3537   No changes.

  3538 

  3539 Version 1.5.5beta01 [July 13, 2011]

  3540   Fixed some typos and made other minor changes in the manual.

  3541   Updated contrib/pngminus/makefile.std (Samuli Souminen)

  3542 

  3543 Version 1.5.5beta02 [July 14, 2011]

  3544   Revised Makefile.am and Makefile.in to look in the right directory for

  3545     pnglibconf.h.prebuilt

  3546 

  3547 Version 1.5.5beta03 [July 27, 2011]

  3548   Enabled compilation with g++ compiler.  This compiler does not recognize

  3549     the file extension, so it always compiles with C++ rules.  Made minor

  3550     changes to pngrutil.c to cast results where C++ expects it but C does not.

  3551   Minor editing of libpng.3 and libpng-manual.txt.

  3552 

  3553 Version 1.5.5beta04 [July 29, 2011]

  3554   Revised CMakeLists.txt (Clifford Yapp)

  3555   Updated commentary about the png_rgb_to_gray() default coefficients

  3556     in the manual and in pngrtran.c

  3557 

  3558 Version 1.5.5beta05 [August 17, 2011]

  3559   Prevent unexpected API exports from non-libpng DLLs on Windows.  The "_DLL"

  3560     is removed from the test of whether a DLL is being built (this erroneously

  3561     caused the libpng APIs to be marked as DLL exports in static builds under

  3562     Microsoft Visual Studio).  Almost all of the libpng building configuration

  3563     is moved from pngconf.h to pngpriv.h, but PNG_DLL_EXPORT remains in

  3564     pngconf.h, though, so that it is colocated with the import definition (it

  3565     is no longer used anywhere in the installed headers).  The VStudio project

  3566     definitions have been cleaned up: "_USRDLL" has been removed from the

  3567     static library builds (this was incorrect), and PNG_USE_DLL has been added

  3568     to pngvalid to test the functionality (pngtest does not supply it,

  3569     deliberately).  The spurious "_EXPORTS" has been removed from the

  3570     libpng build (all these errors were a result of copy/paste between project

  3571     configurations.)

  3572   Added new types and internal functions for CIE RGB end point handling to

  3573     pngpriv.h (functions yet to be implemented).

  3574 

  3575 Version 1.5.5beta06 [August 26, 2011]

  3576   Ensure the CMAKE_LIBRARY_OUTPUT_DIRECTORY is set in CMakeLists.txt

  3577     (Clifford Yap)

  3578   Fixes to rgb_to_gray and cHRM XYZ APIs (John Bowler):

  3579     The rgb_to_gray code had errors when combined with gamma correction.

  3580     Some pixels were treated as true grey when they weren't and such pixels

  3581     and true grey ones were not gamma corrected (the original value of the

  3582     red component was used instead).  APIs to get and set cHRM using color

  3583     space end points have been added and the rgb_to_gray code that defaults

  3584     based on cHRM, and the divide-by-zero bug in png_handle_cHRM (CERT

  3585     VU#477046, CVE-2011-3328, introduced in 1.5.4) have been corrected.

  3586   A considerable number of tests has been added to pngvalid for the

  3587     rgb_to_gray transform.

  3588   Arithmetic errors in rgb_to_gray whereby the calculated gray value was

  3589     truncated to the bit depth rather than rounded have been fixed except in

  3590     the 8-bit non-gamma-corrected case (where consistency seems more important

  3591     than correctness.)  The code still has considerable inaccuracies in the

  3592     8-bit case because 8-bit linear arithmetic is used.

  3593 

  3594 Version 1.5.5beta07 [September 7, 2011]

  3595   Added "$(ARCH)" option to makefile.darwin

  3596   Added SunOS support to configure.ac and Makefile.am

  3597   Changed png_chunk_benign_error() to png_warning() in png.c, in

  3598     png_XYZ_from_xy_checked().

  3599 

  3600 Version 1.5.5beta08 [September 10, 2011]

  3601   Fixed 64-bit compilation errors (gcc). The errors fixed relate

  3602     to conditions where types that are 32 bits in the GCC 32-bit

  3603     world (uLong and png_size_t) become 64 bits in the 64-bit

  3604     world.  This produces potential truncation errors which the

  3605     compiler correctly flags.

  3606   Relocated new HAVE_SOLARIS_LD definition in configure.ac

  3607   Constant changes for 64-bit compatibility (removal of L suffixes). The

  3608     16-bit cases still use "L" as we don't have a 16-bit test system.

  3609 

  3610 Version 1.5.5rc01 [September 15, 2011]

  3611   Removed "L" suffixes in pngpriv.h

  3612 

  3613 Version 1.5.5 [September 22, 2011]

  3614   No changes.

  3615 

  3616 Version 1.5.6beta01 [September 22, 2011]

  3617   Fixed some 64-bit type conversion warnings in pngrtran.c

  3618   Moved row_info from png_struct to a local variable.

  3619   The various interlace mask arrays have been made into arrays of

  3620     bytes and made PNG_CONST and static (previously some arrays were

  3621     marked PNG_CONST and some weren't).

  3622   Additional checks have been added to the transform code to validate the

  3623     pixel depths after the transforms on both read and write.

  3624   Removed some redundant code from pngwrite.c, in png_destroy_write_struct().

  3625   Changed chunk reading/writing code to use png_uint_32 instead of png_byte[4].

  3626     This removes the need to allocate temporary strings for chunk names on

  3627     the stack in the read/write code.  Unknown chunk handling still uses the

  3628     string form because this is exposed in the API.

  3629 

  3630 Version 1.5.6beta02 [September 26, 2011]

  3631   Added a note in the manual the png_read_update_info() must be called only

  3632     once with a particular info_ptr.

  3633   Fixed a typo in the definition of the new PNG_STRING_FROM_CHUNK(s,c) macro.

  3634 

  3635 Version 1.5.6beta03 [September 28, 2011]

  3636   Revised test-pngtest.sh to report FAIL when pngtest fails.

  3637   Added "--strict" option to pngtest, to report FAIL when the failure is

  3638     only because the resulting valid files are different.

  3639   Revised CMakeLists.txt to work with mingw and removed some material from

  3640     CMakeLists.txt that is no longer useful in libpng-1.5.

  3641 

  3642 Version 1.5.6beta04 [October 5, 2011]

  3643   Fixed typo in Makefile.in and Makefile.am ("-M Wl" should be "-M -Wl")."

  3644 

  3645 Version 1.5.6beta05 [October 12, 2011]

  3646   Speed up png_combine_row() for interlaced images. This reduces the generality

  3647     of the code, allowing it to be optimized for Adam7 interlace.  The masks

  3648     passed to png_combine_row() are now generated internally, avoiding

  3649     some code duplication and localizing the interlace handling somewhat.

  3650   Align png_struct::row_buf - previously it was always unaligned, caused by

  3651     a bug in the code that attempted to align it; the code needs to subtract

  3652     one from the pointer to take account of the filter byte prepended to

  3653     each row.

  3654   Optimized png_combine_row() when rows are aligned. This gains a small

  3655     percentage for 16-bit and 32-bit pixels in the typical case where the

  3656     output row buffers are appropriately aligned. The optimization was not

  3657     previously possible because the png_struct buffer was always misaligned.

  3658   Fixed bug in png_write_chunk_header() debug print, introduced in 1.5.6beta01.

  3659 

  3660 Version 1.5.6beta06 [October 17, 2011]

  3661   Removed two redundant tests for unitialized row.

  3662   Fixed a relatively harmless memory overwrite in compressed text writing

  3663     with a 1 byte zlib buffer.

