Issue #3924 has been updated by Xavier Shay.


Progress update:

Fixed the autoload problem, my fork now loads a moderate sized rails app in the same time as ruby 1.9.2 (and far faster that 1.9.3dev 20s compared to 46s), as well as passing all other tests/specs. This proves the algorithm I think, the next step is to optimize the inner functions (such as rb_locate_file_with_extensions), since they still use plenty of rb_funcall to do string manipulation which could be far faster done in C. There is also room for improvement by caching rb_locate_file.

Code is still at https://github.com/xaviershay/ruby/tree/require-performance-fix

I hope to have more time to work on it on Sunday.
----------------------------------------
Bug #3924: Performance bug (in require?)
http://redmine.ruby-lang.org/issues/3924

Author: Carsten Bormann
Status: Open
Priority: Normal
Assignee: 
Category: core
Target version: 1.9.3
ruby -v: ruby 1.9.2p0 (2010-08-18 revision 29036) [x86_64-darwin10] 


=begin
 Running irb < /dev/null in 1.9.2 causes 3016 calls to lstat64.
 
 For instance, there is a sequence of 28 repetitions each of lstat calls to all 6 non-empty path prefixes of /opt/local/lib/ruby1.9/1.9.1/irb.rb -- a total of 170 lstats apparently just to load this file; another set of lstats then occurs later for another 18 (times 6) times.  Clearly, something is running amok in the calling sequence rb_require_safe -> realpath_rec -> lstat.
 
 Another example: Running a simple test with the baretest gem causes 17008 calls to lstat.  According to perftools.rb, 80 % of the 1.2 seconds of CPU is used in Kernel#gem_original_require (and another 12 in GC, some of which may be caused by this).
=end



-- 
http://redmine.ruby-lang.org