Compare commits

..

44 Commits

Author SHA1 Message Date
326b5fb863 update ChangeLog
Change-Id: I794a5d58005bb0934ce0de06483567e8ed6bd8eb
2015-03-10 23:06:09 -07:00
a661e50bcb Disable NEON code on Native Client
The NEON assember in libwebp has not yet been ported
to Native Client. This changes disables it.
Related issue:
https://code.google.com/p/nativeclient/issues/detail?id=3205

(cherry picked from commit ac4f5784a0)

Change-Id: I200291db7aa79d40c1f10cff7622c9b8599e6886
2015-03-10 20:29:08 -07:00
fcd94e925a update ChangeLog
Change-Id: I60c273c650a305fe36564ccc5fb1c8d7ea18118f
2015-03-04 11:30:23 -08:00
569fe5789e update NEWS
Change-Id: Iade773fc3b9961fcd9b4112a3972cfc68e3670f2
2015-03-03 19:19:50 -08:00
bd852f5d81 bump version to 0.4.3
libwebp{,decoder} - 0.4.3
libwebp libtool - 5.3.0
libwebpdecoder libtool - 1.3.0

mux/demux - 0.2.2 (unchanged)
libtool - 1.2.0 (unchanged)

Change-Id: Ie8c35ffc20c1bfd782bdafd99da6c6b1373022c1
2015-03-03 19:05:40 -08:00
2d58b64f51 WebPPictureRescale: add a note about 0 width/height
(cherry picked from commit 0f773693bf)

Change-Id: I3890bb3fd32a148d7dd24c714546160c6c59d4ea
2015-03-03 17:53:49 -08:00
a0d8ca576f examples/Android.mk: add webpmux_example target
renamed from 'webpmux' to avoid name clash with the library module name

(cherry picked from commit 6cef0e4fa4)

Change-Id: I33bbdbdcb25a6f35bd85c9a0dbbb93b2428b05f3
2015-03-03 17:53:49 -08:00
34b1d29e3c Android.mk: add webpmux target
(cherry picked from commit 53c16ff047)

Change-Id: I60fc898fd804e23f08d760694192c5d04adcae91
2015-03-03 17:53:49 -08:00
75619881e6 Android.mk: add webpdemux target
(cherry picked from commit 21852a00a1)

Change-Id: I2fbbefbee59a96c52f5addcfc5bfe1216caad5cc
2015-03-03 17:53:48 -08:00
a98757650a Android.mk: add webpdecoder{,_static} targets
webpdecoder_static is reused to create libwebpdecoder.so and
libwebp.{a,so}

(cherry picked from commit 8697a3bcc8)

Change-Id: I940293cb755040c0ea45dc13f22624de8f355867
2015-03-03 17:53:48 -08:00
a6d4859725 Android.mk: split source lists per-directory
will allow reuse in future targets

(cherry picked from commit 4a67049113)

Conflicts:
	Android.mk

Change-Id: Iededc19d954226e62f2d2383a2b80f268d613647
2015-03-03 17:53:48 -08:00
77544d5f5b fix iOS arm64 build with Xcode 6.3
the standard vtbl functions are available there [1][2].
based on a patch from: aaroncrespo
fixes issue #243.

[1]
http://adcdownload.apple.com//Developer_Tools/Xcode_6.3_beta/Xcode_6.3_beta_Release_Notes.pdf
[2] Apple LLVM Compiler Version 6.1
- Xcode 6.3 updates the Apple LLVM compiler to version 6.1.0.
[...]
Support for the arm64 architecture has been significantly revised to
align with ARM's implementation, where the most visible impact is that a
few of the vector intrinsics have changed to match ARM's specifications.

(cherry picked from commit 602a00f93f)

Change-Id: I79a0016f44b9dbe36d0373f7f00a50ab3c2ca447
2015-03-03 17:53:47 -08:00
6dea15784d doc/webp-container-spec: note MSB order for chunk diagrams
addresses question in issue #241

(cherry picked from commit b510fbfe3b)

Change-Id: Iff6a172d5822f6ec8b9bc0951a1c9cd3f98c9251
2015-03-03 17:53:47 -08:00
f7cd57b23d doc/webp-container-spec: cosmetics
partially normalize indent, vertical whitespace and capitalization with
the copy used on developers.google.com/speed/webp

(cherry picked from commit e7d3df2314)

Change-Id: I8044418eeb9eaf5bd5c799675c74f6f845d503d6
2015-03-03 17:53:47 -08:00
1d6b250b07 vwebp: clear canvas at the beginning of each loop
this is in line with the recommendation in the spec, cf.,
5603947 webp-container-spec: clarify background clear on loop

(cherry picked from commit 1579de3cae)

Change-Id: Id3910395b05a1a1f2804be841b61f97bd4bac593
2015-03-03 17:53:46 -08:00
f97b3f86bf webp-container-spec: clarify background clear on loop
at the beginning of the loop there's an implicit clear of the entire
canvas to the background (or application defined) color. this avoids
adding the final composited frame to the first.

(cherry picked from commit 560394798f)

Change-Id: Ia3a52cf4482c6176334a5c9c99a0ddd07d1776e7
2015-03-03 17:53:46 -08:00
4ba83c1759 vwebp: remove unnecessary static Help() prototype
all uses occur after its declaration

(cherry picked from commit 2c906c407c)

Change-Id: I775642ce6d1dec3bc6da2fa0d5d87490992c7e6c
2015-03-03 17:53:46 -08:00
d34e8e3d18 vwebp/animation: display last frame on end-of-loop
previously the first frame would be redisplayed, which might be
unexpected if the final frame was meant to be a composite, for example.

(cherry picked from commit 0f017b56f3)

Change-Id: I4da795623c71501e2fa426e8fba8fb2ffcbab58a
2015-03-03 17:53:45 -08:00
bbbc524fb4 dec/vp8: clear 'dither_' on skipped blocks
DitherRow() only checks this value, not 'skip_' so previously it was
uninitialized for these blocks.

(cherry picked from commit 66935fb9ee)

Change-Id: I0f698b81854ee9d91edacb51c1e3bdab9cba96f2
2015-03-03 17:53:45 -08:00
0339fa26eb lossless_neon: enable subtract green for aarch64
similar to:
1ba61b0 enable NEON intrinsics in aarch64 builds

vtbl1_u8 is available everywhere but Xcode-based iOS arm64 builds, use
vtbl1q_u8 there.

performance varies based on the input, 1-3% on encode was observed

(cherry picked from commit 416e1cea9b)

Change-Id: Ifec35b37eb856acfcf69ed7f16fa078cd40b7034
2015-03-03 17:53:45 -08:00
5a0c2207f4 Regression fix for lossless decoding
Reported here: https://code.google.com/p/webp/issues/detail?id=239

At the beginning of method 'DecodeImageData', pixels up to
'dec->last_pixel_' are assumed to be already cached. So, at the end of
previous call to that method also, that assumption should hold true.

Hence, we should cache all pixels up to 'src' regardless of 'src_last'.

This affects lossless incremental decoding only, as that is when
src_last and src_end differ.
Note: alpha decoding is implicitly incremental, as alpha decoding of
only the rows 'y_end - y_start' happens during FinishRow() call. So, this bug
affects alpha decoding in non-incremental decoding flow as well.

This bug was introduced in: https://gerrit.chromium.org/gerrit/#/c/59716.

(cherry picked from commit 783a8cda24)

Change-Id: Ide6edfeb2609b02aff701e1bd9fd776da0a16be0
2015-03-03 17:53:44 -08:00
6e3a31d595 wicdec: (msvs) quiet some /analyze warnings
add additional return checks and asserts to avoid:
C6102: Using 'XXX' from failed function call ...

(cherry picked from commit 9b228b5416)

Change-Id: I51f5fa630324e0cd7b2d9fceefecb4f4021474b1
2015-03-03 17:53:44 -08:00
b49a578135 dwebp/WritePNG: mark png variables volatile
these are used on both sides of the setjmp().

(cherry picked from commit 7a191398ca)

Change-Id: I4a789bfb3a5d56946a22286c5a140008d90e1ba2
2015-03-03 17:53:43 -08:00
0a4391a196 dwebp: include setjmp.h w/WEBP_HAVE_PNG
setjmp() is used in WritePNG().

(cherry picked from commit 775dfad297)

Change-Id: Iadd836272fc7d368d635c891507ce2a08c4d3dec
2015-03-03 17:53:43 -08:00
90f1ec58a9 dwebp: correct sign in format strings
width / height are unsigned; fixes a warning with msvs /analyze:
C6340: Mismatch on sign: 'const unsigned int' passed as _Param_(4) when
some signed type is required in call to 'fprintf'.

(cherry picked from commit 47d26be760)

Change-Id: I5f1fad4c93745baf17d70178a5e66579ccd2b155
2015-03-03 17:53:43 -08:00
b61ce861f3 VP8LEncodeStream: add an assert
check enc->argb_ to quiet an msvs /analyze warning:
C6387: 'enc->argb_+y*width' could be '0':  this does not adhere to the
specification for the function 'memcpy'.

(cherry picked from commit f0e0677b87)

Change-Id: I87544e92ee0d3ea38942a475c30c6d552f9877b7
2015-03-03 17:53:42 -08:00
df1081bb82 dsp/cpu: (msvs) add include for __cpuidex
and only use it on x86 / x64 where it's available.
has the side-effect of quieting a msvs /analyze warning:
C6001: Using uninitialized memory 'cpu_info'.

(cherry picked from commit 0de5f33e31)

Change-Id: Iae51be3b22b2ee949cfc473eeea9fd9fb6b3c2cb
2015-03-03 17:53:42 -08:00
39aa055529 dsp/cpu: (msvs) avoid immintrin.h on _M_ARM
_xgetgv() isn't relevant there anyway

broken since:
279e661 Merge "dsp/cpu: add include for _xgetbv() w/MSVS"

(cherry picked from commit 4fbe9cf202)

Change-Id: Iaa7bc0c5be9c06bfffab39e194c64c09bf5b5a27
2015-03-03 17:53:42 -08:00
f814f429ca dsp/cpu: add include for _xgetbv() w/MSVS
explicitly add immintrin.h instead of transitively picking it up via
windows.h presumably. makes the code easier to move around.

