diff options
| author | André Klitzing <aklitzing@gmail.com> | 2017-07-28 18:38:55 +0200 |
|---|---|---|
| committer | André Klitzing <aklitzing@gmail.com> | 2017-09-28 10:06:26 +0000 |
| commit | 9c13ec5fa392930e9a1880ad20c11ba47fdc54ba (patch) | |
| tree | 71355136ba851d004e7f902d13077fdc9388f607 | |
| parent | 0731c092d107415cc300be2209cd2bfb417950e9 (diff) | |
Replace bundled libjpeg by libjpeg-turbo 1.5.2
Most linux distribution uses libjpeg-turbo as a replacement for libjpeg.
Since Qt will be linked against -turbo more frequently than the "original"
libjpeg we should use it for the bundled copy, too.
Also add an import script.
[ChangeLog][Third-Party Code] Replaced bundled libjpeg by libjpeg-turbo 1.5.2
Change-Id: I41081db74f194cbc288820fc783c9fef16545efd
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
133 files changed, 25518 insertions, 31687 deletions
diff --git a/src/3rdparty/libjpeg.pri b/src/3rdparty/libjpeg.pri index 118cc60bcc..a61f28dc5a 100644 --- a/src/3rdparty/libjpeg.pri +++ b/src/3rdparty/libjpeg.pri @@ -1,58 +1,69 @@ winrt: DEFINES += NO_GETENV +DEFINES += \ + C_ARITH_CODING_SUPPORTED=1 \ + D_ARITH_CODING_SUPPORTED=1 \ + BITS_IN_JSAMPLE=8 \ + JPEG_LIB_VERSION=80 \ + SIZEOF_SIZE_T=__SIZEOF_SIZE_T__ + #Disable warnings in 3rdparty code due to unused arguments contains(QMAKE_CC, gcc): { QMAKE_CFLAGS_WARN_ON += -Wno-unused-parameter -Wno-main } -INCLUDEPATH += $$PWD/libjpeg +INCLUDEPATH += $$PWD/libjpeg/src SOURCES += \ - $$PWD/libjpeg/jaricom.c \ - $$PWD/libjpeg/jcapimin.c \ - $$PWD/libjpeg/jcapistd.c \ - $$PWD/libjpeg/jcarith.c \ - $$PWD/libjpeg/jccoefct.c \ - $$PWD/libjpeg/jccolor.c \ - $$PWD/libjpeg/jcdctmgr.c \ - $$PWD/libjpeg/jchuff.c \ - $$PWD/libjpeg/jcinit.c \ - $$PWD/libjpeg/jcmainct.c \ - $$PWD/libjpeg/jcmarker.c \ - $$PWD/libjpeg/jcmaster.c \ - $$PWD/libjpeg/jcomapi.c \ - $$PWD/libjpeg/jcparam.c \ - $$PWD/libjpeg/jcprepct.c \ - $$PWD/libjpeg/jcsample.c \ - $$PWD/libjpeg/jctrans.c \ - $$PWD/libjpeg/jdapimin.c \ - $$PWD/libjpeg/jdapistd.c \ - $$PWD/libjpeg/jdarith.c \ - $$PWD/libjpeg/jdatadst.c \ - $$PWD/libjpeg/jdatasrc.c \ - $$PWD/libjpeg/jdcoefct.c \ - $$PWD/libjpeg/jdcolor.c \ - $$PWD/libjpeg/jddctmgr.c \ - $$PWD/libjpeg/jdhuff.c \ - $$PWD/libjpeg/jdinput.c \ - $$PWD/libjpeg/jdmainct.c \ - $$PWD/libjpeg/jdmarker.c \ - $$PWD/libjpeg/jdmaster.c \ - $$PWD/libjpeg/jdmerge.c \ - $$PWD/libjpeg/jdpostct.c \ - $$PWD/libjpeg/jdsample.c \ - $$PWD/libjpeg/jdtrans.c \ - $$PWD/libjpeg/jerror.c \ - $$PWD/libjpeg/jfdctflt.c \ - $$PWD/libjpeg/jfdctfst.c \ - $$PWD/libjpeg/jfdctint.c \ - $$PWD/libjpeg/jidctflt.c \ - $$PWD/libjpeg/jidctfst.c \ - $$PWD/libjpeg/jidctint.c \ - $$PWD/libjpeg/jquant1.c \ - $$PWD/libjpeg/jquant2.c \ - $$PWD/libjpeg/jutils.c \ - $$PWD/libjpeg/jmemmgr.c \ - $$PWD/libjpeg/jmemnobs.c + $$PWD/libjpeg/src/jaricom.c \ + $$PWD/libjpeg/src/jcapimin.c \ + $$PWD/libjpeg/src/jcapistd.c \ + $$PWD/libjpeg/src/jcarith.c \ + $$PWD/libjpeg/src/jccoefct.c \ + $$PWD/libjpeg/src/jccolor.c \ + $$PWD/libjpeg/src/jcdctmgr.c \ + $$PWD/libjpeg/src/jchuff.c \ + $$PWD/libjpeg/src/jcinit.c \ + $$PWD/libjpeg/src/jcmainct.c \ + $$PWD/libjpeg/src/jcmarker.c \ + $$PWD/libjpeg/src/jcmaster.c \ + $$PWD/libjpeg/src/jcomapi.c \ + $$PWD/libjpeg/src/jcparam.c \ + $$PWD/libjpeg/src/jcprepct.c \ + $$PWD/libjpeg/src/jcsample.c \ + $$PWD/libjpeg/src/jctrans.c \ + $$PWD/libjpeg/src/jdapimin.c \ + $$PWD/libjpeg/src/jdapistd.c \ + $$PWD/libjpeg/src/jdarith.c \ + $$PWD/libjpeg/src/jdatadst.c \ + $$PWD/libjpeg/src/jdatasrc.c \ + $$PWD/libjpeg/src/jdcoefct.c \ + $$PWD/libjpeg/src/jdcolor.c \ + $$PWD/libjpeg/src/jddctmgr.c \ + $$PWD/libjpeg/src/jdhuff.c \ + $$PWD/libjpeg/src/jdinput.c \ + $$PWD/libjpeg/src/jdmainct.c \ + $$PWD/libjpeg/src/jdmarker.c \ + $$PWD/libjpeg/src/jdmaster.c \ + $$PWD/libjpeg/src/jdmerge.c \ + $$PWD/libjpeg/src/jdpostct.c \ + $$PWD/libjpeg/src/jdsample.c \ + $$PWD/libjpeg/src/jdtrans.c \ + $$PWD/libjpeg/src/jerror.c \ + $$PWD/libjpeg/src/jfdctflt.c \ + $$PWD/libjpeg/src/jfdctfst.c \ + $$PWD/libjpeg/src/jfdctint.c \ + $$PWD/libjpeg/src/jidctflt.c \ + $$PWD/libjpeg/src/jidctfst.c \ + $$PWD/libjpeg/src/jidctint.c \ + $$PWD/libjpeg/src/jquant1.c \ + $$PWD/libjpeg/src/jquant2.c \ + $$PWD/libjpeg/src/jutils.c \ + $$PWD/libjpeg/src/jmemmgr.c \ + $$PWD/libjpeg/src/jsimd_none.c \ + $$PWD/libjpeg/src/jcphuff.c \ + $$PWD/libjpeg/src/jidctred.c \ + $$PWD/libjpeg/src/jdphuff.c \ + $$PWD/libjpeg/src/jmemnobs.c TR_EXCLUDE += $$PWD/* diff --git a/src/3rdparty/libjpeg/LICENSE b/src/3rdparty/libjpeg/LICENSE index 797a6d5668..0572390635 100644 --- a/src/3rdparty/libjpeg/LICENSE +++ b/src/3rdparty/libjpeg/LICENSE @@ -1,50 +1,139 @@ -from qtbase/src/3rdparty/libjpeg/README: - -LEGAL ISSUES -============ - -In plain English: - -1. We don't promise that this software works. (But if you find any bugs, - please let us know!) -2. You can use this software for whatever you want. You don't have to pay us. -3. You may not pretend that you wrote this software. If you use it in a - program, you must acknowledge somewhere in your documentation that - you've used the IJG code. - -In legalese: - -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. +libjpeg-turbo Licenses +====================== + +libjpeg-turbo is covered by three compatible BSD-style open source licenses: + +- The IJG (Independent JPEG Group) License, which is listed in + [README.ijg](README.ijg) + + This license applies to the libjpeg API library and associated programs + (any code inherited from libjpeg, and any modifications to that code.) + +- The Modified (3-clause) BSD License, which is listed below + + This license covers the TurboJPEG API library and associated programs. + +- The zlib License, which is listed below + + This license is a subset of the other two, and it covers the libjpeg-turbo + SIMD extensions. + + +Complying with the libjpeg-turbo Licenses +========================================= + +This section provides a roll-up of the libjpeg-turbo licensing terms, to the +best of our understanding. + +1. If you are distributing a modified version of the libjpeg-turbo source, + then: + + 1. You cannot alter or remove any existing copyright or license notices + from the source. + + **Origin** + - Clause 1 of the IJG License + - Clause 1 of the Modified BSD License + - Clauses 1 and 3 of the zlib License + + 2. You must add your own copyright notice to the header of each source + file you modified, so others can tell that you modified that file (if + there is not an existing copyright header in that file, then you can + simply add a notice stating that you modified the file.) + + **Origin** + - Clause 1 of the IJG License + - Clause 2 of the zlib License + + 3. You must include the IJG README file, and you must not alter any of the + copyright or license text in that file. + + **Origin** + - Clause 1 of the IJG License + +2. If you are distributing only libjpeg-turbo binaries without the source, or + if you are distributing an application that statically links with + libjpeg-turbo, then: + + 1. Your product documentation must include a message stating: + + This software is based in part on the work of the Independent JPEG + Group. + + **Origin** + - Clause 2 of the IJG license + + 2. If your binary distribution includes or uses the TurboJPEG API, then + your product documentation must include the text of the Modified BSD + License. + + **Origin** + - Clause 2 of the Modified BSD License + +3. You cannot use the name of the IJG or The libjpeg-turbo Project or the + contributors thereof in advertising, publicity, etc. + + **Origin** + - IJG License + - Clause 3 of the Modified BSD License + +4. The IJG and The libjpeg-turbo Project do not warrant libjpeg-turbo to be + free of defects, nor do we accept any liability for undesirable + consequences resulting from your use of the software. + + **Origin** + - IJG License + - Modified BSD License + - zlib License + + +The Modified (3-clause) BSD License +=================================== + +Copyright (C)\<YEAR\> \<AUTHOR\>. All Rights Reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + +- Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. +- Redistributions in binary form must reproduce the above copyright notice, + this list of conditions and the following disclaimer in the documentation + and/or other materials provided with the distribution. +- Neither the name of the libjpeg-turbo Project nor the names of its + contributors may be used to endorse or promote products derived from this + software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS", +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + + +The zlib License +================ + +Copyright (C) \<YEAR\>, \<AUTHOR\>. + +This software is provided 'as-is', without any express or implied +warranty. In no event will the authors be held liable for any damages +arising from the use of this software. + +Permission is granted to anyone to use this software for any purpose, +including commercial applications, and to alter it and redistribute it +freely, subject to the following restrictions: + +1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. +2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. +3. This notice may not be removed or altered from any source distribution. diff --git a/src/3rdparty/libjpeg/cderror.h b/src/3rdparty/libjpeg/cderror.h deleted file mode 100644 index e19c475c5c..0000000000 --- a/src/3rdparty/libjpeg/cderror.h +++ /dev/null @@ -1,134 +0,0 @@ -/* - * cderror.h - * - * Copyright (C) 1994-1997, Thomas G. Lane. - * Modified 2009 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file defines the error and message codes for the cjpeg/djpeg - * applications. These strings are not needed as part of the JPEG library - * proper. - * Edit this file to add new codes, or to translate the message strings to - * some other language. - */ - -/* - * To define the enum list of message codes, include this file without - * defining macro JMESSAGE. To create a message string table, include it - * again with a suitable JMESSAGE definition (see jerror.c for an example). - */ -#ifndef JMESSAGE -#ifndef CDERROR_H -#define CDERROR_H -/* First time through, define the enum list */ -#define JMAKE_ENUM_LIST -#else -/* Repeated inclusions of this file are no-ops unless JMESSAGE is defined */ -#define JMESSAGE(code,string) -#endif /* CDERROR_H */ -#endif /* JMESSAGE */ - -#ifdef JMAKE_ENUM_LIST - -typedef enum { - -#define JMESSAGE(code,string) code , - -#endif /* JMAKE_ENUM_LIST */ - -JMESSAGE(JMSG_FIRSTADDONCODE=1000, NULL) /* Must be first entry! */ - -#ifdef BMP_SUPPORTED -JMESSAGE(JERR_BMP_BADCMAP, "Unsupported BMP colormap format") -JMESSAGE(JERR_BMP_BADDEPTH, "Only 8- and 24-bit BMP files are supported") -JMESSAGE(JERR_BMP_BADHEADER, "Invalid BMP file: bad header length") -JMESSAGE(JERR_BMP_BADPLANES, "Invalid BMP file: biPlanes not equal to 1") -JMESSAGE(JERR_BMP_COLORSPACE, "BMP output must be grayscale or RGB") -JMESSAGE(JERR_BMP_COMPRESSED, "Sorry, compressed BMPs not yet supported") -JMESSAGE(JERR_BMP_EMPTY, "Empty BMP image") -JMESSAGE(JERR_BMP_NOT, "Not a BMP file - does not start with BM") -JMESSAGE(JTRC_BMP, "%ux%u 24-bit BMP image") -JMESSAGE(JTRC_BMP_MAPPED, "%ux%u 8-bit colormapped BMP image") -JMESSAGE(JTRC_BMP_OS2, "%ux%u 24-bit OS2 BMP image") -JMESSAGE(JTRC_BMP_OS2_MAPPED, "%ux%u 8-bit colormapped OS2 BMP image") -#endif /* BMP_SUPPORTED */ - -#ifdef GIF_SUPPORTED -JMESSAGE(JERR_GIF_BUG, "GIF output got confused") -JMESSAGE(JERR_GIF_CODESIZE, "Bogus GIF codesize %d") -JMESSAGE(JERR_GIF_COLORSPACE, "GIF output must be grayscale or RGB") -JMESSAGE(JERR_GIF_IMAGENOTFOUND, "Too few images in GIF file") -JMESSAGE(JERR_GIF_NOT, "Not a GIF file") -JMESSAGE(JTRC_GIF, "%ux%ux%d GIF image") -JMESSAGE(JTRC_GIF_BADVERSION, - "Warning: unexpected GIF version number '%c%c%c'") -JMESSAGE(JTRC_GIF_EXTENSION, "Ignoring GIF extension block of type 0x%02x") -JMESSAGE(JTRC_GIF_NONSQUARE, "Caution: nonsquare pixels in input") -JMESSAGE(JWRN_GIF_BADDATA, "Corrupt data in GIF file") -JMESSAGE(JWRN_GIF_CHAR, "Bogus char 0x%02x in GIF file, ignoring") -JMESSAGE(JWRN_GIF_ENDCODE, "Premature end of GIF image") -JMESSAGE(JWRN_GIF_NOMOREDATA, "Ran out of GIF bits") -#endif /* GIF_SUPPORTED */ - -#ifdef PPM_SUPPORTED -JMESSAGE(JERR_PPM_COLORSPACE, "PPM output must be grayscale or RGB") -JMESSAGE(JERR_PPM_NONNUMERIC, "Nonnumeric data in PPM file") -JMESSAGE(JERR_PPM_NOT, "Not a PPM/PGM file") -JMESSAGE(JTRC_PGM, "%ux%u PGM image") -JMESSAGE(JTRC_PGM_TEXT, "%ux%u text PGM image") -JMESSAGE(JTRC_PPM, "%ux%u PPM image") -JMESSAGE(JTRC_PPM_TEXT, "%ux%u text PPM image") -#endif /* PPM_SUPPORTED */ - -#ifdef RLE_SUPPORTED -JMESSAGE(JERR_RLE_BADERROR, "Bogus error code from RLE library") -JMESSAGE(JERR_RLE_COLORSPACE, "RLE output must be grayscale or RGB") -JMESSAGE(JERR_RLE_DIMENSIONS, "Image dimensions (%ux%u) too large for RLE") -JMESSAGE(JERR_RLE_EMPTY, "Empty RLE file") -JMESSAGE(JERR_RLE_EOF, "Premature EOF in RLE header") -JMESSAGE(JERR_RLE_MEM, "Insufficient memory for RLE header") -JMESSAGE(JERR_RLE_NOT, "Not an RLE file") -JMESSAGE(JERR_RLE_TOOMANYCHANNELS, "Cannot handle %d output channels for RLE") -JMESSAGE(JERR_RLE_UNSUPPORTED, "Cannot handle this RLE setup") -JMESSAGE(JTRC_RLE, "%ux%u full-color RLE file") -JMESSAGE(JTRC_RLE_FULLMAP, "%ux%u full-color RLE file with map of length %d") -JMESSAGE(JTRC_RLE_GRAY, "%ux%u grayscale RLE file") -JMESSAGE(JTRC_RLE_MAPGRAY, "%ux%u grayscale RLE file with map of length %d") -JMESSAGE(JTRC_RLE_MAPPED, "%ux%u colormapped RLE file with map of length %d") -#endif /* RLE_SUPPORTED */ - -#ifdef TARGA_SUPPORTED -JMESSAGE(JERR_TGA_BADCMAP, "Unsupported Targa colormap format") -JMESSAGE(JERR_TGA_BADPARMS, "Invalid or unsupported Targa file") -JMESSAGE(JERR_TGA_COLORSPACE, "Targa output must be grayscale or RGB") -JMESSAGE(JTRC_TGA, "%ux%u RGB Targa image") -JMESSAGE(JTRC_TGA_GRAY, "%ux%u grayscale Targa image") -JMESSAGE(JTRC_TGA_MAPPED, "%ux%u colormapped Targa image") -#else -JMESSAGE(JERR_TGA_NOTCOMP, "Targa support was not compiled") -#endif /* TARGA_SUPPORTED */ - -JMESSAGE(JERR_BAD_CMAP_FILE, - "Color map file is invalid or of unsupported format") -JMESSAGE(JERR_TOO_MANY_COLORS, - "Output file format cannot handle %d colormap entries") -JMESSAGE(JERR_UNGETC_FAILED, "ungetc failed") -#ifdef TARGA_SUPPORTED -JMESSAGE(JERR_UNKNOWN_FORMAT, - "Unrecognized input file format --- perhaps you need -targa") -#else -JMESSAGE(JERR_UNKNOWN_FORMAT, "Unrecognized input file format") -#endif -JMESSAGE(JERR_UNSUPPORTED_FORMAT, "Unsupported output file format") - -#ifdef JMAKE_ENUM_LIST - - JMSG_LASTADDONCODE -} ADDON_MESSAGE_CODE; - -#undef JMAKE_ENUM_LIST -#endif /* JMAKE_ENUM_LIST */ - -/* Zap JMESSAGE macro so that future re-inclusions do nothing by default */ -#undef JMESSAGE diff --git a/src/3rdparty/libjpeg/cdjpeg.h b/src/3rdparty/libjpeg/cdjpeg.h deleted file mode 100644 index ed024ac3ae..0000000000 --- a/src/3rdparty/libjpeg/cdjpeg.h +++ /dev/null @@ -1,187 +0,0 @@ -/* - * cdjpeg.h - * - * Copyright (C) 1994-1997, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains common declarations for the sample applications - * cjpeg and djpeg. It is NOT used by the core JPEG library. - */ - -#define JPEG_CJPEG_DJPEG /* define proper options in jconfig.h */ -#define JPEG_INTERNAL_OPTIONS /* cjpeg.c,djpeg.c need to see xxx_SUPPORTED */ -#include "jinclude.h" -#include "jpeglib.h" -#include "jerror.h" /* get library error codes too */ -#include "cderror.h" /* get application-specific error codes */ - - -/* - * Object interface for cjpeg's source file decoding modules - */ - -typedef struct cjpeg_source_struct * cjpeg_source_ptr; - -struct cjpeg_source_struct { - JMETHOD(void, start_input, (j_compress_ptr cinfo, - cjpeg_source_ptr sinfo)); - JMETHOD(JDIMENSION, get_pixel_rows, (j_compress_ptr cinfo, - cjpeg_source_ptr sinfo)); - JMETHOD(void, finish_input, (j_compress_ptr cinfo, - cjpeg_source_ptr sinfo)); - - FILE *input_file; - - JSAMPARRAY buffer; - JDIMENSION buffer_height; -}; - - -/* - * Object interface for djpeg's output file encoding modules - */ - -typedef struct djpeg_dest_struct * djpeg_dest_ptr; - -struct djpeg_dest_struct { - /* start_output is called after jpeg_start_decompress finishes. - * The color map will be ready at this time, if one is needed. - */ - JMETHOD(void, start_output, (j_decompress_ptr cinfo, - djpeg_dest_ptr dinfo)); - /* Emit the specified number of pixel rows from the buffer. */ - JMETHOD(void, put_pixel_rows, (j_decompress_ptr cinfo, - djpeg_dest_ptr dinfo, - JDIMENSION rows_supplied)); - /* Finish up at the end of the image. */ - JMETHOD(void, finish_output, (j_decompress_ptr cinfo, - djpeg_dest_ptr dinfo)); - - /* Target file spec; filled in by djpeg.c after object is created. */ - FILE * output_file; - - /* Output pixel-row buffer. Created by module init or start_output. - * Width is cinfo->output_width * cinfo->output_components; - * height is buffer_height. - */ - JSAMPARRAY buffer; - JDIMENSION buffer_height; -}; - - -/* - * cjpeg/djpeg may need to perform extra passes to convert to or from - * the source/destination file format. The JPEG library does not know - * about these passes, but we'd like them to be counted by the progress - * monitor. We use an expanded progress monitor object to hold the - * additional pass count. - */ - -struct cdjpeg_progress_mgr { - struct jpeg_progress_mgr pub; /* fields known to JPEG library */ - int completed_extra_passes; /* extra passes completed */ - int total_extra_passes; /* total extra */ - /* last printed percentage stored here to avoid multiple printouts */ - int percent_done; -}; - -typedef struct cdjpeg_progress_mgr * cd_progress_ptr; - - -/* Short forms of external names for systems with brain-damaged linkers. */ - -#ifdef NEED_SHORT_EXTERNAL_NAMES -#define jinit_read_bmp jIRdBMP -#define jinit_write_bmp jIWrBMP -#define jinit_read_gif jIRdGIF -#define jinit_write_gif jIWrGIF -#define jinit_read_ppm jIRdPPM -#define jinit_write_ppm jIWrPPM -#define jinit_read_rle jIRdRLE -#define jinit_write_rle jIWrRLE -#define jinit_read_targa jIRdTarga -#define jinit_write_targa jIWrTarga -#define read_quant_tables RdQTables -#define read_scan_script RdScnScript -#define set_quality_ratings SetQRates -#define set_quant_slots SetQSlots -#define set_sample_factors SetSFacts -#define read_color_map RdCMap -#define enable_signal_catcher EnSigCatcher -#define start_progress_monitor StProgMon -#define end_progress_monitor EnProgMon -#define read_stdin RdStdin -#define write_stdout WrStdout -#endif /* NEED_SHORT_EXTERNAL_NAMES */ - -/* Module selection routines for I/O modules. */ - -EXTERN(cjpeg_source_ptr) jinit_read_bmp JPP((j_compress_ptr cinfo)); -EXTERN(djpeg_dest_ptr) jinit_write_bmp JPP((j_decompress_ptr cinfo, - boolean is_os2)); -EXTERN(cjpeg_source_ptr) jinit_read_gif JPP((j_compress_ptr cinfo)); -EXTERN(djpeg_dest_ptr) jinit_write_gif JPP((j_decompress_ptr cinfo)); -EXTERN(cjpeg_source_ptr) jinit_read_ppm JPP((j_compress_ptr cinfo)); -EXTERN(djpeg_dest_ptr) jinit_write_ppm JPP((j_decompress_ptr cinfo)); -EXTERN(cjpeg_source_ptr) jinit_read_rle JPP((j_compress_ptr cinfo)); -EXTERN(djpeg_dest_ptr) jinit_write_rle JPP((j_decompress_ptr cinfo)); -EXTERN(cjpeg_source_ptr) jinit_read_targa JPP((j_compress_ptr cinfo)); -EXTERN(djpeg_dest_ptr) jinit_write_targa JPP((j_decompress_ptr cinfo)); - -/* cjpeg support routines (in rdswitch.c) */ - -EXTERN(boolean) read_quant_tables JPP((j_compress_ptr cinfo, char * filename, - boolean force_baseline)); -EXTERN(boolean) read_scan_script JPP((j_compress_ptr cinfo, char * filename)); -EXTERN(boolean) set_quality_ratings JPP((j_compress_ptr cinfo, char *arg, - boolean force_baseline)); -EXTERN(boolean) set_quant_slots JPP((j_compress_ptr cinfo, char *arg)); -EXTERN(boolean) set_sample_factors JPP((j_compress_ptr cinfo, char *arg)); - -/* djpeg support routines (in rdcolmap.c) */ - -EXTERN(void) read_color_map JPP((j_decompress_ptr cinfo, FILE * infile)); - -/* common support routines (in cdjpeg.c) */ - -EXTERN(void) enable_signal_catcher JPP((j_common_ptr cinfo)); -EXTERN(void) start_progress_monitor JPP((j_common_ptr cinfo, - cd_progress_ptr progress)); -EXTERN(void) end_progress_monitor JPP((j_common_ptr cinfo)); -EXTERN(boolean) keymatch JPP((char * arg, const char * keyword, int minchars)); -EXTERN(FILE *) read_stdin JPP((void)); -EXTERN(FILE *) write_stdout JPP((void)); - -/* miscellaneous useful macros */ - -#ifdef DONT_USE_B_MODE /* define mode parameters for fopen() */ -#define READ_BINARY "r" -#define WRITE_BINARY "w" -#else -#ifdef VMS /* VMS is very nonstandard */ -#define READ_BINARY "rb", "ctx=stm" -#define WRITE_BINARY "wb", "ctx=stm" -#else /* standard ANSI-compliant case */ -#define READ_BINARY "rb" -#define WRITE_BINARY "wb" -#endif -#endif - -#ifndef EXIT_FAILURE /* define exit() codes if not provided */ -#define EXIT_FAILURE 1 -#endif -#ifndef EXIT_SUCCESS -#ifdef VMS -#define EXIT_SUCCESS 1 /* VMS is very nonstandard */ -#else -#define EXIT_SUCCESS 0 -#endif -#endif -#ifndef EXIT_WARNING -#ifdef VMS -#define EXIT_WARNING 1 /* VMS is very nonstandard */ -#else -#define EXIT_WARNING 2 -#endif -#endif diff --git a/src/3rdparty/libjpeg/ckconfig.c b/src/3rdparty/libjpeg/ckconfig.c deleted file mode 100644 index e658623fa5..0000000000 --- a/src/3rdparty/libjpeg/ckconfig.c +++ /dev/null @@ -1,402 +0,0 @@ -/* - * ckconfig.c - * - * Copyright (C) 1991-1994, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - */ - -/* - * This program is intended to help you determine how to configure the JPEG - * software for installation on a particular system. The idea is to try to - * compile and execute this program. If your compiler fails to compile the - * program, make changes as indicated in the comments below. Once you can - * compile the program, run it, and it will produce a "jconfig.h" file for - * your system. - * - * As a general rule, each time you try to compile this program, - * pay attention only to the *first* error message you get from the compiler. - * Many C compilers will issue lots of spurious error messages once they - * have gotten confused. Go to the line indicated in the first error message, - * and read the comments preceding that line to see what to change. - * - * Almost all of the edits you may need to make to this program consist of - * changing a line that reads "#define SOME_SYMBOL" to "#undef SOME_SYMBOL", - * or vice versa. This is called defining or undefining that symbol. - */ - - -/* First we must see if your system has the include files we need. - * We start out with the assumption that your system has all the ANSI-standard - * include files. If you get any error trying to include one of these files, - * undefine the corresponding HAVE_xxx symbol. - */ - -#define HAVE_STDDEF_H /* replace 'define' by 'undef' if error here */ -#ifdef HAVE_STDDEF_H /* next line will be skipped if you undef... */ -#include <stddef.h> -#endif - -#define HAVE_STDLIB_H /* same thing for stdlib.h */ -#ifdef HAVE_STDLIB_H -#include <stdlib.h> -#endif - -#include <stdio.h> /* If you ain't got this, you ain't got C. */ - -/* We have to see if your string functions are defined by - * strings.h (old BSD convention) or string.h (everybody else). - * We try the non-BSD convention first; define NEED_BSD_STRINGS - * if the compiler says it can't find string.h. - */ - -#undef NEED_BSD_STRINGS - -#ifdef NEED_BSD_STRINGS -#include <strings.h> -#else -#include <string.h> -#endif - -/* On some systems (especially older Unix machines), type size_t is - * defined only in the include file <sys/types.h>. If you get a failure - * on the size_t test below, try defining NEED_SYS_TYPES_H. - */ - -#undef NEED_SYS_TYPES_H /* start by assuming we don't need it */ -#ifdef NEED_SYS_TYPES_H -#include <sys/types.h> -#endif - - -/* Usually type size_t is defined in one of the include files we've included - * above. If not, you'll get an error on the "typedef size_t my_size_t;" line. - * In that case, first try defining NEED_SYS_TYPES_H just above. - * If that doesn't work, you'll have to search through your system library - * to figure out which include file defines "size_t". Look for a line that - * says "typedef something-or-other size_t;". Then, change the line below - * that says "#include <someincludefile.h>" to instead include the file - * you found size_t in, and define NEED_SPECIAL_INCLUDE. If you can't find - * type size_t anywhere, try replacing "#include <someincludefile.h>" with - * "typedef unsigned int size_t;". - */ - -#undef NEED_SPECIAL_INCLUDE /* assume we DON'T need it, for starters */ - -#ifdef NEED_SPECIAL_INCLUDE -#include <someincludefile.h> -#endif - -typedef size_t my_size_t; /* The payoff: do we have size_t now? */ - - -/* The next question is whether your compiler supports ANSI-style function - * prototypes. You need to know this in order to choose between using - * makefile.ansi and using makefile.unix. - * The #define line below is set to assume you have ANSI function prototypes. - * If you get an error in this group of lines, undefine HAVE_PROTOTYPES. - */ - -#define HAVE_PROTOTYPES - -#ifdef HAVE_PROTOTYPES -int testfunction (int arg1, int * arg2); /* check prototypes */ - -struct methods_struct { /* check method-pointer declarations */ - int (*error_exit) (char *msgtext); - int (*trace_message) (char *msgtext); - int (*another_method) (void); -}; - -int testfunction (int arg1, int * arg2) /* check definitions */ -{ - return arg2[arg1]; -} - -int test2function (void) /* check void arg list */ -{ - return 0; -} -#endif - - -/* Now we want to find out if your compiler knows what "unsigned char" means. - * If you get an error on the "unsigned char un_char;" line, - * then undefine HAVE_UNSIGNED_CHAR. - */ - -#define HAVE_UNSIGNED_CHAR - -#ifdef HAVE_UNSIGNED_CHAR -unsigned char un_char; -#endif - - -/* Now we want to find out if your compiler knows what "unsigned short" means. - * If you get an error on the "unsigned short un_short;" line, - * then undefine HAVE_UNSIGNED_SHORT. - */ - -#define HAVE_UNSIGNED_SHORT - -#ifdef HAVE_UNSIGNED_SHORT -unsigned short un_short; -#endif - - -/* Now we want to find out if your compiler understands type "void". - * If you get an error anywhere in here, undefine HAVE_VOID. - */ - -#define HAVE_VOID - -#ifdef HAVE_VOID -/* Caution: a C++ compiler will insist on complete prototypes */ -typedef void * void_ptr; /* check void * */ -#ifdef HAVE_PROTOTYPES /* check ptr to function returning void */ -typedef void (*void_func) (int a, int b); -#else -typedef void (*void_func) (); -#endif - -#ifdef HAVE_PROTOTYPES /* check void function result */ -void test3function (void_ptr arg1, void_func arg2) -#else -void test3function (arg1, arg2) - void_ptr arg1; - void_func arg2; -#endif -{ - char * locptr = (char *) arg1; /* check casting to and from void * */ - arg1 = (void *) locptr; - (*arg2) (1, 2); /* check call of fcn returning void */ -} -#endif - - -/* Now we want to find out if your compiler knows what "const" means. - * If you get an error here, undefine HAVE_CONST. - */ - -#define HAVE_CONST - -#ifdef HAVE_CONST -static const int carray[3] = {1, 2, 3}; - -#ifdef HAVE_PROTOTYPES -int test4function (const int arg1) -#else -int test4function (arg1) - const int arg1; -#endif -{ - return carray[arg1]; -} -#endif - - -/* If you get an error or warning about this structure definition, - * define INCOMPLETE_TYPES_BROKEN. - */ - -#undef INCOMPLETE_TYPES_BROKEN - -#ifndef INCOMPLETE_TYPES_BROKEN -typedef struct undefined_structure * undef_struct_ptr; -#endif - - -/* If you get an error about duplicate names, - * define NEED_SHORT_EXTERNAL_NAMES. - */ - -#undef NEED_SHORT_EXTERNAL_NAMES - -#ifndef NEED_SHORT_EXTERNAL_NAMES - -int possibly_duplicate_function () -{ - return 0; -} - -int possibly_dupli_function () -{ - return 1; -} - -#endif - - - -/************************************************************************ - * OK, that's it. You should not have to change anything beyond this - * point in order to compile and execute this program. (You might get - * some warnings, but you can ignore them.) - * When you run the program, it will make a couple more tests that it - * can do automatically, and then it will create jconfig.h and print out - * any additional suggestions it has. - ************************************************************************ - */ - - -#ifdef HAVE_PROTOTYPES -int is_char_signed (int arg) -#else -int is_char_signed (arg) - int arg; -#endif -{ - if (arg == 189) { /* expected result for unsigned char */ - return 0; /* type char is unsigned */ - } - else if (arg != -67) { /* expected result for signed char */ - printf("Hmm, it seems 'char' is not eight bits wide on your machine.\n"); - printf("I fear the JPEG software will not work at all.\n\n"); - } - return 1; /* assume char is signed otherwise */ -} - - -#ifdef HAVE_PROTOTYPES -int is_shifting_signed (long arg) -#else -int is_shifting_signed (arg) - long arg; -#endif -/* See whether right-shift on a long is signed or not. */ -{ - long res = arg >> 4; - - if (res == -0x7F7E80CL) { /* expected result for signed shift */ - return 1; /* right shift is signed */ - } - /* see if unsigned-shift hack will fix it. */ - /* we can't just test exact value since it depends on width of long... */ - res |= (~0L) << (32-4); - if (res == -0x7F7E80CL) { /* expected result now? */ - return 0; /* right shift is unsigned */ - } - printf("Right shift isn't acting as I expect it to.\n"); - printf("I fear the JPEG software will not work at all.\n\n"); - return 0; /* try it with unsigned anyway */ -} - - -#ifdef HAVE_PROTOTYPES -int main (int argc, char ** argv) -#else -int main (argc, argv) - int argc; - char ** argv; -#endif -{ - char signed_char_check = (char) (-67); - FILE *outfile; - - /* Attempt to write jconfig.h */ - if ((outfile = fopen("jconfig.h", "w")) == NULL) { - printf("Failed to write jconfig.h\n"); - return 1; - } - - /* Write out all the info */ - fprintf(outfile, "/* jconfig.h --- generated by ckconfig.c */\n"); - fprintf(outfile, "/* see jconfig.txt for explanations */\n\n"); -#ifdef HAVE_PROTOTYPES - fprintf(outfile, "#define HAVE_PROTOTYPES\n"); -#else - fprintf(outfile, "#undef HAVE_PROTOTYPES\n"); -#endif -#ifdef HAVE_UNSIGNED_CHAR - fprintf(outfile, "#define HAVE_UNSIGNED_CHAR\n"); -#else - fprintf(outfile, "#undef HAVE_UNSIGNED_CHAR\n"); -#endif -#ifdef HAVE_UNSIGNED_SHORT - fprintf(outfile, "#define HAVE_UNSIGNED_SHORT\n"); -#else - fprintf(outfile, "#undef HAVE_UNSIGNED_SHORT\n"); -#endif -#ifdef HAVE_VOID - fprintf(outfile, "/* #define void char */\n"); -#else - fprintf(outfile, "#define void char\n"); -#endif -#ifdef HAVE_CONST - fprintf(outfile, "/* #define const */\n"); -#else - fprintf(outfile, "#define const\n"); -#endif - if (is_char_signed((int) signed_char_check)) - fprintf(outfile, "#undef CHAR_IS_UNSIGNED\n"); - else - fprintf(outfile, "#define CHAR_IS_UNSIGNED\n"); -#ifdef HAVE_STDDEF_H - fprintf(outfile, "#define HAVE_STDDEF_H\n"); -#else - fprintf(outfile, "#undef HAVE_STDDEF_H\n"); -#endif -#ifdef HAVE_STDLIB_H - fprintf(outfile, "#define HAVE_STDLIB_H\n"); -#else - fprintf(outfile, "#undef HAVE_STDLIB_H\n"); -#endif -#ifdef NEED_BSD_STRINGS - fprintf(outfile, "#define NEED_BSD_STRINGS\n"); -#else - fprintf(outfile, "#undef NEED_BSD_STRINGS\n"); -#endif -#ifdef NEED_SYS_TYPES_H - fprintf(outfile, "#define NEED_SYS_TYPES_H\n"); -#else - fprintf(outfile, "#undef NEED_SYS_TYPES_H\n"); -#endif - fprintf(outfile, "#undef NEED_FAR_POINTERS\n"); -#ifdef NEED_SHORT_EXTERNAL_NAMES - fprintf(outfile, "#define NEED_SHORT_EXTERNAL_NAMES\n"); -#else - fprintf(outfile, "#undef NEED_SHORT_EXTERNAL_NAMES\n"); -#endif -#ifdef INCOMPLETE_TYPES_BROKEN - fprintf(outfile, "#define INCOMPLETE_TYPES_BROKEN\n"); -#else - fprintf(outfile, "#undef INCOMPLETE_TYPES_BROKEN\n"); -#endif - fprintf(outfile, "\n#ifdef JPEG_INTERNALS\n\n"); - if (is_shifting_signed(-0x7F7E80B1L)) - fprintf(outfile, "#undef RIGHT_SHIFT_IS_UNSIGNED\n"); - else - fprintf(outfile, "#define RIGHT_SHIFT_IS_UNSIGNED\n"); - fprintf(outfile, "\n#endif /* JPEG_INTERNALS */\n"); - fprintf(outfile, "\n#ifdef JPEG_CJPEG_DJPEG\n\n"); - fprintf(outfile, "#define BMP_SUPPORTED /* BMP image file format */\n"); - fprintf(outfile, "#define GIF_SUPPORTED /* GIF image file format */\n"); - fprintf(outfile, "#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */\n"); - fprintf(outfile, "#undef RLE_SUPPORTED /* Utah RLE image file format */\n"); - fprintf(outfile, "#define TARGA_SUPPORTED /* Targa image file format */\n\n"); - fprintf(outfile, "#undef TWO_FILE_COMMANDLINE /* You may need this on non-Unix systems */\n"); - fprintf(outfile, "#undef NEED_SIGNAL_CATCHER /* Define this if you use jmemname.c */\n"); - fprintf(outfile, "#undef DONT_USE_B_MODE\n"); - fprintf(outfile, "/* #define PROGRESS_REPORT */ /* optional */\n"); - fprintf(outfile, "\n#endif /* JPEG_CJPEG_DJPEG */\n"); - - /* Close the jconfig.h file */ - fclose(outfile); - - /* User report */ - printf("Configuration check for Independent JPEG Group's software done.\n"); - printf("\nI have written the jconfig.h file for you.\n\n"); -#ifdef HAVE_PROTOTYPES - printf("You should use makefile.ansi as the starting point for your Makefile.\n"); -#else - printf("You should use makefile.unix as the starting point for your Makefile.\n"); -#endif - -#ifdef NEED_SPECIAL_INCLUDE - printf("\nYou'll need to change jconfig.h to include the system include file\n"); - printf("that you found type size_t in, or add a direct definition of type\n"); - printf("size_t if that's what you used. Just add it to the end.\n"); -#endif - - return 0; -} diff --git a/src/3rdparty/libjpeg/coderules.txt b/src/3rdparty/libjpeg/coderules.txt deleted file mode 100644 index 357929fb44..0000000000 --- a/src/3rdparty/libjpeg/coderules.txt +++ /dev/null @@ -1,118 +0,0 @@ -IJG JPEG LIBRARY: CODING RULES - -Copyright (C) 1991-1996, Thomas G. Lane. -This file is part of the Independent JPEG Group's software. -For conditions of distribution and use, see the accompanying README file. - - -Since numerous people will be contributing code and bug fixes, it's important -to establish a common coding style. The goal of using similar coding styles -is much more important than the details of just what that style is. - -In general we follow the recommendations of "Recommended C Style and Coding -Standards" revision 6.1 (Cannon et al. as modified by Spencer, Keppel and -Brader). This document is available in the IJG FTP archive (see -jpeg/doc/cstyle.ms.tbl.Z, or cstyle.txt.Z for those without nroff/tbl). - -Block comments should be laid out thusly: - -/* - * Block comments in this style. - */ - -We indent statements in K&R style, e.g., - if (test) { - then-part; - } else { - else-part; - } -with two spaces per indentation level. (This indentation convention is -handled automatically by GNU Emacs and many other text editors.) - -Multi-word names should be written in lower case with underscores, e.g., -multi_word_name (not multiWordName). Preprocessor symbols and enum constants -are similar but upper case (MULTI_WORD_NAME). Names should be unique within -the first fifteen characters. (On some older systems, global names must be -unique within six characters. We accommodate this without cluttering the -source code by using macros to substitute shorter names.) - -We use function prototypes everywhere; we rely on automatic source code -transformation to feed prototype-less C compilers. Transformation is done -by the simple and portable tool 'ansi2knr.c' (courtesy of Ghostscript). -ansi2knr is not very bright, so it imposes a format requirement on function -declarations: the function name MUST BEGIN IN COLUMN 1. Thus all functions -should be written in the following style: - -LOCAL(int *) -function_name (int a, char *b) -{ - code... -} - -Note that each function definition must begin with GLOBAL(type), LOCAL(type), -or METHODDEF(type). These macros expand to "static type" or just "type" as -appropriate. They provide a readable indication of the routine's usage and -can readily be changed for special needs. (For instance, special linkage -keywords can be inserted for use in Windows DLLs.) - -ansi2knr does not transform method declarations (function pointers in -structs). We handle these with a macro JMETHOD, defined as - #ifdef HAVE_PROTOTYPES - #define JMETHOD(type,methodname,arglist) type (*methodname) arglist - #else - #define JMETHOD(type,methodname,arglist) type (*methodname) () - #endif -which is used like this: - struct function_pointers { - JMETHOD(void, init_entropy_encoder, (int somearg, jparms *jp)); - JMETHOD(void, term_entropy_encoder, (void)); - }; -Note the set of parentheses surrounding the parameter list. - -A similar solution is used for forward and external function declarations -(see the EXTERN and JPP macros). - -If the code is to work on non-ANSI compilers, we cannot rely on a prototype -declaration to coerce actual parameters into the right types. Therefore, use -explicit casts on actual parameters whenever the actual parameter type is not -identical to the formal parameter. Beware of implicit conversions to "int". - -It seems there are some non-ANSI compilers in which the sizeof() operator -is defined to return int, yet size_t is defined as long. Needless to say, -this is brain-damaged. Always use the SIZEOF() macro in place of sizeof(), -so that the result is guaranteed to be of type size_t. - - -The JPEG library is intended to be used within larger programs. Furthermore, -we want it to be reentrant so that it can be used by applications that process -multiple images concurrently. The following rules support these requirements: - -1. Avoid direct use of file I/O, "malloc", error report printouts, etc; -pass these through the common routines provided. - -2. Minimize global namespace pollution. Functions should be declared static -wherever possible. (Note that our method-based calling conventions help this -a lot: in many modules only the initialization function will ever need to be -called directly, so only that function need be externally visible.) All -global function names should begin with "jpeg_", and should have an -abbreviated name (unique in the first six characters) substituted by macro -when NEED_SHORT_EXTERNAL_NAMES is set. - -3. Don't use global variables; anything that must be used in another module -should be in the common data structures. - -4. Don't use static variables except for read-only constant tables. Variables -that should be private to a module can be placed into private structures (see -the system architecture document, structure.txt). - -5. Source file names should begin with "j" for files that are part of the -library proper; source files that are not part of the library, such as cjpeg.c -and djpeg.c, do not begin with "j". Keep source file names to eight -characters (plus ".c" or ".h", etc) to make life easy for MS-DOSers. Keep -compression and decompression code in separate source files --- some -applications may want only one half of the library. - -Note: these rules (particularly #4) are not followed religiously in the -modules that are used in cjpeg/djpeg but are not part of the JPEG library -proper. Those modules are not really intended to be used in other -applications. diff --git a/src/3rdparty/libjpeg/example.c b/src/3rdparty/libjpeg/example.c deleted file mode 100644 index 1d6f6cc30b..0000000000 --- a/src/3rdparty/libjpeg/example.c +++ /dev/null @@ -1,433 +0,0 @@ -/* - * example.c - * - * This file illustrates how to use the IJG code as a subroutine library - * to read or write JPEG image files. You should look at this code in - * conjunction with the documentation file libjpeg.txt. - * - * This code will not do anything useful as-is, but it may be helpful as a - * skeleton for constructing routines that call the JPEG library. - * - * We present these routines in the same coding style used in the JPEG code - * (ANSI function definitions, etc); but you are of course free to code your - * routines in a different style if you prefer. - */ - -#include <stdio.h> - -/* - * Include file for users of JPEG library. - * You will need to have included system headers that define at least - * the typedefs FILE and size_t before you can include jpeglib.h. - * (stdio.h is sufficient on ANSI-conforming systems.) - * You may also wish to include "jerror.h". - */ - -#include "jpeglib.h" - -/* - * <setjmp.h> is used for the optional error recovery mechanism shown in - * the second part of the example. - */ - -#include <setjmp.h> - - - -/******************** JPEG COMPRESSION SAMPLE INTERFACE *******************/ - -/* This half of the example shows how to feed data into the JPEG compressor. - * We present a minimal version that does not worry about refinements such - * as error recovery (the JPEG code will just exit() if it gets an error). - */ - - -/* - * IMAGE DATA FORMATS: - * - * The standard input image format is a rectangular array of pixels, with - * each pixel having the same number of "component" values (color channels). - * Each pixel row is an array of JSAMPLEs (which typically are unsigned chars). - * If you are working with color data, then the color values for each pixel - * must be adjacent in the row; for example, R,G,B,R,G,B,R,G,B,... for 24-bit - * RGB color. - * - * For this example, we'll assume that this data structure matches the way - * our application has stored the image in memory, so we can just pass a - * pointer to our image buffer. In particular, let's say that the image is - * RGB color and is described by: - */ - -extern JSAMPLE * image_buffer; /* Points to large array of R,G,B-order data */ -extern int image_height; /* Number of rows in image */ -extern int image_width; /* Number of columns in image */ - - -/* - * Sample routine for JPEG compression. We assume that the target file name - * and a compression quality factor are passed in. - */ - -GLOBAL(void) -write_JPEG_file (char * filename, int quality) -{ - /* This struct contains the JPEG compression parameters and pointers to - * working space (which is allocated as needed by the JPEG library). - * It is possible to have several such structures, representing multiple - * compression/decompression processes, in existence at once. We refer - * to any one struct (and its associated working data) as a "JPEG object". - */ - struct jpeg_compress_struct cinfo; - /* This struct represents a JPEG error handler. It is declared separately - * because applications often want to supply a specialized error handler - * (see the second half of this file for an example). But here we just - * take the easy way out and use the standard error handler, which will - * print a message on stderr and call exit() if compression fails. - * Note that this struct must live as long as the main JPEG parameter - * struct, to avoid dangling-pointer problems. - */ - struct jpeg_error_mgr jerr; - /* More stuff */ - FILE * outfile; /* target file */ - JSAMPROW row_pointer[1]; /* pointer to JSAMPLE row[s] */ - int row_stride; /* physical row width in image buffer */ - - /* Step 1: allocate and initialize JPEG compression object */ - - /* We have to set up the error handler first, in case the initialization - * step fails. (Unlikely, but it could happen if you are out of memory.) - * This routine fills in the contents of struct jerr, and returns jerr's - * address which we place into the link field in cinfo. - */ - cinfo.err = jpeg_std_error(&jerr); - /* Now we can initialize the JPEG compression object. */ - jpeg_create_compress(&cinfo); - - /* Step 2: specify data destination (eg, a file) */ - /* Note: steps 2 and 3 can be done in either order. */ - - /* Here we use the library-supplied code to send compressed data to a - * stdio stream. You can also write your own code to do something else. - * VERY IMPORTANT: use "b" option to fopen() if you are on a machine that - * requires it in order to write binary files. - */ - if ((outfile = fopen(filename, "wb")) == NULL) { - fprintf(stderr, "can't open %s\n", filename); - exit(1); - } - jpeg_stdio_dest(&cinfo, outfile); - - /* Step 3: set parameters for compression */ - - /* First we supply a description of the input image. - * Four fields of the cinfo struct must be filled in: - */ - cinfo.image_width = image_width; /* image width and height, in pixels */ - cinfo.image_height = image_height; - cinfo.input_components = 3; /* # of color components per pixel */ - cinfo.in_color_space = JCS_RGB; /* colorspace of input image */ - /* Now use the library's routine to set default compression parameters. - * (You must set at least cinfo.in_color_space before calling this, - * since the defaults depend on the source color space.) - */ - jpeg_set_defaults(&cinfo); - /* Now you can set any non-default parameters you wish to. - * Here we just illustrate the use of quality (quantization table) scaling: - */ - jpeg_set_quality(&cinfo, quality, TRUE /* limit to baseline-JPEG values */); - - /* Step 4: Start compressor */ - - /* TRUE ensures that we will write a complete interchange-JPEG file. - * Pass TRUE unless you are very sure of what you're doing. - */ - jpeg_start_compress(&cinfo, TRUE); - - /* Step 5: while (scan lines remain to be written) */ - /* jpeg_write_scanlines(...); */ - - /* Here we use the library's state variable cinfo.next_scanline as the - * loop counter, so that we don't have to keep track ourselves. - * To keep things simple, we pass one scanline per call; you can pass - * more if you wish, though. - */ - row_stride = image_width * 3; /* JSAMPLEs per row in image_buffer */ - - while (cinfo.next_scanline < cinfo.image_height) { - /* jpeg_write_scanlines expects an array of pointers to scanlines. - * Here the array is only one element long, but you could pass - * more than one scanline at a time if that's more convenient. - */ - row_pointer[0] = & image_buffer[cinfo.next_scanline * row_stride]; - (void) jpeg_write_scanlines(&cinfo, row_pointer, 1); - } - - /* Step 6: Finish compression */ - - jpeg_finish_compress(&cinfo); - /* After finish_compress, we can close the output file. */ - fclose(outfile); - - /* Step 7: release JPEG compression object */ - - /* This is an important step since it will release a good deal of memory. */ - jpeg_destroy_compress(&cinfo); - - /* And we're done! */ -} - - -/* - * SOME FINE POINTS: - * - * In the above loop, we ignored the return value of jpeg_write_scanlines, - * which is the number of scanlines actually written. We could get away - * with this because we were only relying on the value of cinfo.next_scanline, - * which will be incremented correctly. If you maintain additional loop - * variables then you should be careful to increment them properly. - * Actually, for output to a stdio stream you needn't worry, because - * then jpeg_write_scanlines will write all the lines passed (or else exit - * with a fatal error). Partial writes can only occur if you use a data - * destination module that can demand suspension of the compressor. - * (If you don't know what that's for, you don't need it.) - * - * If the compressor requires full-image buffers (for entropy-coding - * optimization or a multi-scan JPEG file), it will create temporary - * files for anything that doesn't fit within the maximum-memory setting. - * (Note that temp files are NOT needed if you use the default parameters.) - * On some systems you may need to set up a signal handler to ensure that - * temporary files are deleted if the program is interrupted. See libjpeg.txt. - * - * Scanlines MUST be supplied in top-to-bottom order if you want your JPEG - * files to be compatible with everyone else's. If you cannot readily read - * your data in that order, you'll need an intermediate array to hold the - * image. See rdtarga.c or rdbmp.c for examples of handling bottom-to-top - * source data using the JPEG code's internal virtual-array mechanisms. - */ - - - -/******************** JPEG DECOMPRESSION SAMPLE INTERFACE *******************/ - -/* This half of the example shows how to read data from the JPEG decompressor. - * It's a bit more refined than the above, in that we show: - * (a) how to modify the JPEG library's standard error-reporting behavior; - * (b) how to allocate workspace using the library's memory manager. - * - * Just to make this example a little different from the first one, we'll - * assume that we do not intend to put the whole image into an in-memory - * buffer, but to send it line-by-line someplace else. We need a one- - * scanline-high JSAMPLE array as a work buffer, and we will let the JPEG - * memory manager allocate it for us. This approach is actually quite useful - * because we don't need to remember to deallocate the buffer separately: it - * will go away automatically when the JPEG object is cleaned up. - */ - - -/* - * ERROR HANDLING: - * - * The JPEG library's standard error handler (jerror.c) is divided into - * several "methods" which you can override individually. This lets you - * adjust the behavior without duplicating a lot of code, which you might - * have to update with each future release. - * - * Our example here shows how to override the "error_exit" method so that - * control is returned to the library's caller when a fatal error occurs, - * rather than calling exit() as the standard error_exit method does. - * - * We use C's setjmp/longjmp facility to return control. This means that the - * routine which calls the JPEG library must first execute a setjmp() call to - * establish the return point. We want the replacement error_exit to do a - * longjmp(). But we need to make the setjmp buffer accessible to the - * error_exit routine. To do this, we make a private extension of the - * standard JPEG error handler object. (If we were using C++, we'd say we - * were making a subclass of the regular error handler.) - * - * Here's the extended error handler struct: - */ - -struct my_error_mgr { - struct jpeg_error_mgr pub; /* "public" fields */ - - jmp_buf setjmp_buffer; /* for return to caller */ -}; - -typedef struct my_error_mgr * my_error_ptr; - -/* - * Here's the routine that will replace the standard error_exit method: - */ - -METHODDEF(void) -my_error_exit (j_common_ptr cinfo) -{ - /* cinfo->err really points to a my_error_mgr struct, so coerce pointer */ - my_error_ptr myerr = (my_error_ptr) cinfo->err; - - /* Always display the message. */ - /* We could postpone this until after returning, if we chose. */ - (*cinfo->err->output_message) (cinfo); - - /* Return control to the setjmp point */ - longjmp(myerr->setjmp_buffer, 1); -} - - -/* - * Sample routine for JPEG decompression. We assume that the source file name - * is passed in. We want to return 1 on success, 0 on error. - */ - - -GLOBAL(int) -read_JPEG_file (char * filename) -{ - /* This struct contains the JPEG decompression parameters and pointers to - * working space (which is allocated as needed by the JPEG library). - */ - struct jpeg_decompress_struct cinfo; - /* We use our private extension JPEG error handler. - * Note that this struct must live as long as the main JPEG parameter - * struct, to avoid dangling-pointer problems. - */ - struct my_error_mgr jerr; - /* More stuff */ - FILE * infile; /* source file */ - JSAMPARRAY buffer; /* Output row buffer */ - int row_stride; /* physical row width in output buffer */ - - /* In this example we want to open the input file before doing anything else, - * so that the setjmp() error recovery below can assume the file is open. - * VERY IMPORTANT: use "b" option to fopen() if you are on a machine that - * requires it in order to read binary files. - */ - - if ((infile = fopen(filename, "rb")) == NULL) { - fprintf(stderr, "can't open %s\n", filename); - return 0; - } - - /* Step 1: allocate and initialize JPEG decompression object */ - - /* We set up the normal JPEG error routines, then override error_exit. */ - cinfo.err = jpeg_std_error(&jerr.pub); - jerr.pub.error_exit = my_error_exit; - /* Establish the setjmp return context for my_error_exit to use. */ - if (setjmp(jerr.setjmp_buffer)) { - /* If we get here, the JPEG code has signaled an error. - * We need to clean up the JPEG object, close the input file, and return. - */ - jpeg_destroy_decompress(&cinfo); - fclose(infile); - return 0; - } - /* Now we can initialize the JPEG decompression object. */ - jpeg_create_decompress(&cinfo); - - /* Step 2: specify data source (eg, a file) */ - - jpeg_stdio_src(&cinfo, infile); - - /* Step 3: read file parameters with jpeg_read_header() */ - - (void) jpeg_read_header(&cinfo, TRUE); - /* We can ignore the return value from jpeg_read_header since - * (a) suspension is not possible with the stdio data source, and - * (b) we passed TRUE to reject a tables-only JPEG file as an error. - * See libjpeg.txt for more info. - */ - - /* Step 4: set parameters for decompression */ - - /* In this example, we don't need to change any of the defaults set by - * jpeg_read_header(), so we do nothing here. - */ - - /* Step 5: Start decompressor */ - - (void) jpeg_start_decompress(&cinfo); - /* We can ignore the return value since suspension is not possible - * with the stdio data source. - */ - - /* We may need to do some setup of our own at this point before reading - * the data. After jpeg_start_decompress() we have the correct scaled - * output image dimensions available, as well as the output colormap - * if we asked for color quantization. - * In this example, we need to make an output work buffer of the right size. - */ - /* JSAMPLEs per row in output buffer */ - row_stride = cinfo.output_width * cinfo.output_components; - /* Make a one-row-high sample array that will go away when done with image */ - buffer = (*cinfo.mem->alloc_sarray) - ((j_common_ptr) &cinfo, JPOOL_IMAGE, row_stride, 1); - - /* Step 6: while (scan lines remain to be read) */ - /* jpeg_read_scanlines(...); */ - - /* Here we use the library's state variable cinfo.output_scanline as the - * loop counter, so that we don't have to keep track ourselves. - */ - while (cinfo.output_scanline < cinfo.output_height) { - /* jpeg_read_scanlines expects an array of pointers to scanlines. - * Here the array is only one element long, but you could ask for - * more than one scanline at a time if that's more convenient. - */ - (void) jpeg_read_scanlines(&cinfo, buffer, 1); - /* Assume put_scanline_someplace wants a pointer and sample count. */ - put_scanline_someplace(buffer[0], row_stride); - } - - /* Step 7: Finish decompression */ - - (void) jpeg_finish_decompress(&cinfo); - /* We can ignore the return value since suspension is not possible - * with the stdio data source. - */ - - /* Step 8: Release JPEG decompression object */ - - /* This is an important step since it will release a good deal of memory. */ - jpeg_destroy_decompress(&cinfo); - - /* After finish_decompress, we can close the input file. - * Here we postpone it until after no more JPEG errors are possible, - * so as to simplify the setjmp error logic above. (Actually, I don't - * think that jpeg_destroy can do an error exit, but why assume anything...) - */ - fclose(infile); - - /* At this point you may want to check to see whether any corrupt-data - * warnings occurred (test whether jerr.pub.num_warnings is nonzero). - */ - - /* And we're done! */ - return 1; -} - - -/* - * SOME FINE POINTS: - * - * In the above code, we ignored the return value of jpeg_read_scanlines, - * which is the number of scanlines actually read. We could get away with - * this because we asked for only one line at a time and we weren't using - * a suspending data source. See libjpeg.txt for more info. - * - * We cheated a bit by calling alloc_sarray() after jpeg_start_decompress(); - * we should have done it beforehand to ensure that the space would be - * counted against the JPEG max_memory setting. In some systems the above - * code would risk an out-of-memory error. However, in general we don't - * know the output image dimensions before jpeg_start_decompress(), unless we - * call jpeg_calc_output_dimensions(). See libjpeg.txt for more about this. - * - * Scanlines are returned in the same order as they appear in the JPEG file, - * which is standardly top-to-bottom. If you must emit data bottom-to-top, - * you can use one of the virtual arrays provided by the JPEG memory manager - * to invert the data. See wrbmp.c for an example. - * - * As with compression, some operating modes may require temporary files. - * On some systems you may need to set up a signal handler to ensure that - * temporary files are deleted if the program is interrupted. See libjpeg.txt. - */ diff --git a/src/3rdparty/libjpeg/filelist.txt b/src/3rdparty/libjpeg/filelist.txt deleted file mode 100644 index 7e053869a6..0000000000 --- a/src/3rdparty/libjpeg/filelist.txt +++ /dev/null @@ -1,215 +0,0 @@ -IJG JPEG LIBRARY: FILE LIST - -Copyright (C) 1994-2009, Thomas G. Lane, Guido Vollbeding. -This file is part of the Independent JPEG Group's software. -For conditions of distribution and use, see the accompanying README file. - - -Here is a road map to the files in the IJG JPEG distribution. The -distribution includes the JPEG library proper, plus two application -programs ("cjpeg" and "djpeg") which use the library to convert JPEG -files to and from some other popular image formats. A third application -"jpegtran" uses the library to do lossless conversion between different -variants of JPEG. There are also two stand-alone applications, -"rdjpgcom" and "wrjpgcom". - - -THE JPEG LIBRARY -================ - -Include files: - -jpeglib.h JPEG library's exported data and function declarations. -jconfig.h Configuration declarations. Note: this file is not present - in the distribution; it is generated during installation. -jmorecfg.h Additional configuration declarations; need not be changed - for a standard installation. -jerror.h Declares JPEG library's error and trace message codes. -jinclude.h Central include file used by all IJG .c files to reference - system include files. -jpegint.h JPEG library's internal data structures. -jdct.h Private declarations for forward & reverse DCT subsystems. -jmemsys.h Private declarations for memory management subsystem. -jversion.h Version information. - -Applications using the library should include jpeglib.h (which in turn -includes jconfig.h and jmorecfg.h). Optionally, jerror.h may be included -if the application needs to reference individual JPEG error codes. The -other include files are intended for internal use and would not normally -be included by an application program. (cjpeg/djpeg/etc do use jinclude.h, -since its function is to improve portability of the whole IJG distribution. -Most other applications will directly include the system include files they -want, and hence won't need jinclude.h.) - - -C source code files: - -These files contain most of the functions intended to be called directly by -an application program: - -jcapimin.c Application program interface: core routines for compression. -jcapistd.c Application program interface: standard compression. -jdapimin.c Application program interface: core routines for decompression. -jdapistd.c Application program interface: standard decompression. -jcomapi.c Application program interface routines common to compression - and decompression. -jcparam.c Compression parameter setting helper routines. -jctrans.c API and library routines for transcoding compression. -jdtrans.c API and library routines for transcoding decompression. - -Compression side of the library: - -jcinit.c Initialization: determines which other modules to use. -jcmaster.c Master control: setup and inter-pass sequencing logic. -jcmainct.c Main buffer controller (preprocessor => JPEG compressor). -jcprepct.c Preprocessor buffer controller. -jccoefct.c Buffer controller for DCT coefficient buffer. -jccolor.c Color space conversion. -jcsample.c Downsampling. -jcdctmgr.c DCT manager (DCT implementation selection & control). -jfdctint.c Forward DCT using slow-but-accurate integer method. -jfdctfst.c Forward DCT using faster, less accurate integer method. -jfdctflt.c Forward DCT using floating-point arithmetic. -jchuff.c Huffman entropy coding. -jcarith.c Arithmetic entropy coding. -jcmarker.c JPEG marker writing. -jdatadst.c Data destination managers for memory and stdio output. - -Decompression side of the library: - -jdmaster.c Master control: determines which other modules to use. -jdinput.c Input controller: controls input processing modules. -jdmainct.c Main buffer controller (JPEG decompressor => postprocessor). -jdcoefct.c Buffer controller for DCT coefficient buffer. -jdpostct.c Postprocessor buffer controller. -jdmarker.c JPEG marker reading. -jdhuff.c Huffman entropy decoding. -jdarith.c Arithmetic entropy decoding. -jddctmgr.c IDCT manager (IDCT implementation selection & control). -jidctint.c Inverse DCT using slow-but-accurate integer method. -jidctfst.c Inverse DCT using faster, less accurate integer method. -jidctflt.c Inverse DCT using floating-point arithmetic. -jdsample.c Upsampling. -jdcolor.c Color space conversion. -jdmerge.c Merged upsampling/color conversion (faster, lower quality). -jquant1.c One-pass color quantization using a fixed-spacing colormap. -jquant2.c Two-pass color quantization using a custom-generated colormap. - Also handles one-pass quantization to an externally given map. -jdatasrc.c Data source managers for memory and stdio input. - -Support files for both compression and decompression: - -jaricom.c Tables for common use in arithmetic entropy encoding and - decoding routines. -jerror.c Standard error handling routines (application replaceable). -jmemmgr.c System-independent (more or less) memory management code. -jutils.c Miscellaneous utility routines. - -jmemmgr.c relies on a system-dependent memory management module. The IJG -distribution includes the following implementations of the system-dependent -module: - -jmemnobs.c "No backing store": assumes adequate virtual memory exists. -jmemansi.c Makes temporary files with ANSI-standard routine tmpfile(). -jmemname.c Makes temporary files with program-generated file names. -jmemdos.c Custom implementation for MS-DOS (16-bit environment only): - can use extended and expanded memory as well as temp files. -jmemmac.c Custom implementation for Apple Macintosh. - -Exactly one of the system-dependent modules should be configured into an -installed JPEG library (see install.txt for hints about which one to use). -On unusual systems you may find it worthwhile to make a special -system-dependent memory manager. - - -Non-C source code files: - -jmemdosa.asm 80x86 assembly code support for jmemdos.c; used only in - MS-DOS-specific configurations of the JPEG library. - - -CJPEG/DJPEG/JPEGTRAN -==================== - -Include files: - -cdjpeg.h Declarations shared by cjpeg/djpeg/jpegtran modules. -cderror.h Additional error and trace message codes for cjpeg et al. -transupp.h Declarations for jpegtran support routines in transupp.c. - -C source code files: - -cjpeg.c Main program for cjpeg. -djpeg.c Main program for djpeg. -jpegtran.c Main program for jpegtran. -cdjpeg.c Utility routines used by all three programs. -rdcolmap.c Code to read a colormap file for djpeg's "-map" switch. -rdswitch.c Code to process some of cjpeg's more complex switches. - Also used by jpegtran. -transupp.c Support code for jpegtran: lossless image manipulations. - -Image file reader modules for cjpeg: - -rdbmp.c BMP file input. -rdgif.c GIF file input (now just a stub). -rdppm.c PPM/PGM file input. -rdrle.c Utah RLE file input. -rdtarga.c Targa file input. - -Image file writer modules for djpeg: - -wrbmp.c BMP file output. -wrgif.c GIF file output (a mere shadow of its former self). -wrppm.c PPM/PGM file output. -wrrle.c Utah RLE file output. -wrtarga.c Targa file output. - - -RDJPGCOM/WRJPGCOM -================= - -C source code files: - -rdjpgcom.c Stand-alone rdjpgcom application. -wrjpgcom.c Stand-alone wrjpgcom application. - -These programs do not depend on the IJG library. They do use -jconfig.h and jinclude.h, only to improve portability. - - -ADDITIONAL FILES -================ - -Documentation (see README for a guide to the documentation files): - -README Master documentation file. -*.txt Other documentation files. -*.1 Documentation in Unix man page format. -change.log Version-to-version change highlights. -example.c Sample code for calling JPEG library. - -Configuration/installation files and programs (see install.txt for more info): - -configure Unix shell script to perform automatic configuration. -configure.ac Source file for use with Autoconf to generate configure. -ltmain.sh Support scripts for configure (from GNU libtool). -config.guess -config.sub -depcomp -missing -install-sh Install shell script for those Unix systems lacking one. -Makefile.in Makefile input for configure. -Makefile.am Source file for use with Automake to generate Makefile.in. -ckconfig.c Program to generate jconfig.h on non-Unix systems. -jconfig.txt Template for making jconfig.h by hand. -mak*.