Your postdoctoral years should be glorious. You’ve proven yourself with a Ph.D., showing you can master your field and find something new to solve, often in collaboration with an adviser, solve it, write it up, and publish it. You have learned the secret handshakes of academia, how to write grant proposals, how to present at meetings, how to meet new people, how to forge collaborations, and how to carve out an exciting niche for yourself.
Your reward is the lollipop stage of an academic career, the postdoc. Postdocs get to focus on research. They have left behind course requirements, dissertation margin rules, and the often capricious rules for a degree at their institution. They do not yet have to face the complex balance of a faculty position with teaching, research, and committee work, to name only the formal stuff.
This is not to say postdocs don’t face challenges. If your funding comes from your mentor, the project may be much more specific than anything you had before. If you have your own funding, with increased independence, challenges in deciding on appropriate methods and approaches will fall more on your shoulders.
You may not have written up all your previous work and need to carve out time to do that if you are to succeed at the next step, a fulfilling faculty position, or something else that uses your considerable skills. A good productivity yardstick is to publish at least a paper a year during your postdoc, mixing work from before, work you take the lead on, and work you are a middle author on. Always choose for publishing the highest level journal you can.
Whether the project is your own or one you joined, a crucial component is independence. For example, no one should pay attention to when you arrive at the lab in the morning, unless you are planning something like a joint pipetting marathon and scheduled it specifically. No one should tell you when you work on the new stuff and when you take time to finish earlier work. No one should look over your shoulder more often than mutually agreed on scheduled meetings, except for if you both enjoy informal contact.
So, this is what it should be like. It is a stage, no more than five years long, ideally, that precedes the next, more permanent step. If you come to hate the project, figure out what to do. Switch projects or identify aspects of it you like. Researchers thrive only with a love of the question, the organism, and the project. You may still be defining your best love in this period and no one will fault you if you decide what first looked glittery is in fact not for you. A considerate mentor will help you with this project.
If you do not get along with your adviser, you should start with working on the relationship. The commonest relationship failure is excessive oversight. Coax the adviser away by offering meetings based on productivity, not random checks for things like open doors. Resist being micro-managed, but offer up written summaries. The micromanaging adviser is probably dealing with consequences of this behavior in a lot of areas of life. One approach that might work may be to turn the adviser’s attention from you personally and your schedule to written accounts of research progress and struggles. Or this may not work. Identify what you can change about the relationship and your interpretation of it, and what you cannot change. If the part that you cannot change is making you miserable, leave.
Keep front and center the belief that the postdoctoral period should be wonderful for everyone. If it is not and you cannot fix it using the tips above, then it is time to move on. No one will hold it against you. Odds are others already know your adviser is difficult. Just remember your job is to flourish academically and intellectually. This is not a shop floor. There is no time clock. There is no boss. Don’t settle for misery!
Sociobiology 
I agree with all of the above except I’m not sure that excessive oversight is the most common problem. Other common problems include stress from long-distance relationships (this is personal, but spills over into work if you are often out of town etc.), too little oversight (eg because you started something new in your postdoc project and really don’t know as much about techniques as perhaps the advisor assumes), quarrels over project ownership (eg if the advisor wants to take your project further but you thought you could take it with you, and variations thereof). Not least, some people only figure out during a postdoc that academia is not for them, or that they are not as effective as they thought in some aspect; either way this will probably lead to strain with the advisor. What is the good/considerate mentor to do? Help you capitalize on your strengths or overcome your weaknesses in the latter case? Either way in my opinion it is best to be as open as possible with your advisor as early as possible. Even advisors who only have their own productivity in mind will try to help you get the most out of your postdoc project.