update ChangeLog

Bug: webp:541
Change-Id: Icd9d0bf1a011263108e899b561f7c8e93dc93e2f

.mailmap

@@ -11,3 +11,5 @@ Vikas Arora <vikasa@google.com>
 Tamar Levy <tamar.levy@intel.com>
 <qrczak@google.com> <qrczak>
 Hui Su <huisu@google.com>
+James Zern <jzern@google.com>
+Roberto Alanis <alanisbaez@google.com>

.pylintrc

@@ -0,0 +1,441 @@
+# This Pylint rcfile contains a best-effort configuration to uphold the
+# best-practices and style described in the Google Python style guide:
+#   https://google.github.io/styleguide/pyguide.html
+#
+# Its canonical open-source location is:
+#   https://google.github.io/styleguide/pylintrc
+
+[MASTER]
+
+# Files or directories to be skipped. They should be base names, not paths.
+ignore=third_party
+
+# Files or directories matching the regex patterns are skipped. The regex
+# matches against base names, not paths.
+ignore-patterns=
+
+# Pickle collected data for later comparisons.
+persistent=no
+
+# List of plugins (as comma separated values of python modules names) to load,
+# usually to register additional checkers.
+load-plugins=
+
+# Use multiple processes to speed up Pylint.
+jobs=4
+
+# Allow loading of arbitrary C extensions. Extensions are imported into the
+# active Python interpreter and may run arbitrary code.
+unsafe-load-any-extension=no
+
+
+[MESSAGES CONTROL]
+
+# Only show warnings with the listed confidence levels. Leave empty to show
+# all. Valid levels: HIGH, INFERENCE, INFERENCE_FAILURE, UNDEFINED
+confidence=
+
+# Enable the message, report, category or checker with the given id(s). You can
+# either give multiple identifier separated by comma (,) or put this option
+# multiple time (only on the command line, not in the configuration file where
+# it should appear only once). See also the "--disable" option for examples.
+#enable=
+
+# Disable the message, report, category or checker with the given id(s). You
+# can either give multiple identifiers separated by comma (,) or put this
+# option multiple times (only on the command line, not in the configuration
+# file where it should appear only once).You can also use "--disable=all" to
+# disable everything first and then reenable specific checks. For example, if
+# you want to run only the similarities checker, you can use "--disable=all
+# --enable=similarities". If you want to run only the classes checker, but have
+# no Warning level messages displayed, use"--disable=all --enable=classes
+# --disable=W"
+disable=abstract-method,
+        apply-builtin,
+        arguments-differ,
+        attribute-defined-outside-init,
+        backtick,
+        bad-option-value,
+        basestring-builtin,
+        buffer-builtin,
+        c-extension-no-member,
+        consider-using-enumerate,
+        cmp-builtin,
+        cmp-method,
+        coerce-builtin,
+        coerce-method,
+        delslice-method,
+        div-method,
+        duplicate-code,
+        eq-without-hash,
+        execfile-builtin,
+        file-builtin,
+        filter-builtin-not-iterating,
+        fixme,
+        getslice-method,
+        global-statement,
+        hex-method,
+        idiv-method,
+        implicit-str-concat-in-sequence,
+        import-error,
+        import-self,
+        import-star-module-level,
+        inconsistent-return-statements,
+        input-builtin,
+        intern-builtin,
+        invalid-str-codec,
+        locally-disabled,
+        long-builtin,
+        long-suffix,
+        map-builtin-not-iterating,
+        misplaced-comparison-constant,
+        missing-function-docstring,
+        metaclass-assignment,
+        next-method-called,
+        next-method-defined,
+        no-absolute-import,
+        no-else-break,
+        no-else-continue,
+        no-else-raise,
+        no-else-return,
+        no-init,  # added
+        no-member,
+        no-name-in-module,
+        no-self-use,
+        nonzero-method,
+        oct-method,
+        old-division,
+        old-ne-operator,
+        old-octal-literal,
+        old-raise-syntax,
+        parameter-unpacking,
+        print-statement,
+        raising-string,
+        range-builtin-not-iterating,
+        raw_input-builtin,
+        rdiv-method,
+        reduce-builtin,
+        relative-import,
+        reload-builtin,
+        round-builtin,
+        setslice-method,
+        signature-differs,
+        standarderror-builtin,
+        suppressed-message,
+        sys-max-int,
+        too-few-public-methods,
+        too-many-ancestors,
+        too-many-arguments,
+        too-many-boolean-expressions,
+        too-many-branches,
+        too-many-instance-attributes,
+        too-many-locals,
+        too-many-nested-blocks,
+        too-many-public-methods,
+        too-many-return-statements,
+        too-many-statements,
+        trailing-newlines,
+        unichr-builtin,
+        unicode-builtin,
+        unnecessary-pass,
+        unpacking-in-except,
+        useless-else-on-loop,
+        useless-object-inheritance,
+        useless-suppression,
+        using-cmp-argument,
+        wrong-import-order,
+        xrange-builtin,
+        zip-builtin-not-iterating,
+
+
+[REPORTS]
+
+# Set the output format. Available formats are text, parseable, colorized, msvs
+# (visual studio) and html. You can also give a reporter class, eg
+# mypackage.mymodule.MyReporterClass.
+output-format=text
+
+# Put messages in a separate file for each module / package specified on the
+# command line instead of printing them on stdout. Reports (if any) will be
+# written in a file name "pylint_global.[txt|html]". This option is deprecated
+# and it will be removed in Pylint 2.0.
+files-output=no
+
+# Tells whether to display a full report or only the messages
+reports=no
+
+# Python expression which should return a note less than 10 (10 is the highest
+# note). You have access to the variables errors warning, statement which
+# respectively contain the number of errors / warnings messages and the total
+# number of statements analyzed. This is used by the global evaluation report
+# (RP0004).
+evaluation=10.0 - ((float(5 * error + warning + refactor + convention) / statement) * 10)
+
+# Template used to display messages. This is a python new-style format string
+# used to format the message information. See doc for all details
+#msg-template=
+
+
+[BASIC]
+
+# Good variable names which should always be accepted, separated by a comma
+good-names=main,_,PRESUBMIT
+
+# Bad variable names which should always be refused, separated by a comma
+bad-names=
+
+# Colon-delimited sets of names that determine each other's naming style when
+# the name regexes allow several styles.
+name-group=
+
+# Include a hint for the correct naming format with invalid-name
+include-naming-hint=no
+
+# List of decorators that produce properties, such as abc.abstractproperty. Add
+# to this list to register other decorators that produce valid properties.
+property-classes=abc.abstractproperty,cached_property.cached_property,cached_property.threaded_cached_property,cached_property.cached_property_with_ttl,cached_property.threaded_cached_property_with_ttl
+
+# Regular expression matching correct function names
+function-rgx=^(?:(?P<exempt>setUp|tearDown|setUpModule|tearDownModule)|(?P<camel_case>_?[A-Z][a-zA-Z0-9]*)|(?P<snake_case>_?[a-z][a-z0-9_]*))$
+
+# Regular expression matching correct variable names
+variable-rgx=^[a-z][a-z0-9_]*$
+
+# Regular expression matching correct constant names
+const-rgx=^(_?[A-Z][A-Z0-9_]*|__[a-z0-9_]+__|_?[a-z][a-z0-9_]*)$
+
+# Regular expression matching correct attribute names
+attr-rgx=^_{0,2}[a-z][a-z0-9_]*$
+
+# Regular expression matching correct argument names
+argument-rgx=^[a-z][a-z0-9_]*$
+
+# Regular expression matching correct class attribute names
+class-attribute-rgx=^(_?[A-Z][A-Z0-9_]*|__[a-z0-9_]+__|_?[a-z][a-z0-9_]*)$
+
+# Regular expression matching correct inline iteration names
+inlinevar-rgx=^[a-z][a-z0-9_]*$
+
+# Regular expression matching correct class names
+class-rgx=^_?[A-Z][a-zA-Z0-9]*$
+
+# Regular expression matching correct module names
+module-rgx=^(_?[a-z][a-z0-9_]*|__init__)$
+
+# Regular expression matching correct method names
+method-rgx=(?x)^(?:(?P<exempt>_[a-z0-9_]+__|runTest|setUp|tearDown|setUpTestCase|tearDownTestCase|setupSelf|tearDownClass|setUpClass|(test|assert)_*[A-Z0-9][a-zA-Z0-9_]*|next)|(?P<camel_case>_{0,2}[A-Z][a-zA-Z0-9_]*)|(?P<snake_case>_{0,2}[a-z][a-z0-9_]*))$
+
+# Regular expression which should only match function or class names that do
+# not require a docstring.
+no-docstring-rgx=(__.*__|main|test.*|.*test|.*Test)$
+
+# Minimum line length for functions/classes that require docstrings, shorter
+# ones are exempt.
+docstring-min-length=10
+
+
+[TYPECHECK]
+
+# List of decorators that produce context managers, such as
+# contextlib.contextmanager. Add to this list to register other decorators that
+# produce valid context managers.
+contextmanager-decorators=contextlib.contextmanager,contextlib2.contextmanager
+
+# Tells whether missing members accessed in mixin class should be ignored. A
+# mixin class is detected if its name ends with "mixin" (case insensitive).
+ignore-mixin-members=yes
+
+# List of module names for which member attributes should not be checked
+# (useful for modules/projects where namespaces are manipulated during runtime
+# and thus existing member attributes cannot be deduced by static analysis. It
+# supports qualified module names, as well as Unix pattern matching.
+ignored-modules=
+
+# List of class names for which member attributes should not be checked (useful
+# for classes with dynamically set attributes). This supports the use of
+# qualified names.
+ignored-classes=optparse.Values,thread._local,_thread._local
+
+# List of members which are set dynamically and missed by pylint inference
+# system, and so shouldn't trigger E1101 when accessed. Python regular
+# expressions are accepted.
+generated-members=
+
+
+[FORMAT]
+
+# Maximum number of characters on a single line.
+max-line-length=80
+
+# TODO(https://github.com/PyCQA/pylint/issues/3352): Direct pylint to exempt
+# lines made too long by directives to pytype.
+
+# Regexp for a line that is allowed to be longer than the limit.
+ignore-long-lines=(?x)(
+  ^\s*(\#\ )?<?https?://\S+>?$|
+  ^\s*(from\s+\S+\s+)?import\s+.+$)
+
+# Allow the body of an if to be on the same line as the test if there is no
+# else.
+single-line-if-stmt=yes
+
+# List of optional constructs for which whitespace checking is disabled. `dict-
+# separator` is used to allow tabulation in dicts, etc.: {1  : 1,\n222: 2}.
+# `trailing-comma` allows a space between comma and closing bracket: (a, ).
+# `empty-line` allows space-only lines.
+no-space-check=
+
+# Maximum number of lines in a module
+max-module-lines=99999
+
+# String used as indentation unit.  The internal Google style guide mandates 2
+# spaces.  Google's externaly-published style guide says 4, consistent with
+# PEP 8.  Here, we use 2 spaces, for conformity with many open-sourced Google
+# projects (like TensorFlow).
+indent-string='  '
+
+# Number of spaces of indent required inside a hanging  or continued line.
+indent-after-paren=4
+
+# Expected format of line ending, e.g. empty (any line ending), LF or CRLF.
+expected-line-ending-format=
+
+
+[MISCELLANEOUS]
+
+# List of note tags to take in consideration, separated by a comma.
+notes=TODO
+
+
+[STRING]
+
+# This flag controls whether inconsistent-quotes generates a warning when the
+# character used as a quote delimiter is used inconsistently within a module.
+check-quote-consistency=yes
+
+
+[VARIABLES]
+
+# Tells whether we should check for unused import in __init__ files.
+init-import=no
+
+# A regular expression matching the name of dummy variables (i.e. expectedly
+# not used).
+dummy-variables-rgx=^\*{0,2}(_$|unused_|dummy_)
+
+# List of additional names supposed to be defined in builtins. Remember that
+# you should avoid to define new builtins when possible.
+additional-builtins=
+
+# List of strings which can identify a callback function by name. A callback
+# name must start or end with one of those strings.
+callbacks=cb_,_cb
+
+# List of qualified module names which can have objects that can redefine
+# builtins.
+redefining-builtins-modules=six,six.moves,past.builtins,future.builtins,functools
+
+
+[LOGGING]
+
+# Logging modules to check that the string format arguments are in logging
+# function parameter format
+logging-modules=logging,absl.logging,tensorflow.io.logging
+
+
+[SIMILARITIES]
+
+# Minimum lines number of a similarity.
+min-similarity-lines=4
+
+# Ignore comments when computing similarities.
+ignore-comments=yes
+
+# Ignore docstrings when computing similarities.
+ignore-docstrings=yes
+
+# Ignore imports when computing similarities.
+ignore-imports=no
+
+
+[SPELLING]
+
+# Spelling dictionary name. Available dictionaries: none. To make it working
+# install python-enchant package.
+spelling-dict=
+
+# List of comma separated words that should not be checked.
+spelling-ignore-words=
+
+# A path to a file that contains private dictionary; one word per line.
+spelling-private-dict-file=
+
+# Tells whether to store unknown words to indicated private dictionary in
+# --spelling-private-dict-file option instead of raising a message.
+spelling-store-unknown-words=no
+
+
+[IMPORTS]
+
+# Deprecated modules which should not be used, separated by a comma
+deprecated-modules=regsub,
+                   TERMIOS,
+                   Bastion,
+                   rexec,
+                   sets
+
+# Create a graph of every (i.e. internal and external) dependencies in the
+# given file (report RP0402 must not be disabled)
+import-graph=
+
+# Create a graph of external dependencies in the given file (report RP0402 must
+# not be disabled)
+ext-import-graph=
+
+# Create a graph of internal dependencies in the given file (report RP0402 must
+# not be disabled)
+int-import-graph=
+
+# Force import order to recognize a module as part of the standard
+# compatibility libraries.
+known-standard-library=
+
+# Force import order to recognize a module as part of a third party library.
+known-third-party=enchant, absl
+
+# Analyse import fallback blocks. This can be used to support both Python 2 and
+# 3 compatible code, which means that the block might have code that exists
+# only in one or another interpreter, leading to false positives when analysed.
+analyse-fallback-blocks=no
+
+
+[CLASSES]
+
+# List of method names used to declare (i.e. assign) instance attributes.
+defining-attr-methods=__init__,
+                      __new__,
+                      setUp
+
+# List of member names, which should be excluded from the protected access
+# warning.
+exclude-protected=_asdict,
+                  _fields,
+                  _replace,
+                  _source,
+                  _make
+
+# List of valid names for the first argument in a class method.
+valid-classmethod-first-arg=cls,
+                            class_
+
+# List of valid names for the first argument in a metaclass class method.
+valid-metaclass-classmethod-first-arg=mcs
+
+
+[EXCEPTIONS]
+
+# Exceptions that will emit a warning when being caught. Defaults to
+# "Exception"
+overgeneral-exceptions=StandardError,
+                       Exception,
+                       BaseException

.style.yapf

@@ -0,0 +1,2 @@
+[style]
+based_on_style = chromium

AUTHORS

@@ -4,8 +4,11 @@ Contributors:
 - Charles Munger (clm at google dot com)
 - Cheng Yi (cyi at google dot com)
 - Christian Duvivier (cduvivier at google dot com)
+- Christopher Degawa (ccom at randomderp dot com)
+- Clement Courbet (courbet at google dot com)
 - Djordje Pesut (djordje dot pesut at imgtec dot com)
 - Hui Su (huisu at google dot com)
+- Ilya Kurdyukov (jpegqs at gmail dot com)
 - Ingvar Stepanyan (rreverser at google dot com)
 - James Zern (jzern at google dot com)
 - Jan Engelhardt (jengelh at medozas dot de)
@@ -29,6 +32,7 @@ Contributors:
 - Pascal Massimino (pascal dot massimino at gmail dot com)
 - Paweł Hajdan, Jr (phajdan dot jr at chromium dot org)
 - Pierre Joye (pierre dot php at gmail dot com)
+- Roberto Alanis (alanisbaez at google dot com)
 - Sam Clegg (sbc at chromium dot org)
 - Scott Hancher (seh at google dot com)
 - Scott LaVarnway (slavarnway at google dot com)
@@ -45,3 +49,4 @@ Contributors:
 - Wan-Teh Chang (wtc at google dot com)
 - Yang Zhang (yang dot zhang at arm dot com)
 - Yannis Guyon (yguyon at google dot com)
+- Zhi An Ng (zhin at chromium dot org)

Android.mk

@@ -74,6 +74,7 @@ dsp_dec_srcs := \
     src/dsp/lossless_msa.c \
     src/dsp/lossless_neon.$(NEON) \
     src/dsp/lossless_sse2.c \
+    src/dsp/lossless_sse41.c \
     src/dsp/rescaler.c \
     src/dsp/rescaler_mips32.c \
     src/dsp/rescaler_mips_dsp_r2.c \

CMakeLists.txt

@@ -1,13 +1,16 @@
-##
-##  Copyright (c) 2020 Google LLC.
-##
-##  Use of this source code is governed by a BSD-style license
-##  that can be found in the LICENSE file in the root of the source
-##  tree. An additional intellectual property rights grant can be found
-##  in the file PATENTS.  All contributing project authors may
-##  be found in the AUTHORS file in the root of the source tree.
+#  Copyright (c) 2020 Google LLC.
+#
+#  Use of this source code is governed by a BSD-style license
+#  that can be found in the LICENSE file in the root of the source
+#  tree. An additional intellectual property rights grant can be found
+#  in the file PATENTS.  All contributing project authors may
+#  be found in the AUTHORS file in the root of the source tree.
 
-cmake_minimum_required(VERSION 3.5)
+cmake_minimum_required(VERSION 3.7)
+
+if(POLICY CMP0072)
+  cmake_policy(SET CMP0072 NEW)
+endif()
 
 project(WebP C)
 
@@ -26,9 +29,11 @@ option(WEBP_BUILD_GIF2WEBP "Build the gif2webp conversion tool." ON)
 option(WEBP_BUILD_IMG2WEBP "Build the img2webp animation tool." ON)
 option(WEBP_BUILD_VWEBP "Build the vwebp viewer tool." ON)
 option(WEBP_BUILD_WEBPINFO "Build the webpinfo command line tool." ON)
+option(WEBP_BUILD_LIBWEBPMUX "Build the libwebpmux library." ON)
 option(WEBP_BUILD_WEBPMUX "Build the webpmux command line tool." ON)
 option(WEBP_BUILD_EXTRAS "Build extras." ON)
 option(WEBP_BUILD_WEBP_JS "Emscripten build of webp.js." OFF)
+option(WEBP_USE_THREAD "Enable threading support" ON)
 option(WEBP_NEAR_LOSSLESS "Enable near-lossless encoding" ON)
 option(WEBP_ENABLE_SWAP_16BIT_CSP "Enable byte swap for 16 bit colorspaces."
        OFF)
@@ -50,6 +55,7 @@ if(WEBP_BUILD_WEBP_JS)
   set(WEBP_BUILD_WEBPINFO OFF)
   set(WEBP_BUILD_WEBPMUX OFF)
   set(WEBP_BUILD_EXTRAS OFF)
+  set(WEBP_USE_THREAD OFF)
 
   if(WEBP_ENABLE_SIMD)
     message("wasm2js does not support SIMD, disabling webp.js generation.")
@@ -277,10 +283,12 @@ if(XCODE)
   libwebp_add_stub_file(webp)
 endif()
 target_link_libraries(webp ${WEBP_DEP_LIBRARIES})
-target_include_directories(webp
-                           PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}
-                                   ${CMAKE_CURRENT_BINARY_DIR}
-                           PUBLIC $<INSTALL_INTERFACE:include>)
+target_include_directories(
+  webp
+  PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}
+          ${CMAKE_CURRENT_BINARY_DIR}
+  PUBLIC $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/src>
+         $<INSTALL_INTERFACE:include>)
 set_target_properties(
   webp
   PROPERTIES PUBLIC_HEADER "${CMAKE_CURRENT_SOURCE_DIR}/src/webp/decode.h;\
@@ -315,10 +323,10 @@ ${CMAKE_CURRENT_SOURCE_DIR}/src/webp/types.h")
 configure_pkg_config("src/demux/libwebpdemux.pc")
 
 # Set the version numbers.
-function(parse_version FILE NAME VAR)
+macro(set_version FILE TARGET_NAME NAME_IN_MAKEFILE)
   file(READ ${CMAKE_CURRENT_SOURCE_DIR}/src/${FILE} SOURCE_FILE)
   string(REGEX MATCH
-               "${NAME}_la_LDFLAGS[^\n]* -version-info [0-9:]+"
+               "${NAME_IN_MAKEFILE}_la_LDFLAGS[^\n]* -version-info [0-9:]+"
                TMP
                ${SOURCE_FILE})
   string(REGEX MATCH
@@ -327,29 +335,35 @@ function(parse_version FILE NAME VAR)
                ${TMP})
   string(REGEX
          REPLACE ":"
-                 "."
-                 VERSION
+                 " "
+                 LT_VERSION
                  ${TMP})
-  set(${VAR} "${VERSION}" PARENT_SCOPE)
-endfunction()
-parse_version(Makefile.am webp WEBP_WEBP_SOVERSION)
-set_target_properties(webp
-                      PROPERTIES VERSION
-                                 ${PACKAGE_VERSION}
-                                 SOVERSION
-                                 ${WEBP_WEBP_SOVERSION})
-parse_version(Makefile.am webpdecoder WEBP_DECODER_SOVERSION)
-set_target_properties(webpdecoder
-                      PROPERTIES VERSION
-                                 ${PACKAGE_VERSION}
-                                 SOVERSION
-                                 ${WEBP_DECODER_SOVERSION})
-parse_version(demux/Makefile.am webpdemux WEBP_DEMUX_SOVERSION)
-set_target_properties(webpdemux
-                      PROPERTIES VERSION
-                                 ${PACKAGE_VERSION}
-                                 SOVERSION
-                                 ${WEBP_DEMUX_SOVERSION})
+
+  # See the libtool docs for more information:
+  # https://www.gnu.org/software/libtool/manual/libtool.html#Updating-version-info
+  #
+  # c=<current>, a=<age>, r=<revision>
+  #
+  # libtool generates a .so file as .so.[c-a].a.r, while -version-info c:r:a is
+  # passed to libtool.
+  #
+  # We set FULL = [c-a].a.r and MAJOR = [c-a].
+  separate_arguments(LT_VERSION)
+  list(GET LT_VERSION 0 LT_CURRENT)
+  list(GET LT_VERSION 1 LT_REVISION)
+  list(GET LT_VERSION 2 LT_AGE)
+  math(EXPR LT_CURRENT_MINUS_AGE "${LT_CURRENT} - ${LT_AGE}")
+
+  set_target_properties(
+    ${TARGET_NAME}
+    PROPERTIES VERSION
+               ${LT_CURRENT_MINUS_AGE}.${LT_AGE}.${LT_REVISION}
+               SOVERSION
+               ${LT_CURRENT_MINUS_AGE})
+endmacro()
+set_version(Makefile.am webp webp)
+set_version(Makefile.am webpdecoder webpdecoder)
+set_version(demux/Makefile.am webpdemux webpdemux)
 file(READ ${CMAKE_CURRENT_SOURCE_DIR}/configure.ac CONFIGURE_FILE)
 string(REGEX MATCH
              "AC_INIT\\([^\n]*\\[[0-9\\.]+\\]"
@@ -377,6 +391,12 @@ foreach(I_FILE RANGE ${WEBP_SIMD_FILES_TO_INCLUDE_RANGE})
                               ${SIMD_COMPILE_FLAG})
 endforeach()
 
+if(NOT WEBP_BUILD_LIBWEBPMUX)
+  set(WEBP_BUILD_GIF2WEBP OFF)
+  set(WEBP_BUILD_IMG2WEBP OFF)
+  set(WEBP_BUILD_WEBPMUX OFF)
+endif()
+
 if(WEBP_BUILD_GIF2WEBP AND NOT GIF_FOUND)
   set(WEBP_BUILD_GIF2WEBP OFF)
 endif()
@@ -449,19 +469,14 @@ if(WEBP_BUILD_CWEBP)
   install(TARGETS cwebp RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR})
 endif()
 
-if(WEBP_BUILD_GIF2WEBP OR WEBP_BUILD_IMG2WEBP)
+if(WEBP_BUILD_LIBWEBPMUX)
   parse_makefile_am(${CMAKE_CURRENT_SOURCE_DIR}/src/mux "WEBP_MUX_SRCS" "")
   add_library(libwebpmux ${WEBP_MUX_SRCS})
   target_link_libraries(libwebpmux webp)
   target_include_directories(libwebpmux
                              PRIVATE ${CMAKE_CURRENT_BINARY_DIR}
                                      ${CMAKE_CURRENT_SOURCE_DIR})
-  parse_version(mux/Makefile.am webpmux WEBP_MUX_SOVERSION)
-  set_target_properties(libwebpmux
-                        PROPERTIES VERSION
-                                   ${PACKAGE_VERSION}
-                                   SOVERSION
-                                   ${WEBP_MUX_SOVERSION})
+  set_version(mux/Makefile.am libwebpmux webpmux)
   set_target_properties(libwebpmux
                         PROPERTIES PUBLIC_HEADER
                                    "${CMAKE_CURRENT_SOURCE_DIR}/src/webp/mux.h;\
@@ -565,9 +580,7 @@ if(WEBP_BUILD_EXTRAS)
   target_link_libraries(get_disto imagedec)
   target_include_directories(get_disto
                              PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}
-                                     ${CMAKE_CURRENT_SOURCE_DIR}/src
                                      ${CMAKE_CURRENT_BINARY_DIR}/src)
-  install(TARGETS get_disto RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR})
 
   # webp_quality
   add_executable(webp_quality ${WEBP_QUALITY_SRCS} ${WEBP_EXTRAS_SRCS})
@@ -575,7 +588,6 @@ if(WEBP_BUILD_EXTRAS)
   target_include_directories(webp_quality
                              PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}
                                      ${CMAKE_CURRENT_BINARY_DIR})
-  install(TARGETS webp_quality RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR})
 
   # vwebp_sdl
   find_package(SDL)
@@ -584,13 +596,11 @@ if(WEBP_BUILD_EXTRAS)
     target_link_libraries(vwebp_sdl ${SDL_LIBRARY} imageioutil webp)
     target_include_directories(vwebp_sdl
                                PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}
-                                       ${CMAKE_CURRENT_SOURCE_DIR}/src
                                        ${CMAKE_CURRENT_BINARY_DIR}
                                        ${CMAKE_CURRENT_BINARY_DIR}/src
                                        ${SDL_INCLUDE_DIR})
     set(WEBP_HAVE_SDL 1)
     target_compile_definitions(vwebp_sdl PUBLIC WEBP_HAVE_SDL)
-    install(TARGETS vwebp_sdl RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR})
   endif()
 endif()
 
@@ -606,7 +616,7 @@ if(WEBP_BUILD_WEBP_JS)
       webp_js
       PROPERTIES LINK_FLAGS "-s WASM=0 \
          -s EXPORTED_FUNCTIONS='[\"_WebpToSDL\"]' -s INVOKE_RUN=0 \
-         -s EXTRA_EXPORTED_RUNTIME_METHODS='[\"cwrap\"]'")
+         -s EXPORTED_RUNTIME_METHODS='[\"cwrap\"]'")
     set_target_properties(webp_js PROPERTIES OUTPUT_NAME webp)
     target_compile_definitions(webp_js PUBLIC EMSCRIPTEN WEBP_HAVE_SDL)
   endif()
@@ -619,7 +629,7 @@ if(WEBP_BUILD_WEBP_JS)
     webp_wasm
     PROPERTIES LINK_FLAGS "-s WASM=1 \
        -s EXPORTED_FUNCTIONS='[\"_WebpToSDL\"]' -s INVOKE_RUN=0 \
-       -s EXTRA_EXPORTED_RUNTIME_METHODS='[\"cwrap\"]'")
+       -s EXPORTED_RUNTIME_METHODS='[\"cwrap\"]'")
   target_compile_definitions(webp_wasm PUBLIC EMSCRIPTEN WEBP_HAVE_SDL)
 
   target_compile_definitions(webpdspdecode PUBLIC EMSCRIPTEN)
@@ -637,7 +647,6 @@ if(WEBP_BUILD_ANIM_UTILS)
                         imageenc
                         imageioutil
                         webp
-                        libwebpmux
                         webpdemux
                         ${WEBP_DEP_GIF_LIBRARIES})
   target_include_directories(anim_diff PRIVATE ${CMAKE_CURRENT_BINARY_DIR}/src)
@@ -653,7 +662,6 @@ if(WEBP_BUILD_ANIM_UTILS)
                         imageenc
                         imageioutil
                         webp
-                        libwebpmux
                         webpdemux
                         ${WEBP_DEP_GIF_LIBRARIES})
   target_include_directories(anim_dump PRIVATE ${CMAKE_CURRENT_BINARY_DIR}/src)

CONTRIBUTING.md

@@ -0,0 +1,29 @@
+# How to Contribute
+
+We'd love to accept your patches and contributions to this project. There are
+just a few small guidelines you need to follow.
+
+## Contributor License Agreement
+
+Contributions to this project must be accompanied by a Contributor License
+Agreement. You (or your employer) retain the copyright to your contribution;
+this simply gives us permission to use and redistribute your contributions as
+part of the project. Head over to <https://cla.developers.google.com/> to see
+your current agreements on file or to sign a new one.
+
+You generally only need to submit a CLA once, so if you've already submitted one
+(even if it was for a different project), you probably don't need to do it
+again.
+
+## Code reviews
+
+All submissions, including submissions by project members, require review. We
+use a [Gerrit](https://www.gerritcodereview.com) instance hosted at
+https://chromium-review.googlesource.com for this purpose. See the
+[WebM Project page](https://www.webmproject.org/code/contribute/submitting-patches/)
+for additional details.
+
+## Community Guidelines
+
+This project follows
+[Google's Open Source Community Guidelines](https://opensource.google.com/conduct/).

