LLVM codgen improvement, count bits intrinsics
Johan Engelen
j at j.nl
Sun Apr 28 11:22:57 UTC 2019
On Saturday, 27 April 2019 at 20:25:01 UTC, NaN wrote:
> Where do you sugest to LLVM people that codegem could be
> improved?
On their mailinglist or in their bug tracker.
> The bit scan forward and reverse both test for zero and do
> jumps (when you want zero defined), when they could be doing
> conditional moves because both instructions st the zero flag if
> the input is zero.
Two remarks:
1. Conditional move is not necessarily faster than branching
2. On recent CPUs `tzcnt` is the better instruction that has
defined output for input 0
-Johan
More information about the digitalmars-d-ldc
mailing list