  3664   Add ability to call png_read_update_info multiple times to pngvalid.c.

  3665   Fixes for multiple calls to png_read_update_info. These fixes attend to

  3666     most of the errors revealed in pngvalid, however doing the gamma work

  3667     twice results in inaccuracies that can't be easily fixed.  There is now

  3668     a warning in the code if this is going to happen.

  3669   Turned on multiple png_read_update_info in pngvalid transform tests.

  3670   Prevent libpng from overwriting unused bits at the end of the image when

  3671     it is not byte aligned, while reading. Prior to libpng-1.5.6 libpng would

  3672     overwrite the partial byte at the end of each row if the row width was not

  3673     an exact multiple of 8 bits and the image is not interlaced.

  3674 

  3675 Version 1.5.6beta07 [October 21, 2011]

  3676   Made png_ptr->prev_row an aligned pointer into png_ptr->big_prev_row

  3677     (Mans Rullgard).

  3678 

  3679 Version 1.5.6rc01 [October 26, 2011]

  3680   Changed misleading "Missing PLTE before cHRM" warning to "Out of place cHRM"

  3681 

  3682 Version 1.5.6rc02 [October 27, 2011]

  3683   Added LSR() macro to defend against buggy compilers that evaluate non-taken

  3684     code branches and complain about out-of-range shifts.

  3685 

  3686 Version 1.5.6rc03 [October 28, 2011]

  3687   Renamed the LSR() macro to PNG_LSR() and added PNG_LSL() macro.

  3688   Fixed compiler warnings with Intel and MSYS compilers. The logical shift

  3689     fix for Microsoft Visual C is required by other compilers, so this

  3690     enables that fix for all compilers when using compile-time constants.

  3691     Under MSYS 'byte' is a name declared in a system header file, so we

  3692     changed the name of a local variable to avoid the warnings that result.

  3693   Added #define PNG_ALIGN_TYPE PNG_ALIGN_NONE to contrib/pngminim/*/pngusr.h

  3694 

  3695 Version 1.5.6 [November 3, 2011]

  3696   No changes.

  3697 

  3698 Version 1.5.7beta01 [November 4, 2011]

  3699   Added support for ARM processor, when decoding all PNG up-filtered rows

  3700     and any other-filtered rows with 3 or 4 bytes per pixel (Mans Rullgard).

  3701   Fixed bug in pngvalid on early allocation failure; fixed type cast in

  3702     pngmem.c; pngvalid would attempt to call png_error() if the allocation

  3703     of a png_struct or png_info failed. This would probably have led to a

  3704     crash.  The pngmem.c implementation of png_malloc() included a cast

  3705     to png_size_t which would fail on large allocations on 16-bit systems.

  3706   Fix for the preprocessor of the Intel C compiler. The preprocessor

  3707     splits adjacent @ signs with a space; this changes the concatentation

  3708     token from @-@-@ to PNG_JOIN; that should work with all compiler

  3709     preprocessors.

  3710   Paeth filter speed improvements from work by Siarhei Siamashka. This

  3711     changes the 'Paeth' reconstruction function to improve the GCC code

  3712     generation on x86. The changes are only part of the suggested ones;

  3713     just the changes that definitely improve speed and remain simple.

  3714     The changes also slightly increase the clarity of the code.

  3715 

  3716 Version 1.5.7beta02 [November 11, 2011]

  3717   Check compression_type parameter in png_get_iCCP and remove spurious

  3718     casts. The compression_type parameter is always assigned to, so must

  3719     be non-NULL. The cast of the profile length potentially truncated the

  3720     value unnecessarily on a 16-bit int system, so the cast of the (byte)

  3721     compression type to (int) is specified by ANSI-C anyway.

  3722   Fixed FP division by zero in pngvalid.c; the 'test_pixel' code left

  3723     the sBIT fields in the test pixel as 0, which resulted in a floating

  3724     point division by zero which was irrelevant but causes systems where

  3725     FP exceptions cause a crash. Added code to pngvalid to turn on FP

  3726     exceptions if the appropriate glibc support is there to ensure this is

  3727     tested in the future.

  3728   Updated scripts/pnglibconf.mak and scripts/makefile.std to handle the

  3729     new PNG_JOIN macro.

  3730   Added versioning to pnglibconf.h comments.

  3731   Simplified read/write API initial version; basic read/write tested on

  3732     a variety of images, limited documentation (in the header file.)

  3733   Installed more accurate linear to sRGB conversion tables. The slightly

  3734     modified tables reduce the number of 16-bit values that

  3735     convert to an off-by-one 8-bit value.  The "makesRGB.c" code that was used

  3736     to generate the tables is now in a contrib/sRGBtables sub-directory.

  3737 

  3738 Version 1.5.7beta03 [November 17, 2011]

  3739   Removed PNG_CONST from the sRGB table declarations in pngpriv.h and png.c

  3740   Added run-time detection of NEON support.

  3741   Added contrib/libtests; includes simplified API test and timing test and

  3742     a color conversion utility for rapid checking of failed 'pngstest' results.

  3743   Multiple transform bug fixes plus a work-round for double gamma correction.

  3744     libpng does not support more than one transform that requires linear data

  3745     at once - if this is tried typically the results is double gamma

  3746     correction. Since the simplified APIs can need rgb to gray combined with

  3747     a compose operation it is necessary to do one of these outside the main

  3748     libpng transform code. This check-in also contains fixes to various bugs

  3749     in the simplified APIs themselves and to some bugs in compose and rgb to

  3750     gray (on palette) itself.

  3751   Fixes for C++ compilation using g++ When libpng source is compiled

  3752     using g++. The compiler imposes C++ rules on the C source; thus it

  3753     is desireable to make the source work with either C or C++ rules

  3754     without throwing away useful error information.  This change adds

  3755     png_voidcast to allow C semantic (void*) cases or the corresponding

  3756     C++ static_cast operation, as appropriate.

  3757   Added --noexecstack to assembler file compilation. GCC does not set

  3758     this on assembler compilation, even though it does on C compilation.

  3759     This creates security issues if assembler code is enabled; the

  3760     work-around is to set it by default in the flags for $(CCAS)

  3761   Work around compilers that don't support declaration of const data. Some

  3762     compilers fault 'extern const' data declarations (because the data is

  3763     not initialized); this turns on const-ness only for compilers where

  3764     this is known to work.

  3765 

  3766 Version 1.5.7beta04 [November 17, 2011]

  3767   Since the gcc driver does not recognize the --noexecstack flag, we must

  3768     use the -Wa prefix to have it passed through to the assembler.

  3769     Also removed a duplicate setting of this flag.

  3770   Added files that were omitted from the libpng-1.5.7beta03 zip distribution.

  3771 

  3772 Version 1.5.7beta05 [November 25, 2011]

  3773   Removed "zTXt" from warning in generic chunk decompression function.

  3774   Validate time settings passed to pngset() and png_convert_to_rfc1123()

  3775     (Frank Busse).

  3776   Added MINGW support to CMakeLists.txt

  3777   Reject invalid compression flag or method when reading the iTXt chunk.

  3778   Backed out 'simplified' API changes. The API seems too complex and there

  3779     is a lack of consensus or enthusiasm for the proposals.  The API also

  3780     reveals significant bugs inside libpng (double gamma correction and the

  3781     known bug of being unable to retrieve a corrected palette). It seems

  3782     better to wait until the bugs, at least, are corrected.

