[PATCH] D50004: [PowerPC] Emit xscpsgndp instead of xxlor when copying floating point scalar registers

Kewen Lin via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Wed Aug 1 23:17:27 PDT 2018


jedilyn added a comment.

In https://reviews.llvm.org/D50004#1182972, @inouehrs wrote:

> XSCPSGNDP has longer latency (6 cycles) than XXLOR (2 cycles) on POWER8 while it has higher throughput with the same latency on POWER9. So XXLOR is preferable for pre-P9.
>
> Also, the two instructions have different behavior for a denormal input value in my understanding; XSCPSGNDP does normalization but XXLOR does not. Does this difference matter?


+1, even for Power9 XSCPSGNDP makes pipeline busy longer than XXLOR, XXLOR is still better.


https://reviews.llvm.org/D50004





More information about the llvm-commits mailing list