DRC [Sun, 11 May 2014 10:09:07 +0000 (10:09 +0000)]
Port the more accurate (and slightly faster) floating point IDCT implementation from jpeg-8a and later. New research revealed that the SSE/SSE2 floating point IDCT implementation was actually more accurate than the jpeg-6b implementation, not less, which is why its mathematical results have always differed from those of the jpeg-6b implementation. This patch brings the accuracy of the C code in line with that of the SSE/SSE2 code.
DRC [Sun, 11 May 2014 09:36:25 +0000 (09:36 +0000)]
Convert tabs to spaces in the libjpeg code and the SIMD code (TurboJPEG retains the use of tabs for historical reasons. They were annoying in the libjpeg code primarily because they were not consistently used and because they were used to format as well as indent the code. In the case of TurboJPEG, tabs are used just to indent the code, so even if the editor assumes a different tab width, the code will still be readable.)
DRC [Sat, 10 May 2014 09:53:34 +0000 (09:53 +0000)]
Using subdirectories unfortunately opened up a can of worms. In order to prevent object name conflicts, it is necessary to use the subdir-objects automake directive, but it simply doesn't work right on some of the versions of automake we still have to support. Another option would be to add a separate Makefile.am file to each subdirectory, but that requires maintaining a completely different set of build rules for each one. Fortunately, however, we're in the 21st century now, so we can use filenames longer than 8.3.
DRC [Fri, 9 May 2014 20:14:26 +0000 (20:14 +0000)]
Re-organize the x86/x86-64 SIMD routines into separate folders by instruction set so we can name each routine similarly to its corresponding C file. This also makes it easier to add support for new instruction sets.
DRC [Fri, 9 May 2014 18:00:32 +0000 (18:00 +0000)]
Convert tabs to spaces in the libjpeg code and the SIMD code (TurboJPEG retains the use of tabs for historical reasons. They were annoying in the libjpeg code primarily because they were not consistently used and because they were used to format as well as indent the code. In the case of TurboJPEG, tabs are used just to indent the code, so even if the editor assumes a different tab width, the code will still be readable.)
DRC [Wed, 7 May 2014 06:02:57 +0000 (06:02 +0000)]
Fix regression that caused 'make test' to fail with non-x86 SIMD code. The round-off error in the SIMD float DCT/IDCT routines only exists in the SSE and SSE2 implementations.
DRC [Tue, 6 May 2014 21:03:35 +0000 (21:03 +0000)]
Replace our custom version of Android.mk with instructions on how to build a libjpeg-turbo SDK for Android using autotools. Upon consulting with AOSP, it appears that Android.mk isn't really necessary except when building libjpeg-turbo for use by the Android platform itself, and it makes more sense for them to maintain the makefile for that purpose rather than for it to be upstreamed. ndk-build has serious limitations that prevent it from being used to generate static libjpeg-turbo libraries (mainly, it isn't possible to combine pre-built objects from one module into a static library for another module, which is necessary because the SIMD extensions sometimes have to be built with different CFLAGS than the rest of the code.) In general, it's just better not to introduce a new build system.
We use __CHAR_UNSIGNED__ (automatically defined by the AC_C_CHAR_UNSIGNED macro) rather than CHAR_IS_UNSIGNED (defined by custom autoconf code in libjpeg that we didn't port over), although I doubt it matters on any of the platforms we support.
We use __CHAR_UNSIGNED__ (automatically defined by the AC_C_CHAR_UNSIGNED macro) rather than CHAR_IS_UNSIGNED (defined by custom autoconf code in libjpeg that we didn't port over), although I doubt it matters on any of the platforms we support.
This patch accomplishes the following:
-- Auto-generates HAVE_LOCALE_H macro and adds it to jconfig.h (this is used by rdjpgcom.c.)
-- Reconciles the description and ordering of macros between config.h.in and jconfig.h.in, so the two files can be easily diffed.
-- Eliminates the use of the autoheader-generated config.h in the project and moves relevant internal-only macros into a new file, jconfigint.h. This is to avoid "already defined" warnings in files that were including both config.h (to get the internal autotools package information or the INLINE definition) and jconfig.h.
This patch accomplishes the following:
-- Auto-generates HAVE_LOCALE_H macro and adds it to jconfig.h (this is used by rdjpgcom.c.)
-- Reconciles the description and ordering of macros between config.h.in and jconfig.h.in, so the two files can be easily diffed.
-- Eliminates the use of the autoheader-generated config.h in the project and moves relevant internal-only macros into a new file, jconfigint.h. This is to avoid "already defined" warnings in files that were including both config.h (to get the internal autotools package information or the INLINE definition) and jconfig.h.
Work around an issue in Visual C++ 2010 and 2013 that was causing the 256-bit bitmap test in the regression tests to fail. More specifically, when optimization is enabled in these versions of Visual C++, the optimizer seems to get confused by the following code block:
Each time cur0 is incremented by delta, the compiled code doubles the value of delta (WTF?!) Thus, by the time the end of the block is reached, cur0 is equal to 15 times its former self, not 7 times its former self as it should be. At any rate, it was a lot simpler to just refactor the code so that it uses multiplication.
Work around an issue in Visual C++ 2010 and 2013 that was causing the 256-bit bitmap test in the regression tests to fail. More specifically, when optimization is enabled in these versions of Visual C++, the optimizer seems to get confused by the following code block:
Each time cur0 is incremented by delta, the compiled code doubles the value of delta (WTF?!) Thus, by the time the end of the block is reached, cur0 is equal to 15 times its former self, not 7 times its former self as it should be. At any rate, it was a lot simpler to just refactor the code so that it uses multiplication.
