OSchip
/
OpenBLAS
Not watched
1
0
Fork 0
Code
Releases 66 Wiki evaluate Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain HPC
Browse Source

Added rot functions. 

Signed-off-by: Shivraj Patil <shivraj.patil@imgtec.com>
tags/v0.2.20^2
Shivraj Patil 9 years ago
parent
e8d0e66982
commit
a4d97d980f
5 changed files with 3995 additions and 0 deletions
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. +7
    -0
      kernel/mips/KERNEL.P5600
  2. +1072
    -0
      kernel/mips/crot_msa.c
  3. +1055
    -0
      kernel/mips/drot_msa.c
  4. +1123
    -0
      kernel/mips/srot_msa.c
  5. +738
    -0
      kernel/mips/zrot_msa.c

+ 7
- 0
kernel/mips/KERNEL.P5600 View File

@@ -83,10 +83,17 @@ DNRM2KERNEL = ../mips/nrm2.c
CNRM2KERNEL = ../mips/znrm2.c
ZNRM2KERNEL = ../mips/znrm2.c

ifdef HAVE_MSA
SROTKERNEL = ../mips/srot_msa.c
DROTKERNEL = ../mips/drot_msa.c
CROTKERNEL = ../mips/crot_msa.c
ZROTKERNEL = ../mips/zrot_msa.c
else
SROTKERNEL = ../mips/rot.c
DROTKERNEL = ../mips/rot.c
CROTKERNEL = ../mips/zrot.c
ZROTKERNEL = ../mips/zrot.c
endif

ifdef HAVE_MSA
SSCALKERNEL = ../mips/sscal_msa.c


+ 1072
- 0
kernel/mips/crot_msa.c
File diff suppressed because it is too large
View File


+ 1055
- 0
kernel/mips/drot_msa.c
File diff suppressed because it is too large
View File


+ 1123
- 0
kernel/mips/srot_msa.c
File diff suppressed because it is too large
View File


+ 738
- 0
kernel/mips/zrot_msa.c View File

@@ -0,0 +1,738 @@
/*******************************************************************************
Copyright (c) 2016, The OpenBLAS Project
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
3. Neither the name of the OpenBLAS project nor the names of
its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*******************************************************************************/

#include "common.h"
#include "macros_msa.h"

