From 9c3ecc56ce333f56b48033ee49d26846e3651d53 Mon Sep 17 00:00:00 2001 From: lkcl Date: Sun, 26 Sep 2021 11:14:42 +0100 Subject: [PATCH] --- 3d_gpu/architecture/dynamic_simd/assign.mdwn | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/3d_gpu/architecture/dynamic_simd/assign.mdwn b/3d_gpu/architecture/dynamic_simd/assign.mdwn index a8c43e450..447d92e86 100644 --- a/3d_gpu/architecture/dynamic_simd/assign.mdwn +++ b/3d_gpu/architecture/dynamic_simd/assign.mdwn @@ -90,7 +90,7 @@ This is similar to the parallel case except A is repeated | 100 | A3A2A1A0 | [A7A7A7A7] | A7A6A5A4 | A3A2A1A0 | | 101 | A3A2A1A0 | A7A6A5A4 | A3A2A1A0 | A3A2A1A0 | | 110 | A3A2A1A0 | A3A2A1A0 | A7A6A5A4 | A3A2A1A0 | -| 111 | A3A2A1A0 | A3A2A1A0 | A3A2A1A0v | A3A2A1A0 | +| 111 | A3A2A1A0 | A3A2A1A0 | A3A2A1A0 | A3A2A1A0 | Note how when the entire partition set is open (1x 16-bit output) that all of A is copied out, and either zero or sign extended @@ -100,4 +100,7 @@ from the first 4 bits of A. Unlike the parallel case, A is not itself partitioned, so is copied over as much as is possible. In some cases such as `1x 4-bit, 1x 12-bit` -(partition mask = `0b100`, above) the 8-bit scalar source will need sign or zero extending. +(partition mask = `0b100`, above) when copying the 8-bit scalar source +into the highest part of B (o3) it is truncated to 4 bis (because +each partition of B is only 4 bits) but for copying to the 12-bit partition +(o2-o1-00) the 8-bit scalar source, A, will need sign or zero extending. -- 2.30.2