It is only recently that I've been able to learn math from a textbook without enormous mental strain. Even a year ago, I don't think I could. And the difference isn't that I've gotten smarter in the past year - it's that I've had to read and understand such difficult material that, through sheer desperation, I learned how.
The key (for me) is to write down what the book says.
Seriously. If the book has a definition, write it. Axioms, copy them out. Theorems, reproduce them on your paper word for word. If there is a proof, write it out, trying to understand each line. (If you don't understand it, just copy it down anyway.)
This seems like it wouldn't work. If you don't understand something when you read it, how can copying it help? I think it's very simple - writing is slow, so it forces you to read the words over and over at a very slow rate.
What about the parts you still don't understand? Again, for me, writing is key. Write down your questions, confusions, and speculation. You may very well answer your questions as you write.
This won't help with math that is too advanced for you to possibly understand, but it's awesome for math that you would understand if someone explained it to you. Of course, it requires work, so no wonder it took me 34 years to figure out. (Yes, I never, ever took notes from a book until recently.)