Move SSE2 and SSE4.1 intrinsics functions to separate files, to be compiled with...
authorJonathan Lennox <jonathan@vidyo.com>
Mon, 3 Aug 2015 21:04:26 +0000 (17:04 -0400)
committerJean-Marc Valin <jmvalin@jmvalin.ca>
Tue, 1 Sep 2015 21:21:31 +0000 (17:21 -0400)
Makefile.am
celt/tests/test_unit_mathops.c
celt/tests/test_unit_rotation.c
celt/x86/pitch_sse.c
celt/x86/pitch_sse2.c [new file with mode: 0644]
celt/x86/pitch_sse4_1.c [new file with mode: 0644]
celt_sources.mk
configure.ac

index 51bce63..efff337 100644 (file)
@@ -30,12 +30,14 @@ else
 OPUS_SOURCES += $(OPUS_SOURCES_FLOAT)
 endif
 
-if HAVE_SSE4_1
-CELT_SOURCES += $(CELT_SOURCES_SSE) $(CELT_SOURCES_SSE4_1)
-else
-if HAVE_SSE2
+if HAVE_SSE
 CELT_SOURCES += $(CELT_SOURCES_SSE)
 endif
+if HAVE_SSE2
+CELT_SOURCES += $(CELT_SOURCES_SSE2)
+endif
+if HAVE_SSE4_1
+CELT_SOURCES += $(CELT_SOURCES_SSE4_1)
 endif
 
 if CPU_ARM
@@ -257,18 +259,29 @@ $(CELT_SOURCES_ARM_ASM:%.s=%-gnu.S): $(top_srcdir)/celt/arm/arm2gnu.pl
 %-gnu.S: %.s
        $(top_srcdir)/celt/arm/arm2gnu.pl @ARM2GNU_PARAMS@ < $< > $@
 
-SSE_OBJ = %_sse.o %_sse.lo %test_unit_mathops.o %test_unit_rotation.o
+OPT_UNIT_TEST_OBJ = $(celt_tests_test_unit_mathops_SOURCES:.c=.o) \
+                    $(celt_tests_test_unit_rotation_SOURCES:.c=.o) \
+                    $(celt_tests_test_unit_mdct_SOURCES:.c=.o) \
+                    $(celt_tests_test_unit_dft_SOURCES:.c=.o)
+
+if HAVE_SSE
+SSE_OBJ = $(CELT_SOURCES_SSE:.c=.lo)
+$(SSE_OBJ) $(OPT_UNIT_TEST_OBJ): CFLAGS += $(OPUS_X86_SSE_CFLAGS)
+endif
 
-if HAVE_SSE4_1
-$(SSE_OBJ): CFLAGS += $(OPUS_X86_SSE4_1_CFLAGS)
-else
 if HAVE_SSE2
-$(SSE_OBJ): CFLAGS += $(OPUS_X86_SSE2_CFLAGS)
+SSE2_OBJ = $(CELT_SOURCES_SSE2:.c=.lo)
+$(SSE2_OBJ) $(OPT_UNIT_TEST_OBJ): CFLAGS += $(OPUS_X86_SSE2_CFLAGS)
 endif
+
+if HAVE_SSE4_1
+SSE4_1_OBJ = $(CELT_SOURCES_SSE4_1:.c=.lo) \
+             $(SILK_SOURCES_SSE4_1:.c=.lo) \
+             $(SILK_SOURCES_FIXED_SSE4_1:.c=.lo)
+$(SSE4_1_OBJ) $(OPT_UNIT_TEST_OBJ): CFLAGS += $(OPUS_X86_SSE4_1_CFLAGS)
 endif
 
 if OPUS_ARM_NEON_INTR
-CELT_ARM_NEON_INTR_OBJ = $(CELT_SOURCES_ARM_NEON_INTR:.c=.lo) \
-                       %test_unit_rotation.o %test_unit_mathops.o
-$(CELT_ARM_NEON_INTR_OBJ): CFLAGS += $(OPUS_ARM_NEON_INTR_CFLAGS)
+CELT_ARM_NEON_INTR_OBJ = $(CELT_SOURCES_ARM_NEON_INTR:.c=.lo)
+$(CELT_ARM_NEON_INTR_OBJ) $(OPT_UNIT_TEST_OBJ): CFLAGS += $(OPUS_ARM_NEON_INTR_CFLAGS)
 endif
index b9b1bcf..2de39ba 100644 (file)
 #include "cwrs.c"
 #include "pitch.c"
 #include "celt_lpc.c"
+#include "celt.c"
 
