chromium-src-build/common.gypi

986 строки
38 KiB
Plaintext
Исходник Обычный вид История

# Copyright (c) 2009 The Chromium Authors. All rights reserved.
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.
# IMPORTANT:
# Please don't directly include this file if you are building via gyp_chromium,
# since gyp_chromium is automatically forcing its inclusion.
{
'variables': {
# .gyp files should set chromium_code to 1 if they build Chromium-specific
# code, as opposed to external code. This variable is used to control
# such things as the set of warnings to enable, and whether warnings are
# treated as errors.
'chromium_code%': 0,
# Variables expected to be overriden on the GYP command line (-D) or by
# ~/.gyp/include.gypi.
# Putting a variables dict inside another variables dict looks kind of
# weird. This is done so that "branding" and "buildtype" are defined as
# variables within the outer variables dict here. This is necessary
# to get these variables defined for the conditions within this variables
# dict that operate on these variables.
'variables': {
# Override branding to select the desired branding flavor.
'branding%': 'Chromium',
# Override buildtype to select the desired build flavor.
# Dev - everyday build for development/testing
# Official - release build (generally implies additional processing)
# TODO(mmoss) Once 'buildtype' is fully supported (e.g. Windows gyp
# conversion is done), some of the things which are now controlled by
# 'branding', such as symbol generation, will need to be refactored based
# on 'buildtype' (i.e. we don't care about saving symbols for non-Official
# builds).
'buildtype%': 'Dev',
# Compute the architecture that we're building for. Default to the
# architecture that we're building on.
'conditions': [
[ 'OS=="linux"', {
# This handles the Linux platforms we generally deal with. Anything
# else gets passed through, which probably won't work very well; such
# hosts should pass an explicit target_arch to gyp.
'target_arch%':
Move the spellchecker to the renderer. The motivation is that this removes the sync IPC on every call to the spellchecker. Also, currently we spellcheck in the IO thread, which frequently needs to go to disk (in particular, the entire spellcheck dictionary starts paged out), so this will block just the single renderer when that happens, rather than the whole IO thread. This breaks the SpellChecker class into two new classes. 1) On the browser side, we have SpellCheckHost. This class handles browser-wide tasks, such as keeping the custom words list in sync with the on-disk custom words dictionary, downloading missing dictionaries, etc. On Posix, it also opens the bdic file since the renderer isn't allowed to open files. SpellCheckHost is created and destroyed on the UI thread. It is initialized on the file thread. 2) On the renderer side, SpellChecker2. This class will one day be renamed SpellChecker. It handles actual checking of the words, memory maps the dictionary file, loads hunspell, etc. There is one SpellChecker2 per RenderThread (hence one per render process). My intention is for this patch to move Linux to this new approach, and follow up with ports for Windows (which will involve passing a dictionary file name rather than a file descriptor through to the renderer) and Mac (which will involve adding sync ViewHost IPC callsfor when the platform spellchecker is enabled). Note that anyone using the platform spellchecker rather than Hunspell will get no benefit out of this refactor. There should be no loss of functionality for Linux (or any other platform) in this patch. The following should all still work: - dictionary is loaded lazily - hunspell is initialized lazily, per renderer - language changes work. - Dynamic downloading of new dictionaries - auto spell correct works (as well as toggling it). - disabling spellcheck works. - custom words work (including adding in one renderer and immediately having it take effect in other renderers, for certain values of "immediate") TODO: - move spellchecker unit tests to test SpellCheck2 - add sync IPC for platform spellchecker; port to Mac - add dictionary location fallback; port to Windows - remove SpellChecker classes from browser/ BUG=25677 Review URL: http://codereview.chromium.org/357003 git-svn-id: http://src.chromium.org/svn/trunk/src/build@31199 4ff67af0-8c30-449e-8e8b-ad334ec8d88c
2009-11-06 06:05:46 +03:00
'<!(uname -m | sed -e "s/i.86/ia32/;s/x86_64/x64/;s/arm.*/arm/")',
}, { # OS!="linux"
'target_arch%': 'ia32',
}],
[ 'OS=="mac"', {
# For now, only Linux and Windows use spellcheck in the renderer.
Move the spellchecker to the renderer. The motivation is that this removes the sync IPC on every call to the spellchecker. Also, currently we spellcheck in the IO thread, which frequently needs to go to disk (in particular, the entire spellcheck dictionary starts paged out), so this will block just the single renderer when that happens, rather than the whole IO thread. This breaks the SpellChecker class into two new classes. 1) On the browser side, we have SpellCheckHost. This class handles browser-wide tasks, such as keeping the custom words list in sync with the on-disk custom words dictionary, downloading missing dictionaries, etc. On Posix, it also opens the bdic file since the renderer isn't allowed to open files. SpellCheckHost is created and destroyed on the UI thread. It is initialized on the file thread. 2) On the renderer side, SpellChecker2. This class will one day be renamed SpellChecker. It handles actual checking of the words, memory maps the dictionary file, loads hunspell, etc. There is one SpellChecker2 per RenderThread (hence one per render process). My intention is for this patch to move Linux to this new approach, and follow up with ports for Windows (which will involve passing a dictionary file name rather than a file descriptor through to the renderer) and Mac (which will involve adding sync ViewHost IPC callsfor when the platform spellchecker is enabled). Note that anyone using the platform spellchecker rather than Hunspell will get no benefit out of this refactor. There should be no loss of functionality for Linux (or any other platform) in this patch. The following should all still work: - dictionary is loaded lazily - hunspell is initialized lazily, per renderer - language changes work. - Dynamic downloading of new dictionaries - auto spell correct works (as well as toggling it). - disabling spellcheck works. - custom words work (including adding in one renderer and immediately having it take effect in other renderers, for certain values of "immediate") TODO: - move spellchecker unit tests to test SpellCheck2 - add sync IPC for platform spellchecker; port to Mac - add dictionary location fallback; port to Windows - remove SpellChecker classes from browser/ BUG=25677 Review URL: http://codereview.chromium.org/357003 git-svn-id: http://src.chromium.org/svn/trunk/src/build@31199 4ff67af0-8c30-449e-8e8b-ad334ec8d88c
2009-11-06 06:05:46 +03:00
'spellchecker_in_renderer%': 0,
}, { # OS!="mac"
