/*
 * jidctflt.c
 *
 * Copyright (C) 1994-1998, Thomas G. Lane.
 * This file is part of the Independent JPEG Group's software.
 *
 * The authors make NO WARRANTY or representation, either express or implied,
 * with respect to this software, its quality, accuracy, merchantability, or
 * fitness for a particular purpose.  This software is provided "AS IS", and you,
 * its user, assume the entire risk as to its quality and accuracy.
 *
 * This software is copyright (C) 1991-1998, Thomas G. Lane.
 * All Rights Reserved except as specified below.
 *
 * Permission is hereby granted to use, copy, modify, and distribute this
 * software (or portions thereof) for any purpose, without fee, subject to these
 * conditions:
 * (1) If any part of the source code for this software is distributed, then this
 * README file must be included, with this copyright and no-warranty notice
 * unaltered; and any additions, deletions, or changes to the original files
 * must be clearly indicated in accompanying documentation.
 * (2) If only executable code is distributed, then the accompanying
 * documentation must state that "this software is based in part on the work of
 * the Independent JPEG Group".
 * (3) Permission for use of this software is granted only if the user accepts
 * full responsibility for any undesirable consequences; the authors accept
 * NO LIABILITY for damages of any kind.
 *
 * These conditions apply to any software derived from or based on the IJG code,
 * not just to the unmodified library.  If you use our work, you ought to
 * acknowledge us.
 *
 * Permission is NOT granted for the use of any IJG author's name or company name
 * in advertising or publicity relating to this software or products derived from
 * it.  This software may be referred to only as "the Independent JPEG Group's
 * software".
 *
 * We specifically permit and encourage the use of this software as the basis of
 * commercial products, provided that all warranty or liability claims are
 * assumed by the product vendor.
 *
 *
 * This file contains a floating-point implementation of the
 * inverse DCT (Discrete Cosine Transform).  In the IJG code, this routine
 * must also perform dequantization of the input coefficients.
 *
 * This implementation should be more accurate than either of the integer
 * IDCT implementations.  However, it may not give the same results on all
 * machines because of differences in roundoff behavior.  Speed will depend
 * on the hardware's floating point capacity.
 *
 * A 2-D IDCT can be done by 1-D IDCT on each column followed by 1-D IDCT
 * on each row (or vice versa, but it's more convenient to emit a row at
 * a time).  Direct algorithms are also available, but they are much more
 * complex and seem not to be any faster when reduced to code.
 *
 * This implementation is based on Arai, Agui, and Nakajima's algorithm for
 * scaled DCT.  Their original paper (Trans. IEICE E-71(11):1095) is in
 * Japanese, but the algorithm is described in the Pennebaker & Mitchell
 * JPEG textbook (see REFERENCES section in file README).  The following code
 * is based directly on figure 4-8 in P&M.
 * While an 8-point DCT cannot be done in less than 11 multiplies, it is
 * possible to arrange the computation so that many of the multiplies are
 * simple scalings of the final outputs.  These multiplies can then be
 * folded into the multiplications or divisions by the JPEG quantization
 * table entries.  The AA&N method leaves only 5 multiplies and 29 adds
 * to be done in the DCT itself.
 * The primary disadvantage of this method is that with a fixed-point
 * implementation, accuracy is lost due to imprecise representation of the
 * scaled quantization values.  However, that problem does not arise if
 * we use floating point arithmetic.
*/

#include <stdint.h>
#include "tinyjpeg-internal.h"

#define FAST_FLOAT float
#define DCTSIZE	   8
#define DCTSIZE2   (DCTSIZE * DCTSIZE)

#define DEQUANTIZE(coef, quantval)  (((FAST_FLOAT) (coef)) * (quantval))

