated our proofs against Coq automatic proof checker. Our development has raised manyquestions, some of them were expected while other ones were surprising.