'spellchecker_in_renderer%': 1,
}],
],
# We do want to build Chromium with Breakpad support in certain
# situations. I.e. for Chrome bot.
'linux_chromium_breakpad%': 0,
# And if we want to dump symbols.
'linux_chromium_dump_symbols%': 0,
# Also see linux_strip_binary below.
# By default, Linux does not use views. To turn on views in Linux,
# set the variable GYP_DEFINES to "toolkit_views=1", or modify
# ~/.gyp/include.gypi .
'toolkit_views%': 0,
# Defaults to a desktop build, overridden via command line/env.
'chromeos%': 0,
# This variable tells WebCore.gyp and JavaScriptCore.gyp whether they are
# are built under a chromium full build (1) or a webkit.org chromium
# build (0).
'inside_chromium_build%': 1,
# Set to 1 to enable fast builds. It disables debug info for fastest
# compilation.
'fastbuild%': 0,
},
# Define branding and buildtype on the basis of their settings within the
# variables sub-dict above, unless overridden.
'branding%': '<(branding)',
'buildtype%': '<(buildtype)',
'target_arch%': '<(target_arch)',
Move the spellchecker to the renderer. The motivation is that this removes the sync IPC on every call to the spellchecker. Also, currently we spellcheck in the IO thread, which frequently needs to go to disk (in particular, the entire spellcheck dictionary starts paged out), so this will block just the single renderer when that happens, rather than the whole IO thread. This breaks the SpellChecker class into two new classes. 1) On the browser side, we have SpellCheckHost. This class handles browser-wide tasks, such as keeping the custom words list in sync with the on-disk custom words dictionary, downloading missing dictionaries, etc. On Posix, it also opens the bdic file since the renderer isn't allowed to open files. SpellCheckHost is created and destroyed on the UI thread. It is initialized on the file thread. 2) On the renderer side, SpellChecker2. This class will one day be renamed SpellChecker. It handles actual checking of the words, memory maps the dictionary file, loads hunspell, etc. There is one SpellChecker2 per RenderThread (hence one per render process). My intention is for this patch to move Linux to this new approach, and follow up with ports for Windows (which will involve passing a dictionary file name rather than a file descriptor through to the renderer) and Mac (which will involve adding sync ViewHost IPC callsfor when the platform spellchecker is enabled). Note that anyone using the platform spellchecker rather than Hunspell will get no benefit out of this refactor. There should be no loss of functionality for Linux (or any other platform) in this patch. The following should all still work: - dictionary is loaded lazily - hunspell is initialized lazily, per renderer - language changes work. - Dynamic downloading of new dictionaries - auto spell correct works (as well as toggling it). - disabling spellcheck works. - custom words work (including adding in one renderer and immediately having it take effect in other renderers, for certain values of "immediate") TODO: - move spellchecker unit tests to test SpellCheck2 - add sync IPC for platform spellchecker; port to Mac - add dictionary location fallback; port to Windows - remove SpellChecker classes from browser/ BUG=25677 Review URL: http://codereview.chromium.org/357003 git-svn-id: http://src.chromium.org/svn/trunk/src/build@31199 4ff67af0-8c30-449e-8e8b-ad334ec8d88c
