From: ts <decoux@...>
Date: 2001-12-06T20:34:05+09:00
Subject: [ruby-talk:27673] Re: Behaviour of Mutex

>>>>> "M" == Max Maischein <corion@informatik.uni-frankfurt.de> writes:

M> So, if anyone can enlighten me about the reason for the current Mutex
M> behaviour, or point me to an implementation that does What I Want<tm>, or
M> tell me that I should write it, I'm happy :-))

 You have try_lock

 Well you are in a special case, try it in a thread

pigeon% cat b.rb
#!/usr/bin/ruby
require 'thread'

# This is the problem
lock = Mutex.new

a = Thread.new do

   p "First synchronize"
   lock.synchronize {
      p "Second synchronize"
      lock.synchronize {
         p "Two times synchronized"
      }
      p "One time synchronized"
   }
   p "Done."
end

a.join
pigeon% 

pigeon% b.rb
"First synchronize"
"Second synchronize"
deadlock 0x401c0528: 2:8 (main) - ./b.rb:20
deadlock 0x401b6230: 2:0  - /usr/local/lib/ruby/1.6/thread.rb:60
/usr/local/lib/ruby/1.6/thread.rb:60: Thread: deadlock (fatal)
pigeon% 


 Look at pthread_mutex_lock (3) for the different kind of mutex (`fast' which
 give a deadlock, `error checking' which give an error, `recursive' which
 allow the same thread to call many times lock, ...)


Guy Decoux