avcodec/mips: MSA (MIPS-SIMD-Arch) optimizations for idctdsp functions

This patch adds MSA (MIPS-SIMD-Arch) optimizations for idctdsp functions in new file idctdsp_msa.c and simple_idct_msa.c
Signed-off-by: Shivraj Patil <shivraj.patil@imgtec.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

libavcodec/idctdsp.c

@@ -305,6 +305,8 @@ av_cold void ff_idctdsp_init(IDCTDSPContext *c, AVCodecContext *avctx)
         ff_idctdsp_init_ppc(c, avctx, high_bit_depth);
     if (ARCH_X86)
         ff_idctdsp_init_x86(c, avctx, high_bit_depth);
+    if (ARCH_MIPS)
+        ff_idctdsp_init_mips(c, avctx, high_bit_depth);
 
     ff_put_pixels_clamped = c->put_pixels_clamped;
     ff_add_pixels_clamped = c->add_pixels_clamped;

libavcodec/idctdsp.h

@@ -108,5 +108,7 @@ void ff_idctdsp_init_ppc(IDCTDSPContext *c, AVCodecContext *avctx,
                          unsigned high_bit_depth);
 void ff_idctdsp_init_x86(IDCTDSPContext *c, AVCodecContext *avctx,
                          unsigned high_bit_depth);
+void ff_idctdsp_init_mips(IDCTDSPContext *c, AVCodecContext *avctx,
+                          unsigned high_bit_depth);
 
 #endif /* AVCODEC_IDCTDSP_H */

libavcodec/mips/Makefile

@@ -29,6 +29,7 @@ OBJS-$(CONFIG_QPELDSP)                    += mips/qpeldsp_init_mips.o
 OBJS-$(CONFIG_HPELDSP)                    += mips/hpeldsp_init_mips.o
 OBJS-$(CONFIG_BLOCKDSP)                   += mips/blockdsp_init_mips.o
 OBJS-$(CONFIG_PIXBLOCKDSP)                += mips/pixblockdsp_init_mips.o
+OBJS-$(CONFIG_IDCTDSP)                    += mips/idctdsp_init_mips.o
 OBJS-$(CONFIG_MPEGVIDEO)                  += mips/mpegvideo_init_mips.o
 OBJS-$(CONFIG_MPEGVIDEOENC)               += mips/mpegvideoencdsp_init_mips.o
 OBJS-$(CONFIG_ME_CMP)                     += mips/me_cmp_init_mips.o
@@ -50,6 +51,8 @@ MSA-OBJS-$(CONFIG_QPELDSP)                += mips/qpeldsp_msa.o
 MSA-OBJS-$(CONFIG_HPELDSP)                += mips/hpeldsp_msa.o
 MSA-OBJS-$(CONFIG_BLOCKDSP)               += mips/blockdsp_msa.o
 MSA-OBJS-$(CONFIG_PIXBLOCKDSP)            += mips/pixblockdsp_msa.o
+MSA-OBJS-$(CONFIG_IDCTDSP)                += mips/idctdsp_msa.o           \
+                                             mips/simple_idct_msa.o
 MSA-OBJS-$(CONFIG_MPEGVIDEO)              += mips/mpegvideo_msa.o
 MSA-OBJS-$(CONFIG_MPEGVIDEOENC)           += mips/mpegvideoencdsp_msa.o
 MSA-OBJS-$(CONFIG_ME_CMP)                 += mips/me_cmp_msa.o