ChangeLog

@@ -1,10 +1,150 @@
+7031946a update NEWS
+973390b6 bump version to 1.2.2
+abd6664f update AUTHORS
+5b7e7930 Merge "add missing USE_{MSA,NEON} checks in headers" into main
+02ca04c3 add missing USE_{MSA,NEON} checks in headers
+e94716e2 xcframeworkbuild.sh: place headers in a subdir
+c846efd8 patch-check: commit subject length check
+b6f756e8 update http links
+8f5cb4c1 update rfc links
+8ea81561 change VP8LPredictorFunc signature to avoid reading 'left'
+6b1d18c3 webpmux: fix the -bgcolor description
+3368d876 Merge "webpmux: add "-set bgcolor A,R,G,B"" into main
+f213abf6 webpinfo: print the number of warnings
+50c97c30 webpmux: add "-set bgcolor A,R,G,B"
+2c206aaf Remove CMakeLists.txt check in compile.sh
+96e3dfef Merge "infra/common.sh: add shard_should_run()" into main
+0e0f74b7 infra/common.sh: add shard_should_run()
+35b7436a Jenkins scripts port: update shell function comments
+21d24b4c webp-container-spec.txt: remove 'experimental' markers
+cdcf8902 Merge "Port Jenkins script: compile" into main
+dc683cde Jenkins scripts port: static analysis
+0858494e Port Jenkins script: compile
+c2cf6a93 Jenkins scripts port: android compilation
+df0e808f presubmit: Add pylint-2.7 and .pylintrc
+676c57db patch-check: shfmt
+7bb7f747 patch-check: Add shellcheck
+abcd1797 Reformat docstrings and imports
+edaf0895 Port Jenkins scripts: compile js
+b9622063 Set CheckPatchFormatted flags to fail on diffs
+e23cd548 dsp.h: enable NEON w/VS2019+ ARM64 targets
+3875c7de CMakeLists.txt: set minimum version to 3.7
+1a8f0d45 Have a hard-coded value for memset in TrellisQuantizeBlock.
+93480160 Speed up TrellisQuantizeBlock
+45eaacc9 Convert deprecated uint32 to uint32_t.
+42592af8 webp,cmake: Remove unnecessary include dirs
+e298e05f Add patch-check steps in PRESUBMIT.py
+29148919 Merge tag 'v1.2.1'
+9ce5843d update ChangeLog (tag: v1.2.1, origin/1.2.1)
+d9191588 fuzzer/*: normalize src/ includes
+c5bc3624 fuzzer/*: normalize src/ includes
+53b6f762 fix indent
+d2caaba4 fix indent
+731246ba update ChangeLog (tag: v1.2.1-rc2)
+d250f01d dsp/*: use WEBP_HAVE_* to determine Init availability
+1fe31625 dsp/*: use WEBP_HAVE_* to determine Init availability
+3a4d3ecd update NEWS
+b2bc8093 bump version to 1.2.1
+e542fc7a update AUTHORS
+e0241154 Merge "libwebp/CMake: Add <BUILD_INTERFACE> to webp incl" into main
+edea6444 libwebp/CMake: Add <BUILD_INTERFACE> to webp incl
+ece18e55 dsp.h: respect --disable-sse2/sse4.1/neon
+a89a3230 wicdec: support alpha from WebP WIC decoder
+26f4aa01 Merge "alpha_processing: fix visual studio warnings" into main
+8f594663 alpha_processing: fix visual studio warnings
+46d844e6 Merge "cpu.cmake: fix compiler flag detection w/3.17.0+" into main
+298d26ea Merge changes I593adf92,If20675e7,Ifac68eac into main
+a1e5dae0 alpha_processing*: use WEBP_RESTRICT qualifier
+327ef24f cpu.cmake: fix compiler flag detection w/3.17.0+
+f70819de configure: enable libwebpmux by default
+dc7e2b42 configure: add informational notices when disabling binaries
+9df23ddd configure: move lib flag checks before binaries
+a2e18f10 Merge "WebPConfig.config.in: correct WEBP_INCLUDE_DIRS" into main
+e1a8d4f3 Merge "bit_reader_inl_utils: uniformly apply WEBP_RESTRICT" into main
+4de35f43 rescaler.c: fix alignment
+0f13eec7 bit_reader_inl_utils: uniformly apply WEBP_RESTRICT
+277d3074 Fix size_t overflow in  WebPRescalerInit
+97adbba5 WebPConfig.config.in: correct WEBP_INCLUDE_DIRS
+b60d4603 advanced_api_fuzzer: add extreme config value coverage
+72fe52f6 anim_encode.c,cosmetics: normalize indent
+116d235c anim_encode: Fix encoded_frames_[] overflow
+6f445b3e CMake: set CMP0072 to NEW
+b1cf887f define WEBP_RESTRICT for MSVC
+3e265136 Add WEBP_RESTRICT & use it in VP8BitReader
+f6d29247 vp8l_dec::ProcessRows: fix int overflow in multiply
+de3b4ba8 CMake: add WEBP_BUILD_LIBWEBPMUX
+7f09d3d1 CMakeLists.txt: rm libwebpmux dep from anim_{diff,dump}
+4edea4a6 Init{RGB,YUV}Rescaler: fix a few more int overflows
+c9e26bdb rescaler_utils: set max valid scaled w/h to INT_MAX/2
+28d488e6 utils.h: add SizeOverflow()
+695bdaa2 Export/EmitRescaledRowsRGBA: fix pointer offset int overflow
+685d073e Init{RGB,YUV}Rescaler: fix int overflows in multiplication
+d38bd0dd WebPFlipBuffer: fix integer overflow
+109ff0f1 utils: allow MALLOC_LIMIT to indicate a max
+a2fce867 WebPRescalerImportRowExpand_C: promote some vals before multiply
+776983d4 AllocateBuffer: fix int multiplication overflow check
+315abbd6 Merge "Revert "Do not use a palette for one color images.""
+eae815d0 Merge changes Ica3bbf75,I82f82954
+afbca5a1 Require Emscripten 2.0.18
+3320416b CMakeLists,emscripten: use EXPORTED_RUNTIME_METHODS
+29145ed6 Update README instructions for using Emscripten
+1f579139 cosmetics: remove use of 'sanity' / 'master'
+29b6129c WebPAnimEncoderNewInternal: remove some unnecessary inits
+b60869a1 Revert "Do not use a palette for one color images."
+6fb4cddc demux: move padded size calc post unpadded validation
+05b72d42 vp8l_enc.c: normalize index types
+b6513fba Do not use a palette for one color images.
+98bbe35b Fix multi-threading with palettes.
+b1674240 Add modified Zeng's method to palette sorting.
+88c90c45 add CONTRIBUTING.md
+6a9916d7 WebPRescalerInit: add missing int64_t promotion
+b6cf52d5 WebPIoInitFromOptions: treat use_scaling as a bool
+3b12b7f4 WebPIoInitFromOptions: treat use_cropping as a bool
+595fa13f add WebPCheckCropDimensions()
+8fdaecb0 Disable cross-color when palette is used.
+8933bac2 WebPIoInitFromOptions: respect incoming bypass_filtering val
+7d416ff0 webpdec,cosmetics: match error text to function call
+ec6cfeb5 Fix typo on WebPPictureAlloc() in README
+7e58a1a2 *.cmake: add license header
+5651a6b2 cmake: fix .so versioning
+25ae67b3 xcframeworkbuild.sh: add arm64 simulator target
+5d4ee4c3 cosmetics: remove use of the term 'dummy'
+01b38ee1 faster CollectColorXXXTransforms_SSE41
+652aa344 Merge "Use BitCtz for FastSLog2Slow_C"
+0320e1e3 add the missing default BitsCtz() code
+8886f620 Use BitCtz for FastSLog2Slow_C
+fae41617 faster CombinedShannonEntropy_SSE2
+5bd2704e Introduce the BitCtz() function.
+fee64287 Merge "wicdec,icc: treat unsupported op as non-fatal"
+33ddb894 lossless_sse{2,41}: remove some unneeded includes
+b27ea852 wicdec,icc: treat unsupported op as non-fatal
+b78494a9 Merge "Fix undefined signed shift."
+e79974cd Fix undefined signed shift.
+a8853394 SSE4.1 versions of BGRA to RGB/BGR color-space conversions
+a09a6472 SSE4.1 version of TransformColorInverse
+401da22b Merge "pngdec: check version before using png_get_chunk_malloc_max"
+26907822 pngdec: check version before using png_get_chunk_malloc_max
+06c1e72e Code cleanup
+8f0d41aa Merge changes Id135bbf4,I99e59797
+373eb170 gif2webp: don't store loop-count if there's only 1 frame
+759b9d5a cmake: add WEBP_USE_THREAD option
+926ce921 cmake: don't install binaries from extras/
+9c367bc6 WebPAnimDecoderNewInternal: validate bitstream before alloc
+47f64f6e filters_sse2: import Chromium change
+cc3577e9 fuzzer/*: use src/ based include paths
+004d77ff Merge tag 'v1.2.0'
+fedac6cc update ChangeLog (tag: v1.2.0-rc3, tag: v1.2.0)
 170a8712 Fix check_c_source_compiles with pthread.
+ceddb5fc Fix check_c_source_compiles with pthread.
 85995719 disable CombinedShannonEntropy_SSE2 on x86
+289757fe TiffDec: enforce stricter mem/dimension limit on tiles
 8af7436f Merge "{ios,xcframework}build.sh: make min version(s) more visible" into 1.2.0
 e56c3c5b pngdec: raise memory limit if needed
+8696147d pngdec: raise memory limit if needed
 13b8e9fe {ios,xcframework}build.sh: make min version(s) more visible
 a9225410 animdecoder_fuzzer: fix memory leak
 d6c2285d update gradle to 6.1.1
+8df77fb1 animdecoder_fuzzer: fix memory leak
 52ce6333 update NEWS
 28c49820 bump version to 1.2.0
 7363dff2 webp/encode.h: restore WEBP_ENCODER_ABI_VERSION to v1.1.0

Makefile.vc

@@ -215,6 +215,7 @@ DSP_DEC_OBJS = \
     $(DIROBJ)\dsp\lossless_msa.obj \
     $(DIROBJ)\dsp\lossless_neon.obj \
     $(DIROBJ)\dsp\lossless_sse2.obj \
+    $(DIROBJ)\dsp\lossless_sse41.obj \
     $(DIROBJ)\dsp\rescaler.obj \
     $(DIROBJ)\dsp\rescaler_mips32.obj \
     $(DIROBJ)\dsp\rescaler_mips_dsp_r2.obj \

NEWS

@@ -1,3 +1,23 @@
+- 12/15/2021: version 1.2.2
+  This is a binary compatible release.
+  * webpmux: add "-set bgcolor A,R,G,B"
+  * add ARM64 NEON support for MSVC builds (#539)
+  * fix duplicate include error in Xcode when using multiple XCFrameworks in a
+    project (#542)
+  * doc updates and bug fixes (#538, #544)
+
+- 7/20/2021: version 1.2.1
+  This is a binary compatible release.
+  * minor lossless encoder improvements and x86 color conversion speed up
+  * add ARM64 simulator support to xcframeworkbuild.sh (#510)
+  * further security related hardening in libwebp & examples
+    (issues: #497, #508, #518)
+    (chromium: #1196480, #1196773, #1196775, #1196777, #1196778, #1196850)
+    (oss-fuzz: #28658, #28978)
+  * toolchain updates and bug fixes (#498, #501, #502, #504, #505, #506, #509,
+                                     #533)
+  * use more inclusive language within the source (#507)
+
 - 12/23/2020: version 1.2.0
   * API changes:
     - libwebp:

PRESUBMIT.py

@@ -0,0 +1,214 @@
+# Copyright (c) 2021, Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#
+#   * Neither the name of Google nor the names of its contributors may
+#     be used to endorse or promote products derived from this software
+#     without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+"""Top-level presubmit script for libwebp.
+
+See https://dev.chromium.org/developers/how-tos/depottools/presubmit-scripts for
+details on the presubmit API built into depot_tools.
+"""
+
+import re
+import subprocess2
+
+USE_PYTHON3 = True
+_BASH_INDENTATION = "2"
+_GIT_COMMIT_SUBJECT_LENGTH = 65
+_INCLUDE_BASH_FILES_ONLY = [r".*\.sh$"]
+_INCLUDE_MAN_FILES_ONLY = [r"man/.+\.1$"]
+_LIBWEBP_MAX_LINE_LENGTH = 80
+
+
+def _CheckCommitSubjectLength(input_api, output_api):
+  """Ensures commit's subject length is no longer than 65 chars."""
+  name = "git-commit subject"
+  cmd = ["git", "log", "-1", "--pretty=%s"]
+  start = input_api.time.time()
+  proc = subprocess2.Popen(
+      cmd,
+      stderr=subprocess2.PIPE,
+      stdout=subprocess2.PIPE,
+      universal_newlines=True)
+
+  stdout, _ = proc.communicate()
+  duration = input_api.time.time() - start
+
+  if not re.match(r"^Revert",
+                  stdout) and (len(stdout) - 1) > _GIT_COMMIT_SUBJECT_LENGTH:
+    failure_msg = (
+        "The commit subject: %s is too long (%d chars)\n"
+        "Try to keep this to 50 or less (up to 65 is permitted for "
+        "non-reverts).\n"
+        "https://www.git-scm.com/book/en/v2/Distributed-Git-Contributing-to-a-"
+        "Project#_commit_guidelines") % (stdout, len(stdout) - 1)
+    return output_api.PresubmitError("%s\n (%4.2fs) failed\n%s" %
+                                     (name, duration, failure_msg))
+
+  return output_api.PresubmitResult("%s\n (%4.2fs) success" % (name, duration))
+
+
+def _GetFilesToSkip(input_api):
+  return list(input_api.DEFAULT_FILES_TO_SKIP) + [
+      r"swig/.*\.py$",
+      r"\.pylintrc$",
+  ]
+
+
+def _RunManCmd(input_api, output_api, man_file):
+  """man command wrapper."""
+  cmd = ["man", "--warnings", "-EUTF-8", "-l", "-Tutf8", "-Z", man_file]
+  name = "Check %s file." % man_file
+  start = input_api.time.time()
+  output, _ = subprocess2.communicate(
+      cmd, stdout=None, stderr=subprocess2.PIPE, universal_newlines=True)
+  duration = input_api.time.time() - start
+  if output[1]:
+    return output_api.PresubmitError("%s\n%s (%4.2fs) failed\n%s" %
+                                     (name, " ".join(cmd), duration, output[1]))
+  return output_api.PresubmitResult("%s\n%s (%4.2fs)\n" %
+                                    (name, " ".join(cmd), duration))
+
+
+def _RunShellCheckCmd(input_api, output_api, bash_file):
+  """shellcheck command wrapper."""
+  cmd = ["shellcheck", "-x", "-oall", "-sbash", bash_file]
+  name = "Check %s file." % bash_file
+  start = input_api.time.time()
+  output, rc = subprocess2.communicate(
+      cmd, stdout=None, stderr=subprocess2.PIPE, universal_newlines=True)
+  duration = input_api.time.time() - start
+  if rc == 0:
+    return output_api.PresubmitResult("%s\n%s (%4.2fs)\n" %
+                                      (name, " ".join(cmd), duration))
+  return output_api.PresubmitError("%s\n%s (%4.2fs) failed\n%s" %
+                                   (name, " ".join(cmd), duration, output[1]))
+
+
+def _RunShfmtCheckCmd(input_api, output_api, bash_file):
+  """shfmt command wrapper."""
+  cmd = [
+      "shfmt", "-i", _BASH_INDENTATION, "-bn", "-ci", "-sr", "-kp", "-d",
+      bash_file
+  ]
+  name = "Check %s file." % bash_file
+  start = input_api.time.time()
+  output, rc = subprocess2.communicate(
+      cmd, stdout=None, stderr=subprocess2.PIPE, universal_newlines=True)
+  duration = input_api.time.time() - start
+  if rc == 0:
+    return output_api.PresubmitResult("%s\n%s (%4.2fs)\n" %
+                                      (name, " ".join(cmd), duration))
+  return output_api.PresubmitError("%s\n%s (%4.2fs) failed\n%s" %
+                                   (name, " ".join(cmd), duration, output[1]))
+
+
+def _RunCmdOnCheckedFiles(input_api, output_api, run_cmd, files_to_check):
+  """Ensure that libwebp/ files are clean."""
+  file_filter = lambda x: input_api.FilterSourceFile(
+      x, files_to_check=files_to_check, files_to_skip=None)
+
+  affected_files = input_api.change.AffectedFiles(file_filter=file_filter)
+  results = [
+      run_cmd(input_api, output_api, f.AbsoluteLocalPath())
+      for f in affected_files
+  ]
+  return results
+
+
+def _CommonChecks(input_api, output_api):
+  """Ensures this patch does not have trailing spaces, extra EOLs,
+     or long lines.
+  """
+  results = []
+  results.extend(
+      input_api.canned_checks.CheckChangeHasNoCrAndHasOnlyOneEol(
+          input_api, output_api))
+  results.extend(
+      input_api.canned_checks.CheckChangeHasNoTabs(input_api, output_api))
+  results.extend(
+      input_api.canned_checks.CheckChangeHasNoStrayWhitespace(
+          input_api, output_api))
+  results.append(_CheckCommitSubjectLength(input_api, output_api))
+
+  source_file_filter = lambda x: input_api.FilterSourceFile(
+      x, files_to_skip=_GetFilesToSkip(input_api))
+  results.extend(
+      input_api.canned_checks.CheckLongLines(
+          input_api,
+          output_api,
+          maxlen=_LIBWEBP_MAX_LINE_LENGTH,
+          source_file_filter=source_file_filter))
+
+  results.extend(
+      input_api.canned_checks.CheckPatchFormatted(
+          input_api,
+          output_api,
+          check_clang_format=False,
+          check_python=True,
+          result_factory=output_api.PresubmitError))
+  results.extend(
+      _RunCmdOnCheckedFiles(input_api, output_api, _RunManCmd,
+                            _INCLUDE_MAN_FILES_ONLY))
+  # Run pylint.
+  results.extend(
+      input_api.canned_checks.RunPylint(
+          input_api,
+          output_api,
+          files_to_skip=_GetFilesToSkip(input_api),
+          pylintrc=".pylintrc",
+          version="2.7"))
+
+  # Binaries shellcheck and shfmt are not installed in depot_tools.
+  # Installation is needed
+  try:
+    subprocess2.communicate(["shellcheck", "--version"])
+    results.extend(
+        _RunCmdOnCheckedFiles(input_api, output_api, _RunShellCheckCmd,
+                              _INCLUDE_BASH_FILES_ONLY))
+    print("shfmt")
+    subprocess2.communicate(["shfmt", "-version"])
+    results.extend(
+        _RunCmdOnCheckedFiles(input_api, output_api, _RunShfmtCheckCmd,
+                              _INCLUDE_BASH_FILES_ONLY))
+  except OSError as os_error:
+    results.append(
+        output_api.PresubmitPromptWarning(
+            "%s\nPlease install missing binaries locally." % os_error.args[0]))
+  return results
+
+
+def CheckChangeOnUpload(input_api, output_api):
+  results = []
+  results.extend(_CommonChecks(input_api, output_api))
+  return results
+
+
+def CheckChangeOnCommit(input_api, output_api):
+  results = []
+  results.extend(_CommonChecks(input_api, output_api))
+  return results

README

@@ -4,7 +4,7 @@
           \__\__/\____/\_____/__/ ____  ___
                 / _/ /    \    \ /  _ \/ _/
                /  \_/   / /   \ \   __/  \__
-               \____/____/\_____/_____/____/v1.2.0
+               \____/____/\_____/_____/____/v1.2.2
 
 Description:
 ============
@@ -13,13 +13,13 @@ WebP codec: library to encode and decode images in WebP format. This package
 contains the library that can be used in other programs to add WebP support,
 as well as the command line tools 'cwebp' and 'dwebp'.
 
-See http://developers.google.com/speed/webp
+See https://developers.google.com/speed/webp
 
 The latest source tree is available at
 https://chromium.googlesource.com/webm/libwebp
 
 It is released under the same license as the WebM project.
-See http://www.webmproject.org/license/software/ or the
+See https://www.webmproject.org/license/software/ or the
 "COPYING" file for details. An additional intellectual
 property rights grant can be found in the file PATENTS.
 
@@ -436,8 +436,8 @@ Prerequisites:
 
 2) (Optional) qcms (Quick Color Management System)
   i. Download qcms from Mozilla / Chromium:
-    http://hg.mozilla.org/mozilla-central/file/0e7639e3bdfb/gfx/qcms
-    http://src.chromium.org/viewvc/chrome/trunk/src/third_party/qcms
+    https://hg.mozilla.org/mozilla-central/file/0e7639e3bdfb/gfx/qcms
+    https://source.chromium.org/chromium/chromium/src/+/main:third_party/qcms/;drc=d4a2f8e1ed461d8fc05ed88d1ae2dc94c9773825
   ii. Build and archive the source files as libqcms.a / qcms.lib
   iii. Update makefile.unix / Makefile.vc
     a) Define WEBP_HAVE_QCMS
@@ -619,7 +619,7 @@ The encoding flow looks like:
   pic.width = width;
   pic.height = height;
   // allocated picture of dimension width x height
-  if (!WebPPictureAllocate(&pic)) {
+  if (!WebPPictureAlloc(&pic)) {
     return 0;   // memory error
   }
   // at this point, 'pic' has been initialized as a container,
@@ -786,10 +786,10 @@ Bugs:
 Please report all bugs to the issue tracker:
     https://bugs.chromium.org/p/webp
 Patches welcome! See this page to get started:
-    http://www.webmproject.org/code/contribute/submitting-patches/
+    https://www.webmproject.org/code/contribute/submitting-patches/
 
 Discuss:
 ========
 
 Email: webp-discuss@webmproject.org
-Web: http://groups.google.com/a/webmproject.org/group/webp-discuss
+Web: https://groups.google.com/a/webmproject.org/group/webp-discuss

README.mux

@@ -1,7 +1,7 @@
           __   __  ____  ____  ____  __ __  _     __ __
          /  \\/  \/  _ \/  _ \/  _ \/  \  \/ \___/_ / _\
          \       /   __/  _  \   __/      /  /  (_/  /__
-          \__\__/\_____/_____/__/  \__//_/\_____/__/___/v1.2.0
+          \__\__/\_____/_____/__/  \__//_/\_____/__/___/v1.2.2
 
 
 Description:
@@ -43,11 +43,12 @@ GET_OPTIONS:
    frame n   get nth frame
 
 SET_OPTIONS:
- Set color profile/metadata:
-   loop LOOP_COUNT   set the loop count
-   icc  file.icc     set ICC profile
-   exif file.exif    set EXIF metadata
-   xmp  file.xmp     set XMP metadata
+ Set color profile/metadata/parameters:
+   loop LOOP_COUNT            set the loop count
+   bgcolor BACKGROUND_COLOR   set the animation background color
+   icc  file.icc              set ICC profile
+   exif file.exif             set EXIF metadata
+   xmp  file.xmp              set XMP metadata
    where:    'file.icc' contains the ICC profile to be set,
              'file.exif' contains the EXIF metadata to be set
              'file.xmp' contains the XMP metadata to be set
@@ -248,10 +249,10 @@ Bugs:
 Please report all bugs to the issue tracker:
     https://bugs.chromium.org/p/webp
 Patches welcome! See this page to get started:
-    http://www.webmproject.org/code/contribute/submitting-patches/
+    https://www.webmproject.org/code/contribute/submitting-patches/
 
 Discuss:
 ========
 
 Email: webp-discuss@webmproject.org
-Web: http://groups.google.com/a/webmproject.org/group/webp-discuss
+Web: https://groups.google.com/a/webmproject.org/group/webp-discuss

README.webp_js

@@ -10,26 +10,18 @@ This file describes the compilation of libwebp into a JavaScript decoder
 using Emscripten and CMake.
 
  - install the Emscripten SDK following the procedure described at:
-   https://kripken.github.io/emscripten-site/docs/getting_started/downloads.html
+   https://emscripten.org/docs/getting_started/downloads.html#installation-instructions-using-the-emsdk-recommended
    After installation, you should have some global variable positioned to the
-   location of the SDK. In particular, $EMSCRIPTEN should point to the
+   location of the SDK. In particular, $EMSDK should point to the
    top-level directory containing Emscripten tools.
 
- - make sure the file $EMSCRIPTEN/cmake/Modules/Platform/Emscripten.cmake is
-   accessible. This is the toolchain file used by CMake to invoke Emscripten.
-   If $EMSCRIPTEN is unset search for Emscripten.cmake under $EMSDK and set
-   $EMSCRIPTEN accordingly, for example:
-   unix-like environments: export EMSCRIPTEN=$EMSDK/upstream/emscripten
-   windows: set EMSCRIPTEN=%EMSDK%\upstream\emscripten
-
  - configure the project 'WEBP_JS' with CMake using:
 
  cd webp_js && \
- cmake -DWEBP_BUILD_WEBP_JS=ON \
-       -DCMAKE_TOOLCHAIN_FILE=$EMSCRIPTEN/cmake/Modules/Platform/Emscripten.cmake \
+ emcmake cmake -DWEBP_BUILD_WEBP_JS=ON \
        ../
 
- - compile webp.js using 'make'.
+ - compile webp.js using 'emmake make'.
 
  - that's it! Upon completion, you should have the webp.js and
    webp.wasm files generated.
@@ -58,7 +50,7 @@ Web-Assembly (WASM) version:
   See webp_js/index_wasm.html for a simple demo page using the WASM version
   of the library.
 
-  You will need a fairly recent version of Emscripten (at least 1.37.8,
+  You will need a fairly recent version of Emscripten (at least 2.0.18,
   latest-upstream is recommended) and of your WASM-enabled browser to run this
   version.
 

build.gradle

@@ -141,6 +141,7 @@ model {
             include "lossless_msa.c"
             include "lossless_neon.$NEON"
             include "lossless_sse2.c"
+            include "lossless_sse41.c"
             include "rescaler.c"
             include "rescaler_mips32.c"
             include "rescaler_mips_dsp_r2.c"

cmake/WebPConfig.cmake.in

@@ -5,7 +5,7 @@ set(WEBP_VERSION ${WebP_VERSION})
 
 include ("${CMAKE_CURRENT_LIST_DIR}/@PROJECT_NAME@Targets.cmake")
 
-set(WebP_INCLUDE_DIRS "webp")
+set(WebP_INCLUDE_DIRS "@CMAKE_INSTALL_FULL_INCLUDEDIR@")
 set(WEBP_INCLUDE_DIRS ${WebP_INCLUDE_DIRS})
 set(WebP_LIBRARIES "@INSTALLED_LIBRARIES@")
 set(WEBP_LIBRARIES "${WebP_LIBRARIES}")

cmake/cpu.cmake

@@ -1,3 +1,11 @@
+#  Copyright (c) 2021 Google LLC.
+#
+#  Use of this source code is governed by a BSD-style license
+#  that can be found in the LICENSE file in the root of the source
+#  tree. An additional intellectual property rights grant can be found
+#  in the file PATENTS.  All contributing project authors may
+#  be found in the AUTHORS file in the root of the source tree.
+
 # Check for SIMD extensions.
 include(CMakePushCheckState)
 
@@ -117,6 +125,12 @@ foreach(I_SIMD RANGE ${WEBP_SIMD_FLAGS_RANGE})
       list(GET SIMD_DISABLE_FLAGS ${I_SIMD} SIMD_COMPILE_FLAG)
       include(CheckCCompilerFlag)
       if(SIMD_COMPILE_FLAG)
+        # Between 3.17.0 and 3.18.2 check_cxx_compiler_flag() sets a normal
+        # variable at parent scope while check_cxx_source_compiles() continues
+        # to set an internal cache variable, so we unset both to avoid the
+        # failure / success state persisting between checks. See
+        # https://gitlab.kitware.com/cmake/cmake/-/issues/21207.
+        unset(HAS_COMPILE_FLAG)
         unset(HAS_COMPILE_FLAG CACHE)
         check_c_compiler_flag(${SIMD_COMPILE_FLAG} HAS_COMPILE_FLAG)
         if(HAS_COMPILE_FLAG)
@@ -134,6 +148,7 @@ foreach(I_SIMD RANGE ${WEBP_SIMD_FLAGS_RANGE})
                                   "warning: argument unused during compilation:"
                                   ${COMMON_PATTERNS})
           if(NOT FLAG_${SIMD_COMPILE_FLAG})
+            unset(HAS_COMPILE_FLAG)
             unset(HAS_COMPILE_FLAG CACHE)
           endif()
         endif()

cmake/deps.cmake

@@ -1,3 +1,11 @@
+#  Copyright (c) 2021 Google LLC.
+#
+#  Use of this source code is governed by a BSD-style license
+#  that can be found in the LICENSE file in the root of the source
+#  tree. An additional intellectual property rights grant can be found
+#  in the file PATENTS.  All contributing project authors may
+#  be found in the AUTHORS file in the root of the source tree.
+
 # Generate the config.h to compile with specific intrinsics / libs.
 