2009-11-06 06:05:46 +03:00
'spellchecker_in_renderer%': '<(spellchecker_in_renderer)',
'toolkit_views%': '<(toolkit_views)',
'chromeos%': '<(chromeos)',
'inside_chromium_build%': '<(inside_chromium_build)',
'fastbuild%': '<(fastbuild)',
# Override chromium_mac_pch and set it to 0 to suppress the use of
# precompiled headers on the Mac. Prefix header injection may still be
# used, but prefix headers will not be precompiled. This is useful when
# using distcc to distribute a build to compile slaves that don't
# share the same compiler executable as the system driving the compilation,
# because precompiled headers rely on pointers into a specific compiler
# executable's image. Setting this to 0 is needed to use an experimental
# Linux-Mac cross compiler distcc farm.
'chromium_mac_pch%': 1,
# Mac OS X SDK and deployment target support.
# The SDK identifies the version of the system headers that will be used,
# and corresponds to the MAC_OS_X_VERSION_MAX_ALLOWED compile-time macro.
# "Maximum allowed" refers to the operating system version whose APIs are
# available in the headers.
# The deployment target identifies the minimum system version that the
# built products are expected to function on. It corresponds to the
# MAC_OS_X_VERSION_MIN_REQUIRED compile-time macro.
# To ensure these macros are available, #include <AvailabilityMacros.h>.
# Additional documentation on these macros is available at
# http://developer.apple.com/mac/library/technotes/tn2002/tn2064.html#SECTION3
# Chrome normally builds with the Mac OS X 10.5 SDK and sets the
# deployment target to 10.5. Other projects, such as O3D, may override
# these defaults.
'mac_sdk%': '10.5',
'mac_deployment_target%': '10.5',
# Set to 1 to enable code coverage. In addition to build changes
# (e.g. extra CFLAGS), also creates a new target in the src/chrome
# project file called "coverage".
# Currently ignored on Windows.
'coverage%': 0,
# Overridable specification for potential use of alternative
# JavaScript engines.
'javascript_engine%': 'v8',
# To do a shared build on linux we need to be able to choose between type
# static_library and shared_library. We default to doing a static build
# but you can override this with "gyp -Dlibrary=shared_library" or you
# can add the following line (without the #) to ~/.gyp/include.gypi
# {'variables': {'library': 'shared_library'}}
# to compile as shared by default
'library%': 'static_library',
# The Google Update appid.
'google_update_appid%': '{8A69D345-D564-463c-AFF1-A69D9E530F96}',
# Whether to add the experimental build define.
'chrome_frame_define%': 0,
# Whether pepper APIs are enabled.
'enable_pepper%': 0,
# Whether usage of OpenMAX is enabled.
'enable_openmax%': 0,
# TODO(bradnelson): eliminate this when possible.
# To allow local gyp files to prevent release.vsprops from being included.
# Yes(1) means include release.vsprops.
# Once all vsprops settings are migrated into gyp, this can go away.
'msvs_use_common_release%': 1,
# TODO(bradnelson): eliminate this when possible.
# To allow local gyp files to override additional linker options for msvs.
# Yes(1) means set use the common linker options.
'msvs_use_common_linker_extras%': 1,
# TODO(sgk): eliminate this if possible.
# It would be nicer to support this via a setting in 'target_defaults'
# in chrome/app/locales/locales.gypi overriding the setting in the
# 'Debug' configuration in the 'target_defaults' dict below,
# but that doesn't work as we'd like.
'msvs_debug_link_incremental%': '2',
# The system root for cross-compiles. Default: none.
'sysroot%': '',
# This is the location of the sandbox binary. Chrome looks for this before
# running the zygote process. If found, and SUID, it will be used to
# sandbox the zygote process and, thus, all renderer processes.
'linux_sandbox_path%': '',
# Set this to true to enable SELinux support.
'selinux%': 0,
# Strip the binary after dumping symbols.
'linux_strip_binary%': 0,
# Enable TCMalloc.
'linux_use_tcmalloc%': 0,
# Set to select the Title Case versions of strings in GRD files.
'use_titlecase_in_grd_files%': 0,
# Used to disable Native Client at compile time, for platforms where it
# isn't supported
'disable_nacl%': 0,
# Set ARM-v7 compilation flags
'armv7%': 0,
'conditions': [
['OS=="linux"', {
# This will set gcc_version to XY if you are running gcc X.Y.*.
# This is used to tweak build flags for gcc 4.4.
'gcc_version%': '<!(python <(DEPTH)/build/compiler_version.py)',
'conditions': [
['branding=="Chrome" or linux_chromium_breakpad==1', {
'linux_breakpad%': 1,
}, {
'linux_breakpad%': 0,
}],
# All Chrome builds have breakpad symbols, but only process the
# symbols from official builds.
# TODO(mmoss) dump_syms segfaults on x64. Enable once dump_syms and
# crash server handle 64-bit symbols.
['linux_chromium_dump_symbols==1 or '
'(branding=="Chrome" and buildtype=="Official" and '
'target_arch=="ia32")', {
'linux_dump_symbols%': 1,
}, {
'linux_dump_symbols%': 0,
}],
['toolkit_views==0', {
# GTK wants Title Case strings
'use_titlecase_in_grd_files%': 1,
}],
],
}], # OS=="linux"
['OS=="mac"', {
# Mac wants Title Case strings
'use_titlecase_in_grd_files%': 1,
'conditions': [
# mac_product_name is set to the name of the .app bundle as it should
# appear on disk. This duplicates data from
# chrome/app/theme/chromium/BRANDING and
# chrome/app/theme/google_chrome/BRANDING, but is necessary to get
# these names into the build system.
['branding=="Chrome"', {
'mac_product_name%': 'Google Chrome',
}, { # else: branding!="Chrome"
'mac_product_name%': 'Chromium',
}],
# Feature variables for enabling Mac Breakpad and Keystone auto-update
# support. Both features are on by default in official builds with
# Chrome branding.
['branding=="Chrome" and buildtype=="Official"', {
'mac_breakpad%': 1,
'mac_keystone%': 1,
}, { # else: branding!="Chrome" or buildtype!="Official"
'mac_breakpad%': 0,
'mac_keystone%': 0,
}],
],
}], # OS=="mac"