  3783   Moved pngvalid.c into contrib/libtests

  3784   Rebuilt Makefile.in, configure, etc., with autoconf-2.68

  3785 

  3786 Version 1.5.7rc01 [December 1, 2011]

  3787   Replaced an "#if" with "#ifdef" in pngrtran.c

  3788   Revised #if PNG_DO_BC block in png.c (use #ifdef and add #else)

  3789 

  3790 Version 1.5.7rc02 [December 5, 2011]

  3791   Revised project files and contrib/pngvalid/pngvalid.c to account for

  3792     the relocation of pngvalid into contrib/libtests.

  3793   Revised pngconf.h to use " __declspec(restrict)" only when MSC_VER >= 1400,

  3794     as in libpng-1.5.4.

  3795   Put CRLF line endings in the owatcom project files.

  3796 

  3797 Version 1.5.7rc03 [December 7, 2011]

  3798   Updated CMakeLists.txt to account for the relocation of pngvalid.c

  3799 

  3800 Version 1.5.7 [December 15, 2011]

  3801   Minor fixes to pngvalid.c for gcc 4.6.2 compatibility to remove warnings

  3802     reported by earlier versions.

  3803   Fixed minor memset/sizeof errors in pngvalid.c.

  3804 

  3805 Version 1.6.0beta01 [December 15, 2011]

  3806   Removed machine-generated configure files from the GIT repository (they will

  3807     continue to appear in the tarball distributions and in the libpng15 and

  3808     earlier GIT branches).

  3809   Restored the new 'simplified' API, which was started in libpng-1.5.7beta02

  3810     but later deleted from libpng-1.5.7beta05.

  3811   Added example programs for the new 'simplified' API.

  3812   Added ANSI-C (C90) headers and require them, and take advantage of the

  3813     change. Also fixed some of the projects/* and contrib/* files that needed

  3814     updates for libpng16 and the move of pngvalid.c.

  3815     With this change the required ANSI-C header files are assumed to exist: the

  3816     implementation must provide float.h, limits.h, stdarg.h and stddef.h and

  3817     libpng relies on limits.h and stddef.h existing and behaving as defined

  3818     (the other two required headers aren't used).  Non-ANSI systems that don't

  3819     have stddef.h or limits.h will have to provide an appropriate fake

  3820     containing the relevant types and #defines.

  3821   The use of FAR/far has been eliminated and the definition of png_alloc_size_t

  3822     is now controlled by a flag so that 'small size_t' systems can select it

  3823     if necessary.  Libpng 1.6 may not currently work on such systems -- it

  3824     seems likely that it will ask 'malloc' for more than 65535 bytes with any

  3825     image that has a sufficiently large row size (rather than simply failing

  3826     to read such images).

  3827   New tools directory containing tools used to generate libpng code.

  3828   Fixed race conditions in parallel make builds. With higher degrees of

  3829     parallelism during 'make' the use of the same temporary file names such

  3830     as 'dfn*' can result in a race where a temporary file from one arm of the

  3831     build is deleted or overwritten in another arm.  This changes the

  3832     temporary files for suffix rules to always use $* and ensures that the

  3833     non-suffix rules use unique file names.

  3834 

  3835 Version 1.6.0beta02 [December 21, 2011]

  3836   Correct configure builds where build and source directories are separate.

  3837     The include path of 'config.h' was erroneously made relative in pngvalid.c

  3838     in libpng 1.5.7.

  3839 

  3840 Version 1.6.0beta03 [December 22, 2011]

  3841   Start-up code size improvements, error handler flexibility. These changes

  3842     alter how the tricky allocation of the initial png_struct and png_info

  3843     structures are handled. png_info is now handled in pretty much the same

  3844     way as everything else, except that the allocations handle NULL return

  3845     silently.  png_struct is changed in a similar way on allocation and on

  3846     deallocation a 'safety' error handler is put in place (which should never

  3847     be required).  The error handler itself is changed to permit mismatches

  3848     in the application and libpng error buffer size; however, this means a

  3849     silent change to the API to return the jmp_buf if the size doesn't match

  3850     the size from the libpng compilation; libpng now allocates the memory and

  3851     this may fail.  Overall these changes result in slight code size

  3852     reductions; however, this is a reduction in code that is always executed

  3853     so is particularly valuable.  Overall on a 64-bit system the libpng DLL

  3854     decreases in code size by 1733 bytes.  pngerror.o increases in size by

  3855     about 465 bytes because of the new functionality.

  3856   Added png_convert_to_rfc1123_buffer() and deprecated png_convert_to_rfc1123()

  3857     to avoid including a spurious buffer in the png_struct.

  3858 

  3859 Version 1.6.0beta04 [December 30, 2011]

  3860   Regenerated configure scripts with automake-1.11.2

  3861   Eliminated png_info_destroy(). It is now used only in png.c and only calls

  3862     one other internal function and memset().

  3863   Enabled png_get_sCAL_fixed() if floating point APIs are enabled. Previously

  3864     it was disabled whenever internal fixed point arithmetic was selected,

  3865     which meant it didn't exist even on systems where FP was available but not

  3866     preferred.

  3867   Added pngvalid.c compile time checks for const APIs.

  3868   Implemented 'restrict' for png_info and png_struct. Because of the way

  3869     libpng works both png_info and png_struct are always accessed via a

  3870     single pointer.  This means adding C99 'restrict' to the pointer gives

  3871     the compiler some opportunity to optimize the code.  This change allows

  3872     that.

  3873   Moved AC_MSG_CHECKING([if libraries can be versioned]) later to the proper

  3874     location in configure.ac (Gilles Espinasse).

  3875   Changed png_memcpy to C assignment where appropriate. Changed all those

  3876     uses of png_memcpy that were doing a simple assignment to assignments

  3877     (all those cases where the thing being copied is a non-array C L-value).

  3878   Added some error checking to png_set_*() routines.

  3879   Removed the reference to the non-exported function png_memcpy() from

  3880     example.c.

  3881   Fixed the Visual C 64-bit build - it requires jmp_buf to be aligned, but

  3882     it had become misaligned.

  3883   Revised contrib/pngminus/pnm2png.c to avoid warnings when png_uint_32

  3884     and unsigned long are of different sizes.

  3885 

  3886 Version 1.6.0beta05 [January 15, 2012]

  3887   Updated manual with description of the simplified API (copied from png.h)

  3888   Fix bug in pngerror.c: some long warnings were being improperly truncated

  3889     (CVE-2011-3464, bug introduced in libpng-1.5.3beta05).

  3890 

  3891 Version 1.6.0beta06 [January 24, 2012]

  3892   Added palette support to the simplified APIs. This commit

  3893     changes some of the macro definitions in png.h, app code

  3894     may need corresponding changes.

  3895   Increased the formatted warning buffer to 192 bytes.

  3896   Added color-map support to simplified API. This is an initial version for

  3897     review; the documentation has not yet been updated.

  3898   Fixed Min/GW uninstall to remove libpng.dll.a

  3899 

  3900 Version 1.6.0beta07 [January 28, 2012]

  3901   Eliminated Intel icc/icl compiler warnings. The Intel (GCC derived)

  3902     compiler issues slightly different warnings from those issued by the

  3903     current vesions of GCC. This eliminates those warnings by

  3904     adding/removing casts and small code rewrites.

