Synchronization article on developerWorks

[Tom Tromey]

>>>>> "Bryce" == Bryce McKinlay <bryce@waitaki.otago.ac.nz> writes:

Bryce> Here's an interesting article about the costs of syncronization. 
Bryce> Especially interesting because GCJ gets mentioned in the performance 
Bryce> tests:
Bryce> http://www-106.ibm.com/developerworks/java/library/j-threads1.html

They use gcj 3.0, which, while a reasonable choice, also will show
more overhead than 3.1.

If you follow the `discussion' link you'll see a comment from our own
AG and a response from the paper's author, both about the gcj results.

Also, they mention something that I've been meaning to address for a
while: double-check locking.  We use this in libgcj, but everything
I've read says it is inherently broken.  We should probably go through
libgcj and remove the instances of it.

Tom