#define PROCESS_ZROT(inc_x2, inc_y2) \
if ((0 == c) && (0 == s)) \
{ \
v2f64 zero = {0, 0}; \
zero = (v2f64) __msa_insert_d((v2i64) zero, 0, 0.0); \
zero = (v2f64) __msa_insert_d((v2i64) zero, 1, 0.0); \
\
/* process 4 floats */ \
for (j = (n >> 1); j--;) \
{ \
ST_DP2_INC(zero, zero, px, inc_x2); \
ST_DP2_INC(zero, zero, py, inc_y2); \
} \
\
if (n & 1) \
{ \
ST_DP(zero, px); \
ST_DP(zero, py); \
} \
} \
else if ((1 == c) && (1 == s)) \
{ \
/* process 8 elements */ \
if (n >> 3) \
{ \
BLASLONG pref_offsetx, pref_offsety; \
\
pref_offsetx = (BLASLONG)px & (L1_DATA_LINESIZE - 1); \
if (pref_offsetx > 0) \
{ \
pref_offsetx = L1_DATA_LINESIZE - pref_offsetx; \
pref_offsetx = pref_offsetx / sizeof(FLOAT); \
} \
\
pref_offsety = (BLASLONG)py & (L1_DATA_LINESIZE - 1); \
if (pref_offsety > 0) \
{ \
pref_offsety = L1_DATA_LINESIZE - pref_offsety; \
pref_offsety = pref_offsety / sizeof(FLOAT); \
} \
\
x0 = LD_DP(px); px += inc_x2; \
x1 = LD_DP(px); px += inc_x2; \
x2 = LD_DP(px); px += inc_x2; \
x3 = LD_DP(px); px += inc_x2; \
y0 = LD_DP(py); py += inc_y2; \
y1 = LD_DP(py); py += inc_y2; \
y2 = LD_DP(py); py += inc_y2; \
y3 = LD_DP(py); py += inc_y2; \
\
for (j = (n >> 3) - 1; j--;) \
{ \
PREFETCH(px + pref_offsetx + 16); \
PREFETCH(px + pref_offsetx + 20); \
PREFETCH(px + pref_offsetx + 24); \
PREFETCH(px + pref_offsetx + 28); \
PREFETCH(py + pref_offsety + 16); \
PREFETCH(py + pref_offsety + 20); \
PREFETCH(py + pref_offsety + 24); \
PREFETCH(py + pref_offsety + 28); \
\
out0 = x0 + y0; \
x4 = LD_DP(px); px += inc_x2; \
out1 = y0 - x0; \
x5 = LD_DP(px); px += inc_x2; \
out2 = x1 + y1; \
x6 = LD_DP(px); px += inc_x2; \
out3 = y1 - x1; \
x7 = LD_DP(px); px += inc_x2; \
out4 = x2 + y2; \
y4 = LD_DP(py); py += inc_y2; \
out5 = y2 - x2; \
y5 = LD_DP(py); py += inc_y2; \
out6 = x3 + y3; \
y6 = LD_DP(py); py += inc_y2; \
out7 = y3 - x3; \
y7 = LD_DP(py); py += inc_y2; \
\
ST_DP(out0, x); x += inc_x2; \
out8 = x4 + y4; \
ST_DP(out1, y); y += inc_y2; \
out9 = y4 - x4; \
ST_DP(out2, x); x += inc_x2; \
out10 = x5 + y5; \
ST_DP(out3, y); y += inc_y2; \
out11 = y5 - x5; \
ST_DP(out4, x); x += inc_x2; \
out12 = x6 + y6; \
ST_DP(out5, y); y += inc_y2; \
out13 = y6 - x6; \
ST_DP(out6, x); x += inc_x2; \
out14 = x7 + y7; \
ST_DP(out7, y); y += inc_y2; \
out15 = y7 - x7; \
\
x0 = LD_DP(px); px += inc_x2; \
ST_DP(out8, x); x += inc_x2; \
x1 = LD_DP(px); px += inc_x2; \
ST_DP(out10, x); x += inc_x2; \
x2 = LD_DP(px); px += inc_x2; \
ST_DP(out12, x); x += inc_x2; \
x3 = LD_DP(px); px += inc_x2; \
ST_DP(out14, x); x += inc_x2; \
\
y0 = LD_DP(py); py += inc_y2; \
