From 592005a6b721c34515023eca0c4c1aa847947e42 Mon Sep 17 00:00:00 2001 From: lkcl Date: Sat, 20 Aug 2022 14:00:21 +0100 Subject: [PATCH] --- openpower/sv/normal.mdwn | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/openpower/sv/normal.mdwn b/openpower/sv/normal.mdwn index c674542c3..3bfc3e8c1 100644 --- a/openpower/sv/normal.mdwn +++ b/openpower/sv/normal.mdwn @@ -199,6 +199,22 @@ truncated based explicitly on whether a test fails. This because it is a precise test on which algorithms will rely. +**Floating-point Exceptions** + +When Floating-point exceptions are enabled VL must be truncated at +the point where the Exception appears not to have occurred. If `VLi` +is set then VL must include the faulting element. Although very strongly +discouraged the Exception Mode that permits Floating Point Exception +notification to arrive too late to unwind is permitted +(under protest, due it violating +the otherwise 100% Deterministic nature of Data-dependent Fail-first).the + +**Use of lax FP Exception Notification Mode could result in parallel +computations proceeding with invalid results that have to be explicitly +detected, whereas with the strict FP Execption Mode enabled, FFirst +truncates VL, allows subsequent parallel computation to avoid +the exceptions entirely** + ## Data-dependent fail-first on CR operations (crand etc) Operations that actually produce or alter CR Field as a result -- 2.30.2