arm: Add NEON optimizations for 10 and 12 bit vp9 loop filter

This work is sponsored by, and copyright, Google.

This is pretty much similar to the 8 bpp version, but in some senses
simpler. All input pixels are 16 bits, and all intermediates also fit
in 16 bits, so there's no lengthening/narrowing in the filter at all.

For the full 16 pixel wide filter, we can only process 4 pixels at a time
(using an implementation very much similar to the one for 8 bpp),
but we can do 8 pixels at a time for the 4 and 8 pixel wide filters with
a different implementation of the core filter.

Examples of relative speedup compared to the C version, from checkasm:
                                   Cortex    A7     A8     A9    A53
vp9_loop_filter_h_4_8_10bpp_neon:          1.83   2.16   1.40   2.09
vp9_loop_filter_h_8_8_10bpp_neon:          1.39   1.67   1.24   1.70
vp9_loop_filter_h_16_8_10bpp_neon:         1.56   1.47   1.10   1.81
vp9_loop_filter_h_16_16_10bpp_neon:        1.94   1.69   1.33   2.24
vp9_loop_filter_mix2_h_44_16_10bpp_neon:   2.01   2.27   1.67   2.39
vp9_loop_filter_mix2_h_48_16_10bpp_neon:   1.84   2.06   1.45   2.19
vp9_loop_filter_mix2_h_84_16_10bpp_neon:   1.89   2.20   1.47   2.29
vp9_loop_filter_mix2_h_88_16_10bpp_neon:   1.69   2.12   1.47   2.08
vp9_loop_filter_mix2_v_44_16_10bpp_neon:   3.16   3.98   2.50   4.05
vp9_loop_filter_mix2_v_48_16_10bpp_neon:   2.84   3.64   2.25   3.77
vp9_loop_filter_mix2_v_84_16_10bpp_neon:   2.65   3.45   2.16   3.54
vp9_loop_filter_mix2_v_88_16_10bpp_neon:   2.55   3.30   2.16   3.55
vp9_loop_filter_v_4_8_10bpp_neon:          2.85   3.97   2.24   3.68
vp9_loop_filter_v_8_8_10bpp_neon:          2.27   3.19   1.96   3.08
vp9_loop_filter_v_16_8_10bpp_neon:         3.42   2.74   2.26   4.40
vp9_loop_filter_v_16_16_10bpp_neon:        2.86   2.44   1.93   3.88

The speedup vs C code measured in checkasm is around 1.1-4x.
These numbers are quite inconclusive though, since the checkasm test
runs multiple filterings on top of each other, so later rounds might
end up with different codepaths (different decisions on which filter
to apply, based on input pixel differences).

Based on START_TIMER/STOP_TIMER wrapping around a few individual
functions, the speedup vs C code is around 2-4x.

Signed-off-by: Martin Storsjö <martin@martin.st>

libavcodec/arm/Makefile

@@ -144,6 +144,7 @@ NEON-OBJS-$(CONFIG_VORBIS_DECODER)     += arm/vorbisdsp_neon.o
 NEON-OBJS-$(CONFIG_VP6_DECODER)        += arm/vp6dsp_neon.o
 NEON-OBJS-$(CONFIG_VP9_DECODER)        += arm/vp9itxfm_16bpp_neon.o     \
                                           arm/vp9itxfm_neon.o           \
+                                          arm/vp9lpf_16bpp_neon.o       \
                                           arm/vp9lpf_neon.o             \
                                           arm/vp9mc_16bpp_neon.o        \
                                           arm/vp9mc_neon.o

libavcodec/arm/vp9dsp_init_16bpp_arm_template.c

@@ -187,8 +187,70 @@ static av_cold void vp9dsp_itxfm_init_arm(VP9DSPContext *dsp)
     }
 }
 
+#define define_loop_filter(dir, wd, size, bpp) \
+void ff_vp9_loop_filter_##dir##_##wd##_##size##_##bpp##_neon(uint8_t *dst, ptrdiff_t stride, int E, int I, int H)
+
+#define define_loop_filters(wd, size, bpp) \
+    define_loop_filter(h, wd, size, bpp);  \
+    define_loop_filter(v, wd, size, bpp)
+
+define_loop_filters(4,  8,  BPP);
+define_loop_filters(8,  8,  BPP);
+define_loop_filters(16, 8,  BPP);
+
+define_loop_filters(16, 16, BPP);
+
+define_loop_filters(44, 16, BPP);
+define_loop_filters(48, 16, BPP);
+define_loop_filters(84, 16, BPP);
+define_loop_filters(88, 16, BPP);
+
+static av_cold void vp9dsp_loopfilter_init_arm(VP9DSPContext *dsp)
+{
+    int cpu_flags = av_get_cpu_flags();
+
+    if (have_neon(cpu_flags)) {
+#define init_lpf_func_8(idx1, idx2, dir, wd, bpp) \
+    dsp->loop_filter_8[idx1][idx2] = ff_vp9_loop_filter_##dir##_##wd##_8_##bpp##_neon
+
+#define init_lpf_func_16(idx, dir, bpp) \
+    dsp->loop_filter_16[idx] = ff_vp9_loop_filter_##dir##_16_16_##bpp##_neon
+
+#define init_lpf_func_mix2(idx1, idx2, idx3, dir, wd, bpp) \
+    dsp->loop_filter_mix2[idx1][idx2][idx3] = ff_vp9_loop_filter_##dir##_##wd##_16_##bpp##_neon
+
+#define init_lpf_funcs_8_wd(idx, wd, bpp) \
+    init_lpf_func_8(idx, 0, h, wd, bpp);  \
+    init_lpf_func_8(idx, 1, v, wd, bpp)
+
+#define init_lpf_funcs_16(bpp)   \
+    init_lpf_func_16(0, h, bpp); \
+    init_lpf_func_16(1, v, bpp)
+
+#define init_lpf_funcs_mix2_wd(idx1, idx2, wd, bpp) \
+    init_lpf_func_mix2(idx1, idx2, 0, h, wd, bpp);  \
+    init_lpf_func_mix2(idx1, idx2, 1, v, wd, bpp)
+
+#define init_lpf_funcs_8(bpp)        \
+    init_lpf_funcs_8_wd(0, 4,  bpp); \
+    init_lpf_funcs_8_wd(1, 8,  bpp); \
+    init_lpf_funcs_8_wd(2, 16, bpp)
+
+#define init_lpf_funcs_mix2(bpp)           \
+    init_lpf_funcs_mix2_wd(0, 0, 44, bpp); \
+    init_lpf_funcs_mix2_wd(0, 1, 48, bpp); \
+    init_lpf_funcs_mix2_wd(1, 0, 84, bpp); \
+    init_lpf_funcs_mix2_wd(1, 1, 88, bpp)
+
+        init_lpf_funcs_8(BPP);
+        init_lpf_funcs_16(BPP);
+        init_lpf_funcs_mix2(BPP);
+    }
+}
+
 av_cold void INIT_FUNC(VP9DSPContext *dsp)
 {
     vp9dsp_mc_init_arm(dsp);
+    vp9dsp_loopfilter_init_arm(dsp);
     vp9dsp_itxfm_init_arm(dsp);
 }