From: lkcl <lkcl@web>
Date: Sun, 17 Sep 2023 07:44:15 +0000 (+0100)
Subject: (no commit message)
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=1911d9d580084550ee4212210b724084b653a297;p=libreriscv.git

---

diff --git a/openpower/sv/cookbook/chacha20.mdwn b/openpower/sv/cookbook/chacha20.mdwn
index c70374740..0a164a504 100644
--- a/openpower/sv/cookbook/chacha20.mdwn
+++ b/openpower/sv/cookbook/chacha20.mdwn
@@ -153,10 +153,17 @@ Let's list the additions only:
     x3  = x3 + x4
     x9  = x9 + x14
 ```
+
+These are clearly regular patterns, and if there was a
+system of "register offsets" available, it would be possible
+to put a loop around one single add instead of unrolling 32
+operations. This is where REMAP Indexing steps in.
     
 ## Introduction to REMAP Indexing
 
 REMAP Indexing performs any arbitrary re-positioning of elements.
+In effect it makes it possible to perform "arrbitrary addressing"
+at runtime of registers.
 Where normally any other Vector Processor would only be able to do a
 sequential element-level series of operations, and if re-ordering
 of the elements is required use a special re-ordering instruction,