 # Check for compiler options.
@@ -22,26 +30,25 @@ check_c_source_compiles("
   " HAVE_BUILTIN_BSWAP64)
 
 # Check for libraries.
-if(NOT WEBP_BUILD_WEBP_JS)
-  # Disable pThreads for WASM.
+if(WEBP_USE_THREAD)
   find_package(Threads)
-endif()
-if(Threads_FOUND)
-  # work around cmake bug on QNX (https://cmake.org/Bug/view.php?id=11333)
-  if(CMAKE_USE_PTHREADS_INIT AND NOT CMAKE_SYSTEM_NAME STREQUAL "QNX")
-    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -pthread")
+  if(Threads_FOUND)
+    # work around cmake bug on QNX (https://cmake.org/Bug/view.php?id=11333)
+    if(CMAKE_USE_PTHREADS_INIT AND NOT CMAKE_SYSTEM_NAME STREQUAL "QNX")
+      set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -pthread")
+    endif()
+    check_c_source_compiles("
+        #include <pthread.h>
+        int main (void) {
+          int attr = PTHREAD_PRIO_INHERIT;
+          return attr;
+        }
+      " FLAG_HAVE_PTHREAD_PRIO_INHERIT)
+    set(HAVE_PTHREAD_PRIO_INHERIT ${FLAG_HAVE_PTHREAD_PRIO_INHERIT})
+    list(APPEND WEBP_DEP_LIBRARIES ${CMAKE_THREAD_LIBS_INIT})
   endif()
-  check_c_source_compiles("
-      #include <pthread.h>
-      int main (void) {
-        int attr = PTHREAD_PRIO_INHERIT;
-        return attr;
-      }
-    " FLAG_HAVE_PTHREAD_PRIO_INHERIT)
-  set(HAVE_PTHREAD_PRIO_INHERIT ${FLAG_HAVE_PTHREAD_PRIO_INHERIT})
-  list(APPEND WEBP_DEP_LIBRARIES ${CMAKE_THREAD_LIBS_INIT})
+  set(WEBP_USE_THREAD ${Threads_FOUND})
 endif()
-set(WEBP_USE_THREAD ${Threads_FOUND})
 
 # TODO: this seems unused, check with autotools.
 set(LT_OBJDIR ".libs/")

configure.ac

@@ -1,6 +1,6 @@
-AC_INIT([libwebp], [1.2.0],
+AC_INIT([libwebp], [1.2.2],
         [https://bugs.chromium.org/p/webp],,
-        [http://developers.google.com/speed/webp])
+        [https://developers.google.com/speed/webp])
 AC_CANONICAL_HOST
 AC_PREREQ([2.60])
 AM_INIT_AUTOMAKE([-Wall foreign subdir-objects])
@@ -32,6 +32,40 @@ AC_ARG_ENABLE([everything],
                SET_IF_UNSET([enable_libwebpextras], [$enableval])
                SET_IF_UNSET([enable_libwebpmux], [$enableval])])
 
+dnl === Check whether libwebpmux should be built
+AC_MSG_CHECKING(whether libwebpmux is to be built)
+AC_ARG_ENABLE([libwebpmux],
+              AS_HELP_STRING([--disable-libwebpmux],
+                             [Disable libwebpmux @<:@default=no@:>@]),
+              [], [enable_libwebpmux=yes])
+AC_MSG_RESULT(${enable_libwebpmux-no})
+AM_CONDITIONAL([BUILD_MUX], [test "$enable_libwebpmux" = "yes"])
+
+dnl === Check whether libwebpdemux should be built
+AC_MSG_CHECKING(whether libwebpdemux is to be built)
+AC_ARG_ENABLE([libwebpdemux],
+              AS_HELP_STRING([--disable-libwebpdemux],
+                             [Disable libwebpdemux @<:@default=no@:>@]),
+              [], [enable_libwebpdemux=yes])
+AC_MSG_RESULT(${enable_libwebpdemux-no})
+AM_CONDITIONAL([BUILD_DEMUX], [test "$enable_libwebpdemux" = "yes"])
+
+dnl === Check whether decoder library should be built.
+AC_MSG_CHECKING(whether decoder library is to be built)
+AC_ARG_ENABLE([libwebpdecoder],
+              AS_HELP_STRING([--enable-libwebpdecoder],
+                             [Build libwebpdecoder @<:@default=no@:>@]))
+AC_MSG_RESULT(${enable_libwebpdecoder-no})
+AM_CONDITIONAL([BUILD_LIBWEBPDECODER], [test "$enable_libwebpdecoder" = "yes"])
+
+dnl === Check whether libwebpextras should be built
+AC_MSG_CHECKING(whether libwebpextras is to be built)
+AC_ARG_ENABLE([libwebpextras],
+              AS_HELP_STRING([--enable-libwebpextras],
+                             [Build libwebpextras @<:@default=no@:>@]))
+AC_MSG_RESULT(${enable_libwebpextras-no})
+AM_CONDITIONAL([BUILD_EXTRAS], [test "$enable_libwebpextras" = "yes"])
+
 dnl === If --enable-asserts is not defined, define NDEBUG
 
 AC_MSG_CHECKING(whether asserts are enabled)
@@ -225,6 +259,9 @@ AS_IF([test "x$enable_neon" != "xno"], [
         *android*) AC_CHECK_HEADERS([cpu-features.h]) ;;
       esac
       ;;
+    aarch64*|arm64*)
+      AC_DEFINE(WEBP_HAVE_NEON, [1], [Set to 1 if NEON is supported])
+      ;;
   esac
   AC_SUBST([NEON_FLAGS])])
 
@@ -409,6 +446,10 @@ AS_IF([test "x$enable_gl" != "xno"], [
 
   if test "$glut_support" = "yes" -a "$enable_libwebpdemux" = "yes"; then
     build_vwebp=yes
+  else
+    AC_MSG_NOTICE(
+      m4_normalize([Not building vwebp.
+                    OpenGL libraries and --enable-libwebpdemux are required.]))
   fi
 ])
 AM_CONDITIONAL([BUILD_VWEBP], [test "$build_vwebp" = "yes"])
@@ -468,7 +509,7 @@ AS_IF([test "x$enable_sdl" != "xno"], [
   if test x"$sdl_support" = "xyes"; then
     build_vwebp_sdl=yes
   else
-    AC_MSG_WARN(Optional SDL library not found)
+    AC_MSG_NOTICE([Not building vwebp-sdl. SDL library is required.])
   fi
 ])
 
@@ -592,11 +633,17 @@ AS_IF([test "x$enable_gif" != "xno"], [
   if test "$gif_support" = "yes" -a \
           "$enable_libwebpdemux" = "yes"; then
     build_anim_diff=yes
+  else
+    AC_MSG_NOTICE(
+      [Not building anim_diff. libgif and --enable-libwebpdemux are required.])
   fi
 
   if test "$gif_support" = "yes" -a \
           "$enable_libwebpmux" = "yes"; then
     build_gif2webp=yes
+  else
+    AC_MSG_NOTICE(
+      [Not building gif2webp. libgif and --enable-libwebpmux are required.])
   fi
 ])
 AM_CONDITIONAL([BUILD_ANIMDIFF], [test "${build_anim_diff}" = "yes"])
@@ -604,11 +651,17 @@ AM_CONDITIONAL([BUILD_GIF2WEBP], [test "${build_gif2webp}" = "yes"])
 
 if test "$enable_libwebpdemux" = "yes" -a "$enable_libwebpmux" = "yes"; then
   build_img2webp=yes
+else
+  AC_MSG_NOTICE(
+    m4_normalize([Not building img2webp.
+                  --enable-libwebpdemux & --enable-libwebpmux are required.]))
 fi
 AM_CONDITIONAL([BUILD_IMG2WEBP], [test "${build_img2webp}" = "yes"])
 
 if test "$enable_libwebpmux" = "yes"; then
   build_webpinfo=yes
+else
+  AC_MSG_NOTICE([Not building webpinfo. --enable-libwebpdemux is required.])
 fi
 AM_CONDITIONAL([BUILD_WEBPINFO], [test "${build_webpinfo}" = "yes"])
 
@@ -692,39 +745,6 @@ else
   AC_MSG_RESULT([no])
 fi
 
-dnl === Check whether libwebpmux should be built
-AC_MSG_CHECKING(whether libwebpmux is to be built)
-AC_ARG_ENABLE([libwebpmux],
-              AS_HELP_STRING([--enable-libwebpmux],
-                             [Build libwebpmux @<:@default=no@:>@]))
-AC_MSG_RESULT(${enable_libwebpmux-no})
-AM_CONDITIONAL([BUILD_MUX], [test "$enable_libwebpmux" = "yes"])
-
-dnl === Check whether libwebpdemux should be built
-AC_MSG_CHECKING(whether libwebpdemux is to be built)
-AC_ARG_ENABLE([libwebpdemux],
-              AS_HELP_STRING([--disable-libwebpdemux],
-                             [Disable libwebpdemux @<:@default=no@:>@]),
-              [], [enable_libwebpdemux=yes])
-AC_MSG_RESULT(${enable_libwebpdemux-no})
-AM_CONDITIONAL([BUILD_DEMUX], [test "$enable_libwebpdemux" = "yes"])
-
-dnl === Check whether decoder library should be built.
-AC_MSG_CHECKING(whether decoder library is to be built)
-AC_ARG_ENABLE([libwebpdecoder],
-              AS_HELP_STRING([--enable-libwebpdecoder],
-                             [Build libwebpdecoder @<:@default=no@:>@]))
-AC_MSG_RESULT(${enable_libwebpdecoder-no})
-AM_CONDITIONAL([BUILD_LIBWEBPDECODER], [test "$enable_libwebpdecoder" = "yes"])
-
-dnl === Check whether libwebpextras should be built
-AC_MSG_CHECKING(whether libwebpextras is to be built)
-AC_ARG_ENABLE([libwebpextras],
-              AS_HELP_STRING([--enable-libwebpextras],
-                             [Build libwebpextras @<:@default=no@:>@]))
-AC_MSG_RESULT(${enable_libwebpextras-no})
-AM_CONDITIONAL([BUILD_EXTRAS], [test "$enable_libwebpextras" = "yes"])
-
 dnl =========================
 
 AC_CONFIG_MACRO_DIR([m4])

doc/README

@@ -6,8 +6,8 @@ HTML generation requires kramdown [1], easily installed as a
 rubygem [2].  Rubygems installation should satisfy dependencies
 automatically.
 
-[1]: http://kramdown.rubyforge.org/
-[2]: http://rubygems.org/
+[1]: https://kramdown.gettalong.org/
+[2]: https://rubygems.org/
 
 HTML generation can then be done from the project root:
 
@@ -26,4 +26,4 @@ $ kramdown doc/webp-lossless-bitstream-spec.txt --template \
 
 Optimally, use kramdown 0.13.7 or newer if syntax highlighting desired.
 
-[3]: http://coderay.rubychan.de/
+[3]: https://github.com/rubychan/coderay

doc/webp-container-spec.txt

@@ -2,7 +2,7 @@
 
 Although you may be viewing an alternate representation, this document
 is sourced in Markdown, a light-duty markup scheme, and is optimized for
-the [kramdown](http://kramdown.rubyforge.org/) transformer.
+the [kramdown](https://kramdown.gettalong.org/) transformer.
 
 See the accompanying README. External link targets are referenced at the
 end of this file.
@@ -53,10 +53,6 @@ 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.
-
 Terminology & Basics
 ------------------------
 
@@ -269,7 +265,7 @@ An extended format file consists of:
 
   * An optional 'XMP ' chunk with XMP metadata.
 
-  * An optional list of [unknown chunks](#unknown-chunks). _\[status: experimental\]_
+  * An optional list of [unknown chunks](#unknown-chunks).
 
 For a _still image_, the _image data_ consists of a single frame, which is made
 up of:
@@ -705,7 +701,7 @@ XMP Metadata: _Chunk Size_ bytes
 Additional guidance about handling metadata can be found in the
 Metadata Working Group's [Guidelines for Handling Metadata][metadata].
 
-#### Unknown Chunks _\[status: experimental\]_
+#### Unknown Chunks
 
 A RIFF chunk (described in [this](#terminology-amp-basics) section) whose _chunk
 tag_ is different from any of the chunks described in this document, is
@@ -815,9 +811,9 @@ RIFF/WEBP
 +- EXIF (metadata)
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-[vp8spec]:  http://tools.ietf.org/html/rfc6386
-[webpllspec]: https://chromium.googlesource.com/webm/libwebp/+/master/doc/webp-lossless-bitstream-spec.txt
-[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
+[vp8spec]:  https://datatracker.ietf.org/doc/html/rfc6386
+[webpllspec]: https://chromium.googlesource.com/webm/libwebp/+/HEAD/doc/webp-lossless-bitstream-spec.txt
+[iccspec]: https://www.color.org/icc_specs2.xalter
+[metadata]: https://web.archive.org/web/20180919181934/http://www.metadataworkinggroup.org/pdf/mwg_guidance.pdf
+[rfc 1166]: https://datatracker.ietf.org/doc/html/rfc1166
+[rfc 2119]: https://datatracker.ietf.org/doc/html/rfc2119

doc/webp-lossless-bitstream-spec.txt

@@ -2,7 +2,7 @@
 
 Although you may be viewing an alternate representation, this document
 is sourced in Markdown, a light-duty markup scheme, and is optimized for
-the [kramdown](http://kramdown.rubyforge.org/) transformer.
+the [kramdown](https://kramdown.gettalong.org/) transformer.
 
 See the accompanying README. External link targets are referenced at the
 end of this file.
@@ -1087,4 +1087,4 @@ A possible example sequence:
 <lz77-coded image>
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-[canonical_huff]: http://en.wikipedia.org/wiki/Canonical_Huffman_code
+[canonical_huff]: https://en.wikipedia.org/wiki/Canonical_Huffman_code

examples/gif2webp.c

@@ -465,8 +465,10 @@ int main(int argc, const char* argv[]) {
     fprintf(stderr, "%s\n", WebPAnimEncoderGetError(enc));
     goto End;
   }
-
-  if (!loop_compatibility) {
+  // If there's only one frame, we don't need to handle loop count.
+  if (frame_number == 1) {
+    loop_count = 0;
+  } else if (!loop_compatibility) {
     if (!stored_loop_count) {
       // if no loop-count element is seen, the default is '1' (loop-once)
       // and we need to signal it explicitly in WebP. Note however that

examples/webpinfo.c

@@ -41,6 +41,7 @@
     if (webp_info->show_diagnosis_) {            \
       fprintf(stderr, "Warning: %s\n", MESSAGE); \
     }                                            \
+    ++webp_info->num_warnings_;                  \
   } while (0)
 
 static const char* const kFormats[3] = {
@@ -116,6 +117,7 @@ typedef struct WebPInfo {
   int is_processing_anim_frame_, seen_alpha_subchunk_, seen_image_subchunk_;
   // Print output control.
   int quiet_, show_diagnosis_, show_summary_;
+  int num_warnings_;
   int parse_bitstream_;
 } WebPInfo;
 
@@ -580,7 +582,7 @@ static WebPInfoStatus ParseAlphaHeader(const ChunkData* const chunk_data,
 // -----------------------------------------------------------------------------
 // Chunk parsing.
 
-static WebPInfoStatus ParseRIFFHeader(const WebPInfo* const webp_info,
+static WebPInfoStatus ParseRIFFHeader(WebPInfo* const webp_info,
                                       MemBuffer* const mem) {
   const size_t min_size = RIFF_HEADER_SIZE + CHUNK_HEADER_SIZE;
   size_t riff_size;
@@ -988,7 +990,7 @@ static WebPInfoStatus ProcessChunk(const ChunkData* const chunk_data,
   return status;
 }
 
-static WebPInfoStatus Validate(const WebPInfo* const webp_info) {
+static WebPInfoStatus Validate(WebPInfo* const webp_info) {
   if (webp_info->num_frames_ < 1) {
     LOG_ERROR("No image/frame detected.");
     return WEBP_INFO_MISSING_DATA;
@@ -1093,16 +1095,14 @@ static WebPInfoStatus AnalyzeWebP(WebPInfo* const webp_info,
     } else {
       printf("Errors detected.\n");
     }
+    if (webp_info->num_warnings_ > 0) {
+      printf("There were %d warning(s).\n", webp_info->num_warnings_);
+    }
   }
   return webp_info_status;
 }
 
-static void HelpShort(void) {
-  printf("Usage: webpinfo [options] in_files\n"
-         "Try -longhelp for an exhaustive list of options.\n");
-}
-
-static void HelpLong(void) {
+static void Help(void) {
   printf("Usage: webpinfo [options] in_files\n"
          "Note: there could be multiple input files;\n"
          "      options must come before input files.\n"
@@ -1123,17 +1123,15 @@ int main(int argc, const char* argv[]) {
   INIT_WARGV(argc, argv);
 
   if (argc == 1) {
-    HelpShort();
+    Help();
     FREE_WARGV_AND_RETURN(WEBP_INFO_OK);
   }
 
   // Parse command-line input.
   for (c = 1; c < argc; ++c) {
-    if (!strcmp(argv[c], "-h") || !strcmp(argv[c], "-help")) {
-      HelpShort();
-      FREE_WARGV_AND_RETURN(WEBP_INFO_OK);
-    } else if (!strcmp(argv[c], "-H") || !strcmp(argv[c], "-longhelp")) {
-      HelpLong();
+    if (!strcmp(argv[c], "-h") || !strcmp(argv[c], "-help") ||
+        !strcmp(argv[c], "-H") || !strcmp(argv[c], "-longhelp")) {
+      Help();
       FREE_WARGV_AND_RETURN(WEBP_INFO_OK);
     } else if (!strcmp(argv[c], "-quiet")) {
       quiet = 1;
@@ -1154,7 +1152,7 @@ int main(int argc, const char* argv[]) {
   }
 
   if (c == argc) {
-    HelpShort();
+    Help();
     FREE_WARGV_AND_RETURN(WEBP_INFO_INVALID_COMMAND);
   }
 

examples/webpmux.c

@@ -99,6 +99,7 @@ typedef enum {
   FEATURE_ANMF,
   FEATURE_DURATION,
   FEATURE_LOOP,
+  FEATURE_BGCOLOR,
   LAST_FEATURE
 } FeatureType;
 
@@ -315,11 +316,12 @@ static void PrintHelp(void) {
 
   printf("\n");
   printf("SET_OPTIONS:\n");
-  printf(" Set color profile/metadata:\n");
-  printf("   loop LOOP_COUNT   set the loop count\n");
-  printf("   icc  file.icc     set ICC profile\n");
-  printf("   exif file.exif    set EXIF metadata\n");
-  printf("   xmp  file.xmp     set XMP metadata\n");
+  printf(" Set color profile/metadata/parameters:\n");
+  printf("   loop LOOP_COUNT            set the loop count\n");
+  printf("   bgcolor BACKGROUND_COLOR   set the animation background color\n");
+  printf("   icc  file.icc              set ICC profile\n");
+  printf("   exif file.exif             set EXIF metadata\n");
+  printf("   xmp  file.xmp              set XMP metadata\n");
   printf("   where:    'file.icc' contains the ICC profile to be set,\n");
   printf("             'file.exif' contains the EXIF metadata to be set\n");
   printf("             'file.xmp' contains the XMP metadata to be set\n");
@@ -472,11 +474,11 @@ static WebPMux* DuplicateMuxHeader(const WebPMux* const mux) {
 }
 
 static int ParseFrameArgs(const char* args, WebPMuxFrameInfo* const info) {
-  int dispose_method, dummy;
+  int dispose_method, unused;
   char plus_minus, blend_method;
   const int num_args = sscanf(args, "+%d+%d+%d+%d%c%c+%d", &info->duration,
                               &info->x_offset, &info->y_offset, &dispose_method,
-                              &plus_minus, &blend_method, &dummy);
+                              &plus_minus, &blend_method, &unused);
   switch (num_args) {
     case 1:
       info->x_offset = info->y_offset = 0;  // fall through
@@ -495,7 +497,7 @@ static int ParseFrameArgs(const char* args, WebPMuxFrameInfo* const info) {
 
   WarnAboutOddOffset(info);
 
-  // Note: The sanity of the following conversion is checked by
+  // Note: The validity of the following conversion is checked by
   // WebPMuxPushFrame().
   info->dispose_method = (WebPMuxAnimDispose)dispose_method;
 
@@ -778,6 +780,13 @@ static int ParseCommandLine(Config* config, const W_CHAR** const unicode_argv) {
         arg->params_ = argv[i + 1];
         ++feature_arg_index;
         i += 2;
+      } else if (!strcmp(argv[i], "bgcolor") &&
+                 (config->action_type_ == ACTION_SET)) {
+        CHECK_NUM_ARGS_AT_LEAST(2, ErrParse);
+        config->type_ = FEATURE_BGCOLOR;
+        arg->params_ = argv[i + 1];
+        ++feature_arg_index;
+        i += 2;
       } else {  // Assume input file.
         if (config->input_ == NULL) {
           config->input_ = wargv[i];
@@ -1053,6 +1062,30 @@ static int Process(const Config* config) {
           }
           break;
         }
+        case FEATURE_BGCOLOR: {
+          WebPMuxAnimParams params = { 0xFFFFFFFF, 0 };
+          uint32_t bgcolor;
+          ok = ParseBgcolorArgs(config->args_[0].params_, &bgcolor);
+          if (!ok) {
+            ERROR_GOTO1("ERROR: Could not parse the background color.\n",
+                        Err2);
+          }
+          ok = CreateMux(config->input_, &mux);
+          if (!ok) goto Err2;
+          ok = (WebPMuxGetAnimationParams(mux, &params) == WEBP_MUX_OK);
+          if (!ok) {
+            ERROR_GOTO1("ERROR: input file does not seem to be an animation.\n",
+                        Err2);
+          }
+          params.bgcolor = bgcolor;
+          err = WebPMuxSetAnimationParams(mux, &params);
+          ok = (err == WEBP_MUX_OK);
+          if (!ok) {
+            ERROR_GOTO2("ERROR (%s): Could not set animation parameters.\n",
+                        ErrorString(err), Err2);
+          }
+          break;
+        }
         default: {
           ERROR_GOTO1("ERROR: Invalid feature for action 'set'.\n", Err2);
           break;

extras/extras.c

@@ -19,7 +19,7 @@
 
 #define XTRA_MAJ_VERSION 1
 #define XTRA_MIN_VERSION 2
-#define XTRA_REV_VERSION 0
+#define XTRA_REV_VERSION 2
 
 //------------------------------------------------------------------------------
 

imageio/image_enc.c

@@ -155,8 +155,8 @@ int WebPWritePNG(const char* out_file_name, int use_stdout,
 }
 
 #elif defined(WEBP_HAVE_PNG)    // !HAVE_WINCODEC_H
-static void PNGAPI PNGErrorFunction(png_structp png, png_const_charp dummy) {
-  (void)dummy;  // remove variable-unused warning
+static void PNGAPI PNGErrorFunction(png_structp png, png_const_charp unused) {
+  (void)unused;  // remove variable-unused warning
   longjmp(png_jmpbuf(png), 1);
 }
 
@@ -320,7 +320,7 @@ int WebPWriteBMP(FILE* fout, const WebPDecBuffer* const buffer) {
   PutLE16(bmp_header + 26, 1);                    // number of planes
   PutLE16(bmp_header + 28, bytes_per_px * 8);     // bits per pixel
   PutLE32(bmp_header + 30, 0);                    // no compression (BI_RGB)
-  PutLE32(bmp_header + 34, 0);                    // image size (dummy)
+  PutLE32(bmp_header + 34, 0);                    // image size (placeholder)
   PutLE32(bmp_header + 38, 2400);                 // x pixels/meter
   PutLE32(bmp_header + 42, 2400);                 // y pixels/meter
   PutLE32(bmp_header + 46, 0);                    // number of palette colors

imageio/image_enc.h

@@ -79,7 +79,7 @@ int WebPWriteTIFF(FILE* fout, const struct WebPDecBuffer* const buffer);
 int WebPWriteAlphaPlane(FILE* fout, const struct WebPDecBuffer* const buffer);
 
 // Save as YUV samples as PGM format (using IMC4 layout).
-// See: http://www.fourcc.org/yuv.php#IMC4.
+// See: https://www.fourcc.org/yuv.php#IMC4.
 // (very convenient format for viewing the samples, esp. for odd dimensions).
 int WebPWritePGM(FILE* fout, const struct WebPDecBuffer* const buffer);
 

imageio/imageio_util.c

@@ -148,9 +148,11 @@ void ImgIoUtilCopyPlane(const uint8_t* src, int src_stride,
 
 // -----------------------------------------------------------------------------
 
-int ImgIoUtilCheckSizeArgumentsOverflow(uint64_t nmemb, size_t size) {
-  const uint64_t total_size = nmemb * size;
+int ImgIoUtilCheckSizeArgumentsOverflow(uint64_t stride, size_t height) {
+  const uint64_t total_size = stride * height;
   int ok = (total_size == (size_t)total_size);
+  // check that 'stride' is representable as int:
+  ok = ok && ((uint64_t)(int)stride == stride);
 #if defined(WEBP_MAX_IMAGE_SIZE)
   ok = ok && (total_size <= (uint64_t)WEBP_MAX_IMAGE_SIZE);
 #endif

imageio/imageio_util.h

@@ -54,8 +54,8 @@ void ImgIoUtilCopyPlane(const uint8_t* src, int src_stride,
 
 //------------------------------------------------------------------------------
 
-// Returns 0 in case of overflow of nmemb * size.
-int ImgIoUtilCheckSizeArgumentsOverflow(uint64_t nmemb, size_t size);
+// Returns 0 in case of overflow, memory over-allocation or excessive dimension.
+int ImgIoUtilCheckSizeArgumentsOverflow(uint64_t stride, size_t height);
 
 #ifdef __cplusplus
 }    // extern "C"

imageio/jpegdec.c

@@ -274,7 +274,7 @@ int ReadJPEG(const uint8_t* const data, size_t data_size,
   ctx.data = data;
   ctx.data_size = data_size;
 
-  memset((j_decompress_ptr)&dinfo, 0, sizeof(dinfo));   // for setjmp sanity
+  memset((j_decompress_ptr)&dinfo, 0, sizeof(dinfo));   // for setjmp safety
   dinfo.err = jpeg_std_error(&jerr.pub);
   jerr.pub.error_exit = my_error_exit;
 

imageio/pngdec.c

@@ -133,7 +133,7 @@ static const struct {
                  MetadataPayload* const payload);
   size_t storage_offset;
 } kPNGMetadataMap[] = {
-  // http://www.sno.phy.queensu.ca/~phil/exiftool/TagNames/PNG.html#TextualData
+  // https://exiftool.org/TagNames/PNG.html#TextualData
   // See also: ExifTool on CPAN.
   { "Raw profile type exif", ProcessRawProfile, METADATA_OFFSET(exif) },
   { "Raw profile type xmp",  ProcessRawProfile, METADATA_OFFSET(xmp) },
@@ -259,11 +259,14 @@ int ReadPNG(const uint8_t* const data, size_t data_size,
     goto End;
   }
 
+#if LOCAL_PNG_PREREQ(1,5) || \
+    (LOCAL_PNG_PREREQ(1,4) && PNG_LIBPNG_VER_RELEASE >= 1)
   // If it looks like the bitstream is going to need more memory than libpng's
   // internal limit (default: 8M), try to (reasonably) raise it.
   if (data_size > png_get_chunk_malloc_max(png) && data_size < (1u << 24)) {
     png_set_chunk_malloc_max(png, data_size);
   }
+#endif
 
   info = png_create_info_struct(png);
   if (info == NULL) goto Error;

imageio/tiffdec.c

@@ -46,7 +46,7 @@ static int ExtractMetadataFromTIFF(TIFF* const tif, Metadata* const metadata) {
         (MetadataPayload*)((uint8_t*)metadata +
                            kTIFFMetadataMap[i].storage_offset);
     void* tag_data;
-    uint32 tag_data_len;
+    uint32_t tag_data_len;
 
     if (TIFFGetField(tif, kTIFFMetadataMap[i].tag, &tag_data_len, &tag_data) &&
         !MetadataCopy((const char*)tag_data, tag_data_len, payload)) {
@@ -157,6 +157,7 @@ int ReadTIFF(const uint8_t* const data, size_t data_size,
   MyData my_data = { data, (toff_t)data_size, 0 };
   TIFF* tif;
   uint32_t image_width, image_height, tile_width, tile_height;
+  uint64_t stride;
   uint16_t samples_per_px = 0;
   uint16_t extra_samples = 0;
   uint16_t* extra_samples_ptr = NULL;
@@ -194,20 +195,31 @@ int ReadTIFF(const uint8_t* const data, size_t data_size,
     fprintf(stderr, "Error! Cannot retrieve TIFF image dimensions.\n");
     goto End;
   }
-  if (!ImgIoUtilCheckSizeArgumentsOverflow((uint64_t)image_width * image_height,
-                                           sizeof(*raster))) {
+  stride = (uint64_t)image_width * sizeof(*raster);
+  if (!ImgIoUtilCheckSizeArgumentsOverflow(stride, image_height)) {
+    fprintf(stderr, "Error! TIFF image dimension (%d x %d) is too large.\n",
+            image_width, image_height);
     goto End;
   }
+
   // According to spec, a tile can be bigger than the image. However it should
-  // be a multiple of 16 and not way too large, so check that it's not more than
-  // twice the image size, for dimensions above some arbitrary minimum 32.
-  if ((TIFFGetField(tif, TIFFTAG_TILEWIDTH, &tile_width) &&
-       tile_width > 32 && tile_width / 2 > image_width) ||
-      (TIFFGetField(tif, TIFFTAG_TILELENGTH, &tile_height) &&
-       tile_height > 32 && tile_height / 2 > image_height)) {
-    fprintf(stderr, "Error! TIFF tile dimensions are too big.\n");
-    goto End;
+  // be a multiple of 16 and not way too large, so check that it's not more
+  // than twice the image size, for dimensions above some arbitrary minimum
+  // 32. We also check that they respect WebP's dimension and memory limit.
+  // Note that a tile can be 6byte/px in some cases. Here we assume
+  // 4byte/px with sizeof(*raster), to be conservative.
+  if (TIFFGetField(tif, TIFFTAG_TILEWIDTH, &tile_width) &&
+      TIFFGetField(tif, TIFFTAG_TILELENGTH, &tile_height)) {
+    if ((tile_width > 32 && tile_width / 2 > image_width) ||
+        (tile_height > 32 && tile_height / 2 > image_height) ||
+        ImgIoUtilCheckSizeArgumentsOverflow(
+            (uint64_t)tile_width * sizeof(*raster), tile_height)) {
+      fprintf(stderr, "Error! TIFF tile dimension (%d x %d) is too large.\n",
+              tile_width, tile_height);
+      goto End;
+    }
   }
+
   if (samples_per_px > 3 && !TIFFGetField(tif, TIFFTAG_EXTRASAMPLES,
                                           &extra_samples, &extra_samples_ptr)) {
     fprintf(stderr, "Error! Cannot retrieve TIFF ExtraSamples info.\n");
@@ -215,15 +227,13 @@ int ReadTIFF(const uint8_t* const data, size_t data_size,
   }
 
   // _Tiffmalloc uses a signed type for size.
-  alloc_size =
-      (int64_t)((uint64_t)image_width * image_height * sizeof(*raster));
+  alloc_size = (int64_t)(stride * image_height);
   if (alloc_size < 0 || alloc_size != (tsize_t)alloc_size) goto End;
 
-  raster = (uint32*)_TIFFmalloc((tsize_t)alloc_size);
+  raster = (uint32_t*)_TIFFmalloc((tsize_t)alloc_size);
   if (raster != NULL) {
     if (TIFFReadRGBAImageOriented(tif, image_width, image_height, raster,
                                   ORIENTATION_TOPLEFT, 1)) {
-      const int stride = image_width * sizeof(*raster);
       pic->width = image_width;
       pic->height = image_height;
       // TIFF data is ABGR
@@ -241,8 +251,8 @@ int ReadTIFF(const uint8_t* const data, size_t data_size,
         }
       }
       ok = keep_alpha
-         ? WebPPictureImportRGBA(pic, (const uint8_t*)raster, stride)
-         : WebPPictureImportRGBX(pic, (const uint8_t*)raster, stride);
+         ? WebPPictureImportRGBA(pic, (const uint8_t*)raster, (int)stride)
+         : WebPPictureImportRGBX(pic, (const uint8_t*)raster, (int)stride);
     }
     _TIFFfree(raster);
   } else {

imageio/webpdec.c

@@ -95,7 +95,7 @@ VP8StatusCode DecodeWebPIncremental(
   {
     WebPIDecoder* const idec = WebPIDecode(data, data_size, config);
     if (idec == NULL) {
-      fprintf(stderr, "Failed during WebPINewDecoder().\n");
+      fprintf(stderr, "Failed during WebPIDecode().\n");
       return VP8_STATUS_OUT_OF_MEMORY;
     } else {
       status = WebPIUpdate(idec, data, data_size);

imageio/wicdec.c

@@ -134,7 +134,10 @@ static HRESULT ExtractICCP(IWICImagingFactory* const factory,
   IWICColorContext** color_contexts;
 