ST_DP(out9, y); y += inc_y2; \
y1 = LD_DP(py); py += inc_y2; \
ST_DP(out11, y); y += inc_y2; \
y2 = LD_DP(py); py += inc_y2; \
ST_DP(out13, y); y += inc_y2; \
y3 = LD_DP(py); py += inc_y2; \
ST_DP(out15, y); y += inc_y2; \
} \
\
x4 = LD_DP(px); px += inc_x2; \
x5 = LD_DP(px); px += inc_x2; \
x6 = LD_DP(px); px += inc_x2; \
x7 = LD_DP(px); px += inc_x2; \
y4 = LD_DP(py); py += inc_y2; \
y5 = LD_DP(py); py += inc_y2; \
y6 = LD_DP(py); py += inc_y2; \
y7 = LD_DP(py); py += inc_y2; \
\
out0 = x0 + y0; \
out1 = y0 - x0; \
out2 = x1 + y1; \
out3 = y1 - x1; \
out4 = x2 + y2; \
out5 = y2 - x2; \
out6 = x3 + y3; \
out7 = y3 - x3; \
out8 = x4 + y4; \
out9 = y4 - x4; \
out10 = x5 + y5; \
out11 = y5 - x5; \
out12 = x6 + y6; \
out13 = y6 - x6; \
out14 = x7 + y7; \
out15 = y7 - x7; \
\
ST_DP8_INC(out0, out2, out4, out6, out8, out10, out12, out14, x, inc_x2); \
ST_DP8_INC(out1, out3, out5, out7, out9, out11, out13, out15, y, inc_y2); \
} \
if (n & 4) \
{ \
LD_DP4_INC(px, inc_x2, x0, x1, x2, x3); \
LD_DP4_INC(py, inc_y2, y0, y1, y2, y3); \
\
out0 = x0 + y0; \
out1 = y0 - x0; \
out2 = x1 + y1; \
out3 = y1 - x1; \
out4 = x2 + y2; \
out5 = y2 - x2; \
out6 = x3 + y3; \
out7 = y3 - x3; \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
} \
if (n & 2) \
{ \
LD_DP2_INC(px, inc_x2, x0, x1); \
LD_DP2_INC(py, inc_y2, y0, y1); \
\
out0 = x0 + y0; \
out1 = y0 - x0; \
out2 = x1 + y1; \
out3 = y1 - x1; \
\
ST_DP2_INC(out0, out2, x, inc_x2); \
ST_DP2_INC(out1, out3, y, inc_y2); \
} \
if (n & 1) \
{ \
x0 = LD_DP(px); \
y0 = LD_DP(py); \
\
out0 = x0 + y0; \
out1 = y0 - x0; \
\
ST_DP(out0, px); \
ST_DP(out1, py); \
} \
} \
else if (0 == s) \
{ \
c0 = COPY_DOUBLE_TO_VECTOR(c); \
\
if (n >> 3) \
{ \
BLASLONG pref_offsetx, pref_offsety; \
\
pref_offsetx = (BLASLONG)px & (L1_DATA_LINESIZE - 1); \
if (pref_offsetx > 0) \
{ \
pref_offsetx = L1_DATA_LINESIZE - pref_offsetx; \
pref_offsetx = pref_offsetx / sizeof(FLOAT); \
} \
\
pref_offsety = (BLASLONG)py & (L1_DATA_LINESIZE - 1); \
if (pref_offsety > 0) \
{ \
pref_offsety = L1_DATA_LINESIZE - pref_offsety; \
pref_offsety = pref_offsety / sizeof(FLOAT); \
} \
\
LD_DP8_INC(px, inc_x2, x0, x1, x2, x3, x4, x5, x6, x7); \
\
for (j = (n >> 3) - 1; j--;) \
{ \
PREFETCH(px + pref_offsetx + 16); \
PREFETCH(px + pref_offsetx + 20); \
PREFETCH(px + pref_offsetx + 24); \
PREFETCH(px + pref_offsetx + 28); \
PREFETCH(py + pref_offsety + 16); \
PREFETCH(py + pref_offsety + 20); \
PREFETCH(py + pref_offsety + 24); \
PREFETCH(py + pref_offsety + 28); \
\
y0 = LD_DP(py); py += inc_y2; \
x0 *= c0; \
y1 = LD_DP(py); py += inc_y2; \
