1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
|
From: Mans Rullgard <mans@mansr.com>
Date: Sun, 24 Aug 2008 21:27:49 +0000 (+0100)
Subject: ARM: NEON optimised h264_idct_add
X-Git-Url: http://git.mansr.com/?p=ffmpeg.mru;a=commitdiff_plain;h=ebfab90234268bb35600a06e9982ca1358ea43f3
ARM: NEON optimised h264_idct_add
---
diff --git a/libavcodec/Makefile b/libavcodec/Makefile
index 36ba158..053e752 100644
--- a/libavcodec/Makefile
+++ b/libavcodec/Makefile
@@ -438,6 +438,7 @@ OBJS-$(HAVE_NEON) += armv4l/dsputil_neon.o \
ASM_OBJS-$(HAVE_NEON) += armv4l/dsputil_neon_s.o \
armv4l/simple_idct_neon.o \
armv4l/h264dsp_neon.o \
+ armv4l/h264idct_neon.o \
OBJS-$(HAVE_VIS) += sparc/dsputil_vis.o \
sparc/simple_idct_vis.o \
diff --git a/libavcodec/armv4l/dsputil_neon.c b/libavcodec/armv4l/dsputil_neon.c
index a932aa9..74f9b4d 100644
--- a/libavcodec/armv4l/dsputil_neon.c
+++ b/libavcodec/armv4l/dsputil_neon.c
@@ -88,6 +88,8 @@ void ff_h264_v_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha,
void ff_h264_h_loop_filter_chroma_neon(uint8_t *pix, int stride, int alpha,
int beta, int8_t *tc0);
+void ff_h264_idct_add_neon(uint8_t *dst, DCTELEM *block, int stride);
+
void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
{
c->put_pixels_tab[0][0] = ff_put_pixels16_neon;
@@ -152,4 +154,6 @@ void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
c->h264_h_loop_filter_luma = ff_h264_h_loop_filter_luma_neon;
c->h264_v_loop_filter_chroma = ff_h264_v_loop_filter_chroma_neon;
c->h264_h_loop_filter_chroma = ff_h264_h_loop_filter_chroma_neon;
+
+ c->h264_idct_add = ff_h264_idct_add_neon;
}
diff --git a/libavcodec/armv4l/h264idct_neon.S b/libavcodec/armv4l/h264idct_neon.S
new file mode 100644
index 0000000..8f456f3
--- /dev/null
+++ b/libavcodec/armv4l/h264idct_neon.S
@@ -0,0 +1,77 @@
+/*
+ * Copyright (c) 2008 Mans Rullgard <mans@mansr.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
+ */
+
+ .fpu neon
+
+ .text
+
+ .global ff_h264_idct_add_neon
+ .func ff_h264_idct_add_neon
+ff_h264_idct_add_neon:
+ mov r3, #(1<<5)
+ vmov.i16 d16, #0
+ vmov.16 d16[0], r3
+ vld1.64 {d0-d3}, [r1,:128]
+ vadd.i16 d0, d0, d16
+
+ vswp d1, d2
+ vadd.i16 d4, d0, d1
+ vshr.s16 q8, q1, #1
+ vsub.i16 d5, d0, d1
+ vadd.i16 d6, d2, d17
+ vsub.i16 d7, d16, d3
+ vadd.i16 q0, q2, q3
+ vsub.i16 q1, q2, q3
+
+ vtrn.16 d0, d1
+ vtrn.16 d3, d2
+ vtrn.32 d0, d3
+ vtrn.32 d1, d2
+
+ vadd.i16 d4, d0, d3
+ vld1.32 {d18[0]}, [r0,:32], r2
+ vswp d1, d3
+ vshr.s16 q8, q1, #1
+ vld1.32 {d19[1]}, [r0,:32], r2
+ vsub.i16 d5, d0, d1
+ vld1.32 {d18[1]}, [r0,:32], r2
+ vadd.i16 d6, d16, d3
+ vld1.32 {d19[0]}, [r0,:32], r2
+ vsub.i16 d7, d2, d17
+ sub r0, r0, r2, lsl #2
+ vadd.i16 q0, q2, q3
+ vsub.i16 q1, q2, q3
+
+ vshr.s16 q0, q0, #6
+ vshr.s16 q1, q1, #6
+
+ vaddw.u8 q0, q0, d18
+ vaddw.u8 q1, q1, d19
+
+ vqmovun.s16 d0, q0
+ vqmovun.s16 d1, q1
+
+ vst1.32 {d0[0]}, [r0,:32], r2
+ vst1.32 {d1[1]}, [r0,:32], r2
+ vst1.32 {d0[1]}, [r0,:32], r2
+ vst1.32 {d1[0]}, [r0,:32], r2
+
+ bx lr
+ .endfunc
|