   IFS(IWICBitmapFrameDecode_GetColorContexts(frame, 0, NULL, &count));
-  if (FAILED(hr) || count == 0) return hr;
+  if (FAILED(hr) || count == 0) {
+    // Treat unsupported operation as a non-fatal error. See crbug.com/webp/506.
+    return (hr == WINCODEC_ERR_UNSUPPORTEDOPERATION) ? S_OK : hr;
+  }
 
   color_contexts = (IWICColorContext**)calloc(count, sizeof(*color_contexts));
   if (color_contexts == NULL) return E_OUTOFMEMORY;
@@ -270,10 +273,15 @@ int ReadPictureWithWIC(const char* const filename,
   WICPixelFormatGUID src_pixel_format = GUID_WICPixelFormatUndefined;
   const WICFormatImporter* importer = NULL;
   GUID src_container_format = GUID_NULL_;
+  // From Windows Kits\10\Include\10.0.19041.0\um\wincodec.h
+  WEBP_DEFINE_GUID(GUID_ContainerFormatWebp_,
+                   0xe094b0e2, 0x67f2, 0x45b3,
+                   0xb0, 0xea, 0x11, 0x53, 0x37, 0xca, 0x7c, 0xf3);
   static const GUID* kAlphaContainers[] = {
     &GUID_ContainerFormatBmp,
     &GUID_ContainerFormatPng,
     &GUID_ContainerFormatTiff,
+    &GUID_ContainerFormatWebp_,
     NULL
   };
   int has_alpha = 0;

infra/common.sh

@@ -0,0 +1,106 @@
+# Copyright (c) 2021, Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#
+#   * Neither the name of Google nor the names of its contributors may
+#     be used to endorse or promote products derived from this software
+#     without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+log_err() {
+  echo "[$(date +'%Y-%m-%dT%H:%M:%S%z')]: $*" >&2
+}
+
+#######################################
+# Create build directory. Build directory will be deleted if it exists.
+# Arguments:
+#   None.
+# Returns:
+#   mkdir result.
+#######################################
+make_build_dir() {
+  if [[ "$#" -ne 1 ]]; then
+    return 1
+  fi
+
+  local build_dir
+  build_dir="$1"
+  rm -rf "${build_dir}"
+  mkdir -p "${build_dir}"
+}
+
+#######################################
+# Cleanup files from the build directory.
+# Globals:
+#   LIBWEBP_ROOT repository's root path.
+# Arguments:
+#   $1 build directory.
+#######################################
+cleanup() {
+  # $1 is not completely removed to allow for binary artifacts to be
+  # extracted.
+  find "${1:?"Build directory not defined"}" \
+    \( -name "*.[ao]" -o -name "*.l[ao]" \) -exec rm -f {} +
+}
+
+#######################################
+# Setup ccache for toolchain.
+# Globals:
+#   PATH
+# Arguments:
+#   None.
+#######################################
+setup_ccache() {
+  if [[ -x "$(command -v ccache)" ]]; then
+    export CCACHE_CPP2=yes
+    export PATH="/usr/lib/ccache:${PATH}"
+  fi
+}
+
+#######################################
+# Detects whether test block should be run in the current test shard.
+# Globals:
+#   TEST_TOTAL_SHARDS: Valid range: [1, N]. Defaults to 1.
+#   TEST_SHARD_INDEX: Valid range: [0, TEST_TOTAL_SHARDS). Defaults to 0.
+#   libwebp_test_id: current test number; incremented with each call.
+# Arguments:
+#   None
+# Returns:
+#   true if the shard is active
+#   false if the shard is inactive
+#######################################
+shard_should_run() {
+  TEST_TOTAL_SHARDS=${TEST_TOTAL_SHARDS:=1}
+  TEST_SHARD_INDEX=${TEST_SHARD_INDEX:=0}
+  libwebp_test_id=${libwebp_test_id:=-1}
+  : $((libwebp_test_id += 1))
+
+  if [[ "${TEST_SHARD_INDEX}" -lt 0 ||
+    "${TEST_SHARD_INDEX}" -ge "${TEST_TOTAL_SHARDS}" ]]; then
+    log_err "Invalid TEST_SHARD_INDEX (${TEST_SHARD_INDEX})!" \
+      "Expected [0, ${TEST_TOTAL_SHARDS})."
+  fi
+
+  [[ "$((libwebp_test_id % TEST_TOTAL_SHARDS))" -eq "${TEST_SHARD_INDEX}" ]]
+}

infra/compile.sh

@@ -0,0 +1,401 @@
+#!/bin/bash
+# Copyright (c) 2021, Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#
+#   * Neither the name of Google nor the names of its contributors may
+#     be used to endorse or promote products derived from this software
+#     without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+set -xe
+LIBWEBP_ROOT="$(realpath "$(dirname "$0")/..")"
+WORKSPACE=${WORKSPACE:-"$(mktemp -d -t webp.XXX)"}
+
+# shellcheck source=infra/common.sh
+source "${LIBWEBP_ROOT}/infra/common.sh"
+
+usage() {
+  cat << EOF
+Usage: compile.sh BUILD_TYPE TARGET
+Options:
+BUILD_TYPE  supported build type: (shared, static, static-debug)
+TARGET      supported target platforms:
+              aarch64-linux-clang
+              aarch64-linux-gnu
+              arm-linux-gnueabi
+              arm-neon-linux-gnueabi
+              cmake
+              cmake-aarch64
+              cmake-arm
+              cmake-clang
+              disable-near-lossless
+              disable-sse4.1
+              disable-stats
+              force-aligned-32
+              force-aligned-64
+              gradle
+              i686-linux-asan
+              i686-linux-clang
+              i686-linux-gnu
+              i686-w64-mingw32
+              mips2el-linux-gnu
+              mips32dspr2el-linux-gnu
+              mips32eb-linux-gnu
+              mips32el-linux-gnu
+              mips32r2el-linux-gnu
+              mips32r5el-linux-gnu
+              mips64r2el-linux-gnu
+              mips64r6el-linux-gnu
+              native
+              reduce-csp
+              reduce-size
+              reduce-size-disable-stats
+              visibility-default-gnu
+              visibility-hidden-clang
+              visibility-hidden-gnu
+              wasm
+              x86_64-linux-clang
+              x86_64-linux-gnu
+              x86_64-linux-msan
+              x86_64-w64-mingw32
+Environment variables:
+WORKSPACE   directory where the build is done
+EOF
+}
+
+################################################################################
+echo "Building libwebp in ${WORKSPACE}"
+
+if [[ ! -d "${WORKSPACE}" ]]; then
+  log_err "${WORKSPACE} directory does not exist"
+  exit 1
+fi
+
+BUILD_TYPE=${1:?"Build type not defined.$(
+  echo
+  usage
+)"}
+TARGET=${2:?"Target not defined.$(
+  echo
+  usage
+)"}
+readonly BUILD_DIR="${WORKSPACE}/build-${BUILD_TYPE}"
+
+trap 'cleanup ${BUILD_DIR}' EXIT
+make_build_dir "${BUILD_DIR}"
+
+config_flags=()
+case "${BUILD_TYPE}" in
+  shared*) ;;  # Valid BUILD_TYPE but no setup required
+  static*) config_flags+=("--disable-shared") ;;
+  experimental) config_flags+=("--enable-experimental") ;;
+  *)
+    log_err "Invalid BUILD_TYPE"
+    usage
+    exit 1
+    ;;
+esac
+
+if grep -m 1 -q "enable-asserts" "${LIBWEBP_ROOT}/configure.ac"; then
+  config_flags+=("--enable-asserts")
+fi
+
+case "${TARGET}" in
+  aarch64-linux-clang)
+    TARGET="aarch64-linux-gnu"
+    CC="clang"
+    CC="${CC} --target=aarch64-linux-gnu"
+    export CC
+    export CFLAGS="-isystem /usr/aarch64-linux-gnu/include"
+    ;;
+  arm-linux-gnueabi)
+    export CFLAGS="-O3 -march=armv7-a -mfloat-abi=softfp -ftree-vectorize"
+    ;;
+  arm-neon-linux-gnueabi)
+    TARGET="arm-linux-gnueabi"
+    CFLAGS="-O3 -march=armv7-a -mfpu=neon -mfloat-abi=softfp -ftree-vectorize"
+    export CFLAGS
+    ;;
+  mips2el-linux-gnu)
+    export CFLAGS="-EL -O2 -mips2"
+    TARGET="mipsel-linux-gnu"
+    ;;
+  mips32el-linux-gnu)
+    export CFLAGS="-EL -O2 -mips32"
+    TARGET="mipsel-linux-gnu"
+    ;;
+  mips32r2el-linux-gnu)
+    export CFLAGS="-EL -O2 -mips32r2"
+    TARGET="mipsel-linux-gnu"
+    ;;
+  mips32dspr2el-linux-gnu)
+    export CFLAGS="-EL -O2 -mdspr2"
+    TARGET="mipsel-linux-gnu"
+    ;;
+  mips32r5el-linux-gnu)
+    export CFLAGS="-EL -O2 -mips32r5 -mmsa"
+    TARGET="mipsel-linux-gnu"
+    ;;
+  mips32eb-linux-gnu)
+    export CFLAGS="-EB -O2 -mips32"
+    TARGET="mips-linux-gnu"
+    ;;
+  mips64r2el-linux-gnu)
+    export CFLAGS="-EL -O2 -mips64r2 -mabi=64"
+    TARGET="mips64el-linux-gnuabi64"
+    ;;
+  mips64r6el-linux-gnu)
+    export CFLAGS="-EL -O2 -mips64r6 -mabi=64 -mmsa"
+    TARGET="mips-img-linux-gnu"
+    ;;
+  i686-linux-gnu)
+    export CC="gcc -m32"
+    ;;
+  i686-linux-clang)
+    TARGET="i686-linux-gnu"
+    export CC="clang -m32"
+    ;;
+  i686-linux-asan)
+    TARGET="i686-linux-gnu"
+    export CC="clang -m32 -fsanitize=address"
+    ;;
+  i686-linux-msan)
+    TARGET="i686-linux-gnu"
+    export CC="clang -m32 -fsanitize=memory"
+    ;;
+  x86_64-linux-clang)
+    TARGET="x86_64-linux-gnu"
+    export CC=clang
+    ;;
+  x86_64-linux-msan)
+    TARGET="x86_64-linux-gnu"
+    export CC="clang -fsanitize=memory"
+    ;;
+  force-aligned-32)
+    config_flags+=("--enable-aligned")
+    TARGET="i686-linux-gnu"
+    export CC="gcc -m32"
+    ;;
+  force-aligned-64)
+    config_flags+=("--enable-aligned")
+    TARGET="x86_64-linux-gnu"
+    ;;
+  visibility-default-*)
+    export CFLAGS="-O2 -g -fvisibility=default"
+    TARGET="x86_64-linux-gnu"
+    ;;
+  visibility-hidden-*)
+    export CFLAGS="-O2 -g -fvisibility=hidden"
+    if [[ "${TARGET}" = "visibility-hidden-clang" ]]; then
+      export CC=clang
+    fi
+    TARGET="x86_64-linux-gnu"
+    ;;
+  disable-sse4.1)
+    grep "${TARGET}" "${LIBWEBP_ROOT}/configure.ac" || exit 0
+    config_flags+=("--${TARGET}")
+    TARGET="x86_64-linux-gnu"
+    ;;
+  disable-near-lossless)
+    grep "${TARGET}" "${LIBWEBP_ROOT}/configure.ac" || exit 0
+    config_flags+=("--${TARGET}")
+    TARGET="x86_64-linux-gnu"
+    ;;
+  disable-stats)
+    git -C "${LIBWEBP_ROOT}" grep WEBP_DISABLE_STATS || exit 0
+    export CFLAGS="-O2 -g -DWEBP_DISABLE_STATS"
+    TARGET="x86_64-linux-gnu"
+    ;;
+  reduce-size)
+    git -C "${LIBWEBP_ROOT}" grep WEBP_REDUCE_SIZE || exit 0
+    export CFLAGS="-O2 -g -DWEBP_REDUCE_SIZE"
+    TARGET="x86_64-linux-gnu"
+    ;;
+  reduce-size-disable-stats)
+    git -C "${LIBWEBP_ROOT}" grep -e WEBP_DISABLE_STATS -e WEBP_REDUCE_SIZE \
+      || exit 0
+    export CFLAGS="-O2 -g -DWEBP_DISABLE_STATS -DWEBP_REDUCE_SIZE"
+    TARGET="x86_64-linux-gnu"
+    ;;
+  reduce-csp)
+    git -C "${LIBWEBP_ROOT}" grep WEBP_REDUCE_CSP || exit 0
+    export CFLAGS="-O2 -g -DWEBP_REDUCE_CSP"
+    TARGET="x86_64-linux-gnu"
+    ;;
+  x86_64-linux-gnu | *mingw32 | aarch64*) ;;  # Default target configuration
+  # non-configure based builds
+  native)
+    setup_ccache
+    # exercise makefile.unix then quit
+    make -C "${LIBWEBP_ROOT}" -f makefile.unix -j all
+    for tgt in extras examples/anim_diff; do
+      grep -q -m 1 "${tgt}" "${LIBWEBP_ROOT}/makefile.unix" \
+        && make -C "${LIBWEBP_ROOT}" -f makefile.unix -j "${tgt}"
+    done
+    [[ -d "${LIBWEBP_ROOT}/tests/fuzzer" ]] \
+      && make -j -C "${LIBWEBP_ROOT}/tests/fuzzer" -f makefile.unix
+    exit 0
+    ;;
+  cmake*)
+    setup_ccache
+    # exercise cmake then quit
+    opts=()
+    case "${TARGET}" in
+      cmake-clang)
+        opts+=("-DCMAKE_C_COMPILER=clang")
+        ;;
+      cmake-arm)
+        opts+=("-DCMAKE_C_COMPILER=arm-linux-gnueabi-gcc")
+        case "${GERRIT_BRANCH:-}" in
+          portable-intrinsics | 0.6.1) exit 0 ;;
+          *) ;;  # Skip configuration
+        esac
+        ;;
+      cmake-aarch64)
+        opts+=("-DCMAKE_C_COMPILER=aarch64-linux-gnu-gcc")
+        case "${GERRIT_BRANCH:-}" in
+          portable-intrinsics | 0.6.1) exit 0 ;;
+          *) ;;  # Skip configuration
+        esac
+        ;;
+      *) ;;  # Skip configuration
+    esac
+    case "${BUILD_TYPE}" in
+      static*)
+        opts+=("-DBUILD_SHARED_LIBS=OFF")
+        ;;
+      experimental)
+        opts+=("-DWEBP_EXPERIMENTAL_FEATURES=ON" "-DBUILD_SHARED_LIBS=ON")
+        ;;
+      *)
+        opts+=("-DBUILD_SHARED_LIBS=ON")
+        ;;
+    esac
+    case "${BUILD_TYPE}" in
+      *debug) opts+=("-DCMAKE_BUILD_TYPE=Debug") ;;
+      *) opts+=("-DCMAKE_BUILD_TYPE=RelWithDebInfo") ;;
+    esac
+    cd "${BUILD_DIR}"
+    opts+=("-DWEBP_BUILD_CWEBP=ON" "-DWEBP_BUILD_DWEBP=ON")
+    grep -m 1 -q WEBP_BUILD_GIF2WEBP "${LIBWEBP_ROOT}/CMakeLists.txt" \
+      && opts+=("-DWEBP_BUILD_GIF2WEBP=ON")
+    grep -m 1 -q WEBP_BUILD_IMG2WEBP "${LIBWEBP_ROOT}/CMakeLists.txt" \
+      && opts+=("-DWEBP_BUILD_IMG2WEBP=ON")
+    cmake "${opts[@]}" "${LIBWEBP_ROOT}"
+    make VERBOSE=1 -j
+    case "${BUILD_TYPE}" in
+      static)
+        mkdir -p examples
+        cp [cd]webp examples
+        ;;
+      *) ;;  # Skip configuration.
+    esac
+
+    grep "install" "${LIBWEBP_ROOT}/CMakeLists.txt" || exit 0
+
+    make DESTDIR="${BUILD_DIR}/webp-install" install/strip
+    mkdir tmp
+    cd tmp
+    cat > CMakeLists.txt << EOF
+cmake_minimum_required(VERSION 2.8.7)
+
+project(libwebp C)
+
+find_package(WebP)
+if (NOT WebP_FOUND)
+  message(FATAL_ERROR "WebP package not found")
+endif ()
+message("WebP_FOUND: \${WebP_FOUND}")
+message("WebP_INCLUDE_DIRS: \${WebP_INCLUDE_DIRS}")
+message("WebP_LIBRARIES: \${WebP_LIBRARIES}")
+message("WEBP_INCLUDE_DIRS: \${WEBP_INCLUDE_DIRS}")
+message("WEBP_LIBRARIES: \${WEBP_LIBRARIES}")
+EOF
+    cmake . "${opts[@]}" \
+      "-DCMAKE_PREFIX_PATH=${BUILD_DIR}/webp-install/usr/local"
+    exit 0
+    ;;
+  gradle)
+    setup_ccache
+    # exercise gradle then quit
+    [[ -f "${LIBWEBP_ROOT}/gradlew" ]] || exit 0
+
+    cd "${BUILD_DIR}"
+    # TODO -g / --gradle-user-home could be used if there's a race between jobs
+    "${LIBWEBP_ROOT}/gradlew" -p "${LIBWEBP_ROOT}" buildAllExecutables
+    exit 0
+    ;;
+  wasm)
+    grep -m 1 -q WEBP_ENABLE_WASM "${LIBWEBP_ROOT}/CMakeLists.txt" || exit 0
+    opts+=("-DCMAKE_C_COMPILER=clang" "-DWEBP_ENABLE_WASM=ON")
+    opts+=("-DWEBP_BUILD_CWEBP=ON" "-DWEBP_BUILD_DWEBP=ON")
+    case "${BUILD_TYPE}" in
+      *debug) opts+=("-DCMAKE_BUILD_TYPE=Debug") ;;
+      *) opts+=("-DCMAKE_BUILD_TYPE=RelWithDebInfo") ;;
+    esac
+    cd "${BUILD_DIR}"
+    cmake "${opts[@]}" "${LIBWEBP_ROOT}"
+    make VERBOSE=1 -j
+    mkdir examples
+    case "${BUILD_TYPE}" in
+      static)
+        mkdir -p examples
+        cp [cd]webp examples
+        ;;
+      *) ;;  # Skip configuration
+    esac
+    exit 0
+    ;;
+  *)
+    log_err "Invalid TARGET"
+    usage
+    exit 1
+    ;;
+esac
+
+case "${TARGET}" in
+  *mingw32) ;;  # Skip configuration
+  *)
+    case "${TARGET}-${CC}" in
+      static-debug-gcc* | static-debug-)
+        CFLAGS="${CFLAGS} -fprofile-arcs -ftest-coverage -O0 -g"
+        CXXFLAGS="${CXXFLAGS} -fprofile-arcs -ftest-coverage -O0 -g"
+        export CFLAGS CXXFLAGS
+        ;;
+      *) ;;  # This case should not be reached.
+    esac
+    ;;
+esac
+
+setup_ccache
+
+cd "${LIBWEBP_ROOT}"
+./autogen.sh
+
+cd "${BUILD_DIR}"
+"${LIBWEBP_ROOT}/configure" \
+  --host "${TARGET}" --build "$("${LIBWEBP_ROOT}/config.guess")" \
+  --enable-everything "${config_flags[@]}"
+make -j V=1

infra/compile_android.sh

@@ -0,0 +1,224 @@
+#!/bin/bash
+# Copyright (c) 2021, Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#
+#   * Neither the name of Google nor the names of its contributors may
+#     be used to endorse or promote products derived from this software
+#     without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+set -xe
+LIBWEBP_ROOT="$(realpath "$(dirname "$0")/..")"
+readonly LIBWEBP_ROOT
+readonly WORKSPACE=${WORKSPACE:-"$(mktemp -d -t webp.android.XXX)"}
+# shellcheck source=infra/common.sh
+source "${LIBWEBP_ROOT}/infra/common.sh"
+
+usage() {
+  cat << EOF
+Usage: $(basename "$0") BUILD_TYPE APP_ABI
+Options:
+BUILD_TYPE  supported build types:
+              static
+              static-debug
+              shared
+              shared-debug
+APP_ABI     supported application binary interfaces:
+              armeabi-v7a
+              arm64-v8a
+              x86
+              x86_64
+Environment variables:
+WORKSPACE   directory where the build is done.
+ANDROID_NDK_DIR directory where the android ndk tools are.
+EOF
+}
+
+################################################################################
+echo "Building libwebp for Android in ${WORKSPACE}"
+
+if [[ ! -d "${WORKSPACE}" ]]; then
+  log_err "${WORKSPACE} directory does not exist."
+  exit 1
+fi
+
+readonly BUILD_TYPE=${1:?"BUILD_TYPE is not defined.$(
+  echo
+  usage
+)"}
+readonly APP_ABI=${2:?"APP_ABI not defined.$(
+  echo
+  usage
+)"}
+readonly ANDROID_NDK_DIR=${ANDROID_NDK_DIR:?"ANDROID_NDK_DIR is not defined.$(
+  echo
+  usage
+)"}
+readonly BUILD_DIR="${WORKSPACE}/build-${BUILD_TYPE}"
+readonly STANDALONE_ANDROID_DIR="${WORKSPACE}/android"
+
+if [[ ! -x "${ANDROID_NDK_DIR}/ndk-build" ]]; then
+  log_err "unable to find ndk-build in ANDROID_NDK_DIR: ${ANDROID_NDK_DIR}."
+  exit 1
+fi
+
+CFLAGS=
+LDFLAGS=
+opts=()
+case "${BUILD_TYPE}" in
+  *debug)
+    readonly APP_OPTIM="debug"
+    CFLAGS="-O0 -g"
+    opts+=("--enable-asserts")
+    ;;
+  static* | shared*)
+    readonly APP_OPTIM="release"
+    CFLAGS="-O2 -g"
+    ;;
+  *)
+    usage
+    exit 1
+    ;;
+esac
+
+case "${BUILD_TYPE}" in
+  shared*) readonly SHARED="1" ;;
+  *)
+    readonly SHARED="0"
+    CFLAGS="${CFLAGS} -fPIE"
+    LDFLAGS="${LDFLAGS} -Wl,-pie"
+    opts+=("--disable-shared")
+    ;;
+esac
+
+# Create a fresh build directory
+make_build_dir "${BUILD_DIR}"
+cd "${BUILD_DIR}"
+ln -s "${LIBWEBP_ROOT}" jni
+
+"${ANDROID_NDK_DIR}/ndk-build" -j2 \
+  APP_ABI="${APP_ABI}" \
+  APP_OPTIM="${APP_OPTIM}" \
+  ENABLE_SHARED="${SHARED}"
+
+cd "${LIBWEBP_ROOT}"
+./autogen.sh
+
+case "${APP_ABI}" in
+  armeabi*) arch="arm" ;;
+  arm64*) arch="arm64" ;;
+  *) arch="${APP_ABI}" ;;
+esac
+# TODO(b/185520507): remove this and use the binaries from
+# toolchains/llvm/prebuilt/ directly.
+rm -rf "${STANDALONE_ANDROID_DIR}"
+"${ANDROID_NDK_DIR}/build/tools/make_standalone_toolchain.py" \
+  --api 24 --arch "${arch}" --stl gnustl --install-dir \
+  "${STANDALONE_ANDROID_DIR}"
+export PATH="${STANDALONE_ANDROID_DIR}/bin:${PATH}"
+
+rm -rf "${BUILD_DIR}"
+make_build_dir "${BUILD_DIR}"
+cd "${BUILD_DIR}"
+
+case "${arch}" in
+  arm)
+    host="arm-linux-androideabi"
+    case "${APP_ABI}" in
+      armeabi) ;;
+      armeabi-v7a)
+        CFLAGS="${CFLAGS} -march=armv7-a -mfpu=neon -mfloat-abi=softfp"
+        ;;
+      *) ;;  # No configuration needed
+    esac
+    ;;
+  arm64)
+    host="aarch64-linux-android"
+    ;;
+  x86)
+    host="i686-linux-android"
+    ;;
+  x86_64)
+    host="x86_64-linux-android"
+    ;;
+  *) ;;  # Skip configuration
+esac
+
+setup_ccache
+CC="clang"
+
+"${LIBWEBP_ROOT}/configure" --host "${host}" --build \
+  "$("${LIBWEBP_ROOT}/config.guess")" CC="${CC}" CFLAGS="${CFLAGS}" \
+  LDFLAGS="${LDFLAGS}" "${opts[@]}"
+make -j
+
+if [[ "${GERRIT_REFSPEC:-}" = "refs/heads/portable-intrinsics" ]] \
+  || [[ "${GERRIT_BRANCH:-}" = "portable-intrinsics" ]]; then
+  cd "${WORKSPACE}"
+  rm -rf build && mkdir build
+  cd build
+  standalone="${WORKSPACE}/android"
+  cmake ../libwebp \
+    -DWEBP_BUILD_DWEBP=1 \
+    -DCMAKE_C_COMPILER="${standalone}/bin/clang" \
+    -DCMAKE_PREFIX_PATH="${standalone}/sysroot/usr/lib" \
+    -DCMAKE_C_FLAGS=-fPIE \
+    -DCMAKE_EXE_LINKER_FLAGS=-Wl,-pie \
+    -DCMAKE_BUILD_TYPE=Release \
+    -DWEBP_ENABLE_WASM=1
+  make -j2
+
+  cd "${WORKSPACE}"
+  make_build_dir "${BUILD_DIR}"
+  cd "${BUILD_DIR}"
+  case "${APP_ABI}" in
+    armeabi-v7a | arm64*)
+      cmake "${LIBWEBP_ROOT}" \
+        -DWEBP_BUILD_DWEBP=1 \
+        -DCMAKE_C_COMPILER="${standalone}/bin/clang" \
+        -DCMAKE_PREFIX_PATH="${standalone}/sysroot/usr/lib" \
+        -DCMAKE_C_FLAGS='-fPIE -DENABLE_NEON_BUILTIN_MULHI_INT16X8' \
+        -DCMAKE_EXE_LINKER_FLAGS=-Wl,-pie \
+        -DCMAKE_BUILD_TYPE=Release \
+        -DWEBP_ENABLE_WASM=1
+      make -j2
+      ;;
+    x86*)
+      cmake "${LIBWEBP_ROOT}" \
+        -DWEBP_BUILD_DWEBP=1 \
+        -DCMAKE_C_COMPILER="${standalone}/bin/clang" \
+        -DCMAKE_PREFIX_PATH="${standalone}/sysroot/usr/lib" \
+        -DCMAKE_C_FLAGS='-fPIE -DENABLE_X86_BUILTIN_MULHI_INT16X8' \
+        -DCMAKE_EXE_LINKER_FLAGS=-Wl,-pie \
+        -DCMAKE_BUILD_TYPE=Release \
+        -DWEBP_ENABLE_WASM=1
+      make -j2
+      ;;
+    *)
+      log_err "APP_ABI not supported."
+      exit 1
+      ;;
+  esac
+fi

infra/compile_js.sh

@@ -0,0 +1,75 @@
+#!/bin/bash
+# Copyright (c) 2021, Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#
+#   * Neither the name of Google nor the names of its contributors may
+#     be used to endorse or promote products derived from this software
+#     without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+set -ex
+
+readonly WORKSPACE="${WORKSPACE:-"$(mktemp -d -t webp.js.XXX)"}"
+readonly BUILD_DIR="${WORKSPACE}/webp_js/"
+readonly LIBWEBP_ROOT="$(realpath "$(dirname "$0")/..")"
+
+# shellcheck source=infra/common.sh
+source "${LIBWEBP_ROOT}/infra/common.sh"
+
+usage() {
+  cat << EOF
+Usage: $(basename "$0")
+Environment variables:
+WORKSPACE directory where the build is done
+EMSDK_DIR directory where emsdk is installed
+EOF
+}
+
+[[ -d "${EMSDK_DIR:?Not defined}" ]] \
+  || (log_err "${EMSDK_DIR} is not a valid directory." && exit 1)
+
+# shellcheck source=/opt/emsdk/emsdk_env.sh
+source "${EMSDK_DIR}/emsdk_env.sh"
+
+readonly EMSCRIPTEN=${EMSCRIPTEN:-"${EMSDK}/upstream/emscripten"}
+readonly \
+  EMSCRIPTEN_CMAKE_FILE="${EMSCRIPTEN}/cmake/Modules/Platform/Emscripten.cmake"
+make_build_dir "${BUILD_DIR}"
+
+pushd "${BUILD_DIR}"
+opts=("-GUnix Makefiles" "-DWEBP_BUILD_WEBP_JS=ON")
+if [[ -z "$(command -v emcmake)" ]]; then
+  opts+=("-DCMAKE_TOOLCHAIN_FILE=${EMSCRIPTEN_CMAKE_FILE}")
+  cmake \
+    "${opts[@]}" \
+    "${LIBWEBP_ROOT}"
+  make -j
+else
+  emcmake cmake \
+    "${opts[@]}" \
+    "${LIBWEBP_ROOT}"
+  emmake make -j
+fi
+popd

infra/run_static_analysis.sh

@@ -0,0 +1,98 @@
+#!/bin/bash
+# Copyright (c) 2021, Google Inc. All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are
+# met:
+#
+#   * Redistributions of source code must retain the above copyright
+#     notice, this list of conditions and the following disclaimer.
+#
+#   * Redistributions in binary form must reproduce the above copyright
+#     notice, this list of conditions and the following disclaimer in
+#     the documentation and/or other materials provided with the
+#     distribution.
+#
+#   * Neither the name of Google nor the names of its contributors may
+#     be used to endorse or promote products derived from this software
+#     without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+set -xe
+
+LIBWEBP_ROOT="$(realpath "$(dirname "$0")/..")"
+readonly LIBWEBP_ROOT
+readonly WORKSPACE=${WORKSPACE:-"$(mktemp -d -t webp.scanbuild.XXX)"}
+
+# shellcheck source=infra/common.sh
+source "${LIBWEBP_ROOT}/infra/common.sh"
+
+usage() {
+  cat << EOF
+Usage: $(basename "$0") MODE
+Options:
+MODE supported scan modes: (shallow|deep)
+Environment variables:
+WORKSPACE directory where the build is done.
+EOF
+}
+
+#######################################
+# Wrap clang-tools scan-build.
+# Globals:
+#   OUTPUT_DIR target directory where scan-build report is generated.
+#   MODE scan-build mode
+# Arguments:
+#   $* scan-build additional args.
+# Returns:
+#   scan-build retcode
+#######################################
+scan_build() {
+  scan-build -o "${OUTPUT_DIR}" --use-analyzer="$(command -v clang)" \
+    -analyzer-config mode="${MODE}" "$*"
+}
+
+MODE=${1:?"MODE is not specified.$(
+  echo
+  usage
+)"}
+
+readonly OUTPUT_DIR="${WORKSPACE}/output-${MODE}"
+readonly BUILD_DIR="${WORKSPACE}/build"
+
+make_build_dir "${OUTPUT_DIR}"
+make_build_dir "${BUILD_DIR}"
+
+cd "${LIBWEBP_ROOT}"
+./autogen.sh
+
+cd "${BUILD_DIR}"
+grep -m 1 -q 'enable-asserts' "${LIBWEBP_ROOT}/configure.ac" \
+    && args='--enable-asserts'
+scan_build "${LIBWEBP_ROOT}/configure" --enable-everything "${args}"
+scan_build make -j4
+
+index="$(find "${OUTPUT_DIR}" -name index.html)"
+if [[ -f "${index}" ]]; then
+  mv "$(dirname "${index}")/"* .
+else
+  # make a empty report to wipe out any old bug reports.
+  cat << EOT > index.html
+<html>
+<body>
+No bugs reported.
+</body>
+</html>
+EOT
+fi

iosbuild.sh

@@ -9,7 +9,7 @@
 # (the previous build will be erased if it exists).
 #
 # This script is inspired by the build script written by Carson McDonald.
-# (http://www.ioncannon.net/programming/1483/using-webp-to-reduce-native-ios-app-size/).
+# (https://www.ioncannon.net/programming/1483/using-webp-to-reduce-native-ios-app-size/).
 