  3905   Updated configure.ac from autoupdate: added --enable-werror option.

  3906     Also some layout regularization and removal of introduced tab characters

  3907     (replaced with 3-character indentation).  Obsolete macros identified by

  3908     autoupdate have been removed; the replacements are all in 2.59 so

  3909     the pre-req hasn't been changed.  --enable-werror checks for support

  3910     for -Werror (or the given argument) in the compiler.  This mimics the

  3911     gcc configure option by allowing -Werror to be turned on safely; without

  3912     the option the tests written in configure itself fail compilation because

  3913     they cause compiler warnings.

  3914   Rewrote autogen.sh to run autoreconf instead of running tools one-by-one.

  3915   Conditionalize the install rules for MINGW and CYGWIN in CMakeLists.txt and

  3916     set CMAKE_LIBRARY_OUTPUT_DIRECTORY to "lib" on all platforms (C. Yapp).

  3917   Freeze libtool files in the 'scripts' directory. This version of autogen.sh

  3918     attempts to dissuade people from running it when it is not, or should not,

  3919     be necessary.  In fact, autogen.sh does not work when run in a libpng

  3920     directory extracted from a tar distribution anymore. You must run it in

  3921     a GIT clone instead.

  3922   Added two images to contrib/pngsuite (1-bit and 2-bit transparent grayscale),

  3923     and renamed three whose names were inconsistent with those in

  3924     pngsuite/README.txt.

  3925 

  3926 Version 1.6.0beta08 [February 1, 2012]

  3927   Fixed Image::colormap misalignment in pngstest.c

  3928   Check libtool/libtoolize version number (2.4.2) in configure.ac

  3929   Divide test-pngstest.sh into separate pngstest runs for basic and

  3930     transparent images.

  3931   Moved automake options to AM_INIT_AUTOMAKE in configure.ac

  3932   Added color-tests, silent-rules (Not yet implemented in Makefile.am) and

  3933     version checking to configure.ac

  3934   Improved pngstest speed by not doing redundant tests and add const to

  3935     the background parameter of png_image_finish_read. The --background

  3936     option is now done automagically only when required, so that commandline

  3937     option no longer exists.

  3938   Cleaned up pngpriv.h to consistently declare all functions and data.

  3939     Also eliminated PNG_CONST_DATA, which is apparently not needed but we

  3940     can't be sure until it is gone.

  3941   Added symbol prefixing that allows all the libpng external symbols

  3942     to be prefixed (suggested by Reuben Hawkins).

  3943   Updated "ftbb*.png" list in the owatcom and vstudio projects.

  3944   Fixed 'prefix' builds on clean systems. The generation of pngprefix.h

  3945     should not require itself.

  3946   Updated INSTALL to explain that autogen.sh must be run in a GIT clone,

  3947     not in a libpng directory extracted from a tar distribution.

  3948 

  3949 Version 1.6.0beta09 [February 1, 2012]

  3950   Reverted the prebuilt configure files to libpng-1.6.0beta05 condition.

  3951 

  3952 Version 1.6.0beta10 [February 3, 2012]

  3953   Added Z_SOLO for zlib-1.2.6+ and correct pngstest tests

  3954   Updated list of test images in CMakeLists.txt

  3955   Updated the prebuilt configure files to current condition.

  3956   Revised INSTALL information about autogen.sh; it works in tar distributions.

  3957 

  3958 Version 1.6.0beta11 [February 16, 2012]

  3959   Fix character count in pngstest command in projects/owatcom/pngstest.tgt

  3960   Revised test-pngstest.sh to report PASS/FAIL for each image.

  3961   Updated documentation about the simplified API.

  3962   Corrected estimate of error in libpng png_set_rgb_to_gray API.  The API is

  3963     extremely inaccurate for sRGB conversions because it uses an 8-bit

  3964     intermediate linear value and it does not use the sRGB transform, so it

  3965     suffers from the known instability in gamma transforms for values close

  3966     to 0 (see Poynton).  The net result is that the calculation has a maximum

  3967     error of 14.99/255; 0.5/255^(1/2.2).  pngstest now uses 15 for the

  3968     permitted 8-bit error. This may still not be enough because of arithmetic

  3969     error.

  3970   Removed some unused arrays (with #ifdef) from png_read_push_finish_row().

  3971   Fixed a memory overwrite bug in simplified read of RGB PNG with

  3972     non-linear gamma Also bugs in the error checking in pngread.c and changed

  3973     quite a lot of the checks in pngstest.c to be correct; either correctly

  3974     written or not over-optimistic.  The pngstest changes are insufficient to

  3975     allow all possible RGB transforms to be passed; pngstest cmppixel needs

  3976     to be rewritten to make it clearer which errors it allows and then changed

  3977     to permit known inaccuracies.

  3978   Removed tests for no-longer-used *_EMPTY_PLTE_SUPPORTED from pngstruct.h

  3979   Fixed fixed/float API export conditionals. 1) If FIXED_POINT or

  3980     FLOATING_POINT options were switched off, png.h ended up with lone ';'

  3981     characters.  This is not valid ANSI-C outside a function.  The ';'

  3982     characters have been moved inside the definition of PNG_FP_EXPORT and

  3983     PNG_FIXED_EXPORT. 2) If either option was switched off, the declaration

  3984     of the corresponding functions were completely omitted, even though some

  3985     of them are still used internally.  The result is still valid, but

  3986     produces warnings from gcc with some warning options (including -Wall). The

  3987     fix is to cause png.h to declare the functions with PNG_INTERNAL_FUNCTION

  3988     when png.h is included from pngpriv.h.

  3989   Check for invalid palette index while reading paletted PNG.  When one is

  3990     found, issue a warning and increase png_ptr->num_palette accordingly.

  3991     Apps are responsible for checking to see if that happened.

  3992 

  3993 Version 1.6.0beta12 [February 18, 2012]

  3994   Do not increase num_palette on invalid_index.

  3995   Relocated check for invalid palette index to pngrtran.c, after unpacking

  3996     the sub-8-bit pixels.

  3997   Fixed CVE-2011-3026 buffer overrun bug.  This bug was introduced when

  3998     iCCP chunk support was added at libpng-1.0.6. Deal more correctly with the

  3999     test on iCCP chunk length. Also removed spurious casts that may hide

  4000     problems on 16-bit systems.

  4001 

  4002 Version 1.6.0beta13 [February 24, 2012]

  4003   Eliminated redundant png_push_read_tEXt|zTXt|iTXt|unknown code from

  4004     pngpread.c and use the sequential png_handle_tEXt, etc., in pngrutil.c;

  4005     now that png_ptr->buffer is inaccessible to applications, the special

  4006     handling is no longer useful.

  4007   Added PNG_SAFE_LIMITS feature to pnglibconf.dfa, pngpriv.h, and new

  4008     pngusr.dfa to reset the user limits to safe ones if PNG_SAFE_LIMITS is

  4009     defined.  To enable, use "CPPFLAGS=-DPNG_SAFE_LIMITS_SUPPORTED=1" on the

  4010     configure command or put #define PNG_SAFE_LIMITS_SUPPORTED in

  4011     pnglibconf.h.prebuilt and pnglibconf.h.

  4012 

  4013 Version 1.6.0beta14 [February 27, 2012]

  4014   Added information about the new limits in the manual.

  4015   Updated Makefile.in

  4016 

  4017 Version 1.6.0beta15 [March 2, 2012]

  4018   Removed unused "current_text" members of png_struct and the png_free()

  4019     of png_ptr->current_text from pngread.c

  4020   Rewrote pngstest.c for substantial speed improvement.

  4021   Fixed transparent pixel and 16-bit rgb tests in pngstest and removed a

  4022     spurious check in pngwrite.c

  4023   Added PNG_IMAGE_FLAG_FAST for the benefit of applications that store

  4024     intermediate files, or intermediate in-memory data, while processing

  4025     image data with the simplified API.  The option makes the files larger

  4026     but faster to write and read.  pngstest now uses this by default; this

  4027     can be disabled with the --slow option.