-#if defined(OPUS_X86_MAY_HAVE_SSE4_1) || defined(OPUS_X86_MAY_HAVE_SSE2)
+#if defined(OPUS_X86_MAY_HAVE_SSE) || defined(OPUS_X86_MAY_HAVE_SSE2) || defined(OPUS_X86_MAY_HAVE_SSE4_1)
+#if defined(OPUS_X86_MAY_HAVE_SSE)
 #include "x86/pitch_sse.c"
+#endif
+#if defined(OPUS_X86_MAY_HAVE_SSE2)
+#include "x86/pitch_sse2.c"
+#endif
 #if defined(OPUS_X86_MAY_HAVE_SSE4_1)
+#include "x86/pitch_sse4_1.c"
 #include "x86/celt_lpc_sse.c"
 #endif
 #include "x86/x86_celt_map.c"
index 5507884..4780005 100644 (file)
 #include "bands.h"
 #include "pitch.c"
 #include "celt_lpc.c"
+#include "celt.c"
 #include <math.h>
 
-#if defined(OPUS_X86_MAY_HAVE_SSE4_1) || defined(OPUS_X86_MAY_HAVE_SSE2)
+#if defined(OPUS_X86_MAY_HAVE_SSE) || defined(OPUS_X86_MAY_HAVE_SSE2) || defined(OPUS_X86_MAY_HAVE_SSE4_1)
+#if defined(OPUS_X86_MAY_HAVE_SSE)
 #include "x86/pitch_sse.c"
+#endif
+#if defined(OPUS_X86_MAY_HAVE_SSE2)
+#include "x86/pitch_sse2.c"
+#endif
 #if defined(OPUS_X86_MAY_HAVE_SSE4_1)
+#include "x86/pitch_sse4_1.c"
 #include "x86/celt_lpc_sse.c"
 #endif
 #include "x86/x86_celt_map.c"
index e3bc6d7..13d739e 100644 (file)
 #include "config.h"
 #endif
 
-#include <xmmintrin.h>
-#include <emmintrin.h>
-
 #include "macros.h"
 #include "celt_lpc.h"
 #include "stack_alloc.h"
 #include "mathops.h"
 #include "pitch.h"
 