(cherry picked from commit b6c0428e8c)

Change-Id: If70d5143ac94fc331da763ce034358858e460e06
2015-03-03 17:53:41 -08:00
8508ab99a7 cpu: fix AVX2 detection for gcc/clang targets
ecx needs to be set to 0; the visual studio builds were already doing
this.

https://software.intel.com/en-us/articles/how-to-detect-new-instruction-support-in-the-4th-generation-intel-core-processor-family

(cherry picked from commit d7c4b02a57)

Change-Id: I95efb115b4d50bbdb6b14fca2aa63d0a24974e55
2015-03-03 17:53:41 -08:00
5769623b6f fix handling of zero-sized partition #0 corner case
reported in https://code.google.com/p/webp/issues/detail?id=237

An empty partition #0 should be indicative of a bitstream error.
The previous code was correct, only an assert was triggered in debug mode.
But we might as well handle the case properly right away...

(cherry picked from commit 205c7f26af)

Change-Id: I4dc31a46191fa9e65659c9a5bf5de9605e93f2f5
2015-03-03 17:53:40 -08:00
b2e71a9080 make the 'last_cpuinfo_used' variable names unique
allows the sources to be #include'd in some hackish builds (don't do
that!)

(cherry picked from commit 67f601cd46)

Conflicts:
	src/dsp/alpha_processing.c
	src/dsp/argb.c
	src/dsp/dec.c
	src/dsp/enc.c
	src/dsp/lossless.c
	src/dsp/upsampling.c
	src/dsp/yuv.c

Change-Id: I0c7a43acbebd0e2d5068845e6daa8ce47361cd91
2015-03-02 18:43:41 -08:00
1273e84517 add -Wformat-nonliteral and -Wformat-security
can be useful, not sure they are a subset of the flags we
use already...

(cherry picked from commit 80d950d94e)

Change-Id: Iec742a99427a791d9527368302a1136df2ff96cd
2015-03-02 18:43:35 -08:00
3ae78eb757 multi-thread fix: lock each entry points with a static var
we compare the current VP8GetCPUInfo pointer to the last used.
This is less code overall and each implementation is still
testable separately (by just changing VP8GetCPUInfo, but not
a separate threads!)

(cherry picked from commit a437694a17)

Conflicts:
	src/dsp/alpha_processing.c
	src/dsp/argb.c
	src/dsp/dec.c
	src/dsp/enc.c
	src/dsp/lossless.c
	src/dsp/upsampling.c
	src/dsp/yuv.c

Change-Id: Ia13fa8ffc4561a884508f6ab71ed0d1b9f1ce59b
2015-03-02 18:43:31 -08:00
5c1eeda922 webp-container-spec: remove references to fragments
this portion of the format was never finalized

(cherry picked from commit a66e66c79d)

Change-Id: I80aa1b27457a0e52b047c7284df2f58b181ca5d8
2015-03-02 18:43:26 -08:00
c5ceea4899 enc_neon: fix building with non-Xcode clang (iOS)
check for __apple_build_version__ to distinguish the two; a version
check could work as Apple bumped Xcode's to 5.x/6.x, but it's unclear
how upstream will deal with their versioning as they go 3.6+, so avoid
it for now.

(cherry picked from commit a3946b8956)

Change-Id: I67cda67c4f68e262a92d805a63cc1496374be063
2015-03-02 18:43:20 -08:00
d0859d69de iosbuild: add x64_64 simulator support
based on the patch here:
https://github.com/pixelkind/webp-ios-build

(cherry picked from commit a96ccf8fde)

Change-Id: Iaa346b751e5f18e8cf13a8e5c4064b0c2a3f5f6c
2015-03-02 18:43:14 -08:00
046732ca65 WebPEncode: Support encoding same pic twice (even if modified)
This wasn't working for this specific scenario:
- Encode an RGBA 'pic' (with trivial alpha) using lossy encoding.
(so that pic->a == NULL after import happens).
- Modify the 'pic->argb' so that it has non-trivial alpha.
- Encode the same 'pic' again.
This used to fail to encode alpha data as pic->a == NULL.

(cherry picked from commit e4f4dddba3)

Change-Id: Ieaaa7bd09825c42f54fbd99e6781d98f0b19cc0c
2015-03-02 18:43:08 -08:00
4426f50179 webp/types.h: use inline for clang++/-std=c++11
at least clang 3.[45] in c++ mode with -std=c++11 define __STRICT_ANSI__
this change set WEBP_INLINE to inline for c++/non-strict-ansi/> c99

fixes crbug.com/428383

(cherry picked from commit 6638710b9e)

Change-Id: Ief2b934353c336a75865c73c90cc3dc5e4f83913
2015-03-02 18:43:02 -08:00
e297fc7171 gif2webp: Use the default hint instead of WEBP_HINT_GRAPH.
This is much faster and the compression is slightly better too.

(cherry picked from commit c94ed49efd)

Change-Id: Ibf0d10eea83bfabfcc44ee497074767462ff41b1
2015-03-02 18:42:54 -08:00
855fe4354b Makefile.vc: add a 'legacy' RTLIBCFG option
disables buffer security checks (/GS-) and any machine optimizations
(e.g., sse2)

fixes issue #228

(cherry picked from commit 34c20c06c8)

Change-Id: I81fa483dc1654199b2017626320383d2d63317dc
2015-03-02 18:42:48 -08:00
b7eb6d55c7 gif2webp: Support GIF_DISPOSE_RESTORE_PREVIOUS
Tweaked the gif2webp_util API to support this.

Requested in: https://code.google.com/p/webp/issues/detail?id=144

(cherry picked from commit 65e5eb8a62)

Change-Id: I0e8c4edc39227355cd8d3acc55795186e25d0c3a
2015-03-02 18:42:40 -08:00
5691bdd9da gif2webp: Handle frames with odd offsets + disposal to background.
Snapping odd offsets in GIF to even offsets in WebP was causing extra row/column
being disposed in such cases.

Code is rewritten to maintain previous and current canvas (it used to maintain
previous canvas and current frame earlier). And we recompute change rectangles
as those from GIF may no longer apply.
Also, this renders methods like ReduceTransparency() and ConvertToKeyFrame()
redundant, as internally maintained current canvas is always independent of
previous canvases.

Disposal method choice: we pick the disposal method that results in the smallest
change rectangle.

(cherry picked from commit e4c829efe9)

Conflicts:
	examples/gif2webp_util.c

Change-Id: Ic31186d98fe1a2a790a89d1571b17e3abd127e79
2015-03-02 18:42:36 -08:00
8301da1380 stopwatch.h: fix includes
WEBP_INLINE -> webp/types.h
memcpy -> string.h

(cherry picked from commit 54edbf65ff)

Change-Id: Iab2ea8b553dc98be75eede751de62ab0292d1f97
2015-03-02 18:42:28 -08:00
37 changed files with 1078 additions and 552 deletions

View File

@ -10,8 +10,6 @@ ifeq ($(APP_OPTIM),release)
endif
endif
include $(CLEAR_VARS)
ifneq ($(findstring armeabi-v7a, $(TARGET_ARCH_ABI)),)
# Setting LOCAL_ARM_NEON will enable -mfpu=neon which may cause illegal
# instructions to be generated for armv7a code. Instead target the neon code
@ -21,7 +19,7 @@ else
NEON := c
endif
LOCAL_SRC_FILES := \
dec_srcs := \
src/dec/alpha.c \
src/dec/buffer.c \
src/dec/frame.c \
@ -32,6 +30,11 @@ LOCAL_SRC_FILES := \
src/dec/vp8.c \
src/dec/vp8l.c \
src/dec/webp.c \
demux_srcs := \
src/demux/demux.c \
dsp_dec_srcs := \
src/dsp/alpha_processing.c \
src/dsp/alpha_processing_sse2.c \
src/dsp/cpu.c \
@ -40,11 +43,6 @@ LOCAL_SRC_FILES := \
src/dsp/dec_mips32.c \
src/dsp/dec_neon.$(NEON) \
src/dsp/dec_sse2.c \
src/dsp/enc.c \
src/dsp/enc_avx2.c \
src/dsp/enc_mips32.c \
src/dsp/enc_neon.$(NEON) \
src/dsp/enc_sse2.c \
src/dsp/lossless.c \
src/dsp/lossless_mips32.c \
src/dsp/lossless_neon.$(NEON) \
@ -55,6 +53,15 @@ LOCAL_SRC_FILES := \
src/dsp/yuv.c \
src/dsp/yuv_mips32.c \
src/dsp/yuv_sse2.c \
dsp_enc_srcs := \
src/dsp/enc.c \
src/dsp/enc_avx2.c \
src/dsp/enc_mips32.c \
src/dsp/enc_neon.$(NEON) \
src/dsp/enc_sse2.c \
enc_srcs := \
src/enc/alpha.c \
src/enc/analysis.c \
src/enc/backward_references.c \
@ -75,19 +82,38 @@ LOCAL_SRC_FILES := \
src/enc/tree.c \
src/enc/vp8l.c \
src/enc/webpenc.c \
mux_srcs := \
src/mux/muxedit.c \
src/mux/muxinternal.c \
src/mux/muxread.c \
utils_dec_srcs := \
src/utils/bit_reader.c \
src/utils/bit_writer.c \
src/utils/color_cache.c \
src/utils/filters.c \
src/utils/huffman.c \
src/utils/huffman_encode.c \
src/utils/quant_levels.c \
src/utils/quant_levels_dec.c \
src/utils/random.c \
src/utils/rescaler.c \
src/utils/thread.c \
src/utils/utils.c \
utils_enc_srcs := \
src/utils/bit_writer.c \
src/utils/huffman_encode.c \
src/utils/quant_levels.c \
################################################################################
# libwebpdecoder
include $(CLEAR_VARS)
LOCAL_SRC_FILES := \
$(dec_srcs) \
$(dsp_dec_srcs) \
$(utils_dec_srcs) \
LOCAL_CFLAGS := $(WEBP_CFLAGS)
LOCAL_C_INCLUDES += $(LOCAL_PATH)/src
@ -96,6 +122,38 @@ LOCAL_ARM_MODE := arm
LOCAL_STATIC_LIBRARIES := cpufeatures
LOCAL_MODULE := webpdecoder_static
include $(BUILD_STATIC_LIBRARY)
ifeq ($(ENABLE_SHARED),1)
include $(CLEAR_VARS)
LOCAL_WHOLE_STATIC_LIBRARIES := webpdecoder_static
LOCAL_MODULE := webpdecoder
include $(BUILD_SHARED_LIBRARY)
endif # ENABLE_SHARED=1
################################################################################
# libwebp
include $(CLEAR_VARS)
LOCAL_SRC_FILES := \
$(dsp_enc_srcs) \
$(enc_srcs) \
$(utils_enc_srcs) \
LOCAL_CFLAGS := $(WEBP_CFLAGS)
LOCAL_C_INCLUDES += $(LOCAL_PATH)/src
# prefer arm over thumb mode for performance gains
LOCAL_ARM_MODE := arm
LOCAL_WHOLE_STATIC_LIBRARIES := webpdecoder_static
LOCAL_MODULE := webp
ifeq ($(ENABLE_SHARED),1)
@ -104,6 +162,54 @@ else
include $(BUILD_STATIC_LIBRARY)
endif
################################################################################
# libwebpdemux
include $(CLEAR_VARS)
LOCAL_SRC_FILES := $(demux_srcs)
LOCAL_CFLAGS := $(WEBP_CFLAGS)
LOCAL_C_INCLUDES += $(LOCAL_PATH)/src
# prefer arm over thumb mode for performance gains
LOCAL_ARM_MODE := arm
LOCAL_MODULE := webpdemux
ifeq ($(ENABLE_SHARED),1)
LOCAL_SHARED_LIBRARIES := webp
include $(BUILD_SHARED_LIBRARY)
else
LOCAL_STATIC_LIBRARIES := webp
include $(BUILD_STATIC_LIBRARY)
endif
################################################################################
# libwebpmux
include $(CLEAR_VARS)
LOCAL_SRC_FILES := $(mux_srcs)
LOCAL_CFLAGS := $(WEBP_CFLAGS)
LOCAL_C_INCLUDES += $(LOCAL_PATH)/src
# prefer arm over thumb mode for performance gains
LOCAL_ARM_MODE := arm
LOCAL_MODULE := webpmux
ifeq ($(ENABLE_SHARED),1)
LOCAL_SHARED_LIBRARIES := webp
include $(BUILD_SHARED_LIBRARY)
else
LOCAL_STATIC_LIBRARIES := webp
include $(BUILD_STATIC_LIBRARY)
endif
################################################################################
include $(LOCAL_PATH)/examples/Android.mk
$(call import-module,android/cpufeatures)