# Whether to use multiple cores to compile with visual studio. This is
# optional because it sometimes causes corruption on VS 2005.
# It is on by default on VS 2008 and off on VS 2005.
['OS=="win"', {
'conditions': [
['MSVS_VERSION=="2005"', {
'msvs_multi_core_compile%': 0,
},{
'msvs_multi_core_compile%': 1,
}],
# Don't do incremental linking for large modules on 32-bit.
['MSVS_OS_BITS==32', {
'msvs_large_module_debug_link_mode%': '1', # No
},{
'msvs_large_module_debug_link_mode%': '2', # Yes
}],
],
}],
],
# NOTE: When these end up in the Mac bundle, we need to replace '-' for '_'
# so Cocoa is happy (http://crbug.com/20441).
'locales': [
'am', 'ar', 'bg', 'bn', 'ca', 'cs', 'da', 'de', 'el', 'en-GB',
'en-US', 'es-419', 'es', 'et', 'fi', 'fil', 'fr', 'gu', 'he',
'hi', 'hr', 'hu', 'id', 'it', 'ja', 'kn', 'ko', 'lt', 'lv',
'ml', 'mr', 'nb', 'nl', 'or', 'pl', 'pt-BR', 'pt-PT', 'ro',
'ru', 'sk', 'sl', 'sr', 'sv', 'sw', 'ta', 'te', 'th', 'tr',
'uk', 'vi', 'zh-CN', 'zh-TW',
],
},
'target_defaults': {
'variables': {
'mac_release_optimization%': '3', # Use -O3 unless overridden
'mac_debug_optimization%': '0', # Use -O0 unless overridden
'release_extra_cflags%': '',
'debug_extra_cflags%': '',
'release_valgrind_build%': 0,
},
'conditions': [
['branding=="Chrome"', {
'defines': ['GOOGLE_CHROME_BUILD'],
}, { # else: branding!="Chrome"
'defines': ['CHROMIUM_BUILD'],
}],
['chrome_frame_define', {
'defines': ['CHROME_FRAME_BUILD'],
}],
['toolkit_views==1', {
'defines': ['TOOLKIT_VIEWS=1'],
}],
['chromeos==1', {
'defines': ['CHROMEOS_TRANSITIONAL=1'],
}],
['chromeos==1 or toolkit_views==1', {
'defines': ['OS_CHROMEOS=1'],
}],
['enable_pepper==1', {
'defines': ['ENABLE_PEPPER=1'],
}],
Move the spellchecker to the renderer. The motivation is that this removes the sync IPC on every call to the spellchecker. Also, currently we spellcheck in the IO thread, which frequently needs to go to disk (in particular, the entire spellcheck dictionary starts paged out), so this will block just the single renderer when that happens, rather than the whole IO thread. This breaks the SpellChecker class into two new classes. 1) On the browser side, we have SpellCheckHost. This class handles browser-wide tasks, such as keeping the custom words list in sync with the on-disk custom words dictionary, downloading missing dictionaries, etc. On Posix, it also opens the bdic file since the renderer isn't allowed to open files. SpellCheckHost is created and destroyed on the UI thread. It is initialized on the file thread. 2) On the renderer side, SpellChecker2. This class will one day be renamed SpellChecker. It handles actual checking of the words, memory maps the dictionary file, loads hunspell, etc. There is one SpellChecker2 per RenderThread (hence one per render process). My intention is for this patch to move Linux to this new approach, and follow up with ports for Windows (which will involve passing a dictionary file name rather than a file descriptor through to the renderer) and Mac (which will involve adding sync ViewHost IPC callsfor when the platform spellchecker is enabled). Note that anyone using the platform spellchecker rather than Hunspell will get no benefit out of this refactor. There should be no loss of functionality for Linux (or any other platform) in this patch. The following should all still work: - dictionary is loaded lazily - hunspell is initialized lazily, per renderer - language changes work. - Dynamic downloading of new dictionaries - auto spell correct works (as well as toggling it). - disabling spellcheck works. - custom words work (including adding in one renderer and immediately having it take effect in other renderers, for certain values of "immediate") TODO: - move spellchecker unit tests to test SpellCheck2 - add sync IPC for platform spellchecker; port to Mac - add dictionary location fallback; port to Windows - remove SpellChecker classes from browser/ BUG=25677 Review URL: http://codereview.chromium.org/357003 git-svn-id: http://src.chromium.org/svn/trunk/src/build@31199 4ff67af0-8c30-449e-8e8b-ad334ec8d88c
2009-11-06 06:05:46 +03:00
['spellchecker_in_renderer==1', {
'defines': ['SPELLCHECKER_IN_RENDERER=1']
}],
['fastbuild!=0', {
'conditions': [
# Finally, for Windows, we simply turn on profiling.
['OS=="win"', {
'msvs_settings': {
'VCLinkerTool': {
'GenerateDebugInformation': 'false',
},
'VCCLCompilerTool': {
'DebugInformationFormat': '0',
}
}
}], # OS==win
], # conditions for fastbuild.
}], # fastbuild!=0
['selinux==1', {
'defines': ['CHROMIUM_SELINUX=1'],
}],
['coverage!=0', {
'conditions': [
['OS=="mac"', {
'xcode_settings': {
'GCC_INSTRUMENT_PROGRAM_FLOW_ARCS': 'YES', # -fprofile-arcs
'GCC_GENERATE_TEST_COVERAGE_FILES': 'YES', # -ftest-coverage
},
# Add -lgcov for executables and shared_libraries, not for
# static_libraries. This is a delayed conditional.
'target_conditions': [
['_type=="executable" or _type=="shared_library"', {
'xcode_settings': { 'OTHER_LDFLAGS': [ '-lgcov' ] },
}],
],
}],
# Linux gyp (into scons) doesn't like target_conditions?
# TODO(???): track down why 'target_conditions' doesn't work
# on Linux gyp into scons like it does on Mac gyp into xcodeproj.
['OS=="linux"', {
'cflags': [ '-ftest-coverage',
'-fprofile-arcs' ],
'link_settings': { 'libraries': [ '-lgcov' ] },
}],
# Finally, for Windows, we simply turn on profiling.
['OS=="win"', {
'msvs_settings': {
'VCLinkerTool': {
'Profile': 'true',
},
'VCCLCompilerTool': {
# /Z7, not /Zi, so coverage is happyb
'DebugInformationFormat': '1',
'AdditionalOptions': '/Yd',
}
}
}], # OS==win
], # conditions for coverage
}], # coverage!=0
], # conditions for 'target_defaults'