-#if defined(OPUS_X86_MAY_HAVE_SSE4_1)
-#include <smmintrin.h>
-#include "x86cpu.h"
-
-opus_val32 celt_inner_prod_sse4_1(const opus_val16 *x, const opus_val16 *y,
-      int N)
-{
-    opus_int  i, dataSize16;
-    opus_int32 sum;
-    __m128i inVec1_76543210, inVec1_FEDCBA98, acc1;
-    __m128i inVec2_76543210, inVec2_FEDCBA98, acc2;
-    __m128i inVec1_3210, inVec2_3210;
-
-    sum = 0;
-    dataSize16 = N & ~15;
-
-    acc1 = _mm_setzero_si128();
-    acc2 = _mm_setzero_si128();
-
-    for (i=0;i<dataSize16;i+=16) {
-        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
-        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
-
-        inVec1_FEDCBA98 = _mm_loadu_si128((__m128i *)(&x[i + 8]));
-        inVec2_FEDCBA98 = _mm_loadu_si128((__m128i *)(&y[i + 8]));
-
-        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
-        inVec1_FEDCBA98 = _mm_madd_epi16(inVec1_FEDCBA98, inVec2_FEDCBA98);
-
-        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
-        acc2 = _mm_add_epi32(acc2, inVec1_FEDCBA98);
-    }
-
-    acc1 = _mm_add_epi32(acc1, acc2);
-
-    if (N - i >= 8)
-    {
-        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
-        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
-
-        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
-
-        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
-        i += 8;
-    }
-
-    if (N - i >= 4)
-    {
-        inVec1_3210 = OP_CVTEPI16_EPI32_M64(&x[i + 0]);
-        inVec2_3210 = OP_CVTEPI16_EPI32_M64(&y[i + 0]);
-
-        inVec1_3210 = _mm_mullo_epi32(inVec1_3210, inVec2_3210);
-
-        acc1 = _mm_add_epi32(acc1, inVec1_3210);
-        i += 4;
-    }
-
-    acc1 = _mm_add_epi32(acc1, _mm_unpackhi_epi64(acc1, acc1));
-    acc1 = _mm_add_epi32(acc1, _mm_shufflelo_epi16(acc1, 0x0E));
-
-    sum += _mm_cvtsi128_si32(acc1);
-
-    for (;i<N;i++)
-    {
-        sum = silk_SMLABB(sum, x[i], y[i]);
-    }
-
-    return sum;
-}
-
-void xcorr_kernel_sse4_1(const opus_val16 * x, const opus_val16 * y, opus_val32 sum[ 4 ], int len)
-{
-    int j;
-
-    __m128i vecX, vecX0, vecX1, vecX2, vecX3;
-    __m128i vecY0, vecY1, vecY2, vecY3;
-    __m128i sum0, sum1, sum2, sum3, vecSum;
-    __m128i initSum;
-
-    celt_assert(len >= 3);
-
-    sum0 = _mm_setzero_si128();
-    sum1 = _mm_setzero_si128();
-    sum2 = _mm_setzero_si128();
-    sum3 = _mm_setzero_si128();
-
-    for (j=0;j<(len-7);j+=8)
-    {
-        vecX = _mm_loadu_si128((__m128i *)(&x[j + 0]));
-        vecY0 = _mm_loadu_si128((__m128i *)(&y[j + 0]));
-        vecY1 = _mm_loadu_si128((__m128i *)(&y[j + 1]));
-        vecY2 = _mm_loadu_si128((__m128i *)(&y[j + 2]));
-        vecY3 = _mm_loadu_si128((__m128i *)(&y[j + 3]));
-
-        sum0 = _mm_add_epi32(sum0, _mm_madd_epi16(vecX, vecY0));
-        sum1 = _mm_add_epi32(sum1, _mm_madd_epi16(vecX, vecY1));
-        sum2 = _mm_add_epi32(sum2, _mm_madd_epi16(vecX, vecY2));
-        sum3 = _mm_add_epi32(sum3, _mm_madd_epi16(vecX, vecY3));
-    }
-
-    sum0 = _mm_add_epi32(sum0, _mm_unpackhi_epi64( sum0, sum0));
-    sum0 = _mm_add_epi32(sum0, _mm_shufflelo_epi16( sum0, 0x0E));
-
-    sum1 = _mm_add_epi32(sum1, _mm_unpackhi_epi64( sum1, sum1));
-    sum1 = _mm_add_epi32(sum1, _mm_shufflelo_epi16( sum1, 0x0E));
-
-    sum2 = _mm_add_epi32(sum2, _mm_unpackhi_epi64( sum2, sum2));
-    sum2 = _mm_add_epi32(sum2, _mm_shufflelo_epi16( sum2, 0x0E));
-
-    sum3 = _mm_add_epi32(sum3, _mm_unpackhi_epi64( sum3, sum3));
-    sum3 = _mm_add_epi32(sum3, _mm_shufflelo_epi16( sum3, 0x0E));
-
-    vecSum = _mm_unpacklo_epi64(_mm_unpacklo_epi32(sum0, sum1),
-          _mm_unpacklo_epi32(sum2, sum3));
-
-    for (;j<(len-3);j+=4)