View File

@ -1,3 +1,47 @@
a661e50 Disable NEON code on Native Client
fcd94e9 update ChangeLog (tag: v0.4.3-rc1)
569fe57 update NEWS
bd852f5 bump version to 0.4.3
2d58b64 WebPPictureRescale: add a note about 0 width/height
a0d8ca5 examples/Android.mk: add webpmux_example target
34b1d29 Android.mk: add webpmux target
7561988 Android.mk: add webpdemux target
a987576 Android.mk: add webpdecoder{,_static} targets
a6d4859 Android.mk: split source lists per-directory
77544d5 fix iOS arm64 build with Xcode 6.3
6dea157 doc/webp-container-spec: note MSB order for chunk diagrams
f7cd57b doc/webp-container-spec: cosmetics
1d6b250 vwebp: clear canvas at the beginning of each loop
f97b3f8 webp-container-spec: clarify background clear on loop
4ba83c1 vwebp: remove unnecessary static Help() prototype
d34e8e3 vwebp/animation: display last frame on end-of-loop
bbbc524 dec/vp8: clear 'dither_' on skipped blocks
0339fa2 lossless_neon: enable subtract green for aarch64
5a0c220 Regression fix for lossless decoding
6e3a31d wicdec: (msvs) quiet some /analyze warnings
b49a578 dwebp/WritePNG: mark png variables volatile
0a4391a dwebp: include setjmp.h w/WEBP_HAVE_PNG
90f1ec5 dwebp: correct sign in format strings
b61ce86 VP8LEncodeStream: add an assert
df1081b dsp/cpu: (msvs) add include for __cpuidex
39aa055 dsp/cpu: (msvs) avoid immintrin.h on _M_ARM
f814f42 dsp/cpu: add include for _xgetbv() w/MSVS
8508ab9 cpu: fix AVX2 detection for gcc/clang targets
5769623 fix handling of zero-sized partition #0 corner case
b2e71a9 make the 'last_cpuinfo_used' variable names unique
1273e84 add -Wformat-nonliteral and -Wformat-security
3ae78eb multi-thread fix: lock each entry points with a static var
5c1eeda webp-container-spec: remove references to fragments
c5ceea4 enc_neon: fix building with non-Xcode clang (iOS)
d0859d6 iosbuild: add x64_64 simulator support
046732c WebPEncode: Support encoding same pic twice (even if modified)
4426f50 webp/types.h: use inline for clang++/-std=c++11
e297fc7 gif2webp: Use the default hint instead of WEBP_HINT_GRAPH.
855fe43 Makefile.vc: add a 'legacy' RTLIBCFG option
b7eb6d5 gif2webp: Support GIF_DISPOSE_RESTORE_PREVIOUS
5691bdd gif2webp: Handle frames with odd offsets + disposal to background.
8301da1 stopwatch.h: fix includes
6a2209a update ChangeLog (tag: v0.4.2, origin/0.4.2, 0.4.2)
36cad6a bit_reader.h: cosmetics: fix a typo
e2ecae6 enc_mips32: workaround gcc-4.9 bug
243e68d update ChangeLog (tag: v0.4.2-rc2)
@ -74,7 +118,7 @@ c2fc52e restore encode API compatibility
793368e restore decode API compatibility
b8984f3 gif2webp: fix compile with giflib 5.1.0
222f9b1 gif2webp: simplify giflib version checking
d2cc61b Extend MakeARGB32() to accept Alpha channel. (master)
d2cc61b Extend MakeARGB32() to accept Alpha channel.
4595b62 Merge "use explicit size of kErrorMessages[] arrays"
157de01 Merge "Actuate memory stats for PRINT_MEMORY_INFO"
fbda2f4 JPEG decoder: delay conversion to YUV to WebPEncode() call

View File

@ -27,7 +27,7 @@ PLATFORM_LDFLAGS = /SAFESEH
NOLOGO = /nologo
CCNODBG = cl.exe $(NOLOGO) /O2 /DNDEBUG
CCDEBUG = cl.exe $(NOLOGO) /Od /Gm /Zi /D_DEBUG /RTC1
CFLAGS = /Isrc $(NOLOGO) /W3 /EHsc /c /GS
CFLAGS = /Isrc $(NOLOGO) /W3 /EHsc /c
CFLAGS = $(CFLAGS) /DWIN32 /D_CRT_SECURE_NO_WARNINGS /DWIN32_LEAN_AND_MEAN
CFLAGS = $(CFLAGS) /DHAVE_WINCODEC_H /DWEBP_USE_THREAD
LDFLAGS = /LARGEADDRESSAWARE /MANIFEST /NXCOMPAT /DYNAMICBASE
@ -54,6 +54,11 @@ AVX2_FLAGS = /arch:AVX2
!IF "$(RTLIBCFG)" == "static"
RTLIB = /MT
RTLIBD = /MTd
!ELSE IF "$(RTLIBCFG)" == "legacy"
RTLIBCFG = static
RTLIB = /MT
RTLIBD = /MTd
CFLAGS = $(CFLAGS) /GS- /arch:IA32
!ELSE
RTLIB = /MD
RTLIBD = /MDd
@ -139,6 +144,7 @@ CFGSET = TRUE
!MESSAGE - all - build (de)mux-based targets for CFG
!MESSAGE
!MESSAGE RTLIBCFG controls the runtime library linkage - 'static' or 'dynamic'.
!MESSAGE 'legacy' will produce a Windows 2000 compatible library.
!MESSAGE OBJDIR is the path where you like to build (obj, bins, etc.),
!MESSAGE defaults to ..\obj

7
NEWS
View File

@ -1,3 +1,10 @@
- 3/3/15: version 0.4.3
This is a binary compatible release.
* Android / gcc / iOS / MSVS build fixes and improvements
* lossless decode fix (issue #239 -- since 0.4.0)
* documentation / vwebp updates for animation
* multi-threading fix (issue #234)
- 10/13/14: version 0.4.2
This is a binary compatible release.
* Android / gcc build fixes

2
README
View File

@ -4,7 +4,7 @@
\__\__/\____/\_____/__/ ____ ___
/ _/ / \ \ / _ \/ _/
/ \_/ / / \ \ __/ \__
\____/____/\_____/_____/____/v0.4.2
\____/____/\_____/_____/____/v0.4.3
Description:
============

View File

@ -1,4 +1,4 @@
AC_INIT([libwebp], [0.4.2],
AC_INIT([libwebp], [0.4.3],
[http://code.google.com/p/webp/issues],,
[http://developers.google.com/speed/webp])
AC_CANONICAL_HOST
@ -54,6 +54,7 @@ AC_DEFUN([TEST_AND_ADD_CFLAGS],
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wall])
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wdeclaration-after-statement])
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wextra])
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wformat-nonliteral])
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wformat-security])
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wmissing-declarations])
TEST_AND_ADD_CFLAGS([AM_CFLAGS], [-Wmissing-prototypes])

