From: lkcl <lkcl@web>
Date: Sat, 4 Jun 2022 17:38:21 +0000 (+0100)
Subject: (no commit message)
X-Git-Tag: opf_rfc_ls005_v1~1978
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=235c83e250cc8854c5d23a2d17235152386dd5c4;p=libreriscv.git

---

diff --git a/openpower/sv/shape_table_format.mdwn b/openpower/sv/shape_table_format.mdwn
index b02e670b5..04374e61c 100644
--- a/openpower/sv/shape_table_format.mdwn
+++ b/openpower/sv/shape_table_format.mdwn
@@ -5,13 +5,15 @@ disabled: the register's elements are a linear (1D) vector.
 | ------ | ------ | ------ | ------ | ------- | ------- | ------- | ------- |
 | 0b00   | skip   | offset | invxyz | permute | zdimsz  | ydimsz  | xdimsz  |
 | 0b01   | submode| offset | invxyz | submode2| rsvd    | rsvd    | xdimsz  |
+| 0b00   | elwidth| offset | invxyz |0b110/0b111| SVGPR | ydimsz  | xdimsz  |
 
 mode sets different behaviours (straight matrix multiply, FFT, DCT).
 
 * **mode=0b00** sets straight Matrix Mode
+* **mode=0b00** with permute=0b110 or 0b111 sets Indexed Mode
 * **mode=0b01** sets "FFT/DCT" mode and activates submodes
 
-**FFT/DCT mode**
+## FFT/DCT mode
 
 submode2=0 is for FFT. For FFT submode the following schedules may be 
 selected:
@@ -42,7 +44,7 @@ is also performed.
 * **submode=0b00** selects the ``j`` offset of the innermost for-loop,
 * **submode=0b01** selects the ``j+1`` offset of the innermost for-loop,
 
-**Matrix Mode**
+## Matrix Mode
 
 In Matrix Mode, skip allows dimensions to be skipped from being included
 in the resultant output index.  this allows sequences to be repeated:
@@ -91,6 +93,8 @@ below shows how the permutation dimensionality order works:
 | 011     | 1,2,0 | (ydim+1)(zdim+1)(xdim+1) |
 | 100     | 2,0,1 | (zdim+1)(xdim+1)(ydim+1) |
 | 101     | 2,1,0 | (zdim+1)(ydim+1)(xdim+1) |
+| 110     | 0,1   | Indexed (xdim+1)(ydim+1) |
+| 110     | 1,0   | Indexed (ydim+1)(xdim+1) |
 
 In other words, the "permute" option changes the order in which
 nested for-loops over the array would be done.  See executable