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

This patch adds MSA (MIPS-SIMD-Arch) optimizations for hpel functions in new file hpeldsp_msa.c
Adds new generic macros (needed for this patch) in libavutil/mips/generic_macros_msa.h
Signed-off-by: Shivraj Patil <shivraj.patil@imgtec.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

libavcodec/hpeldsp.c

@@ -365,4 +365,6 @@ av_cold void ff_hpeldsp_init(HpelDSPContext *c, int flags)
         ff_hpeldsp_init_ppc(c, flags);
     if (ARCH_X86)
         ff_hpeldsp_init_x86(c, flags);
+    if (ARCH_MIPS)
+        ff_hpeldsp_init_mips(c, flags);
 }

libavcodec/hpeldsp.h

@@ -99,5 +99,6 @@ void ff_hpeldsp_init_alpha(HpelDSPContext *c, int flags);
 void ff_hpeldsp_init_arm(HpelDSPContext *c, int flags);
 void ff_hpeldsp_init_ppc(HpelDSPContext *c, int flags);
 void ff_hpeldsp_init_x86(HpelDSPContext *c, int flags);
+void ff_hpeldsp_init_mips(HpelDSPContext *c, int flags);
 
 #endif /* AVCODEC_HPELDSP_H */

libavcodec/mips/Makefile

@@ -26,6 +26,7 @@ OBJS-$(CONFIG_H264CHROMA)                 += mips/h264chroma_init_mips.o
 OBJS-$(CONFIG_H264PRED)                   += mips/h264pred_init_mips.o
 OBJS-$(CONFIG_H263DSP)                    += mips/h263dsp_init_mips.o
 OBJS-$(CONFIG_QPELDSP)                    += mips/qpeldsp_init_mips.o
+OBJS-$(CONFIG_HPELDSP)                    += mips/hpeldsp_init_mips.o
 MSA-OBJS-$(CONFIG_HEVC_DECODER)           += mips/hevcdsp_msa.o            \
                                              mips/hevc_mc_uni_msa.o        \
                                              mips/hevc_mc_uniw_msa.o       \
@@ -41,5 +42,6 @@ MSA-OBJS-$(CONFIG_H264CHROMA)             += mips/h264chroma_msa.o
 MSA-OBJS-$(CONFIG_H264PRED)               += mips/h264pred_msa.o
 MSA-OBJS-$(CONFIG_H263DSP)                += mips/h263dsp_msa.o
 MSA-OBJS-$(CONFIG_QPELDSP)                += mips/qpeldsp_msa.o
+MSA-OBJS-$(CONFIG_HPELDSP)                += mips/hpeldsp_msa.o
 LOONGSON3-OBJS-$(CONFIG_H264DSP)          += mips/h264dsp_mmi.o
 LOONGSON3-OBJS-$(CONFIG_H264CHROMA)       += mips/h264chroma_mmi.o