libavcodec/mips/idctdsp_init_mips.c

+/*
+ * Copyright (c) 2015 Manojkumar Bhosale (Manojkumar.Bhosale@imgtec.com)
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "idctdsp_mips.h"
+
+#if HAVE_MSA
+static av_cold void idctdsp_init_msa(IDCTDSPContext *c, AVCodecContext *avctx,
+                                     unsigned high_bit_depth)
+{
+    if ((avctx->lowres != 1) && (avctx->lowres != 2) && (avctx->lowres != 3) &&
+        (avctx->bits_per_raw_sample != 10) &&
+        (avctx->bits_per_raw_sample != 12) &&
+        (avctx->idct_algo == FF_IDCT_AUTO)) {
+                c->idct_put = ff_simple_idct_put_msa;
+                c->idct_add = ff_simple_idct_add_msa;
+                c->idct = ff_simple_idct_msa;
+                c->perm_type = FF_IDCT_PERM_NONE;
+    }
+
+    c->put_pixels_clamped = ff_put_pixels_clamped_msa;
+    c->put_signed_pixels_clamped = ff_put_signed_pixels_clamped_msa;
+    c->add_pixels_clamped = ff_add_pixels_clamped_msa;
+}
+#endif  // #if HAVE_MSA
+
+void ff_idctdsp_init_mips(IDCTDSPContext *c, AVCodecContext *avctx,
+                          unsigned high_bit_depth)
+{
+#if HAVE_MSA
+    idctdsp_init_msa(c, avctx, high_bit_depth);
+#endif  // #if HAVE_MSA
+}

libavcodec/mips/idctdsp_mips.h

+/*
+ * Copyright (c) 2015 Manojkumar Bhosale (Manojkumar.Bhosale@imgtec.com)
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#ifndef AVCODEC_MIPS_IDCTDSP_MIPS_H
+#define AVCODEC_MIPS_IDCTDSP_MIPS_H
+
+#include "../mpegvideo.h"
+
+void ff_put_pixels_clamped_msa(const int16_t *block,
+                               uint8_t *av_restrict pixels,
+                               ptrdiff_t line_size);
+void ff_put_signed_pixels_clamped_msa(const int16_t *block,
+                                      uint8_t *av_restrict pixels,
+                                      ptrdiff_t line_size);
+void ff_add_pixels_clamped_msa(const int16_t *block,
+                               uint8_t *av_restrict pixels,
+                               ptrdiff_t line_size);
+void ff_j_rev_dct_msa(int16_t *data);
+void ff_jref_idct_put_msa(uint8_t *dest, int32_t stride, int16_t *block);
+void ff_jref_idct_add_msa(uint8_t *dest, int32_t stride, int16_t *block);
+void ff_simple_idct_msa(int16_t *block);
+void ff_simple_idct_put_msa(uint8_t *dest, int32_t stride_dst, int16_t *block);
+void ff_simple_idct_add_msa(uint8_t *dest, int32_t stride_dst, int16_t *block);
+
+#endif  // #ifndef AVCODEC_MIPS_IDCTDSP_MIPS_H