'default_configuration': 'Debug',
'configurations': {
# VCLinkerTool LinkIncremental values below:
# 0 == default
# 1 == /INCREMENTAL:NO
# 2 == /INCREMENTAL
# Debug links incremental, Release does not.
'Common': {
'abstract': 1,
'msvs_configuration_attributes': {
'OutputDirectory': '$(SolutionDir)$(ConfigurationName)',
'IntermediateDirectory': '$(OutDir)\\obj\\$(ProjectName)',
'CharacterSet': '1',
},
'msvs_configuration_platform': 'Win32',
},
'Debug': {
'inherit_from': ['Common'],
'xcode_settings': {
'COPY_PHASE_STRIP': 'NO',
'GCC_OPTIMIZATION_LEVEL': '<(mac_debug_optimization)',
'OTHER_CFLAGS': [ '<@(debug_extra_cflags)', ],
},
'msvs_settings': {
'VCCLCompilerTool': {
'Optimization': '0',
'PreprocessorDefinitions': ['_DEBUG'],
'BasicRuntimeChecks': '3',
'RuntimeLibrary': '1',
},
'VCLinkerTool': {
'LinkIncremental': '<(msvs_debug_link_incremental)',
},
'VCResourceCompilerTool': {
'PreprocessorDefinitions': ['_DEBUG'],
},
},
'conditions': [
['OS=="linux"', {
'cflags': [
'<@(debug_extra_cflags)',
],
}],
],
},
'Release': {
'inherit_from': ['Common'],
'defines': [
'NDEBUG',
],
'xcode_settings': {
'DEAD_CODE_STRIPPING': 'YES', # -Wl,-dead_strip
'GCC_OPTIMIZATION_LEVEL': '<(mac_release_optimization)',
'OTHER_CFLAGS': [ '<@(release_extra_cflags)', ],
},
'msvs_settings': {
'VCLinkerTool': {
'LinkIncremental': '1',
},
},
'conditions': [
['release_valgrind_build==0', {
'defines': ['NVALGRIND'],
}],
['msvs_use_common_release', {
'msvs_props': ['release.vsprops'],
}],
['OS=="linux"', {
'cflags': [
'<@(release_extra_cflags)',
],
}],
],
},
'conditions': [
[ 'OS=="win"', {
# TODO(bradnelson): add a gyp mechanism to make this more graceful.
'Purify': {
'inherit_from': ['Release'],
'defines': [
'PURIFY',
'NO_TCMALLOC',
],
'msvs_settings': {
'VCCLCompilerTool': {
'Optimization': '0',
'RuntimeLibrary': '0',
'BufferSecurityCheck': 'false',
},
'VCLinkerTool': {
'EnableCOMDATFolding': '1',
'LinkIncremental': '1',
},
},
},
'Release - no tcmalloc': {
'inherit_from': ['Release'],
'defines': ['NO_TCMALLOC'],
},
'Debug_x64': {
'inherit_from': ['Debug'],
'msvs_configuration_platform': 'x64',
},
'Release_x64': {
'inherit_from': ['Release'],
'msvs_configuration_platform': 'x64',
},
'Purify_x64': {
'inherit_from': ['Purify'],
'msvs_configuration_platform': 'x64',
},
'Release - no tcmalloc_x64': {
'inherit_from': ['Release - no tcmalloc'],
'msvs_configuration_platform': 'x64',
},
}],
],
},
},
'conditions': [
['OS=="linux"', {
'target_defaults': {
# Enable -Werror by default, but put it in a variable so it can
# be disabled in ~/.gyp/include.gypi on the valgrind builders.
'variables': {
'werror%': '-Werror',
'no_strict_aliasing%': 0,
},
'cflags': [
'<(werror)', # See note above about the werror variable.
