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
|
From f75e9d1868e21dd75ff3a2ca3561546d23877ddb Mon Sep 17 00:00:00 2001
From: Siarhei Siamashka <siarhei.siamashka@nokia.com>
Date: Thu, 10 Dec 2009 00:51:50 +0200
Subject: [PATCH 5/8] ARM: added NEON optimizations for fetch/store r5g6b5 scanline
---
pixman/pixman-arm-neon-asm.S | 20 ++++++++++++++++++++
pixman/pixman-arm-neon.c | 40 ++++++++++++++++++++++++++++++++++++++++
2 files changed, 60 insertions(+), 0 deletions(-)
diff --git a/pixman/pixman-arm-neon-asm.S b/pixman/pixman-arm-neon-asm.S
index e4db5cd..c79ba81 100644
--- a/pixman/pixman-arm-neon-asm.S
+++ b/pixman/pixman-arm-neon-asm.S
@@ -459,6 +459,16 @@ generate_composite_function \
pixman_composite_src_8888_0565_process_pixblock_tail, \
pixman_composite_src_8888_0565_process_pixblock_tail_head
+generate_composite_function_single_scanline \
+ pixman_store_scanline_r5g6b5_asm_neon, 32, 0, 16, \
+ FLAG_DST_WRITEONLY | FLAG_DEINTERLEAVE_32BPP, \
+ 8, /* number of pixels, processed in a single block */ \
+ default_init, \
+ default_cleanup, \
+ pixman_composite_src_8888_0565_process_pixblock_head, \
+ pixman_composite_src_8888_0565_process_pixblock_tail, \
+ pixman_composite_src_8888_0565_process_pixblock_tail_head
+
/******************************************************************************/
.macro pixman_composite_src_0565_8888_process_pixblock_head
@@ -494,6 +504,16 @@ generate_composite_function \
pixman_composite_src_0565_8888_process_pixblock_tail, \
pixman_composite_src_0565_8888_process_pixblock_tail_head
+generate_composite_function_single_scanline \
+ pixman_fetch_scanline_r5g6b5_asm_neon, 16, 0, 32, \
+ FLAG_DST_WRITEONLY | FLAG_DEINTERLEAVE_32BPP, \
+ 8, /* number of pixels, processed in a single block */ \
+ default_init, \
+ default_cleanup, \
+ pixman_composite_src_0565_8888_process_pixblock_head, \
+ pixman_composite_src_0565_8888_process_pixblock_tail, \
+ pixman_composite_src_0565_8888_process_pixblock_tail_head
+
/******************************************************************************/
.macro pixman_composite_add_8_8_process_pixblock_head
diff --git a/pixman/pixman-arm-neon.c b/pixman/pixman-arm-neon.c
index cbfd7cf..f88c8f8 100644
--- a/pixman/pixman-arm-neon.c
+++ b/pixman/pixman-arm-neon.c
@@ -392,6 +392,42 @@ BIND_COMBINE_U (over)
BIND_COMBINE_U (add)
BIND_COMBINE_U (out_reverse)
+void
+pixman_fetch_scanline_r5g6b5_asm_neon (int width,
+ uint32_t *buffer,
+ const uint16_t *pixel);
+void
+pixman_store_scanline_r5g6b5_asm_neon (int width,
+ uint16_t *pixel,
+ const uint32_t *values);
+
+static void
+neon_fetch_scanline_r5g6b5 (pixman_image_t *image,
+ int x,
+ int y,
+ int width,
+ uint32_t * buffer,
+ const uint32_t *mask)
+{
+ const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+ const uint16_t *pixel = (const uint16_t *)bits + x;
+
+ pixman_fetch_scanline_r5g6b5_asm_neon (width, buffer, pixel);
+}
+
+static void
+neon_store_scanline_r5g6b5 (bits_image_t * image,
+ int x,
+ int y,
+ int width,
+ const uint32_t *values)
+{
+ uint32_t *bits = image->bits + image->rowstride * y;
+ uint16_t *pixel = ((uint16_t *) bits) + x;
+
+ pixman_store_scanline_r5g6b5_asm_neon (width, pixel, values);
+}
+
pixman_implementation_t *
_pixman_implementation_create_arm_neon (void)
{
@@ -407,6 +443,10 @@ _pixman_implementation_create_arm_neon (void)
imp->combine_32[PIXMAN_OP_ADD] = neon_combine_add_u;
imp->combine_32[PIXMAN_OP_OUT_REVERSE] = neon_combine_out_reverse_u;
+ _pixman_bits_override_accessors (PIXMAN_r5g6b5,
+ neon_fetch_scanline_r5g6b5,
+ neon_store_scanline_r5g6b5);
+
imp->blt = arm_neon_blt;
imp->fill = arm_neon_fill;
--
1.6.6.1
|
