The constraints imply that the shape of the distribution must be trapezoidal.

You need to find the 2 key values on the y-axis to get the pdf, say k1 and k2.

Now you have 2 equations and 2 unknowns: 1 equation is provided by the fact that the total probability=1.

The 2nd is provided by constraint (c).

Solve simultaneously for k1 and k2.

Now the required probability is f(4) + f(5) where f is the pdf.