libavcodec/mips/hpeldsp_init_mips.c

+/*
+ * Copyright (c) 2015 Parag Salasakar (Parag.Salasakar@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 "../hpeldsp.h"
+#include "libavcodec/mips/hpeldsp_mips.h"
+
+#if HAVE_MSA
+static void ff_hpeldsp_init_msa(HpelDSPContext *c, int flags)
+{
+    c->put_pixels_tab[0][0] = ff_put_pixels16_msa;
+    c->put_pixels_tab[0][1] = ff_put_pixels16_x2_msa;
+    c->put_pixels_tab[0][2] = ff_put_pixels16_y2_msa;
+    c->put_pixels_tab[0][3] = ff_put_pixels16_xy2_msa;
+
+    c->put_pixels_tab[1][0] = ff_put_pixels8_msa;
+    c->put_pixels_tab[1][1] = ff_put_pixels8_x2_msa;
+    c->put_pixels_tab[1][2] = ff_put_pixels8_y2_msa;
+    c->put_pixels_tab[1][3] = ff_put_pixels8_xy2_msa;
+
+    c->put_pixels_tab[2][1] = ff_put_pixels4_x2_msa;
+    c->put_pixels_tab[2][2] = ff_put_pixels4_y2_msa;
+    c->put_pixels_tab[2][3] = ff_put_pixels4_xy2_msa;
+
+    c->put_no_rnd_pixels_tab[0][0] = ff_put_pixels16_msa;
+    c->put_no_rnd_pixels_tab[0][1] = ff_put_no_rnd_pixels16_x2_msa;
+    c->put_no_rnd_pixels_tab[0][2] = ff_put_no_rnd_pixels16_y2_msa;
+    c->put_no_rnd_pixels_tab[0][3] = ff_put_no_rnd_pixels16_xy2_msa;
+
+    c->put_no_rnd_pixels_tab[1][0] = ff_put_pixels8_msa;
+    c->put_no_rnd_pixels_tab[1][1] = ff_put_no_rnd_pixels8_x2_msa;
+    c->put_no_rnd_pixels_tab[1][2] = ff_put_no_rnd_pixels8_y2_msa;
+    c->put_no_rnd_pixels_tab[1][3] = ff_put_no_rnd_pixels8_xy2_msa;
+
+    c->avg_pixels_tab[0][0] = ff_avg_pixels16_msa;
+    c->avg_pixels_tab[0][1] = ff_avg_pixels16_x2_msa;
+    c->avg_pixels_tab[0][2] = ff_avg_pixels16_y2_msa;
+    c->avg_pixels_tab[0][3] = ff_avg_pixels16_xy2_msa;
+
+    c->avg_pixels_tab[1][0] = ff_avg_pixels8_msa;
+    c->avg_pixels_tab[1][1] = ff_avg_pixels8_x2_msa;
+    c->avg_pixels_tab[1][2] = ff_avg_pixels8_y2_msa;
+    c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_msa;
+
+    c->avg_pixels_tab[2][0] = ff_avg_pixels4_msa;
+    c->avg_pixels_tab[2][1] = ff_avg_pixels4_x2_msa;
+    c->avg_pixels_tab[2][2] = ff_avg_pixels4_y2_msa;
+    c->avg_pixels_tab[2][3] = ff_avg_pixels4_xy2_msa;
+}
+#endif  // #if HAVE_MSA
+
+void ff_hpeldsp_init_mips(HpelDSPContext *c, int flags)
+{
+#if HAVE_MSA
+    ff_hpeldsp_init_msa(c, flags);
+#endif  // #if HAVE_MSA
+}

libavcodec/mips/hpeldsp_mips.h

+/*
+ * Copyright (c) 2015 Parag Salasakar (Parag.Salasakar@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_HPELDSP_MIPS_H
+#define AVCODEC_MIPS_HPELDSP_MIPS_H
+
+#include "libavcodec/bit_depth_template.c"
+
+void ff_put_pixels16_msa(uint8_t *block, const uint8_t *pixels,
+                         ptrdiff_t line_size, int32_t h);
+void ff_put_pixels16_x2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_put_pixels16_y2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_put_pixels16_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                             ptrdiff_t line_size, int32_t h);
+void ff_put_pixels8_msa(uint8_t *block, const uint8_t *pixels,
+                        ptrdiff_t line_size, int32_t h);
+void ff_put_pixels8_x2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_put_pixels8_y2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_put_pixels8_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_put_pixels4_msa(uint8_t *block, const uint8_t *pixels,
+                        ptrdiff_t line_size, int32_t h);
+void ff_put_pixels4_x2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_put_pixels4_y2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_put_pixels4_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_put_no_rnd_pixels16_x2_msa(uint8_t *block, const uint8_t *pixels,
+                                   ptrdiff_t line_size, int32_t h);
+void ff_put_no_rnd_pixels16_y2_msa(uint8_t *block, const uint8_t *pixels,
+                                   ptrdiff_t line_size, int32_t h);
+void ff_put_no_rnd_pixels16_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                                    ptrdiff_t line_size, int32_t h);
+void ff_put_no_rnd_pixels8_x2_msa(uint8_t *block, const uint8_t *pixels,
+                                  ptrdiff_t line_size, int32_t h);
+void ff_put_no_rnd_pixels8_y2_msa(uint8_t *block, const uint8_t *pixels,
+                                  ptrdiff_t line_size, int32_t h);
+void ff_put_no_rnd_pixels8_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                                   ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels16_msa(uint8_t *block, const uint8_t *pixels,
+                         ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels16_x2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels16_y2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels16_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                             ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels8_msa(uint8_t *block, const uint8_t *pixels,
+                        ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels8_x2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels8_y2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels8_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels4_msa(uint8_t *block, const uint8_t *pixels,
+                        ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels4_x2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels4_y2_msa(uint8_t *block, const uint8_t *pixels,
+                           ptrdiff_t line_size, int32_t h);
+void ff_avg_pixels4_xy2_msa(uint8_t *block, const uint8_t *pixels,
+                            ptrdiff_t line_size, int32_t h);
+
+#endif  // #ifndef AVCODEC_MIPS_HPELDSP_MIPS_H