#if defined(__GNUC__) && (defined(__i686__) || defined(__x86_64__))

static inline unsigned char descale_and_clamp(int x, int shift)
{
	__asm__ (
		"add %3,%1\n"
		"\tsar %2,%1\n"
		"\tsub $-128,%1\n"
		"\tcmovl %5,%1\n"	/* Use the sub to compare to 0 */
		"\tcmpl %4,%1\n"
		"\tcmovg %4,%1\n"
		: "=r"(x)
		: "0"(x), "Ic"((unsigned char)shift), "ir" (1U << (shift - 1)), "r" (0xff), "r" (0)
		);
	return x;
}

#else
static inline unsigned char descale_and_clamp(int x, int shift)
{
	x += 1UL << (shift - 1);
	if (x < 0)
		x = (x >> shift) | ((~(0UL)) << (32 - (shift)));
	else
		x >>= shift;
	x += 128;
	if (x > 255)
		return 255;
	if (x < 0)
		return 0;
	return x;
}
#endif

/*
 * Perform dequantization and inverse DCT on one block of coefficients.
 */

void tinyjpeg_idct_float(struct component *compptr, uint8_t *output_buf, int stride)
{
	FAST_FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
	FAST_FLOAT tmp10, tmp11, tmp12, tmp13;
	FAST_FLOAT z5, z10, z11, z12, z13;
	int16_t *inptr;
	FAST_FLOAT *quantptr;
	FAST_FLOAT *wsptr;
	uint8_t *outptr;
	int ctr;
	FAST_FLOAT workspace[DCTSIZE2]; /* buffers data between passes */

	/* Pass 1: process columns from input, store into work array. */

	inptr = compptr->DCT;
	quantptr = compptr->Q_table;
	wsptr = workspace;
	for (ctr = DCTSIZE; ctr > 0; ctr--) {
		/* Due to quantization, we will usually find that many of the input
		 * coefficients are zero, especially the AC terms.  We can exploit this
		 * by short-circuiting the IDCT calculation for any column in which all
		 * the AC terms are zero.  In that case each output is equal to the
		 * DC coefficient (with scale factor as needed).
		 * With typical images and quantization tables, half or more of the
		 * column DCT calculations can be simplified this way.
		 */

		if (inptr[DCTSIZE*1] == 0 && inptr[DCTSIZE*2] == 0 &&
				inptr[DCTSIZE*3] == 0 && inptr[DCTSIZE*4] == 0 &&
				inptr[DCTSIZE*5] == 0 && inptr[DCTSIZE*6] == 0 &&
				inptr[DCTSIZE*7] == 0) {
			/* AC terms all zero */
			FAST_FLOAT dcval = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]);

			wsptr[DCTSIZE*0] = dcval;
			wsptr[DCTSIZE*1] = dcval;
			wsptr[DCTSIZE*2] = dcval;
			wsptr[DCTSIZE*3] = dcval;
			wsptr[DCTSIZE*4] = dcval;
			wsptr[DCTSIZE*5] = dcval;
			wsptr[DCTSIZE*6] = dcval;
			wsptr[DCTSIZE*7] = dcval;

			inptr++;			/* advance pointers to next column */
			quantptr++;
			wsptr++;
			continue;
		}

		/* Even part */

		tmp0 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]);
		tmp1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]);
		tmp2 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]);
		tmp3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]);

		tmp10 = tmp0 + tmp2;	/* phase 3 */
		tmp11 = tmp0 - tmp2;

		tmp13 = tmp1 + tmp3;	/* phases 5-3 */
		tmp12 = (tmp1 - tmp3) * ((FAST_FLOAT) 1.414213562) - tmp13; /* 2*c4 */

		tmp0 = tmp10 + tmp13;	/* phase 2 */
		tmp3 = tmp10 - tmp13;
		tmp1 = tmp11 + tmp12;
		tmp2 = tmp11 - tmp12;

