summaryrefslogtreecommitdiffstats
path: root/src/gallium/docs/source
diff options
context:
space:
mode:
authorMarek Olšák <[email protected]>2015-02-28 00:26:31 +0100
committerMarek Olšák <[email protected]>2015-03-16 12:54:18 +0100
commit216543ea547dd0572d9f2f0364f7a239a5aeafe1 (patch)
treea7d6fb2160c1c736e0dcfb2d24790d6cb2ce315d /src/gallium/docs/source
parente92bc6b38e90339a394e95a562bcce35c3ee9696 (diff)
gallium: add FMA and DFMA opcodes (v3)
Needed by ARB_gpu_shader5. v2: select DMAD for FMA with double precision v3: add and select DFMA Reviewed-by: Ilia Mirkin <[email protected]>
Diffstat (limited to 'src/gallium/docs/source')
-rw-r--r--src/gallium/docs/source/screen.rst2
-rw-r--r--src/gallium/docs/source/tgsi.rst26
2 files changed, 28 insertions, 0 deletions
diff --git a/src/gallium/docs/source/screen.rst b/src/gallium/docs/source/screen.rst
index e0fd1a2dbac..26cc9ffc6f7 100644
--- a/src/gallium/docs/source/screen.rst
+++ b/src/gallium/docs/source/screen.rst
@@ -336,6 +336,8 @@ to be 0.
is supported. If it is, DTRUNC/DCEIL/DFLR/DROUND opcodes may be used.
* ``PIPE_SHADER_CAP_TGSI_DFRACEXP_DLDEXP_SUPPORTED``: Whether DFRACEXP and
DLDEXP are supported.
+* ``PIPE_SHADER_CAP_TGSI_FMA_SUPPORTED``: Whether FMA and DFMA (doubles only)
+ are supported.
.. _pipe_compute_cap:
diff --git a/src/gallium/docs/source/tgsi.rst b/src/gallium/docs/source/tgsi.rst
index b0a975aa70a..7771136f167 100644
--- a/src/gallium/docs/source/tgsi.rst
+++ b/src/gallium/docs/source/tgsi.rst
@@ -272,6 +272,21 @@ This instruction replicates its result.
dst.w = src0.w \times src1.w + (1 - src0.w) \times src2.w
+.. opcode:: FMA - Fused Multiply-Add
+
+Perform a * b + c with no intermediate rounding step.
+
+.. math::
+
+ dst.x = src0.x \times src1.x + src2.x
+
+ dst.y = src0.y \times src1.y + src2.y
+
+ dst.z = src0.z \times src1.z + src2.z
+
+ dst.w = src0.w \times src1.w + src2.w
+
+
.. opcode:: DP2A - 2-component Dot Product And Add
.. math::
@@ -1962,6 +1977,17 @@ source is an integer.
dst.zw = src0.zw \times src1.zw + src2.zw
+.. opcode:: DFMA - Fused Multiply-Add
+
+Perform a * b + c with no intermediate rounding step.
+
+.. math::
+
+ dst.xy = src0.xy \times src1.xy + src2.xy
+
+ dst.zw = src0.zw \times src1.zw + src2.zw
+
+
.. opcode:: DRCP - Reciprocal
.. math::