'-pthread',
# We don't use exceptions. By disabling exceptions
# (and asynchronous-unwind-tables), we shave off 2.5mb from
# our resulting binary by not including the eh_frame section.
'-fno-exceptions',
'-fno-asynchronous-unwind-tables',
'-fvisibility=hidden',
'-Wall',
'-D_FILE_OFFSET_BITS=64',
],
'cflags_cc': [
'-fno-rtti',
'-fno-threadsafe-statics',
# Make inline functions have hidden visiblity by default.
# Surprisingly, not covered by -fvisibility=hidden.
'-fvisibility-inlines-hidden',
],
'ldflags': [
'-pthread',
],
'scons_variable_settings': {
'LIBPATH': ['$LIB_DIR'],
# Linking of large files uses lots of RAM, so serialize links
# using the handy flock command from util-linux.
'FLOCK_LINK': ['flock', '$TOP_BUILDDIR/linker.lock', '$LINK'],
'FLOCK_SHLINK': ['flock', '$TOP_BUILDDIR/linker.lock', '$SHLINK'],
'FLOCK_LDMODULE': ['flock', '$TOP_BUILDDIR/linker.lock', '$LDMODULE'],
# We have several cases where archives depend on each other in
# a cyclic fashion. Since the GNU linker does only a single
# pass over the archives we surround the libraries with
# --start-group and --end-group (aka -( and -) ). That causes
# ld to loop over the group until no more undefined symbols
# are found. In an ideal world we would only make groups from
# those libraries which we knew to be in cycles. However,
# that's tough with SCons, so we bodge it by making all the
# archives a group by redefining the linking command here.
#
# TODO: investigate whether we still have cycles that
# require --{start,end}-group. There has been a lot of
# refactoring since this was first coded, which might have
# eliminated the circular dependencies.
#
# Note: $_LIBDIRFLAGS comes before ${LINK,SHLINK,LDMODULE}FLAGS
# so that we prefer our own built libraries (e.g. -lpng) to
# system versions of libraries that pkg-config might turn up.
# TODO(sgk): investigate handling this not by re-ordering the
# flags this way, but by adding a hook to use the SCons
# ParseFlags() option on the output from pkg-config.
'LINKCOM': [['$FLOCK_LINK', '-o', '$TARGET',
'$_LIBDIRFLAGS', '$LINKFLAGS', '$SOURCES',
'-Wl,--start-group', '$_LIBFLAGS', '-Wl,--end-group']],
'SHLINKCOM': [['$FLOCK_SHLINK', '-o', '$TARGET',
'$_LIBDIRFLAGS', '$SHLINKFLAGS', '$SOURCES',
'-Wl,--start-group', '$_LIBFLAGS', '-Wl,--end-group']],
'LDMODULECOM': [['$FLOCK_LDMODULE', '-o', '$TARGET',
'$_LIBDIRFLAGS', '$LDMODULEFLAGS', '$SOURCES',
'-Wl,--start-group', '$_LIBFLAGS', '-Wl,--end-group']],
'IMPLICIT_COMMAND_DEPENDENCIES': 0,
# -rpath is only used when building with shared libraries.
'conditions': [
[ 'library=="shared_library"', {
'RPATH': '$LIB_DIR',
}],
],
},
'scons_import_variables': [
'AS',
'CC',
'CXX',
'LINK',
],
'scons_propagate_variables': [
'AS',
'CC',
'CCACHE_DIR',
'CXX',
'DISTCC_DIR',
'DISTCC_HOSTS',
'HOME',
'INCLUDE_SERVER_ARGS',
'INCLUDE_SERVER_PORT',
'LINK',
'CHROME_BUILD_TYPE',
'CHROMIUM_BUILD',
'OFFICIAL_BUILD',
],
'configurations': {
'Debug': {
'variables': {
'debug_optimize%': '0',
},
'defines': [
'_DEBUG',
],
'cflags': [
'-O>(debug_optimize)',
'-g',
Linux: add Breakpad support This commits a rewrite of the Breakpad Linux client. The old code: * Had a number of plain bugs in it, but those could just have been fixed. * Allocated memory from the heap, which is a no go. * Made libc calls which can enter the dynamic linker - another source of crashes. * Didn't understand some of the tricks needed, like clone() via libc will write to random areas of memory because it assumes that it's only called from libpthread Additionally, we had one more requirement which meant changing the interface: * We need to be able to crash dump the renderers from the browser process. And that last one really needed a rewrite. We intend to try and upstream this new code into Breakpad. The new Breakpad design works like this: When a renderer crashes, a signal handler runs on an alternative stack and collects information about the registers of the thread before the crash. Then we enter Chromium specific code an send a datagram message to a magic file descriptor (4) containing: * the registers and tid of the crashing thread * the active URL * a file descriptor to a socket * a CREDENTIALS structure giving the PID of the renderer. On the other end of the socket is an object on the IO thread (render_crash_handler_host_linux.cc) which reads and parses the datagram. The CREDENTIALS structure is validated by the kernel, so the renderer can't lie about it's PID and try and get the browser to crash dump the wrong process. The browser then ptraces the renderer and extracts all the needed information to write a minidump to a temp file. Then we write a byte to the file descriptor which the renderer gave the browser in the datagram and that's the signal to the renderer to finish dying. It dies by sending itself the same signal which trigger the crash dump in the first place, so it will appear to crash as normal as far as kernel core dumps and waitpid are concerned. The browser then constucts a MIME message in a temp file for upload to the crash service. We then fork out to /usr/bin/wget to actually do the upload (since Debian numbers suggest that 99.8% of users have wget installed.) A second forked child unlinks the temp files once wget has completed. For a browser crash, everything works pretty much the same except that the datagram step is omitted and we clone() off a process to ptrace ourselves and write the minidump. This code is only enabled in Chrome branded builds. Stub source files are substituted in the case of a Chromium build. http://codereview.chromium.org/115526 BUG=9646,10772 TEST=Build a Chrome branded binary. Send SEGV to a renderer and verify that wget output appears on stderr. Send a SEGV to the main binary and verify the same. git-svn-id: http://src.chromium.org/svn/trunk/src/build@16719 4ff67af0-8c30-449e-8e8b-ad334ec8d88c
2009-05-22 07:37:45 +04:00
# One can use '-gstabs' to enable building the debugging
# information in STABS format for breakpad's dumpsyms.
],
'ldflags': [
'-rdynamic', # Allows backtrace to resolve symbols.