  4028   Improved pngstest fine tuning of error numbers, new test file generator.

  4029     The generator generates images that test the full range of sample values,

  4030     allow the error numbers in pngstest to be tuned and checked.  makepng

  4031     also allows generation of images with extra chunks, although this is

  4032     still work-in-progress.

  4033   Added check for invalid palette index while reading.

  4034   Fixed some bugs in ICC profile writing. The code should now accept

  4035     all potentially valid ICC profiles and reject obviously invalid ones.

  4036     It now uses png_error() to do so rather than casually writing a PNG

  4037     without the necessary color data.

  4038   Removed whitespace from the end of lines in all source files and scripts.

  4039 

  4040 Version 1.6.0beta16 [March 6, 2012]

  4041   Relocated palette-index checking function from pngrutil.c to pngtrans.c

  4042   Added palette-index checking while writing.

  4043   Changed png_inflate() and calling routines to avoid overflow problems.

  4044     This is an intermediate check-in that solves the immediate problems and

  4045     introduces one performance improvement (avoiding a copy via png_ptr->zbuf.)

  4046     Further changes will be made to make ICC profile handling more secure.

  4047   Fixed build warnings (MSVC, GCC, GCC v3). Cygwin GCC with default options

  4048     declares 'index' as a global, causing a warning if it is used as a local

  4049     variable.  GCC 64-bit warns about assigning a (size_t) (unsigned 64-bit)

  4050     to an (int) (signed 32-bit).  MSVC, however, warns about using the

  4051     unary '-' operator on an unsigned value (even though it is well defined

  4052     by ANSI-C to be ~x+1).  The padding calculation was changed to use a

  4053     different method.  Removed the tests on png_ptr->pass.

  4054   Added contrib/libtests/tarith.c to test internal arithmetic functions from

  4055     png.c. This is a libpng maintainer program used to validate changes to the

  4056     internal arithmetic functions.

  4057   Made read 'inflate' handling like write 'deflate' handling. The read

  4058     code now claims and releases png_ptr->zstream, like the write code.

  4059     The bug whereby the progressive reader failed to release the zstream

  4060     is now fixed, all initialization is delayed, and the code checks for

  4061     changed parameters on deflate rather than always calling

  4062     deflatedEnd/deflateInit.

  4063   Validate the zTXt strings in pngvalid.

  4064   Added code to validate the windowBits value passed to deflateInit2().

  4065     If the call to deflateInit2() is wrong a png_warning will be issued

  4066     (in fact this is harmless, but the PNG data produced may be sub-optimal).

  4067 

  4068 Version 1.6.0beta17 [March 10, 2012]

  4069   Fixed PNG_LIBPNG_BUILD_BASE_TYPE definition. 

  4070   Reject all iCCP chunks after the first, even if the first one is invalid.

  4071   Deflate/inflate was reworked to move common zlib calls into single

  4072     functions [rw]util.c.  A new shared keyword check routine was also added

  4073     and the 'zbuf' is no longer allocated on progressive read.  It is now

  4074     possible to call png_inflate() incrementally.  A warning is no longer

  4075     issued if the language tag or translated keyword in the iTXt chunk

  4076     has zero length.

  4077   If benign errors are disabled use maximum window on ancilliary inflate.

  4078     This works round a bug introduced in 1.5.4 where compressed ancillary

  4079     chunks could end up with a too-small windowBits value in the deflate

  4080     header.

  4081 

  4082 Version 1.6.0beta18 [March 16, 2012]

  4083   Issue a png_benign_error() instead of png_warning() about bad palette index.

  4084   In pngtest, treat benign errors as errors if "-strict" is present.

  4085   Fixed an off-by-one error in the palette index checking function.

  4086   Fixed a compiler warning under Cygwin (Windows-7, 32-bit system)

  4087   Revised example.c to put text strings in a temporary character array

  4088     instead of directly assigning string constants to png_textp members.

  4089     This avoids compiler warnings when -Wwrite-strings is enabled.

  4090   Added output flushing to aid debugging under Visual Studio. Unfortunately

  4091     this is necessary because the VS2010 output window otherwise simply loses

  4092     the error messages on error (they weren't flushed to the window before

  4093     the process exited, apparently!)

  4094   Added configuration support for benign errors and changed the read

  4095     default. Also changed some warnings in the iCCP and sRGB handling

  4096     from to benign errors. Configuration now makes read benign

  4097     errors warnings and write benign errors to errors by default (thus

  4098     changing the behavior on read).  The simplified API always forces

  4099     read benign errors to warnings (regardless of the system default, unless

  4100     this is disabled in which case the simplified API can't be built.)

  4101 

  4102 Version 1.6.0beta19 [March 18, 2012]

  4103   Work around for duplicate row start calls; added warning messages.

  4104     This turns on PNG_FLAG_DETECT_UNINITIALIZED to detect app code that

  4105     fails to call one of the 'start' routines (not enabled in libpng-1.5

  4106     because it is technically an API change, since it did normally work

  4107     before.)  It also makes duplicate calls to png_read_start_row (an

  4108     internal function called at the start of the image read) benign, as

  4109     they were before changes to use png_inflate_claim. Somehow webkit is

  4110     causing this to happen; this is probably a mis-feature in the zlib

  4111     changes so this commit is only a work-round.

  4112   Removed erroneous setting of DETECT_UNINITIALIZED and added more

  4113     checks. The code now does a png_error if an attempt is made to do the

  4114     row initialization twice; this is an application error and it has

  4115     serious consequences because the transform data in png_struct is

  4116     changed by each call.

  4117   Added application error reporting and added chunk names to read

  4118     benign errors; also added --strict to pngstest - not enabled

  4119     yet because a warning is produced.

  4120   Avoid the double gamma correction warning in the simplified API.

  4121     This allows the --strict option to pass in the pngstest checks

  4122 

  4123 Version 1.6.0beta20 [March 29, 2012]

  4124   Changed chunk handler warnings into benign errors, incrementally load iCCP

  4125   Added checksum-icc.c to contrib/tools

  4126   Prevent PNG_EXPAND+PNG_SHIFT doing the shift twice.

  4127   Recognize known sRGB ICC profiles while reading; prefer writing the

  4128     iCCP profile over writing the sRGB chunk, controlled by the

  4129     PNG_sRGB_PROFILE_CHECKS option.

  4130   Revised png_set_text_2() to avoid potential memory corruption (fixes

  4131     CVE-2011-3048, also known as CVE-2012-3425).

  4132 

  4133 Version 1.6.0beta21 [April 27, 2012]

  4134   Revised scripts/makefile.darwin: use system zlib; remove quotes around

  4135     architecture list; add missing ppc architecture; add architecture options

  4136     to shared library link; don't try to create a shared lib based on missing

  4137     RELEASE variable.

  4138   Enable png_set_check_for_invalid_index() for both read and write.

  4139   Removed #ifdef PNG_HANDLE_AS_UNKNOWN_SUPPORTED in pngpriv.h around

  4140     declaration of png_handle_unknown().

