I have seen more than 30% performance regression for ShortText/LongText
tests in swingMark benchmark.

And it seems the regression was introduced in the last few builds. There was
no such regression in earlier builds like b60-b65.

(compared to jdk 1.3.0).

To reproduce:

% uname -a
SunOS vispro 5.7 Generic_106541-16 sun4u sparc SUNW,Ultra-30

% jdk1.4/bin/java -version
java version "1.4.0-beta_refresh"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.0-beta_refresh-b70)
Java HotSpot(TM) Client VM (build 1.4.0-beta_refresh-b70, mixed mode)


% jdk1.4/bin/java -client -ms256m -mx256m -cp /home/zzhang/bugs perfteam.swingmark.SwingMark -r 5 -q -testfile HTMLTests.txt

ShortHTMLTest 35530 28251 28779 28788 28942
ShortTextTest 30280 29888 31072 30943 30960
LongHTMLTest 325511 321755 322860 319999 320525
LongTextTest 232207 235740 236734 237652 236196

% jdk1.3/bin/java -version
java version "1.3.0"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.0)
Java HotSpot(TM) Client VM (build 1.3.0, mixed mode)

% jdk1.3/bin/java -client -ms256m -mx256m -cp /home/zzhang/bugs perfteam.swingmark.SwingMark -r 5 -q -testfile HTMLTests.txt

ShortHTMLTest 41749 34895 35501 35638 36230
ShortTextTest 21857 21444 22034 22414 23052
LongHTMLTest 330470 324750 324671 324456 324871
LongTextTest 148307 148344 148835 149131 149500

These are numbers over 5 runs. You would see the regressions for
ShortTextTest and LongTextTest above.