[Second Edition, C++ V2.10]
'slvsm2' solves nonlinear equation 19.6.44 at the coarsest level:
the discrete version (19.6.45) is written as:
L(u(i,j))=[u(i+1,j)+u(i-1,j)+u(i,j+1)+u(i,j-1)-4u(i,j)]/h^2+u(i,j)^2-rho(i,j)
for the coarsest level (3x3) all gird points except (possibly) u(1,1) are zero and we seek the solution of:
u(1,1)^2-4/h^2*u(1,1)-rho(1,1)=0,
the solution is (we drop the index) :
u=2/h^2+-sqrt((2/h^2)^2+rho)

slvsm2 however solves
u=-rho/(2/h^2+sqrt((2/h^2)^2+rho))
can the authors kindly confirm? or is there something that I am missing?
thanks

Please disregard my question. I just realized that only one of the solutions to the quadratic equation results in the correct solution. I am assuming that the authors eliminated the second solution by verifying the results.