-    {
-        vecX = OP_CVTEPI16_EPI32_M64(&x[j + 0]);
-        vecX0 = _mm_shuffle_epi32(vecX, 0x00);
-        vecX1 = _mm_shuffle_epi32(vecX, 0x55);
-        vecX2 = _mm_shuffle_epi32(vecX, 0xaa);
-        vecX3 = _mm_shuffle_epi32(vecX, 0xff);
-
-        vecY0 = OP_CVTEPI16_EPI32_M64(&y[j + 0]);
-        vecY1 = OP_CVTEPI16_EPI32_M64(&y[j + 1]);
-        vecY2 = OP_CVTEPI16_EPI32_M64(&y[j + 2]);
-        vecY3 = OP_CVTEPI16_EPI32_M64(&y[j + 3]);
-
-        sum0 = _mm_mullo_epi32(vecX0, vecY0);
-        sum1 = _mm_mullo_epi32(vecX1, vecY1);
-        sum2 = _mm_mullo_epi32(vecX2, vecY2);
-        sum3 = _mm_mullo_epi32(vecX3, vecY3);
-
-        sum0 = _mm_add_epi32(sum0, sum1);
-        sum2 = _mm_add_epi32(sum2, sum3);
-        vecSum = _mm_add_epi32(vecSum, sum0);
-        vecSum = _mm_add_epi32(vecSum, sum2);
-    }
-
-    for (;j<len;j++)
-    {
-        vecX = OP_CVTEPI16_EPI32_M64(&x[j + 0]);
-        vecX0 = _mm_shuffle_epi32(vecX, 0x00);
-
-        vecY0 = OP_CVTEPI16_EPI32_M64(&y[j + 0]);
-
-        sum0 = _mm_mullo_epi32(vecX0, vecY0);
-        vecSum = _mm_add_epi32(vecSum, sum0);
-    }
-
-    initSum = _mm_loadu_si128((__m128i *)(&sum[0]));
-    initSum = _mm_add_epi32(initSum, vecSum);
-    _mm_storeu_si128((__m128i *)sum, initSum);
-}
-#endif
-
-#if defined(OPUS_X86_MAY_HAVE_SSE2)
-opus_val32 celt_inner_prod_sse2(const opus_val16 *x, const opus_val16 *y,
-      int N)
-{
-    opus_int  i, dataSize16;
-    opus_int32 sum;
-
-    __m128i inVec1_76543210, inVec1_FEDCBA98, acc1;
-    __m128i inVec2_76543210, inVec2_FEDCBA98, acc2;
-
-    sum = 0;
-    dataSize16 = N & ~15;
-
-    acc1 = _mm_setzero_si128();
-    acc2 = _mm_setzero_si128();
-
-    for (i=0;i<dataSize16;i+=16)
-    {
-        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
-        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
-
-        inVec1_FEDCBA98 = _mm_loadu_si128((__m128i *)(&x[i + 8]));
-        inVec2_FEDCBA98 = _mm_loadu_si128((__m128i *)(&y[i + 8]));
-
-        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
-        inVec1_FEDCBA98 = _mm_madd_epi16(inVec1_FEDCBA98, inVec2_FEDCBA98);
-
-        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
-        acc2 = _mm_add_epi32(acc2, inVec1_FEDCBA98);
-    }
-
-    acc1 = _mm_add_epi32( acc1, acc2 );
-
-    if (N - i >= 8)
-    {
-        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
-        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
-
-        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
-
-        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
-        i += 8;
-    }
-
-    acc1 = _mm_add_epi32(acc1, _mm_unpackhi_epi64( acc1, acc1));
-    acc1 = _mm_add_epi32(acc1, _mm_shufflelo_epi16( acc1, 0x0E));
-    sum += _mm_cvtsi128_si32(acc1);
-
-    for (;i<N;i++) {
-        sum = silk_SMLABB(sum, x[i], y[i]);
-    }
-
-    return sum;
-}
-#endif
diff --git a/celt/x86/pitch_sse2.c b/celt/x86/pitch_sse2.c
new file mode 100644 (file)
index 0000000..a0e7d1b
--- /dev/null
@@ -0,0 +1,95 @@
+/* Copyright (c) 2014, Cisco Systems, INC
+   Written by XiangMingZhu WeiZhou MinPeng YanWang
+
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - 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.
+
+   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 COPYRIGHT OWNER
+   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.
+*/
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include <xmmintrin.h>
+#include <emmintrin.h>
+
+#include "macros.h"
+#include "celt_lpc.h"
+#include "stack_alloc.h"
+#include "mathops.h"
+#include "pitch.h"
+
+#if defined(OPUS_X86_MAY_HAVE_SSE2) && defined(FIXED_POINT)
+opus_val32 celt_inner_prod_sse2(const opus_val16 *x, const opus_val16 *y,
