You can't use fusion in a car. So fusion is only half of an energy solution: fusion solves the problem of energy production. This leaves another problem unsolved: The problem of storage and transport.
In other words, if you want to use fusion power as a total energy solution, then you have to find some kind of a generalized way of taking energy and storing it for use in a car or somesuch. There are several candidate technologies that could provide this energy transport layer: say electricity stored in batteries, or hydrogen. (Hydrogen is kind of the purest example of an energy transport technology.) However none of these technologies are "ready for prime time".
This is not really a downside of fusion: It is my opinion we have to solve this problem and create an energy transport layer of this sort no matter WHAT our energy production solution. And the many years it may take to get hydrogen etc working doesn't seem like much of a concern considering it will take probably many more years to get fusion working in the first place.
Incidentally there *are* technologies which are capable of "making" oil. For example biofuels are basically an example of this, they are a roundabout way of turning solar energy into oil. The downside of current biofuel technologies is that you have to do all that messy "agriculture" stuff to get them, which means they have high energy and environmental costs; but I guess if we had a perfect source of energy production, like fusion, we wouldn't really care how energy-expensive biofuels were to produce, would we?
There are also interesting speculative technologies, such as thermal depolymerization, which basically "cooks" organic matter (i.e., trash) into oil in a way that mimicks geologic processes (as PhillipH describes) but is much faster. However this is very hard! None of these technologies seem to be going anywhere that I can see, and I find it much less likely that we'll get these working than we will some of the other potential solutions. http://www.theoildrum.com/node/2450 .