* Sample makefiles for particular systems. -jconfig.* Sample jconfig.h for particular systems. -libjpeg.map Script to generate shared library with versioned symbols. -aclocal.m4 M4 macro definitions for use with Autoconf. -ansi2knr.c De-ANSIfier for pre-ANSI C compilers (courtesy of - L. Peter Deutsch and Aladdin Enterprises). - -Test files (see install.txt for test procedure): - -test*.* Source and comparison files for confidence test. - These are binary image files, NOT text files. diff --git a/src/3rdparty/libjpeg/import_from_libjpeg_tarball.sh b/src/3rdparty/libjpeg/import_from_libjpeg_tarball.sh new file mode 100755 index 0000000000..b22f6b0d47 --- /dev/null +++ b/src/3rdparty/libjpeg/import_from_libjpeg_tarball.sh @@ -0,0 +1,165 @@ +#! /bin/sh +############################################################################# +## +## Copyright (C) 2017 André Klitzing +## Contact: https://www.qt.io/licensing/ +## +## This file is the build configuration utility of the Qt Toolkit. +## +## $QT_BEGIN_LICENSE:LGPL$ +## Commercial License Usage +## Licensees holding valid commercial Qt licenses may use this file in +## accordance with the commercial license agreement provided with the +## Software or, alternatively, in accordance with the terms contained in +## a written agreement between you and The Qt Company. For licensing terms +## and conditions see https://www.qt.io/terms-conditions. For further +## information use the contact form at https://www.qt.io/contact-us. +## +## GNU Lesser General Public License Usage +## Alternatively, this file may be used under the terms of the GNU Lesser +## General Public License version 3 as published by the Free Software +## Foundation and appearing in the file LICENSE.LGPL3 included in the +## packaging of this file. Please review the following information to +## ensure the GNU Lesser General Public License version 3 requirements +## will be met: https://www.gnu.org/licenses/lgpl-3.0.html. +## +## GNU General Public License Usage +## Alternatively, this file may be used under the terms of the GNU +## General Public License version 2.0 or (at your option) the GNU General +## Public license version 3 or any later version approved by the KDE Free +## Qt Foundation. The licenses are as published by the Free Software +## Foundation and appearing in the file LICENSE.GPL2 and LICENSE.GPL3 +## included in the packaging of this file. Please review the following +## information to ensure the GNU General Public License requirements will +## be met: https://www.gnu.org/licenses/gpl-2.0.html and +## https://www.gnu.org/licenses/gpl-3.0.html. +## +## $QT_END_LICENSE$ +## +############################################################################# + +# This is a small script to copy the required files from a LIBJPEG tarball +# into 3rdparty/libjpeg/. + +if [ $# -ne 2 ]; then + echo "Usage: $0 LIBJPEG_tarball_dir/ \$QTDIR/src/3rdparty/LIBJPEG/" + exit 1 +fi + +LIBJPEG_DIR=$1 +TARGET_DIR=$2 + +if [ ! -d "$LIBJPEG_DIR" -o ! -r "$LIBJPEG_DIR" -o ! -d "$TARGET_DIR" -o ! -w "$TARGET_DIR" ]; then + echo "Either the LIBJPEG source dir or the target dir do not exist," + echo "are not directories or have the wrong permissions." + exit 2 +fi + +# with 1 argument, copies LIBJPEG_DIR/$1 to TARGET_DIR/$1 +# with 2 arguments, copies LIBJPEG_DIR/$1 to TARGET_DIR/$2 +copy_file() { + if [ $# -lt 1 -o $# -gt 2 ]; then + echo "Wrong number of arguments to copy_file" + exit 3 + fi + + SOURCE_FILE=$1 + if [ -n "$2" ]; then + DEST_FILE=$2 + else + DEST_FILE=$1 + fi + + mkdir -p "$TARGET_DIR/$(dirname "$SOURCE_FILE")" + cp "$LIBJPEG_DIR/$SOURCE_FILE" "$TARGET_DIR/$DEST_FILE" +} + +copy_file "LICENSE.md" "LICENSE" +copy_file "jconfig.txt" "src/jconfig.h" +copy_file "win/jconfigint.h.in" "src/jconfigint.h" + +FILES=" + change.log + ChangeLog.md + README.md + README.ijg + + jaricom.c + jcapimin.c + jcapistd.c + jcarith.c + jccoefct.c + jccolext.c + jccolor.c + jcdctmgr.c + jchuff.c + jchuff.h + jcinit.c + jcmainct.c + jcmarker.c + jcmaster.c + jcomapi.c + jcparam.c + jcphuff.c + jcprepct.c + jcsample.c + jctrans.c + jdapimin.c + jdapistd.c + jdarith.c + jdatadst.c + jdatasrc.c + jdcoefct.c + jdcoefct.h + jdcolext.c + jdcol565.c + jdcolor.c + jdct.h + jddctmgr.c + jdhuff.c + jdhuff.h + jdphuff.c + jdinput.c + jdmainct.c + jdmainct.h + jdmarker.c + jdmaster.c + jdmaster.h + jdmerge.c + jdmrgext.c + jdmrg565.c + jdpostct.c + jdsample.c + jdsample.h + jdtrans.c + jerror.c + jerror.h + jfdctflt.c + jfdctfst.c + jfdctint.c + jidctred.c + jidctflt.c + jidctfst.c + jidctint.c + jinclude.h + jpegcomp.h + jpegint.h + jpeglib.h + jmemmgr.c + jmemnobs.c + jmemsys.h + jmorecfg.h + jpeg_nbits_table.h + jquant1.c + jquant2.c + jsimd.h + jsimd_none.c + jsimddct.h + jstdhuff.c + jutils.c + jversion.h +" + +for i in $FILES; do + copy_file "$i" "src/$i" +done diff --git a/src/3rdparty/libjpeg/jccolor.c b/src/3rdparty/libjpeg/jccolor.c deleted file mode 100644 index 0a8a4b5d13..0000000000 --- a/src/3rdparty/libjpeg/jccolor.c +++ /dev/null @@ -1,459 +0,0 @@ -/* - * jccolor.c - * - * Copyright (C) 1991-1996, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains input colorspace conversion routines. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Private subobject */ - -typedef struct { - struct jpeg_color_converter pub; /* public fields */ - - /* Private state for RGB->YCC conversion */ - INT32 * rgb_ycc_tab; /* => table for RGB to YCbCr conversion */ -} my_color_converter; - -typedef my_color_converter * my_cconvert_ptr; - - -/**************** RGB -> YCbCr conversion: most common case **************/ - -/* - * YCbCr is defined per CCIR 601-1, except that Cb and Cr are - * normalized to the range 0..MAXJSAMPLE rather than -0.5 .. 0.5. - * The conversion equations to be implemented are therefore - * Y = 0.29900 * R + 0.58700 * G + 0.11400 * B - * Cb = -0.16874 * R - 0.33126 * G + 0.50000 * B + CENTERJSAMPLE - * Cr = 0.50000 * R - 0.41869 * G - 0.08131 * B + CENTERJSAMPLE - * (These numbers are derived from TIFF 6.0 section 21, dated 3-June-92.) - * Note: older versions of the IJG code used a zero offset of MAXJSAMPLE/2, - * rather than CENTERJSAMPLE, for Cb and Cr. This gave equal positive and - * negative swings for Cb/Cr, but meant that grayscale values (Cb=Cr=0) - * were not represented exactly. Now we sacrifice exact representation of - * maximum red and maximum blue in order to get exact grayscales. - * - * To avoid floating-point arithmetic, we represent the fractional constants - * as integers scaled up by 2^16 (about 4 digits precision); we have to divide - * the products by 2^16, with appropriate rounding, to get the correct answer. - * - * For even more speed, we avoid doing any multiplications in the inner loop - * by precalculating the constants times R,G,B for all possible values. - * For 8-bit JSAMPLEs this is very reasonable (only 256 entries per table); - * for 12-bit samples it is still acceptable. It's not very reasonable for - * 16-bit samples, but if you want lossless storage you shouldn't be changing - * colorspace anyway. - * The CENTERJSAMPLE offsets and the rounding fudge-factor of 0.5 are included - * in the tables to save adding them separately in the inner loop. - */ - -#define SCALEBITS 16 /* speediest right-shift on some machines */ -#define CBCR_OFFSET ((INT32) CENTERJSAMPLE << SCALEBITS) -#define ONE_HALF ((INT32) 1 << (SCALEBITS-1)) -#define FIX(x) ((INT32) ((x) * (1L<<SCALEBITS) + 0.5)) - -/* We allocate one big table and divide it up into eight parts, instead of - * doing eight alloc_small requests. This lets us use a single table base - * address, which can be held in a register in the inner loops on many - * machines (more than can hold all eight addresses, anyway). - */ - -#define R_Y_OFF 0 /* offset to R => Y section */ -#define G_Y_OFF (1*(MAXJSAMPLE+1)) /* offset to G => Y section */ -#define B_Y_OFF (2*(MAXJSAMPLE+1)) /* etc. */ -#define R_CB_OFF (3*(MAXJSAMPLE+1)) -#define G_CB_OFF (4*(MAXJSAMPLE+1)) -#define B_CB_OFF (5*(MAXJSAMPLE+1)) -#define R_CR_OFF B_CB_OFF /* B=>Cb, R=>Cr are the same */ -#define G_CR_OFF (6*(MAXJSAMPLE+1)) -#define B_CR_OFF (7*(MAXJSAMPLE+1)) -#define TABLE_SIZE (8*(MAXJSAMPLE+1)) - - -/* - * Initialize for RGB->YCC colorspace conversion. - */ - -METHODDEF(void) -rgb_ycc_start (j_compress_ptr cinfo) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - INT32 * rgb_ycc_tab; - INT32 i; - - /* Allocate and fill in the conversion tables. */ - cconvert->rgb_ycc_tab = rgb_ycc_tab = (INT32 *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (TABLE_SIZE * SIZEOF(INT32))); - - for (i = 0; i <= MAXJSAMPLE; i++) { - rgb_ycc_tab[i+R_Y_OFF] = FIX(0.29900) * i; - rgb_ycc_tab[i+G_Y_OFF] = FIX(0.58700) * i; - rgb_ycc_tab[i+B_Y_OFF] = FIX(0.11400) * i + ONE_HALF; - rgb_ycc_tab[i+R_CB_OFF] = (-FIX(0.16874)) * i; - rgb_ycc_tab[i+G_CB_OFF] = (-FIX(0.33126)) * i; - /* We use a rounding fudge-factor of 0.5-epsilon for Cb and Cr. - * This ensures that the maximum output will round to MAXJSAMPLE - * not MAXJSAMPLE+1, and thus that we don't have to range-limit. - */ - rgb_ycc_tab[i+B_CB_OFF] = FIX(0.50000) * i + CBCR_OFFSET + ONE_HALF-1; -/* B=>Cb and R=>Cr tables are the same - rgb_ycc_tab[i+R_CR_OFF] = FIX(0.50000) * i + CBCR_OFFSET + ONE_HALF-1; -*/ - rgb_ycc_tab[i+G_CR_OFF] = (-FIX(0.41869)) * i; - rgb_ycc_tab[i+B_CR_OFF] = (-FIX(0.08131)) * i; - } -} - - -/* - * Convert some rows of samples to the JPEG colorspace. - * - * Note that we change from the application's interleaved-pixel format - * to our internal noninterleaved, one-plane-per-component format. - * The input buffer is therefore three times as wide as the output buffer. - * - * A starting row offset is provided only for the output buffer. The caller - * can easily adjust the passed input_buf value to accommodate any row - * offset required on that side. - */ - -METHODDEF(void) -rgb_ycc_convert (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JSAMPIMAGE output_buf, - JDIMENSION output_row, int num_rows) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - register int r, g, b; - register INT32 * ctab = cconvert->rgb_ycc_tab; - register JSAMPROW inptr; - register JSAMPROW outptr0, outptr1, outptr2; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->image_width; - - while (--num_rows >= 0) { - inptr = *input_buf++; - outptr0 = output_buf[0][output_row]; - outptr1 = output_buf[1][output_row]; - outptr2 = output_buf[2][output_row]; - output_row++; - for (col = 0; col < num_cols; col++) { - r = GETJSAMPLE(inptr[RGB_RED]); - g = GETJSAMPLE(inptr[RGB_GREEN]); - b = GETJSAMPLE(inptr[RGB_BLUE]); - inptr += RGB_PIXELSIZE; - /* If the inputs are 0..MAXJSAMPLE, the outputs of these equations - * must be too; we do not need an explicit range-limiting operation. - * Hence the value being shifted is never negative, and we don't - * need the general RIGHT_SHIFT macro. - */ - /* Y */ - outptr0[col] = (JSAMPLE) - ((ctab[r+R_Y_OFF] + ctab[g+G_Y_OFF] + ctab[b+B_Y_OFF]) - >> SCALEBITS); - /* Cb */ - outptr1[col] = (JSAMPLE) - ((ctab[r+R_CB_OFF] + ctab[g+G_CB_OFF] + ctab[b+B_CB_OFF]) - >> SCALEBITS); - /* Cr */ - outptr2[col] = (JSAMPLE) - ((ctab[r+R_CR_OFF] + ctab[g+G_CR_OFF] + ctab[b+B_CR_OFF]) - >> SCALEBITS); - } - } -} - - -/**************** Cases other than RGB -> YCbCr **************/ - - -/* - * Convert some rows of samples to the JPEG colorspace. - * This version handles RGB->grayscale conversion, which is the same - * as the RGB->Y portion of RGB->YCbCr. - * We assume rgb_ycc_start has been called (we only use the Y tables). - */ - -METHODDEF(void) -rgb_gray_convert (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JSAMPIMAGE output_buf, - JDIMENSION output_row, int num_rows) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - register int r, g, b; - register INT32 * ctab = cconvert->rgb_ycc_tab; - register JSAMPROW inptr; - register JSAMPROW outptr; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->image_width; - - while (--num_rows >= 0) { - inptr = *input_buf++; - outptr = output_buf[0][output_row]; - output_row++; - for (col = 0; col < num_cols; col++) { - r = GETJSAMPLE(inptr[RGB_RED]); - g = GETJSAMPLE(inptr[RGB_GREEN]); - b = GETJSAMPLE(inptr[RGB_BLUE]); - inptr += RGB_PIXELSIZE; - /* Y */ - outptr[col] = (JSAMPLE) - ((ctab[r+R_Y_OFF] + ctab[g+G_Y_OFF] + ctab[b+B_Y_OFF]) - >> SCALEBITS); - } - } -} - - -/* - * Convert some rows of samples to the JPEG colorspace. - * This version handles Adobe-style CMYK->YCCK conversion, - * where we convert R=1-C, G=1-M, and B=1-Y to YCbCr using the same - * conversion as above, while passing K (black) unchanged. - * We assume rgb_ycc_start has been called. - */ - -METHODDEF(void) -cmyk_ycck_convert (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JSAMPIMAGE output_buf, - JDIMENSION output_row, int num_rows) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - register int r, g, b; - register INT32 * ctab = cconvert->rgb_ycc_tab; - register JSAMPROW inptr; - register JSAMPROW outptr0, outptr1, outptr2, outptr3; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->image_width; - - while (--num_rows >= 0) { - inptr = *input_buf++; - outptr0 = output_buf[0][output_row]; - outptr1 = output_buf[1][output_row]; - outptr2 = output_buf[2][output_row]; - outptr3 = output_buf[3][output_row]; - output_row++; - for (col = 0; col < num_cols; col++) { - r = MAXJSAMPLE - GETJSAMPLE(inptr[0]); - g = MAXJSAMPLE - GETJSAMPLE(inptr[1]); - b = MAXJSAMPLE - GETJSAMPLE(inptr[2]); - /* K passes through as-is */ - outptr3[col] = inptr[3]; /* don't need GETJSAMPLE here */ - inptr += 4; - /* If the inputs are 0..MAXJSAMPLE, the outputs of these equations - * must be too; we do not need an explicit range-limiting operation. - * Hence the value being shifted is never negative, and we don't - * need the general RIGHT_SHIFT macro. - */ - /* Y */ - outptr0[col] = (JSAMPLE) - ((ctab[r+R_Y_OFF] + ctab[g+G_Y_OFF] + ctab[b+B_Y_OFF]) - >> SCALEBITS); - /* Cb */ - outptr1[col] = (JSAMPLE) - ((ctab[r+R_CB_OFF] + ctab[g+G_CB_OFF] + ctab[b+B_CB_OFF]) - >> SCALEBITS); - /* Cr */ - outptr2[col] = (JSAMPLE) - ((ctab[r+R_CR_OFF] + ctab[g+G_CR_OFF] + ctab[b+B_CR_OFF]) - >> SCALEBITS); - } - } -} - - -/* - * Convert some rows of samples to the JPEG colorspace. - * This version handles grayscale output with no conversion. - * The source can be either plain grayscale or YCbCr (since Y == gray). - */ - -METHODDEF(void) -grayscale_convert (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JSAMPIMAGE output_buf, - JDIMENSION output_row, int num_rows) -{ - register JSAMPROW inptr; - register JSAMPROW outptr; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->image_width; - int instride = cinfo->input_components; - - while (--num_rows >= 0) { - inptr = *input_buf++; - outptr = output_buf[0][output_row]; - output_row++; - for (col = 0; col < num_cols; col++) { - outptr[col] = inptr[0]; /* don't need GETJSAMPLE() here */ - inptr += instride; - } - } -} - - -/* - * Convert some rows of samples to the JPEG colorspace. - * This version handles multi-component colorspaces without conversion. - * We assume input_components == num_components. - */ - -METHODDEF(void) -null_convert (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JSAMPIMAGE output_buf, - JDIMENSION output_row, int num_rows) -{ - register JSAMPROW inptr; - register JSAMPROW outptr; - register JDIMENSION col; - register int ci; - int nc = cinfo->num_components; - JDIMENSION num_cols = cinfo->image_width; - - while (--num_rows >= 0) { - /* It seems fastest to make a separate pass for each component. */ - for (ci = 0; ci < nc; ci++) { - inptr = *input_buf; - outptr = output_buf[ci][output_row]; - for (col = 0; col < num_cols; col++) { - outptr[col] = inptr[ci]; /* don't need GETJSAMPLE() here */ - inptr += nc; - } - } - input_buf++; - output_row++; - } -} - - -/* - * Empty method for start_pass. - */ - -METHODDEF(void) -null_method (j_compress_ptr cinfo) -{ - /* no work needed */ -} - - -/* - * Module initialization routine for input colorspace conversion. - */ - -GLOBAL(void) -jinit_color_converter (j_compress_ptr cinfo) -{ - my_cconvert_ptr cconvert; - - cconvert = (my_cconvert_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_color_converter)); - cinfo->cconvert = (struct jpeg_color_converter *) cconvert; - /* set start_pass to null method until we find out differently */ - cconvert->pub.start_pass = null_method; - - /* Make sure input_components agrees with in_color_space */ - switch (cinfo->in_color_space) { - case JCS_GRAYSCALE: - if (cinfo->input_components != 1) - ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE); - break; - - case JCS_RGB: -#if RGB_PIXELSIZE != 3 - if (cinfo->input_components != RGB_PIXELSIZE) - ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE); - break; -#endif /* else share code with YCbCr */ - - case JCS_YCbCr: - if (cinfo->input_components != 3) - ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE); - break; - - case JCS_CMYK: - case JCS_YCCK: - if (cinfo->input_components != 4) - ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE); - break; - - default: /* JCS_UNKNOWN can be anything */ - if (cinfo->input_components < 1) - ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE); - break; - } - - /* Check num_components, set conversion method based on requested space */ - switch (cinfo->jpeg_color_space) { - case JCS_GRAYSCALE: - if (cinfo->num_components != 1) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - if (cinfo->in_color_space == JCS_GRAYSCALE) - cconvert->pub.color_convert = grayscale_convert; - else if (cinfo->in_color_space == JCS_RGB) { - cconvert->pub.start_pass = rgb_ycc_start; - cconvert->pub.color_convert = rgb_gray_convert; - } else if (cinfo->in_color_space == JCS_YCbCr) - cconvert->pub.color_convert = grayscale_convert; - else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - case JCS_RGB: - if (cinfo->num_components != 3) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - if (cinfo->in_color_space == JCS_RGB && RGB_PIXELSIZE == 3) - cconvert->pub.color_convert = null_convert; - else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - case JCS_YCbCr: - if (cinfo->num_components != 3) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - if (cinfo->in_color_space == JCS_RGB) { - cconvert->pub.start_pass = rgb_ycc_start; - cconvert->pub.color_convert = rgb_ycc_convert; - } else if (cinfo->in_color_space == JCS_YCbCr) - cconvert->pub.color_convert = null_convert; - else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - case JCS_CMYK: - if (cinfo->num_components != 4) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - if (cinfo->in_color_space == JCS_CMYK) - cconvert->pub.color_convert = null_convert; - else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - case JCS_YCCK: - if (cinfo->num_components != 4) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - if (cinfo->in_color_space == JCS_CMYK) { - cconvert->pub.start_pass = rgb_ycc_start; - cconvert->pub.color_convert = cmyk_ycck_convert; - } else if (cinfo->in_color_space == JCS_YCCK) - cconvert->pub.color_convert = null_convert; - else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - default: /* allow null conversion of JCS_UNKNOWN */ - if (cinfo->jpeg_color_space != cinfo->in_color_space || - cinfo->num_components != cinfo->input_components) - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - cconvert->pub.color_convert = null_convert; - break; - } -} diff --git a/src/3rdparty/libjpeg/jcdctmgr.c b/src/3rdparty/libjpeg/jcdctmgr.c deleted file mode 100644 index 0bbdbb685d..0000000000 --- a/src/3rdparty/libjpeg/jcdctmgr.c +++ /dev/null @@ -1,482 +0,0 @@ -/* - * jcdctmgr.c - * - * Copyright (C) 1994-1996, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains the forward-DCT management logic. - * This code selects a particular DCT implementation to be used, - * and it performs related housekeeping chores including coefficient - * quantization. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" -#include "jdct.h" /* Private declarations for DCT subsystem */ - - -/* Private subobject for this module */ - -typedef struct { - struct jpeg_forward_dct pub; /* public fields */ - - /* Pointer to the DCT routine actually in use */ - forward_DCT_method_ptr do_dct[MAX_COMPONENTS]; - - /* The actual post-DCT divisors --- not identical to the quant table - * entries, because of scaling (especially for an unnormalized DCT). - * Each table is given in normal array order. - */ - DCTELEM * divisors[NUM_QUANT_TBLS]; - -#ifdef DCT_FLOAT_SUPPORTED - /* Same as above for the floating-point case. */ - float_DCT_method_ptr do_float_dct[MAX_COMPONENTS]; - FAST_FLOAT * float_divisors[NUM_QUANT_TBLS]; -#endif -} my_fdct_controller; - -typedef my_fdct_controller * my_fdct_ptr; - - -/* The current scaled-DCT routines require ISLOW-style divisor tables, - * so be sure to compile that code if either ISLOW or SCALING is requested. - */ -#ifdef DCT_ISLOW_SUPPORTED -#define PROVIDE_ISLOW_TABLES -#else -#ifdef DCT_SCALING_SUPPORTED -#define PROVIDE_ISLOW_TABLES -#endif -#endif - - -/* - * Perform forward DCT on one or more blocks of a component. - * - * The input samples are taken from the sample_data[] array starting at - * position start_row/start_col, and moving to the right for any additional - * blocks. The quantized coefficients are returned in coef_blocks[]. - */ - -METHODDEF(void) -forward_DCT (j_compress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY sample_data, JBLOCKROW coef_blocks, - JDIMENSION start_row, JDIMENSION start_col, - JDIMENSION num_blocks) -/* This version is used for integer DCT implementations. */ -{ - /* This routine is heavily used, so it's worth coding it tightly. */ - my_fdct_ptr fdct = (my_fdct_ptr) cinfo->fdct; - forward_DCT_method_ptr do_dct = fdct->do_dct[compptr->component_index]; - DCTELEM * divisors = fdct->divisors[compptr->quant_tbl_no]; - DCTELEM workspace[DCTSIZE2]; /* work area for FDCT subroutine */ - JDIMENSION bi; - - sample_data += start_row; /* fold in the vertical offset once */ - - for (bi = 0; bi < num_blocks; bi++, start_col += compptr->DCT_h_scaled_size) { - /* Perform the DCT */ - (*do_dct) (workspace, sample_data, start_col); - - /* Quantize/descale the coefficients, and store into coef_blocks[] */ - { register DCTELEM temp, qval; - register int i; - register JCOEFPTR output_ptr = coef_blocks[bi]; - - for (i = 0; i < DCTSIZE2; i++) { - qval = divisors[i]; - temp = workspace[i]; - /* Divide the coefficient value by qval, ensuring proper rounding. - * Since C does not specify the direction of rounding for negative - * quotients, we have to force the dividend positive for portability. - * - * In most files, at least half of the output values will be zero - * (at default quantization settings, more like three-quarters...) - * so we should ensure that this case is fast. On many machines, - * a comparison is enough cheaper than a divide to make a special test - * a win. Since both inputs will be nonnegative, we need only test - * for a < b to discover whether a/b is 0. - * If your machine's division is fast enough, define FAST_DIVIDE. - */ -#ifdef FAST_DIVIDE -#define DIVIDE_BY(a,b) a /= b -#else -#define DIVIDE_BY(a,b) if (a >= b) a /= b; else a = 0 -#endif - if (temp < 0) { - temp = -temp; - temp += qval>>1; /* for rounding */ - DIVIDE_BY(temp, qval); - temp = -temp; - } else { - temp += qval>>1; /* for rounding */ - DIVIDE_BY(temp, qval); - } - output_ptr[i] = (JCOEF) temp; - } - } - } -} - - -#ifdef DCT_FLOAT_SUPPORTED - -METHODDEF(void) -forward_DCT_float (j_compress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY sample_data, JBLOCKROW coef_blocks, - JDIMENSION start_row, JDIMENSION start_col, - JDIMENSION num_blocks) -/* This version is used for floating-point DCT implementations. */ -{ - /* This routine is heavily used, so it's worth coding it tightly. */ - my_fdct_ptr fdct = (my_fdct_ptr) cinfo->fdct; - float_DCT_method_ptr do_dct = fdct->do_float_dct[compptr->component_index]; - FAST_FLOAT * divisors = fdct->float_divisors[compptr->quant_tbl_no]; - FAST_FLOAT workspace[DCTSIZE2]; /* work area for FDCT subroutine */ - JDIMENSION bi; - - sample_data += start_row; /* fold in the vertical offset once */ - - for (bi = 0; bi < num_blocks; bi++, start_col += compptr->DCT_h_scaled_size) { - /* Perform the DCT */ - (*do_dct) (workspace, sample_data, start_col); - - /* Quantize/descale the coefficients, and store into coef_blocks[] */ - { register FAST_FLOAT temp; - register int i; - register JCOEFPTR output_ptr = coef_blocks[bi]; - - for (i = 0; i < DCTSIZE2; i++) { - /* Apply the quantization and scaling factor */ - temp = workspace[i] * divisors[i]; - /* Round to nearest integer. - * Since C does not specify the direction of rounding for negative - * quotients, we have to force the dividend positive for portability. - * The maximum coefficient size is +-16K (for 12-bit data), so this - * code should work for either 16-bit or 32-bit ints. - */ - output_ptr[i] = (JCOEF) ((int) (temp + (FAST_FLOAT) 16384.5) - 16384); - } - } - } -} - -#endif /* DCT_FLOAT_SUPPORTED */ - - -/* - * Initialize for a processing pass. - * Verify that all referenced Q-tables are present, and set up - * the divisor table for each one. - * In the current implementation, DCT of all components is done during - * the first pass, even if only some components will be output in the - * first scan. Hence all components should be examined here. - */ - -METHODDEF(void) -start_pass_fdctmgr (j_compress_ptr cinfo) -{ - my_fdct_ptr fdct = (my_fdct_ptr) cinfo->fdct; - int ci, qtblno, i; - jpeg_component_info *compptr; - int method = 0; - JQUANT_TBL * qtbl; - DCTELEM * dtbl; - - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - /* Select the proper DCT routine for this component's scaling */ - switch ((compptr->DCT_h_scaled_size << 8) + compptr->DCT_v_scaled_size) { -#ifdef DCT_SCALING_SUPPORTED - case ((1 << 8) + 1): - fdct->do_dct[ci] = jpeg_fdct_1x1; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((2 << 8) + 2): - fdct->do_dct[ci] = jpeg_fdct_2x2; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((3 << 8) + 3): - fdct->do_dct[ci] = jpeg_fdct_3x3; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((4 << 8) + 4): - fdct->do_dct[ci] = jpeg_fdct_4x4; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((5 << 8) + 5): - fdct->do_dct[ci] = jpeg_fdct_5x5; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((6 << 8) + 6): - fdct->do_dct[ci] = jpeg_fdct_6x6; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((7 << 8) + 7): - fdct->do_dct[ci] = jpeg_fdct_7x7; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((9 << 8) + 9): - fdct->do_dct[ci] = jpeg_fdct_9x9; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((10 << 8) + 10): - fdct->do_dct[ci] = jpeg_fdct_10x10; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((11 << 8) + 11): - fdct->do_dct[ci] = jpeg_fdct_11x11; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((12 << 8) + 12): - fdct->do_dct[ci] = jpeg_fdct_12x12; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((13 << 8) + 13): - fdct->do_dct[ci] = jpeg_fdct_13x13; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((14 << 8) + 14): - fdct->do_dct[ci] = jpeg_fdct_14x14; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((15 << 8) + 15): - fdct->do_dct[ci] = jpeg_fdct_15x15; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((16 << 8) + 16): - fdct->do_dct[ci] = jpeg_fdct_16x16; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((16 << 8) + 8): - fdct->do_dct[ci] = jpeg_fdct_16x8; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((14 << 8) + 7): - fdct->do_dct[ci] = jpeg_fdct_14x7; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((12 << 8) + 6): - fdct->do_dct[ci] = jpeg_fdct_12x6; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((10 << 8) + 5): - fdct->do_dct[ci] = jpeg_fdct_10x5; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((8 << 8) + 4): - fdct->do_dct[ci] = jpeg_fdct_8x4; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((6 << 8) + 3): - fdct->do_dct[ci] = jpeg_fdct_6x3; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((4 << 8) + 2): - fdct->do_dct[ci] = jpeg_fdct_4x2; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((2 << 8) + 1): - fdct->do_dct[ci] = jpeg_fdct_2x1; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((8 << 8) + 16): - fdct->do_dct[ci] = jpeg_fdct_8x16; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((7 << 8) + 14): - fdct->do_dct[ci] = jpeg_fdct_7x14; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((6 << 8) + 12): - fdct->do_dct[ci] = jpeg_fdct_6x12; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((5 << 8) + 10): - fdct->do_dct[ci] = jpeg_fdct_5x10; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((4 << 8) + 8): - fdct->do_dct[ci] = jpeg_fdct_4x8; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((3 << 8) + 6): - fdct->do_dct[ci] = jpeg_fdct_3x6; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((2 << 8) + 4): - fdct->do_dct[ci] = jpeg_fdct_2x4; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; - case ((1 << 8) + 2): - fdct->do_dct[ci] = jpeg_fdct_1x2; - method = JDCT_ISLOW; /* jfdctint uses islow-style table */ - break; -#endif - case ((DCTSIZE << 8) + DCTSIZE): - switch (cinfo->dct_method) { -#ifdef DCT_ISLOW_SUPPORTED - case JDCT_ISLOW: - fdct->do_dct[ci] = jpeg_fdct_islow; - method = JDCT_ISLOW; - break; -#endif -#ifdef DCT_IFAST_SUPPORTED - case JDCT_IFAST: - fdct->do_dct[ci] = jpeg_fdct_ifast; - method = JDCT_IFAST; - break; -#endif -#ifdef DCT_FLOAT_SUPPORTED - case JDCT_FLOAT: - fdct->do_float_dct[ci] = jpeg_fdct_float; - method = JDCT_FLOAT; - break; -#endif - default: - ERREXIT(cinfo, JERR_NOT_COMPILED); - break; - } - break; - default: - ERREXIT2(cinfo, JERR_BAD_DCTSIZE, - compptr->DCT_h_scaled_size, compptr->DCT_v_scaled_size); - break; - } - qtblno = compptr->quant_tbl_no; - /* Make sure specified quantization table is present */ - if (qtblno < 0 || qtblno >= NUM_QUANT_TBLS || - cinfo->quant_tbl_ptrs[qtblno] == NULL) - ERREXIT1(cinfo, JERR_NO_QUANT_TABLE, qtblno); - qtbl = cinfo->quant_tbl_ptrs[qtblno]; - /* Compute divisors for this quant table */ - /* We may do this more than once for same table, but it's not a big deal */ - switch (method) { -#ifdef PROVIDE_ISLOW_TABLES - case JDCT_ISLOW: - /* For LL&M IDCT method, divisors are equal to raw quantization - * coefficients multiplied by 8 (to counteract scaling). - */ - if (fdct->divisors[qtblno] == NULL) { - fdct->divisors[qtblno] = (DCTELEM *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - DCTSIZE2 * SIZEOF(DCTELEM)); - } - dtbl = fdct->divisors[qtblno]; - for (i = 0; i < DCTSIZE2; i++) { - dtbl[i] = ((DCTELEM) qtbl->quantval[i]) << 3; - } - fdct->pub.forward_DCT[ci] = forward_DCT; - break; -#endif -#ifdef DCT_IFAST_SUPPORTED - case JDCT_IFAST: - { - /* For AA&N IDCT method, divisors are equal to quantization - * coefficients scaled by scalefactor[row]*scalefactor[col], where - * scalefactor[0] = 1 - * scalefactor[k] = cos(k*PI/16) * sqrt(2) for k=1..7 - * We apply a further scale factor of 8. - */ -#define CONST_BITS 14 - static const INT16 aanscales[DCTSIZE2] = { - /* precomputed values scaled up by 14 bits */ - 16384, 22725, 21407, 19266, 16384, 12873, 8867, 4520, - 22725, 31521, 29692, 26722, 22725, 17855, 12299, 6270, - 21407, 29692, 27969, 25172, 21407, 16819, 11585, 5906, - 19266, 26722, 25172, 22654, 19266, 15137, 10426, 5315, - 16384, 22725, 21407, 19266, 16384, 12873, 8867, 4520, - 12873, 17855, 16819, 15137, 12873, 10114, 6967, 3552, - 8867, 12299, 11585, 10426, 8867, 6967, 4799, 2446, - 4520, 6270, 5906, 5315, 4520, 3552, 2446, 1247 - }; - SHIFT_TEMPS - - if (fdct->divisors[qtblno] == NULL) { - fdct->divisors[qtblno] = (DCTELEM *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - DCTSIZE2 * SIZEOF(DCTELEM)); - } - dtbl = fdct->divisors[qtblno]; - for (i = 0; i < DCTSIZE2; i++) { - dtbl[i] = (DCTELEM) - DESCALE(MULTIPLY16V16((INT32) qtbl->quantval[i], - (INT32) aanscales[i]), - CONST_BITS-3); - } - } - fdct->pub.forward_DCT[ci] = forward_DCT; - break; -#endif -#ifdef DCT_FLOAT_SUPPORTED - case JDCT_FLOAT: - { - /* For float AA&N IDCT method, divisors are equal to quantization - * coefficients scaled by scalefactor[row]*scalefactor[col], where - * scalefactor[0] = 1 - * scalefactor[k] = cos(k*PI/16) * sqrt(2) for k=1..7 - * We apply a further scale factor of 8. - * What's actually stored is 1/divisor so that the inner loop can - * use a multiplication rather than a division. - */ - FAST_FLOAT * fdtbl; - int row, col; - static const double aanscalefactor[DCTSIZE] = { - 1.0, 1.387039845, 1.306562965, 1.175875602, - 1.0, 0.785694958, 0.541196100, 0.275899379 - }; - - if (fdct->float_divisors[qtblno] == NULL) { - fdct->float_divisors[qtblno] = (FAST_FLOAT *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - DCTSIZE2 * SIZEOF(FAST_FLOAT)); - } - fdtbl = fdct->float_divisors[qtblno]; - i = 0; - for (row = 0; row < DCTSIZE; row++) { - for (col = 0; col < DCTSIZE; col++) { - fdtbl[i] = (FAST_FLOAT) - (1.0 / (((double) qtbl->quantval[i] * - aanscalefactor[row] * aanscalefactor[col] * 8.0))); - i++; - } - } - } - fdct->pub.forward_DCT[ci] = forward_DCT_float; - break; -#endif - default: - ERREXIT(cinfo, JERR_NOT_COMPILED); - break; - } - } -} - - -/* - * Initialize FDCT manager. - */ - -GLOBAL(void) -jinit_forward_dct (j_compress_ptr cinfo) -{ - my_fdct_ptr fdct; - int i; - - fdct = (my_fdct_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_fdct_controller)); - cinfo->fdct = (struct jpeg_forward_dct *) fdct; - fdct->pub.start_pass = start_pass_fdctmgr; - - /* Mark divisor tables unallocated */ - for (i = 0; i < NUM_QUANT_TBLS; i++) { - fdct->divisors[i] = NULL; -#ifdef DCT_FLOAT_SUPPORTED - fdct->float_divisors[i] = NULL; -#endif - } -} diff --git a/src/3rdparty/libjpeg/jchuff.c b/src/3rdparty/libjpeg/jchuff.c deleted file mode 100644 index 257d7aa1f5..0000000000 --- a/src/3rdparty/libjpeg/jchuff.c +++ /dev/null @@ -1,1576 +0,0 @@ -/* - * jchuff.c - * - * Copyright (C) 1991-1997, Thomas G. Lane. - * Modified 2006-2009 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains Huffman entropy encoding routines. - * Both sequential and progressive modes are supported in this single module. - * - * Much of the complexity here has to do with supporting output suspension. - * If the data destination module demands suspension, we want to be able to - * back up to the start of the current MCU. To do this, we copy state - * variables into local working storage, and update them back to the - * permanent JPEG objects only upon successful completion of an MCU. - * - * We do not support output suspension for the progressive JPEG mode, since - * the library currently does not allow multiple-scan files to be written - * with output suspension. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* The legal range of a DCT coefficient is - * -1024 .. +1023 for 8-bit data; - * -16384 .. +16383 for 12-bit data. - * Hence the magnitude should always fit in 10 or 14 bits respectively. - */ - -#if BITS_IN_JSAMPLE == 8 -#define MAX_COEF_BITS 10 -#else -#define MAX_COEF_BITS 14 -#endif - -/* Derived data constructed for each Huffman table */ - -typedef struct { - unsigned int ehufco[256]; /* code for each symbol */ - char ehufsi[256]; /* length of code for each symbol */ - /* If no code has been allocated for a symbol S, ehufsi[S] contains 0 */ -} c_derived_tbl; - - -/* Expanded entropy encoder object for Huffman encoding. - * - * The savable_state subrecord contains fields that change within an MCU, - * but must not be updated permanently until we complete the MCU. - */ - -typedef struct { - INT32 put_buffer; /* current bit-accumulation buffer */ - int put_bits; /* # of bits now in it */ - int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */ -} savable_state; - -/* This macro is to work around compilers with missing or broken - * structure assignment. You'll need to fix this code if you have - * such a compiler and you change MAX_COMPS_IN_SCAN. - */ - -#ifndef NO_STRUCT_ASSIGN -#define ASSIGN_STATE(dest,src) ((dest) = (src)) -#else -#if MAX_COMPS_IN_SCAN == 4 -#define ASSIGN_STATE(dest,src) \ - ((dest).put_buffer = (src).put_buffer, \ - (dest).put_bits = (src).put_bits, \ - (dest).last_dc_val[0] = (src).last_dc_val[0], \ - (dest).last_dc_val[1] = (src).last_dc_val[1], \ - (dest).last_dc_val[2] = (src).last_dc_val[2], \ - (dest).last_dc_val[3] = (src).last_dc_val[3]) -#endif -#endif - - -typedef struct { - struct jpeg_entropy_encoder pub; /* public fields */ - - savable_state saved; /* Bit buffer & DC state at start of MCU */ - - /* These fields are NOT loaded into local working state. */ - unsigned int restarts_to_go; /* MCUs left in this restart interval */ - int next_restart_num; /* next restart number to write (0-7) */ - - /* Pointers to derived tables (these workspaces have image lifespan) */ - c_derived_tbl * dc_derived_tbls[NUM_HUFF_TBLS]; - c_derived_tbl * ac_derived_tbls[NUM_HUFF_TBLS]; - - /* Statistics tables for optimization */ - long * dc_count_ptrs[NUM_HUFF_TBLS]; - long * ac_count_ptrs[NUM_HUFF_TBLS]; - - /* Following fields used only in progressive mode */ - - /* Mode flag: TRUE for optimization, FALSE for actual data output */ - boolean gather_statistics; - - /* next_output_byte/free_in_buffer are local copies of cinfo->dest fields. - */ - JOCTET * next_output_byte; /* => next byte to write in buffer */ - size_t free_in_buffer; /* # of byte spaces remaining in buffer */ - j_compress_ptr cinfo; /* link to cinfo (needed for dump_buffer) */ - - /* Coding status for AC components */ - int ac_tbl_no; /* the table number of the single component */ - unsigned int EOBRUN; /* run length of EOBs */ - unsigned int BE; /* # of buffered correction bits before MCU */ - char * bit_buffer; /* buffer for correction bits (1 per char) */ - /* packing correction bits tightly would save some space but cost time... */ -} huff_entropy_encoder; - -typedef huff_entropy_encoder * huff_entropy_ptr; - -/* Working state while writing an MCU (sequential mode). - * This struct contains all the fields that are needed by subroutines. - */ - -typedef struct { - JOCTET * next_output_byte; /* => next byte to write in buffer */ - size_t free_in_buffer; /* # of byte spaces remaining in buffer */ - savable_state cur; /* Current bit buffer & DC state */ - j_compress_ptr cinfo; /* dump_buffer needs access to this */ -} working_state; - -/* MAX_CORR_BITS is the number of bits the AC refinement correction-bit - * buffer can hold. Larger sizes may slightly improve compression, but - * 1000 is already well into the realm of overkill. - * The minimum safe size is 64 bits. - */ - -#define MAX_CORR_BITS 1000 /* Max # of correction bits I can buffer */ - -/* IRIGHT_SHIFT is like RIGHT_SHIFT, but works on int rather than INT32. - * We assume that int right shift is unsigned if INT32 right shift is, - * which should be safe. - */ - -#ifdef RIGHT_SHIFT_IS_UNSIGNED -#define ISHIFT_TEMPS int ishift_temp; -#define IRIGHT_SHIFT(x,shft) \ - ((ishift_temp = (x)) < 0 ? \ - (ishift_temp >> (shft)) | ((~0) << (16-(shft))) : \ - (ishift_temp >> (shft))) -#else -#define ISHIFT_TEMPS -#define IRIGHT_SHIFT(x,shft) ((x) >> (shft)) -#endif - - -/* - * Compute the derived values for a Huffman table. - * This routine also performs some validation checks on the table. - */ - -LOCAL(void) -jpeg_make_c_derived_tbl (j_compress_ptr cinfo, boolean isDC, int tblno, - c_derived_tbl ** pdtbl) -{ - JHUFF_TBL *htbl; - c_derived_tbl *dtbl; - int p, i, l, lastp, si, maxsymbol; - char huffsize[257]; - unsigned int huffcode[257]; - unsigned int code; - - /* Note that huffsize[] and huffcode[] are filled in code-length order, - * paralleling the order of the symbols themselves in htbl->huffval[]. - */ - - /* Find the input Huffman table */ - if (tblno < 0 || tblno >= NUM_HUFF_TBLS) - ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno); - htbl = - isDC ? cinfo->dc_huff_tbl_ptrs[tblno] : cinfo->ac_huff_tbl_ptrs[tblno]; - if (htbl == NULL) - ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno); - - /* Allocate a workspace if we haven't already done so. */ - if (*pdtbl == NULL) - *pdtbl = (c_derived_tbl *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(c_derived_tbl)); - dtbl = *pdtbl; - - /* Figure C.1: make table of Huffman code length for each symbol */ - - p = 0; - for (l = 1; l <= 16; l++) { - i = (int) htbl->bits[l]; - if (i < 0 || p + i > 256) /* protect against table overrun */ - ERREXIT(cinfo, JERR_BAD_HUFF_TABLE); - while (i--) - huffsize[p++] = (char) l; - } - huffsize[p] = 0; - lastp = p; - - /* Figure C.2: generate the codes themselves */ - /* We also validate that the counts represent a legal Huffman code tree. */ - - code = 0; - si = huffsize[0]; - p = 0; - while (huffsize[p]) { - while (((int) huffsize[p]) == si) { - huffcode[p++] = code; - code++; - } - /* code is now 1 more than the last code used for codelength si; but - * it must still fit in si bits, since no code is allowed to be all ones. - */ - if (((INT32) code) >= (((INT32) 1) << si)) - ERREXIT(cinfo, JERR_BAD_HUFF_TABLE); - code <<= 1; - si++; - } - - /* Figure C.3: generate encoding tables */ - /* These are code and size indexed by symbol value */ - - /* Set all codeless symbols to have code length 0; - * this lets us detect duplicate VAL entries here, and later - * allows emit_bits to detect any attempt to emit such symbols. - */ - MEMZERO(dtbl->ehufsi, SIZEOF(dtbl->ehufsi)); - - /* This is also a convenient place to check for out-of-range - * and duplicated VAL entries. We allow 0..255 for AC symbols - * but only 0..15 for DC. (We could constrain them further - * based on data depth and mode, but this seems enough.) - */ - maxsymbol = isDC ? 15 : 255; - - for (p = 0; p < lastp; p++) { - i = htbl->huffval[p]; - if (i < 0 || i > maxsymbol || dtbl->ehufsi[i]) - ERREXIT(cinfo, JERR_BAD_HUFF_TABLE); - dtbl->ehufco[i] = huffcode[p]; - dtbl->ehufsi[i] = huffsize[p]; - } -} - - -/* Outputting bytes to the file. - * NB: these must be called only when actually outputting, - * that is, entropy->gather_statistics == FALSE. - */ - -/* Emit a byte, taking 'action' if must suspend. */ -#define emit_byte_s(state,val,action) \ - { *(state)->next_output_byte++ = (JOCTET) (val); \ - if (--(state)->free_in_buffer == 0) \ - if (! dump_buffer_s(state)) \ - { action; } } - -/* Emit a byte */ -#define emit_byte_e(entropy,val) \ - { *(entropy)->next_output_byte++ = (JOCTET) (val); \ - if (--(entropy)->free_in_buffer == 0) \ - dump_buffer_e(entropy); } - - -LOCAL(boolean) -dump_buffer_s (working_state * state) -/* Empty the output buffer; return TRUE if successful, FALSE if must suspend */ -{ - struct jpeg_destination_mgr * dest = state->cinfo->dest; - - if (! (*dest->empty_output_buffer) (state->cinfo)) - return FALSE; - /* After a successful buffer dump, must reset buffer pointers */ - state->next_output_byte = dest->next_output_byte; - state->free_in_buffer = dest->free_in_buffer; - return TRUE; -} - - -LOCAL(void) -dump_buffer_e (huff_entropy_ptr entropy) -/* Empty the output buffer; we do not support suspension in this case. */ -{ - struct jpeg_destination_mgr * dest = entropy->cinfo->dest; - - if (! (*dest->empty_output_buffer) (entropy->cinfo)) - ERREXIT(entropy->cinfo, JERR_CANT_SUSPEND); - /* After a successful buffer dump, must reset buffer pointers */ - entropy->next_output_byte = dest->next_output_byte; - entropy->free_in_buffer = dest->free_in_buffer; -} - - -/* Outputting bits to the file */ - -/* Only the right 24 bits of put_buffer are used; the valid bits are - * left-justified in this part. At most 16 bits can be passed to emit_bits - * in one call, and we never retain more than 7 bits in put_buffer - * between calls, so 24 bits are sufficient. - */ - -INLINE -LOCAL(boolean) -emit_bits_s (working_state * state, unsigned int code, int size) -/* Emit some bits; return TRUE if successful, FALSE if must suspend */ -{ - /* This routine is heavily used, so it's worth coding tightly. */ - register INT32 put_buffer = (INT32) code; - register int put_bits = state->cur.put_bits; - - /* if size is 0, caller used an invalid Huffman table entry */ - if (size == 0) - ERREXIT(state->cinfo, JERR_HUFF_MISSING_CODE); - - put_buffer &= (((INT32) 1)<<size) - 1; /* mask off any extra bits in code */ - - put_bits += size; /* new number of bits in buffer */ - - put_buffer <<= 24 - put_bits; /* align incoming bits */ - - put_buffer |= state->cur.put_buffer; /* and merge with old buffer contents */ - - while (put_bits >= 8) { - int c = (int) ((put_buffer >> 16) & 0xFF); - - emit_byte_s(state, c, return FALSE); - if (c == 0xFF) { /* need to stuff a zero byte? */ - emit_byte_s(state, 0, return FALSE); - } - put_buffer <<= 8; - put_bits -= 8; - } - - state->cur.put_buffer = put_buffer; /* update state variables */ - state->cur.put_bits = put_bits; - - return TRUE; -} - - -INLINE -LOCAL(void) -emit_bits_e (huff_entropy_ptr entropy, unsigned int code, int size) -/* Emit some bits, unless we are in gather mode */ -{ - /* This routine is heavily used, so it's worth coding tightly. */ - register INT32 put_buffer = (INT32) code; - register int put_bits = entropy->saved.put_bits; - - /* if size is 0, caller used an invalid Huffman table entry */ - if (size == 0) - ERREXIT(entropy->cinfo, JERR_HUFF_MISSING_CODE); - - if (entropy->gather_statistics) - return; /* do nothing if we're only getting stats */ - - put_buffer &= (((INT32) 1)<<size) - 1; /* mask off any extra bits in code */ - - put_bits += size; /* new number of bits in buffer */ - - put_buffer <<= 24 - put_bits; /* align incoming bits */ - - /* and merge with old buffer contents */ - put_buffer |= entropy->saved.put_buffer; - - while (put_bits >= 8) { - int c = (int) ((put_buffer >> 16) & 0xFF); - - emit_byte_e(entropy, c); - if (c == 0xFF) { /* need to stuff a zero byte? */ - emit_byte_e(entropy, 0); - } - put_buffer <<= 8; - put_bits -= 8; - } - - entropy->saved.put_buffer = put_buffer; /* update variables */ - entropy->saved.put_bits = put_bits; -} - - -LOCAL(boolean) -flush_bits_s (working_state * state) -{ - if (! emit_bits_s(state, 0x7F, 7)) /* fill any partial byte with ones */ - return FALSE; - state->cur.put_buffer = 0; /* and reset bit-buffer to empty */ - state->cur.put_bits = 0; - return TRUE; -} - - -LOCAL(void) -flush_bits_e (huff_entropy_ptr entropy) -{ - emit_bits_e(entropy, 0x7F, 7); /* fill any partial byte with ones */ - entropy->saved.put_buffer = 0; /* and reset bit-buffer to empty */ - entropy->saved.put_bits = 0; -} - - -/* - * Emit (or just count) a Huffman symbol. - */ - -INLINE -LOCAL(void) -emit_dc_symbol (huff_entropy_ptr entropy, int tbl_no, int symbol) -{ - if (entropy->gather_statistics) - entropy->dc_count_ptrs[tbl_no][symbol]++; - else { - c_derived_tbl * tbl = entropy->dc_derived_tbls[tbl_no]; - emit_bits_e(entropy, tbl->ehufco[symbol], tbl->ehufsi[symbol]); - } -} - - -INLINE -LOCAL(void) -emit_ac_symbol (huff_entropy_ptr entropy, int tbl_no, int symbol) -{ - if (entropy->gather_statistics) - entropy->ac_count_ptrs[tbl_no][symbol]++; - else { - c_derived_tbl * tbl = entropy->ac_derived_tbls[tbl_no]; - emit_bits_e(entropy, tbl->ehufco[symbol], tbl->ehufsi[symbol]); - } -} - - -/* - * Emit bits from a correction bit buffer. - */ - -LOCAL(void) -emit_buffered_bits (huff_entropy_ptr entropy, char * bufstart, - unsigned int nbits) -{ - if (entropy->gather_statistics) - return; /* no real work */ - - while (nbits > 0) { - emit_bits_e(entropy, (unsigned int) (*bufstart), 1); - bufstart++; - nbits--; - } -} - - -/* - * Emit any pending EOBRUN symbol. - */ - -LOCAL(void) -emit_eobrun (huff_entropy_ptr entropy) -{ - register int temp, nbits; - - if (entropy->EOBRUN > 0) { /* if there is any pending EOBRUN */ - temp = entropy->EOBRUN; - nbits = 0; - while ((temp >>= 1)) - nbits++; - /* safety check: shouldn't happen given limited correction-bit buffer */ - if (nbits > 14) - ERREXIT(entropy->cinfo, JERR_HUFF_MISSING_CODE); - - emit_ac_symbol(entropy, entropy->ac_tbl_no, nbits << 4); - if (nbits) - emit_bits_e(entropy, entropy->EOBRUN, nbits); - - entropy->EOBRUN = 0; - - /* Emit any buffered correction bits */ - emit_buffered_bits(entropy, entropy->bit_buffer, entropy->BE); - entropy->BE = 0; - } -} - - -/* - * Emit a restart marker & resynchronize predictions. - */ - -LOCAL(boolean) -emit_restart_s (working_state * state, int restart_num) -{ - int ci; - - if (! flush_bits_s(state)) - return FALSE; - - emit_byte_s(state, 0xFF, return FALSE); - emit_byte_s(state, JPEG_RST0 + restart_num, return FALSE); - - /* Re-initialize DC predictions to 0 */ - for (ci = 0; ci < state->cinfo->comps_in_scan; ci++) - state->cur.last_dc_val[ci] = 0; - - /* The restart counter is not updated until we successfully write the MCU. */ - - return TRUE; -} - - -LOCAL(void) -emit_restart_e (huff_entropy_ptr entropy, int restart_num) -{ - int ci; - - emit_eobrun(entropy); - - if (! entropy->gather_statistics) { - flush_bits_e(entropy); - emit_byte_e(entropy, 0xFF); - emit_byte_e(entropy, JPEG_RST0 + restart_num); - } - - if (entropy->cinfo->Ss == 0) { - /* Re-initialize DC predictions to 0 */ - for (ci = 0; ci < entropy->cinfo->comps_in_scan; ci++) - entropy->saved.last_dc_val[ci] = 0; - } else { - /* Re-initialize all AC-related fields to 0 */ - entropy->EOBRUN = 0; - entropy->BE = 0; - } -} - - -/* - * MCU encoding for DC initial scan (either spectral selection, - * or first pass of successive approximation). - */ - -METHODDEF(boolean) -encode_mcu_DC_first (j_compress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - register int temp, temp2; - register int nbits; - int blkn, ci; - int Al = cinfo->Al; - JBLOCKROW block; - jpeg_component_info * compptr; - ISHIFT_TEMPS - - entropy->next_output_byte = cinfo->dest->next_output_byte; - entropy->free_in_buffer = cinfo->dest->free_in_buffer; - - /* Emit restart marker if needed */ - if (cinfo->restart_interval) - if (entropy->restarts_to_go == 0) - emit_restart_e(entropy, entropy->next_restart_num); - - /* Encode the MCU data blocks */ - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - block = MCU_data[blkn]; - ci = cinfo->MCU_membership[blkn]; - compptr = cinfo->cur_comp_info[ci]; - - /* Compute the DC value after the required point transform by Al. - * This is simply an arithmetic right shift. - */ - temp2 = IRIGHT_SHIFT((int) ((*block)[0]), Al); - - /* DC differences are figured on the point-transformed values. */ - temp = temp2 - entropy->saved.last_dc_val[ci]; - entropy->saved.last_dc_val[ci] = temp2; - - /* Encode the DC coefficient difference per section G.1.2.1 */ - temp2 = temp; - if (temp < 0) { - temp = -temp; /* temp is abs value of input */ - /* For a negative input, want temp2 = bitwise complement of abs(input) */ - /* This code assumes we are on a two's complement machine */ - temp2--; - } - - /* Find the number of bits needed for the magnitude of the coefficient */ - nbits = 0; - while (temp) { - nbits++; - temp >>= 1; - } - /* Check for out-of-range coefficient values. - * Since we're encoding a difference, the range limit is twice as much. - */ - if (nbits > MAX_COEF_BITS+1) - ERREXIT(cinfo, JERR_BAD_DCT_COEF); - - /* Count/emit the Huffman-coded symbol for the number of bits */ - emit_dc_symbol(entropy, compptr->dc_tbl_no, nbits); - - /* Emit that number of bits of the value, if positive, */ - /* or the complement of its magnitude, if negative. */ - if (nbits) /* emit_bits rejects calls with size 0 */ - emit_bits_e(entropy, (unsigned int) temp2, nbits); - } - - cinfo->dest->next_output_byte = entropy->next_output_byte; - cinfo->dest->free_in_buffer = entropy->free_in_buffer; - - /* Update restart-interval state too */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) { - entropy->restarts_to_go = cinfo->restart_interval; - entropy->next_restart_num++; - entropy->next_restart_num &= 7; - } - entropy->restarts_to_go--; - } - - return TRUE; -} - - -/* - * MCU encoding for AC initial scan (either spectral selection, - * or first pass of successive approximation). - */ - -METHODDEF(boolean) -encode_mcu_AC_first (j_compress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - register int temp, temp2; - register int nbits; - register int r, k; - int Se, Al; - const int * natural_order; - JBLOCKROW block; - - entropy->next_output_byte = cinfo->dest->next_output_byte; - entropy->free_in_buffer = cinfo->dest->free_in_buffer; - - /* Emit restart marker if needed */ - if (cinfo->restart_interval) - if (entropy->restarts_to_go == 0) - emit_restart_e(entropy, entropy->next_restart_num); - - Se = cinfo->Se; - Al = cinfo->Al; - natural_order = cinfo->natural_order; - - /* Encode the MCU data block */ - block = MCU_data[0]; - - /* Encode the AC coefficients per section G.1.2.2, fig. G.3 */ - - r = 0; /* r = run length of zeros */ - - for (k = cinfo->Ss; k <= Se; k++) { - if ((temp = (*block)[natural_order[k]]) == 0) { - r++; - continue; - } - /* We must apply the point transform by Al. For AC coefficients this - * is an integer division with rounding towards 0. To do this portably - * in C, we shift after obtaining the absolute value; so the code is - * interwoven with finding the abs value (temp) and output bits (temp2). - */ - if (temp < 0) { - temp = -temp; /* temp is abs value of input */ - temp >>= Al; /* apply the point transform */ - /* For a negative coef, want temp2 = bitwise complement of abs(coef) */ - temp2 = ~temp; - } else { - temp >>= Al; /* apply the point transform */ - temp2 = temp; - } - /* Watch out for case that nonzero coef is zero after point transform */ - if (temp == 0) { - r++; - continue; - } - - /* Emit any pending EOBRUN */ - if (entropy->EOBRUN > 0) - emit_eobrun(entropy); - /* if run length > 15, must emit special run-length-16 codes (0xF0) */ - while (r > 15) { - emit_ac_symbol(entropy, entropy->ac_tbl_no, 0xF0); - r -= 16; - } - - /* Find the number of bits needed for the magnitude of the coefficient */ - nbits = 1; /* there must be at least one 1 bit */ - while ((temp >>= 1)) - nbits++; - /* Check for out-of-range coefficient values */ - if (nbits > MAX_COEF_BITS) - ERREXIT(cinfo, JERR_BAD_DCT_COEF); - - /* Count/emit Huffman symbol for run length / number of bits */ - emit_ac_symbol(entropy, entropy->ac_tbl_no, (r << 4) + nbits); - - /* Emit that number of bits of the value, if positive, */ - /* or the complement of its magnitude, if negative. */ - emit_bits_e(entropy, (unsigned int) temp2, nbits); - - r = 0; /* reset zero run length */ - } - - if (r > 0) { /* If there are trailing zeroes, */ - entropy->EOBRUN++; /* count an EOB */ - if (entropy->EOBRUN == 0x7FFF) - emit_eobrun(entropy); /* force it out to avoid overflow */ - } - - cinfo->dest->next_output_byte = entropy->next_output_byte; - cinfo->dest->free_in_buffer = entropy->free_in_buffer; - - /* Update restart-interval state too */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) { - entropy->restarts_to_go = cinfo->restart_interval; - entropy->next_restart_num++; - entropy->next_restart_num &= 7; - } - entropy->restarts_to_go--; - } - - return TRUE; -} - - -/* - * MCU encoding for DC successive approximation refinement scan. - * Note: we assume such scans can be multi-component, although the spec - * is not very clear on the point. - */ - -METHODDEF(boolean) -encode_mcu_DC_refine (j_compress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - register int temp; - int blkn; - int Al = cinfo->Al; - JBLOCKROW block; - - entropy->next_output_byte = cinfo->dest->next_output_byte; - entropy->free_in_buffer = cinfo->dest->free_in_buffer; - - /* Emit restart marker if needed */ - if (cinfo->restart_interval) - if (entropy->restarts_to_go == 0) - emit_restart_e(entropy, entropy->next_restart_num); - - /* Encode the MCU data blocks */ - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - block = MCU_data[blkn]; - - /* We simply emit the Al'th bit of the DC coefficient value. */ - temp = (*block)[0]; - emit_bits_e(entropy, (unsigned int) (temp >> Al), 1); - } - - cinfo->dest->next_output_byte = entropy->next_output_byte; - cinfo->dest->free_in_buffer = entropy->free_in_buffer; - - /* Update restart-interval state too */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) { - entropy->restarts_to_go = cinfo->restart_interval; - entropy->next_restart_num++; - entropy->next_restart_num &= 7; - } - entropy->restarts_to_go--; - } - - return TRUE; -} - - -/* - * MCU encoding for AC successive approximation refinement scan. - */ - -METHODDEF(boolean) -encode_mcu_AC_refine (j_compress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - register int temp; - register int r, k; - int EOB; - char *BR_buffer; - unsigned int BR; - int Se, Al; - const int * natural_order; - JBLOCKROW block; - int absvalues[DCTSIZE2]; - - entropy->next_output_byte = cinfo->dest->next_output_byte; - entropy->free_in_buffer = cinfo->dest->free_in_buffer; - - /* Emit restart marker if needed */ - if (cinfo->restart_interval) - if (entropy->restarts_to_go == 0) - emit_restart_e(entropy, entropy->next_restart_num); - - Se = cinfo->Se; - Al = cinfo->Al; - natural_order = cinfo->natural_order; - - /* Encode the MCU data block */ - block = MCU_data[0]; - - /* It is convenient to make a pre-pass to determine the transformed - * coefficients' absolute values and the EOB position. - */ - EOB = 0; - for (k = cinfo->Ss; k <= Se; k++) { - temp = (*block)[natural_order[k]]; - /* We must apply the point transform by Al. For AC coefficients this - * is an integer division with rounding towards 0. To do this portably - * in C, we shift after obtaining the absolute value. - */ - if (temp < 0) - temp = -temp; /* temp is abs value of input */ - temp >>= Al; /* apply the point transform */ - absvalues[k] = temp; /* save abs value for main pass */ - if (temp == 1) - EOB = k; /* EOB = index of last newly-nonzero coef */ - } - - /* Encode the AC coefficients per section G.1.2.3, fig. G.7 */ - - r = 0; /* r = run length of zeros */ - BR = 0; /* BR = count of buffered bits added now */ - BR_buffer = entropy->bit_buffer + entropy->BE; /* Append bits to buffer */ - - for (k = cinfo->Ss; k <= Se; k++) { - if ((temp = absvalues[k]) == 0) { - r++; - continue; - } - - /* Emit any required ZRLs, but not if they can be folded into EOB */ - while (r > 15 && k <= EOB) { - /* emit any pending EOBRUN and the BE correction bits */ - emit_eobrun(entropy); - /* Emit ZRL */ - emit_ac_symbol(entropy, entropy->ac_tbl_no, 0xF0); - r -= 16; - /* Emit buffered correction bits that must be associated with ZRL */ - emit_buffered_bits(entropy, BR_buffer, BR); - BR_buffer = entropy->bit_buffer; /* BE bits are gone now */ - BR = 0; - } - - /* If the coef was previously nonzero, it only needs a correction bit. - * NOTE: a straight translation of the spec's figure G.7 would suggest - * that we also need to test r > 15. But if r > 15, we can only get here - * if k > EOB, which implies that this coefficient is not 1. - */ - if (temp > 1) { - /* The correction bit is the next bit of the absolute value. */ - BR_buffer[BR++] = (char) (temp & 1); - continue; - } - - /* Emit any pending EOBRUN and the BE correction bits */ - emit_eobrun(entropy); - - /* Count/emit Huffman symbol for run length / number of bits */ - emit_ac_symbol(entropy, entropy->ac_tbl_no, (r << 4) + 1); - - /* Emit output bit for newly-nonzero coef */ - temp = ((*block)[natural_order[k]] < 0) ? 