 set -e
 
@@ -86,7 +86,7 @@ if [[ ! -e ${SRCDIR}/configure ]]; then
 Error creating configure script!
 This script requires the autoconf/automake and libtool to build. MacPorts can
 be used to obtain these:
-http://www.macports.org/install.php
+https://www.macports.org/install.php
 EOF
     exit 1
   fi

makefile.unix

@@ -47,7 +47,7 @@ else
 endif
 
 # To install libraries on Mac OS X:
-# 1. Install MacPorts (http://www.macports.org/install.php)
+# 1. Install MacPorts (https://www.macports.org/install.php)
 # 2. Run "sudo port install jpeg"
 # 3. Run "sudo port install libpng"
 # 4. Run "sudo port install tiff"
@@ -166,6 +166,7 @@ DSP_DEC_OBJS = \
     src/dsp/lossless_msa.o \
     src/dsp/lossless_neon.o \
     src/dsp/lossless_sse2.o \
+    src/dsp/lossless_sse41.o \
     src/dsp/rescaler.o \
     src/dsp/rescaler_mips32.o \
     src/dsp/rescaler_mips_dsp_r2.o \

man/cwebp.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH CWEBP 1 "November 19, 2020"
+.TH CWEBP 1 "November 17, 2021"
 .SH NAME
 cwebp \- compress an image file to a WebP file
 .SH SYNOPSIS
@@ -222,7 +222,7 @@ Compute and report average PSNR (Peak\-Signal\-To\-Noise ratio).
 .TP
 .B \-print_ssim
 Compute and report average SSIM (structural similarity
-metric, see http://en.wikipedia.org/wiki/SSIM for additional details).
+metric, see https://en.wikipedia.org/wiki/SSIM for additional details).
 .TP
 .B \-print_lsim
 Compute and report local similarity metric (sum of lowest error amongst the
@@ -298,7 +298,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting\-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH EXAMPLES
 cwebp \-q 50 -lossless picture.png \-o picture_lossless.webp
@@ -322,5 +322,5 @@ for the Debian project (and may be used by others).
 .BR dwebp (1),
 .BR gif2webp (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.

man/dwebp.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH DWEBP 1 "November 19, 2020"
+.TH DWEBP 1 "November 17, 2021"
 .SH NAME
 dwebp \- decompress a WebP file to an image file
 .SH SYNOPSIS
@@ -113,7 +113,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH EXAMPLES
 dwebp picture.webp \-o output.png
@@ -138,7 +138,7 @@ for the Debian project (and may be used by others).
 .BR gif2webp (1),
 .BR webpmux (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.
 .SS Output file format details
 PAM: http://netpbm.sourceforge.net/doc/pam.html

man/gif2webp.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH GIF2WEBP 1 "May 1, 2020"
+.TH GIF2WEBP 1 "November 17, 2021"
 .SH NAME
 gif2webp \- Convert a GIF image to WebP
 .SH SYNOPSIS
@@ -131,7 +131,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH EXAMPLES
 gif2webp picture.gif \-o picture.webp
@@ -160,5 +160,5 @@ Debian project (and may be used by others).
 .BR dwebp (1),
 .BR webpmux (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.

man/img2webp.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH IMG2WEBP 1 "May 1, 2020"
+.TH IMG2WEBP 1 "November 17, 2021"
 .SH NAME
 img2webp \- create animated WebP file from a sequence of input images.
 .SH SYNOPSIS
@@ -86,7 +86,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting\-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH AUTHORS
 \fBimg2webp\fP is a part of libwebp and was written by the WebP team.
@@ -101,5 +101,5 @@ for the Debian project (and may be used by others).
 .BR webpmux (1),
 .BR gif2webp (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.

man/vwebp.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH VWEBP 1 "June 5, 2019"
+.TH VWEBP 1 "November 17, 2021"
 .SH NAME
 vwebp \- decompress a WebP file and display it in a window
 .SH SYNOPSIS
@@ -77,7 +77,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH EXAMPLES
 vwebp picture.webp
@@ -97,5 +97,5 @@ This manual page was written for the Debian project (and may be used by others).
 .SH SEE ALSO
 .BR dwebp (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.

man/webpinfo.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH WEBPINFO 1 "November 24, 2017"
+.TH WEBPINFO 1 "November 17, 2021"
 .SH NAME
 webpinfo \- print out the chunk level structure of WebP files
 along with basic integrity checks.
@@ -52,7 +52,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting\-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH EXAMPLES
 .br
@@ -76,5 +76,5 @@ for the Debian project (and may be used by others).
 .SH SEE ALSO
 .BR webpmux (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.

man/webpmux.1

@@ -1,5 +1,5 @@
 .\"                                      Hey, EMACS: -*- nroff -*-
-.TH WEBPMUX 1 "December 12, 2020"
+.TH WEBPMUX 1 "November 17, 2021"
 .SH NAME
 webpmux \- create animated WebP files from non\-animated WebP images, extract
 frames from animated WebP images, and manage XMP/EXIF metadata and ICC profile.
@@ -82,6 +82,12 @@ Set loop count on an animated file.
 .P
 Where: 'loop_count' must be in range [0, 65535].
 .TP
+.BI bgcolor " A,R,G,B
+Set the background color of the canvas on an animated file.
+.P
+where: 'A', 'R', 'G' and 'B' are integers in the range 0 to 255 specifying the
+Alpha, Red, Green and Blue component values respectively.
+.TP
 .BI icc " file.icc
 Set ICC profile.
 .P
@@ -185,7 +191,7 @@ Please report all bugs to the issue tracker:
 https://bugs.chromium.org/p/webp
 .br
 Patches welcome! See this page to get started:
-http://www.webmproject.org/code/contribute/submitting\-patches/
+https://www.webmproject.org/code/contribute/submitting\-patches/
 
 .SH EXAMPLES
 .P
@@ -261,5 +267,5 @@ for the Debian project (and may be used by others).
 .BR dwebp (1),
 .BR gif2webp (1)
 .br
-Please refer to http://developers.google.com/speed/webp/ for additional
+Please refer to https://developers.google.com/speed/webp/ for additional
 information.

src/Makefile.am

@@ -36,7 +36,7 @@ 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 8:1:1
+libwebp_la_LDFLAGS = -no-undefined -version-info 8:3:1
 libwebpincludedir = $(includedir)/webp
 pkgconfig_DATA = libwebp.pc
 
@@ -48,7 +48,7 @@ if BUILD_LIBWEBPDECODER
   libwebpdecoder_la_LIBADD += dsp/libwebpdspdecode.la
   libwebpdecoder_la_LIBADD += utils/libwebputilsdecode.la
 
-  libwebpdecoder_la_LDFLAGS = -no-undefined -version-info 4:1:1
+  libwebpdecoder_la_LDFLAGS = -no-undefined -version-info 4:3:1
   pkgconfig_DATA += libwebpdecoder.pc
 endif
 

src/dec/alpha_dec.c

@@ -183,7 +183,7 @@ const uint8_t* VP8DecompressAlphaRows(VP8Decoder* const dec,
   assert(dec != NULL && io != NULL);
 
   if (row < 0 || num_rows <= 0 || row + num_rows > height) {
-    return NULL;    // sanity check.
+    return NULL;
   }
 
   if (!dec->is_alpha_decoded_) {

src/dec/buffer_dec.c

@@ -102,7 +102,7 @@ static VP8StatusCode AllocateBuffer(WebPDecBuffer* const buffer) {
     int stride;
     uint64_t size;
 
-    if ((uint64_t)w * kModeBpp[mode] >= (1ull << 32)) {
+    if ((uint64_t)w * kModeBpp[mode] >= (1ull << 31)) {
       return VP8_STATUS_INVALID_PARAM;
     }
     stride = w * kModeBpp[mode];
@@ -117,7 +117,6 @@ static VP8StatusCode AllocateBuffer(WebPDecBuffer* const buffer) {
     }
     total_size = size + 2 * uv_size + a_size;
 
-    // Security/sanity checks
     output = (uint8_t*)WebPSafeMalloc(total_size, sizeof(*output));
     if (output == NULL) {
       return VP8_STATUS_OUT_OF_MEMORY;
@@ -156,11 +155,11 @@ VP8StatusCode WebPFlipBuffer(WebPDecBuffer* const buffer) {
   }
   if (WebPIsRGBMode(buffer->colorspace)) {
     WebPRGBABuffer* const buf = &buffer->u.RGBA;
-    buf->rgba += (buffer->height - 1) * buf->stride;
+    buf->rgba += (int64_t)(buffer->height - 1) * buf->stride;
     buf->stride = -buf->stride;
   } else {
     WebPYUVABuffer* const buf = &buffer->u.YUVA;
-    const int H = buffer->height;
+    const int64_t H = buffer->height;
     buf->y += (H - 1) * buf->y_stride;
     buf->y_stride = -buf->y_stride;
     buf->u += ((H - 1) >> 1) * buf->u_stride;
@@ -188,8 +187,7 @@ VP8StatusCode WebPAllocateDecBuffer(int width, int height,
       const int ch = options->crop_height;
       const int x = options->crop_left & ~1;
       const int y = options->crop_top & ~1;
-      if (x < 0 || y < 0 || cw <= 0 || ch <= 0 ||
-          x + cw > width || y + ch > height) {
+      if (!WebPCheckCropDimensions(width, height, x, y, cw, ch)) {
         return VP8_STATUS_INVALID_PARAM;   // out of frame boundary.
       }
       width = cw;

src/dec/frame_dec.c

@@ -705,7 +705,7 @@ static int AllocateMemory(VP8Decoder* const dec) {
                         + cache_size + alpha_size + WEBP_ALIGN_CST;
   uint8_t* mem;
 
-  if (needed != (size_t)needed) return 0;  // check for overflow
+  if (!CheckSizeOverflow(needed)) return 0;  // check for overflow
   if (needed > dec->mem_size_) {
     WebPSafeFree(dec->mem_);
     dec->mem_size_ = 0;

src/dec/io_dec.c

@@ -298,46 +298,57 @@ static int InitYUVRescaler(const VP8Io* const io, WebPDecParams* const p) {
   const int uv_out_height = (out_height + 1) >> 1;
   const int uv_in_width  = (io->mb_w + 1) >> 1;
   const int uv_in_height = (io->mb_h + 1) >> 1;
-  const size_t work_size = 2 * out_width;   // scratch memory for luma rescaler
+  // scratch memory for luma rescaler
+  const size_t work_size = 2 * (size_t)out_width;
   const size_t uv_work_size = 2 * uv_out_width;  // and for each u/v ones
-  size_t tmp_size, rescaler_size;
+  uint64_t total_size;
+  size_t rescaler_size;
   rescaler_t* work;
   WebPRescaler* scalers;
   const int num_rescalers = has_alpha ? 4 : 3;
 
-  tmp_size = (work_size + 2 * uv_work_size) * sizeof(*work);
+  total_size = ((uint64_t)work_size + 2 * uv_work_size) * sizeof(*work);
   if (has_alpha) {
-    tmp_size += work_size * sizeof(*work);
+    total_size += (uint64_t)work_size * sizeof(*work);
   }
   rescaler_size = num_rescalers * sizeof(*p->scaler_y) + WEBP_ALIGN_CST;
+  total_size += rescaler_size;
+  if (!CheckSizeOverflow(total_size)) {
+    return 0;
+  }
 
-  p->memory = WebPSafeMalloc(1ULL, tmp_size + rescaler_size);
+  p->memory = WebPSafeMalloc(1ULL, (size_t)total_size);
   if (p->memory == NULL) {
     return 0;   // memory error
   }
   work = (rescaler_t*)p->memory;
 
-  scalers = (WebPRescaler*)WEBP_ALIGN((const uint8_t*)work + tmp_size);
+  scalers = (WebPRescaler*)WEBP_ALIGN(
+      (const uint8_t*)work + total_size - rescaler_size);
   p->scaler_y = &scalers[0];
   p->scaler_u = &scalers[1];
   p->scaler_v = &scalers[2];
   p->scaler_a = has_alpha ? &scalers[3] : NULL;
 
-  WebPRescalerInit(p->scaler_y, io->mb_w, io->mb_h,
-                   buf->y, out_width, out_height, buf->y_stride, 1,
-                   work);
-  WebPRescalerInit(p->scaler_u, uv_in_width, uv_in_height,
-                   buf->u, uv_out_width, uv_out_height, buf->u_stride, 1,
-                   work + work_size);
-  WebPRescalerInit(p->scaler_v, uv_in_width, uv_in_height,
-                   buf->v, uv_out_width, uv_out_height, buf->v_stride, 1,
-                   work + work_size + uv_work_size);
+  if (!WebPRescalerInit(p->scaler_y, io->mb_w, io->mb_h,
+                        buf->y, out_width, out_height, buf->y_stride, 1,
+                        work) ||
+      !WebPRescalerInit(p->scaler_u, uv_in_width, uv_in_height,
+                        buf->u, uv_out_width, uv_out_height, buf->u_stride, 1,
+                        work + work_size) ||
+      !WebPRescalerInit(p->scaler_v, uv_in_width, uv_in_height,
+                        buf->v, uv_out_width, uv_out_height, buf->v_stride, 1,
+                        work + work_size + uv_work_size)) {
+    return 0;
+  }
   p->emit = EmitRescaledYUV;
 
   if (has_alpha) {
-    WebPRescalerInit(p->scaler_a, io->mb_w, io->mb_h,
-                     buf->a, out_width, out_height, buf->a_stride, 1,
-                     work + work_size + 2 * uv_work_size);
+    if (!WebPRescalerInit(p->scaler_a, io->mb_w, io->mb_h,
+                          buf->a, out_width, out_height, buf->a_stride, 1,
+                          work + work_size + 2 * uv_work_size)) {
+      return 0;
+    }
     p->emit_alpha = EmitRescaledAlphaYUV;
     WebPInitAlphaProcessing();
   }
@@ -480,51 +491,58 @@ static int InitRGBRescaler(const VP8Io* const io, WebPDecParams* const p) {
   const int out_height = io->scaled_height;
   const int uv_in_width  = (io->mb_w + 1) >> 1;
   const int uv_in_height = (io->mb_h + 1) >> 1;
-  const size_t work_size = 2 * out_width;   // scratch memory for one rescaler
+  // scratch memory for one rescaler
+  const size_t work_size = 2 * (size_t)out_width;
   rescaler_t* work;  // rescalers work area
   uint8_t* tmp;   // tmp storage for scaled YUV444 samples before RGB conversion
-  size_t tmp_size1, tmp_size2, total_size, rescaler_size;
+  uint64_t tmp_size1, tmp_size2, total_size;
+  size_t rescaler_size;
   WebPRescaler* scalers;
   const int num_rescalers = has_alpha ? 4 : 3;
 
-  tmp_size1 = 3 * work_size;
-  tmp_size2 = 3 * out_width;
-  if (has_alpha) {
-    tmp_size1 += work_size;
-    tmp_size2 += out_width;
-  }
+  tmp_size1 = (uint64_t)num_rescalers * work_size;
+  tmp_size2 = (uint64_t)num_rescalers * out_width;
   total_size = tmp_size1 * sizeof(*work) + tmp_size2 * sizeof(*tmp);
   rescaler_size = num_rescalers * sizeof(*p->scaler_y) + WEBP_ALIGN_CST;
+  total_size += rescaler_size;
+  if (!CheckSizeOverflow(total_size)) {
+    return 0;
+  }
 
-  p->memory = WebPSafeMalloc(1ULL, total_size + rescaler_size);
+  p->memory = WebPSafeMalloc(1ULL, (size_t)total_size);
   if (p->memory == NULL) {
     return 0;   // memory error
   }
   work = (rescaler_t*)p->memory;
   tmp = (uint8_t*)(work + tmp_size1);
 
-  scalers = (WebPRescaler*)WEBP_ALIGN((const uint8_t*)work + total_size);
+  scalers = (WebPRescaler*)WEBP_ALIGN(
+      (const uint8_t*)work + total_size - rescaler_size);
   p->scaler_y = &scalers[0];
   p->scaler_u = &scalers[1];
   p->scaler_v = &scalers[2];
   p->scaler_a = has_alpha ? &scalers[3] : NULL;
 
-  WebPRescalerInit(p->scaler_y, io->mb_w, io->mb_h,
-                   tmp + 0 * out_width, out_width, out_height, 0, 1,
-                   work + 0 * work_size);
-  WebPRescalerInit(p->scaler_u, uv_in_width, uv_in_height,
-                   tmp + 1 * out_width, out_width, out_height, 0, 1,
-                   work + 1 * work_size);
-  WebPRescalerInit(p->scaler_v, uv_in_width, uv_in_height,
-                   tmp + 2 * out_width, out_width, out_height, 0, 1,
-                   work + 2 * work_size);
+  if (!WebPRescalerInit(p->scaler_y, io->mb_w, io->mb_h,
+                        tmp + 0 * out_width, out_width, out_height, 0, 1,
+                        work + 0 * work_size) ||
+      !WebPRescalerInit(p->scaler_u, uv_in_width, uv_in_height,
+                        tmp + 1 * out_width, out_width, out_height, 0, 1,
+                        work + 1 * work_size) ||
+      !WebPRescalerInit(p->scaler_v, uv_in_width, uv_in_height,
+                        tmp + 2 * out_width, out_width, out_height, 0, 1,
+                        work + 2 * work_size)) {
+    return 0;
+  }
   p->emit = EmitRescaledRGB;
   WebPInitYUV444Converters();
 
   if (has_alpha) {
-    WebPRescalerInit(p->scaler_a, io->mb_w, io->mb_h,
-                     tmp + 3 * out_width, out_width, out_height, 0, 1,
-                     work + 3 * work_size);
+    if (!WebPRescalerInit(p->scaler_a, io->mb_w, io->mb_h,
+                          tmp + 3 * out_width, out_width, out_height, 0, 1,
+                          work + 3 * work_size)) {
+      return 0;
+    }
     p->emit_alpha = EmitRescaledAlphaRGB;
     if (p->output->colorspace == MODE_RGBA_4444 ||
         p->output->colorspace == MODE_rgbA_4444) {

src/dec/vp8_dec.c

@@ -335,7 +335,7 @@ int VP8GetHeaders(VP8Decoder* const dec, VP8Io* const io) {
     io->scaled_width = io->width;
     io->scaled_height = io->height;
 
-    io->mb_w = io->width;   // sanity check
+    io->mb_w = io->width;   // for soundness
     io->mb_h = io->height;  // ditto
 
     VP8ResetProba(&dec->proba_);
@@ -403,7 +403,7 @@ static const uint8_t kZigzag[16] = {
   0, 1, 4, 8,  5, 2, 3, 6,  9, 12, 13, 10,  7, 11, 14, 15
 };
 
-// See section 13-2: http://tools.ietf.org/html/rfc6386#section-13.2
+// See section 13-2: https://datatracker.ietf.org/doc/html/rfc6386#section-13.2
 static int GetLargeValue(VP8BitReader* const br, const uint8_t* const p) {
   int v;
   if (!VP8GetBit(br, p[3], "coeffs")) {

src/dec/vp8i_dec.h

@@ -32,7 +32,7 @@ extern "C" {
 // version numbers
 #define DEC_MAJ_VERSION 1
 #define DEC_MIN_VERSION 2
-#define DEC_REV_VERSION 0
+#define DEC_REV_VERSION 2
 
 // YUV-cache parameters. Cache is 32-bytes wide (= one cacheline).
 // Constraints are: We need to store one 16x16 block of luma samples (y),

src/dec/vp8l_dec.c

@@ -84,7 +84,7 @@ static const uint8_t kCodeToPlane[CODE_TO_PLANE_CODES] = {
 // to 256 (green component values) + 24 (length prefix values)
 // + color_cache_size (between 0 and 2048).
 // All values computed for 8-bit first level lookup with Mark Adler's tool:
-// http://www.hdfgroup.org/ftp/lib-external/zlib/zlib-1.2.5/examples/enough.c
+// https://github.com/madler/zlib/blob/v1.2.5/examples/enough.c
 #define FIXED_TABLE_SIZE (630 * 3 + 410)
 static const uint16_t kTableSize[12] = {
   FIXED_TABLE_SIZE + 654,
@@ -559,8 +559,11 @@ static int AllocateAndInitRescaler(VP8LDecoder* const dec, VP8Io* const io) {
   memory += work_size * sizeof(*work);
   scaled_data = (uint32_t*)memory;
 
-  WebPRescalerInit(dec->rescaler, in_width, in_height, (uint8_t*)scaled_data,
-                   out_width, out_height, 0, num_channels, work);
+  if (!WebPRescalerInit(dec->rescaler, in_width, in_height,
+                        (uint8_t*)scaled_data, out_width, out_height,
+                        0, num_channels, work)) {
+    return 0;
+  }
   return 1;
 }
 #endif   // WEBP_REDUCE_SIZE
@@ -574,13 +577,14 @@ static int AllocateAndInitRescaler(VP8LDecoder* const dec, VP8Io* const io) {
 static int Export(WebPRescaler* const rescaler, WEBP_CSP_MODE colorspace,
                   int rgba_stride, uint8_t* const rgba) {
   uint32_t* const src = (uint32_t*)rescaler->dst;
+  uint8_t* dst = rgba;
   const int dst_width = rescaler->dst_width;
   int num_lines_out = 0;
   while (WebPRescalerHasPendingOutput(rescaler)) {
-    uint8_t* const dst = rgba + num_lines_out * rgba_stride;
     WebPRescalerExportRow(rescaler);
     WebPMultARGBRow(src, dst_width, 1);
     VP8LConvertFromBGRA(src, dst_width, colorspace, dst);
+    dst += rgba_stride;
     ++num_lines_out;
   }
   return num_lines_out;
@@ -594,8 +598,8 @@ static int EmitRescaledRowsRGBA(const VP8LDecoder* const dec,
   int num_lines_in = 0;
   int num_lines_out = 0;
   while (num_lines_in < mb_h) {
-    uint8_t* const row_in = in + num_lines_in * in_stride;
-    uint8_t* const row_out = out + num_lines_out * out_stride;
+    uint8_t* const row_in = in + (uint64_t)num_lines_in * in_stride;
+    uint8_t* const row_out = out + (uint64_t)num_lines_out * out_stride;
     const int lines_left = mb_h - num_lines_in;
     const int needed_lines = WebPRescaleNeededLines(dec->rescaler, lines_left);
     int lines_imported;
@@ -796,7 +800,8 @@ static void ProcessRows(VP8LDecoder* const dec, int row) {
       const WebPDecBuffer* const output = dec->output_;
       if (WebPIsRGBMode(output->colorspace)) {  // convert to RGBA
         const WebPRGBABuffer* const buf = &output->u.RGBA;
-        uint8_t* const rgba = buf->rgba + dec->last_out_row_ * buf->stride;
+        uint8_t* const rgba =
+            buf->rgba + (int64_t)dec->last_out_row_ * buf->stride;
         const int num_rows_out =
 #if !defined(WEBP_REDUCE_SIZE)
          io->use_scaling ?
@@ -1514,7 +1519,7 @@ static int AllocateInternalBuffers32b(VP8LDecoder* const dec, int final_width) {
   assert(dec->width_ <= final_width);
   dec->pixels_ = (uint32_t*)WebPSafeMalloc(total_num_pixels, sizeof(uint32_t));
   if (dec->pixels_ == NULL) {
-    dec->argb_cache_ = NULL;    // for sanity check
+    dec->argb_cache_ = NULL;    // for soundness
     dec->status_ = VP8_STATUS_OUT_OF_MEMORY;
     return 0;
   }
@@ -1524,7 +1529,7 @@ static int AllocateInternalBuffers32b(VP8LDecoder* const dec, int final_width) {
 
 static int AllocateInternalBuffers8b(VP8LDecoder* const dec) {
   const uint64_t total_num_pixels = (uint64_t)dec->width_ * dec->height_;
-  dec->argb_cache_ = NULL;    // for sanity check
+  dec->argb_cache_ = NULL;    // for soundness
   dec->pixels_ = (uint32_t*)WebPSafeMalloc(total_num_pixels, sizeof(uint8_t));
   if (dec->pixels_ == NULL) {
     dec->status_ = VP8_STATUS_OUT_OF_MEMORY;
@@ -1666,7 +1671,6 @@ int VP8LDecodeImage(VP8LDecoder* const dec) {
   VP8Io* io = NULL;
   WebPDecParams* params = NULL;
 
-  // Sanity checks.
   if (dec == NULL) return 0;
 
   assert(dec->hdr_.huffman_tables_ != NULL);

src/dec/webp_dec.c

@@ -785,6 +785,13 @@ VP8StatusCode WebPDecode(const uint8_t* data, size_t data_size,
 //------------------------------------------------------------------------------
 // Cropping and rescaling.
 
+int WebPCheckCropDimensions(int image_width, int image_height,
+                            int x, int y, int w, int h) {
+  return !(x < 0 || y < 0 || w <= 0 || h <= 0 ||
+           x >= image_width || w > image_width || w > image_width - x ||
+           y >= image_height || h > image_height || h > image_height - y);
+}
+
 int WebPIoInitFromOptions(const WebPDecoderOptions* const options,
                           VP8Io* const io, WEBP_CSP_MODE src_colorspace) {
   const int W = io->width;
@@ -792,7 +799,7 @@ int WebPIoInitFromOptions(const WebPDecoderOptions* const options,
   int x = 0, y = 0, w = W, h = H;
 
   // Cropping
-  io->use_cropping = (options != NULL) && (options->use_cropping > 0);
+  io->use_cropping = (options != NULL) && options->use_cropping;
   if (io->use_cropping) {
     w = options->crop_width;
     h = options->crop_height;
@@ -802,7 +809,7 @@ int WebPIoInitFromOptions(const WebPDecoderOptions* const options,
       x &= ~1;
       y &= ~1;
     }
-    if (x < 0 || y < 0 || w <= 0 || h <= 0 || x + w > W || y + h > H) {
+    if (!WebPCheckCropDimensions(W, H, x, y, w, h)) {
       return 0;  // out of frame boundary error
     }
   }
@@ -814,7 +821,7 @@ int WebPIoInitFromOptions(const WebPDecoderOptions* const options,
   io->mb_h = h;
 
   // Scaling
-  io->use_scaling = (options != NULL) && (options->use_scaling > 0);
+  io->use_scaling = (options != NULL) && options->use_scaling;
   if (io->use_scaling) {
     int scaled_width = options->scaled_width;
     int scaled_height = options->scaled_height;
@@ -835,8 +842,8 @@ int WebPIoInitFromOptions(const WebPDecoderOptions* const options,
 
   if (io->use_scaling) {
     // disable filter (only for large downscaling ratio).
-    io->bypass_filtering = (io->scaled_width < W * 3 / 4) &&
-                           (io->scaled_height < H * 3 / 4);
+    io->bypass_filtering |= (io->scaled_width < W * 3 / 4) &&
+                            (io->scaled_height < H * 3 / 4);
     io->fancy_upsampling = 0;
   }
   return 1;

src/dec/webpi_dec.h

@@ -77,6 +77,10 @@ VP8StatusCode WebPParseHeaders(WebPHeaderStructure* const headers);
 //------------------------------------------------------------------------------
 // Misc utils
 
+// Returns true if crop dimensions are within image bounds.
+int WebPCheckCropDimensions(int image_width, int image_height,
+                            int x, int y, int w, int h);
+
 // Initializes VP8Io with custom setup, io and teardown functions. The default
 // hooks will use the supplied 'params' as io->opaque handle.
 void WebPInitCustomIo(WebPDecParams* const params, VP8Io* const io);

src/demux/Makefile.am

@@ -13,6 +13,6 @@ noinst_HEADERS =
 noinst_HEADERS += ../webp/format_constants.h
 
 libwebpdemux_la_LIBADD = ../libwebp.la
-libwebpdemux_la_LDFLAGS = -no-undefined -version-info 2:7:0
+libwebpdemux_la_LDFLAGS = -no-undefined -version-info 2:9:0
 libwebpdemuxincludedir = $(includedir)/webp
 pkgconfig_DATA = libwebpdemux.pc

src/demux/anim_decode.c

@@ -87,11 +87,19 @@ WebPAnimDecoder* WebPAnimDecoderNewInternal(
     int abi_version) {
   WebPAnimDecoderOptions options;
   WebPAnimDecoder* dec = NULL;
+  WebPBitstreamFeatures features;
   if (webp_data == NULL ||
       WEBP_ABI_IS_INCOMPATIBLE(abi_version, WEBP_DEMUX_ABI_VERSION)) {
     return NULL;
   }
 
+  // Validate the bitstream before doing expensive allocations. The demuxer may
+  // be more tolerant than the decoder.
+  if (WebPGetFeatures(webp_data->bytes, webp_data->size, &features) !=
+      VP8_STATUS_OK) {
+    return NULL;
+  }
+
   // Note: calloc() so that the pointer members are initialized to NULL.
   dec = (WebPAnimDecoder*)WebPSafeCalloc(1ULL, sizeof(*dec));
   if (dec == NULL) goto Error;
@@ -145,7 +153,7 @@ static int ZeroFillCanvas(uint8_t* buf, uint32_t canvas_width,
                           uint32_t canvas_height) {
   const uint64_t size =
       (uint64_t)canvas_width * canvas_height * NUM_CHANNELS * sizeof(*buf);
-  if (size != (size_t)size) return 0;
+  if (!CheckSizeOverflow(size)) return 0;
   memset(buf, 0, (size_t)size);
   return 1;
 }
@@ -166,7 +174,7 @@ static void ZeroFillFrameRect(uint8_t* buf, int buf_stride, int x_offset,
 static int CopyCanvas(const uint8_t* src, uint8_t* dst,
                       uint32_t width, uint32_t height) {
   const uint64_t size = (uint64_t)width * height * NUM_CHANNELS;
-  if (size != (size_t)size) return 0;
+  if (!CheckSizeOverflow(size)) return 0;
   assert(src != NULL && dst != NULL);
   memcpy(dst, src, (size_t)size);
   return 1;

src/demux/demux.c

@@ -25,7 +25,7 @@
 
 #define DMUX_MAJ_VERSION 1
 #define DMUX_MIN_VERSION 2
-#define DMUX_REV_VERSION 0
+#define DMUX_REV_VERSION 2
 
 typedef struct {
   size_t start_;        // start location of the data
@@ -221,12 +221,16 @@ static ParseStatus StoreFrame(int frame_num, uint32_t min_size,
     const size_t chunk_start_offset = mem->start_;
     const uint32_t fourcc = ReadLE32(mem);
     const uint32_t payload_size = ReadLE32(mem);
-    const uint32_t payload_size_padded = payload_size + (payload_size & 1);
-    const size_t payload_available = (payload_size_padded > MemDataSize(mem))
-                                   ? MemDataSize(mem) : payload_size_padded;
-    const size_t chunk_size = CHUNK_HEADER_SIZE + payload_available;
+    uint32_t payload_size_padded;
+    size_t payload_available;
+    size_t chunk_size;
 
     if (payload_size > MAX_CHUNK_PAYLOAD) return PARSE_ERROR;
+
+    payload_size_padded = payload_size + (payload_size & 1);
+    payload_available = (payload_size_padded > MemDataSize(mem))
+                      ? MemDataSize(mem) : payload_size_padded;
+    chunk_size = CHUNK_HEADER_SIZE + payload_available;
     if (SizeIsInvalid(mem, payload_size_padded)) return PARSE_ERROR;
     if (payload_size_padded > MemDataSize(mem)) status = PARSE_NEED_MORE_DATA;
 
