1 /*
2 * jidctflt.c
3 *
4 * Copyright (C) 1994-1998, Thomas G. Lane.
5 * This file is part of the Independent JPEG Group's software.
6 *
7 * The authors make NO WARRANTY or representation, either express or implied,
8 * with respect to this software, its quality, accuracy, merchantability, or
9 * fitness for a particular purpose. This software is provided "AS IS", and you,
10 * its user, assume the entire risk as to its quality and accuracy.
11 *
12 * This software is copyright (C) 1991-1998, Thomas G. Lane.
13 * All Rights Reserved except as specified below.
14 *
15 * Permission is hereby granted to use, copy, modify, and distribute this
16 * software (or portions thereof) for any purpose, without fee, subject to these
17 * conditions:
18 * (1) If any part of the source code for this software is distributed, then this
19 * README file must be included, with this copyright and no-warranty notice
20 * unaltered; and any additions, deletions, or changes to the original files
21 * must be clearly indicated in accompanying documentation.
22 * (2) If only executable code is distributed, then the accompanying
23 * documentation must state that "this software is based in part on the work of
24 * the Independent JPEG Group".
25 * (3) Permission for use of this software is granted only if the user accepts
26 * full responsibility for any undesirable consequences; the authors accept
27 * NO LIABILITY for damages of any kind.
28 *
29 * These conditions apply to any software derived from or based on the IJG code,
30 * not just to the unmodified library. If you use our work, you ought to
31 * acknowledge us.
32 *
33 * Permission is NOT granted for the use of any IJG author's name or company name
34 * in advertising or publicity relating to this software or products derived from
35 * it. This software may be referred to only as "the Independent JPEG Group's
36 * software".
37 *
38 * We specifically permit and encourage the use of this software as the basis of
39 * commercial products, provided that all warranty or liability claims are
40 * assumed by the product vendor.
41 *
42 *
43 * This file contains a floating-point implementation of the
44 * inverse DCT (Discrete Cosine Transform). In the IJG code, this routine
45 * must also perform dequantization of the input coefficients.
46 *
47 * This implementation should be more accurate than either of the integer
48 * IDCT implementations. However, it may not give the same results on all
49 * machines because of differences in roundoff behavior. Speed will depend
50 * on the hardware's floating point capacity.
51 *
52 * A 2-D IDCT can be done by 1-D IDCT on each column followed by 1-D IDCT
53 * on each row (or vice versa, but it's more convenient to emit a row at
54 * a time). Direct algorithms are also available, but they are much more
55 * complex and seem not to be any faster when reduced to code.
56 *
57 * This implementation is based on Arai, Agui, and Nakajima's algorithm for
58 * scaled DCT. Their original paper (Trans. IEICE E-71(11):1095) is in
59 * Japanese, but the algorithm is described in the Pennebaker & Mitchell
60 * JPEG textbook (see REFERENCES section in file README). The following code
61 * is based directly on figure 4-8 in P&M.
62 * While an 8-point DCT cannot be done in less than 11 multiplies, it is
63 * possible to arrange the computation so that many of the multiplies are
64 * simple scalings of the final outputs. These multiplies can then be
65 * folded into the multiplications or divisions by the JPEG quantization
66 * table entries. The AA&N method leaves only 5 multiplies and 29 adds
67 * to be done in the DCT itself.
68 * The primary disadvantage of this method is that with a fixed-point
69 * implementation, accuracy is lost due to imprecise representation of the
70 * scaled quantization values. However, that problem does not arise if
71 * we use floating point arithmetic.
