Bug in applyYStencil?

[dschwoerer]

    fval.c = block->data[bx.jx][bx.jy][bx.jz];
    fval.m = block->data[bx.jx][bx.jym][bx.jz];
    fval.m2 = block->data[bx.jx][bx.jy2m][bx.jz];
    fval.m3 = block->data[bx.jx][bx.jy+3][bx.jz];
    fval.m4 = block->data[bx.jx][bx.jy+4][bx.jz];

seems suspicious to me.
Is this a bug?
If so, #154 should fix it.

[bendudson]

This certainly looks like a bug in m3 and m4. It would affect high-order one-sided boundary conditions.