libavcodec/mips/idctdsp_msa.c

+/*
+ * Copyright (c) 2015 Manojkumar Bhosale (Manojkumar.Bhosale@imgtec.com)
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/mips/generic_macros_msa.h"
+#include "idctdsp_mips.h"
+
+static void put_pixels_clamped_msa(const int16_t *block, uint8_t *pixels,
+                                   int32_t stride)
+{
+    uint64_t in0_d, in1_d, in2_d, in3_d, in4_d, in5_d, in6_d, in7_d;
+    v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
+
+    LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7);
+    CLIP_SH4_0_255(in0, in1, in2, in3);
+    CLIP_SH4_0_255(in4, in5, in6, in7);
+    PCKEV_B4_SH(in0, in0, in1, in1, in2, in2, in3, in3, in0, in1, in2, in3);
+    PCKEV_B4_SH(in4, in4, in5, in5, in6, in6, in7, in7, in4, in5, in6, in7);
+
+    in0_d = __msa_copy_u_d((v2i64) in0, 0);
+    in1_d = __msa_copy_u_d((v2i64) in1, 0);
+    in2_d = __msa_copy_u_d((v2i64) in2, 0);
+    in3_d = __msa_copy_u_d((v2i64) in3, 0);
+    in4_d = __msa_copy_u_d((v2i64) in4, 0);
+    in5_d = __msa_copy_u_d((v2i64) in5, 0);
+    in6_d = __msa_copy_u_d((v2i64) in6, 0);
+    in7_d = __msa_copy_u_d((v2i64) in7, 0);
+    SD4(in0_d, in1_d, in2_d, in3_d, pixels, stride);
+    pixels += 4 * stride;
+    SD4(in4_d, in5_d, in6_d, in7_d, pixels, stride);
+}
+
+static void put_signed_pixels_clamped_msa(const int16_t *block, uint8_t *pixels,
+                                          int32_t stride)
+{
+    uint64_t in0_d, in1_d, in2_d, in3_d, in4_d, in5_d, in6_d, in7_d;
+    v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
+
+    LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7);
+
+    in0 += 128;
+    in1 += 128;
+    in2 += 128;
+    in3 += 128;
+    in4 += 128;
+    in5 += 128;
+    in6 += 128;
+    in7 += 128;
+
+    CLIP_SH4_0_255(in0, in1, in2, in3);
+    CLIP_SH4_0_255(in4, in5, in6, in7);
+    PCKEV_B4_SH(in0, in0, in1, in1, in2, in2, in3, in3, in0, in1, in2, in3);
+    PCKEV_B4_SH(in4, in4, in5, in5, in6, in6, in7, in7, in4, in5, in6, in7);
+
+    in0_d = __msa_copy_u_d((v2i64) in0, 0);
+    in1_d = __msa_copy_u_d((v2i64) in1, 0);
+    in2_d = __msa_copy_u_d((v2i64) in2, 0);
+    in3_d = __msa_copy_u_d((v2i64) in3, 0);
+    in4_d = __msa_copy_u_d((v2i64) in4, 0);
+    in5_d = __msa_copy_u_d((v2i64) in5, 0);
+    in6_d = __msa_copy_u_d((v2i64) in6, 0);
+    in7_d = __msa_copy_u_d((v2i64) in7, 0);
+    SD4(in0_d, in1_d, in2_d, in3_d, pixels, stride);
+    pixels += 4 * stride;
+    SD4(in4_d, in5_d, in6_d, in7_d, pixels, stride);
+}
+
+static void add_pixels_clamped_msa(const int16_t *block, uint8_t *pixels,
+                                   int32_t stride)
+{
+    uint64_t in0_d, in1_d, in2_d, in3_d, in4_d, in5_d, in6_d, in7_d;
+    v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
+    v16u8 pix_in0, pix_in1, pix_in2, pix_in3;
+    v16u8 pix_in4, pix_in5, pix_in6, pix_in7;
+    v8u16 pix0, pix1, pix2, pix3, pix4, pix5, pix6, pix7;
+    v8i16 zero = { 0 };
+
+    LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7);
+    LD_UB8(pixels, stride, pix_in0, pix_in1, pix_in2,
+           pix_in3, pix_in4, pix_in5, pix_in6, pix_in7);
+
+    ILVR_B4_UH(zero, pix_in0, zero, pix_in1, zero, pix_in2, zero, pix_in3,
+               pix0, pix1, pix2, pix3);
+    ILVR_B4_UH(zero, pix_in4, zero, pix_in5, zero, pix_in6, zero, pix_in7,
+               pix4, pix5, pix6, pix7);
+
+    in0 += (v8i16) pix0;
+    in1 += (v8i16) pix1;
+    in2 += (v8i16) pix2;
+    in3 += (v8i16) pix3;
+    in4 += (v8i16) pix4;
+    in5 += (v8i16) pix5;
+    in6 += (v8i16) pix6;
+    in7 += (v8i16) pix7;
+
+    CLIP_SH4_0_255(in0, in1, in2, in3);
+    CLIP_SH4_0_255(in4, in5, in6, in7);
+    PCKEV_B4_SH(in0, in0, in1, in1, in2, in2, in3, in3, in0, in1, in2, in3);
+    PCKEV_B4_SH(in4, in4, in5, in5, in6, in6, in7, in7, in4, in5, in6, in7);
+
+    in0_d = __msa_copy_u_d((v2i64) in0, 0);
+    in1_d = __msa_copy_u_d((v2i64) in1, 0);
+    in2_d = __msa_copy_u_d((v2i64) in2, 0);
+    in3_d = __msa_copy_u_d((v2i64) in3, 0);
+    in4_d = __msa_copy_u_d((v2i64) in4, 0);
+    in5_d = __msa_copy_u_d((v2i64) in5, 0);
+    in6_d = __msa_copy_u_d((v2i64) in6, 0);
+    in7_d = __msa_copy_u_d((v2i64) in7, 0);
+    SD4(in0_d, in1_d, in2_d, in3_d, pixels, stride);
+    pixels += 4 * stride;
+    SD4(in4_d, in5_d, in6_d, in7_d, pixels, stride);
+}
+
+void ff_put_pixels_clamped_msa(const int16_t *block,
+                               uint8_t *av_restrict pixels,
+                               ptrdiff_t line_size)
+{
+    put_pixels_clamped_msa(block, pixels, line_size);
+}
+
+void ff_put_signed_pixels_clamped_msa(const int16_t *block,
+                                      uint8_t *av_restrict pixels,
+                                      ptrdiff_t line_size)
+{
+    put_signed_pixels_clamped_msa(block, pixels, line_size);
+}
+
+void ff_add_pixels_clamped_msa(const int16_t *block,
+                               uint8_t *av_restrict pixels,
+                               ptrdiff_t line_size)
+{
+    add_pixels_clamped_msa(block, pixels, line_size);
+}

