1/*
2 *  Fast C2P (Chunky-to-Planar) Conversion
3 *
4 *  Copyright (C) 2003-2008 Geert Uytterhoeven
5 *
6 *  This file is subject to the terms and conditions of the GNU General Public
7 *  License. See the file COPYING in the main directory of this archive
8 *  for more details.
9 */
10
11#include <linux/module.h>
12#include <linux/string.h>
13
14#include <asm/unaligned.h>
15
16#include "c2p.h"
17#include "c2p_core.h"
18
19
20    /*
21     *  Perform a full C2P step on 16 8-bit pixels, stored in 4 32-bit words
22     *  containing
23     *    - 16 8-bit chunky pixels on input
24     *    - permutated planar data (2 planes per 32-bit word) on output
25     */
26
27static void c2p_16x8(u32 d[4])
28{
29	transp4(d, 8, 2);
30	transp4(d, 1, 2);
31	transp4x(d, 16, 2);
32	transp4x(d, 2, 2);
33	transp4(d, 4, 1);
34}
35
36
37    /*
38     *  Array containing the permutation indices of the planar data after c2p
39     */
40
41static const int perm_c2p_16x8[4] = { 1, 3, 0, 2 };
42
43
44    /*
45     *  Store a full block of iplan2 data after c2p conversion
46     */
47
48static inline void store_iplan2(void *dst, u32 bpp, u32 d[4])
49{
50	int i;
51
52	for (i = 0; i < bpp/2; i++, dst += 4)
53		put_unaligned_be32(d[perm_c2p_16x8[i]], dst);
54}
55
56
57    /*
58     *  Store a partial block of iplan2 data after c2p conversion
59     */
60
61static inline void store_iplan2_masked(void *dst, u32 bpp, u32 d[4], u32 mask)
62{
63	int i;
64
65	for (i = 0; i < bpp/2; i++, dst += 4)
66		put_unaligned_be32(comp(d[perm_c2p_16x8[i]],
67					get_unaligned_be32(dst), mask),
68				   dst);
69}
70
71
72    /*
73     *  c2p_iplan2 - Copy 8-bit chunky image data to an interleaved planar
74     *  frame buffer with 2 bytes of interleave
75     *  @dst: Starting address of the planar frame buffer
76     *  @dx: Horizontal destination offset (in pixels)
77     *  @dy: Vertical destination offset (in pixels)
78     *  @width: Image width (in pixels)
79     *  @height: Image height (in pixels)
80     *  @dst_nextline: Frame buffer offset to the next line (in bytes)
81     *  @src_nextline: Image offset to the next line (in bytes)
82     *  @bpp: Bits per pixel of the planar frame buffer (2, 4, or 8)
83     */
84
85void c2p_iplan2(void *dst, const void *src, u32 dx, u32 dy, u32 width,
86		u32 height, u32 dst_nextline, u32 src_nextline, u32 bpp)
87{
88	union {
89		u8 pixels[16];
90		u32 words[4];
91	} d;
92	u32 dst_idx, first, last, w;
93	const u8 *c;
94	void *p;
95
96	dst += dy*dst_nextline+(dx & ~15)*bpp;
97	dst_idx = dx % 16;
98	first = 0xffffU >> dst_idx;
99	first |= first << 16;
100	last = 0xffffU ^ (0xffffU >> ((dst_idx+width) % 16));
101	last |= last << 16;
102	while (height--) {
103		c = src;
104		p = dst;
105		w = width;
106		if (dst_idx+width <= 16) {
107			/* Single destination word */
108			first &= last;
109			memset(d.pixels, 0, sizeof(d));
110			memcpy(d.pixels+dst_idx, c, width);
111			c += width;
112			c2p_16x8(d.words);
113			store_iplan2_masked(p, bpp, d.words, first);
114			p += bpp*2;
115		} else {
116			/* Multiple destination words */
117			w = width;
118			/* Leading bits */
119			if (dst_idx) {
120				w = 16 - dst_idx;
121				memset(d.pixels, 0, dst_idx);
122				memcpy(d.pixels+dst_idx, c, w);
123				c += w;
124				c2p_16x8(d.words);
125				store_iplan2_masked(p, bpp, d.words, first);
126				p += bpp*2;
127				w = width-w;
128			}
129			/* Main chunk */
130			while (w >= 16) {
131				memcpy(d.pixels, c, 16);
132				c += 16;
133				c2p_16x8(d.words);
134				store_iplan2(p, bpp, d.words);
135				p += bpp*2;
136				w -= 16;
137			}
138			/* Trailing bits */
139			w %= 16;
140			if (w > 0) {
141				memcpy(d.pixels, c, w);
142				memset(d.pixels+w, 0, 16-w);
143				c2p_16x8(d.words);
144				store_iplan2_masked(p, bpp, d.words, last);
145			}
146		}
147		src += src_nextline;
148		dst += dst_nextline;
149	}
150}
151EXPORT_SYMBOL_GPL(c2p_iplan2);
152
153MODULE_LICENSE("GPL");
154