72 */
73
74 #include <stdint.h>
75 #include "tinyjpeg-internal.h"
76
77 #define FAST_FLOAT float
78 #define DCTSIZE 8
79 #define DCTSIZE2 (DCTSIZE * DCTSIZE)
80
81 #define DEQUANTIZE(coef, quantval) (((FAST_FLOAT) (coef)) * (quantval))
82
83 #if defined(__GNUC__) && (defined(__i686__) || defined(__x86_64__))
84
descale_and_clamp(int x,int shift)85 static inline unsigned char descale_and_clamp(int x, int shift)
86 {
87 __asm__ (
88 "add %3,%1\n"
89 "\tsar %2,%1\n"
90 "\tsub $-128,%1\n"
91 "\tcmovl %5,%1\n" /* Use the sub to compare to 0 */
92 "\tcmpl %4,%1\n"
93 "\tcmovg %4,%1\n"
94 : "=r"(x)
95 : "0"(x), "Ic"((unsigned char)shift), "ir" (1U << (shift - 1)), "r" (0xff), "r" (0)
96 );
97 return x;
98 }
99
100 #else
descale_and_clamp(int x,int shift)101 static inline unsigned char descale_and_clamp(int x, int shift)
102 {
103 x += 1UL << (shift - 1);
104 if (x < 0)
105 x = (x >> shift) | ((~(0UL)) << (32 - (shift)));
106 else
107 x >>= shift;
108 x += 128;
109 if (x > 255)
110 return 255;
111 if (x < 0)
112 return 0;
113 return x;
114 }
115 #endif
116
117 /*
118 * Perform dequantization and inverse DCT on one block of coefficients.
119 */
120
tinyjpeg_idct_float(struct component * compptr,uint8_t * output_buf,int stride)121 void tinyjpeg_idct_float(struct component *compptr, uint8_t *output_buf, int stride)
122 {
123 FAST_FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
124 FAST_FLOAT tmp10, tmp11, tmp12, tmp13;
125 FAST_FLOAT z5, z10, z11, z12, z13;
126 int16_t *inptr;
127 FAST_FLOAT *quantptr;
128 FAST_FLOAT *wsptr;
129 uint8_t *outptr;
130 int ctr;
131 FAST_FLOAT workspace[DCTSIZE2]; /* buffers data between passes */
132
133 /* Pass 1: process columns from input, store into work array. */
134
135 inptr = compptr->DCT;
136 quantptr = compptr->Q_table;
137 wsptr = workspace;
138 for (ctr = DCTSIZE; ctr > 0; ctr--) {
139 /* Due to quantization, we will usually find that many of the input
140 * coefficients are zero, especially the AC terms. We can exploit this
141 * by short-circuiting the IDCT calculation for any column in which all
142 * the AC terms are zero. In that case each output is equal to the
143 * DC coefficient (with scale factor as needed).
144 * With typical images and quantization tables, half or more of the
145 * column DCT calculations can be simplified this way.
146 */
147
148 if (inptr[DCTSIZE*1] == 0 && inptr[DCTSIZE*2] == 0 &&
149 inptr[DCTSIZE*3] == 0 && inptr[DCTSIZE*4] == 0 &&
150 inptr[DCTSIZE*5] == 0 && inptr[DCTSIZE*6] == 0 &&
151 inptr[DCTSIZE*7] == 0) {
152 /* AC terms all zero */
153 FAST_FLOAT dcval = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]);
154
155 wsptr[DCTSIZE*0] = dcval;
156 wsptr[DCTSIZE*1] = dcval;
157 wsptr[DCTSIZE*2] = dcval;
158 wsptr[DCTSIZE*3] = dcval;
159 wsptr[DCTSIZE*4] = dcval;
160 wsptr[DCTSIZE*5] = dcval;
161 wsptr[DCTSIZE*6] = dcval;
162 wsptr[DCTSIZE*7] = dcval;
163
164 inptr++; /* advance pointers to next column */
165 quantptr++;
166 wsptr++;
167 continue;
168 }
169
170 /* Even part */
171
172 tmp0 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]);
173 tmp1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]);
174 tmp2 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]);
175 tmp3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]);
176
177 tmp10 = tmp0 + tmp2; /* phase 3 */
178 tmp11 = tmp0 - tmp2;
179
180 tmp13 = tmp1 + tmp3; /* phases 5-3 */
