You do not need to differentiate the potential energy and use Newton's second law.
You already have conservation of energy: ## \frac {mv^2} {2} + cx^2 = \frac {mv_0^2} {2} ##, thus ## v = \sqrt { v_0^2 - 2 \frac c m x^2} ##, or ## \frac {dx} {dt} = \sqrt { v_0^2 - 2 \frac c m x^2} ##.
Can you solve this?
If you insist on using Newton's second law, then I should say that this step of yours: ## \int 2cxdt = \int m dv \Rightarrow \frac {2cxt} m = v + \text{constant} ## is already wrong, and not really because of the missing minus sign, but because ##x## is a function of time, no a constant, so its integral is not ##xt##.