This change fixes a crash that occurs in `Load_SBit_Png` when
running on a 64-bit Windows OS. A memory access violation exception
would be raised by `setjmp` if the `jmp_buf` is not aligned to a
16-byte memory boundary. This is due to setjmp executing `movdqa`
instructions to store 128-bit XMM registers to memory, which require
correct memory alignment. This problem occurs because
`png_create_read_struct` uses `malloc` and `free` for memory
management, which only guarantees 8-byte alignment on Windows.
Instead, to fix the problem, `png_create_read_struct_2` is used on
64-bit Windows, which allows for user-defined memory allocation and
deallocation callbacks to be specified. These callbacks forward the
allocation and deallocation requests to `_aligned_alloc` and
`_aligned_free`, ensuring that the allocated `png_struct` and
internal `jmp_buf` have the requisite 16-byte alignment.
* src/sfnt/pngshim.c <_WIN64>: Include `malloc.h`.
(malloc_callback, free_callback) <_WIN64>: New functions.
(Load_SBit_Png) <_WIN64>: Use `png_create_read_struct_2` instead of
`png_create_read_struct`
We no longer have to take care of the 8.3 file name limit; this
allows us (a) to introduce longer, meaningful file names, and (b) to
avoid macro names in `#include' lines altogether since some
compilers (most notably Visual C++) doesn't support this properly.
*/*: Replace
#include FOO_H
with
#include <freetype/foo.h>
or something similar. Also update the documentation.
This monster commit was created by applying Nikhil's scripts
`docconverter.py' and `markify.py' to all C header and source files,
followed up by minor manual clean-up.
No change in functionality, of course.
I used commit f7419907bc6044b9b7057f9789866426c804ba82 from
https://github.com/nikramakrishnan/freetype-docs.git.
__builtin_shuffle() was introduced in gcc-4.7. The lowest
gcc to enable vector operation is delayed from 4.6 to 4.7.
* src/sfnt/pngshim.c (premultiply_data): Fix cpp-macro to
enable the vector operation, to change the lowest gcc version
from 4.6 to 4.7.
* src/sfnt/pngshim.c (premultiply_data): Use vectors instead of
scalars.
(vector_shuffle): New macro to take of a different built-in function
name on clang.
This reduces the overhead of `premultiply_data' by 60%.
* src/sfnt/pngshim.c (premultiply_data): Provide code which uses
gcc's (and clang's) `vector_byte' attribute to process 4 pixels at a
time.
It's unavoidable to call the PNG engine, but to get the metrics it
is sufficient to read the PNG image's header only.
* src/sfnt/pngshim.c (Load_SBit_Png): Add argument to control the
allocation of the glyph slot.
* src/sfnt/pngshim.h: Updated.
* src/sfnt/ttsbit.c (tt_sbit_decoder_load_png,
tt_face_load_sbix_image, tt_face_load_sbit_image): Updated.
A new load flag, FT_LOAD_COLOR, makes FreeType load color
embedded-bitmaps, following this draft specification
https://color-emoji.googlecode.com/git/specification/v1.html
which defines two new SFNT tables, `CBDT' and `CBLC' (named and
modeled after `EBDT' and `EBLC', respectively). The color bitmaps
are stored in the new FT_PIXEL_MODE_BGRA format to represent BGRA
pre-multiplied sRGB images. If PNG support is available, PNG color
images as defined in the same proposed specification are supported
also.
Note that color bitmaps are converted to grayscale if client didn't
ask for color.
* builds/unix/configure.raw: Search for libpng.
Add `--without-png' option.
* devel/ftoption.h, include/freetype/config/ftoption.h
(FT_CONFIG_OPTION_USE_PNG): New macro.
* include/freetype/freetype.h (FT_LOAD_COLOR): New load flag.
* include/freetype/ftimage.h (FT_Pixel_Mode): Add
`FT_PIXEL_MODE_BGRA'.
* include/freetype/tttags.h (TTAG_CBDT, TTAG_CBLC): New tags.
* src/base/ftbitmap.c (FT_Bitmap_Embolden): Updated.
(ft_gray_for_premultiplied_srgb_bgra): New function.
(FT_Bitmap_Convert): Handle FT_PIXEL_MODE_BGRA.
* src/sfnt/pngshim.c, src/sfnt/pngshim.h: New files.
* src/sfnt/sfnt.c: Include `pngshim.c'.
* src/sfnt/ttsbit.c: Include FT_BITMAP_H and `pngshim.h'
(tt_face_load_eblc): Load `CBLC'.
(tt_sbit_decoder_init): Load `CBDT'.
(tt_sbit_decoder_alloc_bitmap): Pass load flags to select between
color and grayscale bitmaps.
Set `num_grays'. This is used by `ftview' to choose the blending
algorithm.
(tt_sbit_decoder_load_byte_aligned,
tt_sbit_decoder_load_bit_aligned, tt_sbit_decoder_load_compound,
tt_sbit_decoder_load_image): Pass load flag.
s/write/pwrite/.
Don't call `tt_sbit_decoder_alloc_bitmap'.
Updated.
(tt_sbit_decoder_load_png) [FT_CONFIG_OPTION_USE_PNG]: New function.
(tt_sbit_decoder_load_bitmap): Pass load flag.
Handle new glyph formats 17, 18, and 19.
Call `tt_sbit_decoder_alloc_bitmap'.
Flatten color bitmaps if necessary.
(tt_face_load_sbit_image): Updated.
* src/sfnt/rules.mk (SFNT_DRV_SRC): Add `pngshim.c'.
* docs/CHANGES: Updated.
Jesse Towner