Know your single integration techniques. U-sub, by parts, trig sub, partial fractions, even symmetry arguments. Remember your vector calculus also, like both forms of the line integral, both forms of the surface integral, and the big theorems, Green's, Stoke's, Divergence.
Remember all of your diffeq as well. Separation of variables, exact, slope fields, homogeneous, nonhomogeneous.
For linear algebra, you should know your linear transformation stuff, but the exam likes to focus on eigenstuff. I would recommend Axler's chapter on eigenstuff.
The abstract algebra you need to know is group theory and ring theory. Should really focus on studying ring theory.
There will be questions on probability and combinatorics(including possibly graph theory). You should learn a bit of Python, as there will be a question on reading pseudocode and stating what it does.
Real analysis on the exam is usually some complicated epsilon-delta stuff. I was never good at it.
Topology is point-set, should know all of the definitions and basic theorems, i.e., what you usually cover in Munkres.
There's probably more topics that I missed, but I don't remember right now.
Complex analysis will probably be something on complex integration, residue integration, or Cauchy-Riemann.
I got a 790(78th percentile) on the math GRE. I fucked up one easy computational question which would have given me 800(80th percentile). I don't think my score would have mattered though.