View File

@ -46,25 +46,16 @@ for:
* **Animation.** An image may have multiple frames with pauses between them,
making it an animation.
* **Image Fragmentation.** A single bitstream in WebP has an inherent
limitation for width or height of 2^14 pixels, and, when using VP8, a 512
KiB limit on the size of the first compressed partition. To support larger
images, the format supports images that are composed of multiple fragments,
each encoded as a separate bitstream. All fragments logically form a single
image: they have common metadata, color profile, etc. Image fragmentation
may also improve efficiency for larger images, e.g., grass can be encoded
differently than sky.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in [RFC 2119][].
Bit numbering in chunk diagrams starts at `0` for the most significant bit
('MSB 0') as described in [RFC 1166][].
**Note:** Out of the features mentioned above, lossy compression, lossless
compression, transparency, metadata, color profile and animation are finalized
and are to be considered stable. On the other hand, image fragmentation is
experimental as of now, and is open to discussion, feedback and comments.
The same is indicated using annotation "_status: experimental_" in the relevant
sections of this document.
and are to be considered stable.
Terminology & Basics
------------------------
@ -79,7 +70,7 @@ Below are additional terms used throughout this document:
_Reader/Writer_
: Code that reads WebP files is referred to as a _reader_, while code that
writes them is referred to as a _writer_.
writes them is referred to as a _writer_.
_uint16_
@ -101,10 +92,12 @@ _FourCC_
_1-based_
: An unsigned integer field storing values offset by `-1`. e.g., Such a field
would store value _25_ as _24_.
would store value _25_ as _24_.
RIFF file format
RIFF File Format
----------------
The WebP file format is based on the RIFF (resource interchange file format)
document format.
@ -144,7 +137,8 @@ _ChunkHeader('ABCD')_
chunks that apply to any RIFF file format, while FourCCs specific to a file
format are all lowercase. WebP does not follow this convention.
WebP file header
WebP File Header
----------------
0 1 2 3
@ -164,8 +158,8 @@ WebP file header
File Size: 32 bits (_uint32_)
: The size of the file in bytes starting at offset 8. The maximum value of
this field is 2^32 minus 10 bytes and thus the size of the whole file is at
most 4GiB minus 2 bytes.
this field is 2^32 minus 10 bytes and thus the size of the whole file is at
most 4GiB minus 2 bytes.
'WEBP': 32 bits
@ -177,7 +171,8 @@ the 'WEBP' FourCC. The file SHOULD NOT contain anything after it. As the size
of any chunk is even, the size given by the RIFF header is also even. The
contents of individual chunks will be described in the following sections.
Simple file format (lossy)
Simple File Format (Lossy)
--------------------------
This layout SHOULD be used if the image requires _lossy_ encoding and does not
@ -215,7 +210,8 @@ width and height. That is assumed to be the width and height of the canvas.
The VP8 specification describes how to decode the image into Y'CbCr
format. To convert to RGB, Rec. 601 SHOULD be used.
Simple file format (lossless)
Simple File Format (Lossless)
-----------------------------
**Note:** Older readers may not support files using the lossless format.
@ -253,7 +249,8 @@ The current specification of the VP8L bitstream can be found at
contains the VP8L image width and height. That is assumed to be the width
and height of the canvas.
Extended file format
Extended File Format
--------------------
**Note:** Older readers may not support files using the extended format.
@ -278,10 +275,6 @@ For a _still image_, the _image data_ consists of a single frame, whereas for
an _animated image_, it consists of multiple frames. More details about frames
can be found in the [Animation](#animation) section.
Moreover, each frame can be fragmented or non-fragmented, as will be described
in the [Extended WebP file header](#extended_header) section. More details about
fragments can be found in the [Fragments](#fragments) section.
All chunks SHOULD be placed in the same order as listed above. If a chunk
appears in the wrong place, the file is invalid, but readers MAY parse the
file, ignoring the chunks that come too late.
@ -302,7 +295,7 @@ Extended WebP file header:
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ChunkHeader('VP8X') |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|Rsv|I|L|E|X|A|F| Reserved |
|Rsv|I|L|E|X|A|R| Reserved |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Canvas Width Minus One | ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
@ -320,7 +313,7 @@ ICC profile (I): 1 bit
Alpha (L): 1 bit
: Set if any of the frames of the image contain transparency information
("alpha").
("alpha").
EXIF metadata (E): 1 bit
@ -333,11 +326,11 @@ XMP metadata (X): 1 bit
Animation (A): 1 bit
: Set if this is an animated image. Data in 'ANIM' and 'ANMF' chunks should be
used to control the animation.
used to control the animation.
Image Fragmentation (F): 1 bit _\[status: experimental\]_
Reserved (R): 1 bit
: Set if any of the frames in the image are represented by fragments.
: SHOULD be `0`.
Reserved: 24 bits
@ -382,9 +375,9 @@ animation.
Background Color: 32 bits (_uint32_)
: The default background color of the canvas in \[Blue, Green, Red, Alpha\]
byte order. This color MAY be used to fill the unused space on the canvas around
the frames, as well as the transparent pixels of the first frame. Background
color is also used when disposal method is `1`.
byte order. This color MAY be used to fill the unused space on the canvas
around the frames, as well as the transparent pixels of the first frame.
Background color is also used when disposal method is `1`.
**Note**:
@ -394,6 +387,9 @@ color is also used when disposal method is `1`.
* Viewer applications SHOULD treat the background color value as a hint, and
are not required to use it.
* The canvas is cleared at the start of each loop. The background color MAY be
used to achieve this.
Loop Count: 16 bits (_uint16_)
: The number of times to loop the animation. `0` means infinitely.
@ -402,7 +398,6 @@ This chunk MUST appear if the _Animation_ flag in the VP8X chunk is set.
If the _Animation_ flag is not set and this chunk is present, it
SHOULD be ignored.
ANMF chunk:
For animated images, this chunk contains information about a _single_ frame.
@ -445,8 +440,8 @@ Frame Height Minus One: 24 bits (_uint24_)
Frame Duration: 24 bits (_uint24_)
: The time to wait before displaying the next frame, in 1 millisecond units.
In particular, frame duration of 0 is useful when one wants to update multiple
areas of the canvas at once during the animation.
In particular, frame duration of 0 is useful when one wants to update
multiple areas of the canvas at once during the animation.
Reserved: 6 bits
@ -454,28 +449,28 @@ Reserved: 6 bits
Blending method (B): 1 bit
: Indicates how transparent pixels of _the current frame_ are to be blended with
corresponding pixels of the previous canvas:
: Indicates how transparent pixels of _the current frame_ are to be blended
with corresponding pixels of the previous canvas:
* `0`: Use alpha blending. After disposing of the previous frame, render the
current frame on the canvas using [alpha-blending](#alpha-blending). If the
current frame does not have an alpha channel, assume alpha value of 255,
effectively replacing the rectangle.
* `0`: Use alpha blending. After disposing of the previous frame, render the
current frame on the canvas using [alpha-blending](#alpha-blending). If
the current frame does not have an alpha channel, assume alpha value of
255, effectively replacing the rectangle.
* `1`: Do not blend. After disposing of the previous frame, render the
current frame on the canvas by overwriting the rectangle covered by the
current frame.
* `1`: Do not blend. After disposing of the previous frame, render the
current frame on the canvas by overwriting the rectangle covered by the
current frame.
Disposal method (D): 1 bit
: Indicates how _the current frame_ is to be treated after it has been displayed
(before rendering the next frame) on the canvas:
: Indicates how _the current frame_ is to be treated after it has been
displayed (before rendering the next frame) on the canvas:
* `0`: Do not dispose. Leave the canvas as is.
* `0`: Do not dispose. Leave the canvas as is.
* `1`: Dispose to background color. Fill the _rectangle_ on the canvas covered
by the _current frame_ with background color specified in the
[ANIM chunk](#anim_chunk).
* `1`: Dispose to background color. Fill the _rectangle_ on the canvas
covered by the _current frame_ with background color specified in the
[ANIM chunk](#anim_chunk).
**Notes**:
@ -506,9 +501,7 @@ Disposal method (D): 1 bit
Frame Data: _Chunk Size_ - `16` bytes
: For a fragmented frame, it consists of multiple [fragment chunks](#fragments).
: For a non-fragmented frame, it consists of:
: Consists of:
* An optional [alpha subchunk](#alpha) for the frame.
@ -519,49 +512,6 @@ Frame Data: _Chunk Size_ - `16` bytes
**Note**: The 'ANMF' payload, _Frame Data_ above, consists of individual
_padded_ chunks as described by the [RIFF file format](#riff-file-format).
#### Fragments _\[status: experimental\]_
For images that are represented by fragments, this chunk contains data for
a single fragment. If the _Image Fragmentation Flag_ is not set, then this chunk
SHOULD NOT be present.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ChunkHeader('FRGM') |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Fragment X | ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
... Fragment Y | Fragment Data |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Fragment X: 24 bits (_uint24_)
: The X coordinate of the upper left corner of the fragment is `Fragment X * 2`
Fragment Y: 24 bits (_uint24_)
: The Y coordinate of the upper left corner of the fragment is `Fragment Y * 2`
Fragment Data: _Chunk Size_ - `6` bytes
: It contains:
* An optional [alpha subchunk](#alpha) for the fragment.
* The [bitstream subchunk](#bitstream-vp8vp8l) for the fragment.
* An optional list of [unknown chunks](#unknown-chunks).
Note: The width and height of the fragment is obtained from the bitstream
subchunk.
The fragments of a frame SHOULD have the following properties:
* They collectively cover the whole frame.
* No pair of fragments have any overlapping region on the frame.
* No portion of any fragment should be located outside of the canvas.
#### Alpha
0 1 2 3
@ -579,20 +529,20 @@ Reserved (Rsv): 2 bits
Pre-processing (P): 2 bits
: These INFORMATIVE bits are used to signal the pre-processing that has
been performed during compression. The decoder can use this information to
e.g. dither the values or smooth the gradients prior to display.
been performed during compression. The decoder can use this information to
e.g. dither the values or smooth the gradients prior to display.
* `0`: no pre-processing
* `1`: level reduction
* `0`: no pre-processing
* `1`: level reduction
Filtering method (F): 2 bits
: The filtering method used:
* `0`: None.
* `1`: Horizontal filter.
* `2`: Vertical filter.
* `3`: Gradient filter.
* `0`: None.
* `1`: Horizontal filter.
* `2`: Vertical filter.
* `3`: Gradient filter.
For each pixel, filtering is performed using the following calculations.
Assume the alpha values surrounding the current `X` position are labeled as:
@ -636,15 +586,15 @@ Compression method (C): 2 bits
: The compression method used:
* `0`: No compression.
* `1`: Compressed using the WebP lossless format.
* `0`: No compression.
* `1`: Compressed using the WebP lossless format.
Alpha bitstream: _Chunk Size_ - `1` bytes
: Encoded alpha bitstream.
This optional chunk contains encoded alpha data for this frame/fragment. A
frame/fragment containing a 'VP8L' chunk SHOULD NOT contain this chunk.
This optional chunk contains encoded alpha data for this frame. A frame
containing a 'VP8L' chunk SHOULD NOT contain this chunk.
**Rationale**: The transparency information is already part of the 'VP8L'
chunk.
@ -675,15 +625,15 @@ compression method is '0') or compressed using the lossless format
#### Bitstream (VP8/VP8L)
This chunk contains compressed bitstream data for a single frame/fragment.
This chunk contains compressed bitstream data for a single frame.
A bitstream chunk may be either (i) a VP8 chunk, using "VP8 " (note the
significant fourth-character space) as its tag _or_ (ii) a VP8L chunk, using
"VP8L" as its tag.
The formats of VP8 and VP8L chunks are as described in sections
[Simple file format (lossy)](#simple-file-format-lossy)
and [Simple file format (lossless)](#simple-file-format-lossless) respectively.
[Simple File Format (Lossy)](#simple-file-format-lossy)
and [Simple File Format (Lossless)](#simple-file-format-lossless) respectively.
#### Color profile
@ -731,7 +681,6 @@ EXIF Metadata: _Chunk Size_ bytes
: image metadata in EXIF format.
XMP chunk:
0 1 2 3
@ -762,47 +711,17 @@ A file MAY contain unknown chunks:
* At the end of the file as described in [Extended WebP file
header](#extended_header) section.
* At the end of FRGM and ANMF chunks as described in [Fragments](#fragments)
and [Animation](#animation) sections.
* At the end of ANMF chunks as described in the
[Animation](#animation) section.
Readers SHOULD ignore these chunks. Writers SHOULD preserve them in their
original order (unless they specifically intend to modify these chunks).
### Assembling the Canvas from fragments/frames
### Assembling the Canvas from frames
Here we provide an overview of how a reader should assemble a canvas in case
of a fragmented-image and in case of an animated image. The notation
_VP8X.field_ means the field in the 'VP8X' chunk with the same description.
Displaying a _fragmented image_ canvas MUST be equivalent to the following
pseudocode: _\[status: experimental\]_
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
assert VP8X.flags.hasFragments
canvas ← new black image of size VP8X.canvasWidth x VP8X.canvasHeight.
frgm_params ← nil
for chunk in image_data:
assert chunk.tag is "FRGM"
frgm_params.fragmentX = Fragment X
frgm_params.fragmentY = Fragment Y
for subchunk in 'Fragment Data':
if subchunk.tag == "ALPH":
assert alpha subchunks not found in 'Fragment Data' earlier
frgm_params.alpha = alpha_data
else if subchunk.tag == "VP8 " OR subchunk.tag == "VP8L":
assert bitstream subchunks not found in 'Fragment Data' earlier
frgm_params.bitstream = bitstream_data
frgm_params.fragmentWidth = Width extracted from bitstream subchunk
frgm_params.fragmentHeight = Height extracted from bitstream subchunk
assert VP8X.canvasWidth >=
frgm_params.fragmentX + frgm_params.fragmentWidth
assert VP8X.canvasHeight >=
frgm_params.fragmentY + frgm_params.fragmentHeight
assert fragment has the properties mentioned in "Image Fragments" section.
render fragment with frame_params.alpha and frame_params.bitstream on canvas
with top-left corner in (frgm_params.fragmentX, frgm_params.fragmentY).
canvas contains the decoded canvas.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Here we provide an overview of how a reader should assemble a canvas in the
case of an animated image. The notation _VP8X.field_ means the field in the
'VP8X' chunk with the same description.
Displaying an _animated image_ canvas MUST be equivalent to the following
pseudocode:
@ -810,28 +729,25 @@ pseudocode:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
assert VP8X.flags.hasAnimation
canvas ← new image of size VP8X.canvasWidth x VP8X.canvasHeight with
background color ANIM.background_color.
background color ANIM.background_color.
loop_count ← ANIM.loopCount
dispose_method ← ANIM.disposeMethod
if loop_count == 0:
loop_count = ∞
frame_params ← nil
for loop = 0, ..., loop_count - 1
assert next chunk in image_data is ANMF
frame_params.frameX = Frame X
frame_params.frameY = Frame Y
frame_params.frameWidth = Frame Width Minus One + 1
frame_params.frameHeight = Frame Height Minus One + 1
frame_params.frameDuration = Frame Duration
assert VP8X.canvasWidth >= frame_params.frameX + frame_params.frameWidth
assert VP8X.canvasHeight >= frame_params.frameY + frame_params.frameHeight
if VP8X.flags.hasFragments and first subchunk in 'Frame Data' is FRGM
// Fragmented frame.
frame_params.{bitstream,alpha} = canvas decoded from subchunks in
'Frame Data' as per the pseudocode for
_fragmented image_ above.
else
// Non-fragmented frame.
assert next chunk in image_data is ANMF
for loop = 0..loop_count - 1
clear canvas to ANIM.background_color or application defined color
until eof or non-ANMF chunk
frame_params.frameX = Frame X
frame_params.frameY = Frame Y
frame_params.frameWidth = Frame Width Minus One + 1
frame_params.frameHeight = Frame Height Minus One + 1
frame_params.frameDuration = Frame Duration
frame_right = frame_params.frameX + frame_params.frameWidth
frame_bottom = frame_params.frameY + frame_params.frameHeight
assert VP8X.canvasWidth >= frame_right
assert VP8X.canvasHeight >= frame_bottom
for subchunk in 'Frame Data':
if subchunk.tag == "ALPH":
assert alpha subchunks not found in 'Frame Data' earlier
@ -839,14 +755,15 @@ for loop = 0, ..., loop_count - 1
else if subchunk.tag == "VP8 " OR subchunk.tag == "VP8L":
assert bitstream subchunks not found in 'Frame Data' earlier
frame_params.bitstream = bitstream_data
render frame with frame_params.alpha and frame_params.bitstream on canvas
with top-left corner in (frame_params.frameX, frame_params.frameY), using
dispose method dispose_method.
Show the contents of the image for frame_params.frameDuration * 1ms.
canvas contains the decoded canvas.
render frame with frame_params.alpha and frame_params.bitstream on
canvas with top-left corner at (frame_params.frameX,
frame_params.frameY), using dispose method dispose_method.
canvas contains the decoded image.
Show the contents of the canvas for frame_params.frameDuration * 1ms.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Example file layouts
Example File Layouts
--------------------
A lossy encoded image with alpha may look as follows:
@ -878,17 +795,6 @@ RIFF/WEBP
+- XMP (metadata)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
A fragmented image may look as follows:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
RIFF/WEBP
+- VP8X (descriptions of features used)
+- FRGM (fragment1 parameters + data)
+- FRGM (fragment2 parameters + data)
+- FRGM (fragment3 parameters + data)
+- FRGM (fragment4 parameters + data)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
An animated image with EXIF metadata may look as follows:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@ -906,4 +812,5 @@ RIFF/WEBP
[webpllspec]: https://gerrit.chromium.org/gerrit/gitweb?p=webm/libwebp.git;a=blob;f=doc/webp-lossless-bitstream-spec.txt;hb=master
[iccspec]: http://www.color.org/icc_specs2.xalter
[metadata]: http://www.metadataworkinggroup.org/pdf/mwg_guidance.pdf
[rfc 1166]: http://tools.ietf.org/html/rfc1166
[rfc 2119]: http://tools.ietf.org/html/rfc2119

