From 07d1eac91c7007954fed88332d495a42cd59afef Mon Sep 17 00:00:00 2001 From: Jacob Lifshay Date: Tue, 11 Oct 2022 17:13:09 -0700 Subject: [PATCH] put a reasonable reason to use DOUBLE() lfd using DOUBLE is a completely bogus reason -- lfd doesn't use DOUBLE because it doesn't need to convert from f32 to f64. --- openpower/sv/rfc/ls002/discussion.mdwn | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/openpower/sv/rfc/ls002/discussion.mdwn b/openpower/sv/rfc/ls002/discussion.mdwn index 679a32805..77705be5b 100644 --- a/openpower/sv/rfc/ls002/discussion.mdwn +++ b/openpower/sv/rfc/ls002/discussion.mdwn @@ -225,4 +225,6 @@ comparative analogy as a Programmer's Note. **Should the use of DOUBLE() be bypassed?** -No, because lfd uses it, to normalise the FP value +No, because we specifically want to be able to express all possible f32 values, +including denormal values. those denormal values require normalization to get +the corresponding f64 values. -- 2.30.2