@@ -451,9 +455,11 @@ static ParseStatus ParseVP8XChunks(WebPDemuxer* const dmux) {
     const size_t chunk_start_offset = mem->start_;
     const uint32_t fourcc = ReadLE32(mem);
     const uint32_t chunk_size = ReadLE32(mem);
-    const uint32_t chunk_size_padded = chunk_size + (chunk_size & 1);
+    uint32_t chunk_size_padded;
 
     if (chunk_size > MAX_CHUNK_PAYLOAD) return PARSE_ERROR;
+
+    chunk_size_padded = chunk_size + (chunk_size & 1);
     if (SizeIsInvalid(mem, chunk_size_padded)) return PARSE_ERROR;
 
     switch (fourcc) {

src/demux/libwebpdemux.rc

@@ -6,8 +6,8 @@
 LANGUAGE LANG_ENGLISH, SUBLANG_ENGLISH_US
 
 VS_VERSION_INFO VERSIONINFO
- FILEVERSION 1,0,2,0
- PRODUCTVERSION 1,0,2,0
+ FILEVERSION 1,0,2,2
+ PRODUCTVERSION 1,0,2,2
  FILEFLAGSMASK 0x3fL
 #ifdef _DEBUG
  FILEFLAGS 0x1L
@@ -24,12 +24,12 @@ BEGIN
         BEGIN
             VALUE "CompanyName", "Google, Inc."
             VALUE "FileDescription", "libwebpdemux DLL"
-            VALUE "FileVersion", "1.2.0"
+            VALUE "FileVersion", "1.2.2"
             VALUE "InternalName", "libwebpdemux.dll"
-            VALUE "LegalCopyright", "Copyright (C) 2019"
+            VALUE "LegalCopyright", "Copyright (C) 2021"
             VALUE "OriginalFilename", "libwebpdemux.dll"
             VALUE "ProductName", "WebP Image Demuxer"
-            VALUE "ProductVersion", "1.2.0"
+            VALUE "ProductVersion", "1.2.2"
         END
     END
     BLOCK "VarFileInfo"

src/dsp/Makefile.am

@@ -46,6 +46,7 @@ ENC_SOURCES += ssim.c
 libwebpdspdecode_sse41_la_SOURCES =
 libwebpdspdecode_sse41_la_SOURCES += alpha_processing_sse41.c
 libwebpdspdecode_sse41_la_SOURCES += dec_sse41.c
+libwebpdspdecode_sse41_la_SOURCES += lossless_sse41.c
 libwebpdspdecode_sse41_la_SOURCES += upsampling_sse41.c
 libwebpdspdecode_sse41_la_SOURCES += yuv_sse41.c
 libwebpdspdecode_sse41_la_CPPFLAGS = $(libwebpdsp_la_CPPFLAGS)

src/dsp/alpha_processing.c

@@ -157,7 +157,8 @@ void WebPMultARGBRow_C(uint32_t* const ptr, int width, int inverse) {
   }
 }
 
-void WebPMultRow_C(uint8_t* const ptr, const uint8_t* const alpha,
+void WebPMultRow_C(uint8_t* WEBP_RESTRICT const ptr,
+                   const uint8_t* WEBP_RESTRICT const alpha,
                    int width, int inverse) {
   int x;
   for (x = 0; x < width; ++x) {
@@ -178,7 +179,8 @@ void WebPMultRow_C(uint8_t* const ptr, const uint8_t* const alpha,
 #undef MFIX
 
 void (*WebPMultARGBRow)(uint32_t* const ptr, int width, int inverse);
-void (*WebPMultRow)(uint8_t* const ptr, const uint8_t* const alpha,
+void (*WebPMultRow)(uint8_t* WEBP_RESTRICT const ptr,
+                    const uint8_t* WEBP_RESTRICT const alpha,
                     int width, int inverse);
 
 //------------------------------------------------------------------------------
@@ -193,8 +195,8 @@ void WebPMultARGBRows(uint8_t* ptr, int stride, int width, int num_rows,
   }
 }
 
-void WebPMultRows(uint8_t* ptr, int stride,
-                  const uint8_t* alpha, int alpha_stride,
+void WebPMultRows(uint8_t* WEBP_RESTRICT ptr, int stride,
+                  const uint8_t* WEBP_RESTRICT alpha, int alpha_stride,
                   int width, int num_rows, int inverse) {
   int n;
   for (n = 0; n < num_rows; ++n) {
@@ -290,9 +292,9 @@ static void ApplyAlphaMultiply_16b_C(uint8_t* rgba4444,
 }
 
 #if !WEBP_NEON_OMIT_C_CODE
-static int DispatchAlpha_C(const uint8_t* alpha, int alpha_stride,
+static int DispatchAlpha_C(const uint8_t* WEBP_RESTRICT alpha, int alpha_stride,
                            int width, int height,
-                           uint8_t* dst, int dst_stride) {
+                           uint8_t* WEBP_RESTRICT dst, int dst_stride) {
   uint32_t alpha_mask = 0xff;
   int i, j;
 
@@ -309,9 +311,10 @@ static int DispatchAlpha_C(const uint8_t* alpha, int alpha_stride,
   return (alpha_mask != 0xff);
 }
 
-static void DispatchAlphaToGreen_C(const uint8_t* alpha, int alpha_stride,
-                                   int width, int height,
-                                   uint32_t* dst, int dst_stride) {
+static void DispatchAlphaToGreen_C(const uint8_t* WEBP_RESTRICT alpha,
+                                   int alpha_stride, int width, int height,
+                                   uint32_t* WEBP_RESTRICT dst,
+                                   int dst_stride) {
   int i, j;
   for (j = 0; j < height; ++j) {
     for (i = 0; i < width; ++i) {
@@ -322,9 +325,9 @@ static void DispatchAlphaToGreen_C(const uint8_t* alpha, int alpha_stride,
   }
 }
 
-static int ExtractAlpha_C(const uint8_t* argb, int argb_stride,
+static int ExtractAlpha_C(const uint8_t* WEBP_RESTRICT argb, int argb_stride,
                           int width, int height,
-                          uint8_t* alpha, int alpha_stride) {
+                          uint8_t* WEBP_RESTRICT alpha, int alpha_stride) {
   uint8_t alpha_mask = 0xff;
   int i, j;
 
@@ -340,7 +343,8 @@ static int ExtractAlpha_C(const uint8_t* argb, int argb_stride,
   return (alpha_mask == 0xff);
 }
 
-static void ExtractGreen_C(const uint32_t* argb, uint8_t* alpha, int size) {
+static void ExtractGreen_C(const uint32_t* WEBP_RESTRICT argb,
+                           uint8_t* WEBP_RESTRICT alpha, int size) {
   int i;
   for (i = 0; i < size; ++i) alpha[i] = argb[i] >> 8;
 }
@@ -372,8 +376,11 @@ static WEBP_INLINE uint32_t MakeARGB32(int a, int r, int g, int b) {
 }
 
 #ifdef WORDS_BIGENDIAN
-static void PackARGB_C(const uint8_t* a, const uint8_t* r, const uint8_t* g,
-                       const uint8_t* b, int len, uint32_t* out) {
+static void PackARGB_C(const uint8_t* WEBP_RESTRICT a,
+                       const uint8_t* WEBP_RESTRICT r,
+                       const uint8_t* WEBP_RESTRICT g,
+                       const uint8_t* WEBP_RESTRICT b,
+                       int len, uint32_t* WEBP_RESTRICT out) {
   int i;
   for (i = 0; i < len; ++i) {
     out[i] = MakeARGB32(a[4 * i], r[4 * i], g[4 * i], b[4 * i]);
@@ -381,8 +388,10 @@ static void PackARGB_C(const uint8_t* a, const uint8_t* r, const uint8_t* g,
 }
 #endif
 
-static void PackRGB_C(const uint8_t* r, const uint8_t* g, const uint8_t* b,
-                      int len, int step, uint32_t* out) {
+static void PackRGB_C(const uint8_t* WEBP_RESTRICT r,
+                      const uint8_t* WEBP_RESTRICT g,
+                      const uint8_t* WEBP_RESTRICT b,
+                      int len, int step, uint32_t* WEBP_RESTRICT out) {
   int i, offset = 0;
   for (i = 0; i < len; ++i) {
     out[i] = MakeARGB32(0xff, r[offset], g[offset], b[offset]);
@@ -392,16 +401,22 @@ static void PackRGB_C(const uint8_t* r, const uint8_t* g, const uint8_t* b,
 
 void (*WebPApplyAlphaMultiply)(uint8_t*, int, int, int, int);
 void (*WebPApplyAlphaMultiply4444)(uint8_t*, int, int, int);
-int (*WebPDispatchAlpha)(const uint8_t*, int, int, int, uint8_t*, int);
-void (*WebPDispatchAlphaToGreen)(const uint8_t*, int, int, int, uint32_t*, int);
-int (*WebPExtractAlpha)(const uint8_t*, int, int, int, uint8_t*, int);
-void (*WebPExtractGreen)(const uint32_t* argb, uint8_t* alpha, int size);
+int (*WebPDispatchAlpha)(const uint8_t* WEBP_RESTRICT, int, int, int,
+                         uint8_t* WEBP_RESTRICT, int);
+void (*WebPDispatchAlphaToGreen)(const uint8_t* WEBP_RESTRICT, int, int, int,
+                                 uint32_t* WEBP_RESTRICT, int);
+int (*WebPExtractAlpha)(const uint8_t* WEBP_RESTRICT, int, int, int,
+                        uint8_t* WEBP_RESTRICT, int);
+void (*WebPExtractGreen)(const uint32_t* WEBP_RESTRICT argb,
+                         uint8_t* WEBP_RESTRICT alpha, int size);
 #ifdef WORDS_BIGENDIAN
 void (*WebPPackARGB)(const uint8_t* a, const uint8_t* r, const uint8_t* g,
                      const uint8_t* b, int, uint32_t*);
 #endif
-void (*WebPPackRGB)(const uint8_t* r, const uint8_t* g, const uint8_t* b,
-                    int len, int step, uint32_t* out);
+void (*WebPPackRGB)(const uint8_t* WEBP_RESTRICT r,
+                    const uint8_t* WEBP_RESTRICT g,
+                    const uint8_t* WEBP_RESTRICT b,
+                    int len, int step, uint32_t* WEBP_RESTRICT out);
 
 int (*WebPHasAlpha8b)(const uint8_t* src, int length);
 int (*WebPHasAlpha32b)(const uint8_t* src, int length);
@@ -438,10 +453,10 @@ WEBP_DSP_INIT_FUNC(WebPInitAlphaProcessing) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       WebPInitAlphaProcessingSSE2();
-#if defined(WEBP_USE_SSE41)
+#if defined(WEBP_HAVE_SSE41)
       if (VP8GetCPUInfo(kSSE4_1)) {
         WebPInitAlphaProcessingSSE41();
       }
@@ -455,7 +470,7 @@ WEBP_DSP_INIT_FUNC(WebPInitAlphaProcessing) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     WebPInitAlphaProcessingNEON();

src/dsp/alpha_processing_neon.c

@@ -80,9 +80,9 @@ static void ApplyAlphaMultiply_NEON(uint8_t* rgba, int alpha_first,
 
 //------------------------------------------------------------------------------
 
-static int DispatchAlpha_NEON(const uint8_t* alpha, int alpha_stride,
-                              int width, int height,
-                              uint8_t* dst, int dst_stride) {
+static int DispatchAlpha_NEON(const uint8_t* WEBP_RESTRICT alpha,
+                              int alpha_stride, int width, int height,
+                              uint8_t* WEBP_RESTRICT dst, int dst_stride) {
   uint32_t alpha_mask = 0xffffffffu;
   uint8x8_t mask8 = vdup_n_u8(0xff);
   uint32_t tmp[2];
@@ -112,9 +112,10 @@ static int DispatchAlpha_NEON(const uint8_t* alpha, int alpha_stride,
   return (alpha_mask != 0xffffffffu);
 }
 
-static void DispatchAlphaToGreen_NEON(const uint8_t* alpha, int alpha_stride,
-                                      int width, int height,
-                                      uint32_t* dst, int dst_stride) {
+static void DispatchAlphaToGreen_NEON(const uint8_t* WEBP_RESTRICT alpha,
+                                      int alpha_stride, int width, int height,
+                                      uint32_t* WEBP_RESTRICT dst,
+                                      int dst_stride) {
   int i, j;
   uint8x8x4_t greens;   // leave A/R/B channels zero'd.
   greens.val[0] = vdup_n_u8(0);
@@ -131,9 +132,9 @@ static void DispatchAlphaToGreen_NEON(const uint8_t* alpha, int alpha_stride,
   }
 }
 
-static int ExtractAlpha_NEON(const uint8_t* argb, int argb_stride,
+static int ExtractAlpha_NEON(const uint8_t* WEBP_RESTRICT argb, int argb_stride,
                              int width, int height,
-                             uint8_t* alpha, int alpha_stride) {
+                             uint8_t* WEBP_RESTRICT alpha, int alpha_stride) {
   uint32_t alpha_mask = 0xffffffffu;
   uint8x8_t mask8 = vdup_n_u8(0xff);
   uint32_t tmp[2];
@@ -161,8 +162,8 @@ static int ExtractAlpha_NEON(const uint8_t* argb, int argb_stride,
   return (alpha_mask == 0xffffffffu);
 }
 
-static void ExtractGreen_NEON(const uint32_t* argb,
-                              uint8_t* alpha, int size) {
+static void ExtractGreen_NEON(const uint32_t* WEBP_RESTRICT argb,
+                              uint8_t* WEBP_RESTRICT alpha, int size) {
   int i;
   for (i = 0; i + 16 <= size; i += 16) {
     const uint8x16x4_t rgbX = vld4q_u8((const uint8_t*)(argb + i));

src/dsp/alpha_processing_sse2.c

@@ -18,9 +18,9 @@
 
 //------------------------------------------------------------------------------
 
-static int DispatchAlpha_SSE2(const uint8_t* alpha, int alpha_stride,
-                              int width, int height,
-                              uint8_t* dst, int dst_stride) {
+static int DispatchAlpha_SSE2(const uint8_t* WEBP_RESTRICT alpha,
+                              int alpha_stride, int width, int height,
+                              uint8_t* WEBP_RESTRICT dst, int dst_stride) {
   // alpha_and stores an 'and' operation of all the alpha[] values. The final
   // value is not 0xff if any of the alpha[] is not equal to 0xff.
   uint32_t alpha_and = 0xff;
@@ -72,9 +72,10 @@ static int DispatchAlpha_SSE2(const uint8_t* alpha, int alpha_stride,
   return (alpha_and != 0xff);
 }
 
-static void DispatchAlphaToGreen_SSE2(const uint8_t* alpha, int alpha_stride,
-                                      int width, int height,
-                                      uint32_t* dst, int dst_stride) {
+static void DispatchAlphaToGreen_SSE2(const uint8_t* WEBP_RESTRICT alpha,
+                                      int alpha_stride, int width, int height,
+                                      uint32_t* WEBP_RESTRICT dst,
+                                      int dst_stride) {
   int i, j;
   const __m128i zero = _mm_setzero_si128();
   const int limit = width & ~15;
@@ -98,9 +99,9 @@ static void DispatchAlphaToGreen_SSE2(const uint8_t* alpha, int alpha_stride,
   }
 }
 
-static int ExtractAlpha_SSE2(const uint8_t* argb, int argb_stride,
+static int ExtractAlpha_SSE2(const uint8_t* WEBP_RESTRICT argb, int argb_stride,
                              int width, int height,
-                             uint8_t* alpha, int alpha_stride) {
+                             uint8_t* WEBP_RESTRICT alpha, int alpha_stride) {
   // alpha_and stores an 'and' operation of all the alpha[] values. The final
   // value is not 0xff if any of the alpha[] is not equal to 0xff.
   uint32_t alpha_and = 0xff;
@@ -317,7 +318,8 @@ static void MultARGBRow_SSE2(uint32_t* const ptr, int width, int inverse) {
   if (width > 0) WebPMultARGBRow_C(ptr + x, width, inverse);
 }
 
-static void MultRow_SSE2(uint8_t* const ptr, const uint8_t* const alpha,
+static void MultRow_SSE2(uint8_t* WEBP_RESTRICT const ptr,
+                         const uint8_t* WEBP_RESTRICT const alpha,
                          int width, int inverse) {
   int x = 0;
   if (!inverse) {

src/dsp/alpha_processing_sse41.c

@@ -19,9 +19,9 @@
 
 //------------------------------------------------------------------------------
 
-static int ExtractAlpha_SSE41(const uint8_t* argb, int argb_stride,
-                              int width, int height,
-                              uint8_t* alpha, int alpha_stride) {
+static int ExtractAlpha_SSE41(const uint8_t* WEBP_RESTRICT argb,
+                              int argb_stride, int width, int height,
+                              uint8_t* WEBP_RESTRICT alpha, int alpha_stride) {
   // alpha_and stores an 'and' operation of all the alpha[] values. The final
   // value is not 0xff if any of the alpha[] is not equal to 0xff.
   uint32_t alpha_and = 0xff;

src/dsp/cost.c

@@ -395,12 +395,12 @@ WEBP_DSP_INIT_FUNC(VP8EncDspCostInit) {
       VP8EncDspCostInitMIPSdspR2();
     }
 #endif
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8EncDspCostInitSSE2();
     }
 #endif
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
     if (VP8GetCPUInfo(kNEON)) {
       VP8EncDspCostInitNEON();
     }

src/dsp/cpu.c

@@ -189,17 +189,17 @@ VP8CPUInfo VP8GetCPUInfo = AndroidCPUInfo;
 // Use compile flags as an indicator of SIMD support instead of a runtime check.
 static int wasmCPUInfo(CPUFeature feature) {
   switch (feature) {
-#ifdef WEBP_USE_SSE2
+#ifdef WEBP_HAVE_SSE2
     case kSSE2:
       return 1;
 #endif
-#ifdef WEBP_USE_SSE41
+#ifdef WEBP_HAVE_SSE41
     case kSSE3:
     case kSlowSSSE3:
     case kSSE4_1:
       return 1;
 #endif
-#ifdef WEBP_USE_NEON
+#ifdef WEBP_HAVE_NEON
     case kNEON:
       return 1;
 #endif
@@ -209,9 +209,10 @@ static int wasmCPUInfo(CPUFeature feature) {
   return 0;
 }
 VP8CPUInfo VP8GetCPUInfo = wasmCPUInfo;
-#elif defined(WEBP_USE_NEON)
-// define a dummy function to enable turning off NEON at runtime by setting
-// VP8DecGetCPUInfo = NULL
+#elif defined(WEBP_HAVE_NEON)
+// In most cases this function doesn't check for NEON support (it's assumed by
+// the configuration), but enables turning off NEON at runtime, for testing
+// purposes, by setting VP8DecGetCPUInfo = NULL.
 static int armCPUInfo(CPUFeature feature) {
   if (feature != kNEON) return 0;
 #if defined(__linux__) && defined(WEBP_HAVE_NEON_RTCD)

src/dsp/dec.c

@@ -807,10 +807,10 @@ WEBP_DSP_INIT_FUNC(VP8DspInit) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8DspInitSSE2();
-#if defined(WEBP_USE_SSE41)
+#if defined(WEBP_HAVE_SSE41)
       if (VP8GetCPUInfo(kSSE4_1)) {
         VP8DspInitSSE41();
       }
@@ -834,7 +834,7 @@ WEBP_DSP_INIT_FUNC(VP8DspInit) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     VP8DspInitNEON();

src/dsp/dsp.h

@@ -26,6 +26,23 @@ extern "C" {
 
 #define BPS 32   // this is the common stride for enc/dec
 
+//------------------------------------------------------------------------------
+// WEBP_RESTRICT
+
+// Declares a pointer with the restrict type qualifier if available.
+// This allows code to hint to the compiler that only this pointer references a
+// particular object or memory region within the scope of the block in which it
+// is declared. This may allow for improved optimizations due to the lack of
+// pointer aliasing. See also:
+// https://en.cppreference.com/w/c/language/restrict
+#if defined(__GNUC__)
+#define WEBP_RESTRICT __restrict__
+#elif defined(_MSC_VER)
+#define WEBP_RESTRICT __restrict
+#else
+#define WEBP_RESTRICT
+#endif
+
 //------------------------------------------------------------------------------
 // CPU detection
 
@@ -67,21 +84,31 @@ extern "C" {
 // files without intrinsics, allowing the corresponding Init() to be called.
 // Files containing intrinsics will need to be built targeting the instruction
 // set so should succeed on one of the earlier tests.
-#if defined(__SSE2__) || defined(WEBP_MSC_SSE2) || defined(WEBP_HAVE_SSE2)
+#if (defined(__SSE2__) || defined(WEBP_MSC_SSE2)) && \
+    (!defined(HAVE_CONFIG_H) || defined(WEBP_HAVE_SSE2))
 #define WEBP_USE_SSE2
 #endif
 
-#if defined(__SSE4_1__) || defined(WEBP_MSC_SSE41) || defined(WEBP_HAVE_SSE41)
+#if defined(WEBP_USE_SSE2) && !defined(WEBP_HAVE_SSE2)
+#define WEBP_HAVE_SSE2
+#endif
+
+#if (defined(__SSE4_1__) || defined(WEBP_MSC_SSE41)) && \
+    (!defined(HAVE_CONFIG_H) || defined(WEBP_HAVE_SSE41))
 #define WEBP_USE_SSE41
 #endif
 
+#if defined(WEBP_USE_SSE41) && !defined(WEBP_HAVE_SSE41)
+#define WEBP_HAVE_SSE41
+#endif
+
 #undef WEBP_MSC_SSE41
 #undef WEBP_MSC_SSE2
 
 // 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(__aarch64__) || defined(WEBP_HAVE_NEON)) && \
+#if ((defined(__ARM_NEON__) || defined(__aarch64__)) && \
+     (!defined(HAVE_CONFIG_H) || defined(WEBP_HAVE_NEON))) && \
     !defined(__native_client__)
 #define WEBP_USE_NEON
 #endif
@@ -92,11 +119,20 @@ extern "C" {
 #define WEBP_USE_NEON
 #endif
 
-#if defined(_MSC_VER) && _MSC_VER >= 1700 && defined(_M_ARM)
+// Note: ARM64 is supported in Visual Studio 2017, but requires the direct
+// inclusion of arm64_neon.h; Visual Studio 2019 includes this file in
+// arm_neon.h.
+#if defined(_MSC_VER) && \
+  ((_MSC_VER >= 1700 && defined(_M_ARM)) || \
+   (_MSC_VER >= 1920 && defined(_M_ARM64)))
 #define WEBP_USE_NEON
 #define WEBP_USE_INTRINSICS
 #endif
 
+#if defined(WEBP_USE_NEON) && !defined(WEBP_HAVE_NEON)
+#define WEBP_HAVE_NEON
+#endif
+
 #if defined(__mips__) && !defined(__mips64) && \
     defined(__mips_isa_rev) && (__mips_isa_rev >= 1) && (__mips_isa_rev < 6)
 #define WEBP_USE_MIPS32
@@ -116,7 +152,7 @@ extern "C" {
 #define WEBP_DSP_OMIT_C_CODE 1
 #endif
 
-#if (defined(__aarch64__) || defined(__ARM_NEON__)) && WEBP_DSP_OMIT_C_CODE
+#if defined(WEBP_USE_NEON) && WEBP_DSP_OMIT_C_CODE
 #define WEBP_NEON_OMIT_C_CODE 1
 #else
 #define WEBP_NEON_OMIT_C_CODE 0
@@ -578,26 +614,29 @@ extern void (*WebPApplyAlphaMultiply4444)(
 
 // Dispatch the values from alpha[] plane to the ARGB destination 'dst'.
 // Returns true if alpha[] plane has non-trivial values different from 0xff.
-extern int (*WebPDispatchAlpha)(const uint8_t* alpha, int alpha_stride,
-                                int width, int height,
-                                uint8_t* dst, int dst_stride);
+extern int (*WebPDispatchAlpha)(const uint8_t* WEBP_RESTRICT alpha,
+                                int alpha_stride, int width, int height,
+                                uint8_t* WEBP_RESTRICT dst, int dst_stride);
 
 // Transfer packed 8b alpha[] values to green channel in dst[], zero'ing the
 // A/R/B values. 'dst_stride' is the stride for dst[] in uint32_t units.
-extern void (*WebPDispatchAlphaToGreen)(const uint8_t* alpha, int alpha_stride,
-                                        int width, int height,
-                                        uint32_t* dst, int dst_stride);
+extern void (*WebPDispatchAlphaToGreen)(const uint8_t* WEBP_RESTRICT alpha,
+                                        int alpha_stride, int width, int height,
+                                        uint32_t* WEBP_RESTRICT dst,
+                                        int dst_stride);
 
 // Extract the alpha values from 32b values in argb[] and pack them into alpha[]
 // (this is the opposite of WebPDispatchAlpha).
 // Returns true if there's only trivial 0xff alpha values.
-extern int (*WebPExtractAlpha)(const uint8_t* argb, int argb_stride,
-                               int width, int height,
-                               uint8_t* alpha, int alpha_stride);
+extern int (*WebPExtractAlpha)(const uint8_t* WEBP_RESTRICT argb,
+                               int argb_stride, int width, int height,
+                               uint8_t* WEBP_RESTRICT alpha,
+                               int alpha_stride);
 
 // Extract the green values from 32b values in argb[] and pack them into alpha[]
 // (this is the opposite of WebPDispatchAlphaToGreen).
-extern void (*WebPExtractGreen)(const uint32_t* argb, uint8_t* alpha, int size);
+extern void (*WebPExtractGreen)(const uint32_t* WEBP_RESTRICT argb,
+                                uint8_t* WEBP_RESTRICT alpha, int size);
 
 // Pre-Multiply operation transforms x into x * A / 255  (where x=Y,R,G or B).
 // Un-Multiply operation transforms x into x * 255 / A.
@@ -610,29 +649,35 @@ void WebPMultARGBRows(uint8_t* ptr, int stride, int width, int num_rows,
                       int inverse);
 
 // Same for a row of single values, with side alpha values.
-extern void (*WebPMultRow)(uint8_t* const ptr, const uint8_t* const alpha,
+extern void (*WebPMultRow)(uint8_t* WEBP_RESTRICT const ptr,
+                           const uint8_t* WEBP_RESTRICT const alpha,
                            int width, int inverse);
 
 // Same a WebPMultRow(), but for several 'num_rows' rows.
-void WebPMultRows(uint8_t* ptr, int stride,
-                  const uint8_t* alpha, int alpha_stride,
+void WebPMultRows(uint8_t* WEBP_RESTRICT ptr, int stride,
+                  const uint8_t* WEBP_RESTRICT alpha, int alpha_stride,
                   int width, int num_rows, int inverse);
 
 // Plain-C versions, used as fallback by some implementations.
-void WebPMultRow_C(uint8_t* const ptr, const uint8_t* const alpha,
+void WebPMultRow_C(uint8_t* WEBP_RESTRICT const ptr,
+                   const uint8_t* WEBP_RESTRICT const alpha,
                    int width, int inverse);
 void WebPMultARGBRow_C(uint32_t* const ptr, int width, int inverse);
 
 #ifdef WORDS_BIGENDIAN
 // ARGB packing function: a/r/g/b input is rgba or bgra order.
-extern void (*WebPPackARGB)(const uint8_t* a, const uint8_t* r,
-                            const uint8_t* g, const uint8_t* b, int len,
-                            uint32_t* out);
+extern void (*WebPPackARGB)(const uint8_t* WEBP_RESTRICT a,
+                            const uint8_t* WEBP_RESTRICT r,
+                            const uint8_t* WEBP_RESTRICT g,
+                            const uint8_t* WEBP_RESTRICT b,
+                            int len, uint32_t* WEBP_RESTRICT out);
 #endif
 
 // RGB packing function. 'step' can be 3 or 4. r/g/b input is rgb or bgr order.
-extern void (*WebPPackRGB)(const uint8_t* r, const uint8_t* g, const uint8_t* b,
-                           int len, int step, uint32_t* out);
+extern void (*WebPPackRGB)(const uint8_t* WEBP_RESTRICT r,
+                           const uint8_t* WEBP_RESTRICT g,
+                           const uint8_t* WEBP_RESTRICT b,
+                           int len, int step, uint32_t* WEBP_RESTRICT out);
 
 // This function returns true if src[i] contains a value different from 0xff.
 extern int (*WebPHasAlpha8b)(const uint8_t* src, int length);

src/dsp/enc.c

@@ -773,10 +773,10 @@ WEBP_DSP_INIT_FUNC(VP8EncDspInit) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8EncDspInitSSE2();
-#if defined(WEBP_USE_SSE41)
+#if defined(WEBP_HAVE_SSE41)
       if (VP8GetCPUInfo(kSSE4_1)) {
         VP8EncDspInitSSE41();
       }
@@ -800,7 +800,7 @@ WEBP_DSP_INIT_FUNC(VP8EncDspInit) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     VP8EncDspInitNEON();

src/dsp/enc_neon.c

@@ -9,7 +9,7 @@
 //
 // ARM NEON version of speed-critical encoding functions.
 //
-// adapted from libvpx (http://www.webmproject.org/code/)
+// adapted from libvpx (https://www.webmproject.org/code/)
 
 #include "src/dsp/dsp.h"
 

src/dsp/filters.c

@@ -254,7 +254,7 @@ WEBP_DSP_INIT_FUNC(VP8FiltersInit) {
 #endif
 
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8FiltersInitSSE2();
     }
@@ -271,7 +271,7 @@ WEBP_DSP_INIT_FUNC(VP8FiltersInit) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     VP8FiltersInitNEON();

src/dsp/filters_sse2.c

@@ -320,7 +320,12 @@ extern void VP8FiltersInitSSE2(void);
 
 WEBP_TSAN_IGNORE_FUNCTION void VP8FiltersInitSSE2(void) {
   WebPUnfilters[WEBP_FILTER_HORIZONTAL] = HorizontalUnfilter_SSE2;
+#if defined(CHROMIUM)
+  // TODO(crbug.com/654974)
+  (void)VerticalUnfilter_SSE2;
+#else
   WebPUnfilters[WEBP_FILTER_VERTICAL] = VerticalUnfilter_SSE2;
+#endif
   WebPUnfilters[WEBP_FILTER_GRADIENT] = GradientUnfilter_SSE2;
 
   WebPFilters[WEBP_FILTER_HORIZONTAL] = HorizontalFilter_SSE2;

src/dsp/lossless.c

@@ -107,63 +107,77 @@ static WEBP_INLINE uint32_t Select(uint32_t a, uint32_t b, uint32_t c) {
 //------------------------------------------------------------------------------
 // Predictors
 
