To solve a sparse symmetric indeﬁnite system of linear equations. Given a sparse symmetric matrix and an -vector , this subroutine solves the system .
The method used is a direct method using an factorization, where is unit lower triangular and is block diagonal with blocks of order 1 and 2. Advantage is taken of the extra sparsity available with pivots (blocks of ) with one or both diagonal entries of value zero. The numerical values of the entries are taken in account during the ﬁrst choice of pivots.