Any planing for comparison related performance enhancements ?

[Ulf Zibis]

Hi,

there are 2 RFE's, which would boost the performance of commonly 
comparison use cases:

Bug Id:6932837  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6932837>  - Better use unsigned jump if one of the range limits is 0
Bug Id:6933327  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6933327>  - Use shifted addressing modes instead of shift instructions


Is there some planing, if and when those RFE's would be implemented in 
HotSpot compiler?
There are some character related performance issues, which significantly 
depend on those GIT compiler capabilities.

@Martin: Would you like to publish your webrevs here?

Some other outstanding performance boosts, commonly used in character 
related routines:

Bug Id:6933324  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6933324>  - Always inline methods, which have only 1 call site
Bug Id:6932852  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6932852>  - HotSpot could do much better with do...while loop
Bug Id:6932855  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6932855>  - Save superfluous CMP instruction from while loop
Bug Id:6914113  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6914113>  - Copy int to byte[] in 1 step
Bug Id:6914095  <http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6914095>  - HotSpot should reuse invariant loop parameter


-Ulf

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/attachments/20100629/71dbab51/attachment.html