View File

@ -1,5 +1,8 @@
LOCAL_PATH := $(call my-dir)
################################################################################
# libexample_util
include $(CLEAR_VARS)
LOCAL_SRC_FILES := \
@ -12,6 +15,9 @@ LOCAL_MODULE := example_util
include $(BUILD_STATIC_LIBRARY)
################################################################################
# cwebp
include $(CLEAR_VARS)
# Note: to enable jpeg/png encoding the sources from AOSP can be used with
@ -32,6 +38,9 @@ LOCAL_MODULE := cwebp
include $(BUILD_EXECUTABLE)
################################################################################
# dwebp
include $(CLEAR_VARS)
LOCAL_SRC_FILES := \
@ -44,3 +53,19 @@ LOCAL_STATIC_LIBRARIES := example_util webp
LOCAL_MODULE := dwebp
include $(BUILD_EXECUTABLE)
################################################################################
# webpmux
include $(CLEAR_VARS)
LOCAL_SRC_FILES := \
webpmux.c \
LOCAL_CFLAGS := $(WEBP_CFLAGS)
LOCAL_C_INCLUDES := $(LOCAL_PATH)/../src
LOCAL_STATIC_LIBRARIES := example_util webpmux webp
LOCAL_MODULE := webpmux_example
include $(BUILD_EXECUTABLE)

View File

@ -22,6 +22,7 @@
#ifdef WEBP_HAVE_PNG
#include <png.h>
#include <setjmp.h> // note: this must be included *after* png.h
#endif
#ifdef HAVE_WINCODEC_H
@ -192,8 +193,8 @@ static int WritePNG(FILE* out_file, const WebPDecBuffer* const buffer) {
uint8_t* const rgb = buffer->u.RGBA.rgba;
const int stride = buffer->u.RGBA.stride;
const int has_alpha = (buffer->colorspace == MODE_RGBA);
png_structp png;
png_infop info;
volatile png_structp png;
volatile png_infop info;
png_uint_32 y;
png = png_create_write_struct(PNG_LIBPNG_VER_STRING,
@ -203,11 +204,11 @@ static int WritePNG(FILE* out_file, const WebPDecBuffer* const buffer) {
}
info = png_create_info_struct(png);
if (info == NULL) {
png_destroy_write_struct(&png, NULL);
png_destroy_write_struct((png_structpp)&png, NULL);
return 0;
}
if (setjmp(png_jmpbuf(png))) {
png_destroy_write_struct(&png, &info);
png_destroy_write_struct((png_structpp)&png, (png_infopp)&info);
return 0;
}
png_init_io(png, out_file);
@ -221,7 +222,7 @@ static int WritePNG(FILE* out_file, const WebPDecBuffer* const buffer) {
png_write_rows(png, &row, 1);
}
png_write_end(png, info);
png_destroy_write_struct(&png, &info);
png_destroy_write_struct((png_structpp)&png, (png_infopp)&info);
return 1;
}
#else // !HAVE_WINCODEC_H && !WEBP_HAVE_PNG
@ -244,10 +245,10 @@ static int WritePPM(FILE* fout, const WebPDecBuffer* const buffer, int alpha) {
uint32_t y;
if (alpha) {
fprintf(fout, "P7\nWIDTH %d\nHEIGHT %d\nDEPTH 4\nMAXVAL 255\n"
fprintf(fout, "P7\nWIDTH %u\nHEIGHT %u\nDEPTH 4\nMAXVAL 255\n"
"TUPLTYPE RGB_ALPHA\nENDHDR\n", width, height);
} else {
fprintf(fout, "P6\n%d %d\n255\n", width, height);
fprintf(fout, "P6\n%u %u\n255\n", width, height);
}
for (y = 0; y < height; ++y) {
if (fwrite(rgb + y * stride, width, bytes_per_px, fout) != bytes_per_px) {
@ -404,7 +405,7 @@ static int WriteAlphaPlane(FILE* fout, const WebPDecBuffer* const buffer) {
const int a_stride = buffer->u.YUVA.a_stride;
uint32_t y;
assert(a != NULL);
fprintf(fout, "P5\n%d %d\n255\n", width, height);
fprintf(fout, "P5\n%u %u\n255\n", width, height);
for (y = 0; y < height; ++y) {
if (fwrite(a + y * a_stride, width, 1, fout) != 1) {
return 0;