  4141   Added -lssp_nonshared in a comment in scripts/makefile.freebsd

  4142     and changed deprecated NOOBJ and NOPROFILE to NO_OBJ and NO_PROFILE.

  4143 

  4144 Version 1.6.0beta22 [May 23, 2012]

  4145   Removed need for -Wno-cast-align with clang.  clang correctly warns on

  4146     alignment increasing pointer casts when -Wcast-align is passed. This

  4147     fixes the cases that clang warns about either by eliminating the

  4148     casts from png_bytep to png_uint_16p (pngread.c), or, for pngrutil.c

  4149     where the cast is previously verified or pngstest.c where it is OK, by

  4150     introducing new png_aligncast macros to do the cast in a way that clang

  4151     accepts.

  4152 

  4153 Version 1.6.0beta23 [June 6, 2012]

  4154   Revised CMakeLists.txt to not attempt to make a symlink under mingw.

  4155   Made fixes for new optimization warnings from gcc 4.7.0. The compiler

  4156     performs an optimization which is safe; however it then warns about it.

  4157     Changing the type of 'palette_number' in pngvalid.c removes the warning.

  4158   Do not depend upon a GCC feature macro being available for use in generating

  4159     the linker mapfile symbol prefix.

  4160   Improved performance of new do_check_palette_indexes() function (only

  4161     update the value when it actually increases, move test for whether

  4162     the check is wanted out of the function.

  4163 

  4164 Version 1.6.0beta24 [June 7, 2012]

  4165   Don't check palette indexes if num_palette is 0 (as it can be in MNG files).

  4166 

  4167 Version 1.6.0beta25 [June 16, 2012]

  4168   Revised png_set_keep_unknown_chunks() so num_chunks < 0 means ignore all

  4169     unknown chunks and all known chunks except for IHDR, PLTE, tRNS, IDAT,

  4170     and IEND.  Previously it only meant ignore all unknown chunks, the

  4171     same as num_chunks == 0. Revised png_image_skip_unused_chunks() to

  4172     provide a list of chunks to be processed instead of a list of chunks to

  4173     ignore.  Revised contrib/gregbook/readpng2.c accordingly.

  4174 

  4175 Version 1.6.0beta26 [July 10, 2012]

  4176   Removed scripts/makefile.cegcc from the *.zip and *.7z distributions; it

  4177     depends on configure, which is not included in those archives.

  4178   Moved scripts/chkfmt to contrib/tools.

  4179   Changed "a+w" to "u+w" in Makefile.in to fix CVE-2012-3386.

  4180 

  4181 Version 1.6.0beta27 [August 11, 2012]

  4182   Do not compile PNG_DEPRECATED, PNG_ALLOC and PNG_PRIVATE when __GNUC__ < 3.

  4183   Do not use __restrict when GNUC is <= 3.1

  4184   Removed references to png_zalloc() and png_zfree() from the manual.

  4185   Fixed configurations where floating point is completely disabled.  Because

  4186     of the changes to support symbol prefixing PNG_INTERNAL_FUNCTION declares

  4187     floating point APIs during libpng builds even if they are completely

  4188     disabled. This requires the png floating point types (png_double*) to be

  4189     declared even though the functions are never actually defined.  This

  4190     change provides a dummy definition so that the declarations work, yet any

  4191     implementation will fail to compile because of an incomplete type.

  4192   Re-eliminated the use of strcpy() in pngtest.c.  An unncessary use of

  4193     strcpy() was accidentally re-introduced in libpng16; this change replaces

  4194     it with strncpy().

  4195   Eliminated use of png_sizeof(); use sizeof() instead.

  4196   Use a consistent style for (sizeof type) and (sizeof (array))

  4197   Cleanup of png_set_filler().  This function does very different things on

  4198     read and write.  In libpng 1.6 the two cases can be distinguished and

  4199     considerable code cleanup, and extra error checking, is possible.  This

  4200     makes calls on the write side that have no effect be ignored with a

  4201     png_app_error(), which can be disabled in the app using

  4202     png_set_benign_errors(), and removes the spurious use of usr_channels

  4203     on the read side.

  4204   Insist on autotools 1.12.1 for git builds because there are security issues

  4205     with 1.12 and insisting on anything less would allow 1.12 to be used.

  4206   Removed info_ptr->signature[8] from WRITE-only builds.

  4207   Add some conditions for compiling png_fixed().  This is a small function

  4208     but it requires "-lm" on some platforms.

  4209   Cause pngtest --strict to fail on any warning from libpng (not just errors)

  4210     and cause it not to fail at the comparison step if libpng lacks support

  4211     for writing chunks that it reads from the input (currently only implemented

  4212     for compressed text chunks).

  4213   Make all three "make check" test programs work without READ or WRITE support.

  4214     Now "make check" will succeed even if libpng is compiled with -DPNG_NO_READ

  4215     or -DPNG_NO_WRITE.  The tests performed are reduced, but the basic reading

  4216     and writing of a PNG file is always tested by one or more of the tests.

  4217   Consistently use strlen(), memset(), memcpy(), and memcmp() instead of the

  4218     png_strlen(), png_memset(), png_memcpy(), and png_memcmp() macros.

  4219   Removed the png_sizeof(), png_strlen(), png_memset(), png_memcpy(), and

  4220     png_memcmp() macros.

  4221   Work around gcc 3.x and Microsoft Visual Studio 2010 complaints. Both object

  4222     to the split initialization of num_chunks.

  4223 

  4224 Version 1.6.0beta28 [August 29, 2012]

  4225   Unknown handling fixes and clean up. This adds more correct option

  4226     control of the unknown handling, corrects the pre-existing bug where

  4227     the per-chunk 'keep' setting is ignored and makes it possible to skip

  4228     IDAT chunks in the sequential reader (broken in earlier 1.6 versions).

  4229     There is a new test program, test-unknown.c, which is a work in progress

  4230     (not currently part of the test suite).  Comments in the header files now

  4231     explain how the unknown handling works.

  4232   Allow fine grain control of unknown chunk APIs. This change allows

  4233     png_set_keep_unknown_chunks() to be turned off if not required and causes

  4234     both read and write to behave appropriately (on read this is only possible

  4235     if the user callback is used to handle unknown chunks).  The change

  4236     also removes the support for storing unknown chunks in the info_struct

  4237     if the only unknown handling enabled is via the callback, allowing libpng

  4238     to be configured with callback reading and none of the unnecessary code.

  4239   Corrected fix for unknown handling in pngtest. This reinstates the

  4240     libpng handling of unknown chunks other than vpAg and sTER (including

  4241     unsafe-to-copy chunks which were dropped before) and eliminates the

  4242     repositioning of vpAg and sTER in pngtest.png by changing pngtest.png

  4243     (so the chunks are where libpng would put them).

  4244   Added "tunknown" test and corrected a logic error in png_handle_unknown()

  4245     when SAVE support is absent.  Moved the shell test scripts for

  4246     contrib/libtests from the libpng top directory to contrib/libtests.

  4247     png_handle_unknown() must always read or skip the chunk, if

  4248     SAVE_UNKNOWN_CHUNKS is turned off *and* the application does not set

  4249     a user callback an unknown chunk will not be read, leading to a read

  4250     error, which was revealed by the "tunknown" test.

