The goal is to arrive below a raise 2 degrees Celsius, before 2030. According to the UN, that is the cut-off date to avoid explosion of the world temperature due to positive feedback mechanisms within the world's weather system. The technologies required to achieve this have been identified. It is technically feasible.
The Swedish energy company Vatterfall published their engineering report on climate change. In it, they present this illustrated list of the numerous measures available that reduce CO2 emissions, ordered by cost. On the left there are the measured which save the most money, such as insulation, and to the left those that cost the most money, such as Biodiesel.
The horizontal axis is the cumulative tonnage of emitted CO2 that would be saved, if we applied all the technologies leftward of that point of the axis, given how much opportunities for deployment there are for each of them.
We can achieve our objective by selecting all the technologies, the money saving ones, the cheap ones, the not-so-cheap ones, all the way up to those costing 40 euros per ton. The average cost amounts to 2 euro per ton of CO2. Expensive, sure, but also feasible, economically.
Remains the question of whether it is feasible, politically.
On that question, François offered I read the book Collapse: How Societies Choose to Fail or Succeed, by Jarred Diamond. It is available at the Grande Bibliothèque Nationale, and it will be the next book I read (thanks François).
Update, March 03, 2010:
The study I mentionned above has gone offline. Here is another one with similar conclusions:
Update, Nov 25, 2014: This resource has also gone offline. Here yet another source.