x1 *= c0; \
y2 = LD_DP(py); py += inc_y2; \
x2 *= c0; \
y3 = LD_DP(py); py += inc_y2; \
x3 *= c0; \
y4 = LD_DP(py); py += inc_y2; \
x4 *= c0; \
y5 = LD_DP(py); py += inc_y2; \
x5 *= c0; \
y6 = LD_DP(py); py += inc_y2; \
x6 *= c0; \
y7 = LD_DP(py); py += inc_y2; \
x7 *= c0; \
\
ST_DP(x0, x); x += inc_x2; \
y0 *= c0; \
ST_DP(x1, x); x += inc_x2; \
y1 *= c0; \
ST_DP(x2, x); x += inc_x2; \
y2 *= c0; \
ST_DP(x3, x); x += inc_x2; \
y3 *= c0; \
ST_DP(x4, x); x += inc_x2; \
y4 *= c0; \
ST_DP(x5, x); x += inc_x2; \
y5 *= c0; \
ST_DP(x6, x); x += inc_x2; \
y6 *= c0; \
ST_DP(x7, x); x += inc_x2; \
y7 *= c0; \
\
x0 = LD_DP(px); px += inc_x2; \
ST_DP(y0, y); y += inc_y2; \
x1 = LD_DP(px); px += inc_x2; \
ST_DP(y1, y); y += inc_y2; \
x2 = LD_DP(px); px += inc_x2; \
ST_DP(y2, y); y += inc_y2; \
x3 = LD_DP(px); px += inc_x2; \
ST_DP(y3, y); y += inc_y2; \
x4 = LD_DP(px); px += inc_x2; \
ST_DP(y4, y); y += inc_y2; \
x5 = LD_DP(px); px += inc_x2; \
ST_DP(y5, y); y += inc_y2; \
x6 = LD_DP(px); px += inc_x2; \
ST_DP(y6, y); y += inc_y2; \
x7 = LD_DP(px); px += inc_x2; \
ST_DP(y7, y); y += inc_y2; \
} \
\
LD_DP8_INC(py, inc_y2, y0, y1, y2, y3, y4, y5, y6, y7); \
\
x0 *= c0; \
y0 *= c0; \
x1 *= c0; \
y1 *= c0; \
x2 *= c0; \
y2 *= c0; \
x3 *= c0; \
y3 *= c0; \
x4 *= c0; \
y4 *= c0; \
x5 *= c0; \
y5 *= c0; \
x6 *= c0; \
y6 *= c0; \
x7 *= c0; \
y7 *= c0; \
\
ST_DP8_INC(x0, x1, x2, x3, x4, x5, x6, x7, x, inc_x2); \
ST_DP8_INC(y0, y1, y2, y3, y4, y5, y6, y7, y, inc_y2); \
} \
\
if (n & 4) \
{ \
LD_DP4_INC(px, inc_x2, x0, x1, x2, x3); \
LD_DP4_INC(py, inc_y2, y0, y1, y2, y3); \
\
out0 = c0 * x0; \
out1 = c0 * y0; \
out2 = c0 * x1; \
out3 = c0 * y1; \
out4 = c0 * x2; \
out5 = c0 * y2; \
out6 = c0 * x3; \
out7 = c0 * y3; \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
} \
if (n & 2) \
{ \
LD_DP2_INC(px, inc_x2, x0, x1); \
LD_DP2_INC(py, inc_y2, y0, y1); \
\
out0 = c0 * x0; \
out1 = c0 * y0; \
out2 = c0 * x1; \
out3 = c0 * y1; \
\
ST_DP2_INC(out0, out2, x, inc_x2); \
ST_DP2_INC(out1, out3, y, inc_y2); \
} \
if (n & 1) \
{ \
x0 = LD_DP(px); \
y0 = LD_DP(py); \
\
out0 = c0 * x0; \
out1 = c0 * y0; \
\
ST_DP(out0, px); \
ST_DP(out1, py); \
} \
} \
else if (0 == c) \
{ \
s0 = COPY_DOUBLE_TO_VECTOR(s); \
\
/* process 16 floats */ \
if (n >> 3) \
{ \
BLASLONG pref_offsetx, pref_offsety; \
\
pref_offsetx = (BLASLONG)px & (L1_DATA_LINESIZE - 1); \
if (pref_offsetx > 0) \
{ \
pref_offsetx = L1_DATA_LINESIZE - pref_offsetx; \
pref_offsetx = pref_offsetx / sizeof(FLOAT); \
} \
\
pref_offsety = (BLASLONG)py & (L1_DATA_LINESIZE - 1); \
if (pref_offsety > 0) \
{ \