0 : 1; - emit_bits_e(entropy, (unsigned int) temp, 1); - - /* Emit buffered correction bits that must be associated with this code */ - emit_buffered_bits(entropy, BR_buffer, BR); - BR_buffer = entropy->bit_buffer; /* BE bits are gone now */ - BR = 0; - r = 0; /* reset zero run length */ - } - - if (r > 0 || BR > 0) { /* If there are trailing zeroes, */ - entropy->EOBRUN++; /* count an EOB */ - entropy->BE += BR; /* concat my correction bits to older ones */ - /* We force out the EOB if we risk either: - * 1. overflow of the EOB counter; - * 2. overflow of the correction bit buffer during the next MCU. - */ - if (entropy->EOBRUN == 0x7FFF || entropy->BE > (MAX_CORR_BITS-DCTSIZE2+1)) - emit_eobrun(entropy); - } - - cinfo->dest->next_output_byte = entropy->next_output_byte; - cinfo->dest->free_in_buffer = entropy->free_in_buffer; - - /* Update restart-interval state too */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) { - entropy->restarts_to_go = cinfo->restart_interval; - entropy->next_restart_num++; - entropy->next_restart_num &= 7; - } - entropy->restarts_to_go--; - } - - return TRUE; -} - - -/* Encode a single block's worth of coefficients */ - -LOCAL(boolean) -encode_one_block (working_state * state, JCOEFPTR block, int last_dc_val, - c_derived_tbl *dctbl, c_derived_tbl *actbl) -{ - register int temp, temp2; - register int nbits; - register int k, r, i; - int Se = state->cinfo->lim_Se; - const int * natural_order = state->cinfo->natural_order; - - /* Encode the DC coefficient difference per section F.1.2.1 */ - - temp = temp2 = block[0] - last_dc_val; - - if (temp < 0) { - temp = -temp; /* temp is abs value of input */ - /* For a negative input, want temp2 = bitwise complement of abs(input) */ - /* This code assumes we are on a two's complement machine */ - temp2--; - } - - /* Find the number of bits needed for the magnitude of the coefficient */ - nbits = 0; - while (temp) { - nbits++; - temp >>= 1; - } - /* Check for out-of-range coefficient values. - * Since we're encoding a difference, the range limit is twice as much. - */ - if (nbits > MAX_COEF_BITS+1) - ERREXIT(state->cinfo, JERR_BAD_DCT_COEF); - - /* Emit the Huffman-coded symbol for the number of bits */ - if (! emit_bits_s(state, dctbl->ehufco[nbits], dctbl->ehufsi[nbits])) - return FALSE; - - /* Emit that number of bits of the value, if positive, */ - /* or the complement of its magnitude, if negative. */ - if (nbits) /* emit_bits rejects calls with size 0 */ - if (! emit_bits_s(state, (unsigned int) temp2, nbits)) - return FALSE; - - /* Encode the AC coefficients per section F.1.2.2 */ - - r = 0; /* r = run length of zeros */ - - for (k = 1; k <= Se; k++) { - if ((temp = block[natural_order[k]]) == 0) { - r++; - } else { - /* if run length > 15, must emit special run-length-16 codes (0xF0) */ - while (r > 15) { - if (! emit_bits_s(state, actbl->ehufco[0xF0], actbl->ehufsi[0xF0])) - return FALSE; - r -= 16; - } - - temp2 = temp; - if (temp < 0) { - temp = -temp; /* temp is abs value of input */ - /* This code assumes we are on a two's complement machine */ - temp2--; - } - - /* Find the number of bits needed for the magnitude of the coefficient */ - nbits = 1; /* there must be at least one 1 bit */ - while ((temp >>= 1)) - nbits++; - /* Check for out-of-range coefficient values */ - if (nbits > MAX_COEF_BITS) - ERREXIT(state->cinfo, JERR_BAD_DCT_COEF); - - /* Emit Huffman symbol for run length / number of bits */ - i = (r << 4) + nbits; - if (! emit_bits_s(state, actbl->ehufco[i], actbl->ehufsi[i])) - return FALSE; - - /* Emit that number of bits of the value, if positive, */ - /* or the complement of its magnitude, if negative. */ - if (! emit_bits_s(state, (unsigned int) temp2, nbits)) - return FALSE; - - r = 0; - } - } - - /* If the last coef(s) were zero, emit an end-of-block code */ - if (r > 0) - if (! emit_bits_s(state, actbl->ehufco[0], actbl->ehufsi[0])) - return FALSE; - - return TRUE; -} - - -/* - * Encode and output one MCU's worth of Huffman-compressed coefficients. - */ - -METHODDEF(boolean) -encode_mcu_huff (j_compress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - working_state state; - int blkn, ci; - jpeg_component_info * compptr; - - /* Load up working state */ - state.next_output_byte = cinfo->dest->next_output_byte; - state.free_in_buffer = cinfo->dest->free_in_buffer; - ASSIGN_STATE(state.cur, entropy->saved); - state.cinfo = cinfo; - - /* Emit restart marker if needed */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! emit_restart_s(&state, entropy->next_restart_num)) - return FALSE; - } - - /* Encode the MCU data blocks */ - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - ci = cinfo->MCU_membership[blkn]; - compptr = cinfo->cur_comp_info[ci]; - if (! encode_one_block(&state, - MCU_data[blkn][0], state.cur.last_dc_val[ci], - entropy->dc_derived_tbls[compptr->dc_tbl_no], - entropy->ac_derived_tbls[compptr->ac_tbl_no])) - return FALSE; - /* Update last_dc_val */ - state.cur.last_dc_val[ci] = MCU_data[blkn][0][0]; - } - - /* Completed MCU, so update state */ - cinfo->dest->next_output_byte = state.next_output_byte; - cinfo->dest->free_in_buffer = state.free_in_buffer; - ASSIGN_STATE(entropy->saved, state.cur); - - /* Update restart-interval state too */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) { - entropy->restarts_to_go = cinfo->restart_interval; - entropy->next_restart_num++; - entropy->next_restart_num &= 7; - } - entropy->restarts_to_go--; - } - - return TRUE; -} - - -/* - * Finish up at the end of a Huffman-compressed scan. - */ - -METHODDEF(void) -finish_pass_huff (j_compress_ptr cinfo) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - working_state state; - - if (cinfo->progressive_mode) { - entropy->next_output_byte = cinfo->dest->next_output_byte; - entropy->free_in_buffer = cinfo->dest->free_in_buffer; - - /* Flush out any buffered data */ - emit_eobrun(entropy); - flush_bits_e(entropy); - - cinfo->dest->next_output_byte = entropy->next_output_byte; - cinfo->dest->free_in_buffer = entropy->free_in_buffer; - } else { - /* Load up working state ... flush_bits needs it */ - state.next_output_byte = cinfo->dest->next_output_byte; - state.free_in_buffer = cinfo->dest->free_in_buffer; - ASSIGN_STATE(state.cur, entropy->saved); - state.cinfo = cinfo; - - /* Flush out the last data */ - if (! flush_bits_s(&state)) - ERREXIT(cinfo, JERR_CANT_SUSPEND); - - /* Update state */ - cinfo->dest->next_output_byte = state.next_output_byte; - cinfo->dest->free_in_buffer = state.free_in_buffer; - ASSIGN_STATE(entropy->saved, state.cur); - } -} - - -/* - * Huffman coding optimization. - * - * We first scan the supplied data and count the number of uses of each symbol - * that is to be Huffman-coded. (This process MUST agree with the code above.) - * Then we build a Huffman coding tree for the observed counts. - * Symbols which are not needed at all for the particular image are not - * assigned any code, which saves space in the DHT marker as well as in - * the compressed data. - */ - - -/* Process a single block's worth of coefficients */ - -LOCAL(void) -htest_one_block (j_compress_ptr cinfo, JCOEFPTR block, int last_dc_val, - long dc_counts[], long ac_counts[]) -{ - register int temp; - register int nbits; - register int k, r; - int Se = cinfo->lim_Se; - const int * natural_order = cinfo->natural_order; - - /* Encode the DC coefficient difference per section F.1.2.1 */ - - temp = block[0] - last_dc_val; - if (temp < 0) - temp = -temp; - - /* Find the number of bits needed for the magnitude of the coefficient */ - nbits = 0; - while (temp) { - nbits++; - temp >>= 1; - } - /* Check for out-of-range coefficient values. - * Since we're encoding a difference, the range limit is twice as much. - */ - if (nbits > MAX_COEF_BITS+1) - ERREXIT(cinfo, JERR_BAD_DCT_COEF); - - /* Count the Huffman symbol for the number of bits */ - dc_counts[nbits]++; - - /* Encode the AC coefficients per section F.1.2.2 */ - - r = 0; /* r = run length of zeros */ - - for (k = 1; k <= Se; k++) { - if ((temp = block[natural_order[k]]) == 0) { - r++; - } else { - /* if run length > 15, must emit special run-length-16 codes (0xF0) */ - while (r > 15) { - ac_counts[0xF0]++; - r -= 16; - } - - /* Find the number of bits needed for the magnitude of the coefficient */ - if (temp < 0) - temp = -temp; - - /* Find the number of bits needed for the magnitude of the coefficient */ - nbits = 1; /* there must be at least one 1 bit */ - while ((temp >>= 1)) - nbits++; - /* Check for out-of-range coefficient values */ - if (nbits > MAX_COEF_BITS) - ERREXIT(cinfo, JERR_BAD_DCT_COEF); - - /* Count Huffman symbol for run length / number of bits */ - ac_counts[(r << 4) + nbits]++; - - r = 0; - } - } - - /* If the last coef(s) were zero, emit an end-of-block code */ - if (r > 0) - ac_counts[0]++; -} - - -/* - * Trial-encode one MCU's worth of Huffman-compressed coefficients. - * No data is actually output, so no suspension return is possible. - */ - -METHODDEF(boolean) -encode_mcu_gather (j_compress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int blkn, ci; - jpeg_component_info * compptr; - - /* Take care of restart intervals if needed */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) { - /* Re-initialize DC predictions to 0 */ - for (ci = 0; ci < cinfo->comps_in_scan; ci++) - entropy->saved.last_dc_val[ci] = 0; - /* Update restart state */ - entropy->restarts_to_go = cinfo->restart_interval; - } - entropy->restarts_to_go--; - } - - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - ci = cinfo->MCU_membership[blkn]; - compptr = cinfo->cur_comp_info[ci]; - htest_one_block(cinfo, MCU_data[blkn][0], entropy->saved.last_dc_val[ci], - entropy->dc_count_ptrs[compptr->dc_tbl_no], - entropy->ac_count_ptrs[compptr->ac_tbl_no]); - entropy->saved.last_dc_val[ci] = MCU_data[blkn][0][0]; - } - - return TRUE; -} - - -/* - * Generate the best Huffman code table for the given counts, fill htbl. - * - * The JPEG standard requires that no symbol be assigned a codeword of all - * one bits (so that padding bits added at the end of a compressed segment - * can't look like a valid code). Because of the canonical ordering of - * codewords, this just means that there must be an unused slot in the - * longest codeword length category. Section K.2 of the JPEG spec suggests - * reserving such a slot by pretending that symbol 256 is a valid symbol - * with count 1. In theory that's not optimal; giving it count zero but - * including it in the symbol set anyway should give a better Huffman code. - * But the theoretically better code actually seems to come out worse in - * practice, because it produces more all-ones bytes (which incur stuffed - * zero bytes in the final file). In any case the difference is tiny. - * - * The JPEG standard requires Huffman codes to be no more than 16 bits long. - * If some symbols have a very small but nonzero probability, the Huffman tree - * must be adjusted to meet the code length restriction. We currently use - * the adjustment method suggested in JPEG section K.2. This method is *not* - * optimal; it may not choose the best possible limited-length code. But - * typically only very-low-frequency symbols will be given less-than-optimal - * lengths, so the code is almost optimal. Experimental comparisons against - * an optimal limited-length-code algorithm indicate that the difference is - * microscopic --- usually less than a hundredth of a percent of total size. - * So the extra complexity of an optimal algorithm doesn't seem worthwhile. - */ - -LOCAL(void) -jpeg_gen_optimal_table (j_compress_ptr cinfo, JHUFF_TBL * htbl, long freq[]) -{ -#define MAX_CLEN 32 /* assumed maximum initial code length */ - UINT8 bits[MAX_CLEN+1]; /* bits[k] = # of symbols with code length k */ - int codesize[257]; /* codesize[k] = code length of symbol k */ - int others[257]; /* next symbol in current branch of tree */ - int c1, c2; - int p, i, j; - long v; - - /* This algorithm is explained in section K.2 of the JPEG standard */ - - MEMZERO(bits, SIZEOF(bits)); - MEMZERO(codesize, SIZEOF(codesize)); - for (i = 0; i < 257; i++) - others[i] = -1; /* init links to empty */ - - freq[256] = 1; /* make sure 256 has a nonzero count */ - /* Including the pseudo-symbol 256 in the Huffman procedure guarantees - * that no real symbol is given code-value of all ones, because 256 - * will be placed last in the largest codeword category. - */ - - /* Huffman's basic algorithm to assign optimal code lengths to symbols */ - - for (;;) { - /* Find the smallest nonzero frequency, set c1 = its symbol */ - /* In case of ties, take the larger symbol number */ - c1 = -1; - v = 1000000000L; - for (i = 0; i <= 256; i++) { - if (freq[i] && freq[i] <= v) { - v = freq[i]; - c1 = i; - } - } - - /* Find the next smallest nonzero frequency, set c2 = its symbol */ - /* In case of ties, take the larger symbol number */ - c2 = -1; - v = 1000000000L; - for (i = 0; i <= 256; i++) { - if (freq[i] && freq[i] <= v && i != c1) { - v = freq[i]; - c2 = i; - } - } - - /* Done if we've merged everything into one frequency */ - if (c2 < 0) - break; - - /* Else merge the two counts/trees */ - freq[c1] += freq[c2]; - freq[c2] = 0; - - /* Increment the codesize of everything in c1's tree branch */ - codesize[c1]++; - while (others[c1] >= 0) { - c1 = others[c1]; - codesize[c1]++; - } - - others[c1] = c2; /* chain c2 onto c1's tree branch */ - - /* Increment the codesize of everything in c2's tree branch */ - codesize[c2]++; - while (others[c2] >= 0) { - c2 = others[c2]; - codesize[c2]++; - } - } - - /* Now count the number of symbols of each code length */ - for (i = 0; i <= 256; i++) { - if (codesize[i]) { - /* The JPEG standard seems to think that this can't happen, */ - /* but I'm paranoid... */ - if (codesize[i] > MAX_CLEN) - ERREXIT(cinfo, JERR_HUFF_CLEN_OVERFLOW); - - bits[codesize[i]]++; - } - } - - /* JPEG doesn't allow symbols with code lengths over 16 bits, so if the pure - * Huffman procedure assigned any such lengths, we must adjust the coding. - * Here is what the JPEG spec says about how this next bit works: - * Since symbols are paired for the longest Huffman code, the symbols are - * removed from this length category two at a time. The prefix for the pair - * (which is one bit shorter) is allocated to one of the pair; then, - * skipping the BITS entry for that prefix length, a code word from the next - * shortest nonzero BITS entry is converted into a prefix for two code words - * one bit longer. - */ - - for (i = MAX_CLEN; i > 16; i--) { - while (bits[i] > 0) { - j = i - 2; /* find length of new prefix to be used */ - while (bits[j] == 0) - j--; - - bits[i] -= 2; /* remove two symbols */ - bits[i-1]++; /* one goes in this length */ - bits[j+1] += 2; /* two new symbols in this length */ - bits[j]--; /* symbol of this length is now a prefix */ - } - } - - /* Remove the count for the pseudo-symbol 256 from the largest codelength */ - while (bits[i] == 0) /* find largest codelength still in use */ - i--; - bits[i]--; - - /* Return final symbol counts (only for lengths 0..16) */ - MEMCOPY(htbl->bits, bits, SIZEOF(htbl->bits)); - - /* Return a list of the symbols sorted by code length */ - /* It's not real clear to me why we don't need to consider the codelength - * changes made above, but the JPEG spec seems to think this works. - */ - p = 0; - for (i = 1; i <= MAX_CLEN; i++) { - for (j = 0; j <= 255; j++) { - if (codesize[j] == i) { - htbl->huffval[p] = (UINT8) j; - p++; - } - } - } - - /* Set sent_table FALSE so updated table will be written to JPEG file. */ - htbl->sent_table = FALSE; -} - - -/* - * Finish up a statistics-gathering pass and create the new Huffman tables. - */ - -METHODDEF(void) -finish_pass_gather (j_compress_ptr cinfo) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int ci, tbl; - jpeg_component_info * compptr; - JHUFF_TBL **htblptr; - boolean did_dc[NUM_HUFF_TBLS]; - boolean did_ac[NUM_HUFF_TBLS]; - - /* It's important not to apply jpeg_gen_optimal_table more than once - * per table, because it clobbers the input frequency counts! - */ - if (cinfo->progressive_mode) - /* Flush out buffered data (all we care about is counting the EOB symbol) */ - emit_eobrun(entropy); - - MEMZERO(did_dc, SIZEOF(did_dc)); - MEMZERO(did_ac, SIZEOF(did_ac)); - - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - compptr = cinfo->cur_comp_info[ci]; - /* DC needs no table for refinement scan */ - if (cinfo->Ss == 0 && cinfo->Ah == 0) { - tbl = compptr->dc_tbl_no; - if (! did_dc[tbl]) { - htblptr = & cinfo->dc_huff_tbl_ptrs[tbl]; - if (*htblptr == NULL) - *htblptr = jpeg_alloc_huff_table((j_common_ptr) cinfo); - jpeg_gen_optimal_table(cinfo, *htblptr, entropy->dc_count_ptrs[tbl]); - did_dc[tbl] = TRUE; - } - } - /* AC needs no table when not present */ - if (cinfo->Se) { - tbl = compptr->ac_tbl_no; - if (! did_ac[tbl]) { - htblptr = & cinfo->ac_huff_tbl_ptrs[tbl]; - if (*htblptr == NULL) - *htblptr = jpeg_alloc_huff_table((j_common_ptr) cinfo); - jpeg_gen_optimal_table(cinfo, *htblptr, entropy->ac_count_ptrs[tbl]); - did_ac[tbl] = TRUE; - } - } - } -} - - -/* - * Initialize for a Huffman-compressed scan. - * If gather_statistics is TRUE, we do not output anything during the scan, - * just count the Huffman symbols used and generate Huffman code tables. - */ - -METHODDEF(void) -start_pass_huff (j_compress_ptr cinfo, boolean gather_statistics) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int ci, tbl; - jpeg_component_info * compptr; - - if (gather_statistics) - entropy->pub.finish_pass = finish_pass_gather; - else - entropy->pub.finish_pass = finish_pass_huff; - - if (cinfo->progressive_mode) { - entropy->cinfo = cinfo; - entropy->gather_statistics = gather_statistics; - - /* We assume jcmaster.c already validated the scan parameters. */ - - /* Select execution routine */ - if (cinfo->Ah == 0) { - if (cinfo->Ss == 0) - entropy->pub.encode_mcu = encode_mcu_DC_first; - else - entropy->pub.encode_mcu = encode_mcu_AC_first; - } else { - if (cinfo->Ss == 0) - entropy->pub.encode_mcu = encode_mcu_DC_refine; - else { - entropy->pub.encode_mcu = encode_mcu_AC_refine; - /* AC refinement needs a correction bit buffer */ - if (entropy->bit_buffer == NULL) - entropy->bit_buffer = (char *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - MAX_CORR_BITS * SIZEOF(char)); - } - } - - /* Initialize AC stuff */ - entropy->ac_tbl_no = cinfo->cur_comp_info[0]->ac_tbl_no; - entropy->EOBRUN = 0; - entropy->BE = 0; - } else { - if (gather_statistics) - entropy->pub.encode_mcu = encode_mcu_gather; - else - entropy->pub.encode_mcu = encode_mcu_huff; - } - - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - compptr = cinfo->cur_comp_info[ci]; - /* DC needs no table for refinement scan */ - if (cinfo->Ss == 0 && cinfo->Ah == 0) { - tbl = compptr->dc_tbl_no; - if (gather_statistics) { - /* Check for invalid table index */ - /* (make_c_derived_tbl does this in the other path) */ - if (tbl < 0 || tbl >= NUM_HUFF_TBLS) - ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tbl); - /* Allocate and zero the statistics tables */ - /* Note that jpeg_gen_optimal_table expects 257 entries in each table! */ - if (entropy->dc_count_ptrs[tbl] == NULL) - entropy->dc_count_ptrs[tbl] = (long *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - 257 * SIZEOF(long)); - MEMZERO(entropy->dc_count_ptrs[tbl], 257 * SIZEOF(long)); - } else { - /* Compute derived values for Huffman tables */ - /* We may do this more than once for a table, but it's not expensive */ - jpeg_make_c_derived_tbl(cinfo, TRUE, tbl, - & entropy->dc_derived_tbls[tbl]); - } - /* Initialize DC predictions to 0 */ - entropy->saved.last_dc_val[ci] = 0; - } - /* AC needs no table when not present */ - if (cinfo->Se) { - tbl = compptr->ac_tbl_no; - if (gather_statistics) { - if (tbl < 0 || tbl >= NUM_HUFF_TBLS) - ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tbl); - if (entropy->ac_count_ptrs[tbl] == NULL) - entropy->ac_count_ptrs[tbl] = (long *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - 257 * SIZEOF(long)); - MEMZERO(entropy->ac_count_ptrs[tbl], 257 * SIZEOF(long)); - } else { - jpeg_make_c_derived_tbl(cinfo, FALSE, tbl, - & entropy->ac_derived_tbls[tbl]); - } - } - } - - /* Initialize bit buffer to empty */ - entropy->saved.put_buffer = 0; - entropy->saved.put_bits = 0; - - /* Initialize restart stuff */ - entropy->restarts_to_go = cinfo->restart_interval; - entropy->next_restart_num = 0; -} - - -/* - * Module initialization routine for Huffman entropy encoding. - */ - -GLOBAL(void) -jinit_huff_encoder (j_compress_ptr cinfo) -{ - huff_entropy_ptr entropy; - int i; - - entropy = (huff_entropy_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(huff_entropy_encoder)); - cinfo->entropy = (struct jpeg_entropy_encoder *) entropy; - entropy->pub.start_pass = start_pass_huff; - - /* Mark tables unallocated */ - for (i = 0; i < NUM_HUFF_TBLS; i++) { - entropy->dc_derived_tbls[i] = entropy->ac_derived_tbls[i] = NULL; - entropy->dc_count_ptrs[i] = entropy->ac_count_ptrs[i] = NULL; - } - - if (cinfo->progressive_mode) - entropy->bit_buffer = NULL; /* needed only in AC refinement scan */ -} diff --git a/src/3rdparty/libjpeg/jcmainct.c b/src/3rdparty/libjpeg/jcmainct.c deleted file mode 100644 index 7de75d1675..0000000000 --- a/src/3rdparty/libjpeg/jcmainct.c +++ /dev/null @@ -1,293 +0,0 @@ -/* - * jcmainct.c - * - * Copyright (C) 1994-1996, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains the main buffer controller for compression. - * The main buffer lies between the pre-processor and the JPEG - * compressor proper; it holds downsampled data in the JPEG colorspace. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Note: currently, there is no operating mode in which a full-image buffer - * is needed at this step. If there were, that mode could not be used with - * "raw data" input, since this module is bypassed in that case. However, - * we've left the code here for possible use in special applications. - */ -#undef FULL_MAIN_BUFFER_SUPPORTED - - -/* Private buffer controller object */ - -typedef struct { - struct jpeg_c_main_controller pub; /* public fields */ - - JDIMENSION cur_iMCU_row; /* number of current iMCU row */ - JDIMENSION rowgroup_ctr; /* counts row groups received in iMCU row */ - boolean suspended; /* remember if we suspended output */ - J_BUF_MODE pass_mode; /* current operating mode */ - - /* If using just a strip buffer, this points to the entire set of buffers - * (we allocate one for each component). In the full-image case, this - * points to the currently accessible strips of the virtual arrays. - */ - JSAMPARRAY buffer[MAX_COMPONENTS]; - -#ifdef FULL_MAIN_BUFFER_SUPPORTED - /* If using full-image storage, this array holds pointers to virtual-array - * control blocks for each component. Unused if not full-image storage. - */ - jvirt_sarray_ptr whole_image[MAX_COMPONENTS]; -#endif -} my_main_controller; - -typedef my_main_controller * my_main_ptr; - - -/* Forward declarations */ -METHODDEF(void) process_data_simple_main - JPP((j_compress_ptr cinfo, JSAMPARRAY input_buf, - JDIMENSION *in_row_ctr, JDIMENSION in_rows_avail)); -#ifdef FULL_MAIN_BUFFER_SUPPORTED -METHODDEF(void) process_data_buffer_main - JPP((j_compress_ptr cinfo, JSAMPARRAY input_buf, - JDIMENSION *in_row_ctr, JDIMENSION in_rows_avail)); -#endif - - -/* - * Initialize for a processing pass. - */ - -METHODDEF(void) -start_pass_main (j_compress_ptr cinfo, J_BUF_MODE pass_mode) -{ - my_main_ptr main = (my_main_ptr) cinfo->main; - - /* Do nothing in raw-data mode. */ - if (cinfo->raw_data_in) - return; - - main->cur_iMCU_row = 0; /* initialize counters */ - main->rowgroup_ctr = 0; - main->suspended = FALSE; - main->pass_mode = pass_mode; /* save mode for use by process_data */ - - switch (pass_mode) { - case JBUF_PASS_THRU: -#ifdef FULL_MAIN_BUFFER_SUPPORTED - if (main->whole_image[0] != NULL) - ERREXIT(cinfo, JERR_BAD_BUFFER_MODE); -#endif - main->pub.process_data = process_data_simple_main; - break; -#ifdef FULL_MAIN_BUFFER_SUPPORTED - case JBUF_SAVE_SOURCE: - case JBUF_CRANK_DEST: - case JBUF_SAVE_AND_PASS: - if (main->whole_image[0] == NULL) - ERREXIT(cinfo, JERR_BAD_BUFFER_MODE); - main->pub.process_data = process_data_buffer_main; - break; -#endif - default: - ERREXIT(cinfo, JERR_BAD_BUFFER_MODE); - break; - } -} - - -/* - * Process some data. - * This routine handles the simple pass-through mode, - * where we have only a strip buffer. - */ - -METHODDEF(void) -process_data_simple_main (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JDIMENSION *in_row_ctr, - JDIMENSION in_rows_avail) -{ - my_main_ptr main = (my_main_ptr) cinfo->main; - - while (main->cur_iMCU_row < cinfo->total_iMCU_rows) { - /* Read input data if we haven't filled the main buffer yet */ - if (main->rowgroup_ctr < (JDIMENSION) cinfo->min_DCT_v_scaled_size) - (*cinfo->prep->pre_process_data) (cinfo, - input_buf, in_row_ctr, in_rows_avail, - main->buffer, &main->rowgroup_ctr, - (JDIMENSION) cinfo->min_DCT_v_scaled_size); - - /* If we don't have a full iMCU row buffered, return to application for - * more data. Note that preprocessor will always pad to fill the iMCU row - * at the bottom of the image. - */ - if (main->rowgroup_ctr != (JDIMENSION) cinfo->min_DCT_v_scaled_size) - return; - - /* Send the completed row to the compressor */ - if (! (*cinfo->coef->compress_data) (cinfo, main->buffer)) { - /* If compressor did not consume the whole row, then we must need to - * suspend processing and return to the application. In this situation - * we pretend we didn't yet consume the last input row; otherwise, if - * it happened to be the last row of the image, the application would - * think we were done. - */ - if (! main->suspended) { - (*in_row_ctr)--; - main->suspended = TRUE; - } - return; - } - /* We did finish the row. Undo our little suspension hack if a previous - * call suspended; then mark the main buffer empty. - */ - if (main->suspended) { - (*in_row_ctr)++; - main->suspended = FALSE; - } - main->rowgroup_ctr = 0; - main->cur_iMCU_row++; - } -} - - -#ifdef FULL_MAIN_BUFFER_SUPPORTED - -/* - * Process some data. - * This routine handles all of the modes that use a full-size buffer. - */ - -METHODDEF(void) -process_data_buffer_main (j_compress_ptr cinfo, - JSAMPARRAY input_buf, JDIMENSION *in_row_ctr, - JDIMENSION in_rows_avail) -{ - my_main_ptr main = (my_main_ptr) cinfo->main; - int ci; - jpeg_component_info *compptr; - boolean writing = (main->pass_mode != JBUF_CRANK_DEST); - - while (main->cur_iMCU_row < cinfo->total_iMCU_rows) { - /* Realign the virtual buffers if at the start of an iMCU row. */ - if (main->rowgroup_ctr == 0) { - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - main->buffer[ci] = (*cinfo->mem->access_virt_sarray) - ((j_common_ptr) cinfo, main->whole_image[ci], - main->cur_iMCU_row * (compptr->v_samp_factor * DCTSIZE), - (JDIMENSION) (compptr->v_samp_factor * DCTSIZE), writing); - } - /* In a read pass, pretend we just read some source data. */ - if (! writing) { - *in_row_ctr += cinfo->max_v_samp_factor * DCTSIZE; - main->rowgroup_ctr = DCTSIZE; - } - } - - /* If a write pass, read input data until the current iMCU row is full. */ - /* Note: preprocessor will pad if necessary to fill the last iMCU row. */ - if (writing) { - (*cinfo->prep->pre_process_data) (cinfo, - input_buf, in_row_ctr, in_rows_avail, - main->buffer, &main->rowgroup_ctr, - (JDIMENSION) DCTSIZE); - /* Return to application if we need more data to fill the iMCU row. */ - if (main->rowgroup_ctr < DCTSIZE) - return; - } - - /* Emit data, unless this is a sink-only pass. */ - if (main->pass_mode != JBUF_SAVE_SOURCE) { - if (! (*cinfo->coef->compress_data) (cinfo, main->buffer)) { - /* If compressor did not consume the whole row, then we must need to - * suspend processing and return to the application. In this situation - * we pretend we didn't yet consume the last input row; otherwise, if - * it happened to be the last row of the image, the application would - * think we were done. - */ - if (! main->suspended) { - (*in_row_ctr)--; - main->suspended = TRUE; - } - return; - } - /* We did finish the row. Undo our little suspension hack if a previous - * call suspended; then mark the main buffer empty. - */ - if (main->suspended) { - (*in_row_ctr)++; - main->suspended = FALSE; - } - } - - /* If get here, we are done with this iMCU row. Mark buffer empty. */ - main->rowgroup_ctr = 0; - main->cur_iMCU_row++; - } -} - -#endif /* FULL_MAIN_BUFFER_SUPPORTED */ - - -/* - * Initialize main buffer controller. - */ - -GLOBAL(void) -jinit_c_main_controller (j_compress_ptr cinfo, boolean need_full_buffer) -{ - my_main_ptr main; - int ci; - jpeg_component_info *compptr; - - main = (my_main_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_main_controller)); - cinfo->main = (struct jpeg_c_main_controller *) main; - main->pub.start_pass = start_pass_main; - - /* We don't need to create a buffer in raw-data mode. */ - if (cinfo->raw_data_in) - return; - - /* Create the buffer. It holds downsampled data, so each component - * may be of a different size. - */ - if (need_full_buffer) { -#ifdef FULL_MAIN_BUFFER_SUPPORTED - /* Allocate a full-image virtual array for each component */ - /* Note we pad the bottom to a multiple of the iMCU height */ - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - main->whole_image[ci] = (*cinfo->mem->request_virt_sarray) - ((j_common_ptr) cinfo, JPOOL_IMAGE, FALSE, - compptr->width_in_blocks * compptr->DCT_h_scaled_size, - (JDIMENSION) jround_up((long) compptr->height_in_blocks, - (long) compptr->v_samp_factor) * DCTSIZE, - (JDIMENSION) (compptr->v_samp_factor * compptr->DCT_v_scaled_size)); - } -#else - ERREXIT(cinfo, JERR_BAD_BUFFER_MODE); -#endif - } else { -#ifdef FULL_MAIN_BUFFER_SUPPORTED - main->whole_image[0] = NULL; /* flag for no virtual arrays */ -#endif - /* Allocate a strip buffer for each component */ - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - main->buffer[ci] = (*cinfo->mem->alloc_sarray) - ((j_common_ptr) cinfo, JPOOL_IMAGE, - compptr->width_in_blocks * compptr->DCT_h_scaled_size, - (JDIMENSION) (compptr->v_samp_factor * compptr->DCT_v_scaled_size)); - } - } -} diff --git a/src/3rdparty/libjpeg/jconfig.bcc b/src/3rdparty/libjpeg/jconfig.bcc deleted file mode 100644 index e4da3d72c2..0000000000 --- a/src/3rdparty/libjpeg/jconfig.bcc +++ /dev/null @@ -1,48 +0,0 @@ -/* jconfig.bcc --- jconfig.h for Borland C (Turbo C) on MS-DOS or OS/2. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#ifdef __MSDOS__ -#define NEED_FAR_POINTERS /* for small or medium memory model */ -#endif -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN /* this assumes you have -w-stu in CFLAGS */ - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#ifdef __MSDOS__ -#define USE_MSDOS_MEMMGR /* Define this if you use jmemdos.c */ -#define MAX_ALLOC_CHUNK 65520L /* Maximum request to malloc() */ -#define USE_FMEM /* Borland has _fmemcpy() and _fmemset() */ -#endif - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE -#define USE_SETMODE /* Borland has setmode() */ -#ifdef __MSDOS__ -#define NEED_SIGNAL_CATCHER /* Define this if you use jmemdos.c */ -#endif -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.cfg b/src/3rdparty/libjpeg/jconfig.cfg deleted file mode 100644 index bb7435c9ff..0000000000 --- a/src/3rdparty/libjpeg/jconfig.cfg +++ /dev/null @@ -1,53 +0,0 @@ -/* jconfig.cfg --- source file edited by configure script */ -/* see jconfig.txt for explanations */ - -#undef HAVE_PROTOTYPES -#undef HAVE_UNSIGNED_CHAR -#undef HAVE_UNSIGNED_SHORT -#undef void -#undef const -#undef CHAR_IS_UNSIGNED -#undef HAVE_STDDEF_H -#undef HAVE_STDLIB_H -#undef HAVE_LOCALE_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS -#undef NEED_SHORT_EXTERNAL_NAMES -/* Define this if you get warnings about undefined structures. */ -#undef INCOMPLETE_TYPES_BROKEN - -/* Define "boolean" as unsigned char, not int, on Windows systems. */ -#ifdef _WIN32 -#ifndef __RPCNDR_H__ /* don't conflict if rpcndr.h already read */ -typedef unsigned char boolean; -#endif -#define HAVE_BOOLEAN /* prevent jmorecfg.h from redefining it */ -#endif - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED -#undef INLINE -/* These are for configuring the JPEG memory manager. */ -#undef DEFAULT_MAX_MEM -#undef NO_MKTEMP - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#undef TWO_FILE_COMMANDLINE -#undef NEED_SIGNAL_CATCHER -#undef DONT_USE_B_MODE - -/* Define this if you want percent-done progress reports from cjpeg/djpeg. */ -#undef PROGRESS_REPORT - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.dj b/src/3rdparty/libjpeg/jconfig.dj deleted file mode 100644 index a0d4092f20..0000000000 --- a/src/3rdparty/libjpeg/jconfig.dj +++ /dev/null @@ -1,38 +0,0 @@ -/* jconfig.dj --- jconfig.h for DJGPP (Delorie's GNU C port) on MS-DOS. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS /* DJGPP uses flat 32-bit addressing */ -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#undef TWO_FILE_COMMANDLINE /* optional */ -#define USE_SETMODE /* Needed to make one-file style work in DJGPP */ -#undef NEED_SIGNAL_CATCHER /* Define this if you use jmemname.c */ -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.mac b/src/3rdparty/libjpeg/jconfig.mac deleted file mode 100644 index 70ed66c187..0000000000 --- a/src/3rdparty/libjpeg/jconfig.mac +++ /dev/null @@ -1,43 +0,0 @@ -/* jconfig.mac --- jconfig.h for CodeWarrior on Apple Macintosh */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#define USE_MAC_MEMMGR /* Define this if you use jmemmac.c */ - -#define ALIGN_TYPE long /* Needed for 680x0 Macs */ - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define USE_CCOMMAND /* Command line reader for Macintosh */ -#define TWO_FILE_COMMANDLINE /* Binary I/O thru stdin/stdout doesn't work */ - -#undef NEED_SIGNAL_CATCHER -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.manx b/src/3rdparty/libjpeg/jconfig.manx deleted file mode 100644 index cd529d7d15..0000000000 --- a/src/3rdparty/libjpeg/jconfig.manx +++ /dev/null @@ -1,43 +0,0 @@ -/* jconfig.manx --- jconfig.h for Amiga systems using Manx Aztec C ver 5.x. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#define TEMP_DIRECTORY "JPEGTMP:" /* recommended setting for Amiga */ - -#define SHORTxSHORT_32 /* produces better DCT code with Aztec C */ - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE -#define NEED_SIGNAL_CATCHER -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#define signal_catcher _abort /* hack for Aztec C naming requirements */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.mc6 b/src/3rdparty/libjpeg/jconfig.mc6 deleted file mode 100644 index ad5651b8ce..0000000000 --- a/src/3rdparty/libjpeg/jconfig.mc6 +++ /dev/null @@ -1,52 +0,0 @@ -/* jconfig.mc6 --- jconfig.h for Microsoft C on MS-DOS, version 6.00A & up. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#define NEED_FAR_POINTERS /* for small or medium memory model */ -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#define USE_MSDOS_MEMMGR /* Define this if you use jmemdos.c */ - -#define MAX_ALLOC_CHUNK 65520L /* Maximum request to malloc() */ - -#define USE_FMEM /* Microsoft has _fmemcpy() and _fmemset() */ - -#define NEED_FHEAPMIN /* far heap management routines are broken */ - -#define SHORTxLCONST_32 /* enable compiler-specific DCT optimization */ -/* Note: the above define is known to improve the code with Microsoft C 6.00A. - * I do not know whether it is good for later compiler versions. - * Please report any info on this point to jpeg-info@uc.ag. - */ - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE -#define USE_SETMODE /* Microsoft has setmode() */ -#define NEED_SIGNAL_CATCHER /* Define this if you use jmemdos.c */ -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.sas b/src/3rdparty/libjpeg/jconfig.sas deleted file mode 100644 index b8a1819259..0000000000 --- a/src/3rdparty/libjpeg/jconfig.sas +++ /dev/null @@ -1,43 +0,0 @@ -/* jconfig.sas --- jconfig.h for Amiga systems using SAS C 6.0 and up. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#define TEMP_DIRECTORY "JPEGTMP:" /* recommended setting for Amiga */ - -#define NO_MKTEMP /* SAS C doesn't have mktemp() */ - -#define SHORTxSHORT_32 /* produces better DCT code with SAS C */ - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE -#define NEED_SIGNAL_CATCHER -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.st b/src/3rdparty/libjpeg/jconfig.st deleted file mode 100644 index 5afa0b6ce5..0000000000 --- a/src/3rdparty/libjpeg/jconfig.st +++ /dev/null @@ -1,42 +0,0 @@ -/* jconfig.st --- jconfig.h for Atari ST/STE/TT using Pure C or Turbo C. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS -#undef NEED_SHORT_EXTERNAL_NAMES -#define INCOMPLETE_TYPES_BROKEN /* suppress undefined-structure warnings */ - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#define ALIGN_TYPE long /* apparently double is a weird size? */ - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE /* optional -- undef if you like Unix style */ -/* Note: if you undef TWO_FILE_COMMANDLINE, you may need to define - * USE_SETMODE. Some Atari compilers require it, some do not. - */ -#define NEED_SIGNAL_CATCHER /* needed if you use jmemname.c */ -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.txt b/src/3rdparty/libjpeg/jconfig.txt deleted file mode 100644 index b96d312492..0000000000 --- a/src/3rdparty/libjpeg/jconfig.txt +++ /dev/null @@ -1,164 +0,0 @@ -/* - * jconfig.txt - * - * Copyright (C) 1991-1994, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file documents the configuration options that are required to - * customize the JPEG software for a particular system. - * - * The actual configuration options for a particular installation are stored - * in jconfig.h. On many machines, jconfig.h can be generated automatically - * or copied from one of the "canned" jconfig files that we supply. But if - * you need to generate a jconfig.h file by hand, this file tells you how. - * - * DO NOT EDIT THIS FILE --- IT WON'T ACCOMPLISH ANYTHING. - * EDIT A COPY NAMED JCONFIG.H. - */ - - -/* - * These symbols indicate the properties of your machine or compiler. - * #define the symbol if yes, #undef it if no. - */ - -/* Does your compiler support function prototypes? - * (If not, you also need to use ansi2knr, see install.txt) - */ -#define HAVE_PROTOTYPES - -/* Does your compiler support the declaration "unsigned char" ? - * How about "unsigned short" ? - */ -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT - -/* Define "void" as "char" if your compiler doesn't know about type void. - * NOTE: be sure to define void such that "void *" represents the most general - * pointer type, e.g., that returned by malloc(). - */ -/* #define void char */ - -/* Define "const" as empty if your compiler doesn't know the "const" keyword. - */ -/* #define const */ - -/* Define this if an ordinary "char" type is unsigned. - * If you're not sure, leaving it undefined will work at some cost in speed. - * If you defined HAVE_UNSIGNED_CHAR then the speed difference is minimal. - */ -#undef CHAR_IS_UNSIGNED - -/* Define this if your system has an ANSI-conforming <stddef.h> file. - */ -#define HAVE_STDDEF_H - -/* Define this if your system has an ANSI-conforming <stdlib.h> file. - */ -#define HAVE_STDLIB_H - -/* Define this if your system does not have an ANSI/SysV <string.h>, - * but does have a BSD-style <strings.h>. - */ -#undef NEED_BSD_STRINGS - -/* Define this if your system does not provide typedef size_t in any of the - * ANSI-standard places (stddef.h, stdlib.h, or stdio.h), but places it in - * <sys/types.h> instead. - */ -#undef NEED_SYS_TYPES_H - -/* For 80x86 machines, you need to define NEED_FAR_POINTERS, - * unless you are using a large-data memory model or 80386 flat-memory mode. - * On less brain-damaged CPUs this symbol must not be defined. - * (Defining this symbol causes large data structures to be referenced through - * "far" pointers and to be allocated with a special version of malloc.) - */ -#undef NEED_FAR_POINTERS - -/* Define this if your linker needs global names to be unique in less - * than the first 15 characters. - */ -#undef NEED_SHORT_EXTERNAL_NAMES - -/* Although a real ANSI C compiler can deal perfectly well with pointers to - * unspecified structures (see "incomplete types" in the spec), a few pre-ANSI - * and pseudo-ANSI compilers get confused. To keep one of these bozos happy, - * define INCOMPLETE_TYPES_BROKEN. This is not recommended unless you - * actually get "missing structure definition" warnings or errors while - * compiling the JPEG code. - */ -#undef INCOMPLETE_TYPES_BROKEN - -/* Define "boolean" as unsigned char, not int, on Windows systems. - */ -#ifdef _WIN32 -#ifndef __RPCNDR_H__ /* don't conflict if rpcndr.h already read */ -typedef unsigned char boolean; -#endif -#define HAVE_BOOLEAN /* prevent jmorecfg.h from redefining it */ -#endif - - -/* - * The following options affect code selection within the JPEG library, - * but they don't need to be visible to applications using the library. - * To minimize application namespace pollution, the symbols won't be - * defined unless JPEG_INTERNALS has been defined. - */ - -#ifdef JPEG_INTERNALS - -/* Define this if your compiler implements ">>" on signed values as a logical - * (unsigned) shift; leave it undefined if ">>" is a signed (arithmetic) shift, - * which is the normal and rational definition. - */ -#undef RIGHT_SHIFT_IS_UNSIGNED - - -#endif /* JPEG_INTERNALS */ - - -/* - * The remaining options do not affect the JPEG library proper, - * but only the sample applications cjpeg/djpeg (see cjpeg.c, djpeg.c). - * Other applications can ignore these. - */ - -#ifdef JPEG_CJPEG_DJPEG - -/* These defines indicate which image (non-JPEG) file formats are allowed. */ - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -/* Define this if you want to name both input and output files on the command - * line, rather than using stdout and optionally stdin. You MUST do this if - * your system can't cope with binary I/O to stdin/stdout. See comments at - * head of cjpeg.c or djpeg.c. - */ -#undef TWO_FILE_COMMANDLINE - -/* Define this if your system needs explicit cleanup of temporary files. - * This is crucial under MS-DOS, where the temporary "files" may be areas - * of extended memory; on most other systems it's not as important. - */ -#undef NEED_SIGNAL_CATCHER - -/* By default, we open image files with fopen(...,"rb") or fopen(...,"wb"). - * This is necessary on systems that distinguish text files from binary files, - * and is harmless on most systems that don't. If you have one of the rare - * systems that complains about the "b" spec, define this symbol. - */ -#undef DONT_USE_B_MODE - -/* Define this if you want percent-done progress reports from cjpeg/djpeg. - */ -#undef PROGRESS_REPORT - - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.vc b/src/3rdparty/libjpeg/jconfig.vc deleted file mode 100644 index 679404da4e..0000000000 --- a/src/3rdparty/libjpeg/jconfig.vc +++ /dev/null @@ -1,45 +0,0 @@ -/* jconfig.vc --- jconfig.h for Microsoft Visual C++ on Windows 95 or NT. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS /* we presume a 32-bit flat memory model */ -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -/* Define "boolean" as unsigned char, not int, per Windows custom */ -#ifndef __RPCNDR_H__ /* don't conflict if rpcndr.h already read */ -typedef unsigned char boolean; -#endif -#define HAVE_BOOLEAN /* prevent jmorecfg.h from redefining it */ - - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE /* optional */ -#define USE_SETMODE /* Microsoft has setmode() */ -#undef NEED_SIGNAL_CATCHER -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.vms b/src/3rdparty/libjpeg/jconfig.vms deleted file mode 100644 index 8337b0b69b..0000000000 --- a/src/3rdparty/libjpeg/jconfig.vms +++ /dev/null @@ -1,37 +0,0 @@ -/* jconfig.vms --- jconfig.h for use on Digital VMS. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#undef CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#define TWO_FILE_COMMANDLINE /* Needed on VMS */ -#undef NEED_SIGNAL_CATCHER -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jconfig.wat b/src/3rdparty/libjpeg/jconfig.wat deleted file mode 100644 index 190cc75fd5..0000000000 --- a/src/3rdparty/libjpeg/jconfig.wat +++ /dev/null @@ -1,38 +0,0 @@ -/* jconfig.wat --- jconfig.h for Watcom C/C++ on MS-DOS or OS/2. */ -/* see jconfig.txt for explanations */ - -#define HAVE_PROTOTYPES -#define HAVE_UNSIGNED_CHAR -#define HAVE_UNSIGNED_SHORT -/* #define void char */ -/* #define const */ -#define CHAR_IS_UNSIGNED -#define HAVE_STDDEF_H -#define HAVE_STDLIB_H -#undef NEED_BSD_STRINGS -#undef NEED_SYS_TYPES_H -#undef NEED_FAR_POINTERS /* Watcom uses flat 32-bit addressing */ -#undef NEED_SHORT_EXTERNAL_NAMES -#undef INCOMPLETE_TYPES_BROKEN - -#ifdef JPEG_INTERNALS - -#undef RIGHT_SHIFT_IS_UNSIGNED - -#endif /* JPEG_INTERNALS */ - -#ifdef JPEG_CJPEG_DJPEG - -#define BMP_SUPPORTED /* BMP image file format */ -#define GIF_SUPPORTED /* GIF image file format */ -#define PPM_SUPPORTED /* PBMPLUS PPM/PGM image file format */ -#undef RLE_SUPPORTED /* Utah RLE image file format */ -#define TARGA_SUPPORTED /* Targa image file format */ - -#undef TWO_FILE_COMMANDLINE /* optional */ -#define USE_SETMODE /* Needed to make one-file style work in Watcom */ -#undef NEED_SIGNAL_CATCHER /* Define this if you use jmemname.c */ -#undef DONT_USE_B_MODE -#undef PROGRESS_REPORT /* optional */ - -#endif /* JPEG_CJPEG_DJPEG */ diff --git a/src/3rdparty/libjpeg/jdapistd.c b/src/3rdparty/libjpeg/jdapistd.c deleted file mode 100644 index 9d74537772..0000000000 --- a/src/3rdparty/libjpeg/jdapistd.c +++ /dev/null @@ -1,275 +0,0 @@ -/* - * jdapistd.c - * - * Copyright (C) 1994-1996, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains application interface code for the decompression half - * of the JPEG library. These are the "standard" API routines that are - * used in the normal full-decompression case. They are not used by a - * transcoding-only application. Note that if an application links in - * jpeg_start_decompress, it will end up linking in the entire decompressor. - * We thus must separate this file from jdapimin.c to avoid linking the - * whole decompression library into a transcoder. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Forward declarations */ -LOCAL(boolean) output_pass_setup JPP((j_decompress_ptr cinfo)); - - -/* - * Decompression initialization. - * jpeg_read_header must be completed before calling this. - * - * If a multipass operating mode was selected, this will do all but the - * last pass, and thus may take a great deal of time. - * - * Returns FALSE if suspended. The return value need be inspected only if - * a suspending data source is used. - */ - -GLOBAL(boolean) -jpeg_start_decompress (j_decompress_ptr cinfo) -{ - if (cinfo->global_state == DSTATE_READY) { - /* First call: initialize master control, select active modules */ - jinit_master_decompress(cinfo); - if (cinfo->buffered_image) { - /* No more work here; expecting jpeg_start_output next */ - cinfo->global_state = DSTATE_BUFIMAGE; - return TRUE; - } - cinfo->global_state = DSTATE_PRELOAD; - } - if (cinfo->global_state == DSTATE_PRELOAD) { - /* If file has multiple scans, absorb them all into the coef buffer */ - if (cinfo->inputctl->has_multiple_scans) { -#ifdef D_MULTISCAN_FILES_SUPPORTED - for (;;) { - int retcode; - /* Call progress monitor hook if present */ - if (cinfo->progress != NULL) - (*cinfo->progress->progress_monitor) ((j_common_ptr) cinfo); - /* Absorb some more input */ - retcode = (*cinfo->inputctl->consume_input) (cinfo); - if (retcode == JPEG_SUSPENDED) - return FALSE; - if (retcode == JPEG_REACHED_EOI) - break; - /* Advance progress counter if appropriate */ - if (cinfo->progress != NULL && - (retcode == JPEG_ROW_COMPLETED || retcode == JPEG_REACHED_SOS)) { - if (++cinfo->progress->pass_counter >= cinfo->progress->pass_limit) { - /* jdmaster underestimated number of scans; ratchet up one scan */ - cinfo->progress->pass_limit += (long) cinfo->total_iMCU_rows; - } - } - } -#else - ERREXIT(cinfo, JERR_NOT_COMPILED); -#endif /* D_MULTISCAN_FILES_SUPPORTED */ - } - cinfo->output_scan_number = cinfo->input_scan_number; - } else if (cinfo->global_state != DSTATE_PRESCAN) - ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state); - /* Perform any dummy output passes, and set up for the final pass */ - return output_pass_setup(cinfo); -} - - -/* - * Set up for an output pass, and perform any dummy pass(es) needed. - * Common subroutine for jpeg_start_decompress and jpeg_start_output. - * Entry: global_state = DSTATE_PRESCAN only if previously suspended. - * Exit: If done, returns TRUE and sets global_state for proper output mode. - * If suspended, returns FALSE and sets global_state = DSTATE_PRESCAN. - */ - -LOCAL(boolean) -output_pass_setup (j_decompress_ptr cinfo) -{ - if (cinfo->global_state != DSTATE_PRESCAN) { - /* First call: do pass setup */ - (*cinfo->master->prepare_for_output_pass) (cinfo); - cinfo->output_scanline = 0; - cinfo->global_state = DSTATE_PRESCAN; - } - /* Loop over any required dummy passes */ - while (cinfo->master->is_dummy_pass) { -#ifdef QUANT_2PASS_SUPPORTED - /* Crank through the dummy pass */ - while (cinfo->output_scanline < cinfo->output_height) { - JDIMENSION last_scanline; - /* Call progress monitor hook if present */ - if (cinfo->progress != NULL) { - cinfo->progress->pass_counter = (long) cinfo->output_scanline; - cinfo->progress->pass_limit = (long) cinfo->output_height; - (*cinfo->progress->progress_monitor) ((j_common_ptr) cinfo); - } - /* Process some data */ - last_scanline = cinfo->output_scanline; - (*cinfo->main->process_data) (cinfo, (JSAMPARRAY) NULL, - &cinfo->output_scanline, (JDIMENSION) 0); - if (cinfo->output_scanline == last_scanline) - return FALSE; /* No progress made, must suspend */ - } - /* Finish up dummy pass, and set up for another one */ - (*cinfo->master->finish_output_pass) (cinfo); - (*cinfo->master->prepare_for_output_pass) (cinfo); - cinfo->output_scanline = 0; -#else - ERREXIT(cinfo, JERR_NOT_COMPILED); -#endif /* QUANT_2PASS_SUPPORTED */ - } - /* Ready for application to drive output pass through - * jpeg_read_scanlines or jpeg_read_raw_data. - */ - cinfo->global_state = cinfo->raw_data_out ? DSTATE_RAW_OK : DSTATE_SCANNING; - return TRUE; -} - - -/* - * Read some scanlines of data from the JPEG decompressor. - * - * The return value will be the number of lines actually read. - * This may be less than the number requested in several cases, - * including bottom of image, data source suspension, and operating - * modes that emit multiple scanlines at a time. - * - * Note: we warn about excess calls to jpeg_read_scanlines() since - * this likely signals an application programmer error. However, - * an oversize buffer (max_lines > scanlines remaining) is not an error. - */ - -GLOBAL(JDIMENSION) -jpeg_read_scanlines (j_decompress_ptr cinfo, JSAMPARRAY scanlines, - JDIMENSION max_lines) -{ - JDIMENSION row_ctr; - - if (cinfo->global_state != DSTATE_SCANNING) - ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state); - if (cinfo->output_scanline >= cinfo->output_height) { - WARNMS(cinfo, JWRN_TOO_MUCH_DATA); - return 0; - } - - /* Call progress monitor hook if present */ - if (cinfo->progress != NULL) { - cinfo->progress->pass_counter = (long) cinfo->output_scanline; - cinfo->progress->pass_limit = (long) cinfo->output_height; - (*cinfo->progress->progress_monitor) ((j_common_ptr) cinfo); - } - - /* Process some data */ - row_ctr = 0; - (*cinfo->main->process_data) (cinfo, scanlines, &row_ctr, max_lines); - cinfo->output_scanline += row_ctr; - return row_ctr; -} - - -/* - * Alternate entry point to read raw data. - * Processes exactly one iMCU row per call, unless suspended. - */ - -GLOBAL(JDIMENSION) -jpeg_read_raw_data (j_decompress_ptr cinfo, JSAMPIMAGE data, - JDIMENSION max_lines) -{ - JDIMENSION lines_per_iMCU_row; - - if (cinfo->global_state != DSTATE_RAW_OK) - ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state); - if (cinfo->output_scanline >= cinfo->output_height) { - WARNMS(cinfo, JWRN_TOO_MUCH_DATA); - return 0; - } - - /* Call progress monitor hook if present */ - if (cinfo->progress != NULL) { - cinfo->progress->pass_counter = (long) cinfo->output_scanline; - cinfo->progress->pass_limit = (long) cinfo->output_height; - (*cinfo->progress->progress_monitor) ((j_common_ptr) cinfo); - } - - /* Verify that at least one iMCU row can be returned. */ - lines_per_iMCU_row = cinfo->max_v_samp_factor * cinfo->min_DCT_v_scaled_size; - if (max_lines < lines_per_iMCU_row) - ERREXIT(cinfo, JERR_BUFFER_SIZE); - - /* Decompress directly into user's buffer. */ - if (! (*cinfo->coef->decompress_data) (cinfo, data)) - return 0; /* suspension forced, can do nothing more */ - - /* OK, we processed one iMCU row. */ - cinfo->output_scanline += lines_per_iMCU_row; - return lines_per_iMCU_row; -} - - -/* Additional entry points for buffered-image mode. */ - -#ifdef D_MULTISCAN_FILES_SUPPORTED - -/* - * Initialize for an output pass in buffered-image mode. - */ - -GLOBAL(boolean) -jpeg_start_output (j_decompress_ptr cinfo, int scan_number) -{ - if (cinfo->global_state != DSTATE_BUFIMAGE && - cinfo->global_state != DSTATE_PRESCAN) - ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state); - /* Limit scan number to valid range */ - if (scan_number <= 0) - scan_number = 1; - if (cinfo->inputctl->eoi_reached && - scan_number > cinfo->input_scan_number) - scan_number = cinfo->input_scan_number; - cinfo->output_scan_number = scan_number; - /* Perform any dummy output passes, and set up for the real pass */ - return output_pass_setup(cinfo); -} - - -/* - * Finish up after an output pass in buffered-image mode. - * - * Returns FALSE if suspended. The return value need be inspected only if - * a suspending data source is used. - */ - -GLOBAL(boolean) -jpeg_finish_output (j_decompress_ptr cinfo) -{ - if ((cinfo->global_state == DSTATE_SCANNING || - cinfo->global_state == DSTATE_RAW_OK) && cinfo->buffered_image) { - /* Terminate this pass. */ - /* We do not require the whole pass to have been completed. */ - (*cinfo->master->finish_output_pass) (cinfo); - cinfo->global_state = DSTATE_BUFPOST; - } else if (cinfo->global_state != DSTATE_BUFPOST) { - /* BUFPOST = repeat call after a suspension, anything else is error */ - ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state); - } - /* Read markers looking for SOS or EOI */ - while (cinfo->input_scan_number <= cinfo->output_scan_number && - ! cinfo->inputctl->eoi_reached) { - if ((*cinfo->inputctl->consume_input) (cinfo) == JPEG_SUSPENDED) - return FALSE; /* Suspend, come back later */ - } - cinfo->global_state = DSTATE_BUFIMAGE; - return TRUE; -} - -#endif /* D_MULTISCAN_FILES_SUPPORTED */ diff --git a/src/3rdparty/libjpeg/jdcolor.c b/src/3rdparty/libjpeg/jdcolor.c deleted file mode 100644 index 6c04dfe8aa..0000000000 --- a/src/3rdparty/libjpeg/jdcolor.c +++ /dev/null @@ -1,396 +0,0 @@ -/* - * jdcolor.c - * - * Copyright (C) 1991-1997, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains output colorspace conversion routines. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Private subobject */ - -typedef struct { - struct jpeg_color_deconverter pub; /* public fields */ - - /* Private state for YCC->RGB conversion */ - int * Cr_r_tab; /* => table for Cr to R conversion */ - int * Cb_b_tab; /* => table for Cb to B conversion */ - INT32 * Cr_g_tab; /* => table for Cr to G conversion */ - INT32 * Cb_g_tab; /* => table for Cb to G conversion */ -} my_color_deconverter; - -typedef my_color_deconverter * my_cconvert_ptr; - - -/**************** YCbCr -> RGB conversion: most common case **************/ - -/* - * YCbCr is defined per CCIR 601-1, except that Cb and Cr are - * normalized to the range 0..MAXJSAMPLE rather than -0.5 .. 