+      int N)
+{
+    opus_int  i, dataSize16;
+    opus_int32 sum;
+
+    __m128i inVec1_76543210, inVec1_FEDCBA98, acc1;
+    __m128i inVec2_76543210, inVec2_FEDCBA98, acc2;
+
+    sum = 0;
+    dataSize16 = N & ~15;
+
+    acc1 = _mm_setzero_si128();
+    acc2 = _mm_setzero_si128();
+
+    for (i=0;i<dataSize16;i+=16)
+    {
+        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
+        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
+
+        inVec1_FEDCBA98 = _mm_loadu_si128((__m128i *)(&x[i + 8]));
+        inVec2_FEDCBA98 = _mm_loadu_si128((__m128i *)(&y[i + 8]));
+
+        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
+        inVec1_FEDCBA98 = _mm_madd_epi16(inVec1_FEDCBA98, inVec2_FEDCBA98);
+
+        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
+        acc2 = _mm_add_epi32(acc2, inVec1_FEDCBA98);
+    }
+
+    acc1 = _mm_add_epi32( acc1, acc2 );
+
+    if (N - i >= 8)
+    {
+        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
+        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
+
+        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
+
+        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
+        i += 8;
+    }
+
+    acc1 = _mm_add_epi32(acc1, _mm_unpackhi_epi64( acc1, acc1));
+    acc1 = _mm_add_epi32(acc1, _mm_shufflelo_epi16( acc1, 0x0E));
+    sum += _mm_cvtsi128_si32(acc1);
+
+    for (;i<N;i++) {
+        sum = silk_SMLABB(sum, x[i], y[i]);
+    }
+
+    return sum;
+}
+#endif
diff --git a/celt/x86/pitch_sse4_1.c b/celt/x86/pitch_sse4_1.c
new file mode 100644 (file)
index 0000000..a092c68
--- /dev/null
@@ -0,0 +1,195 @@
+/* Copyright (c) 2014, Cisco Systems, INC
+   Written by XiangMingZhu WeiZhou MinPeng YanWang
+
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - 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.
+
+   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 COPYRIGHT OWNER
+   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.
+*/
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include <xmmintrin.h>
+#include <emmintrin.h>
+
+#include "macros.h"
+#include "celt_lpc.h"
+#include "stack_alloc.h"
+#include "mathops.h"
+#include "pitch.h"
+
+#if defined(OPUS_X86_MAY_HAVE_SSE4_1) && defined(FIXED_POINT)
+#include <smmintrin.h>
+#include "x86cpu.h"
+
+opus_val32 celt_inner_prod_sse4_1(const opus_val16 *x, const opus_val16 *y,
+      int N)
+{
+    opus_int  i, dataSize16;
+    opus_int32 sum;
+    __m128i inVec1_76543210, inVec1_FEDCBA98, acc1;
+    __m128i inVec2_76543210, inVec2_FEDCBA98, acc2;
+    __m128i inVec1_3210, inVec2_3210;
+
+    sum = 0;
+    dataSize16 = N & ~15;
+
+    acc1 = _mm_setzero_si128();
+    acc2 = _mm_setzero_si128();
+
+    for (i=0;i<dataSize16;i+=16) {
+        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
+        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
+
+        inVec1_FEDCBA98 = _mm_loadu_si128((__m128i *)(&x[i + 8]));
+        inVec2_FEDCBA98 = _mm_loadu_si128((__m128i *)(&y[i + 8]));
+
+        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
+        inVec1_FEDCBA98 = _mm_madd_epi16(inVec1_FEDCBA98, inVec2_FEDCBA98);
+
+        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
+        acc2 = _mm_add_epi32(acc2, inVec1_FEDCBA98);
+    }
+
+    acc1 = _mm_add_epi32(acc1, acc2);
+
+    if (N - i >= 8)
+    {
+        inVec1_76543210 = _mm_loadu_si128((__m128i *)(&x[i + 0]));
+        inVec2_76543210 = _mm_loadu_si128((__m128i *)(&y[i + 0]));
+
+        inVec1_76543210 = _mm_madd_epi16(inVec1_76543210, inVec2_76543210);
+
+        acc1 = _mm_add_epi32(acc1, inVec1_76543210);
+        i += 8;
+    }
+
+    if (N - i >= 4)