pref_offsety = L1_DATA_LINESIZE - pref_offsety; \
pref_offsety = pref_offsety / sizeof(FLOAT); \
} \
\
LD_DP4_INC(px, inc_x2, x0, x1, x2, x3); \
LD_DP4_INC(py, inc_y2, y0, y1, y2, y3); \
\
for (j = (n >> 3) - 1; j--;) \
{ \
PREFETCH(px + pref_offsetx + 16); \
PREFETCH(px + pref_offsetx + 20); \
PREFETCH(px + pref_offsetx + 24); \
PREFETCH(px + pref_offsetx + 28); \
PREFETCH(py + pref_offsety + 16); \
PREFETCH(py + pref_offsety + 20); \
PREFETCH(py + pref_offsety + 24); \
PREFETCH(py + pref_offsety + 28); \
\
x4 = LD_DP(px); px += inc_x2; \
out0 = s0 * y0; \
x5 = LD_DP(px); px += inc_x2; \
out2 = s0 * y1; \
x6 = LD_DP(px); px += inc_x2; \
out4 = s0 * y2; \
x7 = LD_DP(px); px += inc_x2; \
out6 = s0 * y3; \
y4 = LD_DP(py); py += inc_y2; \
out1 = -(s0 * x0); \
y5 = LD_DP(py); py += inc_y2; \
out3 = -(s0 * x1); \
y6 = LD_DP(py); py += inc_y2; \
out5 = -(s0 * x2); \
y7 = LD_DP(py); py += inc_y2; \
out7 = -(s0 * x3); \
\
ST_DP(out0, x); x += inc_y2; \
out0 = s0 * y4; \
ST_DP(out2, x); x += inc_y2; \
out2 = s0 * y5; \
ST_DP(out4, x); x += inc_y2; \
out4 = s0 * y6; \
ST_DP(out6, x); x += inc_y2; \
out6 = s0 * y7; \
ST_DP(out1, y); y += inc_y2; \
out1 = -(s0 * x4); \
ST_DP(out3, y); y += inc_y2; \
out3 = -(s0 * x5); \
ST_DP(out5, y); y += inc_y2; \
out5 = -(s0 * x6); \
ST_DP(out7, y); y += inc_y2; \
out7 = -(s0 * x7); \
\
x0 = LD_DP(px); px += inc_x2; \
ST_DP(out0, x); x += inc_y2; \
x1 = LD_DP(px); px += inc_x2; \
ST_DP(out2, x); x += inc_y2; \
x2 = LD_DP(px); px += inc_x2; \
ST_DP(out4, x); x += inc_y2; \
x3 = LD_DP(px); px += inc_x2; \
ST_DP(out6, x); x += inc_y2; \
y0 = LD_DP(py); py += inc_y2; \
ST_DP(out1, y); y += inc_y2; \
y1 = LD_DP(py); py += inc_y2; \
ST_DP(out3, y); y += inc_y2; \
y2 = LD_DP(py); py += inc_y2; \
ST_DP(out5, y); y += inc_y2; \
y3 = LD_DP(py); py += inc_y2; \
ST_DP(out7, y); y += inc_y2; \
} \
\
out0 = s0 * y0; \
out2 = s0 * y1; \
out4 = s0 * y2; \
out6 = s0 * y3; \
out1 = -(s0 * x0); \
out3 = -(s0 * x1); \
out5 = -(s0 * x2); \
out7 = -(s0 * x3); \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
\
LD_DP4_INC(px, inc_x2, x4, x5, x6, x7); \
LD_DP4_INC(py, inc_y2, y4, y5, y6, y7); \
\
out0 = s0 * y4; \
out2 = s0 * y5; \
out4 = s0 * y6; \
out6 = s0 * y7; \
out1 = -(s0 * x4); \
out3 = -(s0 * x5); \
out5 = -(s0 * x6); \
out7 = -(s0 * x7); \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
} \
if (n & 4) \
{ \
LD_DP4_INC(px, inc_x2, x0, x1, x2, x3); \
LD_DP4_INC(py, inc_y2, y0, y1, y2, y3); \
\
out0 = s0 * y0; \
out1 = - (s0 * x0); \
out2 = s0 * y1; \
out3 = - (s0 * x1); \
out4 = s0 * y2; \
out5 = - (s0 * x2); \
