There are some libraries to interface lisp with SDL, and almost all lisp implementations have methods to interface with C code (though you might have to write some wrappers yourself.)
Lisp can be as fast (or, in most cases a bit slower) than C/C++ code when fully optimized, but that is a bit of work. It's a lot easier to write an unoptimized piece of Lisp code and get it to work, and then later go in and optimize as much as needed. This is because you don't declare type specifiers and such in Lisp. It's EXTREMELY easy to write generic code in lisp, and then later optimize it only if necessary.
Granted, there aren't a ton of game oriented libraries ready to go for lisp, and that would be it's biggest drawback. That and, for most simple casual games using C or C++ would probably be faster to develop in the end because of the access to common libraries, and since C/C++ are much more like Blitzmax than Lisp is (Lisp is a whole 'nother world.) If you were writing a strategy or simulation or some game with a lot of AI or some experimental algorithms and such, Lisp would be a great bet.
If you just want to broaden your programming horizons, and forever change your perspective of what is possible with a language, I'd recommend spending a couple months and getting a grasp of what makes Lisp Lisp. It's NOT the perfect tool for every job (although it could be if it was as popular as Python, considering it is many many times faster than Python and Lua and is capable of being used as a scripting language just as they are, except with easier integration to the rest of the code (no glue code).
But, given that it isn't that popular and a lot of libraries haven't been wrapped, I would only recommend Lisp under certain cirucmstances for game development.
One nice thing is, almost all Lisps are multi-platform. One drawback is, the nice Lisp implementations are a commercial product and can set you back over $1000.
Lisp can be as fast (or, in most cases a bit slower) than C/C++ code when fully optimized, but that is a bit of work. It's a lot easier to write an unoptimized piece of Lisp code and get it to work, and then later go in and optimize as much as needed. This is because you don't declare type specifiers and such in Lisp. It's EXTREMELY easy to write generic code in lisp, and then later optimize it only if necessary.
Granted, there aren't a ton of game oriented libraries ready to go for lisp, and that would be it's biggest drawback. That and, for most simple casual games using C or C++ would probably be faster to develop in the end because of the access to common libraries, and since C/C++ are much more like Blitzmax than Lisp is (Lisp is a whole 'nother world.) If you were writing a strategy or simulation or some game with a lot of AI or some experimental algorithms and such, Lisp would be a great bet.
If you just want to broaden your programming horizons, and forever change your perspective of what is possible with a language, I'd recommend spending a couple months and getting a grasp of what makes Lisp Lisp. It's NOT the perfect tool for every job (although it could be if it was as popular as Python, considering it is many many times faster than Python and Lua and is capable of being used as a scripting language just as they are, except with easier integration to the rest of the code (no glue code).
But, given that it isn't that popular and a lot of libraries haven't been wrapped, I would only recommend Lisp under certain cirucmstances for game development.
One nice thing is, almost all Lisps are multi-platform. One drawback is, the nice Lisp implementations are a commercial product and can set you back over $1000.