+    {
+        inVec1_3210 = OP_CVTEPI16_EPI32_M64(&x[i + 0]);
+        inVec2_3210 = OP_CVTEPI16_EPI32_M64(&y[i + 0]);
+
+        inVec1_3210 = _mm_mullo_epi32(inVec1_3210, inVec2_3210);
+
+        acc1 = _mm_add_epi32(acc1, inVec1_3210);
+        i += 4;
+    }
+
+    acc1 = _mm_add_epi32(acc1, _mm_unpackhi_epi64(acc1, acc1));
+    acc1 = _mm_add_epi32(acc1, _mm_shufflelo_epi16(acc1, 0x0E));
+
+    sum += _mm_cvtsi128_si32(acc1);
+
+    for (;i<N;i++)
+    {
+        sum = silk_SMLABB(sum, x[i], y[i]);
+    }
+
+    return sum;
+}
+
+void xcorr_kernel_sse4_1(const opus_val16 * x, const opus_val16 * y, opus_val32 sum[ 4 ], int len)
+{
+    int j;
+
+    __m128i vecX, vecX0, vecX1, vecX2, vecX3;
+    __m128i vecY0, vecY1, vecY2, vecY3;
+    __m128i sum0, sum1, sum2, sum3, vecSum;
+    __m128i initSum;
+
+    celt_assert(len >= 3);
+
+    sum0 = _mm_setzero_si128();
+    sum1 = _mm_setzero_si128();
+    sum2 = _mm_setzero_si128();
+    sum3 = _mm_setzero_si128();
+
+    for (j=0;j<(len-7);j+=8)
+    {
+        vecX = _mm_loadu_si128((__m128i *)(&x[j + 0]));
+        vecY0 = _mm_loadu_si128((__m128i *)(&y[j + 0]));
+        vecY1 = _mm_loadu_si128((__m128i *)(&y[j + 1]));
+        vecY2 = _mm_loadu_si128((__m128i *)(&y[j + 2]));
+        vecY3 = _mm_loadu_si128((__m128i *)(&y[j + 3]));
+
+        sum0 = _mm_add_epi32(sum0, _mm_madd_epi16(vecX, vecY0));
+        sum1 = _mm_add_epi32(sum1, _mm_madd_epi16(vecX, vecY1));
+        sum2 = _mm_add_epi32(sum2, _mm_madd_epi16(vecX, vecY2));
+        sum3 = _mm_add_epi32(sum3, _mm_madd_epi16(vecX, vecY3));
+    }
+
+    sum0 = _mm_add_epi32(sum0, _mm_unpackhi_epi64( sum0, sum0));
+    sum0 = _mm_add_epi32(sum0, _mm_shufflelo_epi16( sum0, 0x0E));
+
+    sum1 = _mm_add_epi32(sum1, _mm_unpackhi_epi64( sum1, sum1));
+    sum1 = _mm_add_epi32(sum1, _mm_shufflelo_epi16( sum1, 0x0E));
+
+    sum2 = _mm_add_epi32(sum2, _mm_unpackhi_epi64( sum2, sum2));
+    sum2 = _mm_add_epi32(sum2, _mm_shufflelo_epi16( sum2, 0x0E));
+
+    sum3 = _mm_add_epi32(sum3, _mm_unpackhi_epi64( sum3, sum3));
+    sum3 = _mm_add_epi32(sum3, _mm_shufflelo_epi16( sum3, 0x0E));
+
+    vecSum = _mm_unpacklo_epi64(_mm_unpacklo_epi32(sum0, sum1),
+          _mm_unpacklo_epi32(sum2, sum3));
+
+    for (;j<(len-3);j+=4)
+    {
+        vecX = OP_CVTEPI16_EPI32_M64(&x[j + 0]);
+        vecX0 = _mm_shuffle_epi32(vecX, 0x00);
+        vecX1 = _mm_shuffle_epi32(vecX, 0x55);
+        vecX2 = _mm_shuffle_epi32(vecX, 0xaa);
+        vecX3 = _mm_shuffle_epi32(vecX, 0xff);
+
+        vecY0 = OP_CVTEPI16_EPI32_M64(&y[j + 0]);
+        vecY1 = OP_CVTEPI16_EPI32_M64(&y[j + 1]);
+        vecY2 = OP_CVTEPI16_EPI32_M64(&y[j + 2]);
+        vecY3 = OP_CVTEPI16_EPI32_M64(&y[j + 3]);
+
+        sum0 = _mm_mullo_epi32(vecX0, vecY0);
+        sum1 = _mm_mullo_epi32(vecX1, vecY1);
+        sum2 = _mm_mullo_epi32(vecX2, vecY2);
+        sum3 = _mm_mullo_epi32(vecX3, vecY3);
+
+        sum0 = _mm_add_epi32(sum0, sum1);
+        sum2 = _mm_add_epi32(sum2, sum3);
+        vecSum = _mm_add_epi32(vecSum, sum0);
+        vecSum = _mm_add_epi32(vecSum, sum2);
+    }
+
+    for (;j<len;j++)
+    {
+        vecX = OP_CVTEPI16_EPI32_M64(&x[j + 0]);
+        vecX0 = _mm_shuffle_epi32(vecX, 0x00);
+
+        vecY0 = OP_CVTEPI16_EPI32_M64(&y[j + 0]);
+
+        sum0 = _mm_mullo_epi32(vecX0, vecY0);
+        vecSum = _mm_add_epi32(vecSum, sum0);
+    }
+
+    initSum = _mm_loadu_si128((__m128i *)(&sum[0]));
+    initSum = _mm_add_epi32(initSum, vecSum);
+    _mm_storeu_si128((__m128i *)sum, initSum);
+}
+#endif
index 29ec937..c92693f 100644 (file)
@@ -21,7 +21,10 @@ CELT_SOURCES_SSE = celt/x86/x86cpu.c \
 celt/x86/x86_celt_map.c \
 celt/x86/pitch_sse.c
 