out6 = s0 * y3; \
out7 = - (s0 * x3); \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
} \
if (n & 2) \
{ \
LD_DP2_INC(px, inc_x2, x0, x1); \
LD_DP2_INC(py, inc_y2, y0, y1); \
\
out0 = s0 * y0; \
out1 = - (s0 * x0); \
out2 = s0 * y1; \
out3 = - (s0 * x1); \
\
ST_DP2_INC(out0, out2, x, inc_x2); \
ST_DP2_INC(out1, out3, y, inc_y2); \
} \
if (n & 1) \
{ \
x0 = LD_DP(px); px += inc_x2; \
y0 = LD_DP(py); py += inc_y2; \
\
out0 = s0 * y0; \
out1 = - (s0 * x0); \
\
ST_DP(out0, x); x += inc_x2; \
ST_DP(out1, y); y += inc_y2; \
} \
} \
else \
{ \
c0 = COPY_DOUBLE_TO_VECTOR(c); \
s0 = COPY_DOUBLE_TO_VECTOR(s); \
\
if (n >> 3) \
{ \
BLASLONG pref_offsetx, pref_offsety; \
\
pref_offsetx = (BLASLONG)px & (L1_DATA_LINESIZE - 1); \
if (pref_offsetx > 0) \
{ \
pref_offsetx = L1_DATA_LINESIZE - pref_offsetx; \
pref_offsetx = pref_offsetx / sizeof(FLOAT); \
} \
\
pref_offsety = (BLASLONG)py & (L1_DATA_LINESIZE - 1); \
if (pref_offsety > 0) \
{ \
pref_offsety = L1_DATA_LINESIZE - pref_offsety; \
pref_offsety = pref_offsety / sizeof(FLOAT); \
} \
\
LD_DP4_INC(px, inc_x2, x0, x1, x2, x3); \
LD_DP4_INC(py, inc_y2, y0, y1, y2, y3); \
\
for (j = (n >> 3) - 1; j--;) \
{ \
PREFETCH(px + pref_offsetx + 16); \
PREFETCH(px + pref_offsetx + 20); \
PREFETCH(px + pref_offsetx + 24); \
PREFETCH(px + pref_offsetx + 28); \
PREFETCH(py + pref_offsety + 16); \
PREFETCH(py + pref_offsety + 20); \
PREFETCH(py + pref_offsety + 24); \
PREFETCH(py + pref_offsety + 28); \
\
x4 = LD_DP(px); px += inc_x2; \
out0 = c0 * x0; \
x5 = LD_DP(px); px += inc_x2; \
out2 = c0 * x1; \
x6 = LD_DP(px); px += inc_x2; \
out4 = c0 * x2; \
x7 = LD_DP(px); px += inc_x2; \
out6 = c0 * x3; \
y4 = LD_DP(py); py += inc_y2; \
out1 = c0 * y0; \
y5 = LD_DP(py); py += inc_y2; \
out3 = c0 * y1; \
y6 = LD_DP(py); py += inc_y2; \
out5 = c0 * y2; \
y7 = LD_DP(py); py += inc_y2; \
out7 = c0 * y3; \
\
out0 += s0 * y0; \
out2 += s0 * y1; \
out4 += s0 * y2; \
out6 += s0 * y3; \
out1 -= s0 * x0; \
out3 -= s0 * x1; \
out5 -= s0 * x2; \
out7 -= s0 * x3; \
\
ST_DP(out0, x); x += inc_x2; \
out0 = c0 * x4; \
ST_DP(out2, x); x += inc_x2; \
out2 = c0 * x5; \
ST_DP(out4, x); x += inc_x2; \
out4 = c0 * x6; \
ST_DP(out6, x); x += inc_x2; \
out6 = c0 * x7; \
ST_DP(out1, y); y += inc_y2; \
out1 = c0 * y4; \
ST_DP(out3, y); y += inc_y2; \
out3 = c0 * y5; \
ST_DP(out5, y); y += inc_y2; \
out5 = c0 * y6; \
ST_DP(out7, y); y += inc_y2; \
out7 = c0 * y7; \
\
x0 = LD_DP(px); px += inc_x2; \
out0 += s0 * y4; \
x1 = LD_DP(px); px += inc_x2; \
out2 += s0 * y5; \
x2 = LD_DP(px); px += inc_x2; \
out4 += s0 * y6; \
x3 = LD_DP(px); px += inc_x2; \