  4251   Cleaned up and corrected ICC profile handling.

  4252     contrib/libtests/makepng: corrected 'rgb' and 'gray' cases.  profile_error

  4253     messages could be truncated; made a correct buffer size calculation and

  4254     adjusted pngerror.c appropriately. png_icc_check_* checking improved;

  4255     changed the functions to receive the correct color type of the PNG on read

  4256     or write and check that it matches the color space of the profile (despite

  4257     what the comments said before, there is danger in assuming the app will

  4258     cope correctly with an RGB profile on a grayscale image and, since it

  4259     violates the PNG spec, allowing it is certain to produce inconsistent

  4260     app behavior and might even cause app crashes.) Check that profiles

  4261     contain the tags needed to process the PNG (tags all required by the ICC

  4262     spec). Removed unused PNG_STATIC from pngpriv.h.

  4263 

  4264 Version 1.6.0beta29 [September 4, 2012]

  4265   Fixed the simplified API example programs to add the *colormap parameter

  4266     to several of he API and improved the error message if the version field

  4267     is not set.

  4268   Added contrib/examples/* to the *.zip and *.7z distributions.

  4269   Updated simplified API synopses and description of the png_image structure

  4270     in the manual.

  4271   Made makepng and pngtest produce identical PNGs, add "--relaxed" option

  4272     to pngtest. The "--relaxed" option turns off the benign errors that are

  4273     enabled by default in pre-RC builds. makepng can now write ICC profiles

  4274     where the length has not been extended to a multiple of 4, and pngtest

  4275     now intercepts all libpng errors, allowing the previously-introduced

  4276     "--strict test" on no warnings to actually work.

  4277   Improved ICC profile handling including cHRM chunk generation and fixed

  4278     Cygwin+MSVC build errors. The ICC profile handling now includes more

  4279     checking.  Several errors that caused rejection of the profile are now

  4280     handled with a warning in such a way that the invalid profiles will be

  4281     read by default in release (but not pre-RC) builds but will not be

  4282     written by default.  The easy part of handling the cHRM chunk is written,

  4283     where the ICC profile contains the required data.  The more difficult

  4284     part plus guessing a gAMA value requires code to pass selected RGB values

  4285     through the profile.

  4286 

  4287 Version 1.6.0beta30 [October 24, 2012]

  4288   Changed ICC profile matrix/vector types to not depend on array type rules.

  4289     By the ANSI-C standard the new types should be identical to the previous

  4290     versions, and all known versions of gcc tested with the previous versions

  4291     except for GCC-4.2.1 work with this version.  The change makes the ANSI-C

  4292     rule that const applied to an array of elements applies instead to the

  4293     elements in the array moot by explicitly applying const to the base

  4294     elements of the png_icc_matrix and png_icc_vector types. The accidental

  4295     (harmless) 'const' previously applied to the parameters of two of the

  4296     functions have also been removed.

  4297   Added a work around for GCC 4.2 optimization bug.

  4298   Marked the broken (bad white point) original HP sRGB profiles correctly and

  4299     correct comments.

  4300   Added -DZ_SOLO to contrib/pngminim/*/makefile to work with zlib-1.2.7

  4301   Use /MDd for vstudio debug builds. Also added pngunkown to the vstudio

  4302     builds, fixed build errors and corrected a minor exit code error in

  4303     pngvalid if the 'touch' file name is invalid.

  4304   Add updated WARNING file to projects/vstudio from libpng 1.5/vstudio

  4305   Fixed build when using #define PNG_NO_READ_GAMMA in png_do_compose() in

  4306     pngrtran.c (Domani Hannes).

  4307 

  4308 Version 1.6.0beta31 [November 1, 2012]

  4309   Undid the erroneous change to vstudio/pngvalid build in libpng-1.6.0beta30.

  4310   Made pngvalid so that it will build outside the libpng source tree.

  4311   Made builds -DPNG_NO_READ_GAMMA compile (the unit tests still fail).

  4312   Made PNG_NO_READ_GAMMA switch off interfaces that depend on READ_GAMMA.

  4313     Prior to 1.6.0 switching off READ_GAMMA did unpredictable things to the

  4314     interfaces that use it (specifically, png_do_background in 1.4 would

  4315     simply display composite for grayscale images but do composition

  4316     with the incorrect arithmetic for color ones). In 1.6 the semantic

  4317     of -DPNG_NO_READ_GAMMA is changed to simply disable any interface that

  4318     depends on it; this obliges people who set it to consider whether they

  4319     really want it off if they happen to use any of the interfaces in

  4320     question (typically most users who disable it won't).

  4321   Fixed GUIDs in projects/vstudio. Some were duplicated or missing,

  4322     resulting in VS2010 having to update the files.

  4323   Removed non-working ICC profile support code that was mostly added to

  4324     libpng-1.6.0beta29 and beta30. There was too much code for too little

  4325     gain; implementing full ICC color correction may be desireable but is left

  4326     up to applications.

  4327 

  4328 Version 1.6.0beta32 [November 25, 2012]

  4329   Fixed an intermittent SEGV in pngstest due to an uninitialized array element.

  4330   Added the ability for contrib/libtests/makepng.c to make a PNG with just one

  4331     color. This is useful for debugging pngstest color inaccuracy reports.

  4332   Fixed error checking in the simplified write API (Olaf van der Spek)

  4333   Made png_user_version_check() ok to use with libpng version 1.10.x and later.

  4334 

  4335 Version 1.6.0beta33 [December 15, 2012]

  4336   Fixed typo in png.c (PNG_SET_CHUNK_MALLOC_MAX should be PNG_CHUNK_MALLOC_MAX)

  4337     that causes the MALLOC_MAX limit not to work (John Bowler)

  4338   Change png_warning() to png_app_error() in pngwrite.c and comment the

  4339     fall-through condition.

  4340   Change png_warning() to png_app_warning() in png_write_tRNS().

  4341   Rearranged the ARM-NEON optimizations: Isolated the machine specific code

  4342     to the hardware subdirectory and added comments to pngrutil.c so that

  4343     implementors of other optimizations know what to do.

  4344   Fixed cases of unquoted DESTDIR in Makefile.am

  4345   Rebuilt Makefile.in, etc., with autoconf-2.69 and automake-1.12.5.

  4346 

  4347 Version 1.6.0beta34 [December 19, 2012]

  4348   Cleaned up whitespace in the synopsis portion of the manpage "libpng.3"

  4349   Disassembled the version number in scripts/options.awk (necessary for

  4350     building on SunOs).

  4351 

  4352 Version 1.6.0beta35 [December 23, 2012]

  4353   Made default Zlib compression settings be configurable. This adds #defines to

  4354     pnglibconf.h to control the defaults.

  4355   Fixed Windows build issues, enabled ARM compilation. Various warnings issued

  4356     by earlier versions of GCC fixed for Cygwin and Min/GW (which both use old

  4357     GCCs.) ARM support is enabled by default in zlib.props (unsupported by

  4358     Microsoft) and ARM compilation is made possible by deleting the check for

  4359     x86. The test programs cannot be run because they are not signed.

  4360 

  4361 Version 1.6.0beta36 [January 2, 2013]

  4362   Discontinued distributing libpng-1.x.x.tar.bz2.

  4363   Discontinued distributing libpng-1.7.0-1.6.0-diff.txt and similar.

  4364   Rebuilt configure with autoconf-2.69 (inadvertently not done in beta33)

  4365   Fixed 'make distcheck' on SUN OS - libpng.so was not being removed

  4366 

  4367 Version 1.6.0beta37 [January 10, 2013]

  4368   Fixed conceivable but difficult to repro overflow. Also added two test

  4369     programs to generate and test a PNG which should have the problem.