-CELT_SOURCES_SSE4_1 = celt/x86/celt_lpc_sse.c
+CELT_SOURCES_SSE2 = celt/x86/pitch_sse2.c
+
+CELT_SOURCES_SSE4_1 = celt/x86/celt_lpc_sse.c \
+celt/x86/pitch_sse4_1.c
 
 CELT_SOURCES_ARM = \
 celt/arm/armcpu.c \
index c19a6a7..ea91ab2 100644 (file)
@@ -348,9 +348,11 @@ AM_CONDITIONAL([OPUS_ARM_INLINE_ASM],
 AM_CONDITIONAL([OPUS_ARM_EXTERNAL_ASM],
     [test x"${asm_optimization%% *}" = x"ARM"])
 
+AM_CONDITIONAL([HAVE_SSE], [false])
 AM_CONDITIONAL([HAVE_SSE2], [false])
 AM_CONDITIONAL([HAVE_SSE4_1], [false])
 
+m4_define([DEFAULT_X86_SSE_CFLAGS], [-msse])
 m4_define([DEFAULT_X86_SSE2_CFLAGS], [-msse2])
 m4_define([DEFAULT_X86_SSE4_1_CFLAGS], [-msse4.1])
 m4_define([DEFAULT_ARM_NEON_INTR_CFLAGS], [-mfpu=neon])
@@ -366,10 +368,12 @@ AS_CASE([$host],
        [arm*eabi*], [AS_VAR_SET([RESOLVED_DEFAULT_ARM_NEON_INTR_CFLAGS], "DEFAULT_ARM_NEON_SOFTFP_INTR_CFLAGS")],
        [AS_VAR_SET([RESOLVED_DEFAULT_ARM_NEON_INTR_CFLAGS], "DEFAULT_ARM_NEON_INTR_CFLAGS")])
 
+AC_ARG_VAR([X86_SSE_CFLAGS], [C compiler flags to compile SSE intrinsics @<:@default=]DEFAULT_X86_SSE_CFLAGS[@:>@])
 AC_ARG_VAR([X86_SSE2_CFLAGS], [C compiler flags to compile SSE2 intrinsics @<:@default=]DEFAULT_X86_SSE2_CFLAGS[@:>@])
 AC_ARG_VAR([X86_SSE4_1_CFLAGS], [C compiler flags to compile SSE4.1 intrinsics @<:@default=]DEFAULT_X86_SSE4_1_CFLAGS[@:>@])
 AC_ARG_VAR([ARM_NEON_INTR_CFLAGS], [C compiler flags to compile ARM NEON intrinsics @<:@default=]DEFAULT_ARM_NEON_INTR_CFLAGS / DEFAULT_ARM_NEON_SOFTFP_INTR_CFLAGS[@:>@])
 