0.5. - * The conversion equations to be implemented are therefore - * R = Y + 1.40200 * Cr - * G = Y - 0.34414 * Cb - 0.71414 * Cr - * B = Y + 1.77200 * Cb - * where Cb and Cr represent the incoming values less CENTERJSAMPLE. - * (These numbers are derived from TIFF 6.0 section 21, dated 3-June-92.) - * - * To avoid floating-point arithmetic, we represent the fractional constants - * as integers scaled up by 2^16 (about 4 digits precision); we have to divide - * the products by 2^16, with appropriate rounding, to get the correct answer. - * Notice that Y, being an integral input, does not contribute any fraction - * so it need not participate in the rounding. - * - * For even more speed, we avoid doing any multiplications in the inner loop - * by precalculating the constants times Cb and Cr for all possible values. - * For 8-bit JSAMPLEs this is very reasonable (only 256 entries per table); - * for 12-bit samples it is still acceptable. It's not very reasonable for - * 16-bit samples, but if you want lossless storage you shouldn't be changing - * colorspace anyway. - * The Cr=>R and Cb=>B values can be rounded to integers in advance; the - * values for the G calculation are left scaled up, since we must add them - * together before rounding. - */ - -#define SCALEBITS 16 /* speediest right-shift on some machines */ -#define ONE_HALF ((INT32) 1 << (SCALEBITS-1)) -#define FIX(x) ((INT32) ((x) * (1L<<SCALEBITS) + 0.5)) - - -/* - * Initialize tables for YCC->RGB colorspace conversion. - */ - -LOCAL(void) -build_ycc_rgb_table (j_decompress_ptr cinfo) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - int i; - INT32 x; - SHIFT_TEMPS - - cconvert->Cr_r_tab = (int *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(int)); - cconvert->Cb_b_tab = (int *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(int)); - cconvert->Cr_g_tab = (INT32 *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(INT32)); - cconvert->Cb_g_tab = (INT32 *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(INT32)); - - for (i = 0, x = -CENTERJSAMPLE; i <= MAXJSAMPLE; i++, x++) { - /* i is the actual input pixel value, in the range 0..MAXJSAMPLE */ - /* The Cb or Cr value we are thinking of is x = i - CENTERJSAMPLE */ - /* Cr=>R value is nearest int to 1.40200 * x */ - cconvert->Cr_r_tab[i] = (int) - RIGHT_SHIFT(FIX(1.40200) * x + ONE_HALF, SCALEBITS); - /* Cb=>B value is nearest int to 1.77200 * x */ - cconvert->Cb_b_tab[i] = (int) - RIGHT_SHIFT(FIX(1.77200) * x + ONE_HALF, SCALEBITS); - /* Cr=>G value is scaled-up -0.71414 * x */ - cconvert->Cr_g_tab[i] = (- FIX(0.71414)) * x; - /* Cb=>G value is scaled-up -0.34414 * x */ - /* We also add in ONE_HALF so that need not do it in inner loop */ - cconvert->Cb_g_tab[i] = (- FIX(0.34414)) * x + ONE_HALF; - } -} - - -/* - * Convert some rows of samples to the output colorspace. - * - * Note that we change from noninterleaved, one-plane-per-component format - * to interleaved-pixel format. The output buffer is therefore three times - * as wide as the input buffer. - * A starting row offset is provided only for the input buffer. The caller - * can easily adjust the passed output_buf value to accommodate any row - * offset required on that side. - */ - -METHODDEF(void) -ycc_rgb_convert (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION input_row, - JSAMPARRAY output_buf, int num_rows) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - register int y, cb, cr; - register JSAMPROW outptr; - register JSAMPROW inptr0, inptr1, inptr2; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->output_width; - /* copy these pointers into registers if possible */ - register JSAMPLE * range_limit = cinfo->sample_range_limit; - register int * Crrtab = cconvert->Cr_r_tab; - register int * Cbbtab = cconvert->Cb_b_tab; - register INT32 * Crgtab = cconvert->Cr_g_tab; - register INT32 * Cbgtab = cconvert->Cb_g_tab; - SHIFT_TEMPS - - while (--num_rows >= 0) { - inptr0 = input_buf[0][input_row]; - inptr1 = input_buf[1][input_row]; - inptr2 = input_buf[2][input_row]; - input_row++; - outptr = *output_buf++; - for (col = 0; col < num_cols; col++) { - y = GETJSAMPLE(inptr0[col]); - cb = GETJSAMPLE(inptr1[col]); - cr = GETJSAMPLE(inptr2[col]); - /* Range-limiting is essential due to noise introduced by DCT losses. */ - outptr[RGB_RED] = range_limit[y + Crrtab[cr]]; - outptr[RGB_GREEN] = range_limit[y + - ((int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], - SCALEBITS))]; - outptr[RGB_BLUE] = range_limit[y + Cbbtab[cb]]; - outptr += RGB_PIXELSIZE; - } - } -} - - -/**************** Cases other than YCbCr -> RGB **************/ - - -/* - * Color conversion for no colorspace change: just copy the data, - * converting from separate-planes to interleaved representation. - */ - -METHODDEF(void) -null_convert (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION input_row, - JSAMPARRAY output_buf, int num_rows) -{ - register JSAMPROW inptr, outptr; - register JDIMENSION count; - register int num_components = cinfo->num_components; - JDIMENSION num_cols = cinfo->output_width; - int ci; - - while (--num_rows >= 0) { - for (ci = 0; ci < num_components; ci++) { - inptr = input_buf[ci][input_row]; - outptr = output_buf[0] + ci; - for (count = num_cols; count > 0; count--) { - *outptr = *inptr++; /* needn't bother with GETJSAMPLE() here */ - outptr += num_components; - } - } - input_row++; - output_buf++; - } -} - - -/* - * Color conversion for grayscale: just copy the data. - * This also works for YCbCr -> grayscale conversion, in which - * we just copy the Y (luminance) component and ignore chrominance. - */ - -METHODDEF(void) -grayscale_convert (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION input_row, - JSAMPARRAY output_buf, int num_rows) -{ - jcopy_sample_rows(input_buf[0], (int) input_row, output_buf, 0, - num_rows, cinfo->output_width); -} - - -/* - * Convert grayscale to RGB: just duplicate the graylevel three times. - * This is provided to support applications that don't want to cope - * with grayscale as a separate case. - */ - -METHODDEF(void) -gray_rgb_convert (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION input_row, - JSAMPARRAY output_buf, int num_rows) -{ - register JSAMPROW inptr, outptr; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->output_width; - - while (--num_rows >= 0) { - inptr = input_buf[0][input_row++]; - outptr = *output_buf++; - for (col = 0; col < num_cols; col++) { - /* We can dispense with GETJSAMPLE() here */ - outptr[RGB_RED] = outptr[RGB_GREEN] = outptr[RGB_BLUE] = inptr[col]; - outptr += RGB_PIXELSIZE; - } - } -} - - -/* - * Adobe-style YCCK->CMYK conversion. - * We convert YCbCr to R=1-C, G=1-M, and B=1-Y using the same - * conversion as above, while passing K (black) unchanged. - * We assume build_ycc_rgb_table has been called. - */ - -METHODDEF(void) -ycck_cmyk_convert (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION input_row, - JSAMPARRAY output_buf, int num_rows) -{ - my_cconvert_ptr cconvert = (my_cconvert_ptr) cinfo->cconvert; - register int y, cb, cr; - register JSAMPROW outptr; - register JSAMPROW inptr0, inptr1, inptr2, inptr3; - register JDIMENSION col; - JDIMENSION num_cols = cinfo->output_width; - /* copy these pointers into registers if possible */ - register JSAMPLE * range_limit = cinfo->sample_range_limit; - register int * Crrtab = cconvert->Cr_r_tab; - register int * Cbbtab = cconvert->Cb_b_tab; - register INT32 * Crgtab = cconvert->Cr_g_tab; - register INT32 * Cbgtab = cconvert->Cb_g_tab; - SHIFT_TEMPS - - while (--num_rows >= 0) { - inptr0 = input_buf[0][input_row]; - inptr1 = input_buf[1][input_row]; - inptr2 = input_buf[2][input_row]; - inptr3 = input_buf[3][input_row]; - input_row++; - outptr = *output_buf++; - for (col = 0; col < num_cols; col++) { - y = GETJSAMPLE(inptr0[col]); - cb = GETJSAMPLE(inptr1[col]); - cr = GETJSAMPLE(inptr2[col]); - /* Range-limiting is essential due to noise introduced by DCT losses. */ - outptr[0] = range_limit[MAXJSAMPLE - (y + Crrtab[cr])]; /* red */ - outptr[1] = range_limit[MAXJSAMPLE - (y + /* green */ - ((int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], - SCALEBITS)))]; - outptr[2] = range_limit[MAXJSAMPLE - (y + Cbbtab[cb])]; /* blue */ - /* K passes through unchanged */ - outptr[3] = inptr3[col]; /* don't need GETJSAMPLE here */ - outptr += 4; - } - } -} - - -/* - * Empty method for start_pass. - */ - -METHODDEF(void) -start_pass_dcolor (j_decompress_ptr cinfo) -{ - /* no work needed */ -} - - -/* - * Module initialization routine for output colorspace conversion. - */ - -GLOBAL(void) -jinit_color_deconverter (j_decompress_ptr cinfo) -{ - my_cconvert_ptr cconvert; - int ci; - - cconvert = (my_cconvert_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_color_deconverter)); - cinfo->cconvert = (struct jpeg_color_deconverter *) cconvert; - cconvert->pub.start_pass = start_pass_dcolor; - - /* Make sure num_components agrees with jpeg_color_space */ - switch (cinfo->jpeg_color_space) { - case JCS_GRAYSCALE: - if (cinfo->num_components != 1) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - break; - - case JCS_RGB: - case JCS_YCbCr: - if (cinfo->num_components != 3) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - break; - - case JCS_CMYK: - case JCS_YCCK: - if (cinfo->num_components != 4) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - break; - - default: /* JCS_UNKNOWN can be anything */ - if (cinfo->num_components < 1) - ERREXIT(cinfo, JERR_BAD_J_COLORSPACE); - break; - } - - /* Set out_color_components and conversion method based on requested space. - * Also clear the component_needed flags for any unused components, - * so that earlier pipeline stages can avoid useless computation. - */ - - switch (cinfo->out_color_space) { - case JCS_GRAYSCALE: - cinfo->out_color_components = 1; - if (cinfo->jpeg_color_space == JCS_GRAYSCALE || - cinfo->jpeg_color_space == JCS_YCbCr) { - cconvert->pub.color_convert = grayscale_convert; - /* For color->grayscale conversion, only the Y (0) component is needed */ - for (ci = 1; ci < cinfo->num_components; ci++) - cinfo->comp_info[ci].component_needed = FALSE; - } else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - case JCS_RGB: - cinfo->out_color_components = RGB_PIXELSIZE; - if (cinfo->jpeg_color_space == JCS_YCbCr) { - cconvert->pub.color_convert = ycc_rgb_convert; - build_ycc_rgb_table(cinfo); - } else if (cinfo->jpeg_color_space == JCS_GRAYSCALE) { - cconvert->pub.color_convert = gray_rgb_convert; - } else if (cinfo->jpeg_color_space == JCS_RGB && RGB_PIXELSIZE == 3) { - cconvert->pub.color_convert = null_convert; - } else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - case JCS_CMYK: - cinfo->out_color_components = 4; - if (cinfo->jpeg_color_space == JCS_YCCK) { - cconvert->pub.color_convert = ycck_cmyk_convert; - build_ycc_rgb_table(cinfo); - } else if (cinfo->jpeg_color_space == JCS_CMYK) { - cconvert->pub.color_convert = null_convert; - } else - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - - default: - /* Permit null conversion to same output space */ - if (cinfo->out_color_space == cinfo->jpeg_color_space) { - cinfo->out_color_components = cinfo->num_components; - cconvert->pub.color_convert = null_convert; - } else /* unsupported non-null conversion */ - ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL); - break; - } - - if (cinfo->quantize_colors) - cinfo->output_components = 1; /* single colormapped output component */ - else - cinfo->output_components = cinfo->out_color_components; -} diff --git a/src/3rdparty/libjpeg/jdct.h b/src/3rdparty/libjpeg/jdct.h deleted file mode 100644 index 360dec80c9..0000000000 --- a/src/3rdparty/libjpeg/jdct.h +++ /dev/null @@ -1,393 +0,0 @@ -/* - * jdct.h - * - * Copyright (C) 1994-1996, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This include file contains common declarations for the forward and - * inverse DCT modules. These declarations are private to the DCT managers - * (jcdctmgr.c, jddctmgr.c) and the individual DCT algorithms. - * The individual DCT algorithms are kept in separate files to ease - * machine-dependent tuning (e.g., assembly coding). - */ - - -/* - * A forward DCT routine is given a pointer to an input sample array and - * a pointer to a work area of type DCTELEM[]; the DCT is to be performed - * in-place in that buffer. Type DCTELEM is int for 8-bit samples, INT32 - * for 12-bit samples. (NOTE: Floating-point DCT implementations use an - * array of type FAST_FLOAT, instead.) - * The input data is to be fetched from the sample array starting at a - * specified column. (Any row offset needed will be applied to the array - * pointer before it is passed to the FDCT code.) - * Note that the number of samples fetched by the FDCT routine is - * DCT_h_scaled_size * DCT_v_scaled_size. - * The DCT outputs are returned scaled up by a factor of 8; they therefore - * have a range of +-8K for 8-bit data, +-128K for 12-bit data. This - * convention improves accuracy in integer implementations and saves some - * work in floating-point ones. - * Quantization of the output coefficients is done by jcdctmgr.c. - */ - -#if BITS_IN_JSAMPLE == 8 -typedef int DCTELEM; /* 16 or 32 bits is fine */ -#else -typedef INT32 DCTELEM; /* must have 32 bits */ -#endif - -typedef JMETHOD(void, forward_DCT_method_ptr, (DCTELEM * data, - JSAMPARRAY sample_data, - JDIMENSION start_col)); -typedef JMETHOD(void, float_DCT_method_ptr, (FAST_FLOAT * data, - JSAMPARRAY sample_data, - JDIMENSION start_col)); - - -/* - * An inverse DCT routine is given a pointer to the input JBLOCK and a pointer - * to an output sample array. The routine must dequantize the input data as - * well as perform the IDCT; for dequantization, it uses the multiplier table - * pointed to by compptr->dct_table. The output data is to be placed into the - * sample array starting at a specified column. (Any row offset needed will - * be applied to the array pointer before it is passed to the IDCT code.) - * Note that the number of samples emitted by the IDCT routine is - * DCT_h_scaled_size * DCT_v_scaled_size. - */ - -/* typedef inverse_DCT_method_ptr is declared in jpegint.h */ - -/* - * Each IDCT routine has its own ideas about the best dct_table element type. - */ - -typedef MULTIPLIER ISLOW_MULT_TYPE; /* short or int, whichever is faster */ -#if BITS_IN_JSAMPLE == 8 -typedef MULTIPLIER IFAST_MULT_TYPE; /* 16 bits is OK, use short if faster */ -#define IFAST_SCALE_BITS 2 /* fractional bits in scale factors */ -#else -typedef INT32 IFAST_MULT_TYPE; /* need 32 bits for scaled quantizers */ -#define IFAST_SCALE_BITS 13 /* fractional bits in scale factors */ -#endif -typedef FAST_FLOAT FLOAT_MULT_TYPE; /* preferred floating type */ - - -/* - * Each IDCT routine is responsible for range-limiting its results and - * converting them to unsigned form (0..MAXJSAMPLE). The raw outputs could - * be quite far out of range if the input data is corrupt, so a bulletproof - * range-limiting step is required. We use a mask-and-table-lookup method - * to do the combined operations quickly. See the comments with - * prepare_range_limit_table (in jdmaster.c) for more info. - */ - -#define IDCT_range_limit(cinfo) ((cinfo)->sample_range_limit + CENTERJSAMPLE) - -#define RANGE_MASK (MAXJSAMPLE * 4 + 3) /* 2 bits wider than legal samples */ - - -/* Short forms of external names for systems with brain-damaged linkers. */ - -#ifdef NEED_SHORT_EXTERNAL_NAMES -#define jpeg_fdct_islow jFDislow -#define jpeg_fdct_ifast jFDifast -#define jpeg_fdct_float jFDfloat -#define jpeg_fdct_7x7 jFD7x7 -#define jpeg_fdct_6x6 jFD6x6 -#define jpeg_fdct_5x5 jFD5x5 -#define jpeg_fdct_4x4 jFD4x4 -#define jpeg_fdct_3x3 jFD3x3 -#define jpeg_fdct_2x2 jFD2x2 -#define jpeg_fdct_1x1 jFD1x1 -#define jpeg_fdct_9x9 jFD9x9 -#define jpeg_fdct_10x10 jFD10x10 -#define jpeg_fdct_11x11 jFD11x11 -#define jpeg_fdct_12x12 jFD12x12 -#define jpeg_fdct_13x13 jFD13x13 -#define jpeg_fdct_14x14 jFD14x14 -#define jpeg_fdct_15x15 jFD15x15 -#define jpeg_fdct_16x16 jFD16x16 -#define jpeg_fdct_16x8 jFD16x8 -#define jpeg_fdct_14x7 jFD14x7 -#define jpeg_fdct_12x6 jFD12x6 -#define jpeg_fdct_10x5 jFD10x5 -#define jpeg_fdct_8x4 jFD8x4 -#define jpeg_fdct_6x3 jFD6x3 -#define jpeg_fdct_4x2 jFD4x2 -#define jpeg_fdct_2x1 jFD2x1 -#define jpeg_fdct_8x16 jFD8x16 -#define jpeg_fdct_7x14 jFD7x14 -#define jpeg_fdct_6x12 jFD6x12 -#define jpeg_fdct_5x10 jFD5x10 -#define jpeg_fdct_4x8 jFD4x8 -#define jpeg_fdct_3x6 jFD3x6 -#define jpeg_fdct_2x4 jFD2x4 -#define jpeg_fdct_1x2 jFD1x2 -#define jpeg_idct_islow jRDislow -#define jpeg_idct_ifast jRDifast -#define jpeg_idct_float jRDfloat -#define jpeg_idct_7x7 jRD7x7 -#define jpeg_idct_6x6 jRD6x6 -#define jpeg_idct_5x5 jRD5x5 -#define jpeg_idct_4x4 jRD4x4 -#define jpeg_idct_3x3 jRD3x3 -#define jpeg_idct_2x2 jRD2x2 -#define jpeg_idct_1x1 jRD1x1 -#define jpeg_idct_9x9 jRD9x9 -#define jpeg_idct_10x10 jRD10x10 -#define jpeg_idct_11x11 jRD11x11 -#define jpeg_idct_12x12 jRD12x12 -#define jpeg_idct_13x13 jRD13x13 -#define jpeg_idct_14x14 jRD14x14 -#define jpeg_idct_15x15 jRD15x15 -#define jpeg_idct_16x16 jRD16x16 -#define jpeg_idct_16x8 jRD16x8 -#define jpeg_idct_14x7 jRD14x7 -#define jpeg_idct_12x6 jRD12x6 -#define jpeg_idct_10x5 jRD10x5 -#define jpeg_idct_8x4 jRD8x4 -#define jpeg_idct_6x3 jRD6x3 -#define jpeg_idct_4x2 jRD4x2 -#define jpeg_idct_2x1 jRD2x1 -#define jpeg_idct_8x16 jRD8x16 -#define jpeg_idct_7x14 jRD7x14 -#define jpeg_idct_6x12 jRD6x12 -#define jpeg_idct_5x10 jRD5x10 -#define jpeg_idct_4x8 jRD4x8 -#define jpeg_idct_3x6 jRD3x8 -#define jpeg_idct_2x4 jRD2x4 -#define jpeg_idct_1x2 jRD1x2 -#endif /* NEED_SHORT_EXTERNAL_NAMES */ - -/* Extern declarations for the forward and inverse DCT routines. */ - -EXTERN(void) jpeg_fdct_islow - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_ifast - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_float - JPP((FAST_FLOAT * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_7x7 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_6x6 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_5x5 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_4x4 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_3x3 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_2x2 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_1x1 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_9x9 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_10x10 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_11x11 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_12x12 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_13x13 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_14x14 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_15x15 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_16x16 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_16x8 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_14x7 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_12x6 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_10x5 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_8x4 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_6x3 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_4x2 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_2x1 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_8x16 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_7x14 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_6x12 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_5x10 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_4x8 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_3x6 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_2x4 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); -EXTERN(void) jpeg_fdct_1x2 - JPP((DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col)); - -EXTERN(void) jpeg_idct_islow - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_ifast - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_float - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_7x7 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_6x6 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_5x5 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_4x4 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_3x3 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_2x2 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_1x1 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_9x9 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_10x10 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_11x11 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_12x12 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_13x13 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_14x14 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_15x15 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_16x16 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_16x8 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_14x7 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_12x6 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_10x5 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_8x4 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_6x3 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_4x2 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_2x1 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_8x16 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_7x14 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_6x12 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_5x10 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_4x8 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_3x6 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_2x4 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); -EXTERN(void) jpeg_idct_1x2 - JPP((j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, JSAMPARRAY output_buf, JDIMENSION output_col)); - - -/* - * Macros for handling fixed-point arithmetic; these are used by many - * but not all of the DCT/IDCT modules. - * - * All values are expected to be of type INT32. - * Fractional constants are scaled left by CONST_BITS bits. - * CONST_BITS is defined within each module using these macros, - * and may differ from one module to the next. - */ - -#define ONE ((INT32) 1) -#define CONST_SCALE (ONE << CONST_BITS) - -/* Convert a positive real constant to an integer scaled by CONST_SCALE. - * Caution: some C compilers fail to reduce "FIX(constant)" at compile time, - * thus causing a lot of useless floating-point operations at run time. - */ - -#define FIX(x) ((INT32) ((x) * CONST_SCALE + 0.5)) - -/* Descale and correctly round an INT32 value that's scaled by N bits. - * We assume RIGHT_SHIFT rounds towards minus infinity, so adding - * the fudge factor is correct for either sign of X. - */ - -#define DESCALE(x,n) RIGHT_SHIFT((x) + (ONE << ((n)-1)), n) - -/* Multiply an INT32 variable by an INT32 constant to yield an INT32 result. - * This macro is used only when the two inputs will actually be no more than - * 16 bits wide, so that a 16x16->32 bit multiply can be used instead of a - * full 32x32 multiply. This provides a useful speedup on many machines. - * Unfortunately there is no way to specify a 16x16->32 multiply portably - * in C, but some C compilers will do the right thing if you provide the - * correct combination of casts. - */ - -#ifdef SHORTxSHORT_32 /* may work if 'int' is 32 bits */ -#define MULTIPLY16C16(var,const) (((INT16) (var)) * ((INT16) (const))) -#endif -#ifdef SHORTxLCONST_32 /* known to work with Microsoft C 6.0 */ -#define MULTIPLY16C16(var,const) (((INT16) (var)) * ((INT32) (const))) -#endif - -#ifndef MULTIPLY16C16 /* default definition */ -#define MULTIPLY16C16(var,const) ((var) * (const)) -#endif - -/* Same except both inputs are variables. */ - -#ifdef SHORTxSHORT_32 /* may work if 'int' is 32 bits */ -#define MULTIPLY16V16(var1,var2) (((INT16) (var1)) * ((INT16) (var2))) -#endif - -#ifndef MULTIPLY16V16 /* default definition */ -#define MULTIPLY16V16(var1,var2) ((var1) * (var2)) -#endif diff --git a/src/3rdparty/libjpeg/jddctmgr.c b/src/3rdparty/libjpeg/jddctmgr.c deleted file mode 100644 index 0ded9d5741..0000000000 --- a/src/3rdparty/libjpeg/jddctmgr.c +++ /dev/null @@ -1,384 +0,0 @@ -/* - * jddctmgr.c - * - * Copyright (C) 1994-1996, Thomas G. Lane. - * Modified 2002-2010 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains the inverse-DCT management logic. - * This code selects a particular IDCT implementation to be used, - * and it performs related housekeeping chores. No code in this file - * is executed per IDCT step, only during output pass setup. - * - * Note that the IDCT routines are responsible for performing coefficient - * dequantization as well as the IDCT proper. This module sets up the - * dequantization multiplier table needed by the IDCT routine. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" -#include "jdct.h" /* Private declarations for DCT subsystem */ - - -/* - * The decompressor input side (jdinput.c) saves away the appropriate - * quantization table for each component at the start of the first scan - * involving that component. (This is necessary in order to correctly - * decode files that reuse Q-table slots.) - * When we are ready to make an output pass, the saved Q-table is converted - * to a multiplier table that will actually be used by the IDCT routine. - * The multiplier table contents are IDCT-method-dependent. To support - * application changes in IDCT method between scans, we can remake the - * multiplier tables if necessary. - * In buffered-image mode, the first output pass may occur before any data - * has been seen for some components, and thus before their Q-tables have - * been saved away. To handle this case, multiplier tables are preset - * to zeroes; the result of the IDCT will be a neutral gray level. - */ - - -/* Private subobject for this module */ - -typedef struct { - struct jpeg_inverse_dct pub; /* public fields */ - - /* This array contains the IDCT method code that each multiplier table - * is currently set up for, or -1 if it's not yet set up. - * The actual multiplier tables are pointed to by dct_table in the - * per-component comp_info structures. - */ - int cur_method[MAX_COMPONENTS]; -} my_idct_controller; - -typedef my_idct_controller * my_idct_ptr; - - -/* Allocated multiplier tables: big enough for any supported variant */ - -typedef union { - ISLOW_MULT_TYPE islow_array[DCTSIZE2]; -#ifdef DCT_IFAST_SUPPORTED - IFAST_MULT_TYPE ifast_array[DCTSIZE2]; -#endif -#ifdef DCT_FLOAT_SUPPORTED - FLOAT_MULT_TYPE float_array[DCTSIZE2]; -#endif -} multiplier_table; - - -/* The current scaled-IDCT routines require ISLOW-style multiplier tables, - * so be sure to compile that code if either ISLOW or SCALING is requested. - */ -#ifdef DCT_ISLOW_SUPPORTED -#define PROVIDE_ISLOW_TABLES -#else -#ifdef IDCT_SCALING_SUPPORTED -#define PROVIDE_ISLOW_TABLES -#endif -#endif - - -/* - * Prepare for an output pass. - * Here we select the proper IDCT routine for each component and build - * a matching multiplier table. - */ - -METHODDEF(void) -start_pass (j_decompress_ptr cinfo) -{ - my_idct_ptr idct = (my_idct_ptr) cinfo->idct; - int ci, i; - jpeg_component_info *compptr; - int method = 0; - inverse_DCT_method_ptr method_ptr = NULL; - JQUANT_TBL * qtbl; - - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - /* Select the proper IDCT routine for this component's scaling */ - switch ((compptr->DCT_h_scaled_size << 8) + compptr->DCT_v_scaled_size) { -#ifdef IDCT_SCALING_SUPPORTED - case ((1 << 8) + 1): - method_ptr = jpeg_idct_1x1; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((2 << 8) + 2): - method_ptr = jpeg_idct_2x2; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((3 << 8) + 3): - method_ptr = jpeg_idct_3x3; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((4 << 8) + 4): - method_ptr = jpeg_idct_4x4; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((5 << 8) + 5): - method_ptr = jpeg_idct_5x5; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((6 << 8) + 6): - method_ptr = jpeg_idct_6x6; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((7 << 8) + 7): - method_ptr = jpeg_idct_7x7; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((9 << 8) + 9): - method_ptr = jpeg_idct_9x9; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((10 << 8) + 10): - method_ptr = jpeg_idct_10x10; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((11 << 8) + 11): - method_ptr = jpeg_idct_11x11; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((12 << 8) + 12): - method_ptr = jpeg_idct_12x12; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((13 << 8) + 13): - method_ptr = jpeg_idct_13x13; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((14 << 8) + 14): - method_ptr = jpeg_idct_14x14; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((15 << 8) + 15): - method_ptr = jpeg_idct_15x15; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((16 << 8) + 16): - method_ptr = jpeg_idct_16x16; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((16 << 8) + 8): - method_ptr = jpeg_idct_16x8; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((14 << 8) + 7): - method_ptr = jpeg_idct_14x7; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((12 << 8) + 6): - method_ptr = jpeg_idct_12x6; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((10 << 8) + 5): - method_ptr = jpeg_idct_10x5; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((8 << 8) + 4): - method_ptr = jpeg_idct_8x4; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((6 << 8) + 3): - method_ptr = jpeg_idct_6x3; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((4 << 8) + 2): - method_ptr = jpeg_idct_4x2; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((2 << 8) + 1): - method_ptr = jpeg_idct_2x1; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((8 << 8) + 16): - method_ptr = jpeg_idct_8x16; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((7 << 8) + 14): - method_ptr = jpeg_idct_7x14; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((6 << 8) + 12): - method_ptr = jpeg_idct_6x12; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((5 << 8) + 10): - method_ptr = jpeg_idct_5x10; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((4 << 8) + 8): - method_ptr = jpeg_idct_4x8; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((3 << 8) + 6): - method_ptr = jpeg_idct_3x6; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((2 << 8) + 4): - method_ptr = jpeg_idct_2x4; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; - case ((1 << 8) + 2): - method_ptr = jpeg_idct_1x2; - method = JDCT_ISLOW; /* jidctint uses islow-style table */ - break; -#endif - case ((DCTSIZE << 8) + DCTSIZE): - switch (cinfo->dct_method) { -#ifdef DCT_ISLOW_SUPPORTED - case JDCT_ISLOW: - method_ptr = jpeg_idct_islow; - method = JDCT_ISLOW; - break; -#endif -#ifdef DCT_IFAST_SUPPORTED - case JDCT_IFAST: - method_ptr = jpeg_idct_ifast; - method = JDCT_IFAST; - break; -#endif -#ifdef DCT_FLOAT_SUPPORTED - case JDCT_FLOAT: - method_ptr = jpeg_idct_float; - method = JDCT_FLOAT; - break; -#endif - default: - ERREXIT(cinfo, JERR_NOT_COMPILED); - break; - } - break; - default: - ERREXIT2(cinfo, JERR_BAD_DCTSIZE, - compptr->DCT_h_scaled_size, compptr->DCT_v_scaled_size); - break; - } - idct->pub.inverse_DCT[ci] = method_ptr; - /* Create multiplier table from quant table. - * However, we can skip this if the component is uninteresting - * or if we already built the table. Also, if no quant table - * has yet been saved for the component, we leave the - * multiplier table all-zero; we'll be reading zeroes from the - * coefficient controller's buffer anyway. - */ - if (! compptr->component_needed || idct->cur_method[ci] == method) - continue; - qtbl = compptr->quant_table; - if (qtbl == NULL) /* happens if no data yet for component */ - continue; - idct->cur_method[ci] = method; - switch (method) { -#ifdef PROVIDE_ISLOW_TABLES - case JDCT_ISLOW: - { - /* For LL&M IDCT method, multipliers are equal to raw quantization - * coefficients, but are stored as ints to ensure access efficiency. - */ - ISLOW_MULT_TYPE * ismtbl = (ISLOW_MULT_TYPE *) compptr->dct_table; - for (i = 0; i < DCTSIZE2; i++) { - ismtbl[i] = (ISLOW_MULT_TYPE) qtbl->quantval[i]; - } - } - break; -#endif -#ifdef DCT_IFAST_SUPPORTED - case JDCT_IFAST: - { - /* For AA&N IDCT method, multipliers are equal to quantization - * coefficients scaled by scalefactor[row]*scalefactor[col], where - * scalefactor[0] = 1 - * scalefactor[k] = cos(k*PI/16) * sqrt(2) for k=1..7 - * For integer operation, the multiplier table is to be scaled by - * IFAST_SCALE_BITS. - */ - IFAST_MULT_TYPE * ifmtbl = (IFAST_MULT_TYPE *) compptr->dct_table; -#define CONST_BITS 14 - static const INT16 aanscales[DCTSIZE2] = { - /* precomputed values scaled up by 14 bits */ - 16384, 22725, 21407, 19266, 16384, 12873, 8867, 4520, - 22725, 31521, 29692, 26722, 22725, 17855, 12299, 6270, - 21407, 29692, 27969, 25172, 21407, 16819, 11585, 5906, - 19266, 26722, 25172, 22654, 19266, 15137, 10426, 5315, - 16384, 22725, 21407, 19266, 16384, 12873, 8867, 4520, - 12873, 17855, 16819, 15137, 12873, 10114, 6967, 3552, - 8867, 12299, 11585, 10426, 8867, 6967, 4799, 2446, - 4520, 6270, 5906, 5315, 4520, 3552, 2446, 1247 - }; - SHIFT_TEMPS - - for (i = 0; i < DCTSIZE2; i++) { - ifmtbl[i] = (IFAST_MULT_TYPE) - DESCALE(MULTIPLY16V16((INT32) qtbl->quantval[i], - (INT32) aanscales[i]), - CONST_BITS-IFAST_SCALE_BITS); - } - } - break; -#endif -#ifdef DCT_FLOAT_SUPPORTED - case JDCT_FLOAT: - { - /* For float AA&N IDCT method, multipliers are equal to quantization - * coefficients scaled by scalefactor[row]*scalefactor[col], where - * scalefactor[0] = 1 - * scalefactor[k] = cos(k*PI/16) * sqrt(2) for k=1..7 - * We apply a further scale factor of 1/8. - */ - FLOAT_MULT_TYPE * fmtbl = (FLOAT_MULT_TYPE *) compptr->dct_table; - int row, col; - static const double aanscalefactor[DCTSIZE] = { - 1.0, 1.387039845, 1.306562965, 1.175875602, - 1.0, 0.785694958, 0.541196100, 0.275899379 - }; - - i = 0; - for (row = 0; row < DCTSIZE; row++) { - for (col = 0; col < DCTSIZE; col++) { - fmtbl[i] = (FLOAT_MULT_TYPE) - ((double) qtbl->quantval[i] * - aanscalefactor[row] * aanscalefactor[col] * 0.125); - i++; - } - } - } - break; -#endif - default: - ERREXIT(cinfo, JERR_NOT_COMPILED); - break; - } - } -} - - -/* - * Initialize IDCT manager. - */ - -GLOBAL(void) -jinit_inverse_dct (j_decompress_ptr cinfo) -{ - my_idct_ptr idct; - int ci; - jpeg_component_info *compptr; - - idct = (my_idct_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_idct_controller)); - cinfo->idct = (struct jpeg_inverse_dct *) idct; - idct->pub.start_pass = start_pass; - - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - /* Allocate and pre-zero a multiplier table for each component */ - compptr->dct_table = - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(multiplier_table)); - MEMZERO(compptr->dct_table, SIZEOF(multiplier_table)); - /* Mark multiplier table not yet set up for any method */ - idct->cur_method[ci] = -1; - } -} diff --git a/src/3rdparty/libjpeg/jdhuff.c b/src/3rdparty/libjpeg/jdhuff.c deleted file mode 100644 index 06f92fe47f..0000000000 --- a/src/3rdparty/libjpeg/jdhuff.c +++ /dev/null @@ -1,1541 +0,0 @@ -/* - * jdhuff.c - * - * Copyright (C) 1991-1997, Thomas G. Lane. - * Modified 2006-2009 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains Huffman entropy decoding routines. - * Both sequential and progressive modes are supported in this single module. - * - * Much of the complexity here has to do with supporting input suspension. - * If the data source module demands suspension, we want to be able to back - * up to the start of the current MCU. To do this, we copy state variables - * into local working storage, and update them back to the permanent - * storage only upon successful completion of an MCU. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Derived data constructed for each Huffman table */ - -#define HUFF_LOOKAHEAD 8 /* # of bits of lookahead */ - -typedef struct { - /* Basic tables: (element [0] of each array is unused) */ - INT32 maxcode[18]; /* largest code of length k (-1 if none) */ - /* (maxcode[17] is a sentinel to ensure jpeg_huff_decode terminates) */ - INT32 valoffset[17]; /* huffval[] offset for codes of length k */ - /* valoffset[k] = huffval[] index of 1st symbol of code length k, less - * the smallest code of length k; so given a code of length k, the - * corresponding symbol is huffval[code + valoffset[k]] - */ - - /* Link to public Huffman table (needed only in jpeg_huff_decode) */ - JHUFF_TBL *pub; - - /* Lookahead tables: indexed by the next HUFF_LOOKAHEAD bits of - * the input data stream. If the next Huffman code is no more - * than HUFF_LOOKAHEAD bits long, we can obtain its length and - * the corresponding symbol directly from these tables. - */ - int look_nbits[1<<HUFF_LOOKAHEAD]; /* # bits, or 0 if too long */ - UINT8 look_sym[1<<HUFF_LOOKAHEAD]; /* symbol, or unused */ -} d_derived_tbl; - - -/* - * Fetching the next N bits from the input stream is a time-critical operation - * for the Huffman decoders. We implement it with a combination of inline - * macros and out-of-line subroutines. Note that N (the number of bits - * demanded at one time) never exceeds 15 for JPEG use. - * - * We read source bytes into get_buffer and dole out bits as needed. - * If get_buffer already contains enough bits, they are fetched in-line - * by the macros CHECK_BIT_BUFFER and GET_BITS. When there aren't enough - * bits, jpeg_fill_bit_buffer is called; it will attempt to fill get_buffer - * as full as possible (not just to the number of bits needed; this - * prefetching reduces the overhead cost of calling jpeg_fill_bit_buffer). - * Note that jpeg_fill_bit_buffer may return FALSE to indicate suspension. - * On TRUE return, jpeg_fill_bit_buffer guarantees that get_buffer contains - * at least the requested number of bits --- dummy zeroes are inserted if - * necessary. - */ - -typedef INT32 bit_buf_type; /* type of bit-extraction buffer */ -#define BIT_BUF_SIZE 32 /* size of buffer in bits */ - -/* If long is > 32 bits on your machine, and shifting/masking longs is - * reasonably fast, making bit_buf_type be long and setting BIT_BUF_SIZE - * appropriately should be a win. Unfortunately we can't define the size - * with something like #define BIT_BUF_SIZE (sizeof(bit_buf_type)*8) - * because not all machines measure sizeof in 8-bit bytes. - */ - -typedef struct { /* Bitreading state saved across MCUs */ - bit_buf_type get_buffer; /* current bit-extraction buffer */ - int bits_left; /* # of unused bits in it */ -} bitread_perm_state; - -typedef struct { /* Bitreading working state within an MCU */ - /* Current data source location */ - /* We need a copy, rather than munging the original, in case of suspension */ - const JOCTET * next_input_byte; /* => next byte to read from source */ - size_t bytes_in_buffer; /* # of bytes remaining in source buffer */ - /* Bit input buffer --- note these values are kept in register variables, - * not in this struct, inside the inner loops. - */ - bit_buf_type get_buffer; /* current bit-extraction buffer */ - int bits_left; /* # of unused bits in it */ - /* Pointer needed by jpeg_fill_bit_buffer. */ - j_decompress_ptr cinfo; /* back link to decompress master record */ -} bitread_working_state; - -/* Macros to declare and load/save bitread local variables. */ -#define BITREAD_STATE_VARS \ - register bit_buf_type get_buffer; \ - register int bits_left; \ - bitread_working_state br_state - -#define BITREAD_LOAD_STATE(cinfop,permstate) \ - br_state.cinfo = cinfop; \ - br_state.next_input_byte = cinfop->src->next_input_byte; \ - br_state.bytes_in_buffer = cinfop->src->bytes_in_buffer; \ - get_buffer = permstate.get_buffer; \ - bits_left = permstate.bits_left; - -#define BITREAD_SAVE_STATE(cinfop,permstate) \ - cinfop->src->next_input_byte = br_state.next_input_byte; \ - cinfop->src->bytes_in_buffer = br_state.bytes_in_buffer; \ - permstate.get_buffer = get_buffer; \ - permstate.bits_left = bits_left - -/* - * These macros provide the in-line portion of bit fetching. - * Use CHECK_BIT_BUFFER to ensure there are N bits in get_buffer - * before using GET_BITS, PEEK_BITS, or DROP_BITS. - * The variables get_buffer and bits_left are assumed to be locals, - * but the state struct might not be (jpeg_huff_decode needs this). - * CHECK_BIT_BUFFER(state,n,action); - * Ensure there are N bits in get_buffer; if suspend, take action. - * val = GET_BITS(n); - * Fetch next N bits. - * val = PEEK_BITS(n); - * Fetch next N bits without removing them from the buffer. - * DROP_BITS(n); - * Discard next N bits. - * The value N should be a simple variable, not an expression, because it - * is evaluated multiple times. - */ - -#define CHECK_BIT_BUFFER(state,nbits,action) \ - { if (bits_left < (nbits)) { \ - if (! jpeg_fill_bit_buffer(&(state),get_buffer,bits_left,nbits)) \ - { action; } \ - get_buffer = (state).get_buffer; bits_left = (state).bits_left; } } - -#define GET_BITS(nbits) \ - (((int) (get_buffer >> (bits_left -= (nbits)))) & BIT_MASK(nbits)) - -#define PEEK_BITS(nbits) \ - (((int) (get_buffer >> (bits_left - (nbits)))) & BIT_MASK(nbits)) - -#define DROP_BITS(nbits) \ - (bits_left -= (nbits)) - - -/* - * Code for extracting next Huffman-coded symbol from input bit stream. - * Again, this is time-critical and we make the main paths be macros. - * - * We use a lookahead table to process codes of up to HUFF_LOOKAHEAD bits - * without looping. Usually, more than 95% of the Huffman codes will be 8 - * or fewer bits long. The few overlength codes are handled with a loop, - * which need not be inline code. - * - * Notes about the HUFF_DECODE macro: - * 1. Near the end of the data segment, we may fail to get enough bits - * for a lookahead. In that case, we do it the hard way. - * 2. If the lookahead table contains no entry, the next code must be - * more than HUFF_LOOKAHEAD bits long. - * 3. jpeg_huff_decode returns -1 if forced to suspend. - */ - -#define HUFF_DECODE(result,state,htbl,failaction,slowlabel) \ -{ register int nb, look; \ - if (bits_left < HUFF_LOOKAHEAD) { \ - if (! jpeg_fill_bit_buffer(&state,get_buffer,bits_left, 0)) {failaction;} \ - get_buffer = state.get_buffer; bits_left = state.bits_left; \ - if (bits_left < HUFF_LOOKAHEAD) { \ - nb = 1; goto slowlabel; \ - } \ - } \ - look = PEEK_BITS(HUFF_LOOKAHEAD); \ - if ((nb = htbl->look_nbits[look]) != 0) { \ - DROP_BITS(nb); \ - result = htbl->look_sym[look]; \ - } else { \ - nb = HUFF_LOOKAHEAD+1; \ -slowlabel: \ - if ((result=jpeg_huff_decode(&state,get_buffer,bits_left,htbl,nb)) < 0) \ - { failaction; } \ - get_buffer = state.get_buffer; bits_left = state.bits_left; \ - } \ -} - - -/* - * Expanded entropy decoder object for Huffman decoding. - * - * The savable_state subrecord contains fields that change within an MCU, - * but must not be updated permanently until we complete the MCU. - */ - -typedef struct { - unsigned int EOBRUN; /* remaining EOBs in EOBRUN */ - int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */ -} savable_state; - -/* This macro is to work around compilers with missing or broken - * structure assignment. You'll need to fix this code if you have - * such a compiler and you change MAX_COMPS_IN_SCAN. - */ - -#ifndef NO_STRUCT_ASSIGN -#define ASSIGN_STATE(dest,src) ((dest) = (src)) -#else -#if MAX_COMPS_IN_SCAN == 4 -#define ASSIGN_STATE(dest,src) \ - ((dest).EOBRUN = (src).EOBRUN, \ - (dest).last_dc_val[0] = (src).last_dc_val[0], \ - (dest).last_dc_val[1] = (src).last_dc_val[1], \ - (dest).last_dc_val[2] = (src).last_dc_val[2], \ - (dest).last_dc_val[3] = (src).last_dc_val[3]) -#endif -#endif - - -typedef struct { - struct jpeg_entropy_decoder pub; /* public fields */ - - /* These fields are loaded into local variables at start of each MCU. - * In case of suspension, we exit WITHOUT updating them. - */ - bitread_perm_state bitstate; /* Bit buffer at start of MCU */ - savable_state saved; /* Other state at start of MCU */ - - /* These fields are NOT loaded into local working state. */ - boolean insufficient_data; /* set TRUE after emitting warning */ - unsigned int restarts_to_go; /* MCUs left in this restart interval */ - - /* Following two fields used only in progressive mode */ - - /* Pointers to derived tables (these workspaces have image lifespan) */ - d_derived_tbl * derived_tbls[NUM_HUFF_TBLS]; - - d_derived_tbl * ac_derived_tbl; /* active table during an AC scan */ - - /* Following fields used only in sequential mode */ - - /* Pointers to derived tables (these workspaces have image lifespan) */ - d_derived_tbl * dc_derived_tbls[NUM_HUFF_TBLS]; - d_derived_tbl * ac_derived_tbls[NUM_HUFF_TBLS]; - - /* Precalculated info set up by start_pass for use in decode_mcu: */ - - /* Pointers to derived tables to be used for each block within an MCU */ - d_derived_tbl * dc_cur_tbls[D_MAX_BLOCKS_IN_MCU]; - d_derived_tbl * ac_cur_tbls[D_MAX_BLOCKS_IN_MCU]; - /* Whether we care about the DC and AC coefficient values for each block */ - int coef_limit[D_MAX_BLOCKS_IN_MCU]; -} huff_entropy_decoder; - -typedef huff_entropy_decoder * huff_entropy_ptr; - - -static const int jpeg_zigzag_order[8][8] = { - { 0, 1, 5, 6, 14, 15, 27, 28 }, - { 2, 4, 7, 13, 16, 26, 29, 42 }, - { 3, 8, 12, 17, 25, 30, 41, 43 }, - { 9, 11, 18, 24, 31, 40, 44, 53 }, - { 10, 19, 23, 32, 39, 45, 52, 54 }, - { 20, 22, 33, 38, 46, 51, 55, 60 }, - { 21, 34, 37, 47, 50, 56, 59, 61 }, - { 35, 36, 48, 49, 57, 58, 62, 63 } -}; - -static const int jpeg_zigzag_order7[7][7] = { - { 0, 1, 5, 6, 14, 15, 27 }, - { 2, 4, 7, 13, 16, 26, 28 }, - { 3, 8, 12, 17, 25, 29, 38 }, - { 9, 11, 18, 24, 30, 37, 39 }, - { 10, 19, 23, 31, 36, 40, 45 }, - { 20, 22, 32, 35, 41, 44, 46 }, - { 21, 33, 34, 42, 43, 47, 48 } -}; - -static const int jpeg_zigzag_order6[6][6] = { - { 0, 1, 5, 6, 14, 15 }, - { 2, 4, 7, 13, 16, 25 }, - { 3, 8, 12, 17, 24, 26 }, - { 9, 11, 18, 23, 27, 32 }, - { 10, 19, 22, 28, 31, 33 }, - { 20, 21, 29, 30, 34, 35 } -}; - -static const int jpeg_zigzag_order5[5][5] = { - { 0, 1, 5, 6, 14 }, - { 2, 4, 7, 13, 15 }, - { 3, 8, 12, 16, 21 }, - { 9, 11, 17, 20, 22 }, - { 10, 18, 19, 23, 24 } -}; - -static const int jpeg_zigzag_order4[4][4] = { - { 0, 1, 5, 6 }, - { 2, 4, 7, 12 }, - { 3, 8, 11, 13 }, - { 9, 10, 14, 15 } -}; - -static const int jpeg_zigzag_order3[3][3] = { - { 0, 1, 5 }, - { 2, 4, 6 }, - { 3, 7, 8 } -}; - -static const int jpeg_zigzag_order2[2][2] = { - { 0, 1 }, - { 2, 3 } -}; - - -/* - * Compute the derived values for a Huffman table. - * This routine also performs some validation checks on the table. - */ - -LOCAL(void) -jpeg_make_d_derived_tbl (j_decompress_ptr cinfo, boolean isDC, int tblno, - d_derived_tbl ** pdtbl) -{ - JHUFF_TBL *htbl; - d_derived_tbl *dtbl; - int p, i, l, si, numsymbols; - int lookbits, ctr; - char huffsize[257]; - unsigned int huffcode[257]; - unsigned int code; - - /* Note that huffsize[] and huffcode[] are filled in code-length order, - * paralleling the order of the symbols themselves in htbl->huffval[]. - */ - - /* Find the input Huffman table */ - if (tblno < 0 || tblno >= NUM_HUFF_TBLS) - ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno); - htbl = - isDC ? cinfo->dc_huff_tbl_ptrs[tblno] : cinfo->ac_huff_tbl_ptrs[tblno]; - if (htbl == NULL) - ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, tblno); - - /* Allocate a workspace if we haven't already done so. */ - if (*pdtbl == NULL) - *pdtbl = (d_derived_tbl *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(d_derived_tbl)); - dtbl = *pdtbl; - dtbl->pub = htbl; /* fill in back link */ - - /* Figure C.1: make table of Huffman code length for each symbol */ - - p = 0; - for (l = 1; l <= 16; l++) { - i = (int) htbl->bits[l]; - if (i < 0 || p + i > 256) /* protect against table overrun */ - ERREXIT(cinfo, JERR_BAD_HUFF_TABLE); - while (i--) - huffsize[p++] = (char) l; - } - huffsize[p] = 0; - numsymbols = p; - - /* Figure C.2: generate the codes themselves */ - /* We also validate that the counts represent a legal Huffman code tree. */ - - code = 0; - si = huffsize[0]; - p = 0; - while (huffsize[p]) { - while (((int) huffsize[p]) == si) { - huffcode[p++] = code; - code++; - } - /* code is now 1 more than the last code used for codelength si; but - * it must still fit in si bits, since no code is allowed to be all ones. - */ - if (((INT32) code) >= (((INT32) 1) << si)) - ERREXIT(cinfo, JERR_BAD_HUFF_TABLE); - code <<= 1; - si++; - } - - /* Figure F.15: generate decoding tables for bit-sequential decoding */ - - p = 0; - for (l = 1; l <= 16; l++) { - if (htbl->bits[l]) { - /* valoffset[l] = huffval[] index of 1st symbol of code length l, - * minus the minimum code of length l - */ - dtbl->valoffset[l] = (INT32) p - (INT32) huffcode[p]; - p += htbl->bits[l]; - dtbl->maxcode[l] = huffcode[p-1]; /* maximum code of length l */ - } else { - dtbl->maxcode[l] = -1; /* -1 if no codes of this length */ - } - } - dtbl->maxcode[17] = 0xFFFFFL; /* ensures jpeg_huff_decode terminates */ - - /* Compute lookahead tables to speed up decoding. - * First we set all the table entries to 0, indicating "too long"; - * then we iterate through the Huffman codes that are short enough and - * fill in all the entries that correspond to bit sequences starting - * with that code. - */ - - MEMZERO(dtbl->look_nbits, SIZEOF(dtbl->look_nbits)); - - p = 0; - for (l = 1; l <= HUFF_LOOKAHEAD; l++) { - for (i = 1; i <= (int) htbl->bits[l]; i++, p++) { - /* l = current code's length, p = its index in huffcode[] & huffval[]. */ - /* Generate left-justified code followed by all possible bit sequences */ - lookbits = huffcode[p] << (HUFF_LOOKAHEAD-l); - for (ctr = 1 << (HUFF_LOOKAHEAD-l); ctr > 0; ctr--) { - dtbl->look_nbits[lookbits] = l; - dtbl->look_sym[lookbits] = htbl->huffval[p]; - lookbits++; - } - } - } - - /* Validate symbols as being reasonable. - * For AC tables, we make no check, but accept all byte values 0..255. - * For DC tables, we require the symbols to be in range 0..15. - * (Tighter bounds could be applied depending on the data depth and mode, - * but this is sufficient to ensure safe decoding.) - */ - if (isDC) { - for (i = 0; i < numsymbols; i++) { - int sym = htbl->huffval[i]; - if (sym < 0 || sym > 15) - ERREXIT(cinfo, JERR_BAD_HUFF_TABLE); - } - } -} - - -/* - * Out-of-line code for bit fetching. - * Note: current values of get_buffer and bits_left are passed as parameters, - * but are returned in the corresponding fields of the state struct. - * - * On most machines MIN_GET_BITS should be 25 to allow the full 32-bit width - * of get_buffer to be used. (On machines with wider words, an even larger - * buffer could be used.) However, on some machines 32-bit shifts are - * quite slow and take time proportional to the number of places shifted. - * (This is true with most PC compilers, for instance.) In this case it may - * be a win to set MIN_GET_BITS to the minimum value of 15. This reduces the - * average shift distance at the cost of more calls to jpeg_fill_bit_buffer. - */ - -#ifdef SLOW_SHIFT_32 -#define MIN_GET_BITS 15 /* minimum allowable value */ -#else -#define MIN_GET_BITS (BIT_BUF_SIZE-7) -#endif - - -LOCAL(boolean) -jpeg_fill_bit_buffer (bitread_working_state * state, - register bit_buf_type get_buffer, register int bits_left, - int nbits) -/* Load up the bit buffer to a depth of at least nbits */ -{ - /* Copy heavily used state fields into locals (hopefully registers) */ - register const JOCTET * next_input_byte = state->next_input_byte; - register size_t bytes_in_buffer = state->bytes_in_buffer; - j_decompress_ptr cinfo = state->cinfo; - - /* Attempt to load at least MIN_GET_BITS bits into get_buffer. */ - /* (It is assumed that no request will be for more than that many bits.) */ - /* We fail to do so only if we hit a marker or are forced to suspend. */ - - if (cinfo->unread_marker == 0) { /* cannot advance past a marker */ - while (bits_left < MIN_GET_BITS) { - register int c; - - /* Attempt to read a byte */ - if (bytes_in_buffer == 0) { - if (! (*cinfo->src->fill_input_buffer) (cinfo)) - return FALSE; - next_input_byte = cinfo->src->next_input_byte; - bytes_in_buffer = cinfo->src->bytes_in_buffer; - } - bytes_in_buffer--; - c = GETJOCTET(*next_input_byte++); - - /* If it's 0xFF, check and discard stuffed zero byte */ - if (c == 0xFF) { - /* Loop here to discard any padding FF's on terminating marker, - * so that we can save a valid unread_marker value. NOTE: we will - * accept multiple FF's followed by a 0 as meaning a single FF data - * byte. This data pattern is not valid according to the standard. - */ - do { - if (bytes_in_buffer == 0) { - if (! (*cinfo->src->fill_input_buffer) (cinfo)) - return FALSE; - next_input_byte = cinfo->src->next_input_byte; - bytes_in_buffer = cinfo->src->bytes_in_buffer; - } - bytes_in_buffer--; - c = GETJOCTET(*next_input_byte++); - } while (c == 0xFF); - - if (c == 0) { - /* Found FF/00, which represents an FF data byte */ - c = 0xFF; - } else { - /* Oops, it's actually a marker indicating end of compressed data. - * Save the marker code for later use. - * Fine point: it might appear that we should save the marker into - * bitread working state, not straight into permanent state. But - * once we have hit a marker, we cannot need to suspend within the - * current MCU, because we will read no more bytes from the data - * source. So it is OK to update permanent state right away. - */ - cinfo->unread_marker = c; - /* See if we need to insert some fake zero bits. */ - goto no_more_bytes; - } - } - - /* OK, load c into get_buffer */ - get_buffer = (get_buffer << 8) | c; - bits_left += 8; - } /* end while */ - } else { - no_more_bytes: - /* We get here if we've read the marker that terminates the compressed - * data segment. There should be enough bits in the buffer register - * to satisfy the request; if so, no problem. - */ - if (nbits > bits_left) { - /* Uh-oh. Report corrupted data to user and stuff zeroes into - * the data stream, so that we can produce some kind of image. - * We use a nonvolatile flag to ensure that only one warning message - * appears per data segment. - */ - if (! ((huff_entropy_ptr) cinfo->entropy)->insufficient_data) { - WARNMS(cinfo, JWRN_HIT_MARKER); - ((huff_entropy_ptr) cinfo->entropy)->insufficient_data = TRUE; - } - /* Fill the buffer with zero bits */ - get_buffer <<= MIN_GET_BITS - bits_left; - bits_left = MIN_GET_BITS; - } - } - - /* Unload the local registers */ - state->next_input_byte = next_input_byte; - state->bytes_in_buffer = bytes_in_buffer; - state->get_buffer = get_buffer; - state->bits_left = bits_left; - - return TRUE; -} - - -/* - * Figure F.12: extend sign bit. - * On some machines, a shift and sub will be faster than a table lookup. - */ - -#ifdef AVOID_TABLES - -#define BIT_MASK(nbits) ((1<<(nbits))-1) -#define HUFF_EXTEND(x,s) ((x) < (1<<((s)-1)) ? (x) - ((1<<(s))-1) : (x)) - -#else - -#define BIT_MASK(nbits) bmask[nbits] -#define HUFF_EXTEND(x,s) ((x) <= bmask[(s) - 1] ? (x) - bmask[s] : (x)) - -static const int bmask[16] = /* bmask[n] is mask for n rightmost bits */ - { 0, 0x0001, 0x0003, 0x0007, 0x000F, 0x001F, 0x003F, 0x007F, 0x00FF, - 0x01FF, 0x03FF, 0x07FF, 0x0FFF, 0x1FFF, 0x3FFF, 0x7FFF }; - -#endif /* AVOID_TABLES */ - - -/* - * Out-of-line code for Huffman code decoding. - */ - -LOCAL(int) -jpeg_huff_decode (bitread_working_state * state, - register bit_buf_type get_buffer, register int bits_left, - d_derived_tbl * htbl, int min_bits) -{ - register int l = min_bits; - register INT32 code; - - /* HUFF_DECODE has determined that the code is at least min_bits */ - /* bits long, so fetch that many bits in one swoop. */ - - CHECK_BIT_BUFFER(*state, l, return -1); - code = GET_BITS(l); - - /* Collect the rest of the Huffman code one bit at a time. */ - /* This is per Figure F.16 in the JPEG spec. */ - - while (code > htbl->maxcode[l]) { - code <<= 1; - CHECK_BIT_BUFFER(*state, 1, return -1); - code |= GET_BITS(1); - l++; - } - - /* Unload the local registers */ - state->get_buffer = get_buffer; - state->bits_left = bits_left; - - /* With garbage input we may reach the sentinel value l = 17. */ - - if (l > 16) { - WARNMS(state->cinfo, JWRN_HUFF_BAD_CODE); - return 0; /* fake a zero as the safest result */ - } - - return htbl->pub->huffval[ (int) (code + htbl->valoffset[l]) ]; -} - - -/* - * Check for a restart marker & resynchronize decoder. - * Returns FALSE if must suspend. - */ - -LOCAL(boolean) -process_restart (j_decompress_ptr cinfo) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int ci; - - /* Throw away any unused bits remaining in bit buffer; */ - /* include any full bytes in next_marker's count of discarded bytes */ - cinfo->marker->discarded_bytes += entropy->bitstate.bits_left / 8; - entropy->bitstate.bits_left = 0; - - /* Advance past the RSTn marker */ - if (! (*cinfo->marker->read_restart_marker) (cinfo)) - return FALSE; - - /* Re-initialize DC predictions to 0 */ - for (ci = 0; ci < cinfo->comps_in_scan; ci++) - entropy->saved.last_dc_val[ci] = 0; - /* Re-init EOB run count, too */ - entropy->saved.EOBRUN = 0; - - /* Reset restart counter */ - entropy->restarts_to_go = cinfo->restart_interval; - - /* Reset out-of-data flag, unless read_restart_marker left us smack up - * against a marker. In that case we will end up treating the next data - * segment as empty, and we can avoid producing bogus output pixels by - * leaving the flag set. - */ - if (cinfo->unread_marker == 0) - entropy->insufficient_data = FALSE; - - return TRUE; -} - - -/* - * Huffman MCU decoding. - * Each of these routines decodes and returns one MCU's worth of - * Huffman-compressed coefficients. - * The coefficients are reordered from zigzag order into natural array order, - * but are not dequantized. - * - * The i'th block of the MCU is stored into the block pointed to by - * MCU_data[i]. WE ASSUME THIS AREA IS INITIALLY ZEROED BY THE CALLER. - * (Wholesale zeroing is usually a little faster than retail...) - * - * We return FALSE if data source requested suspension. In that case no - * changes have been made to permanent state. (Exception: some output - * coefficients may already have been assigned. This is harmless for - * spectral selection, since we'll just re-assign them on the next call. - * Successive approximation AC refinement has to be more careful, however.) - */ - -/* - * MCU decoding for DC initial scan (either spectral selection, - * or first pass of successive approximation). - */ - -METHODDEF(boolean) -decode_mcu_DC_first (j_decompress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int Al = cinfo->Al; - register int s, r; - int blkn, ci; - JBLOCKROW block; - BITREAD_STATE_VARS; - savable_state state; - d_derived_tbl * tbl; - jpeg_component_info * compptr; - - /* Process restart marker if needed; may have to suspend */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! process_restart(cinfo)) - return FALSE; - } - - /* If we've run out of data, just leave the MCU set to zeroes. - * This way, we return uniform gray for the remainder of the segment. - */ - if (! entropy->insufficient_data) { - - /* Load up working state */ - BITREAD_LOAD_STATE(cinfo,entropy->bitstate); - ASSIGN_STATE(state, entropy->saved); - - /* Outer loop handles each block in the MCU */ - - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - block = MCU_data[blkn]; - ci = cinfo->MCU_membership[blkn]; - compptr = cinfo->cur_comp_info[ci]; - tbl = entropy->derived_tbls[compptr->dc_tbl_no]; - - /* Decode a single block's worth of coefficients */ - - /* Section F.2.2.1: decode the DC coefficient difference */ - HUFF_DECODE(s, br_state, tbl, return FALSE, label1); - if (s) { - CHECK_BIT_BUFFER(br_state, s, return FALSE); - r = GET_BITS(s); - s = HUFF_EXTEND(r, s); - } - - /* Convert DC difference to actual value, update last_dc_val */ - s += state.last_dc_val[ci]; - state.last_dc_val[ci] = s; - /* Scale and output the coefficient (assumes jpeg_natural_order[0]=0) */ - (*block)[0] = (JCOEF) (s << Al); - } - - /* Completed MCU, so update state */ - BITREAD_SAVE_STATE(cinfo,entropy->bitstate); - ASSIGN_STATE(entropy->saved, state); - } - - /* Account for restart interval (no-op if not using restarts) */ - entropy->restarts_to_go--; - - return TRUE; -} - - -/* - * MCU decoding for AC initial scan (either spectral selection, - * or first pass of successive approximation). - */ - -METHODDEF(boolean) -decode_mcu_AC_first (j_decompress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - register int s, k, r; - unsigned int EOBRUN; - int Se, Al; - const int * natural_order; - JBLOCKROW block; - BITREAD_STATE_VARS; - d_derived_tbl * tbl; - - /* Process restart marker if needed; may have to suspend */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! process_restart(cinfo)) - return FALSE; - } - - /* If we've run out of data, just leave the MCU set to zeroes. - * This way, we return uniform gray for the remainder of the segment. - */ - if (! entropy->insufficient_data) { - - Se = cinfo->Se; - Al = cinfo->Al; - natural_order = cinfo->natural_order; - - /* Load up working state. - * We can avoid loading/saving bitread state if in an EOB run. - */ - EOBRUN = entropy->saved.EOBRUN; /* only part of saved state we need */ - - /* There is always only one block per MCU */ - - if (EOBRUN > 0) /* if it's a band of zeroes... */ - EOBRUN--; /* ...process it now (we do nothing) */ - else { - BITREAD_LOAD_STATE(cinfo,entropy->bitstate); - block = MCU_data[0]; - tbl = entropy->ac_derived_tbl; - - for (k = cinfo->Ss; k <= Se; k++) { - HUFF_DECODE(s, br_state, tbl, return FALSE, label2); - r = s >> 4; - s &= 15; - if (s) { - k += r; - CHECK_BIT_BUFFER(br_state, s, return FALSE); - r = GET_BITS(s); - s = HUFF_EXTEND(r, s); - /* Scale and output coefficient in natural (dezigzagged) order */ - (*block)[natural_order[k]] = (JCOEF) (s << Al); - } else { - if (r == 15) { /* ZRL */ - k += 15; /* skip 15 zeroes in band */ - } else { /* EOBr, run length is 2^r + appended bits */ - EOBRUN = 1 << r; - if (r) { /* EOBr, r > 0 */ - CHECK_BIT_BUFFER(br_state, r, return FALSE); - r = GET_BITS(r); - EOBRUN += r; - } - EOBRUN--; /* this band is processed at this moment */ - break; /* force end-of-band */ - } - } - } - - BITREAD_SAVE_STATE(cinfo,entropy->bitstate); - } - - /* Completed MCU, so update state */ - entropy->saved.EOBRUN = EOBRUN; /* only part of saved state we need */ - } - - /* Account for restart interval (no-op if not using restarts) */ - entropy->restarts_to_go--; - - return TRUE; -} - - -/* - * MCU decoding for DC successive approximation refinement scan. - * Note: we assume such scans can be multi-component, although the spec - * is not very clear on the point. - */ - -METHODDEF(boolean) -decode_mcu_DC_refine (j_decompress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int p1 = 1 << cinfo->Al; /* 1 in the bit position being coded */ - int blkn; - JBLOCKROW block; - BITREAD_STATE_VARS; - - /* Process restart marker if needed; may have to suspend */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! process_restart(cinfo)) - return FALSE; - } - - /* Not worth the cycles to check insufficient_data here, - * since we will not change the data anyway if we read zeroes. - */ - - /* Load up working state */ - BITREAD_LOAD_STATE(cinfo,entropy->bitstate); - - /* Outer loop handles each block in the MCU */ - - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - block = MCU_data[blkn]; - - /* Encoded data is simply the next bit of the two's-complement DC value */ - CHECK_BIT_BUFFER(br_state, 1, return FALSE); - if (GET_BITS(1)) - (*block)[0] |= p1; - /* Note: since we use |=, repeating the assignment later is safe */ - } - - /* Completed MCU, so update state */ - BITREAD_SAVE_STATE(cinfo,entropy->bitstate); - - /* Account for restart interval (no-op if not using restarts) */ - entropy->restarts_to_go--; - - return TRUE; -} - - -/* - * MCU decoding for AC successive approximation refinement scan. - */ - -METHODDEF(boolean) -decode_mcu_AC_refine (j_decompress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - register int s, k, r; - unsigned int EOBRUN; - int Se, p1, m1; - const int * natural_order; - JBLOCKROW block; - JCOEFPTR thiscoef; - BITREAD_STATE_VARS; - d_derived_tbl * tbl; - int num_newnz; - int newnz_pos[DCTSIZE2]; - - /* Process restart marker if needed; may have to suspend */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! process_restart(cinfo)) - return FALSE; - } - - /* If we've run out of data, don't modify the MCU. - */ - if (! entropy->insufficient_data) { - - Se = cinfo->Se; - p1 = 1 << cinfo->Al; /* 1 in the bit position being coded */ - m1 = (-1) << cinfo->Al; /* -1 in the bit position being coded */ - natural_order = cinfo->natural_order; - - /* Load up working state */ - BITREAD_LOAD_STATE(cinfo,entropy->bitstate); - EOBRUN = entropy->saved.EOBRUN; /* only part of saved state we need */ - - /* There is always only one block per MCU */ - block = MCU_data[0]; - tbl = entropy->ac_derived_tbl; - - /* If we are forced to suspend, we must undo the assignments to any newly - * nonzero coefficients in the block, because otherwise we'd get confused - * next time about which coefficients were already nonzero. - * But we need not undo addition of bits to already-nonzero coefficients; - * instead, we can test the current bit to see if we already did it. - */ - num_newnz = 0; - - /* initialize coefficient loop counter to start of band */ - k = cinfo->Ss; - - if (EOBRUN == 0) { - for (; k <= Se; k++) { - HUFF_DECODE(s, br_state, tbl, goto undoit, label3); - r = s >> 4; - s &= 15; - if (s) { - if (s != 1) /* size of new coef should always be 1 */ - WARNMS(cinfo, JWRN_HUFF_BAD_CODE); - CHECK_BIT_BUFFER(br_state, 1, goto undoit); - if (GET_BITS(1)) - s = p1; /* newly nonzero coef is positive */ - else - s = m1; /* newly nonzero coef is negative */ - } else { - if (r != 15) { - EOBRUN = 1 << r; /* EOBr, run length is 2^r + appended bits */ - if (r) { - CHECK_BIT_BUFFER(br_state, r, goto undoit); - r = GET_BITS(r); - EOBRUN += r; - } - break; /* rest of block is handled by EOB logic */ - } - /* note s = 0 for processing ZRL */ - } - /* Advance over already-nonzero coefs and r still-zero coefs, - * appending correction bits to the nonzeroes. A correction bit is 1 - * if the absolute value of the coefficient must be increased. - */ - do { - thiscoef = *block + natural_order[k]; - if (*thiscoef != 0) { - CHECK_BIT_BUFFER(br_state, 1, goto undoit); - if (GET_BITS(1)) { - if ((*thiscoef & p1) == 0) { /* do nothing if already set it */ - if (*thiscoef >= 0) - *thiscoef += p1; - else - *thiscoef += m1; - } - } - } else { - if (--r < 0) - break; /* reached target zero coefficient */ - } - k++; - } while (k <= Se); - if (s) { - int pos = natural_order[k]; - /* Output newly nonzero coefficient */ - (*block)[pos] = (JCOEF) s; - /* Remember its position in case we have to suspend */ - newnz_pos[num_newnz++] = pos; - } - } - } - - if (EOBRUN > 0) { - /* Scan any remaining coefficient positions after the end-of-band - * (the last newly nonzero coefficient, if any). Append a correction - * bit to each already-nonzero coefficient. A correction bit is 1 - * if the absolute value of the coefficient must be increased. - */ - for (; k <= Se; k++) { - thiscoef = *block + natural_order[k]; - if (*thiscoef != 0) { - CHECK_BIT_BUFFER(br_state, 1, goto undoit); - if (GET_BITS(1)) { - if ((*thiscoef & p1) == 0) { /* do nothing if already changed it */ - if (*thiscoef >= 0) - *thiscoef += p1; - else - *thiscoef += m1; - } - } - } - } - /* Count one block completed in EOB run */ - EOBRUN--; - } - - /* Completed MCU, so update state */ - BITREAD_SAVE_STATE(cinfo,entropy->bitstate); - entropy->saved.EOBRUN = EOBRUN; /* only part of saved state we need */ - } - - /* Account for restart interval (no-op if not using restarts) */ - entropy->restarts_to_go--; - - return TRUE; - -undoit: - /* Re-zero any output coefficients that we made newly nonzero */ - while (num_newnz > 0) - (*block)[newnz_pos[--num_newnz]] = 0; - - return FALSE; -} - - -/* - * Decode one MCU's worth of Huffman-compressed coefficients, - * partial blocks. - */ - -METHODDEF(boolean) -decode_mcu_sub (j_decompress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - const int * natural_order; - int Se, blkn; - BITREAD_STATE_VARS; - savable_state state; - - /* Process restart marker if needed; may have to suspend */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! process_restart(cinfo)) - return FALSE; - } - - /* If we've run out of data, just leave the MCU set to zeroes. - * This way, we return uniform gray for the remainder of the segment. - */ - if (! entropy->insufficient_data) { - - natural_order = cinfo->natural_order; - Se = cinfo->lim_Se; - - /* Load up working state */ - BITREAD_LOAD_STATE(cinfo,entropy->bitstate); - ASSIGN_STATE(state, entropy->saved); - - /* Outer loop handles each block in the MCU */ - - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - JBLOCKROW block = MCU_data[blkn]; - d_derived_tbl * htbl; - register int s, k, r; - int coef_limit, ci; - - /* Decode a single block's worth of coefficients */ - - /* Section F.2.2.1: decode the DC coefficient difference */ - htbl = entropy->dc_cur_tbls[blkn]; - HUFF_DECODE(s, br_state, htbl, return FALSE, label1); - - htbl = entropy->ac_cur_tbls[blkn]; - k = 1; - coef_limit = entropy->coef_limit[blkn]; - if (coef_limit) { - /* Convert DC difference to actual value, update last_dc_val */ - if (s) { - CHECK_BIT_BUFFER(br_state, s, return FALSE); - r = GET_BITS(s); - s = HUFF_EXTEND(r, s); - } - ci = cinfo->MCU_membership[blkn]; - s += state.last_dc_val[ci]; - state.last_dc_val[ci] = s; - /* Output the DC coefficient */ - (*block)[0] = (JCOEF) s; - - /* Section F.2.2.2: decode the AC coefficients */ - /* Since zeroes are skipped, output area must be cleared beforehand */ - for (; k < coef_limit; k++) { - HUFF_DECODE(s, br_state, htbl, return FALSE, label2); - - r = s >> 4; - s &= 15; - - if (s) { - k += r; - CHECK_BIT_BUFFER(br_state, s, return FALSE); - r = GET_BITS(s); - s = HUFF_EXTEND(r, s); - /* Output coefficient in natural (dezigzagged) order. - * Note: the extra entries in natural_order[] will save us - * if k > Se, which could happen if the data is corrupted. - */ - (*block)[natural_order[k]] = (JCOEF) s; - } else { - if (r != 15) - goto EndOfBlock; - k += 15; - } - } - } else { - if (s) { - CHECK_BIT_BUFFER(br_state, s, return FALSE); - DROP_BITS(s); - } - } - - /* Section F.2.2.2: decode the AC coefficients */ - /* In this path we just discard the values */ - for (; k <= Se; k++) { - HUFF_DECODE(s, br_state, htbl, return FALSE, label3); - - r = s >> 4; - s &= 15; - - if (s) { - k += r; - CHECK_BIT_BUFFER(br_state, s, return FALSE); - DROP_BITS(s); - } else { - if (r != 15) - break; - k += 15; - } - } - - EndOfBlock: ; - } - - /* Completed MCU, so update state */ - BITREAD_SAVE_STATE(cinfo,entropy->bitstate); - ASSIGN_STATE(entropy->saved, state); - } - - /* Account for restart interval (no-op if not using restarts) */ - entropy->restarts_to_go--; - - return TRUE; -} - - -/* - * Decode one MCU's worth of Huffman-compressed coefficients, - * full-size blocks. - */ - -METHODDEF(boolean) -decode_mcu (j_decompress_ptr cinfo, JBLOCKROW *MCU_data) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int blkn; - BITREAD_STATE_VARS; - savable_state state; - - /* Process restart marker if needed; may have to suspend */ - if (cinfo->restart_interval) { - if (entropy->restarts_to_go == 0) - if (! process_restart(cinfo)) - return FALSE; - } - - /* If we've run out of data, just leave the MCU set to zeroes. - * This way, we return uniform gray for the remainder of the segment. - */ - if (! entropy->insufficient_data) { - - /* Load up working state */ - BITREAD_LOAD_STATE(cinfo,entropy->bitstate); - ASSIGN_STATE(state, entropy->saved); - - /* Outer loop handles each block in the MCU */ - - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - JBLOCKROW block = MCU_data[blkn]; - d_derived_tbl * htbl; - register int s, k, r; - int coef_limit, ci; - - /* Decode a single block's worth of coefficients */ - - /* Section F.2.2.1: decode the DC coefficient difference */ - htbl = entropy->dc_cur_tbls[blkn]; - HUFF_DECODE(s, br_state, htbl, return FALSE, label1); - - htbl = entropy->ac_cur_tbls[blkn]; - k = 1; - coef_limit = entropy->coef_limit[blkn]; - if (coef_limit) { - /* Convert DC difference to actual value, update last_dc_val */ - if (s) { - CHECK_BIT_BUFFER(br_state, s, return FALSE); - r = GET_BITS(s); - s = HUFF_EXTEND(r, s); - } - ci = cinfo->MCU_membership[blkn]; - s += state.last_dc_val[ci]; - state.last_dc_val[ci] = s; - /* Output the DC coefficient */ - (*block)[0] = (JCOEF) s; - - /* Section F.2.2.2: decode the AC coefficients */ - /* Since zeroes are skipped, output area must be cleared beforehand */ - for (; k < coef_limit; k++) { - HUFF_DECODE(s, br_state, htbl, return FALSE, label2); - - r = s >> 4; - s &= 15; - - if (s) { - k += r; - CHECK_BIT_BUFFER(br_state, s, return FALSE); - r = GET_BITS(s); - s = HUFF_EXTEND(r, s); - /* Output coefficient in natural (dezigzagged) order. - * Note: the extra entries in jpeg_natural_order[] will save us - * if k >= DCTSIZE2, which could happen if the data is corrupted. - */ - (*block)[jpeg_natural_order[k]] = (JCOEF) s; - } else { - if (r != 15) - goto EndOfBlock; - k += 15; - } - } - } else { - if (s) { - CHECK_BIT_BUFFER(br_state, s, return FALSE); - DROP_BITS(s); - } - } - - /* Section F.2.2.2: decode the AC coefficients */ - /* In this path we just discard the values */ - for (; k < DCTSIZE2; k++) { - HUFF_DECODE(s, br_state, htbl, return FALSE, label3); - - r = s >> 4; - s &= 15; - - if (s) { - k += r; - CHECK_BIT_BUFFER(br_state, s, return FALSE); - DROP_BITS(s); - } else { - if (r != 15) - break; - k += 15; - } - } - - EndOfBlock: ; - } - - /* Completed MCU, so update state */ - BITREAD_SAVE_STATE(cinfo,entropy->bitstate); - ASSIGN_STATE(entropy->saved, state); - } - - /* Account for restart interval (no-op if not using restarts) */ - entropy->restarts_to_go--; - - return TRUE; -} - - -/* - * Initialize for a Huffman-compressed scan. - */ - -METHODDEF(void) -start_pass_huff_decoder (j_decompress_ptr cinfo) -{ - huff_entropy_ptr entropy = (huff_entropy_ptr) cinfo->entropy; - int ci, blkn, tbl, i; - jpeg_component_info * compptr; - - if (cinfo->progressive_mode) { - /* Validate progressive scan parameters */ - if (cinfo->Ss == 0) { - if (cinfo->Se != 0) - goto bad; - } else { - /* need not check Ss/Se < 0 since they came from unsigned bytes */ - if (cinfo->Se < cinfo->Ss || cinfo->Se > cinfo->lim_Se) - goto bad; - /* AC scans may have only one component */ - if (cinfo->comps_in_scan != 1) - goto bad; - } - if (cinfo->Ah != 0) { - /* Successive approximation refinement scan: must have Al = Ah-1. */ - if (cinfo->Ah-1 != cinfo->Al) - goto bad; - } - if (cinfo->Al > 13) { /* need not check for < 0 */ - /* Arguably the maximum Al value should be less than 13 for 8-bit precision, - * but the spec doesn't say so, and we try to be liberal about what we - * accept. Note: large Al values could result in out-of-range DC - * coefficients during early scans, leading to bizarre displays due to - * overflows in the IDCT math. But we won't crash. - */ - bad: - ERREXIT4(cinfo, JERR_BAD_PROGRESSION, - cinfo->Ss, cinfo->Se, cinfo->Ah, cinfo->Al); - } - /* Update progression status, and verify that scan order is legal. - * Note that inter-scan inconsistencies are treated as warnings - * not fatal errors ... not clear if this is right way to behave. - */ - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - int coefi, cindex = cinfo->cur_comp_info[ci]->component_index; - int *coef_bit_ptr = & cinfo->coef_bits[cindex][0]; - if (cinfo->Ss && coef_bit_ptr[0] < 0) /* AC without prior DC scan */ - WARNMS2(cinfo, JWRN_BOGUS_PROGRESSION, cindex, 0); - for (coefi = cinfo->Ss; coefi <= cinfo->Se; coefi++) { - int expected = (coef_bit_ptr[coefi] < 0) ? 