View File

@ -46,18 +46,7 @@
//------------------------------------------------------------------------------
static int transparent_index; // Index of transparent color in the map.
static void ResetFrameInfo(WebPMuxFrameInfo* const info) {
WebPDataInit(&info->bitstream);
info->x_offset = 0;
info->y_offset = 0;
info->duration = 0;
info->id = WEBP_CHUNK_ANMF;
info->dispose_method = WEBP_MUX_DISPOSE_NONE;
info->blend_method = WEBP_MUX_BLEND;
transparent_index = -1; // Opaque frame by default.
}
static int transparent_index = -1; // Opaque frame by default.
static void SanitizeKeyFrameIntervals(size_t* const kmin_ptr,
size_t* const kmax_ptr) {
@ -270,7 +259,8 @@ int main(int argc, const char *argv[]) {
GifFileType* gif = NULL;
WebPConfig config;
WebPPicture frame;
WebPMuxFrameInfo info;
int duration = 0;
FrameDisposeMethod orig_dispose = FRAME_DISPOSE_NONE;
WebPMuxAnimParams anim = { WHITE_COLOR, 0 };
WebPFrameCache* cache = NULL;
@ -290,14 +280,11 @@ int main(int argc, const char *argv[]) {
size_t kmax = 0;
int allow_mixed = 0; // If true, each frame can be lossy or lossless.
ResetFrameInfo(&info);
if (!WebPConfigInit(&config) || !WebPPictureInit(&frame)) {
fprintf(stderr, "Error! Version mismatch!\n");
return -1;
}
config.lossless = 1; // Use lossless compression by default.
config.image_hint = WEBP_HINT_GRAPH; // always low-color
if (argc == 1) {
Help();
@ -499,7 +486,8 @@ int main(int argc, const char *argv[]) {
goto End;
}
if (!WebPFrameCacheAddFrame(cache, &config, &gif_rect, &frame, &info)) {
if (!WebPFrameCacheAddFrame(cache, &config, &gif_rect, orig_dispose,
duration, &frame)) {
fprintf(stderr, "Error! Cannot encode frame as WebP\n");
fprintf(stderr, "Error code: %d\n", frame.error_code);
}
@ -515,7 +503,9 @@ int main(int argc, const char *argv[]) {
// In GIF, graphic control extensions are optional for a frame, so we
// may not get one before reading the next frame. To handle this case,
// we reset frame properties to reasonable defaults for the next frame.
ResetFrameInfo(&info);
orig_dispose = FRAME_DISPOSE_NONE;
duration = 0;
transparent_index = -1; // Opaque frame by default.
break;
}
case EXTENSION_RECORD_TYPE: {
@ -533,20 +523,19 @@ int main(int argc, const char *argv[]) {
const int dispose = (flags >> GIF_DISPOSE_SHIFT) & GIF_DISPOSE_MASK;
const int delay = data[2] | (data[3] << 8); // In 10 ms units.
if (data[0] != 4) goto End;
info.duration = delay * 10; // Duration is in 1 ms units for WebP.
if (dispose == 3) {
static int warning_printed = 0;
if (!warning_printed) {
fprintf(stderr, "WARNING: GIF_DISPOSE_RESTORE unsupported.\n");
warning_printed = 1;
}
// failsafe. TODO(urvang): emulate the correct behaviour by
// recoding the whole frame.
info.dispose_method = WEBP_MUX_DISPOSE_BACKGROUND;
} else {
info.dispose_method =
(dispose == 2) ? WEBP_MUX_DISPOSE_BACKGROUND
: WEBP_MUX_DISPOSE_NONE;
duration = delay * 10; // Duration is in 1 ms units for WebP.
switch (dispose) {
case 3:
orig_dispose = FRAME_DISPOSE_RESTORE_PREVIOUS;
break;
case 2:
orig_dispose = FRAME_DISPOSE_BACKGROUND;
break;
case 1:
case 0:
default:
orig_dispose = FRAME_DISPOSE_NONE;
break;
}
transparent_index = (flags & GIF_TRANSPARENT_MASK) ? data[4] : -1;
break;

File diff suppressed because it is too large Load Diff

View File

@ -29,6 +29,13 @@ extern "C" {
struct WebPPicture;
// Includes all disposal methods, even the ones not supported by WebP bitstream.
typedef enum FrameDisposeMethod {
FRAME_DISPOSE_NONE,
FRAME_DISPOSE_BACKGROUND,
FRAME_DISPOSE_RESTORE_PREVIOUS
} FrameDisposeMethod;
typedef struct {
int x_offset, y_offset, width, height;
} WebPFrameRect;
@ -53,15 +60,15 @@ WebPFrameCache* WebPFrameCacheNew(int width, int height,
// Release all the frame data from 'cache' and free 'cache'.
void WebPFrameCacheDelete(WebPFrameCache* const cache);
// Given an image described by 'frame', 'info' and 'orig_rect', optimize it for
// WebP, encode it and add it to 'cache'. 'orig_rect' can be NULL.
// This takes care of frame disposal too, according to 'info->dispose_method'.
// Given an image described by 'frame', 'rect', 'dispose_method' and 'duration',
// optimize it for WebP, encode it and add it to 'cache'. 'rect' can be NULL.
// This takes care of frame disposal too, according to 'dispose_method'.
// Returns false in case of error (and sets frame->error_code accordingly).
int WebPFrameCacheAddFrame(WebPFrameCache* const cache,
const WebPConfig* const config,
const WebPFrameRect* const orig_rect,
WebPPicture* const frame,
WebPMuxFrameInfo* const info);
const WebPFrameRect* const rect,
FrameDisposeMethod dispose_method, int duration,
WebPPicture* const frame);
// Flush the *ready* frames from cache and add them to 'mux'. If 'verbose' is
// true, prints the information about these frames.

View File

@ -14,6 +14,8 @@
#ifndef WEBP_EXAMPLES_STOPWATCH_H_
#define WEBP_EXAMPLES_STOPWATCH_H_
#include "webp/types.h"
#if defined _WIN32 && !defined __GNUC__
#include <windows.h>
@ -37,6 +39,7 @@ static WEBP_INLINE double StopwatchReadAndReset(Stopwatch* watch) {
#else /* !_WIN32 */
#include <string.h> // memcpy
#include <sys/time.h>
typedef struct timeval Stopwatch;

View File

@ -42,8 +42,6 @@
#define snprintf _snprintf
#endif
static void Help(void);
// Unfortunate global variables. Gathered into a struct for comfort.
static struct {
int has_animation;
@ -82,6 +80,16 @@ static void ClearParams(void) {
kParams.dmux = NULL;
}
// Sets the previous frame to the dimensions of the canvas and has it dispose
// to background to cause the canvas to be cleared.
static void ClearPreviousFrame(void) {
WebPIterator* const prev = &kParams.prev_frame;
prev->width = kParams.canvas_width;
prev->height = kParams.canvas_height;
prev->x_offset = prev->y_offset = 0;
prev->dispose_method = WEBP_MUX_DISPOSE_BACKGROUND;
}
// -----------------------------------------------------------------------------
// Color profile handling
static int ApplyColorProfile(const WebPData* const profile,
@ -181,6 +189,8 @@ static void decode_callback(int what) {
if (WebPDemuxGetFrame(kParams.dmux, 1, curr)) {
--kParams.loop_count;
kParams.done = (kParams.loop_count == 0);
if (kParams.done) return;
ClearPreviousFrame();
} else {
kParams.decoding_error = 1;
kParams.done = 1;
@ -395,7 +405,6 @@ int main(int argc, char *argv[]) {
int c;
WebPDecoderConfig* const config = &kParams.config;
WebPIterator* const curr = &kParams.curr_frame;
WebPIterator* const prev = &kParams.prev_frame;
if (!WebPInitDecoderConfig(config)) {
fprintf(stderr, "Library version mismatch!\n");
@ -486,10 +495,7 @@ int main(int argc, char *argv[]) {
printf("Canvas: %d x %d\n", kParams.canvas_width, kParams.canvas_height);
}
prev->width = kParams.canvas_width;
prev->height = kParams.canvas_height;
prev->x_offset = prev->y_offset = 0;
prev->dispose_method = WEBP_MUX_DISPOSE_BACKGROUND;
ClearPreviousFrame();
memset(&kParams.iccp, 0, sizeof(kParams.iccp));
kParams.has_color_profile =

View File

@ -15,6 +15,7 @@
#include "webp/config.h"
#endif
#include <assert.h>
#include <stdio.h>
#ifdef HAVE_WINCODEC_H
@ -109,6 +110,7 @@ static HRESULT ExtractICCP(IWICImagingFactory* const factory,
IFS(IWICBitmapFrameDecode_GetColorContexts(frame,
count, color_contexts,
&num_color_contexts));
assert(FAILED(hr) || num_color_contexts <= count);
for (i = 0; SUCCEEDED(hr) && i < num_color_contexts; ++i) {
WICColorContextType type;
IFS(IWICColorContext_GetType(color_contexts[i], &type));
@ -116,7 +118,7 @@ static HRESULT ExtractICCP(IWICImagingFactory* const factory,
UINT size;
IFS(IWICColorContext_GetProfileBytes(color_contexts[i],
0, NULL, &size));
if (size > 0) {
if (SUCCEEDED(hr) && size > 0) {
iccp->bytes = (uint8_t*)malloc(size);
if (iccp->bytes == NULL) {
hr = E_OUTOFMEMORY;
@ -261,7 +263,7 @@ int ReadPictureWithWIC(const char* const filename,
IFS(IWICBitmapFrameDecode_GetPixelFormat(frame, &src_pixel_format));
IFS(IWICBitmapDecoder_GetContainerFormat(decoder, &src_container_format));
if (keep_alpha) {
if (SUCCEEDED(hr) && keep_alpha) {
const GUID** guid;
for (guid = kAlphaContainers; *guid != NULL; ++guid) {
if (IsEqualGUID(MAKE_REFGUID(src_container_format),