Work around an issue in Visual C++ 2010 and 2013 that was causing the 256-bit bitmap test in the regression tests to fail. More specifically, when optimization is enabled in these versions of Visual C++, the optimizer seems to get confused by the following code block:
Each time cur0 is incremented by delta, the compiled code doubles the value of delta (WTF?!) Thus, by the time the end of the block is reached, cur0 is equal to 15 times its former self, not 7 times its former self as it should be. At any rate, it was a lot simpler to just refactor the code so that it uses multiplication.
Set the BUILD environment variable properly when building with MinGW implementations that lack a Unix shell (TDM-GCC, for instance); Only set CMP0022 if it is supported (older CMake implementations don't support it); Remove trailing spaces
Set the BUILD environment variable properly when building with MinGW implementations that lack a Unix shell (TDM-GCC, for instance); Only set CMP0022 if it is supported (older CMake implementations don't support it); Remove trailing spaces
DRC [Fri, 28 Mar 2014 18:50:30 +0000 (18:50 +0000)]
Integrate a slightly modified version of Mozilla's patch for precomputing the bit-counting LUT. This is useful if the table needs to be shared among multiple processes, although the primary reason for doing that is reduced footprint on mobile devices, which are probably already covered by the clz intrinsic code.
DRC [Thu, 27 Mar 2014 20:47:39 +0000 (20:47 +0000)]
Modify svn:ignore so that it lists only files generated in-tree by autoreconf -fiv. It's pretty much impossible to maintain an up-to-date list of files generated by the actual build, so we'll just assume that the user will do an out-of-tree build if they want to avoid cluttering up their 'svn status' display.
DRC [Thu, 27 Mar 2014 20:43:57 +0000 (20:43 +0000)]
Modify svn:ignore so that it lists only files generated in-tree by autoreconf -fiv. It's pretty much impossible to maintain an up-to-date list of files generated by the actual build, so we'll just assume that the user will do an out-of-tree build if they want to avoid cluttering up their 'svn status' display.
DRC [Thu, 27 Mar 2014 03:36:04 +0000 (03:36 +0000)]
Make iOS build instructions more generic and applicable to all versions of Xcode; modify iOS build procedure for Xcode 5.0 and later to fix a build issue with Xcode 5.1.
DRC [Thu, 27 Mar 2014 03:35:21 +0000 (03:35 +0000)]
Make iOS build instructions more generic and applicable to all versions of Xcode; modify iOS build procedure for Xcode 5.0 and later to fix a build issue with Xcode 5.1.
DRC [Sun, 23 Mar 2014 18:06:11 +0000 (18:06 +0000)]
Remove any claims of support for OS X 10.4 "Tiger" (the packaging system overhaul produces packages that require Leopard or later, and I haven't been able to test Tiger for years anyhow.) Update TurboJPEG shared library version.
DRC [Sun, 23 Mar 2014 15:21:20 +0000 (15:21 +0000)]
Remove the sections about replacing libjpeg at run time and compile time. These were written before O/S distributions started shipping libjpeg-turbo, and they are either pedantic or no longer relevant. Also remove any text that assumes the use of our official project binaries. Notes specific to the official binaries have been moved into the project wiki.
DRC [Sat, 22 Mar 2014 20:51:38 +0000 (20:51 +0000)]
Since we're now maintaining our own Cygwin pseudo-repository directories instead of recommending that users install these packages from a local source, it makes more sense to name the packages according to Cygwin specs, so they can be copied as-is into the pseudo-repository.
DRC [Fri, 21 Mar 2014 23:34:53 +0000 (23:34 +0000)]
Since we're now maintaining our own Cygwin pseudo-repository directories instead of recommending that users install these packages from a local source, it makes more sense to name the packages according to Cygwin specs, so they can be copied as-is into the pseudo-repository.
DRC [Fri, 21 Mar 2014 11:01:00 +0000 (11:01 +0000)]
RHEL 6 (and probably other platforms as well) sets _defaultdocdir=%{_datadir}/doc, which screws things up, since we're overriding _datadir. Since we intend _defaultdocdir to be /usr/share/doc, just be explicit about it.
DRC [Fri, 21 Mar 2014 11:00:00 +0000 (11:00 +0000)]
RHEL 6 (and probably other platforms as well) sets _defaultdocdir=%{_datadir}/doc, which screws things up, since we're overriding _datadir. Since we intend _defaultdocdir to be /usr/share/doc, just be explicit about it.
DRC [Mon, 17 Mar 2014 11:14:52 +0000 (11:14 +0000)]
Extend YUVImage class to allow reuse of the same buffer with different metadata; port TJBench changes that treat YUV encoding/decoding as an intermediate step of the JPEG compression/decompression pipeline rather than a separate test case; add YUV encode/decode tests to the Java version of tjbenchtest
DRC [Mon, 17 Mar 2014 10:13:17 +0000 (10:13 +0000)]
Move the garbage collection of the JPEG tiles into the decompression function to increase the chances that tiled decompression of large images will succeed without an OutOfMemoryError.
DRC [Mon, 17 Mar 2014 10:12:11 +0000 (10:12 +0000)]
Move the garbage collection of the JPEG tiles into the decompression function to increase the chances that tiled decompression of large images will succeed without an OutOfMemoryError.
projects / libjpeg-turbo / log