The goal of profiling is to find out the bottlenecks of the program, such as which function consumes the most time
, which function is called most times etc. in order to improve application performance
by "fixing" those hot spots.
Metering by Mark Kantrowitz .
SLIME provides an interface to the profilers of supported Lisps with its slime-profile-* functions.
Metering has been adapted for the current times by Daniel Kochmański. It's portable and very capable.
Project home and repository.
Works on all active implementations.
SBCL has two: sb-ext:profile which instruments functions, and the statistical sampling sb-sprof contrib (based on CMUCL work).
(sb-sprof:with-profiling (:report :graph)