+AS_VAR_SET_IF([X86_SSE_CFLAGS], [], [AS_VAR_SET([X86_SSE_CFLAGS], "DEFAULT_X86_SSE_CFLAGS")])
 AS_VAR_SET_IF([X86_SSE2_CFLAGS], [], [AS_VAR_SET([X86_SSE2_CFLAGS], "DEFAULT_X86_SSE2_CFLAGS")])
 AS_VAR_SET_IF([X86_SSE4_1_CFLAGS], [], [AS_VAR_SET([X86_SSE4_1_CFLAGS], "DEFAULT_X86_SSE4_1_CFLAGS")])
 AS_VAR_SET_IF([ARM_NEON_INTR_CFLAGS], [], [AS_VAR_SET([ARM_NEON_INTR_CFLAGS], ["$RESOLVED_DEFAULT_ARM_NEON_INTR_CFLAGS"])])
@@ -432,6 +436,24 @@ AS_IF([test x"$enable_intrinsics" = x"yes"],[
    [i?86|x86_64],
    [
       OPUS_CHECK_INTRINSICS(
+         [SSE],
+         [$X86_SSE_CFLAGS],
+         [OPUS_X86_MAY_HAVE_SSE],
+         [OPUS_X86_PRESUME_SSE],
+         [[#include <xmmintrin.h>
+         ]],
+         [[
+             static __m128 mtest;
+             mtest = _mm_setzero_ps();
+         ]]
+      )
+      AS_IF([test x"$OPUS_X86_MAY_HAVE_SSE" = x"1" && test x"$OPUS_X86_PRESUME_SSE" != x"1"],
+          [
+             OPUS_X86_SSE_CFLAGS="$X86_SSE_CFLAGS"
+             AC_SUBST([OPUS_X86_SSE_CFLAGS])
+          ]
+      )
+      OPUS_CHECK_INTRINSICS(
          [SSE2],
          [$X86_SSE2_CFLAGS],
          [OPUS_X86_MAY_HAVE_SSE2],
@@ -473,6 +495,19 @@ AS_IF([test x"$enable_intrinsics" = x"yes"],[
       AS_IF([test x"$enable_float" = x"no"],
       [
          AS_IF([test x"$rtcd_support" = x"no"], [rtcd_support=""])
+         AS_IF([test x"$OPUS_X86_MAY_HAVE_SSE" = x"1"],
+         [
+            AC_DEFINE([OPUS_X86_MAY_HAVE_SSE], 1, [Compiler supports X86 SSE Intrinsics])
+            intrinsics_support="$intrinsics_support SSE"
+
+            AS_IF([test x"$OPUS_X86_PRESUME_SSE" = x"1"],
+               [AC_DEFINE([OPUS_X86_PRESUME_SSE], 1, [Define if binary requires SSE intrinsics support])],
+               [rtcd_support="$rtcd_support SSE"])
+         ],
+         [
+            AC_MSG_WARN([Compiler does not support SSE intrinsics])
+         ])
+
          AS_IF([test x"$OPUS_X86_MAY_HAVE_SSE2" = x"1"],
          [
             AC_DEFINE([OPUS_X86_MAY_HAVE_SSE2], 1, [Compiler supports X86 SSE2 Intrinsics])
@@ -561,6 +596,8 @@ AS_IF([test x"$enable_intrinsics" = x"yes"],[
 AM_CONDITIONAL([CPU_ARM], [test "$cpu_arm" = "yes"])
 AM_CONDITIONAL([OPUS_ARM_NEON_INTR],
     [test x"$OPUS_ARM_MAY_HAVE_NEON_INTR" = x"1"])
+AM_CONDITIONAL([HAVE_SSE],
+    [test x"$OPUS_X86_MAY_HAVE_SSE" = x"1"])
 AM_CONDITIONAL([HAVE_SSE2],
     [test x"$OPUS_X86_MAY_HAVE_SSE2" = x"1"])
 AM_CONDITIONAL([HAVE_SSE4_1],