  4370 

  4371 Version 1.6.0beta39 [January 19, 2013]

  4372   Again corrected attempt at overflow detection in png_set_unknown_chunks()

  4373   (CVE-2013-7353).  Added overflow detection in png_set_sPLT() and

  4374   png_set_text_2() (CVE-2013-7354).

  4375 

  4376 Version 1.6.0beta40 [January 20, 2013]

  4377   Use consistent handling of overflows in text, sPLT and unknown png_set_* APIs

  4378 

  4379 Version 1.6.0rc01 [January 26, 2013]

  4380   No changes.

  4381 

  4382 Version 1.6.0rc02 [February 4, 2013]

  4383   Added png_get_palette_max() function.

  4384 

  4385 Version 1.6.0rc03 [February 5, 2013]

  4386   Fixed the png_get_palette_max API.

  4387 

  4388 Version 1.6.0rc04 [February 7, 2013]

  4389   Turn serial tests back on (recently turned off by autotools upgrade).

  4390 

  4391 Version 1.6.0rc05 [February 8, 2013]

  4392   Update manual about png_get_palette_max().

  4393 

  4394 Version 1.6.0rc06 [February 9, 2013]

  4395   Fixed missing dependency in --prefix builds The intermediate

  4396     internal 'prefix.h' file can only be generated correctly after

  4397     pnglibconf.h, however the dependency was not in Makefile.am.  The

  4398     symptoms are unpredictable depending on the order make chooses to

  4399     build pngprefix.h and pnglibconf.h, often the error goes unnoticed

  4400     because there is a system pnglibconf.h to use instead.

  4401 

  4402 Version 1.6.0rc07 [February 10, 2013]

  4403   Enclosed the new png_get_palette_max in #ifdef PNG_GET_PALETTE_MAX_SUPPORTED

  4404     block, and revised pnglibconf.h and pnglibconf.h.prebuilt accordingly.

  4405 

  4406 Version 1.6.0rc08 [February 10, 2013]

  4407   Fix typo in png.h #ifdef

  4408 

  4409 Version 1.6.0 [February 14, 2013]

  4410   No changes.

  4411 

  4412 Version 1.6.1beta01 [February 16, 2013]

  4413   Made symbol prefixing work with the ARM neon optimizations. Also allow

  4414     pngpriv.h to be included for preprocessor definitions only, so it can

  4415     be used in non-C/C++ files. Back ported from libpng 1.7.

  4416   Made sRGB check numbers consistent.

  4417   Ported libpng 1.5 options.awk/dfn file handling to 1.6, fixed one bug.

  4418   Removed cc -E workround, corrected png_get_palette_max API Tested on

  4419     SUN OS cc 5.9, which demonstrates the tokenization problem previously

  4420     avoided by using /lib/cpp.  Since all .dfn output is now protected in

  4421     double quotes unless it is to be macro substituted the fix should

  4422     work everywhere.

  4423   Enabled parallel tests - back ported from libpng-1.7.

  4424   scripts/pnglibconf.dfa formatting improvements back ported from libpng17.

  4425   Fixed a race condition in the creation of the build 'scripts' directory

  4426     while building with a parallel make.

  4427   Use approved/supported Android method to check for NEON, use Linux/POSIX

  4428     1003.1 API to check /proc/self/auxv avoiding buffer allocation and other

  4429     library calls (ported from libpng15).

  4430 

  4431 Version 1.6.1beta02 [February 19, 2013]

  4432   Use parentheses more consistently in "#if defined(MACRO)" tests.

  4433   Folded long lines.

  4434   Reenabled code to allow zero length PLTE chunks for MNG.

  4435 

  4436 Version 1.6.1beta03 [February 22, 2013]

  4437   Fixed ALIGNED_MEMORY support.

  4438   Allow run-time ARM NEON checking to be disabled. A new configure option:

  4439     --enable-arm-neon=always will stop the run-time checks. New checks

  4440     within arm/arm_init.c will cause the code not to be compiled unless

  4441     __ARM_NEON__ is set. This should make it fail safe (if someone asks

  4442     for it on then the build will fail if it can't be done.)

  4443   Updated the INSTALL document.

  4444 

  4445 Version 1.6.1beta04 [February 27, 2013]

  4446   Revised INSTALL to recommend using CPPFLAGS instead of INCLUDES.

  4447   Revised scripts/makefile.freebsd to respect ZLIBLIB and ZLIBINC.

  4448   Revised scripts/dfn.awk to work with the buggy MSYS awk that has trouble

  4449     with CRLF line endings.

  4450 

  4451 Version 1.6.1beta05 [March 1, 2013]

  4452   Avoid a possible memory leak in contrib/gregbook/readpng.c

  4453 

  4454 Version 1.6.1beta06 [March 4, 2013]

  4455   Better documentation of unknown handling API interactions.

  4456   Corrected Android builds and corrected libpng.vers with symbol

  4457     prefixing. This adds an API to set optimization options externally,

  4458     providing an alternative and general solution for the non-portable

  4459     run-time tests used by the ARM Neon code.  It also makes those tests

  4460     compile and link on Android.

  4461   The order of settings vs options in pnglibconf.h is reversed to allow

  4462     settings to depend on options and options can now set (or override) the

  4463     defaults for settings.

  4464 

  4465 Version 1.6.1beta07 [March 7, 2013]

  4466   Corrected simplified API default gamma for color-mapped output, added

  4467     a flag to change default. In 1.6.0 when the simplified API was used

  4468     to produce color-mapped output from an input image with no gamma

  4469     information the gamma assumed for the input could be different from

  4470     that assumed for non-color-mapped output.  In particular 16-bit depth

  4471     input files were assumed to be sRGB encoded, whereas in the 'direct'

  4472     case they were assumed to have linear data.  This was an error.  The

  4473     fix makes the simplified API treat all input files the same way and

  4474     adds a new flag to the png_image::flags member to allow the

  4475     application/user to specify that 16-bit files contain sRGB data

  4476     rather than the default linear.

  4477   Fixed bugs in the pngpixel and makepng test programs.

  4478 

  4479 Version 1.6.1beta08 [March 7, 2013]

  4480   Fixed CMakelists.txt to allow building a single variant of the library

  4481     (Claudio Bley):

  4482   Introduced a PNG_LIB_TARGETS variable that lists all activated library

  4483     targets.  It is an error if this variable ends up empty, ie. you have

  4484     to build at least one library variant.

  4485   Made the *_COPY targets only depend on library targets actually being build.

  4486   Use PNG_LIB_TARGETS to unify a code path.

  4487   Changed the CREATE_SYMLINK macro to expect the full path to a file as the

  4488     first argument. When symlinking the filename component of that path is

  4489     determined and used as the link target.

  4490   Use copy_if_different in the CREATE_SYMLINK macro.

  4491 

  4492 Version 1.6.1beta09 [March 13, 2013]

  4493   Eliminated two warnings from the Intel C compiler. The warnings are

  4494     technically valid, although a reasonable treatment of division would

  4495     show it to be incorrect.

  4496 

  4497 Version 1.6.1rc01 [March 21, 2013]

  4498   No changes.

  4499 

  4500 Version 1.6.1 [March 28, 2013]

  4501   No changes.

  4502 

  4503 Version 1.6.2beta01 [April 14, 2013]

  4504   Updated documentation of 1.5.x to 1.6.x changes in iCCP chunk handling.

  4505   Fixed incorrect warning of excess deflate data. End condition - the

  4506     warning would be produced if the end of the deflate stream wasn't read

  4507     in the last row.  The warning is harmless.