],
},
'Release': {
'variables': {
'release_optimize%': '2',
},
'cflags': [
'-O>(release_optimize)',
# Don't emit the GCC version ident directives, they just end up
# in the .comment section taking up binary size.
'-fno-ident',
# Put data and code in their own sections, so that unused symbols
# can be removed at link time with --gc-sections.
'-fdata-sections',
'-ffunction-sections',
],
},
},
'variants': {
'coverage': {
'cflags': ['-fprofile-arcs', '-ftest-coverage'],
'ldflags': ['-fprofile-arcs'],
},
'profile': {
'cflags': ['-pg', '-g'],
'ldflags': ['-pg'],
},
'symbols': {
'cflags': ['-g'],
},
},
'conditions': [
[ 'target_arch=="ia32"', {
'asflags': [
# Needed so that libs with .s files (e.g. libicudata.a)
# are compatible with the general 32-bit-ness.
'-32',
],
# All floating-point computations on x87 happens in 80-bit
# precision. Because the C and C++ language standards allow
# the compiler to keep the floating-point values in higher
# precision than what's specified in the source and doing so
# is more efficient than constantly rounding up to 64-bit or
# 32-bit precision as specified in the source, the compiler,
# especially in the optimized mode, tries very hard to keep
# values in x87 floating-point stack (in 80-bit precision)
# as long as possible. This has important side effects, that
# the real value used in computation may change depending on
# how the compiler did the optimization - that is, the value
# kept in 80-bit is different than the value rounded down to
# 64-bit or 32-bit. There are possible compiler options to make
# this behavior consistent (e.g. -ffloat-store would keep all
# floating-values in the memory, thus force them to be rounded
# to its original precision) but they have significant runtime
# performance penalty.
#
# -mfpmath=sse -msse2 makes the compiler use SSE instructions
# which keep floating-point values in SSE registers in its
# native precision (32-bit for single precision, and 64-bit for
# double precision values). This means the floating-point value
# used during computation does not change depending on how the
# compiler optimized the code, since the value is always kept
# in its specified precision.
'conditions': [
['branding=="Chromium"', {
'cflags': [
'-march=pentium4',
'-msse2',
'-mfpmath=sse',
],
}],
],
'cflags': [
'-m32',
],
'ldflags': [
'-m32',
],
}],
['target_arch=="arm"', {
'conditions': [
['armv7==1', {
'target_conditions': [
['_toolset=="target"', {
'cflags': [
'-march=armv7-a',
'-mtune=cortex-a8',
'-mfpu=neon',
'-mfloat-abi=softfp',
],
}],
],
}],
],
}],
['sysroot!=""', {
'target_conditions': [
['_toolset=="target"', {
'cflags': [
'--sysroot=<(sysroot)',
],
'ldflags': [
'--sysroot=<(sysroot)',
],
}]]
}],
['no_strict_aliasing==1', {
'cflags': [
'-fno-strict-aliasing',
],
}],
['linux_breakpad==1', {
'cflags': [ '-gstabs' ],
'defines': ['USE_LINUX_BREAKPAD'],
}],
['library=="shared_library"', {
# When building with shared libraries, remove the visiblity-hiding
# flag.
'cflags!': [ '-fvisibility=hidden' ],
'conditions': [
['target_arch=="x64"', {
# Shared libraries need -fPIC on x86-64
'cflags': ['-fPIC']
}]
],
}],
['linux_use_tcmalloc==1', {
'defines': ['LINUX_USE_TCMALLOC'],
}],
],
},
}],
['OS=="mac"', {
'target_defaults': {
'variables': {
# This should be 'mac_real_dsym%', but there seems to be a bug
# with % in variables that are intended to be set to different
# values in different targets, like this one.
'mac_real_dsym': 0, # Fake .dSYMs are fine in most cases.
},
'mac_bundle': 0,
'xcode_settings': {
'ALWAYS_SEARCH_USER_PATHS': 'NO',
'GCC_C_LANGUAGE_STANDARD': 'c99', # -std=c99
'GCC_CW_ASM_SYNTAX': 'NO', # No -fasm-blocks
'GCC_DYNAMIC_NO_PIC': 'NO', # No -mdynamic-no-pic
# (Equivalent to -fPIC)
'GCC_ENABLE_CPP_EXCEPTIONS': 'NO', # -fno-exceptions
'GCC_ENABLE_CPP_RTTI': 'NO', # -fno-rtti
'GCC_ENABLE_PASCAL_STRINGS': 'NO', # No -mpascal-strings
# GCC_INLINES_ARE_PRIVATE_EXTERN maps to -fvisibility-inlines-hidden
'GCC_INLINES_ARE_PRIVATE_EXTERN': 'YES',
'GCC_OBJC_CALL_CXX_CDTORS': 'YES', # -fobjc-call-cxx-cdtors
'GCC_SYMBOLS_PRIVATE_EXTERN': 'YES', # -fvisibility=hidden
'GCC_THREADSAFE_STATICS': 'NO', # -fno-threadsafe-statics
'GCC_TREAT_WARNINGS_AS_ERRORS': 'YES', # -Werror
'GCC_VERSION': '4.2',
'GCC_WARN_ABOUT_MISSING_NEWLINE': 'YES', # -Wnewline-eof
# MACOSX_DEPLOYMENT_TARGET maps to -mmacosx-version-min
'MACOSX_DEPLOYMENT_TARGET': '<(mac_deployment_target)',
'PREBINDING': 'NO', # No -Wl,-prebind
'USE_HEADERMAP': 'NO',
'WARNING_CFLAGS': ['-Wall', '-Wendif-labels'],
'conditions': [
['chromium_mac_pch', {'GCC_PRECOMPILE_PREFIX_HEADER': 'YES'},
{'GCC_PRECOMPILE_PREFIX_HEADER': 'NO'}
],
],
},
'target_conditions': [
['_type!="static_library"', {
'xcode_settings': {'OTHER_LDFLAGS': ['-Wl,-search_paths_first']},
}],
['_mac_bundle', {
'xcode_settings': {'OTHER_LDFLAGS': ['-Wl,-ObjC']},
}],
['_type=="executable" or _type=="shared_library"', {
'target_conditions': [
['mac_real_dsym == 1', {
# To get a real .dSYM bundle produced by dsymutil, set the
# debug information format to dwarf-with-dsym. Since
# strip_from_xcode will not be used, set Xcode to do the
# stripping as well.
'configurations': {
'Release': {
'xcode_settings': {
'DEBUG_INFORMATION_FORMAT': 'dwarf-with-dsym',
'DEPLOYMENT_POSTPROCESSING': 'YES',
'STRIP_INSTALLED_PRODUCT': 'YES',
'target_conditions': [
['_type=="shared_library"', {
# The Xcode default is to strip debugging symbols
# only (-S). Local symbols should be stripped as
# well, which will be handled by -x. Xcode will
# continue to insert -S when stripping even when
# additional flags are added with STRIPFLAGS.
'STRIPFLAGS': '-x',
}], # _type=="shared_library"
], # target_conditions
}, # xcode_settings
}, # configuration "Release"