View File

@ -26,7 +26,9 @@ readonly OLDPATH=${PATH}
# Add iPhoneOS-V6 to the list of platforms below if you need armv6 support.
# Note that iPhoneOS-V6 support is not available with the iOS6 SDK.
readonly PLATFORMS="iPhoneSimulator iPhoneOS-V7 iPhoneOS-V7s iPhoneOS-V7-arm64"
PLATFORMS="iPhoneSimulator iPhoneSimulator64"
PLATFORMS+=" iPhoneOS-V7 iPhoneOS-V7s iPhoneOS-V7-arm64"
readonly PLATFORMS
readonly SRCDIR=$(dirname $0)
readonly TOPDIR=$(pwd)
readonly BUILDDIR="${TOPDIR}/iosbuild"
@ -78,6 +80,9 @@ for PLATFORM in ${PLATFORMS}; do
elif [[ "${PLATFORM}" == "iPhoneOS-V6" ]]; then
PLATFORM="iPhoneOS"
ARCH="armv6"
elif [[ "${PLATFORM}" == "iPhoneSimulator64" ]]; then
PLATFORM="iPhoneSimulator"
ARCH="x86_64"
else
ARCH="i386"
fi

View File

@ -67,6 +67,8 @@ EXTRA_FLAGS += -Wmissing-prototypes
EXTRA_FLAGS += -Wmissing-declarations
EXTRA_FLAGS += -Wdeclaration-after-statement
EXTRA_FLAGS += -Wshadow
EXTRA_FLAGS += -Wformat-security -Wformat-nonliteral
# EXTRA_FLAGS += -Wvla
# AVX2-specific flags:

View File

@ -35,7 +35,7 @@ libwebp_la_LIBADD += utils/libwebputils.la
# other than the ones listed on the command line, i.e., after linking, it will
# not have unresolved symbols. Some platforms (Windows among them) require all
# symbols in shared libraries to be resolved at library creation.
libwebp_la_LDFLAGS = -no-undefined -version-info 5:2:0
libwebp_la_LDFLAGS = -no-undefined -version-info 5:3:0
libwebpincludedir = $(includedir)/webp
pkgconfig_DATA = libwebp.pc
@ -47,7 +47,7 @@ if BUILD_LIBWEBPDECODER
libwebpdecoder_la_LIBADD += dsp/libwebpdspdecode.la
libwebpdecoder_la_LIBADD += utils/libwebputilsdecode.la
libwebpdecoder_la_LDFLAGS = -no-undefined -version-info 1:2:0
libwebpdecoder_la_LDFLAGS = -no-undefined -version-info 1:3:0
pkgconfig_DATA += libwebpdecoder.pc
endif

View File

@ -357,30 +357,33 @@ static VP8StatusCode DecodeVP8FrameHeader(WebPIDecoder* const idec) {
}
// Partition #0
static int CopyParts0Data(WebPIDecoder* const idec) {
static VP8StatusCode CopyParts0Data(WebPIDecoder* const idec) {
VP8Decoder* const dec = (VP8Decoder*)idec->dec_;
VP8BitReader* const br = &dec->br_;
const size_t psize = br->buf_end_ - br->buf_;
const size_t part_size = br->buf_end_ - br->buf_;
MemBuffer* const mem = &idec->mem_;
assert(!idec->is_lossless_);
assert(mem->part0_buf_ == NULL);
assert(psize > 0);
assert(psize <= mem->part0_size_); // Format limit: no need for runtime check
// the following is a format limitation, no need for runtime check:
assert(part_size <= mem->part0_size_);
if (part_size == 0) { // can't have zero-size partition #0
return VP8_STATUS_BITSTREAM_ERROR;
}
if (mem->mode_ == MEM_MODE_APPEND) {
// We copy and grab ownership of the partition #0 data.
uint8_t* const part0_buf = (uint8_t*)WebPSafeMalloc(1ULL, psize);
uint8_t* const part0_buf = (uint8_t*)WebPSafeMalloc(1ULL, part_size);
if (part0_buf == NULL) {
return 0;
return VP8_STATUS_OUT_OF_MEMORY;
}
memcpy(part0_buf, br->buf_, psize);
memcpy(part0_buf, br->buf_, part_size);
mem->part0_buf_ = part0_buf;
br->buf_ = part0_buf;
br->buf_end_ = part0_buf + psize;
br->buf_end_ = part0_buf + part_size;
} else {
// Else: just keep pointers to the partition #0's data in dec_->br_.
}
mem->start_ += psize;
return 1;
mem->start_ += part_size;
return VP8_STATUS_OK;
}
static VP8StatusCode DecodePartition0(WebPIDecoder* const idec) {
@ -414,8 +417,10 @@ static VP8StatusCode DecodePartition0(WebPIDecoder* const idec) {
dec->mt_method_ = VP8GetThreadMethod(params->options, NULL,
io->width, io->height);
VP8InitDithering(params->options, dec);
if (!CopyParts0Data(idec)) {
return IDecError(idec, VP8_STATUS_OUT_OF_MEMORY);
dec->status_ = CopyParts0Data(idec);
if (dec->status_ != VP8_STATUS_OK) {
return IDecError(idec, dec->status_);
}
// Finish setting up the decoding parameters. Will call io->setup().

View File

@ -562,6 +562,7 @@ int VP8DecodeMB(VP8Decoder* const dec, VP8BitReader* const token_br) {
}
block->non_zero_y_ = 0;
block->non_zero_uv_ = 0;
block->dither_ = 0;
}
if (dec->filter_type_ > 0) { // store filter info

View File

@ -31,7 +31,7 @@ extern "C" {
// version numbers
#define DEC_MAJ_VERSION 0
#define DEC_MIN_VERSION 4
#define DEC_REV_VERSION 2
#define DEC_REV_VERSION 3
// intra prediction modes
enum { B_DC_PRED = 0, // 4x4 modes

View File

@ -900,7 +900,7 @@ static int DecodeImageData(VP8LDecoder* const dec, uint32_t* const data,
process_func(dec, row);
}
}
if (src < src_last) {
if (src < src_end) {
if (col & mask) htree_group = GetHtreeGroupForPos(hdr, col, row);
if (color_cache != NULL) {
while (last_cached < src) {

View File

@ -311,7 +311,12 @@ int (*WebPExtractAlpha)(const uint8_t*, int, int, int, uint8_t*, int);
extern void WebPInitAlphaProcessingSSE2(void);
static volatile VP8CPUInfo alpha_processing_last_cpuinfo_used =
(VP8CPUInfo)&alpha_processing_last_cpuinfo_used;
void WebPInitAlphaProcessing(void) {
if (alpha_processing_last_cpuinfo_used == VP8GetCPUInfo) return;
WebPMultARGBRow = MultARGBRow;
WebPMultRow = MultRow;
WebPApplyAlphaMultiply = ApplyAlphaMultiply;
@ -326,4 +331,5 @@ void WebPInitAlphaProcessing(void) {
}
#endif
}
alpha_processing_last_cpuinfo_used = VP8GetCPUInfo;
}

View File

@ -29,16 +29,18 @@ static WEBP_INLINE void GetCPUInfo(int cpu_info[4], int info_type) {
"cpuid\n"
"xchg %%edi, %%ebx\n"
: "=a"(cpu_info[0]), "=D"(cpu_info[1]), "=c"(cpu_info[2]), "=d"(cpu_info[3])
: "a"(info_type));
: "a"(info_type), "c"(0));
}
#elif defined(__i386__) || defined(__x86_64__)
static WEBP_INLINE void GetCPUInfo(int cpu_info[4], int info_type) {
__asm__ volatile (
"cpuid\n"
: "=a"(cpu_info[0]), "=b"(cpu_info[1]), "=c"(cpu_info[2]), "=d"(cpu_info[3])
: "a"(info_type));
: "a"(info_type), "c"(0));
}
#elif defined(_MSC_FULL_VER) && _MSC_FULL_VER >= 150030729 // >= VS2008 SP1
#elif (defined(_M_X64) || defined(_M_IX86)) && \
defined(_MSC_FULL_VER) && _MSC_FULL_VER >= 150030729 // >= VS2008 SP1
#include <intrin.h>
#define GetCPUInfo(info, type) __cpuidex(info, type, 0) // set ecx=0
#elif defined(WEBP_MSC_SSE2)
#define GetCPUInfo __cpuid
@ -55,7 +57,9 @@ static WEBP_INLINE uint64_t xgetbv(void) {
: "=a"(eax), "=d"(edx) : "c" (ecx));
return ((uint64_t)edx << 32) | eax;
}
#elif defined(_MSC_FULL_VER) && _MSC_FULL_VER >= 160040219 // >= VS2010 SP1
#elif (defined(_M_X64) || defined(_M_IX86)) && \
defined(_MSC_FULL_VER) && _MSC_FULL_VER >= 160040219 // >= VS2010 SP1
#include <immintrin.h>
#define xgetbv() _xgetbv(0)
#elif defined(_MSC_VER) && defined(_M_IX86)
static WEBP_INLINE uint64_t xgetbv(void) {

View File

@ -688,7 +688,12 @@ extern void VP8DspInitSSE2(void);
extern void VP8DspInitNEON(void);
extern void VP8DspInitMIPS32(void);
static volatile VP8CPUInfo dec_last_cpuinfo_used =
(VP8CPUInfo)&dec_last_cpuinfo_used;
void VP8DspInit(void) {
if (dec_last_cpuinfo_used == VP8GetCPUInfo) return;
VP8InitClipTables();
VP8TransformWHT = TransformWHT;
@ -727,5 +732,5 @@ void VP8DspInit(void) {
}
#endif
}
dec_last_cpuinfo_used = VP8GetCPUInfo;
}

View File

@ -66,7 +66,10 @@ extern "C" {
#define WEBP_ANDROID_NEON // Android targets that might support NEON
#endif
#if defined(__ARM_NEON__) || defined(WEBP_ANDROID_NEON) || defined(__aarch64__)
// The intrinsics currently cause compiler errors with arm-nacl-gcc and the
// inline assembly would need to be modified for use with Native Client.
#if (defined(__ARM_NEON__) || defined(WEBP_ANDROID_NEON) || \
defined(__aarch64__)) && !defined(__native_client__)
#define WEBP_USE_NEON
#endif

