rv40: NEON optimised weighted prediction

Signed-off-by: Mans Rullgard <mans@mansr.com>

diff --git a/libavcodec/arm/Makefile b/libavcodec/arm/Makefile
index a948e6d..fc17113 100644 (file)
--- a/libavcodec/arm/Makefile
+++ b/libavcodec/arm/Makefile
@@ -69,6 +69,7 @@ NEON-OBJS-$(CONFIG_RV30_DECODER)       += arm/rv34dsp_init_neon.o       \
 NEON-OBJS-$(CONFIG_RV40_DECODER)       += arm/rv34dsp_init_neon.o       \
                                           arm/rv34dsp_neon.o            \
                                           arm/rv40dsp_init_neon.o       \
+                                          arm/rv40dsp_neon.o            \
                                           arm/h264cmc_neon.o            \
 
 NEON-OBJS-$(CONFIG_VP3_DECODER)        += arm/vp3dsp_neon.o

diff --git a/libavcodec/arm/rv40dsp_init_neon.c b/libavcodec/arm/rv40dsp_init_neon.c
index aa4a88d..3a863e1 100644 (file)
--- a/libavcodec/arm/rv40dsp_init_neon.c
+++ b/libavcodec/arm/rv40dsp_init_neon.c
@@ -29,10 +29,16 @@ void ff_put_rv40_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int);
 void ff_avg_rv40_chroma_mc8_neon(uint8_t *, uint8_t *, int, int, int, int);
 void ff_avg_rv40_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int);
 
+void ff_rv40_weight_func_16_neon(uint8_t *, uint8_t *, uint8_t *, int, int, int);
+void ff_rv40_weight_func_8_neon(uint8_t *, uint8_t *, uint8_t *, int, int, int);
+
 void ff_rv40dsp_init_neon(RV34DSPContext *c, DSPContext* dsp)
 {
     c->put_chroma_pixels_tab[0] = ff_put_rv40_chroma_mc8_neon;
     c->put_chroma_pixels_tab[1] = ff_put_rv40_chroma_mc4_neon;
     c->avg_chroma_pixels_tab[0] = ff_avg_rv40_chroma_mc8_neon;
     c->avg_chroma_pixels_tab[1] = ff_avg_rv40_chroma_mc4_neon;
+
+    c->rv40_weight_pixels_tab[0] = ff_rv40_weight_func_16_neon;
+    c->rv40_weight_pixels_tab[1] = ff_rv40_weight_func_8_neon;
 }

diff --git a/libavcodec/arm/rv40dsp_neon.S b/libavcodec/arm/rv40dsp_neon.S
new file mode 100644 (file)
index 0000000..cafd98a
--- /dev/null
+++ b/libavcodec/arm/rv40dsp_neon.S
@@ -0,0 +1,85 @@
+/*
+ * Copyright (c) 2011 Janne Grunau <janne-libav@jannau.net>
+ *
+ * This file is part of Libav.
+ *
+ * Libav 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.
+ *
+ * Libav 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 Libav; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "asm.S"
+
+.macro  rv40_weight
+        vmovl.u8        q8,  d2
+        vmovl.u8        q9,  d3
+        vmovl.u8        q10, d4
+        vmovl.u8        q11, d5
+        vmull.u16       q2,  d16, d0[2]
+        vmull.u16       q3,  d17, d0[2]
+        vmull.u16       q8,  d18, d0[2]
+        vmull.u16       q9,  d19, d0[2]
+        vmull.u16       q12, d20, d0[0]
+        vmull.u16       q13, d21, d0[0]
+        vmull.u16       q14, d22, d0[0]
+        vmull.u16       q15, d23, d0[0]
+        vshrn.i32       d4,  q2,  #9
+        vshrn.i32       d5,  q3,  #9
+        vshrn.i32       d6,  q8,  #9
+        vshrn.i32       d7,  q9,  #9
+        vshrn.i32       d16, q12, #9
+        vshrn.i32       d17, q13, #9
+        vshrn.i32       d18, q14, #9
+        vshrn.i32       d19, q15, #9
+        vadd.u16        q2,  q2,  q8
+        vadd.u16        q3,  q3,  q9
+        vrshrn.i16      d2,  q2,  #5
+        vrshrn.i16      d3,  q3,  #5
+.endm
+
+/* void ff_rv40_weight_func_16_neon(uint8_t *dst, uint8_t *src1, uint8_t *src2,
+                                    int w1, int w2, int stride) */
+function ff_rv40_weight_func_16_neon, export=1
+        ldr             r12, [sp]
+        vmov            d0,  r3,  r12
+        ldr             r12, [sp, #4]
+        mov             r3,  #16
+1:
+        vld1.8          {q1},     [r1,:128], r12
+        vld1.8          {q2},     [r2,:128], r12
+        rv40_weight
+        vst1.8          {q1},     [r0,:128], r12
+        subs            r3,  r3,  #1
+        bne             1b
+        bx              lr
+endfunc
+
+/* void ff_rv40_weight_func_8_neon(uint8_t *dst, uint8_t *src1, uint8_t *src2,
+                                   int w1, int w2, int stride) */
+function ff_rv40_weight_func_8_neon, export=1
+        ldr             r12, [sp]
+        vmov            d0,  r3,  r12
+        ldr             r12, [sp, #4]
+        mov             r3,  #8
+1:
+        vld1.8          {d2},     [r1,:64], r12
+        vld1.8          {d3},     [r1,:64], r12
+        vld1.8          {d4},     [r2,:64], r12
+        vld1.8          {d5},     [r2,:64], r12
+        rv40_weight
+        vst1.8          {d2},     [r0,:64], r12
+        vst1.8          {d3},     [r0,:64], r12
+        subs            r3,  r3,  #2
+        bne             1b
+        bx              lr
+endfunc