video/fbdev/c2p_planar.c

0001 /*
0002  *  Fast C2P (Chunky-to-Planar) Conversion
0003  *
0004  *  Copyright (C) 2003-2008 Geert Uytterhoeven
0005  *
0006  *  This file is subject to the terms and conditions of the GNU General Public
0007  *  License. See the file COPYING in the main directory of this archive
0008  *  for more details.
0009  */
0010
0011 #include <linux/module.h>
0012 #include <linux/string.h>
0013
0014 #include <asm/unaligned.h>
0015
0016 #include "c2p.h"
0017 #include "c2p_core.h"
0018
0019
0020     /*
0021      *  Perform a full C2P step on 32 8-bit pixels, stored in 8 32-bit words
0022      *  containing
0023      *    - 32 8-bit chunky pixels on input
0024      *    - permutated planar data (1 plane per 32-bit word) on output
0025      */
0026
0027 static void c2p_32x8(u32 d[8])
0028 {
0029     transp8(d, 16, 4);
0030     transp8(d, 8, 2);
0031     transp8(d, 4, 1);
0032     transp8(d, 2, 4);
0033     transp8(d, 1, 2);
0034 }
0035
0036
0037     /*
0038      *  Array containing the permutation indices of the planar data after c2p
0039      */
0040
0041 static const int perm_c2p_32x8[8] = { 7, 5, 3, 1, 6, 4, 2, 0 };
0042
0043
0044     /*
0045      *  Store a full block of planar data after c2p conversion
0046      */
0047
0048 static inline void store_planar(void *dst, u32 dst_inc, u32 bpp, u32 d[8])
0049 {
0050     int i;
0051
0052     for (i = 0; i < bpp; i++, dst += dst_inc)
0053         put_unaligned_be32(d[perm_c2p_32x8[i]], dst);
0054 }
0055
0056
0057     /*
0058      *  Store a partial block of planar data after c2p conversion
0059      */
0060
0061 static inline void store_planar_masked(void *dst, u32 dst_inc, u32 bpp,
0062                        u32 d[8], u32 mask)
0063 {
0064     int i;
0065
0066     for (i = 0; i < bpp; i++, dst += dst_inc)
0067         put_unaligned_be32(comp(d[perm_c2p_32x8[i]],
0068                     get_unaligned_be32(dst), mask),
0069                    dst);
0070 }
0071
0072
0073     /*
0074      *  c2p_planar - Copy 8-bit chunky image data to a planar frame buffer
0075      *  @dst: Starting address of the planar frame buffer
0076      *  @dx: Horizontal destination offset (in pixels)
0077      *  @dy: Vertical destination offset (in pixels)
0078      *  @width: Image width (in pixels)
0079      *  @height: Image height (in pixels)
0080      *  @dst_nextline: Frame buffer offset to the next line (in bytes)
0081      *  @dst_nextplane: Frame buffer offset to the next plane (in bytes)
0082      *  @src_nextline: Image offset to the next line (in bytes)
0083      *  @bpp: Bits per pixel of the planar frame buffer (1-8)
0084      */
0085
0086 void c2p_planar(void *dst, const void *src, u32 dx, u32 dy, u32 width,
0087         u32 height, u32 dst_nextline, u32 dst_nextplane,
0088         u32 src_nextline, u32 bpp)
0089 {
0090     union {
0091         u8 pixels[32];
0092         u32 words[8];
0093     } d;
0094     u32 dst_idx, first, last, w;
0095     const u8 *c;
0096     void *p;
0097
0098     dst += dy*dst_nextline+(dx & ~31);
0099     dst_idx = dx % 32;
0100     first = 0xffffffffU >> dst_idx;
0101     last = ~(0xffffffffU >> ((dst_idx+width) % 32));
0102     while (height--) {
0103         c = src;
0104         p = dst;
0105         w = width;
0106         if (dst_idx+width <= 32) {
0107             /* Single destination word */
0108             first &= last;
0109             memset(d.pixels, 0, sizeof(d));
0110             memcpy(d.pixels+dst_idx, c, width);
0111             c += width;
0112             c2p_32x8(d.words);
0113             store_planar_masked(p, dst_nextplane, bpp, d.words,
0114                         first);
0115             p += 4;
0116         } else {
0117             /* Multiple destination words */
0118             w = width;
0119             /* Leading bits */
0120             if (dst_idx) {
0121                 w = 32 - dst_idx;
0122                 memset(d.pixels, 0, dst_idx);
0123                 memcpy(d.pixels+dst_idx, c, w);
0124                 c += w;
0125                 c2p_32x8(d.words);
0126                 store_planar_masked(p, dst_nextplane, bpp,
0127                             d.words, first);
0128                 p += 4;
0129                 w = width-w;
0130             }
0131             /* Main chunk */
0132             while (w >= 32) {
0133                 memcpy(d.pixels, c, 32);
0134                 c += 32;
0135                 c2p_32x8(d.words);
0136                 store_planar(p, dst_nextplane, bpp, d.words);
0137                 p += 4;
0138                 w -= 32;
0139             }
0140             /* Trailing bits */
0141             w %= 32;
0142             if (w > 0) {
0143                 memcpy(d.pixels, c, w);
0144                 memset(d.pixels+w, 0, 32-w);
0145                 c2p_32x8(d.words);
0146                 store_planar_masked(p, dst_nextplane, bpp,
0147                             d.words, last);
0148             }
0149         }
0150         src += src_nextline;
0151         dst += dst_nextline;
0152     }
0153 }
0154 EXPORT_SYMBOL_GPL(c2p_planar);
0155
0156 MODULE_LICENSE("GPL");