out6 += s0 * y7; \
y0 = LD_DP(py); py += inc_y2; \
out1 -= s0 * x4; \
y1 = LD_DP(py); py += inc_y2; \
out3 -= s0 * x5; \
y2 = LD_DP(py); py += inc_y2; \
out5 -= s0 * x6; \
y3 = LD_DP(py); py += inc_y2; \
out7 -= s0 * x7; \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
} \
\
out0 = c0 * x0; \
out0 += s0 * y0; \
out1 = c0 * y0; \
out1 -= s0 * x0; \
out2 = c0 * x1; \
out2 += s0 * y1; \
out3 = c0 * y1; \
out3 -= s0 * x1; \
out4 = c0 * x2; \
out4 += s0 * y2; \
out5 = c0 * y2; \
out5 -= s0 * x2; \
out6 = c0 * x3; \
out6 += s0 * y3; \
out7 = c0 * y3; \
out7 -= s0 * x3; \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
\
LD_DP4_INC(px, inc_x2, x4, x5, x6, x7); \
LD_DP4_INC(py, inc_y2, y4, y5, y6, y7); \
\
out8 = c0 * x4; \
out8 += s0 * y4; \
out9 = c0 * y4; \
out9 -= s0 * x4; \
out10 = c0 * x5; \
out10 += s0 * y5; \
out11 = c0 * y5; \
out11 -= s0 * x5; \
out12 = c0 * x6; \
out12 += s0 * y6; \
out13 = c0 * y6; \
out13 -= s0 * x6; \
out14 = c0 * x7; \
out14 += s0 * y7; \
out15 = c0 * y7; \
out15 -= s0 * x7; \
\
ST_DP4_INC(out8, out10, out12, out14, x, inc_x2); \
ST_DP4_INC(out9, out11, out13, out15, y, inc_y2); \
} \
if (n & 4) \
{ \
LD_DP4_INC(px, inc_x2, x0, x1, x2, x3); \
LD_DP4_INC(py, inc_y2, y0, y1, y2, y3); \
\
out0 = (c0 * x0) + (s0 * y0); \
out1 = (c0 * y0) - (s0 * x0); \
out2 = (c0 * x1) + (s0 * y1); \
out3 = (c0 * y1) - (s0 * x1); \
out4 = (c0 * x2) + (s0 * y2); \
out5 = (c0 * y2) - (s0 * x2); \
out6 = (c0 * x3) + (s0 * y3); \
out7 = (c0 * y3) - (s0 * x3); \
\
ST_DP4_INC(out0, out2, out4, out6, x, inc_x2); \
ST_DP4_INC(out1, out3, out5, out7, y, inc_y2); \
} \
if (n & 2) \
{ \
LD_DP2_INC(px, inc_x2, x0, x1); \
LD_DP2_INC(py, inc_y2, y0, y1); \
\
out0 = (c0 * x0) + (s0 * y0); \
out1 = (c0 * y0) - (s0 * x0); \
out2 = (c0 * x1) + (s0 * y1); \
out3 = (c0 * y1) - (s0 * x1); \
\
ST_DP2_INC(out0, out2, x, inc_x2); \
ST_DP2_INC(out1, out3, y, inc_y2); \
} \
if (n & 1) \
{ \
x0 = LD_DP(px); \
y0 = LD_DP(py); \
\
out0 = (c0 * x0) + (s0 * y0); \
out1 = (c0 * y0) - (s0 * x0); \
\
ST_DP(out0, px); \
ST_DP(out1, py); \
} \
}

int CNAME(BLASLONG n, FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y,
FLOAT c, FLOAT s)
{
BLASLONG j;
FLOAT *px, *py;
v2f64 x0, x1, x2, x3, x4, x5, x6, x7, y0, y1, y2, y3, y4, y5, y6, y7;
v2f64 out0, out1, out2, out3, out4, out5, out6, out7, c0, s0;
v2f64 out8, out9, out10, out11, out12, out13, out14, out15;

px = x;
py = y;

if ((1 == inc_x) && (1 == inc_y))
{
PROCESS_ZROT(2, 2);
}
else
{
inc_x *= 2;
inc_y *= 2;

PROCESS_ZROT(inc_x, inc_y);
}

return 0;
}

Write Preview
Loading…
Cancel
Save