		/* Odd part */

		tmp4 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]);
		tmp5 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]);
		tmp6 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]);
		tmp7 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]);

		z13 = tmp6 + tmp5;		/* phase 6 */
		z10 = tmp6 - tmp5;
		z11 = tmp4 + tmp7;
		z12 = tmp4 - tmp7;

		tmp7 = z11 + z13;		/* phase 5 */
		tmp11 = (z11 - z13) * ((FAST_FLOAT) 1.414213562); /* 2*c4 */

		z5 = (z10 + z12) * ((FAST_FLOAT) 1.847759065); /* 2*c2 */
		tmp10 = ((FAST_FLOAT) 1.082392200) * z12 - z5; /* 2*(c2-c6) */
		tmp12 = ((FAST_FLOAT) -2.613125930) * z10 + z5; /* -2*(c2+c6) */

		tmp6 = tmp12 - tmp7;	/* phase 2 */
		tmp5 = tmp11 - tmp6;
		tmp4 = tmp10 + tmp5;

		wsptr[DCTSIZE*0] = tmp0 + tmp7;
		wsptr[DCTSIZE*7] = tmp0 - tmp7;
		wsptr[DCTSIZE*1] = tmp1 + tmp6;
		wsptr[DCTSIZE*6] = tmp1 - tmp6;
		wsptr[DCTSIZE*2] = tmp2 + tmp5;
		wsptr[DCTSIZE*5] = tmp2 - tmp5;
		wsptr[DCTSIZE*4] = tmp3 + tmp4;
		wsptr[DCTSIZE*3] = tmp3 - tmp4;

		inptr++;			/* advance pointers to next column */
		quantptr++;
		wsptr++;
	}

	/* Pass 2: process rows from work array, store into output array. */
	/* Note that we must descale the results by a factor of 8 == 2**3. */

	wsptr = workspace;
	outptr = output_buf;
	for (ctr = 0; ctr < DCTSIZE; ctr++) {
		/* Rows of zeroes can be exploited in the same way as we did with columns.
		 * However, the column calculation has created many nonzero AC terms, so
		 * the simplification applies less often (typically 5% to 10% of the time).
		 * And testing floats for zero is relatively expensive, so we don't bother.
		 */

		/* Even part */

		tmp10 = wsptr[0] + wsptr[4];
		tmp11 = wsptr[0] - wsptr[4];

		tmp13 = wsptr[2] + wsptr[6];
		tmp12 = (wsptr[2] - wsptr[6]) * ((FAST_FLOAT) 1.414213562) - tmp13;

		tmp0 = tmp10 + tmp13;
		tmp3 = tmp10 - tmp13;
		tmp1 = tmp11 + tmp12;
		tmp2 = tmp11 - tmp12;

		/* Odd part */

		z13 = wsptr[5] + wsptr[3];
		z10 = wsptr[5] - wsptr[3];
		z11 = wsptr[1] + wsptr[7];
		z12 = wsptr[1] - wsptr[7];

		tmp7 = z11 + z13;
		tmp11 = (z11 - z13) * ((FAST_FLOAT) 1.414213562);

		z5 = (z10 + z12) * ((FAST_FLOAT) 1.847759065); /* 2*c2 */
		tmp10 = ((FAST_FLOAT) 1.082392200) * z12 - z5; /* 2*(c2-c6) */
		tmp12 = ((FAST_FLOAT) -2.613125930) * z10 + z5; /* -2*(c2+c6) */

		tmp6 = tmp12 - tmp7;
		tmp5 = tmp11 - tmp6;
		tmp4 = tmp10 + tmp5;

		/* Final output stage: scale down by a factor of 8 and range-limit */

		outptr[0] = descale_and_clamp((int)(tmp0 + tmp7), 3);
		outptr[7] = descale_and_clamp((int)(tmp0 - tmp7), 3);
		outptr[1] = descale_and_clamp((int)(tmp1 + tmp6), 3);
		outptr[6] = descale_and_clamp((int)(tmp1 - tmp6), 3);
		outptr[2] = descale_and_clamp((int)(tmp2 + tmp5), 3);
		outptr[5] = descale_and_clamp((int)(tmp2 - tmp5), 3);
		outptr[4] = descale_and_clamp((int)(tmp3 + tmp4), 3);
		outptr[3] = descale_and_clamp((int)(tmp3 - tmp4), 3);


		wsptr += DCTSIZE;		/* advance pointer to next row */
		outptr += stride;
	}
}