View File

@ -692,7 +692,12 @@ extern void VP8EncDspInitAVX2(void);
extern void VP8EncDspInitNEON(void);
extern void VP8EncDspInitMIPS32(void);
static volatile VP8CPUInfo enc_last_cpuinfo_used =
(VP8CPUInfo)&enc_last_cpuinfo_used;
void VP8EncDspInit(void) {
if (enc_last_cpuinfo_used == VP8GetCPUInfo) return;
VP8DspInit(); // common inverse transforms
InitTables();
@ -737,5 +742,6 @@ void VP8EncDspInit(void) {
}
#endif
}
enc_last_cpuinfo_used = VP8GetCPUInfo;
}

View File

@ -1012,8 +1012,10 @@ static int QuantizeBlock(int16_t in[16], int16_t out[16],
const int16x8_t out0 = Quantize(in, mtx, 0);
const int16x8_t out1 = Quantize(in, mtx, 8);
uint8x8x4_t shuffles;
// vtbl4_u8 is marked unavailable for iOS arm64, use wider versions there.
#if defined(__APPLE__) && defined(__aarch64__)
// vtbl?_u8 are marked unavailable for iOS arm64 with Xcode < 6.3, use
// non-standard versions there.
#if defined(__APPLE__) && defined(__aarch64__) && \
defined(__apple_build_version__) && (__apple_build_version__< 6020037)
uint8x16x2_t all_out;
INIT_VECTOR2(all_out, vreinterpretq_u8_s16(out0), vreinterpretq_u8_s16(out1));
INIT_VECTOR4(shuffles,

View File

@ -1590,7 +1590,12 @@ extern void VP8LDspInitSSE2(void);
extern void VP8LDspInitNEON(void);
extern void VP8LDspInitMIPS32(void);
static volatile VP8CPUInfo lossless_last_cpuinfo_used =
(VP8CPUInfo)&lossless_last_cpuinfo_used;
void VP8LDspInit(void) {
if (lossless_last_cpuinfo_used == VP8GetCPUInfo) return;
memcpy(VP8LPredictors, kPredictorsC, sizeof(VP8LPredictors));
VP8LSubtractGreenFromBlueAndRed = VP8LSubtractGreenFromBlueAndRed_C;
@ -1634,6 +1639,7 @@ void VP8LDspInit(void) {
}
#endif
}
lossless_last_cpuinfo_used = VP8GetCPUInfo;
}
//------------------------------------------------------------------------------

View File

@ -259,20 +259,45 @@ static uint32_t Predictor13(uint32_t left, const uint32_t* const top) {
//------------------------------------------------------------------------------
// Subtract-Green Transform
// vtbl? are unavailable in iOS/arm64 builds.
#if !defined(__aarch64__)
// vtbl?_u8 are marked unavailable for iOS arm64 with Xcode < 6.3, use
// non-standard versions there.
#if defined(__APPLE__) && defined(__aarch64__) && \
defined(__apple_build_version__) && (__apple_build_version__< 6020037)
#define USE_VTBLQ
#endif
// 255 = byte will be zero'd
#ifdef USE_VTBLQ
// 255 = byte will be zeroed
static const uint8_t kGreenShuffle[16] = {
1, 255, 1, 255, 5, 255, 5, 255, 9, 255, 9, 255, 13, 255, 13, 255
};
static WEBP_INLINE uint8x16_t DoGreenShuffle(const uint8x16_t argb,
const uint8x16_t shuffle) {
return vcombine_u8(vtbl1q_u8(argb, vget_low_u8(shuffle)),
vtbl1q_u8(argb, vget_high_u8(shuffle)));
}
#else // !USE_VTBLQ
// 255 = byte will be zeroed
static const uint8_t kGreenShuffle[8] = { 1, 255, 1, 255, 5, 255, 5, 255 };
static WEBP_INLINE uint8x16_t DoGreenShuffle(const uint8x16_t argb,
const uint8x8_t shuffle) {
return vcombine_u8(vtbl1_u8(vget_low_u8(argb), shuffle),
vtbl1_u8(vget_high_u8(argb), shuffle));
}
#endif // USE_VTBLQ
static void SubtractGreenFromBlueAndRed(uint32_t* argb_data, int num_pixels) {
const uint32_t* const end = argb_data + (num_pixels & ~3);
#ifdef USE_VTBLQ
const uint8x16_t shuffle = vld1q_u8(kGreenShuffle);
#else
const uint8x8_t shuffle = vld1_u8(kGreenShuffle);
#endif
for (; argb_data < end; argb_data += 4) {
const uint8x16_t argb = vld1q_u8((uint8_t*)argb_data);
const uint8x16_t greens =
vcombine_u8(vtbl1_u8(vget_low_u8(argb), shuffle),
vtbl1_u8(vget_high_u8(argb), shuffle));
const uint8x16_t greens = DoGreenShuffle(argb, shuffle);
vst1q_u8((uint8_t*)argb_data, vsubq_u8(argb, greens));
}
// fallthrough and finish off with plain-C
@ -281,19 +306,21 @@ static void SubtractGreenFromBlueAndRed(uint32_t* argb_data, int num_pixels) {
static void AddGreenToBlueAndRed(uint32_t* argb_data, int num_pixels) {
const uint32_t* const end = argb_data + (num_pixels & ~3);
#ifdef USE_VTBLQ
const uint8x16_t shuffle = vld1q_u8(kGreenShuffle);
#else
const uint8x8_t shuffle = vld1_u8(kGreenShuffle);
#endif
for (; argb_data < end; argb_data += 4) {
const uint8x16_t argb = vld1q_u8((uint8_t*)argb_data);
const uint8x16_t greens =
vcombine_u8(vtbl1_u8(vget_low_u8(argb), shuffle),
vtbl1_u8(vget_high_u8(argb), shuffle));
const uint8x16_t greens = DoGreenShuffle(argb, shuffle);
vst1q_u8((uint8_t*)argb_data, vaddq_u8(argb, greens));
}
// fallthrough and finish off with plain-C
VP8LAddGreenToBlueAndRed_C(argb_data, num_pixels & 3);
}
#endif // !__aarch64__
#undef USE_VTBLQ
#endif // USE_INTRINSICS
@ -320,11 +347,9 @@ void VP8LDspInitNEON(void) {
VP8LPredictors[12] = Predictor12;
VP8LPredictors[13] = Predictor13;
#if !defined(__aarch64__)
VP8LSubtractGreenFromBlueAndRed = SubtractGreenFromBlueAndRed;
VP8LAddGreenToBlueAndRed = AddGreenToBlueAndRed;
#endif
#endif
#endif // WEBP_USE_NEON
}

View File

@ -189,7 +189,12 @@ const WebPYUV444Converter WebPYUV444Converters[MODE_LAST] = {
extern void WebPInitUpsamplersSSE2(void);
extern void WebPInitUpsamplersNEON(void);
static volatile VP8CPUInfo upsampling_last_cpuinfo_used2 =
(VP8CPUInfo)&upsampling_last_cpuinfo_used2;
void WebPInitUpsamplers(void) {
if (upsampling_last_cpuinfo_used2 == VP8GetCPUInfo) return;
#ifdef FANCY_UPSAMPLING
WebPUpsamplers[MODE_RGB] = UpsampleRgbLinePair;
WebPUpsamplers[MODE_RGBA] = UpsampleRgbaLinePair;
@ -217,6 +222,7 @@ void WebPInitUpsamplers(void) {
#endif
}
#endif // FANCY_UPSAMPLING
upsampling_last_cpuinfo_used2 = VP8GetCPUInfo;
}
//------------------------------------------------------------------------------

View File

@ -123,7 +123,12 @@ WebPSamplerRowFunc WebPSamplers[MODE_LAST];
extern void WebPInitSamplersSSE2(void);
extern void WebPInitSamplersMIPS32(void);
static volatile VP8CPUInfo yuv_last_cpuinfo_used =
(VP8CPUInfo)&yuv_last_cpuinfo_used;
void WebPInitSamplers(void) {
if (yuv_last_cpuinfo_used == VP8GetCPUInfo) return;
WebPSamplers[MODE_RGB] = YuvToRgbRow;
WebPSamplers[MODE_RGBA] = YuvToRgbaRow;
WebPSamplers[MODE_BGR] = YuvToBgrRow;
@ -149,6 +154,7 @@ void WebPInitSamplers(void) {
}
#endif // WEBP_USE_MIPS32
}
yuv_last_cpuinfo_used = VP8GetCPUInfo;
}
//-----------------------------------------------------------------------------

View File

@ -30,7 +30,7 @@ extern "C" {
// version numbers
#define ENC_MAJ_VERSION 0
#define ENC_MIN_VERSION 4
#define ENC_REV_VERSION 2
#define ENC_REV_VERSION 3
// intra prediction modes
enum { B_DC_PRED = 0, // 4x4 modes

View File

@ -1081,6 +1081,7 @@ WebPEncodingError VP8LEncodeStream(const WebPConfig* const config,
int y;
err = AllocateTransformBuffer(enc, width, height);
if (err != VP8_ENC_OK) goto Error;
assert(enc->argb_ != NULL);
for (y = 0; y < height; ++y) {
memcpy(enc->argb_ + y * width,
picture->argb + y * picture->argb_stride,

View File

@ -326,7 +326,7 @@ int WebPEncode(const WebPConfig* config, WebPPicture* pic) {
if (!config->lossless) {
VP8Encoder* enc = NULL;
if (pic->y == NULL || pic->u == NULL || pic->v == NULL) {
if (pic->use_argb || pic->y == NULL || pic->u == NULL || pic->v == NULL) {
// Make sure we have YUVA samples.
if (config->preprocessing & 4) {
#if WEBP_ENCODER_ABI_VERSION > 0x0204

View File

@ -419,7 +419,9 @@ WEBP_EXTERN(int) WebPPictureView(const WebPPicture* src,
WEBP_EXTERN(int) WebPPictureIsView(const WebPPicture* picture);
// Rescale a picture to new dimension width x height.
// Now gamma correction is applied.
// If either 'width' or 'height' (but not both) is 0 the corresponding
// dimension will be calculated preserving the aspect ratio.
// No gamma correction is applied.
// Returns false in case of error (invalid parameter or insufficient memory).
WEBP_EXTERN(int) WebPPictureRescale(WebPPicture* pic, int width, int height);

View File

@ -18,10 +18,11 @@
#ifndef _MSC_VER
#include <inttypes.h>
#ifdef __STRICT_ANSI__
#define WEBP_INLINE
#else /* __STRICT_ANSI__ */
#if defined(__cplusplus) || !defined(__STRICT_ANSI__) || \
(defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L)
#define WEBP_INLINE inline
#else
#define WEBP_INLINE
#endif
#else
typedef signed char int8_t;