-uint32_t VP8LPredictor0_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor0_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   (void)top;
   (void)left;
   return ARGB_BLACK;
 }
-uint32_t VP8LPredictor1_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor1_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   (void)top;
-  return left;
+  return *left;
 }
-uint32_t VP8LPredictor2_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor2_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   (void)left;
   return top[0];
 }
-uint32_t VP8LPredictor3_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor3_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   (void)left;
   return top[1];
 }
-uint32_t VP8LPredictor4_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor4_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   (void)left;
   return top[-1];
 }
-uint32_t VP8LPredictor5_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average3(left, top[0], top[1]);
+uint32_t VP8LPredictor5_C(const uint32_t* const left,
+                          const uint32_t* const top) {
+  const uint32_t pred = Average3(*left, top[0], top[1]);
   return pred;
 }
-uint32_t VP8LPredictor6_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average2(left, top[-1]);
+uint32_t VP8LPredictor6_C(const uint32_t* const left,
+                          const uint32_t* const top) {
+  const uint32_t pred = Average2(*left, top[-1]);
   return pred;
 }
-uint32_t VP8LPredictor7_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average2(left, top[0]);
+uint32_t VP8LPredictor7_C(const uint32_t* const left,
+                          const uint32_t* const top) {
+  const uint32_t pred = Average2(*left, top[0]);
   return pred;
 }
-uint32_t VP8LPredictor8_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor8_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   const uint32_t pred = Average2(top[-1], top[0]);
   (void)left;
   return pred;
 }
-uint32_t VP8LPredictor9_C(uint32_t left, const uint32_t* const top) {
+uint32_t VP8LPredictor9_C(const uint32_t* const left,
+                          const uint32_t* const top) {
   const uint32_t pred = Average2(top[0], top[1]);
   (void)left;
   return pred;
 }
-uint32_t VP8LPredictor10_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average4(left, top[-1], top[0], top[1]);
+uint32_t VP8LPredictor10_C(const uint32_t* const left,
+                           const uint32_t* const top) {
+  const uint32_t pred = Average4(*left, top[-1], top[0], top[1]);
   return pred;
 }
-uint32_t VP8LPredictor11_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Select(top[0], left, top[-1]);
+uint32_t VP8LPredictor11_C(const uint32_t* const left,
+                           const uint32_t* const top) {
+  const uint32_t pred = Select(top[0], *left, top[-1]);
   return pred;
 }
-uint32_t VP8LPredictor12_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = ClampedAddSubtractFull(left, top[0], top[-1]);
+uint32_t VP8LPredictor12_C(const uint32_t* const left,
+                           const uint32_t* const top) {
+  const uint32_t pred = ClampedAddSubtractFull(*left, top[0], top[-1]);
   return pred;
 }
-uint32_t VP8LPredictor13_C(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = ClampedAddSubtractHalf(left, top[0], top[-1]);
+uint32_t VP8LPredictor13_C(const uint32_t* const left,
+                           const uint32_t* const top) {
+  const uint32_t pred = ClampedAddSubtractHalf(*left, top[0], top[-1]);
   return pred;
 }
 
@@ -575,6 +589,7 @@ VP8LMapARGBFunc VP8LMapColor32b;
 VP8LMapAlphaFunc VP8LMapColor8b;
 
 extern void VP8LDspInitSSE2(void);
+extern void VP8LDspInitSSE41(void);
 extern void VP8LDspInitNEON(void);
 extern void VP8LDspInitMIPSdspR2(void);
 extern void VP8LDspInitMSA(void);
@@ -621,9 +636,14 @@ WEBP_DSP_INIT_FUNC(VP8LDspInit) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8LDspInitSSE2();
+#if defined(WEBP_HAVE_SSE41)
+      if (VP8GetCPUInfo(kSSE4_1)) {
+        VP8LDspInitSSE41();
+      }
+#endif
     }
 #endif
 #if defined(WEBP_USE_MIPS_DSP_R2)
@@ -638,7 +658,7 @@ WEBP_DSP_INIT_FUNC(VP8LDspInit) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     VP8LDspInitNEON();

src/dsp/lossless.h

@@ -28,23 +28,38 @@ extern "C" {
 //------------------------------------------------------------------------------
 // Decoding
 
-typedef uint32_t (*VP8LPredictorFunc)(uint32_t left, const uint32_t* const top);
+typedef uint32_t (*VP8LPredictorFunc)(const uint32_t* const left,
+                                      const uint32_t* const top);
 extern VP8LPredictorFunc VP8LPredictors[16];
 
-uint32_t VP8LPredictor0_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor1_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor2_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor3_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor4_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor5_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor6_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor7_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor8_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor9_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor10_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor11_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor12_C(uint32_t left, const uint32_t* const top);
-uint32_t VP8LPredictor13_C(uint32_t left, const uint32_t* const top);
+uint32_t VP8LPredictor0_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor1_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor2_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor3_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor4_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor5_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor6_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor7_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor8_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor9_C(const uint32_t* const left,
+                          const uint32_t* const top);
+uint32_t VP8LPredictor10_C(const uint32_t* const left,
+                           const uint32_t* const top);
+uint32_t VP8LPredictor11_C(const uint32_t* const left,
+                           const uint32_t* const top);
+uint32_t VP8LPredictor12_C(const uint32_t* const left,
+                           const uint32_t* const top);
+uint32_t VP8LPredictor13_C(const uint32_t* const left,
+                           const uint32_t* const top);
 
 // These Add/Sub function expects upper[-1] and out[-1] to be readable.
 typedef void (*VP8LPredictorAddSubFunc)(const uint32_t* in,

src/dsp/lossless_common.h

@@ -179,7 +179,7 @@ static void PREDICTOR_ADD(const uint32_t* in, const uint32_t* upper, \
   int x;                                                             \
   assert(upper != NULL);                                             \
   for (x = 0; x < num_pixels; ++x) {                                 \
-    const uint32_t pred = (PREDICTOR)(out[x - 1], upper + x);        \
+    const uint32_t pred = (PREDICTOR)(&out[x - 1], upper + x);       \
     out[x] = VP8LAddPixels(in[x], pred);                             \
   }                                                                  \
 }

src/dsp/lossless_enc.c

@@ -329,6 +329,15 @@ const uint8_t kPrefixEncodeExtraBitsValue[PREFIX_LOOKUP_IDX_MAX] = {
 static float FastSLog2Slow_C(uint32_t v) {
   assert(v >= LOG_LOOKUP_IDX_MAX);
   if (v < APPROX_LOG_WITH_CORRECTION_MAX) {
+#if !defined(WEBP_HAVE_SLOW_CLZ_CTZ)
+    // use clz if available
+    const int log_cnt = BitsLog2Floor(v) - 7;
+    const uint32_t y = 1 << log_cnt;
+    int correction = 0;
+    const float v_f = (float)v;
+    const uint32_t orig_v = v;
+    v >>= log_cnt;
+#else
     int log_cnt = 0;
     uint32_t y = 1;
     int correction = 0;
@@ -339,6 +348,7 @@ static float FastSLog2Slow_C(uint32_t v) {
       v = v >> 1;
       y = y << 1;
     } while (v >= LOG_LOOKUP_IDX_MAX);
+#endif
     // vf = (2^log_cnt) * Xf; where y = 2^log_cnt and Xf < 256
     // Xf = floor(Xf) * (1 + (v % y) / v)
     // log2(Xf) = log2(floor(Xf)) + log2(1 + (v % y) / v)
@@ -355,6 +365,14 @@ static float FastSLog2Slow_C(uint32_t v) {
 static float FastLog2Slow_C(uint32_t v) {
   assert(v >= LOG_LOOKUP_IDX_MAX);
   if (v < APPROX_LOG_WITH_CORRECTION_MAX) {
+#if !defined(WEBP_HAVE_SLOW_CLZ_CTZ)
+    // use clz if available
+    const int log_cnt = BitsLog2Floor(v) - 7;
+    const uint32_t y = 1 << log_cnt;
+    const uint32_t orig_v = v;
+    double log_2;
+    v >>= log_cnt;
+#else
     int log_cnt = 0;
     uint32_t y = 1;
     const uint32_t orig_v = v;
@@ -364,6 +382,7 @@ static float FastLog2Slow_C(uint32_t v) {
       v = v >> 1;
       y = y << 1;
     } while (v >= LOG_LOOKUP_IDX_MAX);
+#endif
     log_2 = kLog2Table[v] + log_cnt;
     if (orig_v >= APPROX_LOG_MAX) {
       // Since the division is still expensive, add this correction factor only
@@ -726,7 +745,7 @@ static void PredictorSub##PREDICTOR_I##_C(const uint32_t* in,              \
   assert(upper != NULL);                                                   \
   for (x = 0; x < num_pixels; ++x) {                                       \
     const uint32_t pred =                                                  \
-        VP8LPredictor##PREDICTOR_I##_C(in[x - 1], upper + x);              \
+        VP8LPredictor##PREDICTOR_I##_C(&in[x - 1], upper + x);             \
     out[x] = VP8LSubPixels(in[x], pred);                                   \
   }                                                                        \
 }
@@ -843,10 +862,10 @@ WEBP_DSP_INIT_FUNC(VP8LEncDspInit) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8LEncDspInitSSE2();
-#if defined(WEBP_USE_SSE41)
+#if defined(WEBP_HAVE_SSE41)
       if (VP8GetCPUInfo(kSSE4_1)) {
         VP8LEncDspInitSSE41();
       }
@@ -870,7 +889,7 @@ WEBP_DSP_INIT_FUNC(VP8LEncDspInit) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     VP8LEncDspInitNEON();

src/dsp/lossless_enc_sse2.c

@@ -232,79 +232,55 @@ static void AddVectorEq_SSE2(const uint32_t* a, uint32_t* out, int size) {
 //------------------------------------------------------------------------------
 // Entropy
 
-// Checks whether the X or Y contribution is worth computing and adding.
-// Used in loop unrolling.
-#define ANALYZE_X_OR_Y(x_or_y, j)                                           \
-  do {                                                                      \
-    if ((x_or_y)[i + (j)] != 0) retval -= VP8LFastSLog2((x_or_y)[i + (j)]); \
-  } while (0)
+// TODO(https://crbug.com/webp/499): this function produces different results
+// from the C code due to use of double/float resulting in output differences
+// when compared to -noasm.
+#if !(defined(WEBP_HAVE_SLOW_CLZ_CTZ) || defined(__i386__) || defined(_M_IX86))
 
-// Checks whether the X + Y contribution is worth computing and adding.
-// Used in loop unrolling.
-#define ANALYZE_XY(j)                  \
-  do {                                 \
-    if (tmp[j] != 0) {                 \
-      retval -= VP8LFastSLog2(tmp[j]); \
-      ANALYZE_X_OR_Y(X, j);            \
-    }                                  \
-  } while (0)
-
-#if !(defined(__i386__) || defined(_M_IX86))
 static float CombinedShannonEntropy_SSE2(const int X[256], const int Y[256]) {
   int i;
   double retval = 0.;
-  int sumX, sumXY;
-  int32_t tmp[4];
-  __m128i zero = _mm_setzero_si128();
-  // Sums up X + Y, 4 ints at a time (and will merge it at the end for sumXY).
-  __m128i sumXY_128 = zero;
-  __m128i sumX_128 = zero;
+  int sumX = 0, sumXY = 0;
+  const __m128i zero = _mm_setzero_si128();
 
-  for (i = 0; i < 256; i += 4) {
-    const __m128i x = _mm_loadu_si128((const __m128i*)(X + i));
-    const __m128i y = _mm_loadu_si128((const __m128i*)(Y + i));
-
-    // Check if any X is non-zero: this actually provides a speedup as X is
-    // usually sparse.
-    if (_mm_movemask_epi8(_mm_cmpeq_epi32(x, zero)) != 0xFFFF) {
-      const __m128i xy_128 = _mm_add_epi32(x, y);
-      sumXY_128 = _mm_add_epi32(sumXY_128, xy_128);
-
-      sumX_128 = _mm_add_epi32(sumX_128, x);
-
-      // Analyze the different X + Y.
-      _mm_storeu_si128((__m128i*)tmp, xy_128);
-
-      ANALYZE_XY(0);
-      ANALYZE_XY(1);
-      ANALYZE_XY(2);
-      ANALYZE_XY(3);
-    } else {
-      // X is fully 0, so only deal with Y.
-      sumXY_128 = _mm_add_epi32(sumXY_128, y);
-
-      ANALYZE_X_OR_Y(Y, 0);
-      ANALYZE_X_OR_Y(Y, 1);
-      ANALYZE_X_OR_Y(Y, 2);
-      ANALYZE_X_OR_Y(Y, 3);
+  for (i = 0; i < 256; i += 16) {
+    const __m128i x0 = _mm_loadu_si128((const __m128i*)(X + i +  0));
+    const __m128i y0 = _mm_loadu_si128((const __m128i*)(Y + i +  0));
+    const __m128i x1 = _mm_loadu_si128((const __m128i*)(X + i +  4));
+    const __m128i y1 = _mm_loadu_si128((const __m128i*)(Y + i +  4));
+    const __m128i x2 = _mm_loadu_si128((const __m128i*)(X + i +  8));
+    const __m128i y2 = _mm_loadu_si128((const __m128i*)(Y + i +  8));
+    const __m128i x3 = _mm_loadu_si128((const __m128i*)(X + i + 12));
+    const __m128i y3 = _mm_loadu_si128((const __m128i*)(Y + i + 12));
+    const __m128i x4 = _mm_packs_epi16(_mm_packs_epi32(x0, x1),
+                                       _mm_packs_epi32(x2, x3));
+    const __m128i y4 = _mm_packs_epi16(_mm_packs_epi32(y0, y1),
+                                       _mm_packs_epi32(y2, y3));
+    const int32_t mx = _mm_movemask_epi8(_mm_cmpgt_epi8(x4, zero));
+    int32_t my = _mm_movemask_epi8(_mm_cmpgt_epi8(y4, zero)) | mx;
+    while (my) {
+      const int32_t j = BitsCtz(my);
+      int xy;
+      if ((mx >> j) & 1) {
+        const int x = X[i + j];
+        sumXY += x;
+        retval -= VP8LFastSLog2(x);
+      }
+      xy = X[i + j] + Y[i + j];
+      sumX += xy;
+      retval -= VP8LFastSLog2(xy);
+      my &= my - 1;
     }
   }
-
-  // Sum up sumX_128 to get sumX.
-  _mm_storeu_si128((__m128i*)tmp, sumX_128);
-  sumX = tmp[3] + tmp[2] + tmp[1] + tmp[0];
-
-  // Sum up sumXY_128 to get sumXY.
-  _mm_storeu_si128((__m128i*)tmp, sumXY_128);
-  sumXY = tmp[3] + tmp[2] + tmp[1] + tmp[0];
-
   retval += VP8LFastSLog2(sumX) + VP8LFastSLog2(sumXY);
   return (float)retval;
 }
-#endif  // !(defined(__i386__) || defined(_M_IX86))
 
-#undef ANALYZE_X_OR_Y
-#undef ANALYZE_XY
+#else
+
+#define DONT_USE_COMBINED_SHANNON_ENTROPY_SSE2_FUNC   // won't be faster
+
+#endif
 
 //------------------------------------------------------------------------------
 
@@ -662,10 +638,7 @@ WEBP_TSAN_IGNORE_FUNCTION void VP8LEncDspInitSSE2(void) {
   VP8LCollectColorRedTransforms = CollectColorRedTransforms_SSE2;
   VP8LAddVector = AddVector_SSE2;
   VP8LAddVectorEq = AddVectorEq_SSE2;
-  // TODO(https://crbug.com/webp/499): this function produces different results
-  // from the C code due to use of double/float resulting in output differences
-  // when compared to -noasm.
-#if !(defined(__i386__) || defined(_M_IX86))
+#if !defined(DONT_USE_COMBINED_SHANNON_ENTROPY_SSE2_FUNC)
   VP8LCombinedShannonEntropy = CombinedShannonEntropy_SSE2;
 #endif
   VP8LVectorMismatch = VectorMismatch_SSE2;

src/dsp/lossless_enc_sse41.c

@@ -44,46 +44,47 @@ static void SubtractGreenFromBlueAndRed_SSE41(uint32_t* argb_data,
 //------------------------------------------------------------------------------
 // Color Transform
 
-#define SPAN 8
+#define MK_CST_16(HI, LO) \
+  _mm_set1_epi32((int)(((uint32_t)(HI) << 16) | ((LO) & 0xffff)))
+
 static void CollectColorBlueTransforms_SSE41(const uint32_t* argb, int stride,
                                              int tile_width, int tile_height,
                                              int green_to_blue, int red_to_blue,
                                              int histo[]) {
-  const __m128i mults_r = _mm_set1_epi16(CST_5b(red_to_blue));
-  const __m128i mults_g = _mm_set1_epi16(CST_5b(green_to_blue));
-  const __m128i mask_g = _mm_set1_epi16((short)0xff00);   // green mask
-  const __m128i mask_gb = _mm_set1_epi32(0xffff);         // green/blue mask
-  const __m128i mask_b = _mm_set1_epi16(0x00ff);          // blue mask
-  const __m128i shuffler_lo = _mm_setr_epi8(-1, 2, -1, 6, -1, 10, -1, 14, -1,
-                                            -1, -1, -1, -1, -1, -1, -1);
-  const __m128i shuffler_hi = _mm_setr_epi8(-1, -1, -1, -1, -1, -1, -1, -1, -1,
-                                            2, -1, 6, -1, 10, -1, 14);
-  int y;
-  for (y = 0; y < tile_height; ++y) {
-    const uint32_t* const src = argb + y * stride;
-    int i, x;
-    for (x = 0; x + SPAN <= tile_width; x += SPAN) {
-      uint16_t values[SPAN];
-      const __m128i in0 = _mm_loadu_si128((__m128i*)&src[x + 0]);
-      const __m128i in1 = _mm_loadu_si128((__m128i*)&src[x + SPAN / 2]);
-      const __m128i r0 = _mm_shuffle_epi8(in0, shuffler_lo);
-      const __m128i r1 = _mm_shuffle_epi8(in1, shuffler_hi);
-      const __m128i r = _mm_or_si128(r0, r1);         // r 0
-      const __m128i gb0 = _mm_and_si128(in0, mask_gb);
-      const __m128i gb1 = _mm_and_si128(in1, mask_gb);
-      const __m128i gb = _mm_packus_epi32(gb0, gb1);  // g b
-      const __m128i g = _mm_and_si128(gb, mask_g);    // g 0
-      const __m128i A = _mm_mulhi_epi16(r, mults_r);  // x dbr
-      const __m128i B = _mm_mulhi_epi16(g, mults_g);  // x dbg
-      const __m128i C = _mm_sub_epi8(gb, B);          // x b'
-      const __m128i D = _mm_sub_epi8(C, A);           // x b''
-      const __m128i E = _mm_and_si128(D, mask_b);     // 0 b''
-      _mm_storeu_si128((__m128i*)values, E);
-      for (i = 0; i < SPAN; ++i) ++histo[values[i]];
+  const __m128i mult =
+      MK_CST_16(CST_5b(red_to_blue) + 256,CST_5b(green_to_blue));
+  const __m128i perm =
+      _mm_setr_epi8(-1, 1, -1, 2, -1, 5, -1, 6, -1, 9, -1, 10, -1, 13, -1, 14);
+  if (tile_width >= 4) {
+    int y;
+    for (y = 0; y < tile_height; ++y) {
+      const uint32_t* const src = argb + y * stride;
+      const __m128i A1 = _mm_loadu_si128((const __m128i*)src);
+      const __m128i B1 = _mm_shuffle_epi8(A1, perm);
+      const __m128i C1 = _mm_mulhi_epi16(B1, mult);
+      const __m128i D1 = _mm_sub_epi16(A1, C1);
+      __m128i E = _mm_add_epi16(_mm_srli_epi32(D1, 16), D1);
+      int x;
+      for (x = 4; x + 4 <= tile_width; x += 4) {
+        const __m128i A2 = _mm_loadu_si128((const __m128i*)(src + x));
+        __m128i B2, C2, D2;
+        ++histo[_mm_extract_epi8(E,  0)];
+        B2 = _mm_shuffle_epi8(A2, perm);
+        ++histo[_mm_extract_epi8(E,  4)];
+        C2 = _mm_mulhi_epi16(B2, mult);
+        ++histo[_mm_extract_epi8(E,  8)];
+        D2 = _mm_sub_epi16(A2, C2);
+        ++histo[_mm_extract_epi8(E, 12)];
+        E = _mm_add_epi16(_mm_srli_epi32(D2, 16), D2);
+      }
+      ++histo[_mm_extract_epi8(E,  0)];
+      ++histo[_mm_extract_epi8(E,  4)];
+      ++histo[_mm_extract_epi8(E,  8)];
+      ++histo[_mm_extract_epi8(E, 12)];
     }
   }
   {
-    const int left_over = tile_width & (SPAN - 1);
+    const int left_over = tile_width & 3;
     if (left_over > 0) {
       VP8LCollectColorBlueTransforms_C(argb + tile_width - left_over, stride,
                                        left_over, tile_height,
@@ -95,33 +96,37 @@ static void CollectColorBlueTransforms_SSE41(const uint32_t* argb, int stride,
 static void CollectColorRedTransforms_SSE41(const uint32_t* argb, int stride,
                                             int tile_width, int tile_height,
                                             int green_to_red, int histo[]) {
-  const __m128i mults_g = _mm_set1_epi16(CST_5b(green_to_red));
-  const __m128i mask_g = _mm_set1_epi32(0x00ff00);  // green mask
-  const __m128i mask = _mm_set1_epi16(0xff);
 
-  int y;
-  for (y = 0; y < tile_height; ++y) {
-    const uint32_t* const src = argb + y * stride;
-    int i, x;
-    for (x = 0; x + SPAN <= tile_width; x += SPAN) {
-      uint16_t values[SPAN];
-      const __m128i in0 = _mm_loadu_si128((__m128i*)&src[x + 0]);
-      const __m128i in1 = _mm_loadu_si128((__m128i*)&src[x + SPAN / 2]);
-      const __m128i g0 = _mm_and_si128(in0, mask_g);  // 0 0  | g 0
-      const __m128i g1 = _mm_and_si128(in1, mask_g);
-      const __m128i g = _mm_packus_epi32(g0, g1);     // g 0
-      const __m128i A0 = _mm_srli_epi32(in0, 16);     // 0 0  | x r
-      const __m128i A1 = _mm_srli_epi32(in1, 16);
-      const __m128i A = _mm_packus_epi32(A0, A1);     // x r
-      const __m128i B = _mm_mulhi_epi16(g, mults_g);  // x dr
-      const __m128i C = _mm_sub_epi8(A, B);           // x r'
-      const __m128i D = _mm_and_si128(C, mask);       // 0 r'
-      _mm_storeu_si128((__m128i*)values, D);
-      for (i = 0; i < SPAN; ++i) ++histo[values[i]];
+  const __m128i mult = MK_CST_16(0, CST_5b(green_to_red));
+  const __m128i mask_g = _mm_set1_epi32(0x0000ff00);
+  if (tile_width >= 4) {
+    int y;
+    for (y = 0; y < tile_height; ++y) {
+      const uint32_t* const src = argb + y * stride;
+      const __m128i A1 = _mm_loadu_si128((const __m128i*)src);
+      const __m128i B1 = _mm_and_si128(A1, mask_g);
+      const __m128i C1 = _mm_madd_epi16(B1, mult);
+      __m128i D = _mm_sub_epi16(A1, C1);
+      int x;
+      for (x = 4; x + 4 <= tile_width; x += 4) {
+        const __m128i A2 = _mm_loadu_si128((const __m128i*)(src + x));
+        __m128i B2, C2;
+        ++histo[_mm_extract_epi8(D,  2)];
+        B2 = _mm_and_si128(A2, mask_g);
+        ++histo[_mm_extract_epi8(D,  6)];
+        C2 = _mm_madd_epi16(B2, mult);
+        ++histo[_mm_extract_epi8(D, 10)];
+        ++histo[_mm_extract_epi8(D, 14)];
+        D = _mm_sub_epi16(A2, C2);
+      }
+      ++histo[_mm_extract_epi8(D,  2)];
+      ++histo[_mm_extract_epi8(D,  6)];
+      ++histo[_mm_extract_epi8(D, 10)];
+      ++histo[_mm_extract_epi8(D, 14)];
     }
   }
   {
-    const int left_over = tile_width & (SPAN - 1);
+    const int left_over = tile_width & 3;
     if (left_over > 0) {
       VP8LCollectColorRedTransforms_C(argb + tile_width - left_over, stride,
                                       left_over, tile_height, green_to_red,
@@ -130,6 +135,8 @@ static void CollectColorRedTransforms_SSE41(const uint32_t* argb, int stride,
   }
 }
 
+#undef MK_CST_16
+
 //------------------------------------------------------------------------------
 // Entry point
 

src/dsp/lossless_mips_dsp_r2.c

@@ -188,46 +188,51 @@ static WEBP_INLINE uint32_t Average4(uint32_t a0, uint32_t a1,
   return Average2(Average2(a0, a1), Average2(a2, a3));
 }
 
-static uint32_t Predictor5_MIPSdspR2(uint32_t left, const uint32_t* const top) {
-  return Average3(left, top[0], top[1]);
+static uint32_t Predictor5_MIPSdspR2(const uint32_t* const left,
+                                     const uint32_t* const top) {
+  return Average3(*left, top[0], top[1]);
 }
 
-static uint32_t Predictor6_MIPSdspR2(uint32_t left, const uint32_t* const top) {
-  return Average2(left, top[-1]);
+static uint32_t Predictor6_MIPSdspR2(const uint32_t* const left,
+                                     const uint32_t* const top) {
+  return Average2(*left, top[-1]);
 }
 
-static uint32_t Predictor7_MIPSdspR2(uint32_t left, const uint32_t* const top) {
-  return Average2(left, top[0]);
+static uint32_t Predictor7_MIPSdspR2(const uint32_t* const left,
+                                     const uint32_t* const top) {
+  return Average2(*left, top[0]);
 }
 
-static uint32_t Predictor8_MIPSdspR2(uint32_t left, const uint32_t* const top) {
+static uint32_t Predictor8_MIPSdspR2(const uint32_t* const left,
+                                     const uint32_t* const top) {
   (void)left;
   return Average2(top[-1], top[0]);
 }
 
-static uint32_t Predictor9_MIPSdspR2(uint32_t left, const uint32_t* const top) {
+static uint32_t Predictor9_MIPSdspR2(const uint32_t* const left,
+                                     const uint32_t* const top) {
   (void)left;
   return Average2(top[0], top[1]);
 }
 
-static uint32_t Predictor10_MIPSdspR2(uint32_t left,
+static uint32_t Predictor10_MIPSdspR2(const uint32_t* const left,
                                       const uint32_t* const top) {
-  return Average4(left, top[-1], top[0], top[1]);
+  return Average4(*left, top[-1], top[0], top[1]);
 }
 
-static uint32_t Predictor11_MIPSdspR2(uint32_t left,
+static uint32_t Predictor11_MIPSdspR2(const uint32_t* const left,
                                       const uint32_t* const top) {
-  return Select(top[0], left, top[-1]);
+  return Select(top[0], *left, top[-1]);
 }
 
-static uint32_t Predictor12_MIPSdspR2(uint32_t left,
+static uint32_t Predictor12_MIPSdspR2(const uint32_t* const left,
                                       const uint32_t* const top) {
-  return ClampedAddSubtractFull(left, top[0], top[-1]);
+  return ClampedAddSubtractFull(*left, top[0], top[-1]);
 }
 
-static uint32_t Predictor13_MIPSdspR2(uint32_t left,
+static uint32_t Predictor13_MIPSdspR2(const uint32_t* const left,
                                       const uint32_t* const top) {
-  return ClampedAddSubtractHalf(left, top[0], top[-1]);
+  return ClampedAddSubtractHalf(*left, top[0], top[-1]);
 }
 
 // Add green to blue and red channels (i.e. perform the inverse transform of

src/dsp/lossless_neon.c

@@ -188,17 +188,21 @@ static WEBP_INLINE uint32_t Average3_NEON(uint32_t a0, uint32_t a1,
   return avg;
 }
 
-static uint32_t Predictor5_NEON(uint32_t left, const uint32_t* const top) {
-  return Average3_NEON(left, top[0], top[1]);
+static uint32_t Predictor5_NEON(const uint32_t* const left,
+                                const uint32_t* const top) {
+  return Average3_NEON(*left, top[0], top[1]);
 }
-static uint32_t Predictor6_NEON(uint32_t left, const uint32_t* const top) {
-  return Average2_NEON(left, top[-1]);
+static uint32_t Predictor6_NEON(const uint32_t* const left,
+                                const uint32_t* const top) {
+  return Average2_NEON(*left, top[-1]);
 }
-static uint32_t Predictor7_NEON(uint32_t left, const uint32_t* const top) {
-  return Average2_NEON(left, top[0]);
+static uint32_t Predictor7_NEON(const uint32_t* const left,
+                                const uint32_t* const top) {
+  return Average2_NEON(*left, top[0]);
 }
-static uint32_t Predictor13_NEON(uint32_t left, const uint32_t* const top) {
-  return ClampedAddSubtractHalf_NEON(left, top[0], top[-1]);
+static uint32_t Predictor13_NEON(const uint32_t* const left,
+                                 const uint32_t* const top) {
+  return ClampedAddSubtractHalf_NEON(*left, top[0], top[-1]);
 }
 
 // Batch versions of those functions.

src/dsp/lossless_sse2.c

@@ -18,7 +18,6 @@
 #include "src/dsp/common_sse2.h"
 #include "src/dsp/lossless.h"
 #include "src/dsp/lossless_common.h"
-#include <assert.h>
 #include <emmintrin.h>
 
 //------------------------------------------------------------------------------
@@ -139,42 +138,51 @@ static WEBP_INLINE uint32_t Average4_SSE2(uint32_t a0, uint32_t a1,
   return output;
 }
 
