From: lkcl <lkcl@web>
Date: Sat, 15 Apr 2023 15:13:00 +0000 (+0100)
Subject: (no commit message)
X-Git-Tag: opf_rfc_ls009_v1~72
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=9372559bbb750ac460d8cd143ed84f17acd76697;p=libreriscv.git

---

diff --git a/openpower/sv/rfc/ls009.mdwn b/openpower/sv/rfc/ls009.mdwn
index 47427a919..64fe2aefb 100644
--- a/openpower/sv/rfc/ls009.mdwn
+++ b/openpower/sv/rfc/ls009.mdwn
@@ -1576,7 +1576,7 @@ offset is put through each algorithm to determine the actual Element Offset.
 Alternative implementations producing different ordering
 is prohibited as software will be critically relying on these Deterministic Schedules.
 
-## REMAP 2D/3D Matrix pseudocode
+### REMAP 2D/3D Matrix
 
 The following stand-alone executable source code is the Canonical
 Specification for Matrix (2D/3D) REMAP. 
@@ -1674,7 +1674,7 @@ if __name__ == '__main__':
     demo()
 ```
 
-## REMAP Parallel Reduction pseudocode
+### REMAP Parallel Reduction pseudocode
 
 The python3 program below is stand-alone executable and is the Canonical Specification
 for Parallel Reduction REMAP.
@@ -1765,7 +1765,7 @@ if __name__ == '__main__':
     demo()
 ```
 
-## REMAP FFT pseudocode
+### REMAP FFT pseudocode
 
 The FFT REMAP is RADIX2 only.
 
@@ -1937,24 +1937,20 @@ if __name__ == '__main__':
     demo()
 ```
 
-## DCT REMAP pseudocode
+### DCT REMAP
 
-The DCT REMAP is RADIX2 only.  Convolutions may be applied as usual
+DCT REMAP is RADIX2 only.  Convolutions may be applied as usual
 to create non-RADIX2 DCT. Combined with appropriate Twin-butterfly
-instructions the algorithm below, written in python3, becomes part
-of an in-place in-registers Vectorised DCT.
+instructions the algorithm below (written in python3), becomes part
+of an in-place in-registers Vectorised DCT.  The algorithms work
+by loading data such that as the nested loops progress the result
+is sorted into correct sequential order.
 
 ```
 # DCT "REMAP" scheduler to create an in-place iterative DCT.
 #
-# Original fastdctlee.py by Nayuki:
-# Copyright (c) 2020 Project Nayuki. (MIT License)
-# https://www.nayuki.io/page/fast-discrete-cosine-transform-algorithms
 
-from copy import deepcopy
-import math
-
-# bits of the integer 'val'.
+# bits of the integer 'val' of width 'width' are reversed
 def reverse_bits(val, width):
     result = 0
     for _ in range(width):
@@ -1963,7 +1959,7 @@ def reverse_bits(val, width):
     return result
 
 
-# iterative version of [recursively-applied] half-rev.
+# iterative version of [recursively-applied] half-reversing
 # turns out this is Gray-Encoding.
 def halfrev2(vec, pre_rev=True):
     res = []