181 tmp12 = (tmp1 - tmp3) * ((FAST_FLOAT) 1.414213562) - tmp13; /* 2*c4 */
182
183 tmp0 = tmp10 + tmp13; /* phase 2 */
184 tmp3 = tmp10 - tmp13;
185 tmp1 = tmp11 + tmp12;
186 tmp2 = tmp11 - tmp12;
187
188 /* Odd part */
189
190 tmp4 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]);
191 tmp5 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]);
192 tmp6 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]);
193 tmp7 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]);
194
195 z13 = tmp6 + tmp5; /* phase 6 */
196 z10 = tmp6 - tmp5;
197 z11 = tmp4 + tmp7;
198 z12 = tmp4 - tmp7;
199
200 tmp7 = z11 + z13; /* phase 5 */
201 tmp11 = (z11 - z13) * ((FAST_FLOAT) 1.414213562); /* 2*c4 */
202
203 z5 = (z10 + z12) * ((FAST_FLOAT) 1.847759065); /* 2*c2 */
204 tmp10 = ((FAST_FLOAT) 1.082392200) * z12 - z5; /* 2*(c2-c6) */
205 tmp12 = ((FAST_FLOAT) -2.613125930) * z10 + z5; /* -2*(c2+c6) */
206
207 tmp6 = tmp12 - tmp7; /* phase 2 */
208 tmp5 = tmp11 - tmp6;
209 tmp4 = tmp10 + tmp5;
210
211 wsptr[DCTSIZE*0] = tmp0 + tmp7;
212 wsptr[DCTSIZE*7] = tmp0 - tmp7;
213 wsptr[DCTSIZE*1] = tmp1 + tmp6;
214 wsptr[DCTSIZE*6] = tmp1 - tmp6;
215 wsptr[DCTSIZE*2] = tmp2 + tmp5;
216 wsptr[DCTSIZE*5] = tmp2 - tmp5;
217 wsptr[DCTSIZE*4] = tmp3 + tmp4;
218 wsptr[DCTSIZE*3] = tmp3 - tmp4;
219
220 inptr++; /* advance pointers to next column */
221 quantptr++;
222 wsptr++;
223 }
224
225 /* Pass 2: process rows from work array, store into output array. */
226 /* Note that we must descale the results by a factor of 8 == 2**3. */
227
228 wsptr = workspace;
229 outptr = output_buf;
230 for (ctr = 0; ctr < DCTSIZE; ctr++) {
231 /* Rows of zeroes can be exploited in the same way as we did with columns.
232 * However, the column calculation has created many nonzero AC terms, so
233 * the simplification applies less often (typically 5% to 10% of the time).
234 * And testing floats for zero is relatively expensive, so we don't bother.
235 */
236
237 /* Even part */
238
239 tmp10 = wsptr[0] + wsptr[4];
240 tmp11 = wsptr[0] - wsptr[4];
241
242 tmp13 = wsptr[2] + wsptr[6];
243 tmp12 = (wsptr[2] - wsptr[6]) * ((FAST_FLOAT) 1.414213562) - tmp13;
244
245 tmp0 = tmp10 + tmp13;
246 tmp3 = tmp10 - tmp13;
247 tmp1 = tmp11 + tmp12;
248 tmp2 = tmp11 - tmp12;
249
250 /* Odd part */
251
252 z13 = wsptr[5] + wsptr[3];
253 z10 = wsptr[5] - wsptr[3];
254 z11 = wsptr[1] + wsptr[7];
255 z12 = wsptr[1] - wsptr[7];
256
257 tmp7 = z11 + z13;
258 tmp11 = (z11 - z13) * ((FAST_FLOAT) 1.414213562);
259
260 z5 = (z10 + z12) * ((FAST_FLOAT) 1.847759065); /* 2*c2 */
261 tmp10 = ((FAST_FLOAT) 1.082392200) * z12 - z5; /* 2*(c2-c6) */
262 tmp12 = ((FAST_FLOAT) -2.613125930) * z10 + z5; /* -2*(c2+c6) */
263
264 tmp6 = tmp12 - tmp7;
265 tmp5 = tmp11 - tmp6;
266 tmp4 = tmp10 + tmp5;
267
268 /* Final output stage: scale down by a factor of 8 and range-limit */
269
270 outptr[0] = descale_and_clamp((int)(tmp0 + tmp7), 3);
271 outptr[7] = descale_and_clamp((int)(tmp0 - tmp7), 3);
272 outptr[1] = descale_and_clamp((int)(tmp1 + tmp6), 3);
273 outptr[6] = descale_and_clamp((int)(tmp1 - tmp6), 3);
274 outptr[2] = descale_and_clamp((int)(tmp2 + tmp5), 3);
275 outptr[5] = descale_and_clamp((int)(tmp2 - tmp5), 3);
276 outptr[4] = descale_and_clamp((int)(tmp3 + tmp4), 3);
277 outptr[3] = descale_and_clamp((int)(tmp3 - tmp4), 3);
278
279
280 wsptr += DCTSIZE; /* advance pointer to next row */
281 outptr += stride;
282 }
283 }
284
285