-static uint32_t Predictor5_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average3_SSE2(left, top[0], top[1]);
+static uint32_t Predictor5_SSE2(const uint32_t* const left,
+                                const uint32_t* const top) {
+  const uint32_t pred = Average3_SSE2(*left, top[0], top[1]);
   return pred;
 }
-static uint32_t Predictor6_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average2_SSE2(left, top[-1]);
+static uint32_t Predictor6_SSE2(const uint32_t* const left,
+                                const uint32_t* const top) {
+  const uint32_t pred = Average2_SSE2(*left, top[-1]);
   return pred;
 }
-static uint32_t Predictor7_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average2_SSE2(left, top[0]);
+static uint32_t Predictor7_SSE2(const uint32_t* const left,
+                                const uint32_t* const top) {
+  const uint32_t pred = Average2_SSE2(*left, top[0]);
   return pred;
 }
-static uint32_t Predictor8_SSE2(uint32_t left, const uint32_t* const top) {
+static uint32_t Predictor8_SSE2(const uint32_t* const left,
+                                const uint32_t* const top) {
   const uint32_t pred = Average2_SSE2(top[-1], top[0]);
   (void)left;
   return pred;
 }
-static uint32_t Predictor9_SSE2(uint32_t left, const uint32_t* const top) {
+static uint32_t Predictor9_SSE2(const uint32_t* const left,
+                                const uint32_t* const top) {
   const uint32_t pred = Average2_SSE2(top[0], top[1]);
   (void)left;
   return pred;
 }
-static uint32_t Predictor10_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Average4_SSE2(left, top[-1], top[0], top[1]);
+static uint32_t Predictor10_SSE2(const uint32_t* const left,
+                                 const uint32_t* const top) {
+  const uint32_t pred = Average4_SSE2(*left, top[-1], top[0], top[1]);
   return pred;
 }
-static uint32_t Predictor11_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = Select_SSE2(top[0], left, top[-1]);
+static uint32_t Predictor11_SSE2(const uint32_t* const left,
+                                 const uint32_t* const top) {
+  const uint32_t pred = Select_SSE2(top[0], *left, top[-1]);
   return pred;
 }
-static uint32_t Predictor12_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = ClampedAddSubtractFull_SSE2(left, top[0], top[-1]);
+static uint32_t Predictor12_SSE2(const uint32_t* const left,
+                                 const uint32_t* const top) {
+  const uint32_t pred = ClampedAddSubtractFull_SSE2(*left, top[0], top[-1]);
   return pred;
 }
-static uint32_t Predictor13_SSE2(uint32_t left, const uint32_t* const top) {
-  const uint32_t pred = ClampedAddSubtractHalf_SSE2(left, top[0], top[-1]);
+static uint32_t Predictor13_SSE2(const uint32_t* const left,
+                                 const uint32_t* const top) {
+  const uint32_t pred = ClampedAddSubtractHalf_SSE2(*left, top[0], top[-1]);
   return pred;
 }
 

src/dsp/lossless_sse41.c

@@ -0,0 +1,132 @@
+// Copyright 2021 Google Inc. All Rights Reserved.
+//
+// Use of this source code is governed by a BSD-style license
+// that can be found in the COPYING file in the root of the source
+// tree. An additional intellectual property rights grant can be found
+// in the file PATENTS. All contributing project authors may
+// be found in the AUTHORS file in the root of the source tree.
+// -----------------------------------------------------------------------------
+//
+// SSE41 variant of methods for lossless decoder
+
+#include "src/dsp/dsp.h"
+
+#if defined(WEBP_USE_SSE41)
+
+#include "src/dsp/common_sse41.h"
+#include "src/dsp/lossless.h"
+#include "src/dsp/lossless_common.h"
+
+//------------------------------------------------------------------------------
+// Color-space conversion functions
+
+static void TransformColorInverse_SSE41(const VP8LMultipliers* const m,
+                                        const uint32_t* const src,
+                                        int num_pixels, uint32_t* dst) {
+// sign-extended multiplying constants, pre-shifted by 5.
+#define CST(X)  (((int16_t)(m->X << 8)) >> 5)   // sign-extend
+  const __m128i mults_rb = _mm_set1_epi32((uint32_t)CST(green_to_red_) << 16 |
+                                          (CST(green_to_blue_) & 0xffff));
+  const __m128i mults_b2 = _mm_set1_epi32(CST(red_to_blue_));
+#undef CST
+  const __m128i mask_ag = _mm_set1_epi32(0xff00ff00);
+  const __m128i perm1 = _mm_setr_epi8(-1, 1, -1, 1, -1, 5, -1, 5,
+                                      -1, 9, -1, 9, -1, 13, -1, 13);
+  const __m128i perm2 = _mm_setr_epi8(-1, 2, -1, -1, -1, 6, -1, -1,
+                                      -1, 10, -1, -1, -1, 14, -1, -1);
+  int i;
+  for (i = 0; i + 4 <= num_pixels; i += 4) {
+    const __m128i A = _mm_loadu_si128((const __m128i*)(src + i));
+    const __m128i B = _mm_shuffle_epi8(A, perm1); // argb -> g0g0
+    const __m128i C = _mm_mulhi_epi16(B, mults_rb);
+    const __m128i D = _mm_add_epi8(A, C);
+    const __m128i E = _mm_shuffle_epi8(D, perm2);
+    const __m128i F = _mm_mulhi_epi16(E, mults_b2);
+    const __m128i G = _mm_add_epi8(D, F);
+    const __m128i out = _mm_blendv_epi8(G, A, mask_ag);
+    _mm_storeu_si128((__m128i*)&dst[i], out);
+  }
+  // Fall-back to C-version for left-overs.
+  if (i != num_pixels) {
+    VP8LTransformColorInverse_C(m, src + i, num_pixels - i, dst + i);
+  }
+}
+
+//------------------------------------------------------------------------------
+
+#define ARGB_TO_RGB_SSE41 do {                        \
+  while (num_pixels >= 16) {                          \
+    const __m128i in0 = _mm_loadu_si128(in + 0);      \
+    const __m128i in1 = _mm_loadu_si128(in + 1);      \
+    const __m128i in2 = _mm_loadu_si128(in + 2);      \
+    const __m128i in3 = _mm_loadu_si128(in + 3);      \
+    const __m128i a0 = _mm_shuffle_epi8(in0, perm0);  \
+    const __m128i a1 = _mm_shuffle_epi8(in1, perm1);  \
+    const __m128i a2 = _mm_shuffle_epi8(in2, perm2);  \
+    const __m128i a3 = _mm_shuffle_epi8(in3, perm3);  \
+    const __m128i b0 = _mm_blend_epi16(a0, a1, 0xc0); \
+    const __m128i b1 = _mm_blend_epi16(a1, a2, 0xf0); \
+    const __m128i b2 = _mm_blend_epi16(a2, a3, 0xfc); \
+    _mm_storeu_si128(out + 0, b0);                    \
+    _mm_storeu_si128(out + 1, b1);                    \
+    _mm_storeu_si128(out + 2, b2);                    \
+    in += 4;                                          \
+    out += 3;                                         \
+    num_pixels -= 16;                                 \
+  }                                                   \
+} while (0)
+
+static void ConvertBGRAToRGB_SSE41(const uint32_t* src, int num_pixels,
+                                   uint8_t* dst) {
+  const __m128i* in = (const __m128i*)src;
+  __m128i* out = (__m128i*)dst;
+  const __m128i perm0 = _mm_setr_epi8(2, 1, 0, 6, 5, 4, 10, 9,
+                                      8, 14, 13, 12, -1, -1, -1, -1);
+  const __m128i perm1 = _mm_shuffle_epi32(perm0, 0x39);
+  const __m128i perm2 = _mm_shuffle_epi32(perm0, 0x4e);
+  const __m128i perm3 = _mm_shuffle_epi32(perm0, 0x93);
+
+  ARGB_TO_RGB_SSE41;
+
+  // left-overs
+  if (num_pixels > 0) {
+    VP8LConvertBGRAToRGB_C((const uint32_t*)in, num_pixels, (uint8_t*)out);
+  }
+}
+
+static void ConvertBGRAToBGR_SSE41(const uint32_t* src,
+                                   int num_pixels, uint8_t* dst) {
+  const __m128i* in = (const __m128i*)src;
+  __m128i* out = (__m128i*)dst;
+  const __m128i perm0 = _mm_setr_epi8(0, 1, 2, 4, 5, 6, 8, 9, 10,
+                                      12, 13, 14, -1, -1, -1, -1);
+  const __m128i perm1 = _mm_shuffle_epi32(perm0, 0x39);
+  const __m128i perm2 = _mm_shuffle_epi32(perm0, 0x4e);
+  const __m128i perm3 = _mm_shuffle_epi32(perm0, 0x93);
+
+  ARGB_TO_RGB_SSE41;
+
+  // left-overs
+  if (num_pixels > 0) {
+    VP8LConvertBGRAToBGR_C((const uint32_t*)in, num_pixels, (uint8_t*)out);
+  }
+}
+
+#undef ARGB_TO_RGB_SSE41
+
+//------------------------------------------------------------------------------
+// Entry point
+
+extern void VP8LDspInitSSE41(void);
+
+WEBP_TSAN_IGNORE_FUNCTION void VP8LDspInitSSE41(void) {
+  VP8LTransformColorInverse = TransformColorInverse_SSE41;
+  VP8LConvertBGRAToRGB = ConvertBGRAToRGB_SSE41;
+  VP8LConvertBGRAToBGR = ConvertBGRAToBGR_SSE41;
+}
+
+#else  // !WEBP_USE_SSE41
+
+WEBP_DSP_INIT_STUB(VP8LDspInitSSE41)
+
+#endif  // WEBP_USE_SSE41

src/dsp/msa_macro.h

@@ -14,6 +14,10 @@
 #ifndef WEBP_DSP_MSA_MACRO_H_
 #define WEBP_DSP_MSA_MACRO_H_
 
+#include "src/dsp/dsp.h"
+
+#if defined(WEBP_USE_MSA)
+
 #include <stdint.h>
 #include <msa.h>
 
@@ -1389,4 +1393,5 @@ static WEBP_INLINE uint32_t func_hadd_uh_u32(v8u16 in) {
 } while (0)
 #define AVER_UB2_UB(...) AVER_UB2(v16u8, __VA_ARGS__)
 
+#endif  // WEBP_USE_MSA
 #endif  // WEBP_DSP_MSA_MACRO_H_

src/dsp/neon.h

@@ -12,10 +12,12 @@
 #ifndef WEBP_DSP_NEON_H_
 #define WEBP_DSP_NEON_H_
 
-#include <arm_neon.h>
-
 #include "src/dsp/dsp.h"
 
+#if defined(WEBP_USE_NEON)
+
+#include <arm_neon.h>
+
 // Right now, some intrinsics functions seem slower, so we disable them
 // everywhere except newer clang/gcc or aarch64 where the inline assembly is
 // incompatible.
@@ -98,4 +100,5 @@ static WEBP_INLINE int32x4x4_t Transpose4x4_NEON(const int32x4x4_t rows) {
 } while (0)
 #endif
 
+#endif  // WEBP_USE_NEON
 #endif  // WEBP_DSP_NEON_H_

src/dsp/rescaler.c

@@ -38,8 +38,9 @@ void WebPRescalerImportRowExpand_C(WebPRescaler* const wrk,
     int x_out = channel;
     // simple bilinear interpolation
     int accum = wrk->x_add;
-    int left = src[x_in];
-    int right = (wrk->src_width > 1) ? src[x_in + x_stride] : left;
+    rescaler_t left = (rescaler_t)src[x_in];
+    rescaler_t right =
+        (wrk->src_width > 1) ? (rescaler_t)src[x_in + x_stride] : left;
     x_in += x_stride;
     while (1) {
       wrk->frow[x_out] = right * wrk->x_add + (left - right) * accum;
@@ -50,7 +51,7 @@ void WebPRescalerImportRowExpand_C(WebPRescaler* const wrk,
         left = right;
         x_in += x_stride;
         assert(x_in < wrk->src_width * x_stride);
-        right = src[x_in];
+        right = (rescaler_t)src[x_in];
         accum += wrk->x_add;
       }
     }
@@ -213,7 +214,7 @@ WEBP_DSP_INIT_FUNC(WebPRescalerDspInit) {
   WebPRescalerImportRowShrink = WebPRescalerImportRowShrink_C;
 
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       WebPRescalerDspInitSSE2();
     }
@@ -235,7 +236,7 @@ WEBP_DSP_INIT_FUNC(WebPRescalerDspInit) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     WebPRescalerDspInitNEON();

src/dsp/ssim.c

@@ -150,7 +150,7 @@ WEBP_DSP_INIT_FUNC(VP8SSIMDspInit) {
 #endif
 
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       VP8SSIMDspInitSSE2();
     }

src/dsp/upsampling.c

@@ -233,12 +233,12 @@ WEBP_DSP_INIT_FUNC(WebPInitYUV444Converters) {
   WebPYUV444Converters[MODE_rgbA_4444] = WebPYuv444ToRgba4444_C;
 
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       WebPInitYUV444ConvertersSSE2();
     }
 #endif
-#if defined(WEBP_USE_SSE41)
+#if defined(WEBP_HAVE_SSE41)
     if (VP8GetCPUInfo(kSSE4_1)) {
       WebPInitYUV444ConvertersSSE41();
     }
@@ -278,12 +278,12 @@ WEBP_DSP_INIT_FUNC(WebPInitUpsamplers) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       WebPInitUpsamplersSSE2();
     }
 #endif
-#if defined(WEBP_USE_SSE41)
+#if defined(WEBP_HAVE_SSE41)
     if (VP8GetCPUInfo(kSSE4_1)) {
       WebPInitUpsamplersSSE41();
     }
@@ -300,7 +300,7 @@ WEBP_DSP_INIT_FUNC(WebPInitUpsamplers) {
 #endif
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     WebPInitUpsamplersNEON();

src/dsp/yuv.c

@@ -90,16 +90,16 @@ WEBP_DSP_INIT_FUNC(WebPInitSamplers) {
 
   // If defined, use CPUInfo() to overwrite some pointers with faster versions.
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       WebPInitSamplersSSE2();
     }
-#endif  // WEBP_USE_SSE2
-#if defined(WEBP_USE_SSE41)
+#endif  // WEBP_HAVE_SSE2
+#if defined(WEBP_HAVE_SSE41)
     if (VP8GetCPUInfo(kSSE4_1)) {
       WebPInitSamplersSSE41();
     }
-#endif  // WEBP_USE_SSE41
+#endif  // WEBP_HAVE_SSE41
 #if defined(WEBP_USE_MIPS32)
     if (VP8GetCPUInfo(kMIPS32)) {
       WebPInitSamplersMIPS32();
@@ -276,26 +276,26 @@ WEBP_DSP_INIT_FUNC(WebPInitConvertARGBToYUV) {
 #endif
 
   if (VP8GetCPUInfo != NULL) {
-#if defined(WEBP_USE_SSE2)
+#if defined(WEBP_HAVE_SSE2)
     if (VP8GetCPUInfo(kSSE2)) {
       WebPInitConvertARGBToYUVSSE2();
       WebPInitSharpYUVSSE2();
     }
-#endif  // WEBP_USE_SSE2
-#if defined(WEBP_USE_SSE41)
+#endif  // WEBP_HAVE_SSE2
+#if defined(WEBP_HAVE_SSE41)
     if (VP8GetCPUInfo(kSSE4_1)) {
       WebPInitConvertARGBToYUVSSE41();
     }
-#endif  // WEBP_USE_SSE41
+#endif  // WEBP_HAVE_SSE41
   }
 
-#if defined(WEBP_USE_NEON)
+#if defined(WEBP_HAVE_NEON)
   if (WEBP_NEON_OMIT_C_CODE ||
       (VP8GetCPUInfo != NULL && VP8GetCPUInfo(kNEON))) {
     WebPInitConvertARGBToYUVNEON();
     WebPInitSharpYUVNEON();
   }
-#endif  // WEBP_USE_NEON
+#endif  // WEBP_HAVE_NEON
 
   assert(WebPConvertARGBToY != NULL);
   assert(WebPConvertARGBToUV != NULL);

src/dsp/yuv.h

@@ -10,7 +10,7 @@
 // inline YUV<->RGB conversion function
 //
 // The exact naming is Y'CbCr, following the ITU-R BT.601 standard.
-// More information at: http://en.wikipedia.org/wiki/YCbCr
+// More information at: https://en.wikipedia.org/wiki/YCbCr
 // Y = 0.2569 * R + 0.5044 * G + 0.0979 * B + 16
 // U = -0.1483 * R - 0.2911 * G + 0.4394 * B + 128
 // V = 0.4394 * R - 0.3679 * G - 0.0715 * B + 128

src/enc/alpha_enc.c

@@ -303,7 +303,7 @@ static int EncodeAlpha(VP8Encoder* const enc,
   int ok = 1;
   const int reduce_levels = (quality < 100);
 
-  // quick sanity checks
+  // quick correctness checks
   assert((uint64_t)data_size == (uint64_t)width * height);  // as per spec
   assert(enc != NULL && pic != NULL && pic->a != NULL);
   assert(output != NULL && output_size != NULL);
@@ -361,7 +361,7 @@ static int EncodeAlpha(VP8Encoder* const enc,
 //------------------------------------------------------------------------------
 // Main calls
 
-static int CompressAlphaJob(void* arg1, void* dummy) {
+static int CompressAlphaJob(void* arg1, void* unused) {
   VP8Encoder* const enc = (VP8Encoder*)arg1;
   const WebPConfig* config = enc->config_;
   uint8_t* alpha_data = NULL;
@@ -375,13 +375,13 @@ static int CompressAlphaJob(void* arg1, void* dummy) {
                    filter, effort_level, &alpha_data, &alpha_size)) {
     return 0;
   }
-  if (alpha_size != (uint32_t)alpha_size) {  // Sanity check.
+  if (alpha_size != (uint32_t)alpha_size) {  // Soundness check.
     WebPSafeFree(alpha_data);
     return 0;
   }
   enc->alpha_data_size_ = (uint32_t)alpha_size;
   enc->alpha_data_ = alpha_data;
-  (void)dummy;
+  (void)unused;
   return 1;
 }
 

src/enc/histogram_enc.c

@@ -1171,13 +1171,15 @@ static void RemoveEmptyHistograms(VP8LHistogramSet* const image_histo) {
 int VP8LGetHistoImageSymbols(int xsize, int ysize,
                              const VP8LBackwardRefs* const refs,
                              int quality, int low_effort,
-                             int histo_bits, int cache_bits,
+                             int histogram_bits, int cache_bits,
                              VP8LHistogramSet* const image_histo,
                              VP8LHistogram* const tmp_histo,
                              uint16_t* const histogram_symbols) {
   int ok = 0;
-  const int histo_xsize = histo_bits ? VP8LSubSampleSize(xsize, histo_bits) : 1;
-  const int histo_ysize = histo_bits ? VP8LSubSampleSize(ysize, histo_bits) : 1;
+  const int histo_xsize =
+      histogram_bits ? VP8LSubSampleSize(xsize, histogram_bits) : 1;
+  const int histo_ysize =
+      histogram_bits ? VP8LSubSampleSize(ysize, histogram_bits) : 1;
   const int image_histo_raw_size = histo_xsize * histo_ysize;
   VP8LHistogramSet* const orig_histo =
       VP8LAllocateHistogramSet(image_histo_raw_size, cache_bits);
@@ -1193,7 +1195,7 @@ int VP8LGetHistoImageSymbols(int xsize, int ysize,
   if (orig_histo == NULL || map_tmp == NULL) goto Error;
 
   // Construct the histograms from backward references.
-  HistogramBuild(xsize, histo_bits, refs, orig_histo);
+  HistogramBuild(xsize, histogram_bits, refs, orig_histo);
   // Copies the histograms and computes its bit_cost.
   // histogram_symbols is optimized
   HistogramCopyAndAnalyze(orig_histo, image_histo, &num_used,

src/enc/histogram_enc.h

@@ -64,8 +64,8 @@ void VP8LHistogramCreate(VP8LHistogram* const p,
                          const VP8LBackwardRefs* const refs,
                          int palette_code_bits);
 
-// Return the size of the histogram for a given palette_code_bits.
-int VP8LGetHistogramSize(int palette_code_bits);
+// Return the size of the histogram for a given cache_bits.
+int VP8LGetHistogramSize(int cache_bits);
 
 // Set the palette_code_bits and reset the stats.
 // If init_arrays is true, the arrays are also filled with 0's.
@@ -110,7 +110,7 @@ int VP8LGetHistoImageSymbols(int xsize, int ysize,
                              const VP8LBackwardRefs* const refs,
                              int quality, int low_effort,
                              int histogram_bits, int cache_bits,
-                             VP8LHistogramSet* const image_in,
+                             VP8LHistogramSet* const image_histo,
                              VP8LHistogram* const tmp_histo,
                              uint16_t* const histogram_symbols);
 

src/enc/picture_rescale_enc.c

@@ -164,22 +164,25 @@ int WebPPictureCrop(WebPPicture* pic,
 //------------------------------------------------------------------------------
 // Simple picture rescaler
 
-static void RescalePlane(const uint8_t* src,
-                         int src_width, int src_height, int src_stride,
-                         uint8_t* dst,
-                         int dst_width, int dst_height, int dst_stride,
-                         rescaler_t* const work,
-                         int num_channels) {
+static int RescalePlane(const uint8_t* src,
+                        int src_width, int src_height, int src_stride,
+                        uint8_t* dst,
+                        int dst_width, int dst_height, int dst_stride,
+                        rescaler_t* const work,
+                        int num_channels) {
   WebPRescaler rescaler;
   int y = 0;
-  WebPRescalerInit(&rescaler, src_width, src_height,
-                   dst, dst_width, dst_height, dst_stride,
-                   num_channels, work);
+  if (!WebPRescalerInit(&rescaler, src_width, src_height,
+                        dst, dst_width, dst_height, dst_stride,
+                        num_channels, work)) {
+    return 0;
+  }
   while (y < src_height) {
     y += WebPRescalerImport(&rescaler, src_height - y,
                             src + y * src_stride, src_stride);
     WebPRescalerExport(&rescaler);
   }
+  return 1;
 }
 
 static void AlphaMultiplyARGB(WebPPicture* const pic, int inverse) {
@@ -222,25 +225,28 @@ int WebPPictureRescale(WebPPicture* pic, int width, int height) {
     // If present, we need to rescale alpha first (for AlphaMultiplyY).
     if (pic->a != NULL) {
       WebPInitAlphaProcessing();
-      RescalePlane(pic->a, prev_width, prev_height, pic->a_stride,
-                   tmp.a, width, height, tmp.a_stride, work, 1);
+      if (!RescalePlane(pic->a, prev_width, prev_height, pic->a_stride,
+                        tmp.a, width, height, tmp.a_stride, work, 1)) {
+        return 0;
+      }
     }
 
     // We take transparency into account on the luma plane only. That's not
     // totally exact blending, but still is a good approximation.
     AlphaMultiplyY(pic, 0);
-    RescalePlane(pic->y, prev_width, prev_height, pic->y_stride,
-                 tmp.y, width, height, tmp.y_stride, work, 1);
+    if (!RescalePlane(pic->y, prev_width, prev_height, pic->y_stride,
+                      tmp.y, width, height, tmp.y_stride, work, 1) ||
+        !RescalePlane(pic->u,
+                      HALVE(prev_width), HALVE(prev_height), pic->uv_stride,
+                      tmp.u,
+                      HALVE(width), HALVE(height), tmp.uv_stride, work, 1) ||
+        !RescalePlane(pic->v,
+                      HALVE(prev_width), HALVE(prev_height), pic->uv_stride,
+                      tmp.v,
+                      HALVE(width), HALVE(height), tmp.uv_stride, work, 1)) {
+      return 0;
+    }
     AlphaMultiplyY(&tmp, 1);
-
-    RescalePlane(pic->u,
-                 HALVE(prev_width), HALVE(prev_height), pic->uv_stride,
-                 tmp.u,
-                 HALVE(width), HALVE(height), tmp.uv_stride, work, 1);
-    RescalePlane(pic->v,
-                 HALVE(prev_width), HALVE(prev_height), pic->uv_stride,
-                 tmp.v,
-                 HALVE(width), HALVE(height), tmp.uv_stride, work, 1);
   } else {
     work = (rescaler_t*)WebPSafeMalloc(2ULL * width * 4, sizeof(*work));
     if (work == NULL) {
@@ -252,11 +258,12 @@ int WebPPictureRescale(WebPPicture* pic, int width, int height) {
     // the premultiplication afterward (while preserving the alpha channel).
     WebPInitAlphaProcessing();
     AlphaMultiplyARGB(pic, 0);
-    RescalePlane((const uint8_t*)pic->argb, prev_width, prev_height,
-                 pic->argb_stride * 4,
-                 (uint8_t*)tmp.argb, width, height,
-                 tmp.argb_stride * 4,
-                 work, 4);
+    if (!RescalePlane((const uint8_t*)pic->argb, prev_width, prev_height,
+                      pic->argb_stride * 4,
+                      (uint8_t*)tmp.argb, width, height,
+                      tmp.argb_stride * 4, work, 4)) {
+      return 0;
+    }
     AlphaMultiplyARGB(&tmp, 1);
   }
   WebPPictureFree(pic);

src/enc/predictor_enc.c

@@ -249,7 +249,7 @@ static WEBP_INLINE void GetResidual(
       } else if (x == 0) {
         predict = upper_row[x];  // Top.
       } else {
-        predict = pred_func(current_row[x - 1], upper_row + x);
+        predict = pred_func(&current_row[x - 1], upper_row + x);
       }
 #if (WEBP_NEAR_LOSSLESS == 1)
       if (max_quantization == 1 || mode == 0 || y == 0 || y == height - 1 ||

src/enc/quant_enc.c

@@ -585,6 +585,9 @@ static WEBP_INLINE score_t RDScoreTrellis(int lambda, score_t rate,
   return rate * lambda + RD_DISTO_MULT * distortion;
 }
 
+// Coefficient type.
+enum { TYPE_I16_AC = 0, TYPE_I16_DC = 1, TYPE_CHROMA_A = 2, TYPE_I4_AC = 3 };
+
 static int TrellisQuantizeBlock(const VP8Encoder* const enc,
                                 int16_t in[16], int16_t out[16],
                                 int ctx0, int coeff_type,
@@ -593,7 +596,7 @@ static int TrellisQuantizeBlock(const VP8Encoder* const enc,
   const ProbaArray* const probas = enc->proba_.coeffs_[coeff_type];
   CostArrayPtr const costs =
       (CostArrayPtr)enc->proba_.remapped_costs_[coeff_type];
-  const int first = (coeff_type == 0) ? 1 : 0;
+  const int first = (coeff_type == TYPE_I16_AC) ? 1 : 0;
   Node nodes[16][NUM_NODES];
   ScoreState score_states[2][NUM_NODES];
   ScoreState* ss_cur = &SCORE_STATE(0, MIN_DELTA);
@@ -657,16 +660,17 @@ static int TrellisQuantizeBlock(const VP8Encoder* const enc,
     // test all alternate level values around level0.
     for (m = -MIN_DELTA; m <= MAX_DELTA; ++m) {
       Node* const cur = &NODE(n, m);
-      int level = level0 + m;
+      const int level = level0 + m;
       const int ctx = (level > 2) ? 2 : level;
       const int band = VP8EncBands[n + 1];
       score_t base_score;
-      score_t best_cur_score = MAX_COST;
-      int best_prev = 0;   // default, in case
+      score_t best_cur_score;
+      int best_prev;
+      score_t cost, score;
 
-      ss_cur[m].score = MAX_COST;
       ss_cur[m].costs = costs[n + 1][ctx];
       if (level < 0 || level > thresh_level) {
+        ss_cur[m].score = MAX_COST;
         // Node is dead.
         continue;
       }
@@ -682,18 +686,24 @@ static int TrellisQuantizeBlock(const VP8Encoder* const enc,
       }
 
       // Inspect all possible non-dead predecessors. Retain only the best one.
-      for (p = -MIN_DELTA; p <= MAX_DELTA; ++p) {
+      // The base_score is added to all scores so it is only added for the final
+      // value after the loop.
+      cost = VP8LevelCost(ss_prev[-MIN_DELTA].costs, level);
+      best_cur_score =
+          ss_prev[-MIN_DELTA].score + RDScoreTrellis(lambda, cost, 0);
+      best_prev = -MIN_DELTA;
+      for (p = -MIN_DELTA + 1; p <= MAX_DELTA; ++p) {
         // Dead nodes (with ss_prev[p].score >= MAX_COST) are automatically
         // eliminated since their score can't be better than the current best.
-        const score_t cost = VP8LevelCost(ss_prev[p].costs, level);
+        cost = VP8LevelCost(ss_prev[p].costs, level);
         // Examine node assuming it's a non-terminal one.
-        const score_t score =
-            base_score + ss_prev[p].score + RDScoreTrellis(lambda, cost, 0);
+        score = ss_prev[p].score + RDScoreTrellis(lambda, cost, 0);
         if (score < best_cur_score) {
           best_cur_score = score;
           best_prev = p;
         }
       }
+      best_cur_score += base_score;
       // Store best finding in current node.
       cur->sign = sign;
       cur->level = level;
@@ -701,11 +711,11 @@ static int TrellisQuantizeBlock(const VP8Encoder* const enc,
       ss_cur[m].score = best_cur_score;
 
       // Now, record best terminal node (and thus best entry in the graph).
-      if (level != 0) {
+      if (level != 0 && best_cur_score < best_score) {
         const score_t last_pos_cost =
             (n < 15) ? VP8BitCost(0, probas[band][ctx][0]) : 0;
         const score_t last_pos_score = RDScoreTrellis(lambda, last_pos_cost, 0);
-        const score_t score = best_cur_score + last_pos_score;
+        score = best_cur_score + last_pos_score;
         if (score < best_score) {
           best_score = score;
           best_path[0] = n;                     // best eob position
@@ -717,10 +727,16 @@ static int TrellisQuantizeBlock(const VP8Encoder* const enc,
   }
 
   // Fresh start
-  memset(in + first, 0, (16 - first) * sizeof(*in));
-  memset(out + first, 0, (16 - first) * sizeof(*out));
+  // Beware! We must preserve in[0]/out[0] value for TYPE_I16_AC case.
+  if (coeff_type == TYPE_I16_AC) {
+    memset(in + 1, 0, 15 * sizeof(*in));
+    memset(out + 1, 0, 15 * sizeof(*out));
+  } else {
+    memset(in, 0, 16 * sizeof(*in));
+    memset(out, 0, 16 * sizeof(*out));
+  }
   if (best_path[0] == -1) {
-    return 0;   // skip!
+    return 0;  // skip!
   }
 
   {
@@ -775,9 +791,9 @@ static int ReconstructIntra16(VP8EncIterator* const it,
     for (y = 0, n = 0; y < 4; ++y) {
       for (x = 0; x < 4; ++x, ++n) {
         const int ctx = it->top_nz_[x] + it->left_nz_[y];
-        const int non_zero =
-            TrellisQuantizeBlock(enc, tmp[n], rd->y_ac_levels[n], ctx, 0,
-                                 &dqm->y1_, dqm->lambda_trellis_i16_);
+        const int non_zero = TrellisQuantizeBlock(
+            enc, tmp[n], rd->y_ac_levels[n], ctx, TYPE_I16_AC, &dqm->y1_,
+            dqm->lambda_trellis_i16_);
         it->top_nz_[x] = it->left_nz_[y] = non_zero;
         rd->y_ac_levels[n][0] = 0;
         nz |= non_zero << n;
@@ -818,7 +834,7 @@ static int ReconstructIntra4(VP8EncIterator* const it,
   if (DO_TRELLIS_I4 && it->do_trellis_) {
     const int x = it->i4_ & 3, y = it->i4_ >> 2;
     const int ctx = it->top_nz_[x] + it->left_nz_[y];
-    nz = TrellisQuantizeBlock(enc, tmp, levels, ctx, 3, &dqm->y1_,
+    nz = TrellisQuantizeBlock(enc, tmp, levels, ctx, TYPE_I4_AC, &dqm->y1_,
                               dqm->lambda_trellis_i4_);
   } else {
     nz = VP8EncQuantizeBlock(tmp, levels, &dqm->y1_);
@@ -927,9 +943,9 @@ static int ReconstructUV(VP8EncIterator* const it, VP8ModeScore* const rd,
       for (y = 0; y < 2; ++y) {
         for (x = 0; x < 2; ++x, ++n) {
           const int ctx = it->top_nz_[4 + ch + x] + it->left_nz_[4 + ch + y];
-          const int non_zero =
-              TrellisQuantizeBlock(enc, tmp[n], rd->uv_levels[n], ctx, 2,
-                                   &dqm->uv_, dqm->lambda_trellis_uv_);
+          const int non_zero = TrellisQuantizeBlock(
+              enc, tmp[n], rd->uv_levels[n], ctx, TYPE_CHROMA_A, &dqm->uv_,
+              dqm->lambda_trellis_uv_);
           it->top_nz_[4 + ch + x] = it->left_nz_[4 + ch + y] = non_zero;
           nz |= non_zero << n;
         }

src/enc/syntax_enc.c

@@ -349,7 +349,7 @@ int VP8EncWrite(VP8Encoder* const enc) {
                                        (enc->alpha_data_size_ & 1);
     riff_size += CHUNK_HEADER_SIZE + padded_alpha_size;
   }
-  // Sanity check.
+  // RIFF size should fit in 32-bits.
   if (riff_size > 0xfffffffeU) {
     return WebPEncodingSetError(pic, VP8_ENC_ERROR_FILE_TOO_BIG);
   }