libavutil/mips/generic_macros_msa.h

@@ -507,6 +507,14 @@
     ST_SW(in0, (pdst));                 \
     ST_SW(in1, (pdst) + stride);        \
 }
+#define ST_SW8(in0, in1, in2, in3, in4, in5, in6, in7,  \
+               pdst, stride)                            \
+{                                                       \
+    ST_SW2(in0, in1, (pdst), stride);                   \
+    ST_SW2(in2, in3, (pdst) + 2 * stride, stride);      \
+    ST_SW2(in4, in5, (pdst) + 4 * stride, stride);      \
+    ST_SW2(in6, in7, (pdst) + 6 * stride, stride);      \
+}
 
 /* Description : Store as 2x4 byte block to destination memory from input vector
    Arguments   : Inputs  - in, stidx, pdst, stride
@@ -2382,6 +2390,35 @@
     out7 = in0 - in7;                                                \
 }
 
+/* Description : Butterfly of 16 input vectors
+   Arguments   : Inputs  - in0 ...  in15
+                 Outputs - out0 .. out15
+   Details     : Butterfly operation
+*/
+#define BUTTERFLY_16(in0, in1, in2, in3, in4, in5, in6, in7,                \
+                     in8, in9,  in10, in11, in12, in13, in14, in15,         \
+                     out0, out1, out2, out3, out4, out5, out6, out7,        \
+                     out8, out9, out10, out11, out12, out13, out14, out15)  \
+{                                                                           \
+    out0 = in0 + in15;                                                      \
+    out1 = in1 + in14;                                                      \
+    out2 = in2 + in13;                                                      \
+    out3 = in3 + in12;                                                      \
+    out4 = in4 + in11;                                                      \
+    out5 = in5 + in10;                                                      \
+    out6 = in6 + in9;                                                       \
+    out7 = in7 + in8;                                                       \
+                                                                            \
+    out8 = in7 - in8;                                                       \
+    out9 = in6 - in9;                                                       \
+    out10 = in5 - in10;                                                     \
+    out11 = in4 - in11;                                                     \
+    out12 = in3 - in12;                                                     \
+    out13 = in2 - in13;                                                     \
+    out14 = in1 - in14;                                                     \
+    out15 = in0 - in15;                                                     \
+}
+
 /* Description : Transposes input 4x4 byte block
    Arguments   : Inputs  - in0, in1, in2, in3      (input 4x4 byte block)
                  Outputs - out0, out1, out2, out3  (output 4x4 byte block)