0 : coef_bit_ptr[coefi]; - if (cinfo->Ah != expected) - WARNMS2(cinfo, JWRN_BOGUS_PROGRESSION, cindex, coefi); - coef_bit_ptr[coefi] = cinfo->Al; - } - } - - /* Select MCU decoding routine */ - if (cinfo->Ah == 0) { - if (cinfo->Ss == 0) - entropy->pub.decode_mcu = decode_mcu_DC_first; - else - entropy->pub.decode_mcu = decode_mcu_AC_first; - } else { - if (cinfo->Ss == 0) - entropy->pub.decode_mcu = decode_mcu_DC_refine; - else - entropy->pub.decode_mcu = decode_mcu_AC_refine; - } - - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - compptr = cinfo->cur_comp_info[ci]; - /* Make sure requested tables are present, and compute derived tables. - * We may build same derived table more than once, but it's not expensive. - */ - if (cinfo->Ss == 0) { - if (cinfo->Ah == 0) { /* DC refinement needs no table */ - tbl = compptr->dc_tbl_no; - jpeg_make_d_derived_tbl(cinfo, TRUE, tbl, - & entropy->derived_tbls[tbl]); - } - } else { - tbl = compptr->ac_tbl_no; - jpeg_make_d_derived_tbl(cinfo, FALSE, tbl, - & entropy->derived_tbls[tbl]); - /* remember the single active table */ - entropy->ac_derived_tbl = entropy->derived_tbls[tbl]; - } - /* Initialize DC predictions to 0 */ - entropy->saved.last_dc_val[ci] = 0; - } - - /* Initialize private state variables */ - entropy->saved.EOBRUN = 0; - } else { - /* Check that the scan parameters Ss, Se, Ah/Al are OK for sequential JPEG. - * This ought to be an error condition, but we make it a warning because - * there are some baseline files out there with all zeroes in these bytes. - */ - if (cinfo->Ss != 0 || cinfo->Ah != 0 || cinfo->Al != 0 || - ((cinfo->is_baseline || cinfo->Se < DCTSIZE2) && - cinfo->Se != cinfo->lim_Se)) - WARNMS(cinfo, JWRN_NOT_SEQUENTIAL); - - /* Select MCU decoding routine */ - /* We retain the hard-coded case for full-size blocks. - * This is not necessary, but it appears that this version is slightly - * more performant in the given implementation. - * With an improved implementation we would prefer a single optimized - * function. - */ - if (cinfo->lim_Se != DCTSIZE2-1) - entropy->pub.decode_mcu = decode_mcu_sub; - else - entropy->pub.decode_mcu = decode_mcu; - - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - compptr = cinfo->cur_comp_info[ci]; - /* Compute derived values for Huffman tables */ - /* We may do this more than once for a table, but it's not expensive */ - tbl = compptr->dc_tbl_no; - jpeg_make_d_derived_tbl(cinfo, TRUE, tbl, - & entropy->dc_derived_tbls[tbl]); - if (cinfo->lim_Se) { /* AC needs no table when not present */ - tbl = compptr->ac_tbl_no; - jpeg_make_d_derived_tbl(cinfo, FALSE, tbl, - & entropy->ac_derived_tbls[tbl]); - } - /* Initialize DC predictions to 0 */ - entropy->saved.last_dc_val[ci] = 0; - } - - /* Precalculate decoding info for each block in an MCU of this scan */ - for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) { - ci = cinfo->MCU_membership[blkn]; - compptr = cinfo->cur_comp_info[ci]; - /* Precalculate which table to use for each block */ - entropy->dc_cur_tbls[blkn] = entropy->dc_derived_tbls[compptr->dc_tbl_no]; - entropy->ac_cur_tbls[blkn] = entropy->ac_derived_tbls[compptr->ac_tbl_no]; - /* Decide whether we really care about the coefficient values */ - if (compptr->component_needed) { - ci = compptr->DCT_v_scaled_size; - i = compptr->DCT_h_scaled_size; - switch (cinfo->lim_Se) { - case (1*1-1): - entropy->coef_limit[blkn] = 1; - break; - case (2*2-1): - if (ci <= 0 || ci > 2) ci = 2; - if (i <= 0 || i > 2) i = 2; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order2[ci - 1][i - 1]; - break; - case (3*3-1): - if (ci <= 0 || ci > 3) ci = 3; - if (i <= 0 || i > 3) i = 3; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order3[ci - 1][i - 1]; - break; - case (4*4-1): - if (ci <= 0 || ci > 4) ci = 4; - if (i <= 0 || i > 4) i = 4; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order4[ci - 1][i - 1]; - break; - case (5*5-1): - if (ci <= 0 || ci > 5) ci = 5; - if (i <= 0 || i > 5) i = 5; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order5[ci - 1][i - 1]; - break; - case (6*6-1): - if (ci <= 0 || ci > 6) ci = 6; - if (i <= 0 || i > 6) i = 6; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order6[ci - 1][i - 1]; - break; - case (7*7-1): - if (ci <= 0 || ci > 7) ci = 7; - if (i <= 0 || i > 7) i = 7; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order7[ci - 1][i - 1]; - break; - default: - if (ci <= 0 || ci > 8) ci = 8; - if (i <= 0 || i > 8) i = 8; - entropy->coef_limit[blkn] = 1 + jpeg_zigzag_order[ci - 1][i - 1]; - break; - } - } else { - entropy->coef_limit[blkn] = 0; - } - } - } - - /* Initialize bitread state variables */ - entropy->bitstate.bits_left = 0; - entropy->bitstate.get_buffer = 0; /* unnecessary, but keeps Purify quiet */ - entropy->insufficient_data = FALSE; - - /* Initialize restart counter */ - entropy->restarts_to_go = cinfo->restart_interval; -} - - -/* - * Module initialization routine for Huffman entropy decoding. - */ - -GLOBAL(void) -jinit_huff_decoder (j_decompress_ptr cinfo) -{ - huff_entropy_ptr entropy; - int i; - - entropy = (huff_entropy_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(huff_entropy_decoder)); - cinfo->entropy = (struct jpeg_entropy_decoder *) entropy; - entropy->pub.start_pass = start_pass_huff_decoder; - - if (cinfo->progressive_mode) { - /* Create progression status table */ - int *coef_bit_ptr, ci; - cinfo->coef_bits = (int (*)[DCTSIZE2]) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - cinfo->num_components*DCTSIZE2*SIZEOF(int)); - coef_bit_ptr = & cinfo->coef_bits[0][0]; - for (ci = 0; ci < cinfo->num_components; ci++) - for (i = 0; i < DCTSIZE2; i++) - *coef_bit_ptr++ = -1; - - /* Mark derived tables unallocated */ - for (i = 0; i < NUM_HUFF_TBLS; i++) { - entropy->derived_tbls[i] = NULL; - } - } else { - /* Mark tables unallocated */ - for (i = 0; i < NUM_HUFF_TBLS; i++) { - entropy->dc_derived_tbls[i] = entropy->ac_derived_tbls[i] = NULL; - } - } -} diff --git a/src/3rdparty/libjpeg/jdinput.c b/src/3rdparty/libjpeg/jdinput.c deleted file mode 100644 index 2c5c717b9c..0000000000 --- a/src/3rdparty/libjpeg/jdinput.c +++ /dev/null @@ -1,661 +0,0 @@ -/* - * jdinput.c - * - * Copyright (C) 1991-1997, Thomas G. Lane. - * Modified 2002-2009 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains input control logic for the JPEG decompressor. - * These routines are concerned with controlling the decompressor's input - * processing (marker reading and coefficient decoding). The actual input - * reading is done in jdmarker.c, jdhuff.c, and jdarith.c. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Private state */ - -typedef struct { - struct jpeg_input_controller pub; /* public fields */ - - int inheaders; /* Nonzero until first SOS is reached */ -} my_input_controller; - -typedef my_input_controller * my_inputctl_ptr; - - -/* Forward declarations */ -METHODDEF(int) consume_markers JPP((j_decompress_ptr cinfo)); - - -/* - * Routines to calculate various quantities related to the size of the image. - */ - - -/* - * Compute output image dimensions and related values. - * NOTE: this is exported for possible use by application. - * Hence it mustn't do anything that can't be done twice. - */ - -GLOBAL(void) -jpeg_core_output_dimensions (j_decompress_ptr cinfo) -/* Do computations that are needed before master selection phase. - * This function is used for transcoding and full decompression. - */ -{ -#ifdef IDCT_SCALING_SUPPORTED - int ci; - jpeg_component_info *compptr; - - /* Compute actual output image dimensions and DCT scaling choices. */ - if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom) { - /* Provide 1/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 1; - cinfo->min_DCT_v_scaled_size = 1; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 2) { - /* Provide 2/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 2L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 2L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 2; - cinfo->min_DCT_v_scaled_size = 2; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 3) { - /* Provide 3/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 3L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 3L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 3; - cinfo->min_DCT_v_scaled_size = 3; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 4) { - /* Provide 4/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 4L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 4L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 4; - cinfo->min_DCT_v_scaled_size = 4; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 5) { - /* Provide 5/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 5L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 5L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 5; - cinfo->min_DCT_v_scaled_size = 5; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 6) { - /* Provide 6/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 6L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 6L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 6; - cinfo->min_DCT_v_scaled_size = 6; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 7) { - /* Provide 7/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 7L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 7L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 7; - cinfo->min_DCT_v_scaled_size = 7; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 8) { - /* Provide 8/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 8L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 8L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 8; - cinfo->min_DCT_v_scaled_size = 8; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 9) { - /* Provide 9/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 9L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 9L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 9; - cinfo->min_DCT_v_scaled_size = 9; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 10) { - /* Provide 10/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 10L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 10L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 10; - cinfo->min_DCT_v_scaled_size = 10; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 11) { - /* Provide 11/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 11L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 11L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 11; - cinfo->min_DCT_v_scaled_size = 11; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 12) { - /* Provide 12/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 12L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 12L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 12; - cinfo->min_DCT_v_scaled_size = 12; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 13) { - /* Provide 13/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 13L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 13L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 13; - cinfo->min_DCT_v_scaled_size = 13; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 14) { - /* Provide 14/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 14L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 14L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 14; - cinfo->min_DCT_v_scaled_size = 14; - } else if (cinfo->scale_num * cinfo->block_size <= cinfo->scale_denom * 15) { - /* Provide 15/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 15L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 15L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 15; - cinfo->min_DCT_v_scaled_size = 15; - } else { - /* Provide 16/block_size scaling */ - cinfo->output_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * 16L, (long) cinfo->block_size); - cinfo->output_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * 16L, (long) cinfo->block_size); - cinfo->min_DCT_h_scaled_size = 16; - cinfo->min_DCT_v_scaled_size = 16; - } - - /* Recompute dimensions of components */ - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - compptr->DCT_h_scaled_size = cinfo->min_DCT_h_scaled_size; - compptr->DCT_v_scaled_size = cinfo->min_DCT_v_scaled_size; - } - -#else /* !IDCT_SCALING_SUPPORTED */ - - /* Hardwire it to "no scaling" */ - cinfo->output_width = cinfo->image_width; - cinfo->output_height = cinfo->image_height; - /* jdinput.c has already initialized DCT_scaled_size, - * and has computed unscaled downsampled_width and downsampled_height. - */ - -#endif /* IDCT_SCALING_SUPPORTED */ -} - - -LOCAL(void) -initial_setup (j_decompress_ptr cinfo) -/* Called once, when first SOS marker is reached */ -{ - int ci; - jpeg_component_info *compptr; - - /* Make sure image isn't bigger than I can handle */ - if ((long) cinfo->image_height > (long) JPEG_MAX_DIMENSION || - (long) cinfo->image_width > (long) JPEG_MAX_DIMENSION) - ERREXIT1(cinfo, JERR_IMAGE_TOO_BIG, (unsigned int) JPEG_MAX_DIMENSION); - - /* For now, precision must match compiled-in value... */ - if (cinfo->data_precision != BITS_IN_JSAMPLE) - ERREXIT1(cinfo, JERR_BAD_PRECISION, cinfo->data_precision); - - /* Check that number of components won't exceed internal array sizes */ - if (cinfo->num_components > MAX_COMPONENTS) - ERREXIT2(cinfo, JERR_COMPONENT_COUNT, cinfo->num_components, - MAX_COMPONENTS); - - /* Compute maximum sampling factors; check factor validity */ - cinfo->max_h_samp_factor = 1; - cinfo->max_v_samp_factor = 1; - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - if (compptr->h_samp_factor<=0 || compptr->h_samp_factor>MAX_SAMP_FACTOR || - compptr->v_samp_factor<=0 || compptr->v_samp_factor>MAX_SAMP_FACTOR) - ERREXIT(cinfo, JERR_BAD_SAMPLING); - cinfo->max_h_samp_factor = MAX(cinfo->max_h_samp_factor, - compptr->h_samp_factor); - cinfo->max_v_samp_factor = MAX(cinfo->max_v_samp_factor, - compptr->v_samp_factor); - } - - /* Derive block_size, natural_order, and lim_Se */ - if (cinfo->is_baseline || (cinfo->progressive_mode && - cinfo->comps_in_scan)) { /* no pseudo SOS marker */ - cinfo->block_size = DCTSIZE; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - } else - switch (cinfo->Se) { - case (1*1-1): - cinfo->block_size = 1; - cinfo->natural_order = jpeg_natural_order; /* not needed */ - cinfo->lim_Se = cinfo->Se; - break; - case (2*2-1): - cinfo->block_size = 2; - cinfo->natural_order = jpeg_natural_order2; - cinfo->lim_Se = cinfo->Se; - break; - case (3*3-1): - cinfo->block_size = 3; - cinfo->natural_order = jpeg_natural_order3; - cinfo->lim_Se = cinfo->Se; - break; - case (4*4-1): - cinfo->block_size = 4; - cinfo->natural_order = jpeg_natural_order4; - cinfo->lim_Se = cinfo->Se; - break; - case (5*5-1): - cinfo->block_size = 5; - cinfo->natural_order = jpeg_natural_order5; - cinfo->lim_Se = cinfo->Se; - break; - case (6*6-1): - cinfo->block_size = 6; - cinfo->natural_order = jpeg_natural_order6; - cinfo->lim_Se = cinfo->Se; - break; - case (7*7-1): - cinfo->block_size = 7; - cinfo->natural_order = jpeg_natural_order7; - cinfo->lim_Se = cinfo->Se; - break; - case (8*8-1): - cinfo->block_size = 8; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (9*9-1): - cinfo->block_size = 9; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (10*10-1): - cinfo->block_size = 10; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (11*11-1): - cinfo->block_size = 11; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (12*12-1): - cinfo->block_size = 12; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (13*13-1): - cinfo->block_size = 13; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (14*14-1): - cinfo->block_size = 14; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (15*15-1): - cinfo->block_size = 15; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - case (16*16-1): - cinfo->block_size = 16; - cinfo->natural_order = jpeg_natural_order; - cinfo->lim_Se = DCTSIZE2-1; - break; - default: - ERREXIT4(cinfo, JERR_BAD_PROGRESSION, - cinfo->Ss, cinfo->Se, cinfo->Ah, cinfo->Al); - break; - } - - /* We initialize DCT_scaled_size and min_DCT_scaled_size to block_size. - * In the full decompressor, - * this will be overridden by jpeg_calc_output_dimensions in jdmaster.c; - * but in the transcoder, - * jpeg_calc_output_dimensions is not used, so we must do it here. - */ - cinfo->min_DCT_h_scaled_size = cinfo->block_size; - cinfo->min_DCT_v_scaled_size = cinfo->block_size; - - /* Compute dimensions of components */ - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - compptr->DCT_h_scaled_size = cinfo->block_size; - compptr->DCT_v_scaled_size = cinfo->block_size; - /* Size in DCT blocks */ - compptr->width_in_blocks = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * (long) compptr->h_samp_factor, - (long) (cinfo->max_h_samp_factor * cinfo->block_size)); - compptr->height_in_blocks = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * (long) compptr->v_samp_factor, - (long) (cinfo->max_v_samp_factor * cinfo->block_size)); - /* downsampled_width and downsampled_height will also be overridden by - * jdmaster.c if we are doing full decompression. The transcoder library - * doesn't use these values, but the calling application might. - */ - /* Size in samples */ - compptr->downsampled_width = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width * (long) compptr->h_samp_factor, - (long) cinfo->max_h_samp_factor); - compptr->downsampled_height = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height * (long) compptr->v_samp_factor, - (long) cinfo->max_v_samp_factor); - /* Mark component needed, until color conversion says otherwise */ - compptr->component_needed = TRUE; - /* Mark no quantization table yet saved for component */ - compptr->quant_table = NULL; - } - - /* Compute number of fully interleaved MCU rows. */ - cinfo->total_iMCU_rows = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height, - (long) (cinfo->max_v_samp_factor * cinfo->block_size)); - - /* Decide whether file contains multiple scans */ - if (cinfo->comps_in_scan < cinfo->num_components || cinfo->progressive_mode) - cinfo->inputctl->has_multiple_scans = TRUE; - else - cinfo->inputctl->has_multiple_scans = FALSE; -} - - -LOCAL(void) -per_scan_setup (j_decompress_ptr cinfo) -/* Do computations that are needed before processing a JPEG scan */ -/* cinfo->comps_in_scan and cinfo->cur_comp_info[] were set from SOS marker */ -{ - int ci, mcublks, tmp; - jpeg_component_info *compptr; - - if (cinfo->comps_in_scan == 1) { - - /* Noninterleaved (single-component) scan */ - compptr = cinfo->cur_comp_info[0]; - - /* Overall image size in MCUs */ - cinfo->MCUs_per_row = compptr->width_in_blocks; - cinfo->MCU_rows_in_scan = compptr->height_in_blocks; - - /* For noninterleaved scan, always one block per MCU */ - compptr->MCU_width = 1; - compptr->MCU_height = 1; - compptr->MCU_blocks = 1; - compptr->MCU_sample_width = compptr->DCT_h_scaled_size; - compptr->last_col_width = 1; - /* For noninterleaved scans, it is convenient to define last_row_height - * as the number of block rows present in the last iMCU row. - */ - tmp = (int) (compptr->height_in_blocks % compptr->v_samp_factor); - if (tmp == 0) tmp = compptr->v_samp_factor; - compptr->last_row_height = tmp; - - /* Prepare array describing MCU composition */ - cinfo->blocks_in_MCU = 1; - cinfo->MCU_membership[0] = 0; - - } else { - - /* Interleaved (multi-component) scan */ - if (cinfo->comps_in_scan <= 0 || cinfo->comps_in_scan > MAX_COMPS_IN_SCAN) - ERREXIT2(cinfo, JERR_COMPONENT_COUNT, cinfo->comps_in_scan, - MAX_COMPS_IN_SCAN); - - /* Overall image size in MCUs */ - cinfo->MCUs_per_row = (JDIMENSION) - jdiv_round_up((long) cinfo->image_width, - (long) (cinfo->max_h_samp_factor * cinfo->block_size)); - cinfo->MCU_rows_in_scan = (JDIMENSION) - jdiv_round_up((long) cinfo->image_height, - (long) (cinfo->max_v_samp_factor * cinfo->block_size)); - - cinfo->blocks_in_MCU = 0; - - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - compptr = cinfo->cur_comp_info[ci]; - /* Sampling factors give # of blocks of component in each MCU */ - compptr->MCU_width = compptr->h_samp_factor; - compptr->MCU_height = compptr->v_samp_factor; - compptr->MCU_blocks = compptr->MCU_width * compptr->MCU_height; - compptr->MCU_sample_width = compptr->MCU_width * compptr->DCT_h_scaled_size; - /* Figure number of non-dummy blocks in last MCU column & row */ - tmp = (int) (compptr->width_in_blocks % compptr->MCU_width); - if (tmp == 0) tmp = compptr->MCU_width; - compptr->last_col_width = tmp; - tmp = (int) (compptr->height_in_blocks % compptr->MCU_height); - if (tmp == 0) tmp = compptr->MCU_height; - compptr->last_row_height = tmp; - /* Prepare array describing MCU composition */ - mcublks = compptr->MCU_blocks; - if (cinfo->blocks_in_MCU + mcublks > D_MAX_BLOCKS_IN_MCU) - ERREXIT(cinfo, JERR_BAD_MCU_SIZE); - while (mcublks-- > 0) { - cinfo->MCU_membership[cinfo->blocks_in_MCU++] = ci; - } - } - - } -} - - -/* - * Save away a copy of the Q-table referenced by each component present - * in the current scan, unless already saved during a prior scan. - * - * In a multiple-scan JPEG file, the encoder could assign different components - * the same Q-table slot number, but change table definitions between scans - * so that each component uses a different Q-table. (The IJG encoder is not - * currently capable of doing this, but other encoders might.) Since we want - * to be able to dequantize all the components at the end of the file, this - * means that we have to save away the table actually used for each component. - * We do this by copying the table at the start of the first scan containing - * the component. - * The JPEG spec prohibits the encoder from changing the contents of a Q-table - * slot between scans of a component using that slot. If the encoder does so - * anyway, this decoder will simply use the Q-table values that were current - * at the start of the first scan for the component. - * - * The decompressor output side looks only at the saved quant tables, - * not at the current Q-table slots. - */ - -LOCAL(void) -latch_quant_tables (j_decompress_ptr cinfo) -{ - int ci, qtblno; - jpeg_component_info *compptr; - JQUANT_TBL * qtbl; - - for (ci = 0; ci < cinfo->comps_in_scan; ci++) { - compptr = cinfo->cur_comp_info[ci]; - /* No work if we already saved Q-table for this component */ - if (compptr->quant_table != NULL) - continue; - /* Make sure specified quantization table is present */ - qtblno = compptr->quant_tbl_no; - if (qtblno < 0 || qtblno >= NUM_QUANT_TBLS || - cinfo->quant_tbl_ptrs[qtblno] == NULL) - ERREXIT1(cinfo, JERR_NO_QUANT_TABLE, qtblno); - /* OK, save away the quantization table */ - qtbl = (JQUANT_TBL *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(JQUANT_TBL)); - MEMCOPY(qtbl, cinfo->quant_tbl_ptrs[qtblno], SIZEOF(JQUANT_TBL)); - compptr->quant_table = qtbl; - } -} - - -/* - * Initialize the input modules to read a scan of compressed data. - * The first call to this is done by jdmaster.c after initializing - * the entire decompressor (during jpeg_start_decompress). - * Subsequent calls come from consume_markers, below. - */ - -METHODDEF(void) -start_input_pass (j_decompress_ptr cinfo) -{ - per_scan_setup(cinfo); - latch_quant_tables(cinfo); - (*cinfo->entropy->start_pass) (cinfo); - (*cinfo->coef->start_input_pass) (cinfo); - cinfo->inputctl->consume_input = cinfo->coef->consume_data; -} - - -/* - * Finish up after inputting a compressed-data scan. - * This is called by the coefficient controller after it's read all - * the expected data of the scan. - */ - -METHODDEF(void) -finish_input_pass (j_decompress_ptr cinfo) -{ - cinfo->inputctl->consume_input = consume_markers; -} - - -/* - * Read JPEG markers before, between, or after compressed-data scans. - * Change state as necessary when a new scan is reached. - * Return value is JPEG_SUSPENDED, JPEG_REACHED_SOS, or JPEG_REACHED_EOI. - * - * The consume_input method pointer points either here or to the - * coefficient controller's consume_data routine, depending on whether - * we are reading a compressed data segment or inter-segment markers. - * - * Note: This function should NOT return a pseudo SOS marker (with zero - * component number) to the caller. A pseudo marker received by - * read_markers is processed and then skipped for other markers. - */ - -METHODDEF(int) -consume_markers (j_decompress_ptr cinfo) -{ - my_inputctl_ptr inputctl = (my_inputctl_ptr) cinfo->inputctl; - int val; - - if (inputctl->pub.eoi_reached) /* After hitting EOI, read no further */ - return JPEG_REACHED_EOI; - - for (;;) { /* Loop to pass pseudo SOS marker */ - val = (*cinfo->marker->read_markers) (cinfo); - - switch (val) { - case JPEG_REACHED_SOS: /* Found SOS */ - if (inputctl->inheaders) { /* 1st SOS */ - if (inputctl->inheaders == 1) - initial_setup(cinfo); - if (cinfo->comps_in_scan == 0) { /* pseudo SOS marker */ - inputctl->inheaders = 2; - break; - } - inputctl->inheaders = 0; - /* Note: start_input_pass must be called by jdmaster.c - * before any more input can be consumed. jdapimin.c is - * responsible for enforcing this sequencing. - */ - } else { /* 2nd or later SOS marker */ - if (! inputctl->pub.has_multiple_scans) - ERREXIT(cinfo, JERR_EOI_EXPECTED); /* Oops, I wasn't expecting this! */ - if (cinfo->comps_in_scan == 0) /* unexpected pseudo SOS marker */ - break; - start_input_pass(cinfo); - } - return val; - case JPEG_REACHED_EOI: /* Found EOI */ - inputctl->pub.eoi_reached = TRUE; - if (inputctl->inheaders) { /* Tables-only datastream, apparently */ - if (cinfo->marker->saw_SOF) - ERREXIT(cinfo, JERR_SOF_NO_SOS); - } else { - /* Prevent infinite loop in coef ctlr's decompress_data routine - * if user set output_scan_number larger than number of scans. - */ - if (cinfo->output_scan_number > cinfo->input_scan_number) - cinfo->output_scan_number = cinfo->input_scan_number; - } - return val; - case JPEG_SUSPENDED: - return val; - default: - return val; - } - } -} - - -/* - * Reset state to begin a fresh datastream. - */ - -METHODDEF(void) -reset_input_controller (j_decompress_ptr cinfo) -{ - my_inputctl_ptr inputctl = (my_inputctl_ptr) cinfo->inputctl; - - inputctl->pub.consume_input = consume_markers; - inputctl->pub.has_multiple_scans = FALSE; /* "unknown" would be better */ - inputctl->pub.eoi_reached = FALSE; - inputctl->inheaders = 1; - /* Reset other modules */ - (*cinfo->err->reset_error_mgr) ((j_common_ptr) cinfo); - (*cinfo->marker->reset_marker_reader) (cinfo); - /* Reset progression state -- would be cleaner if entropy decoder did this */ - cinfo->coef_bits = NULL; -} - - -/* - * Initialize the input controller module. - * This is called only once, when the decompression object is created. - */ - -GLOBAL(void) -jinit_input_controller (j_decompress_ptr cinfo) -{ - my_inputctl_ptr inputctl; - - /* Create subobject in permanent pool */ - inputctl = (my_inputctl_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_PERMANENT, - SIZEOF(my_input_controller)); - cinfo->inputctl = (struct jpeg_input_controller *) inputctl; - /* Initialize method pointers */ - inputctl->pub.consume_input = consume_markers; - inputctl->pub.reset_input_controller = reset_input_controller; - inputctl->pub.start_input_pass = start_input_pass; - inputctl->pub.finish_input_pass = finish_input_pass; - /* Initialize state: can't use reset_input_controller since we don't - * want to try to reset other modules yet. - */ - inputctl->pub.has_multiple_scans = FALSE; /* "unknown" would be better */ - inputctl->pub.eoi_reached = FALSE; - inputctl->inheaders = 1; -} diff --git a/src/3rdparty/libjpeg/jdmerge.c b/src/3rdparty/libjpeg/jdmerge.c deleted file mode 100644 index 37444468c2..0000000000 --- a/src/3rdparty/libjpeg/jdmerge.c +++ /dev/null @@ -1,400 +0,0 @@ -/* - * jdmerge.c - * - * Copyright (C) 1994-1996, Thomas G. Lane. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains code for merged upsampling/color conversion. - * - * This file combines functions from jdsample.c and jdcolor.c; - * read those files first to understand what's going on. - * - * When the chroma components are to be upsampled by simple replication - * (ie, box filtering), we can save some work in color conversion by - * calculating all the output pixels corresponding to a pair of chroma - * samples at one time. In the conversion equations - * R = Y + K1 * Cr - * G = Y + K2 * Cb + K3 * Cr - * B = Y + K4 * Cb - * only the Y term varies among the group of pixels corresponding to a pair - * of chroma samples, so the rest of the terms can be calculated just once. - * At typical sampling ratios, this eliminates half or three-quarters of the - * multiplications needed for color conversion. - * - * This file currently provides implementations for the following cases: - * YCbCr => RGB color conversion only. - * Sampling ratios of 2h1v or 2h2v. - * No scaling needed at upsample time. - * Corner-aligned (non-CCIR601) sampling alignment. - * Other special cases could be added, but in most applications these are - * the only common cases. (For uncommon cases we fall back on the more - * general code in jdsample.c and jdcolor.c.) - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - -#ifdef UPSAMPLE_MERGING_SUPPORTED - - -/* Private subobject */ - -typedef struct { - struct jpeg_upsampler pub; /* public fields */ - - /* Pointer to routine to do actual upsampling/conversion of one row group */ - JMETHOD(void, upmethod, (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr, - JSAMPARRAY output_buf)); - - /* Private state for YCC->RGB conversion */ - int * Cr_r_tab; /* => table for Cr to R conversion */ - int * Cb_b_tab; /* => table for Cb to B conversion */ - INT32 * Cr_g_tab; /* => table for Cr to G conversion */ - INT32 * Cb_g_tab; /* => table for Cb to G conversion */ - - /* For 2:1 vertical sampling, we produce two output rows at a time. - * We need a "spare" row buffer to hold the second output row if the - * application provides just a one-row buffer; we also use the spare - * to discard the dummy last row if the image height is odd. - */ - JSAMPROW spare_row; - boolean spare_full; /* T if spare buffer is occupied */ - - JDIMENSION out_row_width; /* samples per output row */ - JDIMENSION rows_to_go; /* counts rows remaining in image */ -} my_upsampler; - -typedef my_upsampler * my_upsample_ptr; - -#define SCALEBITS 16 /* speediest right-shift on some machines */ -#define ONE_HALF ((INT32) 1 << (SCALEBITS-1)) -#define FIX(x) ((INT32) ((x) * (1L<<SCALEBITS) + 0.5)) - - -/* - * Initialize tables for YCC->RGB colorspace conversion. - * This is taken directly from jdcolor.c; see that file for more info. - */ - -LOCAL(void) -build_ycc_rgb_table (j_decompress_ptr cinfo) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - int i; - INT32 x; - SHIFT_TEMPS - - upsample->Cr_r_tab = (int *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(int)); - upsample->Cb_b_tab = (int *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(int)); - upsample->Cr_g_tab = (INT32 *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(INT32)); - upsample->Cb_g_tab = (INT32 *) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (MAXJSAMPLE+1) * SIZEOF(INT32)); - - for (i = 0, x = -CENTERJSAMPLE; i <= MAXJSAMPLE; i++, x++) { - /* i is the actual input pixel value, in the range 0..MAXJSAMPLE */ - /* The Cb or Cr value we are thinking of is x = i - CENTERJSAMPLE */ - /* Cr=>R value is nearest int to 1.40200 * x */ - upsample->Cr_r_tab[i] = (int) - RIGHT_SHIFT(FIX(1.40200) * x + ONE_HALF, SCALEBITS); - /* Cb=>B value is nearest int to 1.77200 * x */ - upsample->Cb_b_tab[i] = (int) - RIGHT_SHIFT(FIX(1.77200) * x + ONE_HALF, SCALEBITS); - /* Cr=>G value is scaled-up -0.71414 * x */ - upsample->Cr_g_tab[i] = (- FIX(0.71414)) * x; - /* Cb=>G value is scaled-up -0.34414 * x */ - /* We also add in ONE_HALF so that need not do it in inner loop */ - upsample->Cb_g_tab[i] = (- FIX(0.34414)) * x + ONE_HALF; - } -} - - -/* - * Initialize for an upsampling pass. - */ - -METHODDEF(void) -start_pass_merged_upsample (j_decompress_ptr cinfo) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - - /* Mark the spare buffer empty */ - upsample->spare_full = FALSE; - /* Initialize total-height counter for detecting bottom of image */ - upsample->rows_to_go = cinfo->output_height; -} - - -/* - * Control routine to do upsampling (and color conversion). - * - * The control routine just handles the row buffering considerations. - */ - -METHODDEF(void) -merged_2v_upsample (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION *in_row_group_ctr, - JDIMENSION in_row_groups_avail, - JSAMPARRAY output_buf, JDIMENSION *out_row_ctr, - JDIMENSION out_rows_avail) -/* 2:1 vertical sampling case: may need a spare row. */ -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - JSAMPROW work_ptrs[2]; - JDIMENSION num_rows; /* number of rows returned to caller */ - - if (upsample->spare_full) { - /* If we have a spare row saved from a previous cycle, just return it. */ - jcopy_sample_rows(& upsample->spare_row, 0, output_buf + *out_row_ctr, 0, - 1, upsample->out_row_width); - num_rows = 1; - upsample->spare_full = FALSE; - } else { - /* Figure number of rows to return to caller. */ - num_rows = 2; - /* Not more than the distance to the end of the image. */ - if (num_rows > upsample->rows_to_go) - num_rows = upsample->rows_to_go; - /* And not more than what the client can accept: */ - out_rows_avail -= *out_row_ctr; - if (num_rows > out_rows_avail) - num_rows = out_rows_avail; - /* Create output pointer array for upsampler. */ - work_ptrs[0] = output_buf[*out_row_ctr]; - if (num_rows > 1) { - work_ptrs[1] = output_buf[*out_row_ctr + 1]; - } else { - work_ptrs[1] = upsample->spare_row; - upsample->spare_full = TRUE; - } - /* Now do the upsampling. */ - (*upsample->upmethod) (cinfo, input_buf, *in_row_group_ctr, work_ptrs); - } - - /* Adjust counts */ - *out_row_ctr += num_rows; - upsample->rows_to_go -= num_rows; - /* When the buffer is emptied, declare this input row group consumed */ - if (! upsample->spare_full) - (*in_row_group_ctr)++; -} - - -METHODDEF(void) -merged_1v_upsample (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION *in_row_group_ctr, - JDIMENSION in_row_groups_avail, - JSAMPARRAY output_buf, JDIMENSION *out_row_ctr, - JDIMENSION out_rows_avail) -/* 1:1 vertical sampling case: much easier, never need a spare row. */ -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - - /* Just do the upsampling. */ - (*upsample->upmethod) (cinfo, input_buf, *in_row_group_ctr, - output_buf + *out_row_ctr); - /* Adjust counts */ - (*out_row_ctr)++; - (*in_row_group_ctr)++; -} - - -/* - * These are the routines invoked by the control routines to do - * the actual upsampling/conversion. One row group is processed per call. - * - * Note: since we may be writing directly into application-supplied buffers, - * we have to be honest about the output width; we can't assume the buffer - * has been rounded up to an even width. - */ - - -/* - * Upsample and color convert for the case of 2:1 horizontal and 1:1 vertical. - */ - -METHODDEF(void) -h2v1_merged_upsample (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr, - JSAMPARRAY output_buf) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - register int y, cred, cgreen, cblue; - int cb, cr; - register JSAMPROW outptr; - JSAMPROW inptr0, inptr1, inptr2; - JDIMENSION col; - /* copy these pointers into registers if possible */ - register JSAMPLE * range_limit = cinfo->sample_range_limit; - int * Crrtab = upsample->Cr_r_tab; - int * Cbbtab = upsample->Cb_b_tab; - INT32 * Crgtab = upsample->Cr_g_tab; - INT32 * Cbgtab = upsample->Cb_g_tab; - SHIFT_TEMPS - - inptr0 = input_buf[0][in_row_group_ctr]; - inptr1 = input_buf[1][in_row_group_ctr]; - inptr2 = input_buf[2][in_row_group_ctr]; - outptr = output_buf[0]; - /* Loop for each pair of output pixels */ - for (col = cinfo->output_width >> 1; col > 0; col--) { - /* Do the chroma part of the calculation */ - cb = GETJSAMPLE(*inptr1++); - cr = GETJSAMPLE(*inptr2++); - cred = Crrtab[cr]; - cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS); - cblue = Cbbtab[cb]; - /* Fetch 2 Y values and emit 2 pixels */ - y = GETJSAMPLE(*inptr0++); - outptr[RGB_RED] = range_limit[y + cred]; - outptr[RGB_GREEN] = range_limit[y + cgreen]; - outptr[RGB_BLUE] = range_limit[y + cblue]; - outptr += RGB_PIXELSIZE; - y = GETJSAMPLE(*inptr0++); - outptr[RGB_RED] = range_limit[y + cred]; - outptr[RGB_GREEN] = range_limit[y + cgreen]; - outptr[RGB_BLUE] = range_limit[y + cblue]; - outptr += RGB_PIXELSIZE; - } - /* If image width is odd, do the last output column separately */ - if (cinfo->output_width & 1) { - cb = GETJSAMPLE(*inptr1); - cr = GETJSAMPLE(*inptr2); - cred = Crrtab[cr]; - cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS); - cblue = Cbbtab[cb]; - y = GETJSAMPLE(*inptr0); - outptr[RGB_RED] = range_limit[y + cred]; - outptr[RGB_GREEN] = range_limit[y + cgreen]; - outptr[RGB_BLUE] = range_limit[y + cblue]; - } -} - - -/* - * Upsample and color convert for the case of 2:1 horizontal and 2:1 vertical. - */ - -METHODDEF(void) -h2v2_merged_upsample (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr, - JSAMPARRAY output_buf) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - register int y, cred, cgreen, cblue; - int cb, cr; - register JSAMPROW outptr0, outptr1; - JSAMPROW inptr00, inptr01, inptr1, inptr2; - JDIMENSION col; - /* copy these pointers into registers if possible */ - register JSAMPLE * range_limit = cinfo->sample_range_limit; - int * Crrtab = upsample->Cr_r_tab; - int * Cbbtab = upsample->Cb_b_tab; - INT32 * Crgtab = upsample->Cr_g_tab; - INT32 * Cbgtab = upsample->Cb_g_tab; - SHIFT_TEMPS - - inptr00 = input_buf[0][in_row_group_ctr*2]; - inptr01 = input_buf[0][in_row_group_ctr*2 + 1]; - inptr1 = input_buf[1][in_row_group_ctr]; - inptr2 = input_buf[2][in_row_group_ctr]; - outptr0 = output_buf[0]; - outptr1 = output_buf[1]; - /* Loop for each group of output pixels */ - for (col = cinfo->output_width >> 1; col > 0; col--) { - /* Do the chroma part of the calculation */ - cb = GETJSAMPLE(*inptr1++); - cr = GETJSAMPLE(*inptr2++); - cred = Crrtab[cr]; - cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS); - cblue = Cbbtab[cb]; - /* Fetch 4 Y values and emit 4 pixels */ - y = GETJSAMPLE(*inptr00++); - outptr0[RGB_RED] = range_limit[y + cred]; - outptr0[RGB_GREEN] = range_limit[y + cgreen]; - outptr0[RGB_BLUE] = range_limit[y + cblue]; - outptr0 += RGB_PIXELSIZE; - y = GETJSAMPLE(*inptr00++); - outptr0[RGB_RED] = range_limit[y + cred]; - outptr0[RGB_GREEN] = range_limit[y + cgreen]; - outptr0[RGB_BLUE] = range_limit[y + cblue]; - outptr0 += RGB_PIXELSIZE; - y = GETJSAMPLE(*inptr01++); - outptr1[RGB_RED] = range_limit[y + cred]; - outptr1[RGB_GREEN] = range_limit[y + cgreen]; - outptr1[RGB_BLUE] = range_limit[y + cblue]; - outptr1 += RGB_PIXELSIZE; - y = GETJSAMPLE(*inptr01++); - outptr1[RGB_RED] = range_limit[y + cred]; - outptr1[RGB_GREEN] = range_limit[y + cgreen]; - outptr1[RGB_BLUE] = range_limit[y + cblue]; - outptr1 += RGB_PIXELSIZE; - } - /* If image width is odd, do the last output column separately */ - if (cinfo->output_width & 1) { - cb = GETJSAMPLE(*inptr1); - cr = GETJSAMPLE(*inptr2); - cred = Crrtab[cr]; - cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS); - cblue = Cbbtab[cb]; - y = GETJSAMPLE(*inptr00); - outptr0[RGB_RED] = range_limit[y + cred]; - outptr0[RGB_GREEN] = range_limit[y + cgreen]; - outptr0[RGB_BLUE] = range_limit[y + cblue]; - y = GETJSAMPLE(*inptr01); - outptr1[RGB_RED] = range_limit[y + cred]; - outptr1[RGB_GREEN] = range_limit[y + cgreen]; - outptr1[RGB_BLUE] = range_limit[y + cblue]; - } -} - - -/* - * Module initialization routine for merged upsampling/color conversion. - * - * NB: this is called under the conditions determined by use_merged_upsample() - * in jdmaster.c. That routine MUST correspond to the actual capabilities - * of this module; no safety checks are made here. - */ - -GLOBAL(void) -jinit_merged_upsampler (j_decompress_ptr cinfo) -{ - my_upsample_ptr upsample; - - upsample = (my_upsample_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_upsampler)); - cinfo->upsample = (struct jpeg_upsampler *) upsample; - upsample->pub.start_pass = start_pass_merged_upsample; - upsample->pub.need_context_rows = FALSE; - - upsample->out_row_width = cinfo->output_width * cinfo->out_color_components; - - if (cinfo->max_v_samp_factor == 2) { - upsample->pub.upsample = merged_2v_upsample; - upsample->upmethod = h2v2_merged_upsample; - /* Allocate a spare row buffer */ - upsample->spare_row = (JSAMPROW) - (*cinfo->mem->alloc_large) ((j_common_ptr) cinfo, JPOOL_IMAGE, - (size_t) (upsample->out_row_width * SIZEOF(JSAMPLE))); - } else { - upsample->pub.upsample = merged_1v_upsample; - upsample->upmethod = h2v1_merged_upsample; - /* No spare row needed */ - upsample->spare_row = NULL; - } - - build_ycc_rgb_table(cinfo); -} - -#endif /* UPSAMPLE_MERGING_SUPPORTED */ diff --git a/src/3rdparty/libjpeg/jdsample.c b/src/3rdparty/libjpeg/jdsample.c deleted file mode 100644 index 7bc8885b02..0000000000 --- a/src/3rdparty/libjpeg/jdsample.c +++ /dev/null @@ -1,361 +0,0 @@ -/* - * jdsample.c - * - * Copyright (C) 1991-1996, Thomas G. Lane. - * Modified 2002-2008 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains upsampling routines. - * - * Upsampling input data is counted in "row groups". A row group - * is defined to be (v_samp_factor * DCT_v_scaled_size / min_DCT_v_scaled_size) - * sample rows of each component. Upsampling will normally produce - * max_v_samp_factor pixel rows from each row group (but this could vary - * if the upsampler is applying a scale factor of its own). - * - * An excellent reference for image resampling is - * Digital Image Warping, George Wolberg, 1990. - * Pub. by IEEE Computer Society Press, Los Alamitos, CA. ISBN 0-8186-8944-7. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" - - -/* Pointer to routine to upsample a single component */ -typedef JMETHOD(void, upsample1_ptr, - (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY input_data, JSAMPARRAY * output_data_ptr)); - -/* Private subobject */ - -typedef struct { - struct jpeg_upsampler pub; /* public fields */ - - /* Color conversion buffer. When using separate upsampling and color - * conversion steps, this buffer holds one upsampled row group until it - * has been color converted and output. - * Note: we do not allocate any storage for component(s) which are full-size, - * ie do not need rescaling. The corresponding entry of color_buf[] is - * simply set to point to the input data array, thereby avoiding copying. - */ - JSAMPARRAY color_buf[MAX_COMPONENTS]; - - /* Per-component upsampling method pointers */ - upsample1_ptr methods[MAX_COMPONENTS]; - - int next_row_out; /* counts rows emitted from color_buf */ - JDIMENSION rows_to_go; /* counts rows remaining in image */ - - /* Height of an input row group for each component. */ - int rowgroup_height[MAX_COMPONENTS]; - - /* These arrays save pixel expansion factors so that int_expand need not - * recompute them each time. They are unused for other upsampling methods. - */ - UINT8 h_expand[MAX_COMPONENTS]; - UINT8 v_expand[MAX_COMPONENTS]; -} my_upsampler; - -typedef my_upsampler * my_upsample_ptr; - - -/* - * Initialize for an upsampling pass. - */ - -METHODDEF(void) -start_pass_upsample (j_decompress_ptr cinfo) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - - /* Mark the conversion buffer empty */ - upsample->next_row_out = cinfo->max_v_samp_factor; - /* Initialize total-height counter for detecting bottom of image */ - upsample->rows_to_go = cinfo->output_height; -} - - -/* - * Control routine to do upsampling (and color conversion). - * - * In this version we upsample each component independently. - * We upsample one row group into the conversion buffer, then apply - * color conversion a row at a time. - */ - -METHODDEF(void) -sep_upsample (j_decompress_ptr cinfo, - JSAMPIMAGE input_buf, JDIMENSION *in_row_group_ctr, - JDIMENSION in_row_groups_avail, - JSAMPARRAY output_buf, JDIMENSION *out_row_ctr, - JDIMENSION out_rows_avail) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - int ci; - jpeg_component_info * compptr; - JDIMENSION num_rows; - - /* Fill the conversion buffer, if it's empty */ - if (upsample->next_row_out >= cinfo->max_v_samp_factor) { - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - /* Invoke per-component upsample method. Notice we pass a POINTER - * to color_buf[ci], so that fullsize_upsample can change it. - */ - (*upsample->methods[ci]) (cinfo, compptr, - input_buf[ci] + (*in_row_group_ctr * upsample->rowgroup_height[ci]), - upsample->color_buf + ci); - } - upsample->next_row_out = 0; - } - - /* Color-convert and emit rows */ - - /* How many we have in the buffer: */ - num_rows = (JDIMENSION) (cinfo->max_v_samp_factor - upsample->next_row_out); - /* Not more than the distance to the end of the image. Need this test - * in case the image height is not a multiple of max_v_samp_factor: - */ - if (num_rows > upsample->rows_to_go) - num_rows = upsample->rows_to_go; - /* And not more than what the client can accept: */ - out_rows_avail -= *out_row_ctr; - if (num_rows > out_rows_avail) - num_rows = out_rows_avail; - - (*cinfo->cconvert->color_convert) (cinfo, upsample->color_buf, - (JDIMENSION) upsample->next_row_out, - output_buf + *out_row_ctr, - (int) num_rows); - - /* Adjust counts */ - *out_row_ctr += num_rows; - upsample->rows_to_go -= num_rows; - upsample->next_row_out += num_rows; - /* When the buffer is emptied, declare this input row group consumed */ - if (upsample->next_row_out >= cinfo->max_v_samp_factor) - (*in_row_group_ctr)++; -} - - -/* - * These are the routines invoked by sep_upsample to upsample pixel values - * of a single component. One row group is processed per call. - */ - - -/* - * For full-size components, we just make color_buf[ci] point at the - * input buffer, and thus avoid copying any data. Note that this is - * safe only because sep_upsample doesn't declare the input row group - * "consumed" until we are done color converting and emitting it. - */ - -METHODDEF(void) -fullsize_upsample (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY input_data, JSAMPARRAY * output_data_ptr) -{ - *output_data_ptr = input_data; -} - - -/* - * This is a no-op version used for "uninteresting" components. - * These components will not be referenced by color conversion. - */ - -METHODDEF(void) -noop_upsample (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY input_data, JSAMPARRAY * output_data_ptr) -{ - *output_data_ptr = NULL; /* safety check */ -} - - -/* - * This version handles any integral sampling ratios. - * This is not used for typical JPEG files, so it need not be fast. - * Nor, for that matter, is it particularly accurate: the algorithm is - * simple replication of the input pixel onto the corresponding output - * pixels. The hi-falutin sampling literature refers to this as a - * "box filter". A box filter tends to introduce visible artifacts, - * so if you are actually going to use 3:1 or 4:1 sampling ratios - * you would be well advised to improve this code. - */ - -METHODDEF(void) -int_upsample (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY input_data, JSAMPARRAY * output_data_ptr) -{ - my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample; - JSAMPARRAY output_data = *output_data_ptr; - register JSAMPROW inptr, outptr; - register JSAMPLE invalue; - register int h; - JSAMPROW outend; - int h_expand, v_expand; - int inrow, outrow; - - h_expand = upsample->h_expand[compptr->component_index]; - v_expand = upsample->v_expand[compptr->component_index]; - - inrow = outrow = 0; - while (outrow < cinfo->max_v_samp_factor) { - /* Generate one output row with proper horizontal expansion */ - inptr = input_data[inrow]; - outptr = output_data[outrow]; - outend = outptr + cinfo->output_width; - while (outptr < outend) { - invalue = *inptr++; /* don't need GETJSAMPLE() here */ - for (h = h_expand; h > 0; h--) { - *outptr++ = invalue; - } - } - /* Generate any additional output rows by duplicating the first one */ - if (v_expand > 1) { - jcopy_sample_rows(output_data, outrow, output_data, outrow+1, - v_expand-1, cinfo->output_width); - } - inrow++; - outrow += v_expand; - } -} - - -/* - * Fast processing for the common case of 2:1 horizontal and 1:1 vertical. - * It's still a box filter. - */ - -METHODDEF(void) -h2v1_upsample (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY input_data, JSAMPARRAY * output_data_ptr) -{ - JSAMPARRAY output_data = *output_data_ptr; - register JSAMPROW inptr, outptr; - register JSAMPLE invalue; - JSAMPROW outend; - int outrow; - - for (outrow = 0; outrow < cinfo->max_v_samp_factor; outrow++) { - inptr = input_data[outrow]; - outptr = output_data[outrow]; - outend = outptr + cinfo->output_width; - while (outptr < outend) { - invalue = *inptr++; /* don't need GETJSAMPLE() here */ - *outptr++ = invalue; - *outptr++ = invalue; - } - } -} - - -/* - * Fast processing for the common case of 2:1 horizontal and 2:1 vertical. - * It's still a box filter. - */ - -METHODDEF(void) -h2v2_upsample (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JSAMPARRAY input_data, JSAMPARRAY * output_data_ptr) -{ - JSAMPARRAY output_data = *output_data_ptr; - register JSAMPROW inptr, outptr; - register JSAMPLE invalue; - JSAMPROW outend; - int inrow, outrow; - - inrow = outrow = 0; - while (outrow < cinfo->max_v_samp_factor) { - inptr = input_data[inrow]; - outptr = output_data[outrow]; - outend = outptr + cinfo->output_width; - while (outptr < outend) { - invalue = *inptr++; /* don't need GETJSAMPLE() here */ - *outptr++ = invalue; - *outptr++ = invalue; - } - jcopy_sample_rows(output_data, outrow, output_data, outrow+1, - 1, cinfo->output_width); - inrow++; - outrow += 2; - } -} - - -/* - * Module initialization routine for upsampling. - */ - -GLOBAL(void) -jinit_upsampler (j_decompress_ptr cinfo) -{ - my_upsample_ptr upsample; - int ci; - jpeg_component_info * compptr; - boolean need_buffer; - int h_in_group, v_in_group, h_out_group, v_out_group; - - upsample = (my_upsample_ptr) - (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE, - SIZEOF(my_upsampler)); - cinfo->upsample = (struct jpeg_upsampler *) upsample; - upsample->pub.start_pass = start_pass_upsample; - upsample->pub.upsample = sep_upsample; - upsample->pub.need_context_rows = FALSE; /* until we find out differently */ - - if (cinfo->CCIR601_sampling) /* this isn't supported */ - ERREXIT(cinfo, JERR_CCIR601_NOTIMPL); - - /* Verify we can handle the sampling factors, select per-component methods, - * and create storage as needed. - */ - for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components; - ci++, compptr++) { - /* Compute size of an "input group" after IDCT scaling. This many samples - * are to be converted to max_h_samp_factor * max_v_samp_factor pixels. - */ - h_in_group = (compptr->h_samp_factor * compptr->DCT_h_scaled_size) / - cinfo->min_DCT_h_scaled_size; - v_in_group = (compptr->v_samp_factor * compptr->DCT_v_scaled_size) / - cinfo->min_DCT_v_scaled_size; - h_out_group = cinfo->max_h_samp_factor; - v_out_group = cinfo->max_v_samp_factor; - upsample->rowgroup_height[ci] = v_in_group; /* save for use later */ - need_buffer = TRUE; - if (! compptr->component_needed) { - /* Don't bother to upsample an uninteresting component. */ - upsample->methods[ci] = noop_upsample; - need_buffer = FALSE; - } else if (h_in_group == h_out_group && v_in_group == v_out_group) { - /* Fullsize components can be processed without any work. */ - upsample->methods[ci] = fullsize_upsample; - need_buffer = FALSE; - } else if (h_in_group * 2 == h_out_group && - v_in_group == v_out_group) { - /* Special case for 2h1v upsampling */ - upsample->methods[ci] = h2v1_upsample; - } else if (h_in_group * 2 == h_out_group && - v_in_group * 2 == v_out_group) { - /* Special case for 2h2v upsampling */ - upsample->methods[ci] = h2v2_upsample; - } else if ((h_out_group % h_in_group) == 0 && - (v_out_group % v_in_group) == 0) { - /* Generic integral-factors upsampling method */ - upsample->methods[ci] = int_upsample; - upsample->h_expand[ci] = (UINT8) (h_out_group / h_in_group); - upsample->v_expand[ci] = (UINT8) (v_out_group / v_in_group); - } else - ERREXIT(cinfo, JERR_FRACT_SAMPLE_NOTIMPL); - if (need_buffer) { - upsample->color_buf[ci] = (*cinfo->mem->alloc_sarray) - ((j_common_ptr) cinfo, JPOOL_IMAGE, - (JDIMENSION) jround_up((long) cinfo->output_width, - (long) cinfo->max_h_samp_factor), - (JDIMENSION) cinfo->max_v_samp_factor); - } - } -} diff --git a/src/3rdparty/libjpeg/jfdctint.c b/src/3rdparty/libjpeg/jfdctint.c deleted file mode 100644 index 1dde58c499..0000000000 --- a/src/3rdparty/libjpeg/jfdctint.c +++ /dev/null @@ -1,4348 +0,0 @@ -/* - * jfdctint.c - * - * Copyright (C) 1991-1996, Thomas G. Lane. - * Modification developed 2003-2009 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains a slow-but-accurate integer implementation of the - * forward DCT (Discrete Cosine Transform). - * - * A 2-D DCT can be done by 1-D DCT on each row followed by 1-D DCT - * on each column. 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 an algorithm described in - * C. Loeffler, A. Ligtenberg and G. Moschytz, "Practical Fast 1-D DCT - * Algorithms with 11 Multiplications", Proc. Int'l. Conf. on Acoustics, - * Speech, and Signal Processing 1989 (ICASSP '89), pp. 988-991. - * The primary algorithm described there uses 11 multiplies and 29 adds. - * We use their alternate method with 12 multiplies and 32 adds. - * The advantage of this method is that no data path contains more than one - * multiplication; this allows a very simple and accurate implementation in - * scaled fixed-point arithmetic, with a minimal number of shifts. - * - * We also provide FDCT routines with various input sample block sizes for - * direct resolution reduction or enlargement and for direct resolving the - * common 2x1 and 1x2 subsampling cases without additional resampling: NxN - * (N=1...16), 2NxN, and Nx2N (N=1...8) pixels for one 8x8 output DCT block. - * - * For N<8 we fill the remaining block coefficients with zero. - * For N>8 we apply a partial N-point FDCT on the input samples, computing - * just the lower 8 frequency coefficients and discarding the rest. - * - * We must scale the output coefficients of the N-point FDCT appropriately - * to the standard 8-point FDCT level by 8/N per 1-D pass. This scaling - * is folded into the constant multipliers (pass 2) and/or final/initial - * shifting. - * - * CAUTION: We rely on the FIX() macro except for the N=1,2,4,8 cases - * since there would be too many additional constants to pre-calculate. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" -#include "jdct.h" /* Private declarations for DCT subsystem */ - -#ifdef DCT_ISLOW_SUPPORTED - - -/* - * This module is specialized to the case DCTSIZE = 8. - */ - -#if DCTSIZE != 8 - Sorry, this code only copes with 8x8 DCT blocks. /* deliberate syntax err */ -#endif - - -/* - * The poop on this scaling stuff is as follows: - * - * Each 1-D DCT step produces outputs which are a factor of sqrt(N) - * larger than the true DCT outputs. The final outputs are therefore - * a factor of N larger than desired; since N=8 this can be cured by - * a simple right shift at the end of the algorithm. The advantage of - * this arrangement is that we save two multiplications per 1-D DCT, - * because the y0 and y4 outputs need not be divided by sqrt(N). - * In the IJG code, this factor of 8 is removed by the quantization step - * (in jcdctmgr.c), NOT in this module. - * - * We have to do addition and subtraction of the integer inputs, which - * is no problem, and multiplication by fractional constants, which is - * a problem to do in integer arithmetic. We multiply all the constants - * by CONST_SCALE and convert them to integer constants (thus retaining - * CONST_BITS bits of precision in the constants). After doing a - * multiplication we have to divide the product by CONST_SCALE, with proper - * rounding, to produce the correct output. This division can be done - * cheaply as a right shift of CONST_BITS bits. We postpone shifting - * as long as possible so that partial sums can be added together with - * full fractional precision. - * - * The outputs of the first pass are scaled up by PASS1_BITS bits so that - * they are represented to better-than-integral precision. These outputs - * require BITS_IN_JSAMPLE + PASS1_BITS + 3 bits; this fits in a 16-bit word - * with the recommended scaling. (For 12-bit sample data, the intermediate - * array is INT32 anyway.) - * - * To avoid overflow of the 32-bit intermediate results in pass 2, we must - * have BITS_IN_JSAMPLE + CONST_BITS + PASS1_BITS <= 26. Error analysis - * shows that the values given below are the most effective. - */ - -#if BITS_IN_JSAMPLE == 8 -#define CONST_BITS 13 -#define PASS1_BITS 2 -#else -#define CONST_BITS 13 -#define PASS1_BITS 1 /* lose a little precision to avoid overflow */ -#endif - -/* Some C compilers fail to reduce "FIX(constant)" at compile time, thus - * causing a lot of useless floating-point operations at run time. - * To get around this we use the following pre-calculated constants. - * If you change CONST_BITS you may want to add appropriate values. - * (With a reasonable C compiler, you can just rely on the FIX() macro...) - */ - -#if CONST_BITS == 13 -#define FIX_0_298631336 ((INT32) 2446) /* FIX(0.298631336) */ -#define FIX_0_390180644 ((INT32) 3196) /* FIX(0.390180644) */ -#define FIX_0_541196100 ((INT32) 4433) /* FIX(0.541196100) */ -#define FIX_0_765366865 ((INT32) 6270) /* FIX(0.765366865) */ -#define FIX_0_899976223 ((INT32) 7373) /* FIX(0.899976223) */ -#define FIX_1_175875602 ((INT32) 9633) /* FIX(1.175875602) */ -#define FIX_1_501321110 ((INT32) 12299) /* FIX(1.501321110) */ -#define FIX_1_847759065 ((INT32) 15137) /* FIX(1.847759065) */ -#define FIX_1_961570560 ((INT32) 16069) /* FIX(1.961570560) */ -#define FIX_2_053119869 ((INT32) 16819) /* FIX(2.053119869) */ -#define FIX_2_562915447 ((INT32) 20995) /* FIX(2.562915447) */ -#define FIX_3_072711026 ((INT32) 25172) /* FIX(3.072711026) */ -#else -#define FIX_0_298631336 FIX(0.298631336) -#define FIX_0_390180644 FIX(0.390180644) -#define FIX_0_541196100 FIX(0.541196100) -#define FIX_0_765366865 FIX(0.765366865) -#define FIX_0_899976223 FIX(0.899976223) -#define FIX_1_175875602 FIX(1.175875602) -#define FIX_1_501321110 FIX(1.501321110) -#define FIX_1_847759065 FIX(1.847759065) -#define FIX_1_961570560 FIX(1.961570560) -#define FIX_2_053119869 FIX(2.053119869) -#define FIX_2_562915447 FIX(2.562915447) -#define FIX_3_072711026 FIX(3.072711026) -#endif - - -/* Multiply an INT32 variable by an INT32 constant to yield an INT32 result. - * For 8-bit samples with the recommended scaling, all the variable - * and constant values involved are no more than 16 bits wide, so a - * 16x16->32 bit multiply can be used instead of a full 32x32 multiply. - * For 12-bit samples, a full 32-bit multiplication will be needed. - */ - -#if BITS_IN_JSAMPLE == 8 -#define MULTIPLY(var,const) MULTIPLY16C16(var,const) -#else -#define MULTIPLY(var,const) ((var) * (const)) -#endif - - -/* - * Perform the forward DCT on one block of samples. - */ - -GLOBAL(void) -jpeg_fdct_islow (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3; - INT32 tmp10, tmp11, tmp12, tmp13; - INT32 z1; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - - dataptr = data; - for (ctr = 0; ctr < DCTSIZE; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part per LL&M figure 1 --- note that published figure is faulty; - * rotator "sqrt(2)*c1" should be "sqrt(2)*c6". - */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[7]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[6]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[5]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[4]); - - tmp10 = tmp0 + tmp3; - tmp12 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp13 = tmp1 - tmp2; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[7]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[6]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[5]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[4]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) ((tmp10 + tmp11 - 8 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[4] = (DCTELEM) ((tmp10 - tmp11) << PASS1_BITS); - - z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-1); - dataptr[2] = (DCTELEM) RIGHT_SHIFT(z1 + MULTIPLY(tmp12, FIX_0_765366865), - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) RIGHT_SHIFT(z1 - MULTIPLY(tmp13, FIX_1_847759065), - CONST_BITS-PASS1_BITS); - - /* Odd part per figure 8 --- note paper omits factor of sqrt(2). - * cK represents sqrt(2) * cos(K*pi/16). - * i0..i3 in the paper are tmp0..tmp3 here. - */ - - tmp10 = tmp0 + tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp0 + tmp2; - tmp13 = tmp1 + tmp3; - z1 = MULTIPLY(tmp12 + tmp13, FIX_1_175875602); /* c3 */ - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-1); - - tmp0 = MULTIPLY(tmp0, FIX_1_501321110); /* c1+c3-c5-c7 */ - tmp1 = MULTIPLY(tmp1, FIX_3_072711026); /* c1+c3+c5-c7 */ - tmp2 = MULTIPLY(tmp2, FIX_2_053119869); /* c1+c3-c5+c7 */ - tmp3 = MULTIPLY(tmp3, FIX_0_298631336); /* -c1+c3+c5-c7 */ - tmp10 = MULTIPLY(tmp10, - FIX_0_899976223); /* c7-c3 */ - tmp11 = MULTIPLY(tmp11, - FIX_2_562915447); /* -c1-c3 */ - tmp12 = MULTIPLY(tmp12, - FIX_0_390180644); /* c5-c3 */ - tmp13 = MULTIPLY(tmp13, - FIX_1_961570560); /* -c3-c5 */ - - tmp12 += z1; - tmp13 += z1; - - dataptr[1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + tmp10 + tmp12, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) - RIGHT_SHIFT(tmp1 + tmp11 + tmp13, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) - RIGHT_SHIFT(tmp2 + tmp11 + tmp12, CONST_BITS-PASS1_BITS); - dataptr[7] = (DCTELEM) - RIGHT_SHIFT(tmp3 + tmp10 + tmp13, CONST_BITS-PASS1_BITS); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - */ - - dataptr = data; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part per LL&M figure 1 --- note that published figure is faulty; - * rotator "sqrt(2)*c1" should be "sqrt(2)*c6". - */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*4]; - - /* Add fudge factor here for final descale. */ - tmp10 = tmp0 + tmp3 + (ONE << (PASS1_BITS-1)); - tmp12 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp13 = tmp1 - tmp2; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*4]; - - dataptr[DCTSIZE*0] = (DCTELEM) RIGHT_SHIFT(tmp10 + tmp11, PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) RIGHT_SHIFT(tmp10 - tmp11, PASS1_BITS); - - z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS+PASS1_BITS-1); - dataptr[DCTSIZE*2] = (DCTELEM) - RIGHT_SHIFT(z1 + MULTIPLY(tmp12, FIX_0_765366865), CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*6] = (DCTELEM) - RIGHT_SHIFT(z1 - MULTIPLY(tmp13, FIX_1_847759065), CONST_BITS+PASS1_BITS); - - /* Odd part per figure 8 --- note paper omits factor of sqrt(2). - * cK represents sqrt(2) * cos(K*pi/16). - * i0..i3 in the paper are tmp0..tmp3 here. - */ - - tmp10 = tmp0 + tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp0 + tmp2; - tmp13 = tmp1 + tmp3; - z1 = MULTIPLY(tmp12 + tmp13, FIX_1_175875602); /* c3 */ - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS+PASS1_BITS-1); - - tmp0 = MULTIPLY(tmp0, FIX_1_501321110); /* c1+c3-c5-c7 */ - tmp1 = MULTIPLY(tmp1, FIX_3_072711026); /* c1+c3+c5-c7 */ - tmp2 = MULTIPLY(tmp2, FIX_2_053119869); /* c1+c3-c5+c7 */ - tmp3 = MULTIPLY(tmp3, FIX_0_298631336); /* -c1+c3+c5-c7 */ - tmp10 = MULTIPLY(tmp10, - FIX_0_899976223); /* c7-c3 */ - tmp11 = MULTIPLY(tmp11, - FIX_2_562915447); /* -c1-c3 */ - tmp12 = MULTIPLY(tmp12, - FIX_0_390180644); /* c5-c3 */ - tmp13 = MULTIPLY(tmp13, - FIX_1_961570560); /* -c3-c5 */ - - tmp12 += z1; - tmp13 += z1; - - dataptr[DCTSIZE*1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + tmp10 + tmp12, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - RIGHT_SHIFT(tmp1 + tmp11 + tmp13, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*5] = (DCTELEM) - RIGHT_SHIFT(tmp2 + tmp11 + tmp12, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*7] = (DCTELEM) - RIGHT_SHIFT(tmp3 + tmp10 + tmp13, CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - -#ifdef DCT_SCALING_SUPPORTED - - -/* - * Perform the forward DCT on a 7x7 sample block. - */ - -GLOBAL(void) -jpeg_fdct_7x7 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3; - INT32 tmp10, tmp11, tmp12; - INT32 z1, z2, z3; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* cK represents sqrt(2) * cos(K*pi/14). */ - - dataptr = data; - for (ctr = 0; ctr < 7; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[6]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[5]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[4]); - tmp3 = GETJSAMPLE(elemptr[3]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[6]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[5]); - tmp12 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[4]); - - z1 = tmp0 + tmp2; - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((z1 + tmp1 + tmp3 - 7 * CENTERJSAMPLE) << PASS1_BITS); - tmp3 += tmp3; - z1 -= tmp3; - z1 -= tmp3; - z1 = MULTIPLY(z1, FIX(0.353553391)); /* (c2+c6-c4)/2 */ - z2 = MULTIPLY(tmp0 - tmp2, FIX(0.920609002)); /* (c2+c4-c6)/2 */ - z3 = MULTIPLY(tmp1 - tmp2, FIX(0.314692123)); /* c6 */ - dataptr[2] = (DCTELEM) DESCALE(z1 + z2 + z3, CONST_BITS-PASS1_BITS); - z1 -= z2; - z2 = MULTIPLY(tmp0 - tmp1, FIX(0.881747734)); /* c4 */ - dataptr[4] = (DCTELEM) - DESCALE(z2 + z3 - MULTIPLY(tmp1 - tmp3, FIX(0.707106781)), /* c2+c6-c4 */ - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) DESCALE(z1 + z2, CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(0.935414347)); /* (c3+c1-c5)/2 */ - tmp2 = MULTIPLY(tmp10 - tmp11, FIX(0.170262339)); /* (c3+c5-c1)/2 */ - tmp0 = tmp1 - tmp2; - tmp1 += tmp2; - tmp2 = MULTIPLY(tmp11 + tmp12, - FIX(1.378756276)); /* -c1 */ - tmp1 += tmp2; - tmp3 = MULTIPLY(tmp10 + tmp12, FIX(0.613604268)); /* c5 */ - tmp0 += tmp3; - tmp2 += tmp3 + MULTIPLY(tmp12, FIX(1.870828693)); /* c3+c1-c5 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp0, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp1, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp2, CONST_BITS-PASS1_BITS); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/7)**2 = 64/49, which we fold - * into the constant multipliers: - * cK now represents sqrt(2) * cos(K*pi/14) * 64/49. - */ - - dataptr = data; - for (ctr = 0; ctr < 7; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*6]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*5]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*4]; - tmp3 = dataptr[DCTSIZE*3]; - - tmp10 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*6]; - tmp11 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*5]; - tmp12 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*4]; - - z1 = tmp0 + tmp2; - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(z1 + tmp1 + tmp3, FIX(1.306122449)), /* 64/49 */ - CONST_BITS+PASS1_BITS); - tmp3 += tmp3; - z1 -= tmp3; - z1 -= tmp3; - z1 = MULTIPLY(z1, FIX(0.461784020)); /* (c2+c6-c4)/2 */ - z2 = MULTIPLY(tmp0 - tmp2, FIX(1.202428084)); /* (c2+c4-c6)/2 */ - z3 = MULTIPLY(tmp1 - tmp2, FIX(0.411026446)); /* c6 */ - dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(z1 + z2 + z3, CONST_BITS+PASS1_BITS); - z1 -= z2; - z2 = MULTIPLY(tmp0 - tmp1, FIX(1.151670509)); /* c4 */ - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(z2 + z3 - MULTIPLY(tmp1 - tmp3, FIX(0.923568041)), /* c2+c6-c4 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(z1 + z2, CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(1.221765677)); /* (c3+c1-c5)/2 */ - tmp2 = MULTIPLY(tmp10 - tmp11, FIX(0.222383464)); /* (c3+c5-c1)/2 */ - tmp0 = tmp1 - tmp2; - tmp1 += tmp2; - tmp2 = MULTIPLY(tmp11 + tmp12, - FIX(1.800824523)); /* -c1 */ - tmp1 += tmp2; - tmp3 = MULTIPLY(tmp10 + tmp12, FIX(0.801442310)); /* c5 */ - tmp0 += tmp3; - tmp2 += tmp3 + MULTIPLY(tmp12, FIX(2.443531355)); /* c3+c1-c5 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp0, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp1, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp2, CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 6x6 sample block. - */ - -GLOBAL(void) -jpeg_fdct_6x6 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2; - INT32 tmp10, tmp11, tmp12; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* cK represents sqrt(2) * cos(K*pi/12). */ - - dataptr = data; - for (ctr = 0; ctr < 6; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[5]); - tmp11 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[4]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[3]); - - tmp10 = tmp0 + tmp2; - tmp12 = tmp0 - tmp2; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[5]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[4]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[3]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 - 6 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp12, FIX(1.224744871)), /* c2 */ - CONST_BITS-PASS1_BITS); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp11, FIX(0.707106781)), /* c4 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp10 = DESCALE(MULTIPLY(tmp0 + tmp2, FIX(0.366025404)), /* c5 */ - CONST_BITS-PASS1_BITS); - - dataptr[1] = (DCTELEM) (tmp10 + ((tmp0 + tmp1) << PASS1_BITS)); - dataptr[3] = (DCTELEM) ((tmp0 - tmp1 - tmp2) << PASS1_BITS); - dataptr[5] = (DCTELEM) (tmp10 + ((tmp2 - tmp1) << PASS1_BITS)); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/6)**2 = 16/9, which we fold - * into the constant multipliers: - * cK now represents sqrt(2) * cos(K*pi/12) * 16/9. - */ - - dataptr = data; - for (ctr = 0; ctr < 6; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*5]; - tmp11 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*3]; - - tmp10 = tmp0 + tmp2; - tmp12 = tmp0 - tmp2; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*3]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp12, FIX(2.177324216)), /* c2 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp11, FIX(1.257078722)), /* c4 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp2, FIX(0.650711829)); /* c5 */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0 + tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp2, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp2 - tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 5x5 sample block. - */ - -GLOBAL(void) -jpeg_fdct_5x5 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2; - INT32 tmp10, tmp11; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We scale the results further by 2 as part of output adaption */ - /* scaling for different DCT size. */ - /* cK represents sqrt(2) * cos(K*pi/10). */ - - dataptr = data; - for (ctr = 0; ctr < 5; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[4]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[3]); - tmp2 = GETJSAMPLE(elemptr[2]); - - tmp10 = tmp0 + tmp1; - tmp11 = tmp0 - tmp1; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[4]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[3]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp2 - 5 * CENTERJSAMPLE) << (PASS1_BITS+1)); - tmp11 = MULTIPLY(tmp11, FIX(0.790569415)); /* (c2+c4)/2 */ - tmp10 -= tmp2 << 2; - tmp10 = MULTIPLY(tmp10, FIX(0.353553391)); /* (c2-c4)/2 */ - dataptr[2] = (DCTELEM) DESCALE(tmp11 + tmp10, CONST_BITS-PASS1_BITS-1); - dataptr[4] = (DCTELEM) DESCALE(tmp11 - tmp10, CONST_BITS-PASS1_BITS-1); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp1, FIX(0.831253876)); /* c3 */ - - dataptr[1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0, FIX(0.513743148)), /* c1-c3 */ - CONST_BITS-PASS1_BITS-1); - dataptr[3] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp1, FIX(2.176250899)), /* c1+c3 */ - CONST_BITS-PASS1_BITS-1); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/5)**2 = 64/25, which we partially - * fold into the constant multipliers (other part was done in pass 1): - * cK now represents sqrt(2) * cos(K*pi/10) * 32/25. - */ - - dataptr = data; - for (ctr = 0; ctr < 5; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*4]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*3]; - tmp2 = dataptr[DCTSIZE*2]; - - tmp10 = tmp0 + tmp1; - tmp11 = tmp0 - tmp1; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*4]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*3]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp2, FIX(1.28)), /* 32/25 */ - CONST_BITS+PASS1_BITS); - tmp11 = MULTIPLY(tmp11, FIX(1.011928851)); /* (c2+c4)/2 */ - tmp10 -= tmp2 << 2; - tmp10 = MULTIPLY(tmp10, FIX(0.452548340)); /* (c2-c4)/2 */ - dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(tmp11 + tmp10, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(tmp11 - tmp10, CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp1, FIX(1.064004961)); /* c3 */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0, FIX(0.657591230)), /* c1-c3 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp1, FIX(2.785601151)), /* c1+c3 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 4x4 sample block. - */ - -GLOBAL(void) -jpeg_fdct_4x4 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1; - INT32 tmp10, tmp11; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We must also scale the output by (8/4)**2 = 2**2, which we add here. */ - /* cK represents sqrt(2) * cos(K*pi/16) [refers to 8-point FDCT]. */ - - dataptr = data; - for (ctr = 0; ctr < 4; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[3]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[2]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[3]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[2]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp0 + tmp1 - 4 * CENTERJSAMPLE) << (PASS1_BITS+2)); - dataptr[2] = (DCTELEM) ((tmp0 - tmp1) << (PASS1_BITS+2)); - - /* Odd part */ - - tmp0 = MULTIPLY(tmp10 + tmp11, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-PASS1_BITS-3); - - dataptr[1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + MULTIPLY(tmp10, FIX_0_765366865), /* c2-c6 */ - CONST_BITS-PASS1_BITS-2); - dataptr[3] = (DCTELEM) - RIGHT_SHIFT(tmp0 - MULTIPLY(tmp11, FIX_1_847759065), /* c2+c6 */ - CONST_BITS-PASS1_BITS-2); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - */ - - dataptr = data; - for (ctr = 0; ctr < 4; ctr++) { - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*3] + (ONE << (PASS1_BITS-1)); - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*2]; - - tmp10 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*3]; - tmp11 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*2]; - - dataptr[DCTSIZE*0] = (DCTELEM) RIGHT_SHIFT(tmp0 + tmp1, PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) RIGHT_SHIFT(tmp0 - tmp1, PASS1_BITS); - - /* Odd part */ - - tmp0 = MULTIPLY(tmp10 + tmp11, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS+PASS1_BITS-1); - - dataptr[DCTSIZE*1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + MULTIPLY(tmp10, FIX_0_765366865), /* c2-c6 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - RIGHT_SHIFT(tmp0 - MULTIPLY(tmp11, FIX_1_847759065), /* c2+c6 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 3x3 sample block. - */ - -GLOBAL(void) -jpeg_fdct_3x3 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We scale the results further by 2**2 as part of output adaption */ - /* scaling for different DCT size. */ - /* cK represents sqrt(2) * cos(K*pi/6). */ - - dataptr = data; - for (ctr = 0; ctr < 3; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[2]); - tmp1 = GETJSAMPLE(elemptr[1]); - - tmp2 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[2]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp0 + tmp1 - 3 * CENTERJSAMPLE) << (PASS1_BITS+2)); - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp1, FIX(0.707106781)), /* c2 */ - CONST_BITS-PASS1_BITS-2); - - /* Odd part */ - - dataptr[1] = (DCTELEM) - DESCALE(MULTIPLY(tmp2, FIX(1.224744871)), /* c1 */ - CONST_BITS-PASS1_BITS-2); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/3)**2 = 64/9, which we partially - * fold into the constant multipliers (other part was done in pass 1): - * cK now represents sqrt(2) * cos(K*pi/6) * 16/9. - */ - - dataptr = data; - for (ctr = 0; ctr < 3; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*2]; - tmp1 = dataptr[DCTSIZE*1]; - - tmp2 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*2]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 + tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp1, FIX(1.257078722)), /* c2 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(MULTIPLY(tmp2, FIX(2.177324216)), /* c1 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 2x2 sample block. - */ - -GLOBAL(void) -jpeg_fdct_2x2 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3; - JSAMPROW elemptr; - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT. */ - - /* Row 0 */ - elemptr = sample_data[0] + start_col; - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[1]); - tmp1 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[1]); - - /* Row 1 */ - elemptr = sample_data[1] + start_col; - - tmp2 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[1]); - tmp3 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[1]); - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/2)**2 = 2**4. - */ - - /* Column 0 */ - /* Apply unsigned->signed conversion */ - data[DCTSIZE*0] = (DCTELEM) ((tmp0 + tmp2 - 4 * CENTERJSAMPLE) << 4); - data[DCTSIZE*1] = (DCTELEM) ((tmp0 - tmp2) << 4); - - /* Column 1 */ - data[DCTSIZE*0+1] = (DCTELEM) ((tmp1 + tmp3) << 4); - data[DCTSIZE*1+1] = (DCTELEM) ((tmp1 - tmp3) << 4); -} - - -/* - * Perform the forward DCT on a 1x1 sample block. - */ - -GLOBAL(void) -jpeg_fdct_1x1 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* We leave the result scaled up by an overall factor of 8. */ - /* We must also scale the output by (8/1)**2 = 2**6. */ - /* Apply unsigned->signed conversion */ - data[0] = (DCTELEM) - ((GETJSAMPLE(sample_data[0][start_col]) - CENTERJSAMPLE) << 6); -} - - -/* - * Perform the forward DCT on a 9x9 sample block. - */ - -GLOBAL(void) -jpeg_fdct_9x9 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4; - INT32 tmp10, tmp11, tmp12, tmp13; - INT32 z1, z2; - DCTELEM workspace[8]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* we scale the results further by 2 as part of output adaption */ - /* scaling for different DCT size. */ - /* cK represents sqrt(2) * cos(K*pi/18). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[8]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[7]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[6]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[5]); - tmp4 = GETJSAMPLE(elemptr[4]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[8]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[7]); - tmp12 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[6]); - tmp13 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[5]); - - z1 = tmp0 + tmp2 + tmp3; - z2 = tmp1 + tmp4; - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) ((z1 + z2 - 9 * CENTERJSAMPLE) << 1); - dataptr[6] = (DCTELEM) - DESCALE(MULTIPLY(z1 - z2 - z2, FIX(0.707106781)), /* c6 */ - CONST_BITS-1); - z1 = MULTIPLY(tmp0 - tmp2, FIX(1.328926049)); /* c2 */ - z2 = MULTIPLY(tmp1 - tmp4 - tmp4, FIX(0.707106781)); /* c6 */ - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp2 - tmp3, FIX(1.083350441)) /* c4 */ - + z1 + z2, CONST_BITS-1); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp3 - tmp0, FIX(0.245575608)) /* c8 */ - + z1 - z2, CONST_BITS-1); - - /* Odd part */ - - dataptr[3] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12 - tmp13, FIX(1.224744871)), /* c3 */ - CONST_BITS-1); - - tmp11 = MULTIPLY(tmp11, FIX(1.224744871)); /* c3 */ - tmp0 = MULTIPLY(tmp10 + tmp12, FIX(0.909038955)); /* c5 */ - tmp1 = MULTIPLY(tmp10 + tmp13, FIX(0.483689525)); /* c7 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp11 + tmp0 + tmp1, CONST_BITS-1); - - tmp2 = MULTIPLY(tmp12 - tmp13, FIX(1.392728481)); /* c1 */ - - dataptr[5] = (DCTELEM) DESCALE(tmp0 - tmp11 - tmp2, CONST_BITS-1); - dataptr[7] = (DCTELEM) DESCALE(tmp1 - tmp11 + tmp2, CONST_BITS-1); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 9) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/9)**2 = 64/81, which we partially - * fold into the constant multipliers and final/initial shifting: - * cK now represents sqrt(2) * cos(K*pi/18) * 128/81. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*0]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*7]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*6]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*5]; - tmp4 = dataptr[DCTSIZE*4]; - - tmp10 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*0]; - tmp11 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*7]; - tmp12 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*6]; - tmp13 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*5]; - - z1 = tmp0 + tmp2 + tmp3; - z2 = tmp1 + tmp4; - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(z1 + z2, FIX(1.580246914)), /* 128/81 */ - CONST_BITS+2); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(MULTIPLY(z1 - z2 - z2, FIX(1.117403309)), /* c6 */ - CONST_BITS+2); - z1 = MULTIPLY(tmp0 - tmp2, FIX(2.100031287)); /* c2 */ - z2 = MULTIPLY(tmp1 - tmp4 - tmp4, FIX(1.117403309)); /* c6 */ - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp2 - tmp3, FIX(1.711961190)) /* c4 */ - + z1 + z2, CONST_BITS+2); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp3 - tmp0, FIX(0.388070096)) /* c8 */ - + z1 - z2, CONST_BITS+2); - - /* Odd part */ - - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12 - tmp13, FIX(1.935399303)), /* c3 */ - CONST_BITS+2); - - tmp11 = MULTIPLY(tmp11, FIX(1.935399303)); /* c3 */ - tmp0 = MULTIPLY(tmp10 + tmp12, FIX(1.436506004)); /* c5 */ - tmp1 = MULTIPLY(tmp10 + tmp13, FIX(0.764348879)); /* c7 */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp11 + tmp0 + tmp1, CONST_BITS+2); - - tmp2 = MULTIPLY(tmp12 - tmp13, FIX(2.200854883)); /* c1 */ - - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(tmp0 - tmp11 - tmp2, CONST_BITS+2); - dataptr[DCTSIZE*7] = (DCTELEM) - DESCALE(tmp1 - tmp11 + tmp2, CONST_BITS+2); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 10x10 sample block. - */ - -GLOBAL(void) -jpeg_fdct_10x10 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14; - DCTELEM workspace[8*2]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* we scale the results further by 2 as part of output adaption */ - /* scaling for different DCT size. */ - /* cK represents sqrt(2) * cos(K*pi/20). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[9]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[8]); - tmp12 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[7]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[6]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[5]); - - tmp10 = tmp0 + tmp4; - tmp13 = tmp0 - tmp4; - tmp11 = tmp1 + tmp3; - tmp14 = tmp1 - tmp3; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[9]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[8]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[7]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[6]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[5]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 + tmp12 - 10 * CENTERJSAMPLE) << 1); - tmp12 += tmp12; - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.144122806)) - /* c4 */ - MULTIPLY(tmp11 - tmp12, FIX(0.437016024)), /* c8 */ - CONST_BITS-1); - tmp10 = MULTIPLY(tmp13 + tmp14, FIX(0.831253876)); /* c6 */ - dataptr[2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp13, FIX(0.513743148)), /* c2-c6 */ - CONST_BITS-1); - dataptr[6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(2.176250899)), /* c2+c6 */ - CONST_BITS-1); - - /* Odd part */ - - tmp10 = tmp0 + tmp4; - tmp11 = tmp1 - tmp3; - dataptr[5] = (DCTELEM) ((tmp10 - tmp11 - tmp2) << 1); - tmp2 <<= CONST_BITS; - dataptr[1] = (DCTELEM) - DESCALE(MULTIPLY(tmp0, FIX(1.396802247)) + /* c1 */ - MULTIPLY(tmp1, FIX(1.260073511)) + tmp2 + /* c3 */ - MULTIPLY(tmp3, FIX(0.642039522)) + /* c7 */ - MULTIPLY(tmp4, FIX(0.221231742)), /* c9 */ - CONST_BITS-1); - tmp12 = MULTIPLY(tmp0 - tmp4, FIX(0.951056516)) - /* (c3+c7)/2 */ - MULTIPLY(tmp1 + tmp3, FIX(0.587785252)); /* (c1-c9)/2 */ - tmp13 = MULTIPLY(tmp10 + tmp11, FIX(0.309016994)) + /* (c3-c7)/2 */ - (tmp11 << (CONST_BITS - 1)) - tmp2; - dataptr[3] = (DCTELEM) DESCALE(tmp12 + tmp13, CONST_BITS-1); - dataptr[7] = (DCTELEM) DESCALE(tmp12 - tmp13, CONST_BITS-1); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 10) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/10)**2 = 16/25, which we partially - * fold into the constant multipliers and final/initial shifting: - * cK now represents sqrt(2) * cos(K*pi/20) * 32/25. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*1]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*0]; - tmp12 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*7]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*6]; - tmp4 = dataptr[DCTSIZE*4] + dataptr[DCTSIZE*5]; - - tmp10 = tmp0 + tmp4; - tmp13 = tmp0 - tmp4; - tmp11 = tmp1 + tmp3; - tmp14 = tmp1 - tmp3; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*1]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*0]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*7]; - tmp3 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*6]; - tmp4 = dataptr[DCTSIZE*4] - dataptr[DCTSIZE*5]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11 + tmp12, FIX(1.28)), /* 32/25 */ - CONST_BITS+2); - tmp12 += tmp12; - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.464477191)) - /* c4 */ - MULTIPLY(tmp11 - tmp12, FIX(0.559380511)), /* c8 */ - CONST_BITS+2); - tmp10 = MULTIPLY(tmp13 + tmp14, FIX(1.064004961)); /* c6 */ - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp13, FIX(0.657591230)), /* c2-c6 */ - CONST_BITS+2); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(2.785601151)), /* c2+c6 */ - CONST_BITS+2); - - /* Odd part */ - - tmp10 = tmp0 + tmp4; - tmp11 = tmp1 - tmp3; - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp2, FIX(1.28)), /* 32/25 */ - CONST_BITS+2); - tmp2 = MULTIPLY(tmp2, FIX(1.28)); /* 32/25 */ - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(MULTIPLY(tmp0, FIX(1.787906876)) + /* c1 */ - MULTIPLY(tmp1, FIX(1.612894094)) + tmp2 + /* c3 */ - MULTIPLY(tmp3, FIX(0.821810588)) + /* c7 */ - MULTIPLY(tmp4, FIX(0.283176630)), /* c9 */ - CONST_BITS+2); - tmp12 = MULTIPLY(tmp0 - tmp4, FIX(1.217352341)) - /* (c3+c7)/2 */ - MULTIPLY(tmp1 + tmp3, FIX(0.752365123)); /* (c1-c9)/2 */ - tmp13 = MULTIPLY(tmp10 + tmp11, FIX(0.395541753)) + /* (c3-c7)/2 */ - MULTIPLY(tmp11, FIX(0.64)) - tmp2; /* 16/25 */ - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp12 + tmp13, CONST_BITS+2); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp12 - tmp13, CONST_BITS+2); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on an 11x11 sample block. - */ - -GLOBAL(void) -jpeg_fdct_11x11 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14; - INT32 z1, z2, z3; - DCTELEM workspace[8*3]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* we scale the results further by 2 as part of output adaption */ - /* scaling for different DCT size. */ - /* cK represents sqrt(2) * cos(K*pi/22). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[10]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[9]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[8]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[7]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[6]); - tmp5 = GETJSAMPLE(elemptr[5]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[10]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[9]); - tmp12 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[8]); - tmp13 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[7]); - tmp14 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[6]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp0 + tmp1 + tmp2 + tmp3 + tmp4 + tmp5 - 11 * CENTERJSAMPLE) << 1); - tmp5 += tmp5; - tmp0 -= tmp5; - tmp1 -= tmp5; - tmp2 -= tmp5; - tmp3 -= tmp5; - tmp4 -= tmp5; - z1 = MULTIPLY(tmp0 + tmp3, FIX(1.356927976)) + /* c2 */ - MULTIPLY(tmp2 + tmp4, FIX(0.201263574)); /* c10 */ - z2 = MULTIPLY(tmp1 - tmp3, FIX(0.926112931)); /* c6 */ - z3 = MULTIPLY(tmp0 - tmp1, FIX(1.189712156)); /* c4 */ - dataptr[2] = (DCTELEM) - DESCALE(z1 + z2 - MULTIPLY(tmp3, FIX(1.018300590)) /* c2+c8-c6 */ - - MULTIPLY(tmp4, FIX(1.390975730)), /* c4+c10 */ - CONST_BITS-1); - dataptr[4] = (DCTELEM) - DESCALE(z2 + z3 + MULTIPLY(tmp1, FIX(0.062335650)) /* c4-c6-c10 */ - - MULTIPLY(tmp2, FIX(1.356927976)) /* c2 */ - + MULTIPLY(tmp4, FIX(0.587485545)), /* c8 */ - CONST_BITS-1); - dataptr[6] = (DCTELEM) - DESCALE(z1 + z3 - MULTIPLY(tmp0, FIX(1.620527200)) /* c2+c4-c6 */ - - MULTIPLY(tmp2, FIX(0.788749120)), /* c8+c10 */ - CONST_BITS-1); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(1.286413905)); /* c3 */ - tmp2 = MULTIPLY(tmp10 + tmp12, FIX(1.068791298)); /* c5 */ - tmp3 = MULTIPLY(tmp10 + tmp13, FIX(0.764581576)); /* c7 */ - tmp0 = tmp1 + tmp2 + tmp3 - MULTIPLY(tmp10, FIX(1.719967871)) /* c7+c5+c3-c1 */ - + MULTIPLY(tmp14, FIX(0.398430003)); /* c9 */ - tmp4 = MULTIPLY(tmp11 + tmp12, - FIX(0.764581576)); /* -c7 */ - tmp5 = MULTIPLY(tmp11 + tmp13, - FIX(1.399818907)); /* -c1 */ - tmp1 += tmp4 + tmp5 + MULTIPLY(tmp11, FIX(1.276416582)) /* c9+c7+c1-c3 */ - - MULTIPLY(tmp14, FIX(1.068791298)); /* c5 */ - tmp10 = MULTIPLY(tmp12 + tmp13, FIX(0.398430003)); /* c9 */ - tmp2 += tmp4 + tmp10 - MULTIPLY(tmp12, FIX(1.989053629)) /* c9+c5+c3-c7 */ - + MULTIPLY(tmp14, FIX(1.399818907)); /* c1 */ - tmp3 += tmp5 + tmp10 + MULTIPLY(tmp13, FIX(1.305598626)) /* c1+c5-c9-c7 */ - - MULTIPLY(tmp14, FIX(1.286413905)); /* c3 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp0, CONST_BITS-1); - dataptr[3] = (DCTELEM) DESCALE(tmp1, CONST_BITS-1); - dataptr[5] = (DCTELEM) DESCALE(tmp2, CONST_BITS-1); - dataptr[7] = (DCTELEM) DESCALE(tmp3, CONST_BITS-1); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 11) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/11)**2 = 64/121, which we partially - * fold into the constant multipliers and final/initial shifting: - * cK now represents sqrt(2) * cos(K*pi/22) * 128/121. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*2]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*1]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*0]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*7]; - tmp4 = dataptr[DCTSIZE*4] + dataptr[DCTSIZE*6]; - tmp5 = dataptr[DCTSIZE*5]; - - tmp10 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*2]; - tmp11 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*1]; - tmp12 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*0]; - tmp13 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*7]; - tmp14 = dataptr[DCTSIZE*4] - dataptr[DCTSIZE*6]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 + tmp1 + tmp2 + tmp3 + tmp4 + tmp5, - FIX(1.057851240)), /* 128/121 */ - CONST_BITS+2); - tmp5 += tmp5; - tmp0 -= tmp5; - tmp1 -= tmp5; - tmp2 -= tmp5; - tmp3 -= tmp5; - tmp4 -= tmp5; - z1 = MULTIPLY(tmp0 + tmp3, FIX(1.435427942)) + /* c2 */ - MULTIPLY(tmp2 + tmp4, FIX(0.212906922)); /* c10 */ - z2 = MULTIPLY(tmp1 - tmp3, FIX(0.979689713)); /* c6 */ - z3 = MULTIPLY(tmp0 - tmp1, FIX(1.258538479)); /* c4 */ - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(z1 + z2 - MULTIPLY(tmp3, FIX(1.077210542)) /* c2+c8-c6 */ - - MULTIPLY(tmp4, FIX(1.471445400)), /* c4+c10 */ - CONST_BITS+2); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(z2 + z3 + MULTIPLY(tmp1, FIX(0.065941844)) /* c4-c6-c10 */ - - MULTIPLY(tmp2, FIX(1.435427942)) /* c2 */ - + MULTIPLY(tmp4, FIX(0.621472312)), /* c8 */ - CONST_BITS+2); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(z1 + z3 - MULTIPLY(tmp0, FIX(1.714276708)) /* c2+c4-c6 */ - - MULTIPLY(tmp2, FIX(0.834379234)), /* c8+c10 */ - CONST_BITS+2); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(1.360834544)); /* c3 */ - tmp2 = MULTIPLY(tmp10 + tmp12, FIX(1.130622199)); /* c5 */ - tmp3 = MULTIPLY(tmp10 + tmp13, FIX(0.808813568)); /* c7 */ - tmp0 = tmp1 + tmp2 + tmp3 - MULTIPLY(tmp10, FIX(1.819470145)) /* c7+c5+c3-c1 */ - + MULTIPLY(tmp14, FIX(0.421479672)); /* c9 */ - tmp4 = MULTIPLY(tmp11 + tmp12, - FIX(0.808813568)); /* -c7 */ - tmp5 = MULTIPLY(tmp11 + tmp13, - FIX(1.480800167)); /* -c1 */ - tmp1 += tmp4 + tmp5 + MULTIPLY(tmp11, FIX(1.350258864)) /* c9+c7+c1-c3 */ - - MULTIPLY(tmp14, FIX(1.130622199)); /* c5 */ - tmp10 = MULTIPLY(tmp12 + tmp13, FIX(0.421479672)); /* c9 */ - tmp2 += tmp4 + tmp10 - MULTIPLY(tmp12, FIX(2.104122847)) /* c9+c5+c3-c7 */ - + MULTIPLY(tmp14, FIX(1.480800167)); /* c1 */ - tmp3 += tmp5 + tmp10 + MULTIPLY(tmp13, FIX(1.381129125)) /* c1+c5-c9-c7 */ - - MULTIPLY(tmp14, FIX(1.360834544)); /* c3 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp0, CONST_BITS+2); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp1, CONST_BITS+2); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp2, CONST_BITS+2); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp3, CONST_BITS+2); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 12x12 sample block. - */ - -GLOBAL(void) -jpeg_fdct_12x12 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15; - DCTELEM workspace[8*4]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT. */ - /* cK represents sqrt(2) * cos(K*pi/24). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[11]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[10]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[9]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[8]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[7]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[6]); - - tmp10 = tmp0 + tmp5; - tmp13 = tmp0 - tmp5; - tmp11 = tmp1 + tmp4; - tmp14 = tmp1 - tmp4; - tmp12 = tmp2 + tmp3; - tmp15 = tmp2 - tmp3; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[11]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[10]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[9]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[8]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[7]); - tmp5 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[6]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) (tmp10 + tmp11 + tmp12 - 12 * CENTERJSAMPLE); - dataptr[6] = (DCTELEM) (tmp13 - tmp14 - tmp15); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.224744871)), /* c4 */ - CONST_BITS); - dataptr[2] = (DCTELEM) - DESCALE(tmp14 - tmp15 + MULTIPLY(tmp13 + tmp15, FIX(1.366025404)), /* c2 */ - CONST_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp1 + tmp4, FIX_0_541196100); /* c9 */ - tmp14 = tmp10 + MULTIPLY(tmp1, FIX_0_765366865); /* c3-c9 */ - tmp15 = tmp10 - MULTIPLY(tmp4, FIX_1_847759065); /* c3+c9 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(1.121971054)); /* c5 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(0.860918669)); /* c7 */ - tmp10 = tmp12 + tmp13 + tmp14 - MULTIPLY(tmp0, FIX(0.580774953)) /* c5+c7-c1 */ - + MULTIPLY(tmp5, FIX(0.184591911)); /* c11 */ - tmp11 = MULTIPLY(tmp2 + tmp3, - FIX(0.184591911)); /* -c11 */ - tmp12 += tmp11 - tmp15 - MULTIPLY(tmp2, FIX(2.339493912)) /* c1+c5-c11 */ - + MULTIPLY(tmp5, FIX(0.860918669)); /* c7 */ - tmp13 += tmp11 - tmp14 + MULTIPLY(tmp3, FIX(0.725788011)) /* c1+c11-c7 */ - - MULTIPLY(tmp5, FIX(1.121971054)); /* c5 */ - tmp11 = tmp15 + MULTIPLY(tmp0 - tmp3, FIX(1.306562965)) /* c3 */ - - MULTIPLY(tmp2 + tmp5, FIX_0_541196100); /* c9 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp10, CONST_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp11, CONST_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp12, CONST_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp13, CONST_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 12) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/12)**2 = 4/9, which we partially - * fold into the constant multipliers and final shifting: - * cK now represents sqrt(2) * cos(K*pi/24) * 8/9. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*3]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*2]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*1]; - tmp3 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*0]; - tmp4 = dataptr[DCTSIZE*4] + dataptr[DCTSIZE*7]; - tmp5 = dataptr[DCTSIZE*5] + dataptr[DCTSIZE*6]; - - tmp10 = tmp0 + tmp5; - tmp13 = tmp0 - tmp5; - tmp11 = tmp1 + tmp4; - tmp14 = tmp1 - tmp4; - tmp12 = tmp2 + tmp3; - tmp15 = tmp2 - tmp3; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*3]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*2]; - tmp2 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*1]; - tmp3 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*0]; - tmp4 = dataptr[DCTSIZE*4] - dataptr[DCTSIZE*7]; - tmp5 = dataptr[DCTSIZE*5] - dataptr[DCTSIZE*6]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11 + tmp12, FIX(0.888888889)), /* 8/9 */ - CONST_BITS+1); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(MULTIPLY(tmp13 - tmp14 - tmp15, FIX(0.888888889)), /* 8/9 */ - CONST_BITS+1); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.088662108)), /* c4 */ - CONST_BITS+1); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp14 - tmp15, FIX(0.888888889)) + /* 8/9 */ - MULTIPLY(tmp13 + tmp15, FIX(1.214244803)), /* c2 */ - CONST_BITS+1); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp1 + tmp4, FIX(0.481063200)); /* c9 */ - tmp14 = tmp10 + MULTIPLY(tmp1, FIX(0.680326102)); /* c3-c9 */ - tmp15 = tmp10 - MULTIPLY(tmp4, FIX(1.642452502)); /* c3+c9 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(0.997307603)); /* c5 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(0.765261039)); /* c7 */ - tmp10 = tmp12 + tmp13 + tmp14 - MULTIPLY(tmp0, FIX(0.516244403)) /* c5+c7-c1 */ - + MULTIPLY(tmp5, FIX(0.164081699)); /* c11 */ - tmp11 = MULTIPLY(tmp2 + tmp3, - FIX(0.164081699)); /* -c11 */ - tmp12 += tmp11 - tmp15 - MULTIPLY(tmp2, FIX(2.079550144)) /* c1+c5-c11 */ - + MULTIPLY(tmp5, FIX(0.765261039)); /* c7 */ - tmp13 += tmp11 - tmp14 + MULTIPLY(tmp3, FIX(0.645144899)) /* c1+c11-c7 */ - - MULTIPLY(tmp5, FIX(0.997307603)); /* c5 */ - tmp11 = tmp15 + MULTIPLY(tmp0 - tmp3, FIX(1.161389302)) /* c3 */ - - MULTIPLY(tmp2 + tmp5, FIX(0.481063200)); /* c9 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp10, CONST_BITS+1); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp11, CONST_BITS+1); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp12, CONST_BITS+1); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp13, CONST_BITS+1); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 13x13 sample block. - */ - -GLOBAL(void) -jpeg_fdct_13x13 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15; - INT32 z1, z2; - DCTELEM workspace[8*5]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT. */ - /* cK represents sqrt(2) * cos(K*pi/26). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[12]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[11]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[10]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[9]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[8]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[7]); - tmp6 = GETJSAMPLE(elemptr[6]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[12]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[11]); - tmp12 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[10]); - tmp13 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[9]); - tmp14 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[8]); - tmp15 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[7]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - (tmp0 + tmp1 + tmp2 + tmp3 + tmp4 + tmp5 + tmp6 - 13 * CENTERJSAMPLE); - tmp6 += tmp6; - tmp0 -= tmp6; - tmp1 -= tmp6; - tmp2 -= tmp6; - tmp3 -= tmp6; - tmp4 -= tmp6; - tmp5 -= tmp6; - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp0, FIX(1.373119086)) + /* c2 */ - MULTIPLY(tmp1, FIX(1.058554052)) + /* c6 */ - MULTIPLY(tmp2, FIX(0.501487041)) - /* c10 */ - MULTIPLY(tmp3, FIX(0.170464608)) - /* c12 */ - MULTIPLY(tmp4, FIX(0.803364869)) - /* c8 */ - MULTIPLY(tmp5, FIX(1.252223920)), /* c4 */ - CONST_BITS); - z1 = MULTIPLY(tmp0 - tmp2, FIX(1.155388986)) - /* (c4+c6)/2 */ - MULTIPLY(tmp3 - tmp4, FIX(0.435816023)) - /* (c2-c10)/2 */ - MULTIPLY(tmp1 - tmp5, FIX(0.316450131)); /* (c8-c12)/2 */ - z2 = MULTIPLY(tmp0 + tmp2, FIX(0.096834934)) - /* (c4-c6)/2 */ - MULTIPLY(tmp3 + tmp4, FIX(0.937303064)) + /* (c2+c10)/2 */ - MULTIPLY(tmp1 + tmp5, FIX(0.486914739)); /* (c8+c12)/2 */ - - dataptr[4] = (DCTELEM) DESCALE(z1 + z2, CONST_BITS); - dataptr[6] = (DCTELEM) DESCALE(z1 - z2, CONST_BITS); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(1.322312651)); /* c3 */ - tmp2 = MULTIPLY(tmp10 + tmp12, FIX(1.163874945)); /* c5 */ - tmp3 = MULTIPLY(tmp10 + tmp13, FIX(0.937797057)) + /* c7 */ - MULTIPLY(tmp14 + tmp15, FIX(0.338443458)); /* c11 */ - tmp0 = tmp1 + tmp2 + tmp3 - - MULTIPLY(tmp10, FIX(2.020082300)) + /* c3+c5+c7-c1 */ - MULTIPLY(tmp14, FIX(0.318774355)); /* c9-c11 */ - tmp4 = MULTIPLY(tmp14 - tmp15, FIX(0.937797057)) - /* c7 */ - MULTIPLY(tmp11 + tmp12, FIX(0.338443458)); /* c11 */ - tmp5 = MULTIPLY(tmp11 + tmp13, - FIX(1.163874945)); /* -c5 */ - tmp1 += tmp4 + tmp5 + - MULTIPLY(tmp11, FIX(0.837223564)) - /* c5+c9+c11-c3 */ - MULTIPLY(tmp14, FIX(2.341699410)); /* c1+c7 */ - tmp6 = MULTIPLY(tmp12 + tmp13, - FIX(0.657217813)); /* -c9 */ - tmp2 += tmp4 + tmp6 - - MULTIPLY(tmp12, FIX(1.572116027)) + /* c1+c5-c9-c11 */ - MULTIPLY(tmp15, FIX(2.260109708)); /* c3+c7 */ - tmp3 += tmp5 + tmp6 + - MULTIPLY(tmp13, FIX(2.205608352)) - /* c3+c5+c9-c7 */ - MULTIPLY(tmp15, FIX(1.742345811)); /* c1+c11 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp0, CONST_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp1, CONST_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp2, CONST_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp3, CONST_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 13) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/13)**2 = 64/169, which we partially - * fold into the constant multipliers and final shifting: - * cK now represents sqrt(2) * cos(K*pi/26) * 128/169. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*4]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*3]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*2]; - tmp3 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*1]; - tmp4 = dataptr[DCTSIZE*4] + wsptr[DCTSIZE*0]; - tmp5 = dataptr[DCTSIZE*5] + dataptr[DCTSIZE*7]; - tmp6 = dataptr[DCTSIZE*6]; - - tmp10 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*4]; - tmp11 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*3]; - tmp12 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*2]; - tmp13 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*1]; - tmp14 = dataptr[DCTSIZE*4] - wsptr[DCTSIZE*0]; - tmp15 = dataptr[DCTSIZE*5] - dataptr[DCTSIZE*7]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 + tmp1 + tmp2 + tmp3 + tmp4 + tmp5 + tmp6, - FIX(0.757396450)), /* 128/169 */ - CONST_BITS+1); - tmp6 += tmp6; - tmp0 -= tmp6; - tmp1 -= tmp6; - tmp2 -= tmp6; - tmp3 -= tmp6; - tmp4 -= tmp6; - tmp5 -= tmp6; - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp0, FIX(1.039995521)) + /* c2 */ - MULTIPLY(tmp1, FIX(0.801745081)) + /* c6 */ - MULTIPLY(tmp2, FIX(0.379824504)) - /* c10 */ - MULTIPLY(tmp3, FIX(0.129109289)) - /* c12 */ - MULTIPLY(tmp4, FIX(0.608465700)) - /* c8 */ - MULTIPLY(tmp5, FIX(0.948429952)), /* c4 */ - CONST_BITS+1); - z1 = MULTIPLY(tmp0 - tmp2, FIX(0.875087516)) - /* (c4+c6)/2 */ - MULTIPLY(tmp3 - tmp4, FIX(0.330085509)) - /* (c2-c10)/2 */ - MULTIPLY(tmp1 - tmp5, FIX(0.239678205)); /* (c8-c12)/2 */ - z2 = MULTIPLY(tmp0 + tmp2, FIX(0.073342435)) - /* (c4-c6)/2 */ - MULTIPLY(tmp3 + tmp4, FIX(0.709910013)) + /* (c2+c10)/2 */ - MULTIPLY(tmp1 + tmp5, FIX(0.368787494)); /* (c8+c12)/2 */ - - dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(z1 + z2, CONST_BITS+1); - dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(z1 - z2, CONST_BITS+1); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(1.001514908)); /* c3 */ - tmp2 = MULTIPLY(tmp10 + tmp12, FIX(0.881514751)); /* c5 */ - tmp3 = MULTIPLY(tmp10 + tmp13, FIX(0.710284161)) + /* c7 */ - MULTIPLY(tmp14 + tmp15, FIX(0.256335874)); /* c11 */ - tmp0 = tmp1 + tmp2 + tmp3 - - MULTIPLY(tmp10, FIX(1.530003162)) + /* c3+c5+c7-c1 */ - MULTIPLY(tmp14, FIX(0.241438564)); /* c9-c11 */ - tmp4 = MULTIPLY(tmp14 - tmp15, FIX(0.710284161)) - /* c7 */ - MULTIPLY(tmp11 + tmp12, FIX(0.256335874)); /* c11 */ - tmp5 = MULTIPLY(tmp11 + tmp13, - FIX(0.881514751)); /* -c5 */ - tmp1 += tmp4 + tmp5 + - MULTIPLY(tmp11, FIX(0.634110155)) - /* c5+c9+c11-c3 */ - MULTIPLY(tmp14, FIX(1.773594819)); /* c1+c7 */ - tmp6 = MULTIPLY(tmp12 + tmp13, - FIX(0.497774438)); /* -c9 */ - tmp2 += tmp4 + tmp6 - - MULTIPLY(tmp12, FIX(1.190715098)) + /* c1+c5-c9-c11 */ - MULTIPLY(tmp15, FIX(1.711799069)); /* c3+c7 */ - tmp3 += tmp5 + tmp6 + - MULTIPLY(tmp13, FIX(1.670519935)) - /* c3+c5+c9-c7 */ - MULTIPLY(tmp15, FIX(1.319646532)); /* c1+c11 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp0, CONST_BITS+1); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp1, CONST_BITS+1); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp2, CONST_BITS+1); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp3, CONST_BITS+1); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 14x14 sample block. - */ - -GLOBAL(void) -jpeg_fdct_14x14 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16; - DCTELEM workspace[8*6]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT. */ - /* cK represents sqrt(2) * cos(K*pi/28). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[13]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[12]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[11]); - tmp13 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[10]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[9]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[8]); - tmp6 = GETJSAMPLE(elemptr[6]) + GETJSAMPLE(elemptr[7]); - - tmp10 = tmp0 + tmp6; - tmp14 = tmp0 - tmp6; - tmp11 = tmp1 + tmp5; - tmp15 = tmp1 - tmp5; - tmp12 = tmp2 + tmp4; - tmp16 = tmp2 - tmp4; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[13]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[12]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[11]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[10]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[9]); - tmp5 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[8]); - tmp6 = GETJSAMPLE(elemptr[6]) - GETJSAMPLE(elemptr[7]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - (tmp10 + tmp11 + tmp12 + tmp13 - 14 * CENTERJSAMPLE); - tmp13 += tmp13; - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(1.274162392)) + /* c4 */ - MULTIPLY(tmp11 - tmp13, FIX(0.314692123)) - /* c12 */ - MULTIPLY(tmp12 - tmp13, FIX(0.881747734)), /* c8 */ - CONST_BITS); - - tmp10 = MULTIPLY(tmp14 + tmp15, FIX(1.105676686)); /* c6 */ - - dataptr[2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp14, FIX(0.273079590)) /* c2-c6 */ - + MULTIPLY(tmp16, FIX(0.613604268)), /* c10 */ - CONST_BITS); - dataptr[6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp15, FIX(1.719280954)) /* c6+c10 */ - - MULTIPLY(tmp16, FIX(1.378756276)), /* c2 */ - CONST_BITS); - - /* Odd part */ - - tmp10 = tmp1 + tmp2; - tmp11 = tmp5 - tmp4; - dataptr[7] = (DCTELEM) (tmp0 - tmp10 + tmp3 - tmp11 - tmp6); - tmp3 <<= CONST_BITS; - tmp10 = MULTIPLY(tmp10, - FIX(0.158341681)); /* -c13 */ - tmp11 = MULTIPLY(tmp11, FIX(1.405321284)); /* c1 */ - tmp10 += tmp11 - tmp3; - tmp11 = MULTIPLY(tmp0 + tmp2, FIX(1.197448846)) + /* c5 */ - MULTIPLY(tmp4 + tmp6, FIX(0.752406978)); /* c9 */ - dataptr[5] = (DCTELEM) - DESCALE(tmp10 + tmp11 - MULTIPLY(tmp2, FIX(2.373959773)) /* c3+c5-c13 */ - + MULTIPLY(tmp4, FIX(1.119999435)), /* c1+c11-c9 */ - CONST_BITS); - tmp12 = MULTIPLY(tmp0 + tmp1, FIX(1.334852607)) + /* c3 */ - MULTIPLY(tmp5 - tmp6, FIX(0.467085129)); /* c11 */ - dataptr[3] = (DCTELEM) - DESCALE(tmp10 + tmp12 - MULTIPLY(tmp1, FIX(0.424103948)) /* c3-c9-c13 */ - - MULTIPLY(tmp5, FIX(3.069855259)), /* c1+c5+c11 */ - CONST_BITS); - dataptr[1] = (DCTELEM) - DESCALE(tmp11 + tmp12 + tmp3 + tmp6 - - MULTIPLY(tmp0 + tmp6, FIX(1.126980169)), /* c3+c5-c1 */ - CONST_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 14) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/14)**2 = 16/49, which we partially - * fold into the constant multipliers and final shifting: - * cK now represents sqrt(2) * cos(K*pi/28) * 32/49. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*3]; - tmp13 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*2]; - tmp4 = dataptr[DCTSIZE*4] + wsptr[DCTSIZE*1]; - tmp5 = dataptr[DCTSIZE*5] + wsptr[DCTSIZE*0]; - tmp6 = dataptr[DCTSIZE*6] + dataptr[DCTSIZE*7]; - - tmp10 = tmp0 + tmp6; - tmp14 = tmp0 - tmp6; - tmp11 = tmp1 + tmp5; - tmp15 = tmp1 - tmp5; - tmp12 = tmp2 + tmp4; - tmp16 = tmp2 - tmp4; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*3]; - tmp3 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*2]; - tmp4 = dataptr[DCTSIZE*4] - wsptr[DCTSIZE*1]; - tmp5 = dataptr[DCTSIZE*5] - wsptr[DCTSIZE*0]; - tmp6 = dataptr[DCTSIZE*6] - dataptr[DCTSIZE*7]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11 + tmp12 + tmp13, - FIX(0.653061224)), /* 32/49 */ - CONST_BITS+1); - tmp13 += tmp13; - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(0.832106052)) + /* c4 */ - MULTIPLY(tmp11 - tmp13, FIX(0.205513223)) - /* c12 */ - MULTIPLY(tmp12 - tmp13, FIX(0.575835255)), /* c8 */ - CONST_BITS+1); - - tmp10 = MULTIPLY(tmp14 + tmp15, FIX(0.722074570)); /* c6 */ - - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp14, FIX(0.178337691)) /* c2-c6 */ - + MULTIPLY(tmp16, FIX(0.400721155)), /* c10 */ - CONST_BITS+1); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp15, FIX(1.122795725)) /* c6+c10 */ - - MULTIPLY(tmp16, FIX(0.900412262)), /* c2 */ - CONST_BITS+1); - - /* Odd part */ - - tmp10 = tmp1 + tmp2; - tmp11 = tmp5 - tmp4; - dataptr[DCTSIZE*7] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp10 + tmp3 - tmp11 - tmp6, - FIX(0.653061224)), /* 32/49 */ - CONST_BITS+1); - tmp3 = MULTIPLY(tmp3 , FIX(0.653061224)); /* 32/49 */ - tmp10 = MULTIPLY(tmp10, - FIX(0.103406812)); /* -c13 */ - tmp11 = MULTIPLY(tmp11, FIX(0.917760839)); /* c1 */ - tmp10 += tmp11 - tmp3; - tmp11 = MULTIPLY(tmp0 + tmp2, FIX(0.782007410)) + /* c5 */ - MULTIPLY(tmp4 + tmp6, FIX(0.491367823)); /* c9 */ - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(tmp10 + tmp11 - MULTIPLY(tmp2, FIX(1.550341076)) /* c3+c5-c13 */ - + MULTIPLY(tmp4, FIX(0.731428202)), /* c1+c11-c9 */ - CONST_BITS+1); - tmp12 = MULTIPLY(tmp0 + tmp1, FIX(0.871740478)) + /* c3 */ - MULTIPLY(tmp5 - tmp6, FIX(0.305035186)); /* c11 */ - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(tmp10 + tmp12 - MULTIPLY(tmp1, FIX(0.276965844)) /* c3-c9-c13 */ - - MULTIPLY(tmp5, FIX(2.004803435)), /* c1+c5+c11 */ - CONST_BITS+1); - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp11 + tmp12 + tmp3 - - MULTIPLY(tmp0, FIX(0.735987049)) /* c3+c5-c1 */ - - MULTIPLY(tmp6, FIX(0.082925825)), /* c9-c11-c13 */ - CONST_BITS+1); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 15x15 sample block. - */ - -GLOBAL(void) -jpeg_fdct_15x15 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16; - INT32 z1, z2, z3; - DCTELEM workspace[8*7]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT. */ - /* cK represents sqrt(2) * cos(K*pi/30). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[14]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[13]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[12]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[11]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[10]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[9]); - tmp6 = GETJSAMPLE(elemptr[6]) + GETJSAMPLE(elemptr[8]); - tmp7 = GETJSAMPLE(elemptr[7]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[14]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[13]); - tmp12 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[12]); - tmp13 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[11]); - tmp14 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[10]); - tmp15 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[9]); - tmp16 = GETJSAMPLE(elemptr[6]) - GETJSAMPLE(elemptr[8]); - - z1 = tmp0 + tmp4 + tmp5; - z2 = tmp1 + tmp3 + tmp6; - z3 = tmp2 + tmp7; - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) (z1 + z2 + z3 - 15 * CENTERJSAMPLE); - z3 += z3; - dataptr[6] = (DCTELEM) - DESCALE(MULTIPLY(z1 - z3, FIX(1.144122806)) - /* c6 */ - MULTIPLY(z2 - z3, FIX(0.437016024)), /* c12 */ - CONST_BITS); - tmp2 += ((tmp1 + tmp4) >> 1) - tmp7 - tmp7; - z1 = MULTIPLY(tmp3 - tmp2, FIX(1.531135173)) - /* c2+c14 */ - MULTIPLY(tmp6 - tmp2, FIX(2.238241955)); /* c4+c8 */ - z2 = MULTIPLY(tmp5 - tmp2, FIX(0.798468008)) - /* c8-c14 */ - MULTIPLY(tmp0 - tmp2, FIX(0.091361227)); /* c2-c4 */ - z3 = MULTIPLY(tmp0 - tmp3, FIX(1.383309603)) + /* c2 */ - MULTIPLY(tmp6 - tmp5, FIX(0.946293579)) + /* c8 */ - MULTIPLY(tmp1 - tmp4, FIX(0.790569415)); /* (c6+c12)/2 */ - - dataptr[2] = (DCTELEM) DESCALE(z1 + z3, CONST_BITS); - dataptr[4] = (DCTELEM) DESCALE(z2 + z3, CONST_BITS); - - /* Odd part */ - - tmp2 = MULTIPLY(tmp10 - tmp12 - tmp13 + tmp15 + tmp16, - FIX(1.224744871)); /* c5 */ - tmp1 = MULTIPLY(tmp10 - tmp14 - tmp15, FIX(1.344997024)) + /* c3 */ - MULTIPLY(tmp11 - tmp13 - tmp16, FIX(0.831253876)); /* c9 */ - tmp12 = MULTIPLY(tmp12, FIX(1.224744871)); /* c5 */ - tmp4 = MULTIPLY(tmp10 - tmp16, FIX(1.406466353)) + /* c1 */ - MULTIPLY(tmp11 + tmp14, FIX(1.344997024)) + /* c3 */ - MULTIPLY(tmp13 + tmp15, FIX(0.575212477)); /* c11 */ - tmp0 = MULTIPLY(tmp13, FIX(0.475753014)) - /* c7-c11 */ - MULTIPLY(tmp14, FIX(0.513743148)) + /* c3-c9 */ - MULTIPLY(tmp16, FIX(1.700497885)) + tmp4 + tmp12; /* c1+c13 */ - tmp3 = MULTIPLY(tmp10, - FIX(0.355500862)) - /* -(c1-c7) */ - MULTIPLY(tmp11, FIX(2.176250899)) - /* c3+c9 */ - MULTIPLY(tmp15, FIX(0.869244010)) + tmp4 - tmp12; /* c11+c13 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp0, CONST_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp1, CONST_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp2, CONST_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp3, CONST_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 15) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/15)**2 = 64/225, which we partially - * fold into the constant multipliers and final shifting: - * cK now represents sqrt(2) * cos(K*pi/30) * 256/225. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*6]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*5]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*4]; - tmp3 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*3]; - tmp4 = dataptr[DCTSIZE*4] + wsptr[DCTSIZE*2]; - tmp5 = dataptr[DCTSIZE*5] + wsptr[DCTSIZE*1]; - tmp6 = dataptr[DCTSIZE*6] + wsptr[DCTSIZE*0]; - tmp7 = dataptr[DCTSIZE*7]; - - tmp10 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*6]; - tmp11 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*5]; - tmp12 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*4]; - tmp13 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*3]; - tmp14 = dataptr[DCTSIZE*4] - wsptr[DCTSIZE*2]; - tmp15 = dataptr[DCTSIZE*5] - wsptr[DCTSIZE*1]; - tmp16 = dataptr[DCTSIZE*6] - wsptr[DCTSIZE*0]; - - z1 = tmp0 + tmp4 + tmp5; - z2 = tmp1 + tmp3 + tmp6; - z3 = tmp2 + tmp7; - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(z1 + z2 + z3, FIX(1.137777778)), /* 256/225 */ - CONST_BITS+2); - z3 += z3; - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(MULTIPLY(z1 - z3, FIX(1.301757503)) - /* c6 */ - MULTIPLY(z2 - z3, FIX(0.497227121)), /* c12 */ - CONST_BITS+2); - tmp2 += ((tmp1 + tmp4) >> 1) - tmp7 - tmp7; - z1 = MULTIPLY(tmp3 - tmp2, FIX(1.742091575)) - /* c2+c14 */ - MULTIPLY(tmp6 - tmp2, FIX(2.546621957)); /* c4+c8 */ - z2 = MULTIPLY(tmp5 - tmp2, FIX(0.908479156)) - /* c8-c14 */ - MULTIPLY(tmp0 - tmp2, FIX(0.103948774)); /* c2-c4 */ - z3 = MULTIPLY(tmp0 - tmp3, FIX(1.573898926)) + /* c2 */ - MULTIPLY(tmp6 - tmp5, FIX(1.076671805)) + /* c8 */ - MULTIPLY(tmp1 - tmp4, FIX(0.899492312)); /* (c6+c12)/2 */ - - dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(z1 + z3, CONST_BITS+2); - dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(z2 + z3, CONST_BITS+2); - - /* Odd part */ - - tmp2 = MULTIPLY(tmp10 - tmp12 - tmp13 + tmp15 + tmp16, - FIX(1.393487498)); /* c5 */ - tmp1 = MULTIPLY(tmp10 - tmp14 - tmp15, FIX(1.530307725)) + /* c3 */ - MULTIPLY(tmp11 - tmp13 - tmp16, FIX(0.945782187)); /* c9 */ - tmp12 = MULTIPLY(tmp12, FIX(1.393487498)); /* c5 */ - tmp4 = MULTIPLY(tmp10 - tmp16, FIX(1.600246161)) + /* c1 */ - MULTIPLY(tmp11 + tmp14, FIX(1.530307725)) + /* c3 */ - MULTIPLY(tmp13 + tmp15, FIX(0.654463974)); /* c11 */ - tmp0 = MULTIPLY(tmp13, FIX(0.541301207)) - /* c7-c11 */ - MULTIPLY(tmp14, FIX(0.584525538)) + /* c3-c9 */ - MULTIPLY(tmp16, FIX(1.934788705)) + tmp4 + tmp12; /* c1+c13 */ - tmp3 = MULTIPLY(tmp10, - FIX(0.404480980)) - /* -(c1-c7) */ - MULTIPLY(tmp11, FIX(2.476089912)) - /* c3+c9 */ - MULTIPLY(tmp15, FIX(0.989006518)) + tmp4 - tmp12; /* c11+c13 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp0, CONST_BITS+2); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp1, CONST_BITS+2); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp2, CONST_BITS+2); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp3, CONST_BITS+2); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 16x16 sample block. - */ - -GLOBAL(void) -jpeg_fdct_16x16 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16, tmp17; - DCTELEM workspace[DCTSIZE2]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* cK represents sqrt(2) * cos(K*pi/32). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[15]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[14]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[13]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[12]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[11]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[10]); - tmp6 = GETJSAMPLE(elemptr[6]) + GETJSAMPLE(elemptr[9]); - tmp7 = GETJSAMPLE(elemptr[7]) + GETJSAMPLE(elemptr[8]); - - tmp10 = tmp0 + tmp7; - tmp14 = tmp0 - tmp7; - tmp11 = tmp1 + tmp6; - tmp15 = tmp1 - tmp6; - tmp12 = tmp2 + tmp5; - tmp16 = tmp2 - tmp5; - tmp13 = tmp3 + tmp4; - tmp17 = tmp3 - tmp4; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[15]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[14]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[13]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[12]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[11]); - tmp5 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[10]); - tmp6 = GETJSAMPLE(elemptr[6]) - GETJSAMPLE(elemptr[9]); - tmp7 = GETJSAMPLE(elemptr[7]) - GETJSAMPLE(elemptr[8]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 + tmp12 + tmp13 - 16 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(1.306562965)) + /* c4[16] = c2[8] */ - MULTIPLY(tmp11 - tmp12, FIX_0_541196100), /* c12[16] = c6[8] */ - CONST_BITS-PASS1_BITS); - - tmp10 = MULTIPLY(tmp17 - tmp15, FIX(0.275899379)) + /* c14[16] = c7[8] */ - MULTIPLY(tmp14 - tmp16, FIX(1.387039845)); /* c2[16] = c1[8] */ - - dataptr[2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp15, FIX(1.451774982)) /* c6+c14 */ - + MULTIPLY(tmp16, FIX(2.172734804)), /* c2+c10 */ - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(0.211164243)) /* c2-c6 */ - - MULTIPLY(tmp17, FIX(1.061594338)), /* c10+c14 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp11 = MULTIPLY(tmp0 + tmp1, FIX(1.353318001)) + /* c3 */ - MULTIPLY(tmp6 - tmp7, FIX(0.410524528)); /* c13 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(1.247225013)) + /* c5 */ - MULTIPLY(tmp5 + tmp7, FIX(0.666655658)); /* c11 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(1.093201867)) + /* c7 */ - MULTIPLY(tmp4 - tmp7, FIX(0.897167586)); /* c9 */ - tmp14 = MULTIPLY(tmp1 + tmp2, FIX(0.138617169)) + /* c15 */ - MULTIPLY(tmp6 - tmp5, FIX(1.407403738)); /* c1 */ - tmp15 = MULTIPLY(tmp1 + tmp3, - FIX(0.666655658)) + /* -c11 */ - MULTIPLY(tmp4 + tmp6, - FIX(1.247225013)); /* -c5 */ - tmp16 = MULTIPLY(tmp2 + tmp3, - FIX(1.353318001)) + /* -c3 */ - MULTIPLY(tmp5 - tmp4, FIX(0.410524528)); /* c13 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(tmp0, FIX(2.286341144)) + /* c7+c5+c3-c1 */ - MULTIPLY(tmp7, FIX(0.779653625)); /* c15+c13-c11+c9 */ - tmp11 += tmp14 + tmp15 + MULTIPLY(tmp1, FIX(0.071888074)) /* c9-c3-c15+c11 */ - - MULTIPLY(tmp6, FIX(1.663905119)); /* c7+c13+c1-c5 */ - tmp12 += tmp14 + tmp16 - MULTIPLY(tmp2, FIX(1.125726048)) /* c7+c5+c15-c3 */ - + MULTIPLY(tmp5, FIX(1.227391138)); /* c9-c11+c1-c13 */ - tmp13 += tmp15 + tmp16 + MULTIPLY(tmp3, FIX(1.065388962)) /* c15+c3+c11-c7 */ - + MULTIPLY(tmp4, FIX(2.167985692)); /* c1+c13+c5-c9 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp10, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp11, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp12, CONST_BITS-PASS1_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp13, CONST_BITS-PASS1_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == DCTSIZE * 2) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/16)**2 = 1/2**2. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*4]; - tmp4 = dataptr[DCTSIZE*4] + wsptr[DCTSIZE*3]; - tmp5 = dataptr[DCTSIZE*5] + wsptr[DCTSIZE*2]; - tmp6 = dataptr[DCTSIZE*6] + wsptr[DCTSIZE*1]; - tmp7 = dataptr[DCTSIZE*7] + wsptr[DCTSIZE*0]; - - tmp10 = tmp0 + tmp7; - tmp14 = tmp0 - tmp7; - tmp11 = tmp1 + tmp6; - tmp15 = tmp1 - tmp6; - tmp12 = tmp2 + tmp5; - tmp16 = tmp2 - tmp5; - tmp13 = tmp3 + tmp4; - tmp17 = tmp3 - tmp4; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*4]; - tmp4 = dataptr[DCTSIZE*4] - wsptr[DCTSIZE*3]; - tmp5 = dataptr[DCTSIZE*5] - wsptr[DCTSIZE*2]; - tmp6 = dataptr[DCTSIZE*6] - wsptr[DCTSIZE*1]; - tmp7 = dataptr[DCTSIZE*7] - wsptr[DCTSIZE*0]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(tmp10 + tmp11 + tmp12 + tmp13, PASS1_BITS+2); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(1.306562965)) + /* c4[16] = c2[8] */ - MULTIPLY(tmp11 - tmp12, FIX_0_541196100), /* c12[16] = c6[8] */ - CONST_BITS+PASS1_BITS+2); - - tmp10 = MULTIPLY(tmp17 - tmp15, FIX(0.275899379)) + /* c14[16] = c7[8] */ - MULTIPLY(tmp14 - tmp16, FIX(1.387039845)); /* c2[16] = c1[8] */ - - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp15, FIX(1.451774982)) /* c6+c14 */ - + MULTIPLY(tmp16, FIX(2.172734804)), /* c2+10 */ - CONST_BITS+PASS1_BITS+2); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(0.211164243)) /* c2-c6 */ - - MULTIPLY(tmp17, FIX(1.061594338)), /* c10+c14 */ - CONST_BITS+PASS1_BITS+2); - - /* Odd part */ - - tmp11 = MULTIPLY(tmp0 + tmp1, FIX(1.353318001)) + /* c3 */ - MULTIPLY(tmp6 - tmp7, FIX(0.410524528)); /* c13 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(1.247225013)) + /* c5 */ - MULTIPLY(tmp5 + tmp7, FIX(0.666655658)); /* c11 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(1.093201867)) + /* c7 */ - MULTIPLY(tmp4 - tmp7, FIX(0.897167586)); /* c9 */ - tmp14 = MULTIPLY(tmp1 + tmp2, FIX(0.138617169)) + /* c15 */ - MULTIPLY(tmp6 - tmp5, FIX(1.407403738)); /* c1 */ - tmp15 = MULTIPLY(tmp1 + tmp3, - FIX(0.666655658)) + /* -c11 */ - MULTIPLY(tmp4 + tmp6, - FIX(1.247225013)); /* -c5 */ - tmp16 = MULTIPLY(tmp2 + tmp3, - FIX(1.353318001)) + /* -c3 */ - MULTIPLY(tmp5 - tmp4, FIX(0.410524528)); /* c13 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(tmp0, FIX(2.286341144)) + /* c7+c5+c3-c1 */ - MULTIPLY(tmp7, FIX(0.779653625)); /* c15+c13-c11+c9 */ - tmp11 += tmp14 + tmp15 + MULTIPLY(tmp1, FIX(0.071888074)) /* c9-c3-c15+c11 */ - - MULTIPLY(tmp6, FIX(1.663905119)); /* c7+c13+c1-c5 */ - tmp12 += tmp14 + tmp16 - MULTIPLY(tmp2, FIX(1.125726048)) /* c7+c5+c15-c3 */ - + MULTIPLY(tmp5, FIX(1.227391138)); /* c9-c11+c1-c13 */ - tmp13 += tmp15 + tmp16 + MULTIPLY(tmp3, FIX(1.065388962)) /* c15+c3+c11-c7 */ - + MULTIPLY(tmp4, FIX(2.167985692)); /* c1+c13+c5-c9 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp10, CONST_BITS+PASS1_BITS+2); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp11, CONST_BITS+PASS1_BITS+2); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp12, CONST_BITS+PASS1_BITS+2); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp13, CONST_BITS+PASS1_BITS+2); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 16x8 sample block. - * - * 16-point FDCT in pass 1 (rows), 8-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_16x8 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16, tmp17; - INT32 z1; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 16-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/32). */ - - dataptr = data; - ctr = 0; - for (ctr = 0; ctr < DCTSIZE; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[15]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[14]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[13]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[12]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[11]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[10]); - tmp6 = GETJSAMPLE(elemptr[6]) + GETJSAMPLE(elemptr[9]); - tmp7 = GETJSAMPLE(elemptr[7]) + GETJSAMPLE(elemptr[8]); - - tmp10 = tmp0 + tmp7; - tmp14 = tmp0 - tmp7; - tmp11 = tmp1 + tmp6; - tmp15 = tmp1 - tmp6; - tmp12 = tmp2 + tmp5; - tmp16 = tmp2 - tmp5; - tmp13 = tmp3 + tmp4; - tmp17 = tmp3 - tmp4; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[15]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[14]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[13]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[12]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[11]); - tmp5 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[10]); - tmp6 = GETJSAMPLE(elemptr[6]) - GETJSAMPLE(elemptr[9]); - tmp7 = GETJSAMPLE(elemptr[7]) - GETJSAMPLE(elemptr[8]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 + tmp12 + tmp13 - 16 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(1.306562965)) + /* c4[16] = c2[8] */ - MULTIPLY(tmp11 - tmp12, FIX_0_541196100), /* c12[16] = c6[8] */ - CONST_BITS-PASS1_BITS); - - tmp10 = MULTIPLY(tmp17 - tmp15, FIX(0.275899379)) + /* c14[16] = c7[8] */ - MULTIPLY(tmp14 - tmp16, FIX(1.387039845)); /* c2[16] = c1[8] */ - - dataptr[2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp15, FIX(1.451774982)) /* c6+c14 */ - + MULTIPLY(tmp16, FIX(2.172734804)), /* c2+c10 */ - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(0.211164243)) /* c2-c6 */ - - MULTIPLY(tmp17, FIX(1.061594338)), /* c10+c14 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp11 = MULTIPLY(tmp0 + tmp1, FIX(1.353318001)) + /* c3 */ - MULTIPLY(tmp6 - tmp7, FIX(0.410524528)); /* c13 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(1.247225013)) + /* c5 */ - MULTIPLY(tmp5 + tmp7, FIX(0.666655658)); /* c11 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(1.093201867)) + /* c7 */ - MULTIPLY(tmp4 - tmp7, FIX(0.897167586)); /* c9 */ - tmp14 = MULTIPLY(tmp1 + tmp2, FIX(0.138617169)) + /* c15 */ - MULTIPLY(tmp6 - tmp5, FIX(1.407403738)); /* c1 */ - tmp15 = MULTIPLY(tmp1 + tmp3, - FIX(0.666655658)) + /* -c11 */ - MULTIPLY(tmp4 + tmp6, - FIX(1.247225013)); /* -c5 */ - tmp16 = MULTIPLY(tmp2 + tmp3, - FIX(1.353318001)) + /* -c3 */ - MULTIPLY(tmp5 - tmp4, FIX(0.410524528)); /* c13 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(tmp0, FIX(2.286341144)) + /* c7+c5+c3-c1 */ - MULTIPLY(tmp7, FIX(0.779653625)); /* c15+c13-c11+c9 */ - tmp11 += tmp14 + tmp15 + MULTIPLY(tmp1, FIX(0.071888074)) /* c9-c3-c15+c11 */ - - MULTIPLY(tmp6, FIX(1.663905119)); /* c7+c13+c1-c5 */ - tmp12 += tmp14 + tmp16 - MULTIPLY(tmp2, FIX(1.125726048)) /* c7+c5+c15-c3 */ - + MULTIPLY(tmp5, FIX(1.227391138)); /* c9-c11+c1-c13 */ - tmp13 += tmp15 + tmp16 + MULTIPLY(tmp3, FIX(1.065388962)) /* c15+c3+c11-c7 */ - + MULTIPLY(tmp4, FIX(2.167985692)); /* c1+c13+c5-c9 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp10, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp11, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp12, CONST_BITS-PASS1_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp13, CONST_BITS-PASS1_BITS); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by 8/16 = 1/2. - */ - - dataptr = data; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part per LL&M figure 1 --- note that published figure is faulty; - * rotator "sqrt(2)*c1" should be "sqrt(2)*c6". - */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*4]; - - tmp10 = tmp0 + tmp3; - tmp12 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp13 = tmp1 - tmp2; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*4]; - - dataptr[DCTSIZE*0] = (DCTELEM) DESCALE(tmp10 + tmp11, PASS1_BITS+1); - dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(tmp10 - tmp11, PASS1_BITS+1); - - z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); - dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, FIX_0_765366865), - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(z1 - MULTIPLY(tmp13, FIX_1_847759065), - CONST_BITS+PASS1_BITS+1); - - /* Odd part per figure 8 --- note paper omits factor of sqrt(2). - * 8-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/16). - * i0..i3 in the paper are tmp0..tmp3 here. - */ - - tmp10 = tmp0 + tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp0 + tmp2; - tmp13 = tmp1 + tmp3; - z1 = MULTIPLY(tmp12 + tmp13, FIX_1_175875602); /* c3 */ - - tmp0 = MULTIPLY(tmp0, FIX_1_501321110); /* c1+c3-c5-c7 */ - tmp1 = MULTIPLY(tmp1, FIX_3_072711026); /* c1+c3+c5-c7 */ - tmp2 = MULTIPLY(tmp2, FIX_2_053119869); /* c1+c3-c5+c7 */ - tmp3 = MULTIPLY(tmp3, FIX_0_298631336); /* -c1+c3+c5-c7 */ - tmp10 = MULTIPLY(tmp10, - FIX_0_899976223); /* c7-c3 */ - tmp11 = MULTIPLY(tmp11, - FIX_2_562915447); /* -c1-c3 */ - tmp12 = MULTIPLY(tmp12, - FIX_0_390180644); /* c5-c3 */ - tmp13 = MULTIPLY(tmp13, - FIX_1_961570560); /* -c3-c5 */ - - tmp12 += z1; - tmp13 += z1; - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp0 + tmp10 + tmp12, - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp1 + tmp11 + tmp13, - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp2 + tmp11 + tmp12, - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp3 + tmp10 + tmp13, - CONST_BITS+PASS1_BITS+1); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 14x7 sample block. - * - * 14-point FDCT in pass 1 (rows), 7-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_14x7 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16; - INT32 z1, z2, z3; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Zero bottom row of output coefficient block. */ - MEMZERO(&data[DCTSIZE*7], SIZEOF(DCTELEM) * DCTSIZE); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 14-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/28). */ - - dataptr = data; - for (ctr = 0; ctr < 7; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[13]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[12]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[11]); - tmp13 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[10]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[9]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[8]); - tmp6 = GETJSAMPLE(elemptr[6]) + GETJSAMPLE(elemptr[7]); - - tmp10 = tmp0 + tmp6; - tmp14 = tmp0 - tmp6; - tmp11 = tmp1 + tmp5; - tmp15 = tmp1 - tmp5; - tmp12 = tmp2 + tmp4; - tmp16 = tmp2 - tmp4; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[13]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[12]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[11]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[10]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[9]); - tmp5 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[8]); - tmp6 = GETJSAMPLE(elemptr[6]) - GETJSAMPLE(elemptr[7]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 + tmp12 + tmp13 - 14 * CENTERJSAMPLE) << PASS1_BITS); - tmp13 += tmp13; - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(1.274162392)) + /* c4 */ - MULTIPLY(tmp11 - tmp13, FIX(0.314692123)) - /* c12 */ - MULTIPLY(tmp12 - tmp13, FIX(0.881747734)), /* c8 */ - CONST_BITS-PASS1_BITS); - - tmp10 = MULTIPLY(tmp14 + tmp15, FIX(1.105676686)); /* c6 */ - - dataptr[2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp14, FIX(0.273079590)) /* c2-c6 */ - + MULTIPLY(tmp16, FIX(0.613604268)), /* c10 */ - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp15, FIX(1.719280954)) /* c6+c10 */ - - MULTIPLY(tmp16, FIX(1.378756276)), /* c2 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp10 = tmp1 + tmp2; - tmp11 = tmp5 - tmp4; - dataptr[7] = (DCTELEM) ((tmp0 - tmp10 + tmp3 - tmp11 - tmp6) << PASS1_BITS); - tmp3 <<= CONST_BITS; - tmp10 = MULTIPLY(tmp10, - FIX(0.158341681)); /* -c13 */ - tmp11 = MULTIPLY(tmp11, FIX(1.405321284)); /* c1 */ - tmp10 += tmp11 - tmp3; - tmp11 = MULTIPLY(tmp0 + tmp2, FIX(1.197448846)) + /* c5 */ - MULTIPLY(tmp4 + tmp6, FIX(0.752406978)); /* c9 */ - dataptr[5] = (DCTELEM) - DESCALE(tmp10 + tmp11 - MULTIPLY(tmp2, FIX(2.373959773)) /* c3+c5-c13 */ - + MULTIPLY(tmp4, FIX(1.119999435)), /* c1+c11-c9 */ - CONST_BITS-PASS1_BITS); - tmp12 = MULTIPLY(tmp0 + tmp1, FIX(1.334852607)) + /* c3 */ - MULTIPLY(tmp5 - tmp6, FIX(0.467085129)); /* c11 */ - dataptr[3] = (DCTELEM) - DESCALE(tmp10 + tmp12 - MULTIPLY(tmp1, FIX(0.424103948)) /* c3-c9-c13 */ - - MULTIPLY(tmp5, FIX(3.069855259)), /* c1+c5+c11 */ - CONST_BITS-PASS1_BITS); - dataptr[1] = (DCTELEM) - DESCALE(tmp11 + tmp12 + tmp3 + tmp6 - - MULTIPLY(tmp0 + tmp6, FIX(1.126980169)), /* c3+c5-c1 */ - CONST_BITS-PASS1_BITS); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/14)*(8/7) = 32/49, which we - * partially fold into the constant multipliers and final shifting: - * 7-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/14) * 64/49. - */ - - dataptr = data; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*6]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*5]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*4]; - tmp3 = dataptr[DCTSIZE*3]; - - tmp10 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*6]; - tmp11 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*5]; - tmp12 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*4]; - - z1 = tmp0 + tmp2; - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(z1 + tmp1 + tmp3, FIX(1.306122449)), /* 64/49 */ - CONST_BITS+PASS1_BITS+1); - tmp3 += tmp3; - z1 -= tmp3; - z1 -= tmp3; - z1 = MULTIPLY(z1, FIX(0.461784020)); /* (c2+c6-c4)/2 */ - z2 = MULTIPLY(tmp0 - tmp2, FIX(1.202428084)); /* (c2+c4-c6)/2 */ - z3 = MULTIPLY(tmp1 - tmp2, FIX(0.411026446)); /* c6 */ - dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(z1 + z2 + z3, CONST_BITS+PASS1_BITS+1); - z1 -= z2; - z2 = MULTIPLY(tmp0 - tmp1, FIX(1.151670509)); /* c4 */ - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(z2 + z3 - MULTIPLY(tmp1 - tmp3, FIX(0.923568041)), /* c2+c6-c4 */ - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(z1 + z2, CONST_BITS+PASS1_BITS+1); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(1.221765677)); /* (c3+c1-c5)/2 */ - tmp2 = MULTIPLY(tmp10 - tmp11, FIX(0.222383464)); /* (c3+c5-c1)/2 */ - tmp0 = tmp1 - tmp2; - tmp1 += tmp2; - tmp2 = MULTIPLY(tmp11 + tmp12, - FIX(1.800824523)); /* -c1 */ - tmp1 += tmp2; - tmp3 = MULTIPLY(tmp10 + tmp12, FIX(0.801442310)); /* c5 */ - tmp0 += tmp3; - tmp2 += tmp3 + MULTIPLY(tmp12, FIX(2.443531355)); /* c3+c1-c5 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp0, CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp1, CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp2, CONST_BITS+PASS1_BITS+1); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 12x6 sample block. - * - * 12-point FDCT in pass 1 (rows), 6-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_12x6 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Zero 2 bottom rows of output coefficient block. */ - MEMZERO(&data[DCTSIZE*6], SIZEOF(DCTELEM) * DCTSIZE * 2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 12-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/24). */ - - dataptr = data; - for (ctr = 0; ctr < 6; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[11]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[10]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[9]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[8]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[7]); - tmp5 = GETJSAMPLE(elemptr[5]) + GETJSAMPLE(elemptr[6]); - - tmp10 = tmp0 + tmp5; - tmp13 = tmp0 - tmp5; - tmp11 = tmp1 + tmp4; - tmp14 = tmp1 - tmp4; - tmp12 = tmp2 + tmp3; - tmp15 = tmp2 - tmp3; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[11]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[10]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[9]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[8]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[7]); - tmp5 = GETJSAMPLE(elemptr[5]) - GETJSAMPLE(elemptr[6]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 + tmp12 - 12 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[6] = (DCTELEM) ((tmp13 - tmp14 - tmp15) << PASS1_BITS); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.224744871)), /* c4 */ - CONST_BITS-PASS1_BITS); - dataptr[2] = (DCTELEM) - DESCALE(tmp14 - tmp15 + MULTIPLY(tmp13 + tmp15, FIX(1.366025404)), /* c2 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp1 + tmp4, FIX_0_541196100); /* c9 */ - tmp14 = tmp10 + MULTIPLY(tmp1, FIX_0_765366865); /* c3-c9 */ - tmp15 = tmp10 - MULTIPLY(tmp4, FIX_1_847759065); /* c3+c9 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(1.121971054)); /* c5 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(0.860918669)); /* c7 */ - tmp10 = tmp12 + tmp13 + tmp14 - MULTIPLY(tmp0, FIX(0.580774953)) /* c5+c7-c1 */ - + MULTIPLY(tmp5, FIX(0.184591911)); /* c11 */ - tmp11 = MULTIPLY(tmp2 + tmp3, - FIX(0.184591911)); /* -c11 */ - tmp12 += tmp11 - tmp15 - MULTIPLY(tmp2, FIX(2.339493912)) /* c1+c5-c11 */ - + MULTIPLY(tmp5, FIX(0.860918669)); /* c7 */ - tmp13 += tmp11 - tmp14 + MULTIPLY(tmp3, FIX(0.725788011)) /* c1+c11-c7 */ - - MULTIPLY(tmp5, FIX(1.121971054)); /* c5 */ - tmp11 = tmp15 + MULTIPLY(tmp0 - tmp3, FIX(1.306562965)) /* c3 */ - - MULTIPLY(tmp2 + tmp5, FIX_0_541196100); /* c9 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp10, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp11, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp12, CONST_BITS-PASS1_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp13, CONST_BITS-PASS1_BITS); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/12)*(8/6) = 8/9, which we - * partially fold into the constant multipliers and final shifting: - * 6-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/12) * 16/9. - */ - - dataptr = data; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*5]; - tmp11 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*3]; - - tmp10 = tmp0 + tmp2; - tmp12 = tmp0 - tmp2; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*3]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp12, FIX(2.177324216)), /* c2 */ - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp11, FIX(1.257078722)), /* c4 */ - CONST_BITS+PASS1_BITS+1); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp2, FIX(0.650711829)); /* c5 */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0 + tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp2, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp2 - tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS+1); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 10x5 sample block. - * - * 10-point FDCT in pass 1 (rows), 5-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_10x5 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Zero 3 bottom rows of output coefficient block. */ - MEMZERO(&data[DCTSIZE*5], SIZEOF(DCTELEM) * DCTSIZE * 3); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 10-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/20). */ - - dataptr = data; - for (ctr = 0; ctr < 5; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[9]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[8]); - tmp12 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[7]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[6]); - tmp4 = GETJSAMPLE(elemptr[4]) + GETJSAMPLE(elemptr[5]); - - tmp10 = tmp0 + tmp4; - tmp13 = tmp0 - tmp4; - tmp11 = tmp1 + tmp3; - tmp14 = tmp1 - tmp3; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[9]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[8]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[7]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[6]); - tmp4 = GETJSAMPLE(elemptr[4]) - GETJSAMPLE(elemptr[5]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 + tmp12 - 10 * CENTERJSAMPLE) << PASS1_BITS); - tmp12 += tmp12; - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.144122806)) - /* c4 */ - MULTIPLY(tmp11 - tmp12, FIX(0.437016024)), /* c8 */ - CONST_BITS-PASS1_BITS); - tmp10 = MULTIPLY(tmp13 + tmp14, FIX(0.831253876)); /* c6 */ - dataptr[2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp13, FIX(0.513743148)), /* c2-c6 */ - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(2.176250899)), /* c2+c6 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp10 = tmp0 + tmp4; - tmp11 = tmp1 - tmp3; - dataptr[5] = (DCTELEM) ((tmp10 - tmp11 - tmp2) << PASS1_BITS); - tmp2 <<= CONST_BITS; - dataptr[1] = (DCTELEM) - DESCALE(MULTIPLY(tmp0, FIX(1.396802247)) + /* c1 */ - MULTIPLY(tmp1, FIX(1.260073511)) + tmp2 + /* c3 */ - MULTIPLY(tmp3, FIX(0.642039522)) + /* c7 */ - MULTIPLY(tmp4, FIX(0.221231742)), /* c9 */ - CONST_BITS-PASS1_BITS); - tmp12 = MULTIPLY(tmp0 - tmp4, FIX(0.951056516)) - /* (c3+c7)/2 */ - MULTIPLY(tmp1 + tmp3, FIX(0.587785252)); /* (c1-c9)/2 */ - tmp13 = MULTIPLY(tmp10 + tmp11, FIX(0.309016994)) + /* (c3-c7)/2 */ - (tmp11 << (CONST_BITS - 1)) - tmp2; - dataptr[3] = (DCTELEM) DESCALE(tmp12 + tmp13, CONST_BITS-PASS1_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp12 - tmp13, CONST_BITS-PASS1_BITS); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/10)*(8/5) = 32/25, which we - * fold into the constant multipliers: - * 5-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/10) * 32/25. - */ - - dataptr = data; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*4]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*3]; - tmp2 = dataptr[DCTSIZE*2]; - - tmp10 = tmp0 + tmp1; - tmp11 = tmp0 - tmp1; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*4]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*3]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp2, FIX(1.28)), /* 32/25 */ - CONST_BITS+PASS1_BITS); - tmp11 = MULTIPLY(tmp11, FIX(1.011928851)); /* (c2+c4)/2 */ - tmp10 -= tmp2 << 2; - tmp10 = MULTIPLY(tmp10, FIX(0.452548340)); /* (c2-c4)/2 */ - dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(tmp11 + tmp10, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(tmp11 - tmp10, CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp1, FIX(1.064004961)); /* c3 */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0, FIX(0.657591230)), /* c1-c3 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp1, FIX(2.785601151)), /* c1+c3 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on an 8x4 sample block. - * - * 8-point FDCT in pass 1 (rows), 4-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_8x4 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3; - INT32 tmp10, tmp11, tmp12, tmp13; - INT32 z1; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Zero 4 bottom rows of output coefficient block. */ - MEMZERO(&data[DCTSIZE*4], SIZEOF(DCTELEM) * DCTSIZE * 4); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We must also scale the output by 8/4 = 2, which we add here. */ - - dataptr = data; - for (ctr = 0; ctr < 4; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part per LL&M figure 1 --- note that published figure is faulty; - * rotator "sqrt(2)*c1" should be "sqrt(2)*c6". - */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[7]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[6]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[5]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[4]); - - tmp10 = tmp0 + tmp3; - tmp12 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp13 = tmp1 - tmp2; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[7]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[6]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[5]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[4]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 - 8 * CENTERJSAMPLE) << (PASS1_BITS+1)); - dataptr[4] = (DCTELEM) ((tmp10 - tmp11) << (PASS1_BITS+1)); - - z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-2); - dataptr[2] = (DCTELEM) RIGHT_SHIFT(z1 + MULTIPLY(tmp12, FIX_0_765366865), - CONST_BITS-PASS1_BITS-1); - dataptr[6] = (DCTELEM) RIGHT_SHIFT(z1 - MULTIPLY(tmp13, FIX_1_847759065), - CONST_BITS-PASS1_BITS-1); - - /* Odd part per figure 8 --- note paper omits factor of sqrt(2). - * 8-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/16). - * i0..i3 in the paper are tmp0..tmp3 here. - */ - - tmp10 = tmp0 + tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp0 + tmp2; - tmp13 = tmp1 + tmp3; - z1 = MULTIPLY(tmp12 + tmp13, FIX_1_175875602); /* c3 */ - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-2); - - tmp0 = MULTIPLY(tmp0, FIX_1_501321110); /* c1+c3-c5-c7 */ - tmp1 = MULTIPLY(tmp1, FIX_3_072711026); /* c1+c3+c5-c7 */ - tmp2 = MULTIPLY(tmp2, FIX_2_053119869); /* c1+c3-c5+c7 */ - tmp3 = MULTIPLY(tmp3, FIX_0_298631336); /* -c1+c3+c5-c7 */ - tmp10 = MULTIPLY(tmp10, - FIX_0_899976223); /* c7-c3 */ - tmp11 = MULTIPLY(tmp11, - FIX_2_562915447); /* -c1-c3 */ - tmp12 = MULTIPLY(tmp12, - FIX_0_390180644); /* c5-c3 */ - tmp13 = MULTIPLY(tmp13, - FIX_1_961570560); /* -c3-c5 */ - - tmp12 += z1; - tmp13 += z1; - - dataptr[1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + tmp10 + tmp12, CONST_BITS-PASS1_BITS-1); - dataptr[3] = (DCTELEM) - RIGHT_SHIFT(tmp1 + tmp11 + tmp13, CONST_BITS-PASS1_BITS-1); - dataptr[5] = (DCTELEM) - RIGHT_SHIFT(tmp2 + tmp11 + tmp12, CONST_BITS-PASS1_BITS-1); - dataptr[7] = (DCTELEM) - RIGHT_SHIFT(tmp3 + tmp10 + tmp13, CONST_BITS-PASS1_BITS-1); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * 4-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/16). - */ - - dataptr = data; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*3] + (ONE << (PASS1_BITS-1)); - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*2]; - - tmp10 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*3]; - tmp11 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*2]; - - dataptr[DCTSIZE*0] = (DCTELEM) RIGHT_SHIFT(tmp0 + tmp1, PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) RIGHT_SHIFT(tmp0 - tmp1, PASS1_BITS); - - /* Odd part */ - - tmp0 = MULTIPLY(tmp10 + tmp11, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS+PASS1_BITS-1); - - dataptr[DCTSIZE*1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + MULTIPLY(tmp10, FIX_0_765366865), /* c2-c6 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - RIGHT_SHIFT(tmp0 - MULTIPLY(tmp11, FIX_1_847759065), /* c2+c6 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 6x3 sample block. - * - * 6-point FDCT in pass 1 (rows), 3-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_6x3 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2; - INT32 tmp10, tmp11, tmp12; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We scale the results further by 2 as part of output adaption */ - /* scaling for different DCT size. */ - /* 6-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/12). */ - - dataptr = data; - for (ctr = 0; ctr < 3; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[5]); - tmp11 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[4]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[3]); - - tmp10 = tmp0 + tmp2; - tmp12 = tmp0 - tmp2; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[5]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[4]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[3]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 - 6 * CENTERJSAMPLE) << (PASS1_BITS+1)); - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp12, FIX(1.224744871)), /* c2 */ - CONST_BITS-PASS1_BITS-1); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp11, FIX(0.707106781)), /* c4 */ - CONST_BITS-PASS1_BITS-1); - - /* Odd part */ - - tmp10 = DESCALE(MULTIPLY(tmp0 + tmp2, FIX(0.366025404)), /* c5 */ - CONST_BITS-PASS1_BITS-1); - - dataptr[1] = (DCTELEM) (tmp10 + ((tmp0 + tmp1) << (PASS1_BITS+1))); - dataptr[3] = (DCTELEM) ((tmp0 - tmp1 - tmp2) << (PASS1_BITS+1)); - dataptr[5] = (DCTELEM) (tmp10 + ((tmp2 - tmp1) << (PASS1_BITS+1))); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/6)*(8/3) = 32/9, which we partially - * fold into the constant multipliers (other part was done in pass 1): - * 3-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/6) * 16/9. - */ - - dataptr = data; - for (ctr = 0; ctr < 6; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*2]; - tmp1 = dataptr[DCTSIZE*1]; - - tmp2 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*2]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 + tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp1, FIX(1.257078722)), /* c2 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(MULTIPLY(tmp2, FIX(2.177324216)), /* c1 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 4x2 sample block. - * - * 4-point FDCT in pass 1 (rows), 2-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_4x2 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1; - INT32 tmp10, tmp11; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We must also scale the output by (8/4)*(8/2) = 2**3, which we add here. */ - /* 4-point FDCT kernel, */ - /* cK represents sqrt(2) * cos(K*pi/16) [refers to 8-point FDCT]. */ - - dataptr = data; - for (ctr = 0; ctr < 2; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[3]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[2]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[3]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[2]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp0 + tmp1 - 4 * CENTERJSAMPLE) << (PASS1_BITS+3)); - dataptr[2] = (DCTELEM) ((tmp0 - tmp1) << (PASS1_BITS+3)); - - /* Odd part */ - - tmp0 = MULTIPLY(tmp10 + tmp11, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-PASS1_BITS-4); - - dataptr[1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + MULTIPLY(tmp10, FIX_0_765366865), /* c2-c6 */ - CONST_BITS-PASS1_BITS-3); - dataptr[3] = (DCTELEM) - RIGHT_SHIFT(tmp0 - MULTIPLY(tmp11, FIX_1_847759065), /* c2+c6 */ - CONST_BITS-PASS1_BITS-3); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - */ - - dataptr = data; - for (ctr = 0; ctr < 4; ctr++) { - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = dataptr[DCTSIZE*0] + (ONE << (PASS1_BITS-1)); - tmp1 = dataptr[DCTSIZE*1]; - - dataptr[DCTSIZE*0] = (DCTELEM) RIGHT_SHIFT(tmp0 + tmp1, PASS1_BITS); - - /* Odd part */ - - dataptr[DCTSIZE*1] = (DCTELEM) RIGHT_SHIFT(tmp0 - tmp1, PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 2x1 sample block. - * - * 2-point FDCT in pass 1 (rows), 1-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_2x1 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1; - JSAMPROW elemptr; - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - elemptr = sample_data[0] + start_col; - - tmp0 = GETJSAMPLE(elemptr[0]); - tmp1 = GETJSAMPLE(elemptr[1]); - - /* We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/2)*(8/1) = 2**5. - */ - - /* Even part */ - /* Apply unsigned->signed conversion */ - data[0] = (DCTELEM) ((tmp0 + tmp1 - 2 * CENTERJSAMPLE) << 5); - - /* Odd part */ - data[1] = (DCTELEM) ((tmp0 - tmp1) << 5); -} - - -/* - * Perform the forward DCT on an 8x16 sample block. - * - * 8-point FDCT in pass 1 (rows), 16-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_8x16 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16, tmp17; - INT32 z1; - DCTELEM workspace[DCTSIZE2]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part per LL&M figure 1 --- note that published figure is faulty; - * rotator "sqrt(2)*c1" should be "sqrt(2)*c6". - */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[7]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[6]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[5]); - tmp3 = GETJSAMPLE(elemptr[3]) + GETJSAMPLE(elemptr[4]); - - tmp10 = tmp0 + tmp3; - tmp12 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp13 = tmp1 - tmp2; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[7]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[6]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[5]); - tmp3 = GETJSAMPLE(elemptr[3]) - GETJSAMPLE(elemptr[4]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) ((tmp10 + tmp11 - 8 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[4] = (DCTELEM) ((tmp10 - tmp11) << PASS1_BITS); - - z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); - dataptr[2] = (DCTELEM) DESCALE(z1 + MULTIPLY(tmp12, FIX_0_765366865), - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) DESCALE(z1 - MULTIPLY(tmp13, FIX_1_847759065), - CONST_BITS-PASS1_BITS); - - /* Odd part per figure 8 --- note paper omits factor of sqrt(2). - * 8-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/16). - * i0..i3 in the paper are tmp0..tmp3 here. - */ - - tmp10 = tmp0 + tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp0 + tmp2; - tmp13 = tmp1 + tmp3; - z1 = MULTIPLY(tmp12 + tmp13, FIX_1_175875602); /* c3 */ - - tmp0 = MULTIPLY(tmp0, FIX_1_501321110); /* c1+c3-c5-c7 */ - tmp1 = MULTIPLY(tmp1, FIX_3_072711026); /* c1+c3+c5-c7 */ - tmp2 = MULTIPLY(tmp2, FIX_2_053119869); /* c1+c3-c5+c7 */ - tmp3 = MULTIPLY(tmp3, FIX_0_298631336); /* -c1+c3+c5-c7 */ - tmp10 = MULTIPLY(tmp10, - FIX_0_899976223); /* c7-c3 */ - tmp11 = MULTIPLY(tmp11, - FIX_2_562915447); /* -c1-c3 */ - tmp12 = MULTIPLY(tmp12, - FIX_0_390180644); /* c5-c3 */ - tmp13 = MULTIPLY(tmp13, - FIX_1_961570560); /* -c3-c5 */ - - tmp12 += z1; - tmp13 += z1; - - dataptr[1] = (DCTELEM) DESCALE(tmp0 + tmp10 + tmp12, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp1 + tmp11 + tmp13, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp2 + tmp11 + tmp12, CONST_BITS-PASS1_BITS); - dataptr[7] = (DCTELEM) DESCALE(tmp3 + tmp10 + tmp13, CONST_BITS-PASS1_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == DCTSIZE * 2) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by 8/16 = 1/2. - * 16-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/32). - */ - - dataptr = data; - wsptr = workspace; - for (ctr = DCTSIZE-1; ctr >= 0; ctr--) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*4]; - tmp4 = dataptr[DCTSIZE*4] + wsptr[DCTSIZE*3]; - tmp5 = dataptr[DCTSIZE*5] + wsptr[DCTSIZE*2]; - tmp6 = dataptr[DCTSIZE*6] + wsptr[DCTSIZE*1]; - tmp7 = dataptr[DCTSIZE*7] + wsptr[DCTSIZE*0]; - - tmp10 = tmp0 + tmp7; - tmp14 = tmp0 - tmp7; - tmp11 = tmp1 + tmp6; - tmp15 = tmp1 - tmp6; - tmp12 = tmp2 + tmp5; - tmp16 = tmp2 - tmp5; - tmp13 = tmp3 + tmp4; - tmp17 = tmp3 - tmp4; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*4]; - tmp4 = dataptr[DCTSIZE*4] - wsptr[DCTSIZE*3]; - tmp5 = dataptr[DCTSIZE*5] - wsptr[DCTSIZE*2]; - tmp6 = dataptr[DCTSIZE*6] - wsptr[DCTSIZE*1]; - tmp7 = dataptr[DCTSIZE*7] - wsptr[DCTSIZE*0]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(tmp10 + tmp11 + tmp12 + tmp13, PASS1_BITS+1); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(1.306562965)) + /* c4[16] = c2[8] */ - MULTIPLY(tmp11 - tmp12, FIX_0_541196100), /* c12[16] = c6[8] */ - CONST_BITS+PASS1_BITS+1); - - tmp10 = MULTIPLY(tmp17 - tmp15, FIX(0.275899379)) + /* c14[16] = c7[8] */ - MULTIPLY(tmp14 - tmp16, FIX(1.387039845)); /* c2[16] = c1[8] */ - - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp15, FIX(1.451774982)) /* c6+c14 */ - + MULTIPLY(tmp16, FIX(2.172734804)), /* c2+c10 */ - CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(0.211164243)) /* c2-c6 */ - - MULTIPLY(tmp17, FIX(1.061594338)), /* c10+c14 */ - CONST_BITS+PASS1_BITS+1); - - /* Odd part */ - - tmp11 = MULTIPLY(tmp0 + tmp1, FIX(1.353318001)) + /* c3 */ - MULTIPLY(tmp6 - tmp7, FIX(0.410524528)); /* c13 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(1.247225013)) + /* c5 */ - MULTIPLY(tmp5 + tmp7, FIX(0.666655658)); /* c11 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(1.093201867)) + /* c7 */ - MULTIPLY(tmp4 - tmp7, FIX(0.897167586)); /* c9 */ - tmp14 = MULTIPLY(tmp1 + tmp2, FIX(0.138617169)) + /* c15 */ - MULTIPLY(tmp6 - tmp5, FIX(1.407403738)); /* c1 */ - tmp15 = MULTIPLY(tmp1 + tmp3, - FIX(0.666655658)) + /* -c11 */ - MULTIPLY(tmp4 + tmp6, - FIX(1.247225013)); /* -c5 */ - tmp16 = MULTIPLY(tmp2 + tmp3, - FIX(1.353318001)) + /* -c3 */ - MULTIPLY(tmp5 - tmp4, FIX(0.410524528)); /* c13 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(tmp0, FIX(2.286341144)) + /* c7+c5+c3-c1 */ - MULTIPLY(tmp7, FIX(0.779653625)); /* c15+c13-c11+c9 */ - tmp11 += tmp14 + tmp15 + MULTIPLY(tmp1, FIX(0.071888074)) /* c9-c3-c15+c11 */ - - MULTIPLY(tmp6, FIX(1.663905119)); /* c7+c13+c1-c5 */ - tmp12 += tmp14 + tmp16 - MULTIPLY(tmp2, FIX(1.125726048)) /* c7+c5+c15-c3 */ - + MULTIPLY(tmp5, FIX(1.227391138)); /* c9-c11+c1-c13 */ - tmp13 += tmp15 + tmp16 + MULTIPLY(tmp3, FIX(1.065388962)) /* c15+c3+c11-c7 */ - + MULTIPLY(tmp4, FIX(2.167985692)); /* c1+c13+c5-c9 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp10, CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp11, CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp12, CONST_BITS+PASS1_BITS+1); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp13, CONST_BITS+PASS1_BITS+1); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 7x14 sample block. - * - * 7-point FDCT in pass 1 (rows), 14-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_7x14 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16; - INT32 z1, z2, z3; - DCTELEM workspace[8*6]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 7-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/14). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[6]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[5]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[4]); - tmp3 = GETJSAMPLE(elemptr[3]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[6]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[5]); - tmp12 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[4]); - - z1 = tmp0 + tmp2; - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((z1 + tmp1 + tmp3 - 7 * CENTERJSAMPLE) << PASS1_BITS); - tmp3 += tmp3; - z1 -= tmp3; - z1 -= tmp3; - z1 = MULTIPLY(z1, FIX(0.353553391)); /* (c2+c6-c4)/2 */ - z2 = MULTIPLY(tmp0 - tmp2, FIX(0.920609002)); /* (c2+c4-c6)/2 */ - z3 = MULTIPLY(tmp1 - tmp2, FIX(0.314692123)); /* c6 */ - dataptr[2] = (DCTELEM) DESCALE(z1 + z2 + z3, CONST_BITS-PASS1_BITS); - z1 -= z2; - z2 = MULTIPLY(tmp0 - tmp1, FIX(0.881747734)); /* c4 */ - dataptr[4] = (DCTELEM) - DESCALE(z2 + z3 - MULTIPLY(tmp1 - tmp3, FIX(0.707106781)), /* c2+c6-c4 */ - CONST_BITS-PASS1_BITS); - dataptr[6] = (DCTELEM) DESCALE(z1 + z2, CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp1 = MULTIPLY(tmp10 + tmp11, FIX(0.935414347)); /* (c3+c1-c5)/2 */ - tmp2 = MULTIPLY(tmp10 - tmp11, FIX(0.170262339)); /* (c3+c5-c1)/2 */ - tmp0 = tmp1 - tmp2; - tmp1 += tmp2; - tmp2 = MULTIPLY(tmp11 + tmp12, - FIX(1.378756276)); /* -c1 */ - tmp1 += tmp2; - tmp3 = MULTIPLY(tmp10 + tmp12, FIX(0.613604268)); /* c5 */ - tmp0 += tmp3; - tmp2 += tmp3 + MULTIPLY(tmp12, FIX(1.870828693)); /* c3+c1-c5 */ - - dataptr[1] = (DCTELEM) DESCALE(tmp0, CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) DESCALE(tmp1, CONST_BITS-PASS1_BITS); - dataptr[5] = (DCTELEM) DESCALE(tmp2, CONST_BITS-PASS1_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 14) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/7)*(8/14) = 32/49, which we - * fold into the constant multipliers: - * 14-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/28) * 32/49. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = 0; ctr < 7; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*3]; - tmp13 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*2]; - tmp4 = dataptr[DCTSIZE*4] + wsptr[DCTSIZE*1]; - tmp5 = dataptr[DCTSIZE*5] + wsptr[DCTSIZE*0]; - tmp6 = dataptr[DCTSIZE*6] + dataptr[DCTSIZE*7]; - - tmp10 = tmp0 + tmp6; - tmp14 = tmp0 - tmp6; - tmp11 = tmp1 + tmp5; - tmp15 = tmp1 - tmp5; - tmp12 = tmp2 + tmp4; - tmp16 = tmp2 - tmp4; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*3]; - tmp3 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*2]; - tmp4 = dataptr[DCTSIZE*4] - wsptr[DCTSIZE*1]; - tmp5 = dataptr[DCTSIZE*5] - wsptr[DCTSIZE*0]; - tmp6 = dataptr[DCTSIZE*6] - dataptr[DCTSIZE*7]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11 + tmp12 + tmp13, - FIX(0.653061224)), /* 32/49 */ - CONST_BITS+PASS1_BITS); - tmp13 += tmp13; - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp13, FIX(0.832106052)) + /* c4 */ - MULTIPLY(tmp11 - tmp13, FIX(0.205513223)) - /* c12 */ - MULTIPLY(tmp12 - tmp13, FIX(0.575835255)), /* c8 */ - CONST_BITS+PASS1_BITS); - - tmp10 = MULTIPLY(tmp14 + tmp15, FIX(0.722074570)); /* c6 */ - - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp14, FIX(0.178337691)) /* c2-c6 */ - + MULTIPLY(tmp16, FIX(0.400721155)), /* c10 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp15, FIX(1.122795725)) /* c6+c10 */ - - MULTIPLY(tmp16, FIX(0.900412262)), /* c2 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = tmp1 + tmp2; - tmp11 = tmp5 - tmp4; - dataptr[DCTSIZE*7] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp10 + tmp3 - tmp11 - tmp6, - FIX(0.653061224)), /* 32/49 */ - CONST_BITS+PASS1_BITS); - tmp3 = MULTIPLY(tmp3 , FIX(0.653061224)); /* 32/49 */ - tmp10 = MULTIPLY(tmp10, - FIX(0.103406812)); /* -c13 */ - tmp11 = MULTIPLY(tmp11, FIX(0.917760839)); /* c1 */ - tmp10 += tmp11 - tmp3; - tmp11 = MULTIPLY(tmp0 + tmp2, FIX(0.782007410)) + /* c5 */ - MULTIPLY(tmp4 + tmp6, FIX(0.491367823)); /* c9 */ - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(tmp10 + tmp11 - MULTIPLY(tmp2, FIX(1.550341076)) /* c3+c5-c13 */ - + MULTIPLY(tmp4, FIX(0.731428202)), /* c1+c11-c9 */ - CONST_BITS+PASS1_BITS); - tmp12 = MULTIPLY(tmp0 + tmp1, FIX(0.871740478)) + /* c3 */ - MULTIPLY(tmp5 - tmp6, FIX(0.305035186)); /* c11 */ - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(tmp10 + tmp12 - MULTIPLY(tmp1, FIX(0.276965844)) /* c3-c9-c13 */ - - MULTIPLY(tmp5, FIX(2.004803435)), /* c1+c5+c11 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp11 + tmp12 + tmp3 - - MULTIPLY(tmp0, FIX(0.735987049)) /* c3+c5-c1 */ - - MULTIPLY(tmp6, FIX(0.082925825)), /* c9-c11-c13 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 6x12 sample block. - * - * 6-point FDCT in pass 1 (rows), 12-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_6x12 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15; - DCTELEM workspace[8*4]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 6-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/12). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[5]); - tmp11 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[4]); - tmp2 = GETJSAMPLE(elemptr[2]) + GETJSAMPLE(elemptr[3]); - - tmp10 = tmp0 + tmp2; - tmp12 = tmp0 - tmp2; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[5]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[4]); - tmp2 = GETJSAMPLE(elemptr[2]) - GETJSAMPLE(elemptr[3]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp11 - 6 * CENTERJSAMPLE) << PASS1_BITS); - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp12, FIX(1.224744871)), /* c2 */ - CONST_BITS-PASS1_BITS); - dataptr[4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp11, FIX(0.707106781)), /* c4 */ - CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp10 = DESCALE(MULTIPLY(tmp0 + tmp2, FIX(0.366025404)), /* c5 */ - CONST_BITS-PASS1_BITS); - - dataptr[1] = (DCTELEM) (tmp10 + ((tmp0 + tmp1) << PASS1_BITS)); - dataptr[3] = (DCTELEM) ((tmp0 - tmp1 - tmp2) << PASS1_BITS); - dataptr[5] = (DCTELEM) (tmp10 + ((tmp2 - tmp1) << PASS1_BITS)); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 12) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/6)*(8/12) = 8/9, which we - * fold into the constant multipliers: - * 12-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/24) * 8/9. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = 0; ctr < 6; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*3]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*2]; - tmp2 = dataptr[DCTSIZE*2] + wsptr[DCTSIZE*1]; - tmp3 = dataptr[DCTSIZE*3] + wsptr[DCTSIZE*0]; - tmp4 = dataptr[DCTSIZE*4] + dataptr[DCTSIZE*7]; - tmp5 = dataptr[DCTSIZE*5] + dataptr[DCTSIZE*6]; - - tmp10 = tmp0 + tmp5; - tmp13 = tmp0 - tmp5; - tmp11 = tmp1 + tmp4; - tmp14 = tmp1 - tmp4; - tmp12 = tmp2 + tmp3; - tmp15 = tmp2 - tmp3; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*3]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*2]; - tmp2 = dataptr[DCTSIZE*2] - wsptr[DCTSIZE*1]; - tmp3 = dataptr[DCTSIZE*3] - wsptr[DCTSIZE*0]; - tmp4 = dataptr[DCTSIZE*4] - dataptr[DCTSIZE*7]; - tmp5 = dataptr[DCTSIZE*5] - dataptr[DCTSIZE*6]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11 + tmp12, FIX(0.888888889)), /* 8/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(MULTIPLY(tmp13 - tmp14 - tmp15, FIX(0.888888889)), /* 8/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.088662108)), /* c4 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp14 - tmp15, FIX(0.888888889)) + /* 8/9 */ - MULTIPLY(tmp13 + tmp15, FIX(1.214244803)), /* c2 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp1 + tmp4, FIX(0.481063200)); /* c9 */ - tmp14 = tmp10 + MULTIPLY(tmp1, FIX(0.680326102)); /* c3-c9 */ - tmp15 = tmp10 - MULTIPLY(tmp4, FIX(1.642452502)); /* c3+c9 */ - tmp12 = MULTIPLY(tmp0 + tmp2, FIX(0.997307603)); /* c5 */ - tmp13 = MULTIPLY(tmp0 + tmp3, FIX(0.765261039)); /* c7 */ - tmp10 = tmp12 + tmp13 + tmp14 - MULTIPLY(tmp0, FIX(0.516244403)) /* c5+c7-c1 */ - + MULTIPLY(tmp5, FIX(0.164081699)); /* c11 */ - tmp11 = MULTIPLY(tmp2 + tmp3, - FIX(0.164081699)); /* -c11 */ - tmp12 += tmp11 - tmp15 - MULTIPLY(tmp2, FIX(2.079550144)) /* c1+c5-c11 */ - + MULTIPLY(tmp5, FIX(0.765261039)); /* c7 */ - tmp13 += tmp11 - tmp14 + MULTIPLY(tmp3, FIX(0.645144899)) /* c1+c11-c7 */ - - MULTIPLY(tmp5, FIX(0.997307603)); /* c5 */ - tmp11 = tmp15 + MULTIPLY(tmp0 - tmp3, FIX(1.161389302)) /* c3 */ - - MULTIPLY(tmp2 + tmp5, FIX(0.481063200)); /* c9 */ - - dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp10, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp11, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp12, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp13, CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 5x10 sample block. - * - * 5-point FDCT in pass 1 (rows), 10-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_5x10 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4; - INT32 tmp10, tmp11, tmp12, tmp13, tmp14; - DCTELEM workspace[8*2]; - DCTELEM *dataptr; - DCTELEM *wsptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* 5-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/10). */ - - dataptr = data; - ctr = 0; - for (;;) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[4]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[3]); - tmp2 = GETJSAMPLE(elemptr[2]); - - tmp10 = tmp0 + tmp1; - tmp11 = tmp0 - tmp1; - - tmp0 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[4]); - tmp1 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[3]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp10 + tmp2 - 5 * CENTERJSAMPLE) << PASS1_BITS); - tmp11 = MULTIPLY(tmp11, FIX(0.790569415)); /* (c2+c4)/2 */ - tmp10 -= tmp2 << 2; - tmp10 = MULTIPLY(tmp10, FIX(0.353553391)); /* (c2-c4)/2 */ - dataptr[2] = (DCTELEM) DESCALE(tmp11 + tmp10, CONST_BITS-PASS1_BITS); - dataptr[4] = (DCTELEM) DESCALE(tmp11 - tmp10, CONST_BITS-PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp1, FIX(0.831253876)); /* c3 */ - - dataptr[1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0, FIX(0.513743148)), /* c1-c3 */ - CONST_BITS-PASS1_BITS); - dataptr[3] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp1, FIX(2.176250899)), /* c1+c3 */ - CONST_BITS-PASS1_BITS); - - ctr++; - - if (ctr != DCTSIZE) { - if (ctr == 10) - break; /* Done. */ - dataptr += DCTSIZE; /* advance pointer to next row */ - } else - dataptr = workspace; /* switch pointer to extended workspace */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/5)*(8/10) = 32/25, which we - * fold into the constant multipliers: - * 10-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/20) * 32/25. - */ - - dataptr = data; - wsptr = workspace; - for (ctr = 0; ctr < 5; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + wsptr[DCTSIZE*1]; - tmp1 = dataptr[DCTSIZE*1] + wsptr[DCTSIZE*0]; - tmp12 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*7]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*6]; - tmp4 = dataptr[DCTSIZE*4] + dataptr[DCTSIZE*5]; - - tmp10 = tmp0 + tmp4; - tmp13 = tmp0 - tmp4; - tmp11 = tmp1 + tmp3; - tmp14 = tmp1 - tmp3; - - tmp0 = dataptr[DCTSIZE*0] - wsptr[DCTSIZE*1]; - tmp1 = dataptr[DCTSIZE*1] - wsptr[DCTSIZE*0]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*7]; - tmp3 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*6]; - tmp4 = dataptr[DCTSIZE*4] - dataptr[DCTSIZE*5]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11 + tmp12, FIX(1.28)), /* 32/25 */ - CONST_BITS+PASS1_BITS); - tmp12 += tmp12; - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp12, FIX(1.464477191)) - /* c4 */ - MULTIPLY(tmp11 - tmp12, FIX(0.559380511)), /* c8 */ - CONST_BITS+PASS1_BITS); - tmp10 = MULTIPLY(tmp13 + tmp14, FIX(1.064004961)); /* c6 */ - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp13, FIX(0.657591230)), /* c2-c6 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*6] = (DCTELEM) - DESCALE(tmp10 - MULTIPLY(tmp14, FIX(2.785601151)), /* c2+c6 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = tmp0 + tmp4; - tmp11 = tmp1 - tmp3; - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp2, FIX(1.28)), /* 32/25 */ - CONST_BITS+PASS1_BITS); - tmp2 = MULTIPLY(tmp2, FIX(1.28)); /* 32/25 */ - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(MULTIPLY(tmp0, FIX(1.787906876)) + /* c1 */ - MULTIPLY(tmp1, FIX(1.612894094)) + tmp2 + /* c3 */ - MULTIPLY(tmp3, FIX(0.821810588)) + /* c7 */ - MULTIPLY(tmp4, FIX(0.283176630)), /* c9 */ - CONST_BITS+PASS1_BITS); - tmp12 = MULTIPLY(tmp0 - tmp4, FIX(1.217352341)) - /* (c3+c7)/2 */ - MULTIPLY(tmp1 + tmp3, FIX(0.752365123)); /* (c1-c9)/2 */ - tmp13 = MULTIPLY(tmp10 + tmp11, FIX(0.395541753)) + /* (c3-c7)/2 */ - MULTIPLY(tmp11, FIX(0.64)) - tmp2; /* 16/25 */ - dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp12 + tmp13, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp12 - tmp13, CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - wsptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 4x8 sample block. - * - * 4-point FDCT in pass 1 (rows), 8-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_4x8 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3; - INT32 tmp10, tmp11, tmp12, tmp13; - INT32 z1; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We must also scale the output by 8/4 = 2, which we add here. */ - /* 4-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/16). */ - - dataptr = data; - for (ctr = 0; ctr < DCTSIZE; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[3]); - tmp1 = GETJSAMPLE(elemptr[1]) + GETJSAMPLE(elemptr[2]); - - tmp10 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[3]); - tmp11 = GETJSAMPLE(elemptr[1]) - GETJSAMPLE(elemptr[2]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp0 + tmp1 - 4 * CENTERJSAMPLE) << (PASS1_BITS+1)); - dataptr[2] = (DCTELEM) ((tmp0 - tmp1) << (PASS1_BITS+1)); - - /* Odd part */ - - tmp0 = MULTIPLY(tmp10 + tmp11, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-PASS1_BITS-2); - - dataptr[1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + MULTIPLY(tmp10, FIX_0_765366865), /* c2-c6 */ - CONST_BITS-PASS1_BITS-1); - dataptr[3] = (DCTELEM) - RIGHT_SHIFT(tmp0 - MULTIPLY(tmp11, FIX_1_847759065), /* c2+c6 */ - CONST_BITS-PASS1_BITS-1); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - */ - - dataptr = data; - for (ctr = 0; ctr < 4; ctr++) { - /* Even part per LL&M figure 1 --- note that published figure is faulty; - * rotator "sqrt(2)*c1" should be "sqrt(2)*c6". - */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*4]; - - /* Add fudge factor here for final descale. */ - tmp10 = tmp0 + tmp3 + (ONE << (PASS1_BITS-1)); - tmp12 = tmp0 - tmp3; - tmp11 = tmp1 + tmp2; - tmp13 = tmp1 - tmp2; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*7]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*6]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*5]; - tmp3 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*4]; - - dataptr[DCTSIZE*0] = (DCTELEM) RIGHT_SHIFT(tmp10 + tmp11, PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) RIGHT_SHIFT(tmp10 - tmp11, PASS1_BITS); - - z1 = MULTIPLY(tmp12 + tmp13, FIX_0_541196100); - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS+PASS1_BITS-1); - dataptr[DCTSIZE*2] = (DCTELEM) - RIGHT_SHIFT(z1 + MULTIPLY(tmp12, FIX_0_765366865), CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*6] = (DCTELEM) - RIGHT_SHIFT(z1 - MULTIPLY(tmp13, FIX_1_847759065), CONST_BITS+PASS1_BITS); - - /* Odd part per figure 8 --- note paper omits factor of sqrt(2). - * 8-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/16). - * i0..i3 in the paper are tmp0..tmp3 here. - */ - - tmp10 = tmp0 + tmp3; - tmp11 = tmp1 + tmp2; - tmp12 = tmp0 + tmp2; - tmp13 = tmp1 + tmp3; - z1 = MULTIPLY(tmp12 + tmp13, FIX_1_175875602); /* c3 */ - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS+PASS1_BITS-1); - - tmp0 = MULTIPLY(tmp0, FIX_1_501321110); /* c1+c3-c5-c7 */ - tmp1 = MULTIPLY(tmp1, FIX_3_072711026); /* c1+c3+c5-c7 */ - tmp2 = MULTIPLY(tmp2, FIX_2_053119869); /* c1+c3-c5+c7 */ - tmp3 = MULTIPLY(tmp3, FIX_0_298631336); /* -c1+c3+c5-c7 */ - tmp10 = MULTIPLY(tmp10, - FIX_0_899976223); /* c7-c3 */ - tmp11 = MULTIPLY(tmp11, - FIX_2_562915447); /* -c1-c3 */ - tmp12 = MULTIPLY(tmp12, - FIX_0_390180644); /* c5-c3 */ - tmp13 = MULTIPLY(tmp13, - FIX_1_961570560); /* -c3-c5 */ - - tmp12 += z1; - tmp13 += z1; - - dataptr[DCTSIZE*1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + tmp10 + tmp12, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - RIGHT_SHIFT(tmp1 + tmp11 + tmp13, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*5] = (DCTELEM) - RIGHT_SHIFT(tmp2 + tmp11 + tmp12, CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*7] = (DCTELEM) - RIGHT_SHIFT(tmp3 + tmp10 + tmp13, CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 3x6 sample block. - * - * 3-point FDCT in pass 1 (rows), 6-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_3x6 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1, tmp2; - INT32 tmp10, tmp11, tmp12; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - /* We scale the results further by 2 as part of output adaption */ - /* scaling for different DCT size. */ - /* 3-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/6). */ - - dataptr = data; - for (ctr = 0; ctr < 6; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]) + GETJSAMPLE(elemptr[2]); - tmp1 = GETJSAMPLE(elemptr[1]); - - tmp2 = GETJSAMPLE(elemptr[0]) - GETJSAMPLE(elemptr[2]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) - ((tmp0 + tmp1 - 3 * CENTERJSAMPLE) << (PASS1_BITS+1)); - dataptr[2] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp1, FIX(0.707106781)), /* c2 */ - CONST_BITS-PASS1_BITS-1); - - /* Odd part */ - - dataptr[1] = (DCTELEM) - DESCALE(MULTIPLY(tmp2, FIX(1.224744871)), /* c1 */ - CONST_BITS-PASS1_BITS-1); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We remove the PASS1_BITS scaling, but leave the results scaled up - * by an overall factor of 8. - * We must also scale the output by (8/6)*(8/3) = 32/9, which we partially - * fold into the constant multipliers (other part was done in pass 1): - * 6-point FDCT kernel, cK represents sqrt(2) * cos(K*pi/12) * 16/9. - */ - - dataptr = data; - for (ctr = 0; ctr < 3; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*5]; - tmp11 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*3]; - - tmp10 = tmp0 + tmp2; - tmp12 = tmp0 - tmp2; - - tmp0 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*5]; - tmp1 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*4]; - tmp2 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*3]; - - dataptr[DCTSIZE*0] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 + tmp11, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*2] = (DCTELEM) - DESCALE(MULTIPLY(tmp12, FIX(2.177324216)), /* c2 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*4] = (DCTELEM) - DESCALE(MULTIPLY(tmp10 - tmp11 - tmp11, FIX(1.257078722)), /* c4 */ - CONST_BITS+PASS1_BITS); - - /* Odd part */ - - tmp10 = MULTIPLY(tmp0 + tmp2, FIX(0.650711829)); /* c5 */ - - dataptr[DCTSIZE*1] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp0 + tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - DESCALE(MULTIPLY(tmp0 - tmp1 - tmp2, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - dataptr[DCTSIZE*5] = (DCTELEM) - DESCALE(tmp10 + MULTIPLY(tmp2 - tmp1, FIX(1.777777778)), /* 16/9 */ - CONST_BITS+PASS1_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 2x4 sample block. - * - * 2-point FDCT in pass 1 (rows), 4-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_2x4 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1; - INT32 tmp10, tmp11; - DCTELEM *dataptr; - JSAMPROW elemptr; - int ctr; - SHIFT_TEMPS - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - /* Pass 1: process rows. */ - /* Note results are scaled up by sqrt(8) compared to a true DCT. */ - /* We must also scale the output by (8/2)*(8/4) = 2**3, which we add here. */ - - dataptr = data; - for (ctr = 0; ctr < 4; ctr++) { - elemptr = sample_data[ctr] + start_col; - - /* Even part */ - - tmp0 = GETJSAMPLE(elemptr[0]); - tmp1 = GETJSAMPLE(elemptr[1]); - - /* Apply unsigned->signed conversion */ - dataptr[0] = (DCTELEM) ((tmp0 + tmp1 - 2 * CENTERJSAMPLE) << 3); - - /* Odd part */ - - dataptr[1] = (DCTELEM) ((tmp0 - tmp1) << 3); - - dataptr += DCTSIZE; /* advance pointer to next row */ - } - - /* Pass 2: process columns. - * We leave the results scaled up by an overall factor of 8. - * 4-point FDCT kernel, - * cK represents sqrt(2) * cos(K*pi/16) [refers to 8-point FDCT]. - */ - - dataptr = data; - for (ctr = 0; ctr < 2; ctr++) { - /* Even part */ - - tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*3]; - tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*2]; - - tmp10 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*3]; - tmp11 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*2]; - - dataptr[DCTSIZE*0] = (DCTELEM) (tmp0 + tmp1); - dataptr[DCTSIZE*2] = (DCTELEM) (tmp0 - tmp1); - - /* Odd part */ - - tmp0 = MULTIPLY(tmp10 + tmp11, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-1); - - dataptr[DCTSIZE*1] = (DCTELEM) - RIGHT_SHIFT(tmp0 + MULTIPLY(tmp10, FIX_0_765366865), /* c2-c6 */ - CONST_BITS); - dataptr[DCTSIZE*3] = (DCTELEM) - RIGHT_SHIFT(tmp0 - MULTIPLY(tmp11, FIX_1_847759065), /* c2+c6 */ - CONST_BITS); - - dataptr++; /* advance pointer to next column */ - } -} - - -/* - * Perform the forward DCT on a 1x2 sample block. - * - * 1-point FDCT in pass 1 (rows), 2-point in pass 2 (columns). - */ - -GLOBAL(void) -jpeg_fdct_1x2 (DCTELEM * data, JSAMPARRAY sample_data, JDIMENSION start_col) -{ - INT32 tmp0, tmp1; - - /* Pre-zero output coefficient block. */ - MEMZERO(data, SIZEOF(DCTELEM) * DCTSIZE2); - - tmp0 = GETJSAMPLE(sample_data[0][start_col]); - tmp1 = GETJSAMPLE(sample_data[1][start_col]); - - /* We leave the results scaled up by an overall factor of 8. - * We must also scale the output by (8/1)*(8/2) = 2**5. - */ - - /* Even part */ - /* Apply unsigned->signed conversion */ - data[DCTSIZE*0] = (DCTELEM) ((tmp0 + tmp1 - 2 * CENTERJSAMPLE) << 5); - - /* Odd part */ - data[DCTSIZE*1] = (DCTELEM) ((tmp0 - tmp1) << 5); -} - -#endif /* DCT_SCALING_SUPPORTED */ -#endif /* DCT_ISLOW_SUPPORTED */ diff --git a/src/3rdparty/libjpeg/jidctint.c b/src/3rdparty/libjpeg/jidctint.c deleted file mode 100644 index dcdf7ce454..0000000000 --- a/src/3rdparty/libjpeg/jidctint.c +++ /dev/null @@ -1,5137 +0,0 @@ -/* - * jidctint.c - * - * Copyright (C) 1991-1998, Thomas G. Lane. - * Modification developed 2002-2009 by Guido Vollbeding. - * This file is part of the Independent JPEG Group's software. - * For conditions of distribution and use, see the accompanying README file. - * - * This file contains a slow-but-accurate integer implementation of the - * inverse DCT (Discrete Cosine Transform). In the IJG code, this routine - * must also perform dequantization of the input coefficients. - * - * 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 an algorithm described in - * C. Loeffler, A. Ligtenberg and G. Moschytz, "Practical Fast 1-D DCT - * Algorithms with 11 Multiplications", Proc. Int'l. Conf. on Acoustics, - * Speech, and Signal Processing 1989 (ICASSP '89), pp. 988-991. - * The primary algorithm described there uses 11 multiplies and 29 adds. - * We use their alternate method with 12 multiplies and 32 adds. - * The advantage of this method is that no data path contains more than one - * multiplication; this allows a very simple and accurate implementation in - * scaled fixed-point arithmetic, with a minimal number of shifts. - * - * We also provide IDCT routines with various output sample block sizes for - * direct resolution reduction or enlargement and for direct resolving the - * common 2x1 and 1x2 subsampling cases without additional resampling: NxN - * (N=1...16), 2NxN, and Nx2N (N=1...8) pixels for one 8x8 input DCT block. - * - * For N<8 we simply take the corresponding low-frequency coefficients of - * the 8x8 input DCT block and apply an NxN point IDCT on the sub-block - * to yield the downscaled outputs. - * This can be seen as direct low-pass downsampling from the DCT domain - * point of view rather than the usual spatial domain point of view, - * yielding significant computational savings and results at least - * as good as common bilinear (averaging) spatial downsampling. - * - * For N>8 we apply a partial NxN IDCT on the 8 input coefficients as - * lower frequencies and higher frequencies assumed to be zero. - * It turns out that the computational effort is similar to the 8x8 IDCT - * regarding the output size. - * Furthermore, the scaling and descaling is the same for all IDCT sizes. - * - * CAUTION: We rely on the FIX() macro except for the N=1,2,4,8 cases - * since there would be too many additional constants to pre-calculate. - */ - -#define JPEG_INTERNALS -#include "jinclude.h" -#include "jpeglib.h" -#include "jdct.h" /* Private declarations for DCT subsystem */ - -#ifdef DCT_ISLOW_SUPPORTED - - -/* - * This module is specialized to the case DCTSIZE = 8. - */ - -#if DCTSIZE != 8 - Sorry, this code only copes with 8x8 DCT blocks. /* deliberate syntax err */ -#endif - - -/* - * The poop on this scaling stuff is as follows: - * - * Each 1-D IDCT step produces outputs which are a factor of sqrt(N) - * larger than the true IDCT outputs. The final outputs are therefore - * a factor of N larger than desired; since N=8 this can be cured by - * a simple right shift at the end of the algorithm. The advantage of - * this arrangement is that we save two multiplications per 1-D IDCT, - * because the y0 and y4 inputs need not be divided by sqrt(N). - * - * We have to do addition and subtraction of the integer inputs, which - * is no problem, and multiplication by fractional constants, which is - * a problem to do in integer arithmetic. We multiply all the constants - * by CONST_SCALE and convert them to integer constants (thus retaining - * CONST_BITS bits of precision in the constants). After doing a - * multiplication we have to divide the product by CONST_SCALE, with proper - * rounding, to produce the correct output. This division can be done - * cheaply as a right shift of CONST_BITS bits. We postpone shifting - * as long as possible so that partial sums can be added together with - * full fractional precision. - * - * The outputs of the first pass are scaled up by PASS1_BITS bits so that - * they are represented to better-than-integral precision. These outputs - * require BITS_IN_JSAMPLE + PASS1_BITS + 3 bits; this fits in a 16-bit word - * with the recommended scaling. (To scale up 12-bit sample data further, an - * intermediate INT32 array would be needed.) - * - * To avoid overflow of the 32-bit intermediate results in pass 2, we must - * have BITS_IN_JSAMPLE + CONST_BITS + PASS1_BITS <= 26. Error analysis - * shows that the values given below are the most effective. - */ - -#if BITS_IN_JSAMPLE == 8 -#define CONST_BITS 13 -#define PASS1_BITS 2 -#else -#define CONST_BITS 13 -#define PASS1_BITS 1 /* lose a little precision to avoid overflow */ -#endif - -/* Some C compilers fail to reduce "FIX(constant)" at compile time, thus - * causing a lot of useless floating-point operations at run time. - * To get around this we use the following pre-calculated constants. - * If you change CONST_BITS you may want to add appropriate values. - * (With a reasonable C compiler, you can just rely on the FIX() macro...) - */ - -#if CONST_BITS == 13 -#define FIX_0_298631336 ((INT32) 2446) /* FIX(0.298631336) */ -#define FIX_0_390180644 ((INT32) 3196) /* FIX(0.390180644) */ -#define FIX_0_541196100 ((INT32) 4433) /* FIX(0.541196100) */ -#define FIX_0_765366865 ((INT32) 6270) /* FIX(0.765366865) */ -#define FIX_0_899976223 ((INT32) 7373) /* FIX(0.899976223) */ -#define FIX_1_175875602 ((INT32) 9633) /* FIX(1.175875602) */ -#define FIX_1_501321110 ((INT32) 12299) /* FIX(1.501321110) */ -#define FIX_1_847759065 ((INT32) 15137) /* FIX(1.847759065) */ -#define FIX_1_961570560 ((INT32) 16069) /* FIX(1.961570560) */ -#define FIX_2_053119869 ((INT32) 16819) /* FIX(2.053119869) */ -#define FIX_2_562915447 ((INT32) 20995) /* FIX(2.562915447) */ -#define FIX_3_072711026 ((INT32) 25172) /* FIX(3.072711026) */ -#else -#define FIX_0_298631336 FIX(0.298631336) -#define FIX_0_390180644 FIX(0.390180644) -#define FIX_0_541196100 FIX(0.541196100) -#define FIX_0_765366865 FIX(0.765366865) -#define FIX_0_899976223 FIX(0.899976223) -#define FIX_1_175875602 FIX(1.175875602) -#define FIX_1_501321110 FIX(1.501321110) -#define FIX_1_847759065 FIX(1.847759065) -#define FIX_1_961570560 FIX(1.961570560) -#define FIX_2_053119869 FIX(2.053119869) -#define FIX_2_562915447 FIX(2.562915447) -#define FIX_3_072711026 FIX(3.072711026) -#endif - - -/* Multiply an INT32 variable by an INT32 constant to yield an INT32 result. - * For 8-bit samples with the recommended scaling, all the variable - * and constant values involved are no more than 16 bits wide, so a - * 16x16->32 bit multiply can be used instead of a full 32x32 multiply. - * For 12-bit samples, a full 32-bit multiplication will be needed. - */ - -#if BITS_IN_JSAMPLE == 8 -#define MULTIPLY(var,const) MULTIPLY16C16(var,const) -#else -#define MULTIPLY(var,const) ((var) * (const)) -#endif - - -/* Dequantize a coefficient by multiplying it by the multiplier-table - * entry; produce an int result. In this module, both inputs and result - * are 16 bits or less, so either int or short multiply will work. - */ - -#define DEQUANTIZE(coef,quantval) (((ISLOW_MULT_TYPE) (coef)) * (quantval)) - - -/* - * Perform dequantization and inverse DCT on one block of coefficients. - */ - -GLOBAL(void) -jpeg_idct_islow (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3; - INT32 tmp10, tmp11, tmp12, tmp13; - INT32 z1, z2, z3; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[DCTSIZE2]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - /* Note results are scaled up by sqrt(8) compared to a true IDCT; */ - /* furthermore, we scale the results by 2**PASS1_BITS. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_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 */ - int dcval = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]) << PASS1_BITS; - - 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: reverse the even part of the forward DCT. */ - /* The rotator is sqrt(2)*c(-6). */ - - z2 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - z1 = MULTIPLY(z2 + z3, FIX_0_541196100); - tmp2 = z1 + MULTIPLY(z2, FIX_0_765366865); - tmp3 = z1 - MULTIPLY(z3, FIX_1_847759065); - - z2 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - z3 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z2 <<= CONST_BITS; - z3 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - z2 += ONE << (CONST_BITS-PASS1_BITS-1); - - tmp0 = z2 + z3; - tmp1 = z2 - z3; - - tmp10 = tmp0 + tmp2; - tmp13 = tmp0 - tmp2; - tmp11 = tmp1 + tmp3; - tmp12 = tmp1 - tmp3; - - /* Odd part per figure 8; the matrix is unitary and hence its - * transpose is its inverse. i0..i3 are y7,y5,y3,y1 respectively. - */ - - tmp0 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - tmp1 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - tmp2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - tmp3 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - - z2 = tmp0 + tmp2; - z3 = tmp1 + tmp3; - - z1 = MULTIPLY(z2 + z3, FIX_1_175875602); /* sqrt(2) * c3 */ - z2 = MULTIPLY(z2, - FIX_1_961570560); /* sqrt(2) * (-c3-c5) */ - z3 = MULTIPLY(z3, - FIX_0_390180644); /* sqrt(2) * (c5-c3) */ - z2 += z1; - z3 += z1; - - z1 = MULTIPLY(tmp0 + tmp3, - FIX_0_899976223); /* sqrt(2) * (c7-c3) */ - tmp0 = MULTIPLY(tmp0, FIX_0_298631336); /* sqrt(2) * (-c1+c3+c5-c7) */ - tmp3 = MULTIPLY(tmp3, FIX_1_501321110); /* sqrt(2) * ( c1+c3-c5-c7) */ - tmp0 += z1 + z2; - tmp3 += z1 + z3; - - z1 = MULTIPLY(tmp1 + tmp2, - FIX_2_562915447); /* sqrt(2) * (-c1-c3) */ - tmp1 = MULTIPLY(tmp1, FIX_2_053119869); /* sqrt(2) * ( c1+c3-c5+c7) */ - tmp2 = MULTIPLY(tmp2, FIX_3_072711026); /* sqrt(2) * ( c1+c3+c5-c7) */ - tmp1 += z1 + z3; - tmp2 += z1 + z2; - - /* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */ - - wsptr[DCTSIZE*0] = (int) RIGHT_SHIFT(tmp10 + tmp3, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*7] = (int) RIGHT_SHIFT(tmp10 - tmp3, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*1] = (int) RIGHT_SHIFT(tmp11 + tmp2, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*6] = (int) RIGHT_SHIFT(tmp11 - tmp2, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*2] = (int) RIGHT_SHIFT(tmp12 + tmp1, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*5] = (int) RIGHT_SHIFT(tmp12 - tmp1, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*3] = (int) RIGHT_SHIFT(tmp13 + tmp0, CONST_BITS-PASS1_BITS); - wsptr[DCTSIZE*4] = (int) RIGHT_SHIFT(tmp13 - tmp0, CONST_BITS-PASS1_BITS); - - 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, */ - /* and also undo the PASS1_BITS scaling. */ - - wsptr = workspace; - for (ctr = 0; ctr < DCTSIZE; ctr++) { - outptr = output_buf[ctr] + output_col; - /* 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). - * On machines with very fast multiplication, it's possible that the - * test takes more time than it's worth. In that case this section - * may be commented out. - */ - -#ifndef NO_ZERO_ROW_TEST - if (wsptr[1] == 0 && wsptr[2] == 0 && wsptr[3] == 0 && wsptr[4] == 0 && - wsptr[5] == 0 && wsptr[6] == 0 && wsptr[7] == 0) { - /* AC terms all zero */ - JSAMPLE dcval = range_limit[(int) DESCALE((INT32) wsptr[0], PASS1_BITS+3) - & RANGE_MASK]; - - outptr[0] = dcval; - outptr[1] = dcval; - outptr[2] = dcval; - outptr[3] = dcval; - outptr[4] = dcval; - outptr[5] = dcval; - outptr[6] = dcval; - outptr[7] = dcval; - - wsptr += DCTSIZE; /* advance pointer to next row */ - continue; - } -#endif - - /* Even part: reverse the even part of the forward DCT. */ - /* The rotator is sqrt(2)*c(-6). */ - - z2 = (INT32) wsptr[2]; - z3 = (INT32) wsptr[6]; - - z1 = MULTIPLY(z2 + z3, FIX_0_541196100); - tmp2 = z1 + MULTIPLY(z2, FIX_0_765366865); - tmp3 = z1 - MULTIPLY(z3, FIX_1_847759065); - - /* Add fudge factor here for final descale. */ - z2 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - z3 = (INT32) wsptr[4]; - - tmp0 = (z2 + z3) << CONST_BITS; - tmp1 = (z2 - z3) << CONST_BITS; - - tmp10 = tmp0 + tmp2; - tmp13 = tmp0 - tmp2; - tmp11 = tmp1 + tmp3; - tmp12 = tmp1 - tmp3; - - /* Odd part per figure 8; the matrix is unitary and hence its - * transpose is its inverse. i0..i3 are y7,y5,y3,y1 respectively. - */ - - tmp0 = (INT32) wsptr[7]; - tmp1 = (INT32) wsptr[5]; - tmp2 = (INT32) wsptr[3]; - tmp3 = (INT32) wsptr[1]; - - z2 = tmp0 + tmp2; - z3 = tmp1 + tmp3; - - z1 = MULTIPLY(z2 + z3, FIX_1_175875602); /* sqrt(2) * c3 */ - z2 = MULTIPLY(z2, - FIX_1_961570560); /* sqrt(2) * (-c3-c5) */ - z3 = MULTIPLY(z3, - FIX_0_390180644); /* sqrt(2) * (c5-c3) */ - z2 += z1; - z3 += z1; - - z1 = MULTIPLY(tmp0 + tmp3, - FIX_0_899976223); /* sqrt(2) * (c7-c3) */ - tmp0 = MULTIPLY(tmp0, FIX_0_298631336); /* sqrt(2) * (-c1+c3+c5-c7) */ - tmp3 = MULTIPLY(tmp3, FIX_1_501321110); /* sqrt(2) * ( c1+c3-c5-c7) */ - tmp0 += z1 + z2; - tmp3 += z1 + z3; - - z1 = MULTIPLY(tmp1 + tmp2, - FIX_2_562915447); /* sqrt(2) * (-c1-c3) */ - tmp1 = MULTIPLY(tmp1, FIX_2_053119869); /* sqrt(2) * ( c1+c3-c5+c7) */ - tmp2 = MULTIPLY(tmp2, FIX_3_072711026); /* sqrt(2) * ( c1+c3+c5-c7) */ - tmp1 += z1 + z3; - tmp2 += z1 + z2; - - /* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp3, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp3, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp11 + tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp11 - tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp12 + tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp12 - tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp13 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp13 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += DCTSIZE; /* advance pointer to next row */ - } -} - -#ifdef IDCT_SCALING_SUPPORTED - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 7x7 output block. - * - * Optimized algorithm with 12 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/14). - */ - -GLOBAL(void) -jpeg_idct_7x7 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp1, tmp2, tmp10, tmp11, tmp12, tmp13; - INT32 z1, z2, z3; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[7*7]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 7; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp13 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp13 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - tmp13 += ONE << (CONST_BITS-PASS1_BITS-1); - - z1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z2 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - tmp10 = MULTIPLY(z2 - z3, FIX(0.881747734)); /* c4 */ - tmp12 = MULTIPLY(z1 - z2, FIX(0.314692123)); /* c6 */ - tmp11 = tmp10 + tmp12 + tmp13 - MULTIPLY(z2, FIX(1.841218003)); /* c2+c4-c6 */ - tmp0 = z1 + z3; - z2 -= tmp0; - tmp0 = MULTIPLY(tmp0, FIX(1.274162392)) + tmp13; /* c2 */ - tmp10 += tmp0 - MULTIPLY(z3, FIX(0.077722536)); /* c2-c4-c6 */ - tmp12 += tmp0 - MULTIPLY(z1, FIX(2.470602249)); /* c2+c4+c6 */ - tmp13 += MULTIPLY(z2, FIX(1.414213562)); /* c0 */ - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - - tmp1 = MULTIPLY(z1 + z2, FIX(0.935414347)); /* (c3+c1-c5)/2 */ - tmp2 = MULTIPLY(z1 - z2, FIX(0.170262339)); /* (c3+c5-c1)/2 */ - tmp0 = tmp1 - tmp2; - tmp1 += tmp2; - tmp2 = MULTIPLY(z2 + z3, - FIX(1.378756276)); /* -c1 */ - tmp1 += tmp2; - z2 = MULTIPLY(z1 + z3, FIX(0.613604268)); /* c5 */ - tmp0 += z2; - tmp2 += z2 + MULTIPLY(z3, FIX(1.870828693)); /* c3+c1-c5 */ - - /* Final output stage */ - - wsptr[7*0] = (int) RIGHT_SHIFT(tmp10 + tmp0, CONST_BITS-PASS1_BITS); - wsptr[7*6] = (int) RIGHT_SHIFT(tmp10 - tmp0, CONST_BITS-PASS1_BITS); - wsptr[7*1] = (int) RIGHT_SHIFT(tmp11 + tmp1, CONST_BITS-PASS1_BITS); - wsptr[7*5] = (int) RIGHT_SHIFT(tmp11 - tmp1, CONST_BITS-PASS1_BITS); - wsptr[7*2] = (int) RIGHT_SHIFT(tmp12 + tmp2, CONST_BITS-PASS1_BITS); - wsptr[7*4] = (int) RIGHT_SHIFT(tmp12 - tmp2, CONST_BITS-PASS1_BITS); - wsptr[7*3] = (int) RIGHT_SHIFT(tmp13, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 7 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 7; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp13 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp13 <<= CONST_BITS; - - z1 = (INT32) wsptr[2]; - z2 = (INT32) wsptr[4]; - z3 = (INT32) wsptr[6]; - - tmp10 = MULTIPLY(z2 - z3, FIX(0.881747734)); /* c4 */ - tmp12 = MULTIPLY(z1 - z2, FIX(0.314692123)); /* c6 */ - tmp11 = tmp10 + tmp12 + tmp13 - MULTIPLY(z2, FIX(1.841218003)); /* c2+c4-c6 */ - tmp0 = z1 + z3; - z2 -= tmp0; - tmp0 = MULTIPLY(tmp0, FIX(1.274162392)) + tmp13; /* c2 */ - tmp10 += tmp0 - MULTIPLY(z3, FIX(0.077722536)); /* c2-c4-c6 */ - tmp12 += tmp0 - MULTIPLY(z1, FIX(2.470602249)); /* c2+c4+c6 */ - tmp13 += MULTIPLY(z2, FIX(1.414213562)); /* c0 */ - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - - tmp1 = MULTIPLY(z1 + z2, FIX(0.935414347)); /* (c3+c1-c5)/2 */ - tmp2 = MULTIPLY(z1 - z2, FIX(0.170262339)); /* (c3+c5-c1)/2 */ - tmp0 = tmp1 - tmp2; - tmp1 += tmp2; - tmp2 = MULTIPLY(z2 + z3, - FIX(1.378756276)); /* -c1 */ - tmp1 += tmp2; - z2 = MULTIPLY(z1 + z3, FIX(0.613604268)); /* c5 */ - tmp0 += z2; - tmp2 += z2 + MULTIPLY(z3, FIX(1.870828693)); /* c3+c1-c5 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp11 + tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp11 - tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp12 + tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp12 - tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 7; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a reduced-size 6x6 output block. - * - * Optimized algorithm with 3 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/12). - */ - -GLOBAL(void) -jpeg_idct_6x6 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp1, tmp2, tmp10, tmp11, tmp12; - INT32 z1, z2, z3; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[6*6]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 6; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp0 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp0 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-PASS1_BITS-1); - tmp2 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - tmp10 = MULTIPLY(tmp2, FIX(0.707106781)); /* c4 */ - tmp1 = tmp0 + tmp10; - tmp11 = RIGHT_SHIFT(tmp0 - tmp10 - tmp10, CONST_BITS-PASS1_BITS); - tmp10 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - tmp0 = MULTIPLY(tmp10, FIX(1.224744871)); /* c2 */ - tmp10 = tmp1 + tmp0; - tmp12 = tmp1 - tmp0; - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - tmp1 = MULTIPLY(z1 + z3, FIX(0.366025404)); /* c5 */ - tmp0 = tmp1 + ((z1 + z2) << CONST_BITS); - tmp2 = tmp1 + ((z3 - z2) << CONST_BITS); - tmp1 = (z1 - z2 - z3) << PASS1_BITS; - - /* Final output stage */ - - wsptr[6*0] = (int) RIGHT_SHIFT(tmp10 + tmp0, CONST_BITS-PASS1_BITS); - wsptr[6*5] = (int) RIGHT_SHIFT(tmp10 - tmp0, CONST_BITS-PASS1_BITS); - wsptr[6*1] = (int) (tmp11 + tmp1); - wsptr[6*4] = (int) (tmp11 - tmp1); - wsptr[6*2] = (int) RIGHT_SHIFT(tmp12 + tmp2, CONST_BITS-PASS1_BITS); - wsptr[6*3] = (int) RIGHT_SHIFT(tmp12 - tmp2, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 6 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 6; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp0 <<= CONST_BITS; - tmp2 = (INT32) wsptr[4]; - tmp10 = MULTIPLY(tmp2, FIX(0.707106781)); /* c4 */ - tmp1 = tmp0 + tmp10; - tmp11 = tmp0 - tmp10 - tmp10; - tmp10 = (INT32) wsptr[2]; - tmp0 = MULTIPLY(tmp10, FIX(1.224744871)); /* c2 */ - tmp10 = tmp1 + tmp0; - tmp12 = tmp1 - tmp0; - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - tmp1 = MULTIPLY(z1 + z3, FIX(0.366025404)); /* c5 */ - tmp0 = tmp1 + ((z1 + z2) << CONST_BITS); - tmp2 = tmp1 + ((z3 - z2) << CONST_BITS); - tmp1 = (z1 - z2 - z3) << CONST_BITS; - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp11 + tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp11 - tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp12 + tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp12 - tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 6; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a reduced-size 5x5 output block. - * - * Optimized algorithm with 5 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/10). - */ - -GLOBAL(void) -jpeg_idct_5x5 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp1, tmp10, tmp11, tmp12; - INT32 z1, z2, z3; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[5*5]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 5; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp12 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp12 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - tmp12 += ONE << (CONST_BITS-PASS1_BITS-1); - tmp0 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - tmp1 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z1 = MULTIPLY(tmp0 + tmp1, FIX(0.790569415)); /* (c2+c4)/2 */ - z2 = MULTIPLY(tmp0 - tmp1, FIX(0.353553391)); /* (c2-c4)/2 */ - z3 = tmp12 + z2; - tmp10 = z3 + z1; - tmp11 = z3 - z1; - tmp12 -= z2 << 2; - - /* Odd part */ - - z2 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z3 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - - z1 = MULTIPLY(z2 + z3, FIX(0.831253876)); /* c3 */ - tmp0 = z1 + MULTIPLY(z2, FIX(0.513743148)); /* c1-c3 */ - tmp1 = z1 - MULTIPLY(z3, FIX(2.176250899)); /* c1+c3 */ - - /* Final output stage */ - - wsptr[5*0] = (int) RIGHT_SHIFT(tmp10 + tmp0, CONST_BITS-PASS1_BITS); - wsptr[5*4] = (int) RIGHT_SHIFT(tmp10 - tmp0, CONST_BITS-PASS1_BITS); - wsptr[5*1] = (int) RIGHT_SHIFT(tmp11 + tmp1, CONST_BITS-PASS1_BITS); - wsptr[5*3] = (int) RIGHT_SHIFT(tmp11 - tmp1, CONST_BITS-PASS1_BITS); - wsptr[5*2] = (int) RIGHT_SHIFT(tmp12, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 5 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 5; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp12 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp12 <<= CONST_BITS; - tmp0 = (INT32) wsptr[2]; - tmp1 = (INT32) wsptr[4]; - z1 = MULTIPLY(tmp0 + tmp1, FIX(0.790569415)); /* (c2+c4)/2 */ - z2 = MULTIPLY(tmp0 - tmp1, FIX(0.353553391)); /* (c2-c4)/2 */ - z3 = tmp12 + z2; - tmp10 = z3 + z1; - tmp11 = z3 - z1; - tmp12 -= z2 << 2; - - /* Odd part */ - - z2 = (INT32) wsptr[1]; - z3 = (INT32) wsptr[3]; - - z1 = MULTIPLY(z2 + z3, FIX(0.831253876)); /* c3 */ - tmp0 = z1 + MULTIPLY(z2, FIX(0.513743148)); /* c1-c3 */ - tmp1 = z1 - MULTIPLY(z3, FIX(2.176250899)); /* c1+c3 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp11 + tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp11 - tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 5; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a reduced-size 4x4 output block. - * - * Optimized algorithm with 3 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/16) [refers to 8-point IDCT]. - */ - -GLOBAL(void) -jpeg_idct_4x4 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp2, tmp10, tmp12; - INT32 z1, z2, z3; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[4*4]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 4; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp0 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp2 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - - tmp10 = (tmp0 + tmp2) << PASS1_BITS; - tmp12 = (tmp0 - tmp2) << PASS1_BITS; - - /* Odd part */ - /* Same rotation as in the even part of the 8x8 LL&M IDCT */ - - z2 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z3 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - - z1 = MULTIPLY(z2 + z3, FIX_0_541196100); /* c6 */ - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-1); - tmp0 = RIGHT_SHIFT(z1 + MULTIPLY(z2, FIX_0_765366865), /* c2-c6 */ - CONST_BITS-PASS1_BITS); - tmp2 = RIGHT_SHIFT(z1 - MULTIPLY(z3, FIX_1_847759065), /* c2+c6 */ - CONST_BITS-PASS1_BITS); - - /* Final output stage */ - - wsptr[4*0] = (int) (tmp10 + tmp0); - wsptr[4*3] = (int) (tmp10 - tmp0); - wsptr[4*1] = (int) (tmp12 + tmp2); - wsptr[4*2] = (int) (tmp12 - tmp2); - } - - /* Pass 2: process 4 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 4; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp2 = (INT32) wsptr[2]; - - tmp10 = (tmp0 + tmp2) << CONST_BITS; - tmp12 = (tmp0 - tmp2) << CONST_BITS; - - /* Odd part */ - /* Same rotation as in the even part of the 8x8 LL&M IDCT */ - - z2 = (INT32) wsptr[1]; - z3 = (INT32) wsptr[3]; - - z1 = MULTIPLY(z2 + z3, FIX_0_541196100); /* c6 */ - tmp0 = z1 + MULTIPLY(z2, FIX_0_765366865); /* c2-c6 */ - tmp2 = z1 - MULTIPLY(z3, FIX_1_847759065); /* c2+c6 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp12 + tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp12 - tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 4; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a reduced-size 3x3 output block. - * - * Optimized algorithm with 2 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/6). - */ - -GLOBAL(void) -jpeg_idct_3x3 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp2, tmp10, tmp12; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[3*3]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 3; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp0 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp0 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-PASS1_BITS-1); - tmp2 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - tmp12 = MULTIPLY(tmp2, FIX(0.707106781)); /* c2 */ - tmp10 = tmp0 + tmp12; - tmp2 = tmp0 - tmp12 - tmp12; - - /* Odd part */ - - tmp12 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - tmp0 = MULTIPLY(tmp12, FIX(1.224744871)); /* c1 */ - - /* Final output stage */ - - wsptr[3*0] = (int) RIGHT_SHIFT(tmp10 + tmp0, CONST_BITS-PASS1_BITS); - wsptr[3*2] = (int) RIGHT_SHIFT(tmp10 - tmp0, CONST_BITS-PASS1_BITS); - wsptr[3*1] = (int) RIGHT_SHIFT(tmp2, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 3 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 3; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp0 <<= CONST_BITS; - tmp2 = (INT32) wsptr[2]; - tmp12 = MULTIPLY(tmp2, FIX(0.707106781)); /* c2 */ - tmp10 = tmp0 + tmp12; - tmp2 = tmp0 - tmp12 - tmp12; - - /* Odd part */ - - tmp12 = (INT32) wsptr[1]; - tmp0 = MULTIPLY(tmp12, FIX(1.224744871)); /* c1 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 3; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a reduced-size 2x2 output block. - * - * Multiplication-less algorithm. - */ - -GLOBAL(void) -jpeg_idct_2x2 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp4, tmp5; - ISLOW_MULT_TYPE * quantptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - SHIFT_TEMPS - - /* Pass 1: process columns from input. */ - - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - - /* Column 0 */ - tmp4 = DEQUANTIZE(coef_block[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp5 = DEQUANTIZE(coef_block[DCTSIZE*1], quantptr[DCTSIZE*1]); - /* Add fudge factor here for final descale. */ - tmp4 += ONE << 2; - - tmp0 = tmp4 + tmp5; - tmp2 = tmp4 - tmp5; - - /* Column 1 */ - tmp4 = DEQUANTIZE(coef_block[DCTSIZE*0+1], quantptr[DCTSIZE*0+1]); - tmp5 = DEQUANTIZE(coef_block[DCTSIZE*1+1], quantptr[DCTSIZE*1+1]); - - tmp1 = tmp4 + tmp5; - tmp3 = tmp4 - tmp5; - - /* Pass 2: process 2 rows, store into output array. */ - - /* Row 0 */ - outptr = output_buf[0] + output_col; - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp0 + tmp1, 3) & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp0 - tmp1, 3) & RANGE_MASK]; - - /* Row 1 */ - outptr = output_buf[1] + output_col; - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp2 + tmp3, 3) & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp2 - tmp3, 3) & RANGE_MASK]; -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a reduced-size 1x1 output block. - * - * We hardly need an inverse DCT routine for this: just take the - * average pixel value, which is one-eighth of the DC coefficient. - */ - -GLOBAL(void) -jpeg_idct_1x1 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - int dcval; - ISLOW_MULT_TYPE * quantptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - SHIFT_TEMPS - - /* 1x1 is trivial: just take the DC coefficient divided by 8. */ - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - dcval = DEQUANTIZE(coef_block[0], quantptr[0]); - dcval = (int) DESCALE((INT32) dcval, 3); - - output_buf[0][output_col] = range_limit[dcval & RANGE_MASK]; -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 9x9 output block. - * - * Optimized algorithm with 10 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/18). - */ - -GLOBAL(void) -jpeg_idct_9x9 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp0, tmp1, tmp2, tmp3, tmp10, tmp11, tmp12, tmp13, tmp14; - INT32 z1, z2, z3, z4; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*9]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp0 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp0 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - tmp0 += ONE << (CONST_BITS-PASS1_BITS-1); - - z1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z2 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - tmp3 = MULTIPLY(z3, FIX(0.707106781)); /* c6 */ - tmp1 = tmp0 + tmp3; - tmp2 = tmp0 - tmp3 - tmp3; - - tmp0 = MULTIPLY(z1 - z2, FIX(0.707106781)); /* c6 */ - tmp11 = tmp2 + tmp0; - tmp14 = tmp2 - tmp0 - tmp0; - - tmp0 = MULTIPLY(z1 + z2, FIX(1.328926049)); /* c2 */ - tmp2 = MULTIPLY(z1, FIX(1.083350441)); /* c4 */ - tmp3 = MULTIPLY(z2, FIX(0.245575608)); /* c8 */ - - tmp10 = tmp1 + tmp0 - tmp3; - tmp12 = tmp1 - tmp0 + tmp2; - tmp13 = tmp1 - tmp2 + tmp3; - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - - z2 = MULTIPLY(z2, - FIX(1.224744871)); /* -c3 */ - - tmp2 = MULTIPLY(z1 + z3, FIX(0.909038955)); /* c5 */ - tmp3 = MULTIPLY(z1 + z4, FIX(0.483689525)); /* c7 */ - tmp0 = tmp2 + tmp3 - z2; - tmp1 = MULTIPLY(z3 - z4, FIX(1.392728481)); /* c1 */ - tmp2 += z2 - tmp1; - tmp3 += z2 + tmp1; - tmp1 = MULTIPLY(z1 - z3 - z4, FIX(1.224744871)); /* c3 */ - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp10 + tmp0, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp10 - tmp0, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp11 + tmp1, CONST_BITS-PASS1_BITS); - wsptr[8*7] = (int) RIGHT_SHIFT(tmp11 - tmp1, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) RIGHT_SHIFT(tmp12 + tmp2, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp12 - tmp2, CONST_BITS-PASS1_BITS); - wsptr[8*3] = (int) RIGHT_SHIFT(tmp13 + tmp3, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp13 - tmp3, CONST_BITS-PASS1_BITS); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp14, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 9 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 9; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp0 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp0 <<= CONST_BITS; - - z1 = (INT32) wsptr[2]; - z2 = (INT32) wsptr[4]; - z3 = (INT32) wsptr[6]; - - tmp3 = MULTIPLY(z3, FIX(0.707106781)); /* c6 */ - tmp1 = tmp0 + tmp3; - tmp2 = tmp0 - tmp3 - tmp3; - - tmp0 = MULTIPLY(z1 - z2, FIX(0.707106781)); /* c6 */ - tmp11 = tmp2 + tmp0; - tmp14 = tmp2 - tmp0 - tmp0; - - tmp0 = MULTIPLY(z1 + z2, FIX(1.328926049)); /* c2 */ - tmp2 = MULTIPLY(z1, FIX(1.083350441)); /* c4 */ - tmp3 = MULTIPLY(z2, FIX(0.245575608)); /* c8 */ - - tmp10 = tmp1 + tmp0 - tmp3; - tmp12 = tmp1 - tmp0 + tmp2; - tmp13 = tmp1 - tmp2 + tmp3; - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - z4 = (INT32) wsptr[7]; - - z2 = MULTIPLY(z2, - FIX(1.224744871)); /* -c3 */ - - tmp2 = MULTIPLY(z1 + z3, FIX(0.909038955)); /* c5 */ - tmp3 = MULTIPLY(z1 + z4, FIX(0.483689525)); /* c7 */ - tmp0 = tmp2 + tmp3 - z2; - tmp1 = MULTIPLY(z3 - z4, FIX(1.392728481)); /* c1 */ - tmp2 += z2 - tmp1; - tmp3 += z2 + tmp1; - tmp1 = MULTIPLY(z1 - z3 - z4, FIX(1.224744871)); /* c3 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp10 + tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[8] = range_limit[(int) RIGHT_SHIFT(tmp10 - tmp0, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp11 + tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp11 - tmp1, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp12 + tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp12 - tmp2, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp13 + tmp3, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp13 - tmp3, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 8; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 10x10 output block. - * - * Optimized algorithm with 12 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/20). - */ - -GLOBAL(void) -jpeg_idct_10x10 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp10, tmp11, tmp12, tmp13, tmp14; - INT32 tmp20, tmp21, tmp22, tmp23, tmp24; - INT32 z1, z2, z3, z4, z5; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*10]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - z3 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - z3 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - z3 += ONE << (CONST_BITS-PASS1_BITS-1); - z4 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z1 = MULTIPLY(z4, FIX(1.144122806)); /* c4 */ - z2 = MULTIPLY(z4, FIX(0.437016024)); /* c8 */ - tmp10 = z3 + z1; - tmp11 = z3 - z2; - - tmp22 = RIGHT_SHIFT(z3 - ((z1 - z2) << 1), /* c0 = (c4-c8)*2 */ - CONST_BITS-PASS1_BITS); - - z2 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - z1 = MULTIPLY(z2 + z3, FIX(0.831253876)); /* c6 */ - tmp12 = z1 + MULTIPLY(z2, FIX(0.513743148)); /* c2-c6 */ - tmp13 = z1 - MULTIPLY(z3, FIX(2.176250899)); /* c2+c6 */ - - tmp20 = tmp10 + tmp12; - tmp24 = tmp10 - tmp12; - tmp21 = tmp11 + tmp13; - tmp23 = tmp11 - tmp13; - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - - tmp11 = z2 + z4; - tmp13 = z2 - z4; - - tmp12 = MULTIPLY(tmp13, FIX(0.309016994)); /* (c3-c7)/2 */ - z5 = z3 << CONST_BITS; - - z2 = MULTIPLY(tmp11, FIX(0.951056516)); /* (c3+c7)/2 */ - z4 = z5 + tmp12; - - tmp10 = MULTIPLY(z1, FIX(1.396802247)) + z2 + z4; /* c1 */ - tmp14 = MULTIPLY(z1, FIX(0.221231742)) - z2 + z4; /* c9 */ - - z2 = MULTIPLY(tmp11, FIX(0.587785252)); /* (c1-c9)/2 */ - z4 = z5 - tmp12 - (tmp13 << (CONST_BITS - 1)); - - tmp12 = (z1 - tmp13 - z3) << PASS1_BITS; - - tmp11 = MULTIPLY(z1, FIX(1.260073511)) - z2 - z4; /* c3 */ - tmp13 = MULTIPLY(z1, FIX(0.642039522)) - z2 + z4; /* c7 */ - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp20 + tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*9] = (int) RIGHT_SHIFT(tmp20 - tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp21 + tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp21 - tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) (tmp22 + tmp12); - wsptr[8*7] = (int) (tmp22 - tmp12); - wsptr[8*3] = (int) RIGHT_SHIFT(tmp23 + tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp23 - tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp24 + tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp24 - tmp14, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 10 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 10; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - z3 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - z3 <<= CONST_BITS; - z4 = (INT32) wsptr[4]; - z1 = MULTIPLY(z4, FIX(1.144122806)); /* c4 */ - z2 = MULTIPLY(z4, FIX(0.437016024)); /* c8 */ - tmp10 = z3 + z1; - tmp11 = z3 - z2; - - tmp22 = z3 - ((z1 - z2) << 1); /* c0 = (c4-c8)*2 */ - - z2 = (INT32) wsptr[2]; - z3 = (INT32) wsptr[6]; - - z1 = MULTIPLY(z2 + z3, FIX(0.831253876)); /* c6 */ - tmp12 = z1 + MULTIPLY(z2, FIX(0.513743148)); /* c2-c6 */ - tmp13 = z1 - MULTIPLY(z3, FIX(2.176250899)); /* c2+c6 */ - - tmp20 = tmp10 + tmp12; - tmp24 = tmp10 - tmp12; - tmp21 = tmp11 + tmp13; - tmp23 = tmp11 - tmp13; - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - z3 <<= CONST_BITS; - z4 = (INT32) wsptr[7]; - - tmp11 = z2 + z4; - tmp13 = z2 - z4; - - tmp12 = MULTIPLY(tmp13, FIX(0.309016994)); /* (c3-c7)/2 */ - - z2 = MULTIPLY(tmp11, FIX(0.951056516)); /* (c3+c7)/2 */ - z4 = z3 + tmp12; - - tmp10 = MULTIPLY(z1, FIX(1.396802247)) + z2 + z4; /* c1 */ - tmp14 = MULTIPLY(z1, FIX(0.221231742)) - z2 + z4; /* c9 */ - - z2 = MULTIPLY(tmp11, FIX(0.587785252)); /* (c1-c9)/2 */ - z4 = z3 - tmp12 - (tmp13 << (CONST_BITS - 1)); - - tmp12 = ((z1 - tmp13) << CONST_BITS) - z3; - - tmp11 = MULTIPLY(z1, FIX(1.260073511)) - z2 - z4; /* c3 */ - tmp13 = MULTIPLY(z1, FIX(0.642039522)) - z2 + z4; /* c7 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp20 + tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[9] = range_limit[(int) RIGHT_SHIFT(tmp20 - tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp21 + tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[8] = range_limit[(int) RIGHT_SHIFT(tmp21 - tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp22 + tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp22 - tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp23 + tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp23 - tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp24 + tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp24 - tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 8; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 11x11 output block. - * - * Optimized algorithm with 24 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/22). - */ - -GLOBAL(void) -jpeg_idct_11x11 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp10, tmp11, tmp12, tmp13, tmp14; - INT32 tmp20, tmp21, tmp22, tmp23, tmp24, tmp25; - INT32 z1, z2, z3, z4; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*11]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - tmp10 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - tmp10 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - tmp10 += ONE << (CONST_BITS-PASS1_BITS-1); - - z1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z2 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z3 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - tmp20 = MULTIPLY(z2 - z3, FIX(2.546640132)); /* c2+c4 */ - tmp23 = MULTIPLY(z2 - z1, FIX(0.430815045)); /* c2-c6 */ - z4 = z1 + z3; - tmp24 = MULTIPLY(z4, - FIX(1.155664402)); /* -(c2-c10) */ - z4 -= z2; - tmp25 = tmp10 + MULTIPLY(z4, FIX(1.356927976)); /* c2 */ - tmp21 = tmp20 + tmp23 + tmp25 - - MULTIPLY(z2, FIX(1.821790775)); /* c2+c4+c10-c6 */ - tmp20 += tmp25 + MULTIPLY(z3, FIX(2.115825087)); /* c4+c6 */ - tmp23 += tmp25 - MULTIPLY(z1, FIX(1.513598477)); /* c6+c8 */ - tmp24 += tmp25; - tmp22 = tmp24 - MULTIPLY(z3, FIX(0.788749120)); /* c8+c10 */ - tmp24 += MULTIPLY(z2, FIX(1.944413522)) - /* c2+c8 */ - MULTIPLY(z1, FIX(1.390975730)); /* c4+c10 */ - tmp25 = tmp10 - MULTIPLY(z4, FIX(1.414213562)); /* c0 */ - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - - tmp11 = z1 + z2; - tmp14 = MULTIPLY(tmp11 + z3 + z4, FIX(0.398430003)); /* c9 */ - tmp11 = MULTIPLY(tmp11, FIX(0.887983902)); /* c3-c9 */ - tmp12 = MULTIPLY(z1 + z3, FIX(0.670361295)); /* c5-c9 */ - tmp13 = tmp14 + MULTIPLY(z1 + z4, FIX(0.366151574)); /* c7-c9 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(z1, FIX(0.923107866)); /* c7+c5+c3-c1-2*c9 */ - z1 = tmp14 - MULTIPLY(z2 + z3, FIX(1.163011579)); /* c7+c9 */ - tmp11 += z1 + MULTIPLY(z2, FIX(2.073276588)); /* c1+c7+3*c9-c3 */ - tmp12 += z1 - MULTIPLY(z3, FIX(1.192193623)); /* c3+c5-c7-c9 */ - z1 = MULTIPLY(z2 + z4, - FIX(1.798248910)); /* -(c1+c9) */ - tmp11 += z1; - tmp13 += z1 + MULTIPLY(z4, FIX(2.102458632)); /* c1+c5+c9-c7 */ - tmp14 += MULTIPLY(z2, - FIX(1.467221301)) + /* -(c5+c9) */ - MULTIPLY(z3, FIX(1.001388905)) - /* c1-c9 */ - MULTIPLY(z4, FIX(1.684843907)); /* c3+c9 */ - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp20 + tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*10] = (int) RIGHT_SHIFT(tmp20 - tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp21 + tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*9] = (int) RIGHT_SHIFT(tmp21 - tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) RIGHT_SHIFT(tmp22 + tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp22 - tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*3] = (int) RIGHT_SHIFT(tmp23 + tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*7] = (int) RIGHT_SHIFT(tmp23 - tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp24 + tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp24 - tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp25, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 11 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 11; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - tmp10 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - tmp10 <<= CONST_BITS; - - z1 = (INT32) wsptr[2]; - z2 = (INT32) wsptr[4]; - z3 = (INT32) wsptr[6]; - - tmp20 = MULTIPLY(z2 - z3, FIX(2.546640132)); /* c2+c4 */ - tmp23 = MULTIPLY(z2 - z1, FIX(0.430815045)); /* c2-c6 */ - z4 = z1 + z3; - tmp24 = MULTIPLY(z4, - FIX(1.155664402)); /* -(c2-c10) */ - z4 -= z2; - tmp25 = tmp10 + MULTIPLY(z4, FIX(1.356927976)); /* c2 */ - tmp21 = tmp20 + tmp23 + tmp25 - - MULTIPLY(z2, FIX(1.821790775)); /* c2+c4+c10-c6 */ - tmp20 += tmp25 + MULTIPLY(z3, FIX(2.115825087)); /* c4+c6 */ - tmp23 += tmp25 - MULTIPLY(z1, FIX(1.513598477)); /* c6+c8 */ - tmp24 += tmp25; - tmp22 = tmp24 - MULTIPLY(z3, FIX(0.788749120)); /* c8+c10 */ - tmp24 += MULTIPLY(z2, FIX(1.944413522)) - /* c2+c8 */ - MULTIPLY(z1, FIX(1.390975730)); /* c4+c10 */ - tmp25 = tmp10 - MULTIPLY(z4, FIX(1.414213562)); /* c0 */ - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - z4 = (INT32) wsptr[7]; - - tmp11 = z1 + z2; - tmp14 = MULTIPLY(tmp11 + z3 + z4, FIX(0.398430003)); /* c9 */ - tmp11 = MULTIPLY(tmp11, FIX(0.887983902)); /* c3-c9 */ - tmp12 = MULTIPLY(z1 + z3, FIX(0.670361295)); /* c5-c9 */ - tmp13 = tmp14 + MULTIPLY(z1 + z4, FIX(0.366151574)); /* c7-c9 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(z1, FIX(0.923107866)); /* c7+c5+c3-c1-2*c9 */ - z1 = tmp14 - MULTIPLY(z2 + z3, FIX(1.163011579)); /* c7+c9 */ - tmp11 += z1 + MULTIPLY(z2, FIX(2.073276588)); /* c1+c7+3*c9-c3 */ - tmp12 += z1 - MULTIPLY(z3, FIX(1.192193623)); /* c3+c5-c7-c9 */ - z1 = MULTIPLY(z2 + z4, - FIX(1.798248910)); /* -(c1+c9) */ - tmp11 += z1; - tmp13 += z1 + MULTIPLY(z4, FIX(2.102458632)); /* c1+c5+c9-c7 */ - tmp14 += MULTIPLY(z2, - FIX(1.467221301)) + /* -(c5+c9) */ - MULTIPLY(z3, FIX(1.001388905)) - /* c1-c9 */ - MULTIPLY(z4, FIX(1.684843907)); /* c3+c9 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp20 + tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[10] = range_limit[(int) RIGHT_SHIFT(tmp20 - tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp21 + tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[9] = range_limit[(int) RIGHT_SHIFT(tmp21 - tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp22 + tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[8] = range_limit[(int) RIGHT_SHIFT(tmp22 - tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp23 + tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp23 - tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp24 + tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp24 - tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp25, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 8; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 12x12 output block. - * - * Optimized algorithm with 15 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/24). - */ - -GLOBAL(void) -jpeg_idct_12x12 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15; - INT32 tmp20, tmp21, tmp22, tmp23, tmp24, tmp25; - INT32 z1, z2, z3, z4; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*12]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - z3 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - z3 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - z3 += ONE << (CONST_BITS-PASS1_BITS-1); - - z4 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z4 = MULTIPLY(z4, FIX(1.224744871)); /* c4 */ - - tmp10 = z3 + z4; - tmp11 = z3 - z4; - - z1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z4 = MULTIPLY(z1, FIX(1.366025404)); /* c2 */ - z1 <<= CONST_BITS; - z2 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - z2 <<= CONST_BITS; - - tmp12 = z1 - z2; - - tmp21 = z3 + tmp12; - tmp24 = z3 - tmp12; - - tmp12 = z4 + z2; - - tmp20 = tmp10 + tmp12; - tmp25 = tmp10 - tmp12; - - tmp12 = z4 - z1 - z2; - - tmp22 = tmp11 + tmp12; - tmp23 = tmp11 - tmp12; - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - - tmp11 = MULTIPLY(z2, FIX(1.306562965)); /* c3 */ - tmp14 = MULTIPLY(z2, - FIX_0_541196100); /* -c9 */ - - tmp10 = z1 + z3; - tmp15 = MULTIPLY(tmp10 + z4, FIX(0.860918669)); /* c7 */ - tmp12 = tmp15 + MULTIPLY(tmp10, FIX(0.261052384)); /* c5-c7 */ - tmp10 = tmp12 + tmp11 + MULTIPLY(z1, FIX(0.280143716)); /* c1-c5 */ - tmp13 = MULTIPLY(z3 + z4, - FIX(1.045510580)); /* -(c7+c11) */ - tmp12 += tmp13 + tmp14 - MULTIPLY(z3, FIX(1.478575242)); /* c1+c5-c7-c11 */ - tmp13 += tmp15 - tmp11 + MULTIPLY(z4, FIX(1.586706681)); /* c1+c11 */ - tmp15 += tmp14 - MULTIPLY(z1, FIX(0.676326758)) - /* c7-c11 */ - MULTIPLY(z4, FIX(1.982889723)); /* c5+c7 */ - - z1 -= z4; - z2 -= z3; - z3 = MULTIPLY(z1 + z2, FIX_0_541196100); /* c9 */ - tmp11 = z3 + MULTIPLY(z1, FIX_0_765366865); /* c3-c9 */ - tmp14 = z3 - MULTIPLY(z2, FIX_1_847759065); /* c3+c9 */ - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp20 + tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*11] = (int) RIGHT_SHIFT(tmp20 - tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp21 + tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*10] = (int) RIGHT_SHIFT(tmp21 - tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) RIGHT_SHIFT(tmp22 + tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*9] = (int) RIGHT_SHIFT(tmp22 - tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*3] = (int) RIGHT_SHIFT(tmp23 + tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp23 - tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp24 + tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*7] = (int) RIGHT_SHIFT(tmp24 - tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp25 + tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp25 - tmp15, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 12 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 12; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - z3 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - z3 <<= CONST_BITS; - - z4 = (INT32) wsptr[4]; - z4 = MULTIPLY(z4, FIX(1.224744871)); /* c4 */ - - tmp10 = z3 + z4; - tmp11 = z3 - z4; - - z1 = (INT32) wsptr[2]; - z4 = MULTIPLY(z1, FIX(1.366025404)); /* c2 */ - z1 <<= CONST_BITS; - z2 = (INT32) wsptr[6]; - z2 <<= CONST_BITS; - - tmp12 = z1 - z2; - - tmp21 = z3 + tmp12; - tmp24 = z3 - tmp12; - - tmp12 = z4 + z2; - - tmp20 = tmp10 + tmp12; - tmp25 = tmp10 - tmp12; - - tmp12 = z4 - z1 - z2; - - tmp22 = tmp11 + tmp12; - tmp23 = tmp11 - tmp12; - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - z4 = (INT32) wsptr[7]; - - tmp11 = MULTIPLY(z2, FIX(1.306562965)); /* c3 */ - tmp14 = MULTIPLY(z2, - FIX_0_541196100); /* -c9 */ - - tmp10 = z1 + z3; - tmp15 = MULTIPLY(tmp10 + z4, FIX(0.860918669)); /* c7 */ - tmp12 = tmp15 + MULTIPLY(tmp10, FIX(0.261052384)); /* c5-c7 */ - tmp10 = tmp12 + tmp11 + MULTIPLY(z1, FIX(0.280143716)); /* c1-c5 */ - tmp13 = MULTIPLY(z3 + z4, - FIX(1.045510580)); /* -(c7+c11) */ - tmp12 += tmp13 + tmp14 - MULTIPLY(z3, FIX(1.478575242)); /* c1+c5-c7-c11 */ - tmp13 += tmp15 - tmp11 + MULTIPLY(z4, FIX(1.586706681)); /* c1+c11 */ - tmp15 += tmp14 - MULTIPLY(z1, FIX(0.676326758)) - /* c7-c11 */ - MULTIPLY(z4, FIX(1.982889723)); /* c5+c7 */ - - z1 -= z4; - z2 -= z3; - z3 = MULTIPLY(z1 + z2, FIX_0_541196100); /* c9 */ - tmp11 = z3 + MULTIPLY(z1, FIX_0_765366865); /* c3-c9 */ - tmp14 = z3 - MULTIPLY(z2, FIX_1_847759065); /* c3+c9 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp20 + tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[11] = range_limit[(int) RIGHT_SHIFT(tmp20 - tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp21 + tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[10] = range_limit[(int) RIGHT_SHIFT(tmp21 - tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp22 + tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[9] = range_limit[(int) RIGHT_SHIFT(tmp22 - tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp23 + tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[8] = range_limit[(int) RIGHT_SHIFT(tmp23 - tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp24 + tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp24 - tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp25 + tmp15, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp25 - tmp15, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 8; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 13x13 output block. - * - * Optimized algorithm with 29 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/26). - */ - -GLOBAL(void) -jpeg_idct_13x13 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15; - INT32 tmp20, tmp21, tmp22, tmp23, tmp24, tmp25, tmp26; - INT32 z1, z2, z3, z4; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*13]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - z1 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-1); - - z2 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z3 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z4 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - tmp10 = z3 + z4; - tmp11 = z3 - z4; - - tmp12 = MULTIPLY(tmp10, FIX(1.155388986)); /* (c4+c6)/2 */ - tmp13 = MULTIPLY(tmp11, FIX(0.096834934)) + z1; /* (c4-c6)/2 */ - - tmp20 = MULTIPLY(z2, FIX(1.373119086)) + tmp12 + tmp13; /* c2 */ - tmp22 = MULTIPLY(z2, FIX(0.501487041)) - tmp12 + tmp13; /* c10 */ - - tmp12 = MULTIPLY(tmp10, FIX(0.316450131)); /* (c8-c12)/2 */ - tmp13 = MULTIPLY(tmp11, FIX(0.486914739)) + z1; /* (c8+c12)/2 */ - - tmp21 = MULTIPLY(z2, FIX(1.058554052)) - tmp12 + tmp13; /* c6 */ - tmp25 = MULTIPLY(z2, - FIX(1.252223920)) + tmp12 + tmp13; /* c4 */ - - tmp12 = MULTIPLY(tmp10, FIX(0.435816023)); /* (c2-c10)/2 */ - tmp13 = MULTIPLY(tmp11, FIX(0.937303064)) - z1; /* (c2+c10)/2 */ - - tmp23 = MULTIPLY(z2, - FIX(0.170464608)) - tmp12 - tmp13; /* c12 */ - tmp24 = MULTIPLY(z2, - FIX(0.803364869)) + tmp12 - tmp13; /* c8 */ - - tmp26 = MULTIPLY(tmp11 - z2, FIX(1.414213562)) + z1; /* c0 */ - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - - tmp11 = MULTIPLY(z1 + z2, FIX(1.322312651)); /* c3 */ - tmp12 = MULTIPLY(z1 + z3, FIX(1.163874945)); /* c5 */ - tmp15 = z1 + z4; - tmp13 = MULTIPLY(tmp15, FIX(0.937797057)); /* c7 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(z1, FIX(2.020082300)); /* c7+c5+c3-c1 */ - tmp14 = MULTIPLY(z2 + z3, - FIX(0.338443458)); /* -c11 */ - tmp11 += tmp14 + MULTIPLY(z2, FIX(0.837223564)); /* c5+c9+c11-c3 */ - tmp12 += tmp14 - MULTIPLY(z3, FIX(1.572116027)); /* c1+c5-c9-c11 */ - tmp14 = MULTIPLY(z2 + z4, - FIX(1.163874945)); /* -c5 */ - tmp11 += tmp14; - tmp13 += tmp14 + MULTIPLY(z4, FIX(2.205608352)); /* c3+c5+c9-c7 */ - tmp14 = MULTIPLY(z3 + z4, - FIX(0.657217813)); /* -c9 */ - tmp12 += tmp14; - tmp13 += tmp14; - tmp15 = MULTIPLY(tmp15, FIX(0.338443458)); /* c11 */ - tmp14 = tmp15 + MULTIPLY(z1, FIX(0.318774355)) - /* c9-c11 */ - MULTIPLY(z2, FIX(0.466105296)); /* c1-c7 */ - z1 = MULTIPLY(z3 - z2, FIX(0.937797057)); /* c7 */ - tmp14 += z1; - tmp15 += z1 + MULTIPLY(z3, FIX(0.384515595)) - /* c3-c7 */ - MULTIPLY(z4, FIX(1.742345811)); /* c1+c11 */ - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp20 + tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*12] = (int) RIGHT_SHIFT(tmp20 - tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp21 + tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*11] = (int) RIGHT_SHIFT(tmp21 - tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) RIGHT_SHIFT(tmp22 + tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*10] = (int) RIGHT_SHIFT(tmp22 - tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*3] = (int) RIGHT_SHIFT(tmp23 + tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*9] = (int) RIGHT_SHIFT(tmp23 - tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp24 + tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp24 - tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp25 + tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*7] = (int) RIGHT_SHIFT(tmp25 - tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp26, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 13 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 13; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - z1 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - z1 <<= CONST_BITS; - - z2 = (INT32) wsptr[2]; - z3 = (INT32) wsptr[4]; - z4 = (INT32) wsptr[6]; - - tmp10 = z3 + z4; - tmp11 = z3 - z4; - - tmp12 = MULTIPLY(tmp10, FIX(1.155388986)); /* (c4+c6)/2 */ - tmp13 = MULTIPLY(tmp11, FIX(0.096834934)) + z1; /* (c4-c6)/2 */ - - tmp20 = MULTIPLY(z2, FIX(1.373119086)) + tmp12 + tmp13; /* c2 */ - tmp22 = MULTIPLY(z2, FIX(0.501487041)) - tmp12 + tmp13; /* c10 */ - - tmp12 = MULTIPLY(tmp10, FIX(0.316450131)); /* (c8-c12)/2 */ - tmp13 = MULTIPLY(tmp11, FIX(0.486914739)) + z1; /* (c8+c12)/2 */ - - tmp21 = MULTIPLY(z2, FIX(1.058554052)) - tmp12 + tmp13; /* c6 */ - tmp25 = MULTIPLY(z2, - FIX(1.252223920)) + tmp12 + tmp13; /* c4 */ - - tmp12 = MULTIPLY(tmp10, FIX(0.435816023)); /* (c2-c10)/2 */ - tmp13 = MULTIPLY(tmp11, FIX(0.937303064)) - z1; /* (c2+c10)/2 */ - - tmp23 = MULTIPLY(z2, - FIX(0.170464608)) - tmp12 - tmp13; /* c12 */ - tmp24 = MULTIPLY(z2, - FIX(0.803364869)) + tmp12 - tmp13; /* c8 */ - - tmp26 = MULTIPLY(tmp11 - z2, FIX(1.414213562)) + z1; /* c0 */ - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - z4 = (INT32) wsptr[7]; - - tmp11 = MULTIPLY(z1 + z2, FIX(1.322312651)); /* c3 */ - tmp12 = MULTIPLY(z1 + z3, FIX(1.163874945)); /* c5 */ - tmp15 = z1 + z4; - tmp13 = MULTIPLY(tmp15, FIX(0.937797057)); /* c7 */ - tmp10 = tmp11 + tmp12 + tmp13 - - MULTIPLY(z1, FIX(2.020082300)); /* c7+c5+c3-c1 */ - tmp14 = MULTIPLY(z2 + z3, - FIX(0.338443458)); /* -c11 */ - tmp11 += tmp14 + MULTIPLY(z2, FIX(0.837223564)); /* c5+c9+c11-c3 */ - tmp12 += tmp14 - MULTIPLY(z3, FIX(1.572116027)); /* c1+c5-c9-c11 */ - tmp14 = MULTIPLY(z2 + z4, - FIX(1.163874945)); /* -c5 */ - tmp11 += tmp14; - tmp13 += tmp14 + MULTIPLY(z4, FIX(2.205608352)); /* c3+c5+c9-c7 */ - tmp14 = MULTIPLY(z3 + z4, - FIX(0.657217813)); /* -c9 */ - tmp12 += tmp14; - tmp13 += tmp14; - tmp15 = MULTIPLY(tmp15, FIX(0.338443458)); /* c11 */ - tmp14 = tmp15 + MULTIPLY(z1, FIX(0.318774355)) - /* c9-c11 */ - MULTIPLY(z2, FIX(0.466105296)); /* c1-c7 */ - z1 = MULTIPLY(z3 - z2, FIX(0.937797057)); /* c7 */ - tmp14 += z1; - tmp15 += z1 + MULTIPLY(z3, FIX(0.384515595)) - /* c3-c7 */ - MULTIPLY(z4, FIX(1.742345811)); /* c1+c11 */ - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp20 + tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[12] = range_limit[(int) RIGHT_SHIFT(tmp20 - tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp21 + tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[11] = range_limit[(int) RIGHT_SHIFT(tmp21 - tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp22 + tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[10] = range_limit[(int) RIGHT_SHIFT(tmp22 - tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp23 + tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[9] = range_limit[(int) RIGHT_SHIFT(tmp23 - tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp24 + tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[8] = range_limit[(int) RIGHT_SHIFT(tmp24 - tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp25 + tmp15, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp25 - tmp15, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp26, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 8; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 14x14 output block. - * - * Optimized algorithm with 20 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/28). - */ - -GLOBAL(void) -jpeg_idct_14x14 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16; - INT32 tmp20, tmp21, tmp22, tmp23, tmp24, tmp25, tmp26; - INT32 z1, z2, z3, z4; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*14]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - z1 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-1); - z4 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z2 = MULTIPLY(z4, FIX(1.274162392)); /* c4 */ - z3 = MULTIPLY(z4, FIX(0.314692123)); /* c12 */ - z4 = MULTIPLY(z4, FIX(0.881747734)); /* c8 */ - - tmp10 = z1 + z2; - tmp11 = z1 + z3; - tmp12 = z1 - z4; - - tmp23 = RIGHT_SHIFT(z1 - ((z2 + z3 - z4) << 1), /* c0 = (c4+c12-c8)*2 */ - CONST_BITS-PASS1_BITS); - - z1 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z2 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - z3 = MULTIPLY(z1 + z2, FIX(1.105676686)); /* c6 */ - - tmp13 = z3 + MULTIPLY(z1, FIX(0.273079590)); /* c2-c6 */ - tmp14 = z3 - MULTIPLY(z2, FIX(1.719280954)); /* c6+c10 */ - tmp15 = MULTIPLY(z1, FIX(0.613604268)) - /* c10 */ - MULTIPLY(z2, FIX(1.378756276)); /* c2 */ - - tmp20 = tmp10 + tmp13; - tmp26 = tmp10 - tmp13; - tmp21 = tmp11 + tmp14; - tmp25 = tmp11 - tmp14; - tmp22 = tmp12 + tmp15; - tmp24 = tmp12 - tmp15; - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z3 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - tmp13 = z4 << CONST_BITS; - - tmp14 = z1 + z3; - tmp11 = MULTIPLY(z1 + z2, FIX(1.334852607)); /* c3 */ - tmp12 = MULTIPLY(tmp14, FIX(1.197448846)); /* c5 */ - tmp10 = tmp11 + tmp12 + tmp13 - MULTIPLY(z1, FIX(1.126980169)); /* c3+c5-c1 */ - tmp14 = MULTIPLY(tmp14, FIX(0.752406978)); /* c9 */ - tmp16 = tmp14 - MULTIPLY(z1, FIX(1.061150426)); /* c9+c11-c13 */ - z1 -= z2; - tmp15 = MULTIPLY(z1, FIX(0.467085129)) - tmp13; /* c11 */ - tmp16 += tmp15; - z1 += z4; - z4 = MULTIPLY(z2 + z3, - FIX(0.158341681)) - tmp13; /* -c13 */ - tmp11 += z4 - MULTIPLY(z2, FIX(0.424103948)); /* c3-c9-c13 */ - tmp12 += z4 - MULTIPLY(z3, FIX(2.373959773)); /* c3+c5-c13 */ - z4 = MULTIPLY(z3 - z2, FIX(1.405321284)); /* c1 */ - tmp14 += z4 + tmp13 - MULTIPLY(z3, FIX(1.6906431334)); /* c1+c9-c11 */ - tmp15 += z4 + MULTIPLY(z2, FIX(0.674957567)); /* c1+c11-c5 */ - - tmp13 = (z1 - z3) << PASS1_BITS; - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp20 + tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*13] = (int) RIGHT_SHIFT(tmp20 - tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp21 + tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*12] = (int) RIGHT_SHIFT(tmp21 - tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) RIGHT_SHIFT(tmp22 + tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*11] = (int) RIGHT_SHIFT(tmp22 - tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*3] = (int) (tmp23 + tmp13); - wsptr[8*10] = (int) (tmp23 - tmp13); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp24 + tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*9] = (int) RIGHT_SHIFT(tmp24 - tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp25 + tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp25 - tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp26 + tmp16, CONST_BITS-PASS1_BITS); - wsptr[8*7] = (int) RIGHT_SHIFT(tmp26 - tmp16, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 14 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 14; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - z1 = (INT32) wsptr[0] + (ONE << (PASS1_BITS+2)); - z1 <<= CONST_BITS; - z4 = (INT32) wsptr[4]; - z2 = MULTIPLY(z4, FIX(1.274162392)); /* c4 */ - z3 = MULTIPLY(z4, FIX(0.314692123)); /* c12 */ - z4 = MULTIPLY(z4, FIX(0.881747734)); /* c8 */ - - tmp10 = z1 + z2; - tmp11 = z1 + z3; - tmp12 = z1 - z4; - - tmp23 = z1 - ((z2 + z3 - z4) << 1); /* c0 = (c4+c12-c8)*2 */ - - z1 = (INT32) wsptr[2]; - z2 = (INT32) wsptr[6]; - - z3 = MULTIPLY(z1 + z2, FIX(1.105676686)); /* c6 */ - - tmp13 = z3 + MULTIPLY(z1, FIX(0.273079590)); /* c2-c6 */ - tmp14 = z3 - MULTIPLY(z2, FIX(1.719280954)); /* c6+c10 */ - tmp15 = MULTIPLY(z1, FIX(0.613604268)) - /* c10 */ - MULTIPLY(z2, FIX(1.378756276)); /* c2 */ - - tmp20 = tmp10 + tmp13; - tmp26 = tmp10 - tmp13; - tmp21 = tmp11 + tmp14; - tmp25 = tmp11 - tmp14; - tmp22 = tmp12 + tmp15; - tmp24 = tmp12 - tmp15; - - /* Odd part */ - - z1 = (INT32) wsptr[1]; - z2 = (INT32) wsptr[3]; - z3 = (INT32) wsptr[5]; - z4 = (INT32) wsptr[7]; - z4 <<= CONST_BITS; - - tmp14 = z1 + z3; - tmp11 = MULTIPLY(z1 + z2, FIX(1.334852607)); /* c3 */ - tmp12 = MULTIPLY(tmp14, FIX(1.197448846)); /* c5 */ - tmp10 = tmp11 + tmp12 + z4 - MULTIPLY(z1, FIX(1.126980169)); /* c3+c5-c1 */ - tmp14 = MULTIPLY(tmp14, FIX(0.752406978)); /* c9 */ - tmp16 = tmp14 - MULTIPLY(z1, FIX(1.061150426)); /* c9+c11-c13 */ - z1 -= z2; - tmp15 = MULTIPLY(z1, FIX(0.467085129)) - z4; /* c11 */ - tmp16 += tmp15; - tmp13 = MULTIPLY(z2 + z3, - FIX(0.158341681)) - z4; /* -c13 */ - tmp11 += tmp13 - MULTIPLY(z2, FIX(0.424103948)); /* c3-c9-c13 */ - tmp12 += tmp13 - MULTIPLY(z3, FIX(2.373959773)); /* c3+c5-c13 */ - tmp13 = MULTIPLY(z3 - z2, FIX(1.405321284)); /* c1 */ - tmp14 += tmp13 + z4 - MULTIPLY(z3, FIX(1.6906431334)); /* c1+c9-c11 */ - tmp15 += tmp13 + MULTIPLY(z2, FIX(0.674957567)); /* c1+c11-c5 */ - - tmp13 = ((z1 - z3) << CONST_BITS) + z4; - - /* Final output stage */ - - outptr[0] = range_limit[(int) RIGHT_SHIFT(tmp20 + tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[13] = range_limit[(int) RIGHT_SHIFT(tmp20 - tmp10, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[1] = range_limit[(int) RIGHT_SHIFT(tmp21 + tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[12] = range_limit[(int) RIGHT_SHIFT(tmp21 - tmp11, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[2] = range_limit[(int) RIGHT_SHIFT(tmp22 + tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[11] = range_limit[(int) RIGHT_SHIFT(tmp22 - tmp12, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[3] = range_limit[(int) RIGHT_SHIFT(tmp23 + tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[10] = range_limit[(int) RIGHT_SHIFT(tmp23 - tmp13, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[4] = range_limit[(int) RIGHT_SHIFT(tmp24 + tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[9] = range_limit[(int) RIGHT_SHIFT(tmp24 - tmp14, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[5] = range_limit[(int) RIGHT_SHIFT(tmp25 + tmp15, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[8] = range_limit[(int) RIGHT_SHIFT(tmp25 - tmp15, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[6] = range_limit[(int) RIGHT_SHIFT(tmp26 + tmp16, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - outptr[7] = range_limit[(int) RIGHT_SHIFT(tmp26 - tmp16, - CONST_BITS+PASS1_BITS+3) - & RANGE_MASK]; - - wsptr += 8; /* advance pointer to next row */ - } -} - - -/* - * Perform dequantization and inverse DCT on one block of coefficients, - * producing a 15x15 output block. - * - * Optimized algorithm with 22 multiplications in the 1-D kernel. - * cK represents sqrt(2) * cos(K*pi/30). - */ - -GLOBAL(void) -jpeg_idct_15x15 (j_decompress_ptr cinfo, jpeg_component_info * compptr, - JCOEFPTR coef_block, - JSAMPARRAY output_buf, JDIMENSION output_col) -{ - INT32 tmp10, tmp11, tmp12, tmp13, tmp14, tmp15, tmp16; - INT32 tmp20, tmp21, tmp22, tmp23, tmp24, tmp25, tmp26, tmp27; - INT32 z1, z2, z3, z4; - JCOEFPTR inptr; - ISLOW_MULT_TYPE * quantptr; - int * wsptr; - JSAMPROW outptr; - JSAMPLE *range_limit = IDCT_range_limit(cinfo); - int ctr; - int workspace[8*15]; /* buffers data between passes */ - SHIFT_TEMPS - - /* Pass 1: process columns from input, store into work array. */ - - inptr = coef_block; - quantptr = (ISLOW_MULT_TYPE *) compptr->dct_table; - wsptr = workspace; - for (ctr = 0; ctr < 8; ctr++, inptr++, quantptr++, wsptr++) { - /* Even part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*0], quantptr[DCTSIZE*0]); - z1 <<= CONST_BITS; - /* Add fudge factor here for final descale. */ - z1 += ONE << (CONST_BITS-PASS1_BITS-1); - - z2 = DEQUANTIZE(inptr[DCTSIZE*2], quantptr[DCTSIZE*2]); - z3 = DEQUANTIZE(inptr[DCTSIZE*4], quantptr[DCTSIZE*4]); - z4 = DEQUANTIZE(inptr[DCTSIZE*6], quantptr[DCTSIZE*6]); - - tmp10 = MULTIPLY(z4, FIX(0.437016024)); /* c12 */ - tmp11 = MULTIPLY(z4, FIX(1.144122806)); /* c6 */ - - tmp12 = z1 - tmp10; - tmp13 = z1 + tmp11; - z1 -= (tmp11 - tmp10) << 1; /* c0 = (c6-c12)*2 */ - - z4 = z2 - z3; - z3 += z2; - tmp10 = MULTIPLY(z3, FIX(1.337628990)); /* (c2+c4)/2 */ - tmp11 = MULTIPLY(z4, FIX(0.045680613)); /* (c2-c4)/2 */ - z2 = MULTIPLY(z2, FIX(1.439773946)); /* c4+c14 */ - - tmp20 = tmp13 + tmp10 + tmp11; - tmp23 = tmp12 - tmp10 + tmp11 + z2; - - tmp10 = MULTIPLY(z3, FIX(0.547059574)); /* (c8+c14)/2 */ - tmp11 = MULTIPLY(z4, FIX(0.399234004)); /* (c8-c14)/2 */ - - tmp25 = tmp13 - tmp10 - tmp11; - tmp26 = tmp12 + tmp10 - tmp11 - z2; - - tmp10 = MULTIPLY(z3, FIX(0.790569415)); /* (c6+c12)/2 */ - tmp11 = MULTIPLY(z4, FIX(0.353553391)); /* (c6-c12)/2 */ - - tmp21 = tmp12 + tmp10 + tmp11; - tmp24 = tmp13 - tmp10 + tmp11; - tmp11 += tmp11; - tmp22 = z1 + tmp11; /* c10 = c6-c12 */ - tmp27 = z1 - tmp11 - tmp11; /* c0 = (c6-c12)*2 */ - - /* Odd part */ - - z1 = DEQUANTIZE(inptr[DCTSIZE*1], quantptr[DCTSIZE*1]); - z2 = DEQUANTIZE(inptr[DCTSIZE*3], quantptr[DCTSIZE*3]); - z4 = DEQUANTIZE(inptr[DCTSIZE*5], quantptr[DCTSIZE*5]); - z3 = MULTIPLY(z4, FIX(1.224744871)); /* c5 */ - z4 = DEQUANTIZE(inptr[DCTSIZE*7], quantptr[DCTSIZE*7]); - - tmp13 = z2 - z4; - tmp15 = MULTIPLY(z1 + tmp13, FIX(0.831253876)); /* c9 */ - tmp11 = tmp15 + MULTIPLY(z1, FIX(0.513743148)); /* c3-c9 */ - tmp14 = tmp15 - MULTIPLY(tmp13, FIX(2.176250899)); /* c3+c9 */ - - tmp13 = MULTIPLY(z2, - FIX(0.831253876)); /* -c9 */ - tmp15 = MULTIPLY(z2, - FIX(1.344997024)); /* -c3 */ - z2 = z1 - z4; - tmp12 = z3 + MULTIPLY(z2, FIX(1.406466353)); /* c1 */ - - tmp10 = tmp12 + MULTIPLY(z4, FIX(2.457431844)) - tmp15; /* c1+c7 */ - tmp16 = tmp12 - MULTIPLY(z1, FIX(1.112434820)) + tmp13; /* c1-c13 */ - tmp12 = MULTIPLY(z2, FIX(1.224744871)) - z3; /* c5 */ - z2 = MULTIPLY(z1 + z4, FIX(0.575212477)); /* c11 */ - tmp13 += z2 + MULTIPLY(z1, FIX(0.475753014)) - z3; /* c7-c11 */ - tmp15 += z2 - MULTIPLY(z4, FIX(0.869244010)) + z3; /* c11+c13 */ - - /* Final output stage */ - - wsptr[8*0] = (int) RIGHT_SHIFT(tmp20 + tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*14] = (int) RIGHT_SHIFT(tmp20 - tmp10, CONST_BITS-PASS1_BITS); - wsptr[8*1] = (int) RIGHT_SHIFT(tmp21 + tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*13] = (int) RIGHT_SHIFT(tmp21 - tmp11, CONST_BITS-PASS1_BITS); - wsptr[8*2] = (int) RIGHT_SHIFT(tmp22 + tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*12] = (int) RIGHT_SHIFT(tmp22 - tmp12, CONST_BITS-PASS1_BITS); - wsptr[8*3] = (int) RIGHT_SHIFT(tmp23 + tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*11] = (int) RIGHT_SHIFT(tmp23 - tmp13, CONST_BITS-PASS1_BITS); - wsptr[8*4] = (int) RIGHT_SHIFT(tmp24 + tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*10] = (int) RIGHT_SHIFT(tmp24 - tmp14, CONST_BITS-PASS1_BITS); - wsptr[8*5] = (int) RIGHT_SHIFT(tmp25 + tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*9] = (int) RIGHT_SHIFT(tmp25 - tmp15, CONST_BITS-PASS1_BITS); - wsptr[8*6] = (int) RIGHT_SHIFT(tmp26 + tmp16, CONST_BITS-PASS1_BITS); - wsptr[8*8] = (int) RIGHT_SHIFT(tmp26 - tmp16, CONST_BITS-PASS1_BITS); - wsptr[8*7] = (int) RIGHT_SHIFT(tmp27, CONST_BITS-PASS1_BITS); - } - - /* Pass 2: process 15 rows from work array, store into output array. */ - - wsptr = workspace; - for (ctr = 0; ctr < 15; ctr++) { - outptr = output_buf[ctr] + output_col; - - /* Even part */ - - /* Add fudge factor here for final descale. */ - z1 = (INT32) wsptr[0] + (ONE << (PA |