Slow code due to many AbstractVector::check and AbstractShuffle::checkIndexes checks in C2

John Rose john.r.rose at
Fri May 22 22:36:05 UTC 2020

On May 22, 2020, at 3:25 PM, Paul Sandoz <paul.sandoz at> wrote:
> The checkIndexes of the shuffle.checkIndexes is likely to have a cost, it's not optimized right now. 

After it’s converted to vector operations, it should be hoistable
from loops.  “Should be” != “is”, of course.

More information about the panama-dev mailing list