video/fbdev/c2p_iplan2.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 16 8-bit pixels, stored in 4 32-bit words
0022      *  containing
0023      *    - 16 8-bit chunky pixels on input
0024      *    - permutated planar data (2 planes per 32-bit word) on output
0025      */
0026
0027 static void c2p_16x8(u32 d[4])
0028 {
0029     transp4(d, 8, 2);
0030     transp4(d, 1, 2);
0031     transp4x(d, 16, 2);
0032     transp4x(d, 2, 2);
0033     transp4(d, 4, 1);
0034 }
0035
0036
0037     /*
0038      *  Array containing the permutation indices of the planar data after c2p
0039      */
0040
0041 static const int perm_c2p_16x8[4] = { 1, 3, 0, 2 };
0042
0043
0044     /*
0045      *  Store a full block of iplan2 data after c2p conversion
0046      */
0047
0048 static inline void store_iplan2(void *dst, u32 bpp, u32 d[4])
0049 {
0050     int i;
0051
0052     for (i = 0; i < bpp/2; i++, dst += 4)
0053         put_unaligned_be32(d[perm_c2p_16x8[i]], dst);
0054 }
0055
0056
0057     /*
0058      *  Store a partial block of iplan2 data after c2p conversion
0059      */
0060
0061 static inline void store_iplan2_masked(void *dst, u32 bpp, u32 d[4], u32 mask)
0062 {
0063     int i;
0064
0065     for (i = 0; i < bpp/2; i++, dst += 4)
0066         put_unaligned_be32(comp(d[perm_c2p_16x8[i]],
0067                     get_unaligned_be32(dst), mask),
0068                    dst);
0069 }
0070
0071
0072     /*
0073      *  c2p_iplan2 - Copy 8-bit chunky image data to an interleaved planar
0074      *  frame buffer with 2 bytes of interleave
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      *  @src_nextline: Image offset to the next line (in bytes)
0082      *  @bpp: Bits per pixel of the planar frame buffer (2, 4, or 8)
0083      */
0084
0085 void c2p_iplan2(void *dst, const void *src, u32 dx, u32 dy, u32 width,
0086         u32 height, u32 dst_nextline, u32 src_nextline, u32 bpp)
0087 {
0088     union {
0089         u8 pixels[16];
0090         u32 words[4];
0091     } d;
0092     u32 dst_idx, first, last, w;
0093     const u8 *c;
0094     void *p;
0095
0096     dst += dy*dst_nextline+(dx & ~15)*bpp;
0097     dst_idx = dx % 16;
0098     first = 0xffffU >> dst_idx;
0099     first |= first << 16;
0100     last = 0xffffU ^ (0xffffU >> ((dst_idx+width) % 16));
0101     last |= last << 16;
0102     while (height--) {
0103         c = src;
0104         p = dst;
0105         w = width;
0106         if (dst_idx+width <= 16) {
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_16x8(d.words);
0113             store_iplan2_masked(p, bpp, d.words, first);
0114             p += bpp*2;
0115         } else {
0116             /* Multiple destination words */
0117             w = width;
0118             /* Leading bits */
0119             if (dst_idx) {
0120                 w = 16 - dst_idx;
0121                 memset(d.pixels, 0, dst_idx);
0122                 memcpy(d.pixels+dst_idx, c, w);
0123                 c += w;
0124                 c2p_16x8(d.words);
0125                 store_iplan2_masked(p, bpp, d.words, first);
0126                 p += bpp*2;
0127                 w = width-w;
0128             }
0129             /* Main chunk */
0130             while (w >= 16) {
0131                 memcpy(d.pixels, c, 16);
0132                 c += 16;
0133                 c2p_16x8(d.words);
0134                 store_iplan2(p, bpp, d.words);
0135                 p += bpp*2;
0136                 w -= 16;
0137             }
0138             /* Trailing bits */
0139             w %= 16;
0140             if (w > 0) {
0141                 memcpy(d.pixels, c, w);
0142                 memset(d.pixels+w, 0, 16-w);
0143                 c2p_16x8(d.words);
0144                 store_iplan2_masked(p, bpp, d.words, last);
0145             }
0146         }
0147         src += src_nextline;
0148         dst += dst_nextline;
0149     }
0150 }
0151 EXPORT_SYMBOL_GPL(c2p_iplan2);
0152
0153 MODULE_LICENSE("GPL");