-findirect-dispatch performance

[Andrew Haley <aph at redhat dot com>]

Now that the gc bitmap generation bug is fixed, I've been measuring
the performance difference with -findirect-dispatch.

Bottom line: it seems to me that, as far as I can measure, there is no
perforance difference.

My machine has many daemons, so it's hard to be perfectly
accurate, but here is an embedded CaffeineMark run:

With -findirect-dispatch:

cuddles:~/tmpzz/copy $ ./CaffeineMarkEmbeddedApp
Sieve score = 13919 (98)
Loop score = 58683 (2017)
Logic score = 149024 (0)
String score = 12227 (708)
Float score = 17880 (185)
Method score = 18890 (166650)
Overall score = 28197

Without:

cuddles:~/tmpzz/copy $ ./CaffeineMarkEmbeddedApp
Sieve score = 14462 (98)
Loop score = 61232 (2017)
Logic score = 153833 (0)
String score = 11679 (708)
Float score = 17474 (185)
Method score = 18479 (166650)
Overall score = 28300

The difference between these figures is less than the measurement
noise.  So, I'd have to put my machine into single user mode to
determine which was fastest.

Now, the embedded CaffeineMark is not the best Java benchmark program,
so if anyone can suggest anything better I'm interested.  However, on
the basis of these timings I think we can agree that this is good
enough.

Andrew.