review (S) for 6885584: A particular class structure causes large allocation spike for jit (original) (raw)
Vladimir Kozlov Vladimir.Kozlov at Sun.COM
Tue Oct 6 17:14:51 PDT 2009
- Previous message: review (S) for 6885584: A particular class structure causes large allocation spike for jit
- Next message: review (S) for 6885584: A particular class structure causes large allocation spike for jit
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Tom,
For longs it is possible that max < max_juint. So can we change the next condition:
if (_hi < max_juint)
return make(_lo, max_juint, WidenMax);
else
return make(_lo, max, WidenMax);
to next?:
if (_hi < max_juint && max_juint < max)
return make(_lo, max_juint, WidenMax);
else
return make(_lo, max, WidenMax);
Vladimir
Tom Rodriguez wrote:
On Oct 6, 2009, at 4:42 PM, Vladimir Kozlov wrote:
Tom, why changes are different for Int and Long? For longs you replaced maxjlong, minjlong with max,min I did them separately and didn't notice that they were different. I've fixed them to be the same. tom in make() parameters and conditions.
Vladimir Tom Rodriguez wrote: http://cr.openjdk.java.net/~never/6885584
- Previous message: review (S) for 6885584: A particular class structure causes large allocation spike for jit
- Next message: review (S) for 6885584: A particular class structure causes large allocation spike for jit
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the hotspot-compiler-dev mailing list