}, # configurations
}, { # mac_real_dsym != 1
# To get a fast fake .dSYM bundle, use a post-build step to
# produce the .dSYM and strip the executable. strip_from_xcode
# only operates in the Release configuration.
'postbuilds': [
{
'variables': {
# Define strip_from_xcode in a variable ending in _path
# so that gyp understands it's a path and performs proper
# relativization during dict merging.
'strip_from_xcode_path': 'mac/strip_from_xcode',
},
'postbuild_name': 'Strip If Needed',
'action': ['<(strip_from_xcode_path)'],
},
], # postbuilds
}], # mac_real_dsym
], # target_conditions
}], # _type=="executable" or _type=="shared_library"
], # target_conditions
}, # target_defaults
}], # OS=="mac"
['OS=="win"', {
'target_defaults': {
'defines': [
'_WIN32_WINNT=0x0600',
'WINVER=0x0600',
'WIN32',
'_WINDOWS',
'_HAS_EXCEPTIONS=0',
'NOMINMAX',
'_CRT_RAND_S',
'CERT_CHAIN_PARA_HAS_EXTRA_FIELDS',
'WIN32_LEAN_AND_MEAN',
'_SECURE_ATL',
'_HAS_TR1=0',
],
'msvs_system_include_dirs': [
'<(DEPTH)/third_party/platformsdk_win2008_6_1/files/Include',
'$(VSInstallDir)/VC/atlmfc/include',
],
'msvs_cygwin_dirs': ['<(DEPTH)/third_party/cygwin'],
'msvs_disabled_warnings': [4396, 4503, 4819],
'msvs_settings': {
'VCCLCompilerTool': {
'MinimalRebuild': 'false',
'ExceptionHandling': '0',
'BufferSecurityCheck': 'true',
'EnableFunctionLevelLinking': 'true',
'RuntimeTypeInfo': 'false',
'WarningLevel': '3',
'WarnAsError': 'true',
'DebugInformationFormat': '3',
'conditions': [
[ 'msvs_multi_core_compile', {
'AdditionalOptions': '/MP',
}],
],
},
'VCLibrarianTool': {
'AdditionalOptions': '/ignore:4221',
'AdditionalLibraryDirectories':
['<(DEPTH)/third_party/platformsdk_win2008_6_1/files/Lib'],
},
'VCLinkerTool': {
'AdditionalDependencies': [
'wininet.lib',
'version.lib',
'msimg32.lib',
'ws2_32.lib',
'usp10.lib',
'psapi.lib',
'dbghelp.lib',
],
'AdditionalLibraryDirectories':
['<(DEPTH)/third_party/platformsdk_win2008_6_1/files/Lib'],
'GenerateDebugInformation': 'true',
'MapFileName': '$(OutDir)\\$(TargetName).map',
'ImportLibrary': '$(OutDir)\\lib\\$(TargetName).lib',
'TargetMachine': '1',
'FixedBaseAddress': '1',
# SubSystem values:
# 0 == not set
# 1 == /SUBSYSTEM:CONSOLE
# 2 == /SUBSYSTEM:WINDOWS
# Most of the executables we'll ever create are tests
# and utilities with console output.
'SubSystem': '1',
},
'VCMIDLTool': {
'GenerateStublessProxies': 'true',
'TypeLibraryName': '$(InputName).tlb',
'OutputDirectory': '$(IntDir)',
'HeaderFileName': '$(InputName).h',
'DLLDataFileName': 'dlldata.c',
'InterfaceIdentifierFileName': '$(InputName)_i.c',
'ProxyFileName': '$(InputName)_p.c',
},
'VCResourceCompilerTool': {
'Culture' : '1033',
'AdditionalIncludeDirectories': ['<(DEPTH)'],
},
},
},
}],
['chromium_code==0', {
# This section must follow the other conditon sections above because
# external_code.gypi expects to be merged into those settings.
'includes': [
'external_code.gypi',
],
}, {
'target_defaults': {
# In Chromium code, we define __STDC_FORMAT_MACROS in order to get the
# C99 macros on Mac and Linux.
'defines': [
'__STDC_FORMAT_MACROS',
],
},
}],
['disable_nacl==1', {
'target_defaults': {
'defines': [
'DISABLE_NACL',
],
},
}],
['msvs_use_common_linker_extras', {
'target_defaults': {
'msvs_settings': {
'VCLinkerTool': {
'AdditionalOptions':
'/safeseh /dynamicbase /ignore:4199 /ignore:4221 /nxcompat',
'DelayLoadDLLs': [
'dbghelp.dll',
'dwmapi.dll',
'uxtheme.dll',
],
},
},
},
}],
],
'scons_settings': {
'sconsbuild_dir': '<(DEPTH)/sconsbuild',
'tools': ['ar', 'as', 'gcc', 'g++', 'gnulink', 'chromium_builders'],
},
'xcode_settings': {
# DON'T ADD ANYTHING NEW TO THIS BLOCK UNLESS YOU REALLY REALLY NEED IT!
# This block adds *project-wide* configuration settings to each project
# file. It's almost always wrong to put things here. Specify your
# custom xcode_settings in target_defaults to add them to targets instead.
# In an Xcode Project Info window, the "Base SDK for All Configurations"
# setting sets the SDK on a project-wide basis. In order to get the
# configured SDK to show properly in the Xcode UI, SDKROOT must be set
# here at the project level.
'SDKROOT': 'macosx<(mac_sdk)', # -isysroot
# The Xcode generator will look for an xcode_settings section at the root
# of each dict and use it to apply settings on a file-wide basis. Most
# settings should not be here, they should be in target-specific
# xcode_settings sections, or better yet, should use non-Xcode-specific
# settings in target dicts. SYMROOT is a special case, because many other
# Xcode variables depend on it, including variables such as
# PROJECT_DERIVED_FILE_DIR. When a source group corresponding to something
# like PROJECT_DERIVED_FILE_DIR is added to a project, in order for the
# files to appear (when present) in the UI as actual files and not red
# red "missing file" proxies, the correct path to PROJECT_DERIVED_FILE_DIR,
# and therefore SYMROOT, needs to be set at the project level.
'SYMROOT': '<(DEPTH)/xcodebuild',
},
}
# Local Variables:
# tab-width:2
# indent-tabs-mode:nil
# End:
# vim: set expandtab tabstop=2 shiftwidth=2: