jcdctmgr.c


Log

Author Commit Date CI Message
MIYASAKA Masaru a2e6a9dd 2006-02-04T00:00:00 IJG R6b with x86SIMD V1.02 Independent JPEG Group's JPEG software release 6b with x86 SIMD extension for IJG JPEG library version 1.02
Thomas G. Lane 489583f5 1996-02-07T00:00:00 The Independent JPEG Group's JPEG software v6a
Thomas G. Lane bc79e068 1995-08-02T00:00:00 The Independent JPEG Group's JPEG software v6
Thomas G. Lane 36a4cccc 1994-09-24T00:00:00 The Independent JPEG Group's JPEG software v5
Guido Vollbeding 5996a25e 2009-06-27T00:00:00 The Independent JPEG Group's JPEG software v7
DRC a73e870a 2012-12-31T02:52:30 Change the copyright notices to make it clear that our modified files are not part of the IJG's software. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/branches/1.2.x@873 632fc199-4ca6-4c93-a231-07263d6284db
DRC 0f0fd751 2012-02-07T23:27:14 Compiler warnings git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/branches/1.1.x@784 632fc199-4ca6-4c93-a231-07263d6284db
DRC d65d99a9 2012-01-31T03:39:23 Compiler warnings git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/branches/1.2.x@758 632fc199-4ca6-4c93-a231-07263d6284db
DRC a49c4e52 2011-02-18T20:50:08 The SIMD quantization algorithm does not produce correct results with the fast forward integer DCT and JPEG qualities >= 98, so for now, use the non-SIMD quantization function under those circumstances. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@395 632fc199-4ca6-4c93-a231-07263d6284db
DRC 6ca69537 2011-02-18T03:31:11 The SIMD quantization algorithm does not produce correct results with the fast forward integer DCT and JPEG qualities >= 98, so for now, use the non-SIMD quantization function under those circumstances. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/branches/1.0.x@383 632fc199-4ca6-4c93-a231-07263d6284db
DRC 72abc297 2011-02-18T01:45:24 The SIMD quantization algorithm does not produce correct results with the fast forward integer DCT and JPEG qualities >= 98, so for now, use the non-SIMD quantization function under those circumstances. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/branches/1.1.x@378 632fc199-4ca6-4c93-a231-07263d6284db
DRC fc5dc4fa 2009-10-01T22:26:14 Some systems (notably OS X Leopard) have fls() already, so rename ours to avoid conflict git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@66 632fc199-4ca6-4c93-a231-07263d6284db
Pierre Ossman dc5db14a 2009-03-13T12:17:26 Move variable init around a bit to please crappy compilers. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@26 632fc199-4ca6-4c93-a231-07263d6284db
Pierre Ossman 35c47196 2009-03-09T13:29:37 Make sure the work space memory is properly aligned We use the heap allocators to avoid having more than one implementation of the alignment logic. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@19 632fc199-4ca6-4c93-a231-07263d6284db
Pierre Ossman dedc42e2 2009-03-09T13:23:04 "Optimise" quantization step by replacing the division by a multiplication. This has no measurable difference right now but makes it possible to do SIMD implementations of this stage. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@16 632fc199-4ca6-4c93-a231-07263d6284db
Pierre Ossman 59a3938b 2009-03-09T13:15:56 Framework for supporting SIMD acceleration Designed to impose minimal changes on the "normal" code. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@14 632fc199-4ca6-4c93-a231-07263d6284db
Pierre Ossman 49dcbfbf 2009-03-09T10:37:20 Split up the forward DCT routine into three stages Divide it into sample conversion, DCT and quantization in order to easily provide alternative implementations of each stage. git-svn-id: svn+ssh://svn.code.sf.net/p/libjpeg-turbo/code/trunk@13 632fc199-4ca6-4c93-a231-07263d6284db