1: \begin{abstract}

2:   In this paper we prove a theorem that provides an upper bound for

3:   the density of packings of congruent copies of a given convex body

4:   in~$\R^n$; this theorem is a generalization of the linear

5:   programming bound for sphere packings. We illustrate its use by

6:   computing an upper bound for the maximum density of packings of

7:   regular pentagons in the plane. Our computational approach is

8:   numerical and uses a combination of semidefinite programming, sums

9:   of squares, and the harmonic analysis of the Euclidean motion group.

10:   We show how, with some extra work, the bounds so obtained can be

11:   made rigorous.

12: \end{abstract}

13: