kris/boringssl
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
966284337d
boringssl/CMakeLists.txt

278 lines
11 KiB
CMake
Raw Normal View History

Do a cursory conversion of a few tests to GTest. For now, this is the laziest conversion possible. The intent is to just get the build setup ready so that we can get everything working in our consumers. The intended end state is: - The standalone build produces three test targets, one per library: {crypto,ssl,decrepit}_tests. - Each FOO_test is made up of: FOO/**/*_test.cc crypto/test/gtest_main.cc test_support - generate_build_files.py emits variables crypto_test_sources and ssl_test_sources. These variables are populated with FindCFiles, looking for *_test.cc. - The consuming file assembles those variables into the two test targets (plus decrepit) from there. This avoids having generate_build_files.py emit actual build rules. - Our standalone builders, Chromium, and Android just run the top-level test targets using whatever GTest-based reporting story they have. In transition, we start by converting one of two tests in each library to populate the three test targets. Those are added to all_tests.json and all_tests.go hacked to handle them transparently. This keeps our standalone builder working. generate_build_files.py, to start with, populates the new source lists manually and subtracts them out of the old machinery. We emit both for the time being. When this change rolls in, we'll write all the build glue needed to build the GTest-based tests and add it to consumers' continuous builders. Next, we'll subsume a file-based test and get the consumers working with that. (I.e. make sure the GTest targets can depend on a data file.) Once that's all done, we'll be sure all this will work. At that point, we start subsuming the remaining tests into the GTest targets and, asynchronously, rewriting tests to use GTest properly rather than cursory conversion here. When all non-GTest tests are gone, the old generate_build_files.py hooks will be removed, consumers updated to not depend on them, and standalone builders converted to not rely on all_tests.go, which can then be removed. (Unless bits end up being needed as a malloc test driver. I'm thinking we'll want to do something with --gtest_filter.) As part of this CL, I've bumped the CMake requirements (for target_include_directories) and added a few suppressions for warnings that GTest doesn't pass. BUG=129 Change-Id: I881b26b07a8739cc0b52dbb51a30956908e1b71a Reviewed-on: https://boringssl-review.googlesource.com/13232 Reviewed-by: Adam Langley <agl@google.com>
2017-01-20 00:05:47 +00:00
cmake_minimum_required (VERSION 2.8.11)
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
Prefer MSVC over GCC if both are in %PATH%. Notably, putting Strawberry Perl in %PATH% will usually end up putting a copy of gcc in %PATH%, which trips up people trying to build on Windows. This is arguably misusing the variable (normally set by the generator), but it should work. Change-Id: I13a011eb33688ae928a56cce266edd2759a3cb32 Reviewed-on: https://boringssl-review.googlesource.com/7070 Reviewed-by: Adam Langley <agl@google.com>
2016-02-06 02:58:39 +00:00
# Defer enabling C and CXX languages.
project (BoringSSL NONE)
if(WIN32)
# On Windows, prefer cl over gcc if both are available. By default most of
# the CMake generators prefer gcc, even on Windows.
set(CMAKE_GENERATOR_CC cl)
endif()
enable_language(C)
enable_language(CXX)
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
Add support for building with the Android NDK. Previously I've been using the Linaro toolchains and just building static binaries. However, the Linaro toolchains have a broken pthread_rwlock_wrlock—it does nothing and then unlocking corrupts the lock. Building with the Android NDK avoids this. These build instructions depend on https://github.com/taka-no-me/android-cmake which people will need to clone into util/ if they want to use the Android NDK. Change-Id: Ic64919f9399af2a57e8df4fb4b3400865ddb2427 Reviewed-on: https://boringssl-review.googlesource.com/4600 Reviewed-by: Adam Langley <agl@google.com>
2015-04-29 01:46:58 +01:00
if(ANDROID)
# Android-NDK CMake files reconfigure the path and so Go and Perl won't be
# found. However, ninja will still find them in $PATH if we just name them.
Fix discovery rule for perl and go for Android We don't use find_package/find_program on android to find go/perl because the android toolchain reconfigure the $PATH. The pervious way of solving this was to let ninja look for go/perl on the $PATH but this approach prevented us from specifying explicit go/perl executables what is needed for hermetic build using prebuilts. This CL changes the Android specific discovery rule to only set GO_EXECUTABLE and PERL_EXECUTABLE if they are not specified on the command line or inside the toolchain file. Change-Id: Ib6ef69707749073f2b79244ebb301502b2a5a34a Reviewed-on: https://boringssl-review.googlesource.com/8000 Reviewed-by: David Benjamin <davidben@google.com>
2016-05-19 14:28:14 +01:00
if(NOT PERL_EXECUTABLE)
set(PERL_EXECUTABLE "perl")
endif()
if(NOT GO_EXECUTABLE)
set(GO_EXECUTABLE "go")
endif()
Add support for building with the Android NDK. Previously I've been using the Linaro toolchains and just building static binaries. However, the Linaro toolchains have a broken pthread_rwlock_wrlock—it does nothing and then unlocking corrupts the lock. Building with the Android NDK avoids this. These build instructions depend on https://github.com/taka-no-me/android-cmake which people will need to clone into util/ if they want to use the Android NDK. Change-Id: Ic64919f9399af2a57e8df4fb4b3400865ddb2427 Reviewed-on: https://boringssl-review.googlesource.com/4600 Reviewed-by: Adam Langley <agl@google.com>
2015-04-29 01:46:58 +01:00
else()
find_package(Perl REQUIRED)
find_program(GO_EXECUTABLE go)
endif()
Find perl using the CMake's standard FindPerl module. This gives a standard PERL_EXECUTABLE configuration knob which is useful for specifying a perl to use without having it in PATH. Change-Id: I4b196b77e0b4666081a3f291fee3654c47925844 Reviewed-on: https://boringssl-review.googlesource.com/3570 Reviewed-by: Adam Langley <agl@google.com>
2015-02-23 18:06:19 +00:00
Generate error data at build time. This avoids cluttering up the diff and making merge conflicts a pain. It does, however, mean we need to generate err_data.c ahead of time in Chromium and likely other downstream builds. It also adds a build dependency on Go. Change-Id: I6e0513ed9f50cfb030f7a523ea28519590977104 Reviewed-on: https://boringssl-review.googlesource.com/3790 Reviewed-by: Adam Langley <agl@google.com>
2015-03-05 06:19:27 +00:00
if (NOT GO_EXECUTABLE)
message(FATAL_ERROR "Could not find Go")
endif()
Improve compatibility with CMake 3.0 and later CMake 3.0 changed the identifier for Apple-supplied Clang to AppleClang. CMake 3.1 changed the behavior of variable expansion in quoted strings and complains with warning CMP0054 twice without these changes. BUG=crbug.com/451610 Change-Id: I0f1514ec302cf5f1b5cfc2c5a1c71c9e20d5e855 Reviewed-on: https://boringssl-review.googlesource.com/3011 Reviewed-by: Adam Langley <agl@google.com>
2015-01-24 01:25:44 +00:00
if(CMAKE_COMPILER_IS_GNUCXX OR CMAKE_CXX_COMPILER_ID MATCHES "Clang")
Add top-level BUILD file (in util/). When we have *-with-bazel branches this BUILD file will be copied to the top-level for consumers that want to use Bazel. From empirical testing, x86-64 on Linux is spelt “k8” and x86-64 on macOS is spelt “darwin”. I've not tried to enable assembly for any other cases yet. Change-Id: Ic6cb739565f145db20756fb57c0d087227fd9e18 Reviewed-on: https://boringssl-review.googlesource.com/8571 Reviewed-by: Adam Langley <agl@google.com>
2016-06-30 22:16:59 +01:00
set(C_CXX_FLAGS "-Wall -Werror -Wformat=2 -Wsign-compare -Wmissing-field-initializers -Wwrite-strings -ggdb -fvisibility=hidden -fno-common")
Be -Wnewline-eof clean. Change-Id: I85216184f9277ce0c0caae31e379b638683e28c5 Reviewed-on: https://boringssl-review.googlesource.com/10703 Commit-Queue: David Benjamin <davidben@google.com> Commit-Queue: Adam Langley <agl@google.com> Reviewed-by: Adam Langley <agl@google.com>
2016-08-29 19:59:31 +01:00
if(CMAKE_CXX_COMPILER_ID MATCHES "Clang")
set(C_CXX_FLAGS "${C_CXX_FLAGS} -Wnewline-eof")
endif()
Add top-level BUILD file (in util/). When we have *-with-bazel branches this BUILD file will be copied to the top-level for consumers that want to use Bazel. From empirical testing, x86-64 on Linux is spelt “k8” and x86-64 on macOS is spelt “darwin”. I've not tried to enable assembly for any other cases yet. Change-Id: Ic6cb739565f145db20756fb57c0d087227fd9e18 Reviewed-on: https://boringssl-review.googlesource.com/8571 Reviewed-by: Adam Langley <agl@google.com>
2016-06-30 22:16:59 +01:00
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${C_CXX_FLAGS} -Wmissing-prototypes -Wold-style-definition -Wstrict-prototypes")
Spell C++11 as C++11, not C++0x. We already require GCC 4.8+, so -std=c++11 should work fine. Change-Id: I07d46d7dcccb695b5df97a702f0d5007fdff3385 Reviewed-on: https://boringssl-review.googlesource.com/8245 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com>
2016-06-13 18:42:04 +01:00
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 ${C_CXX_FLAGS} -Wmissing-declarations")
Don't pass -Wa,-g on clang. clang's integrated assembler does not appear to support debug symbols yet. (And somehow CMake has another statefulness bug that causes this to only manifest on a second run, so the try bots aren't noticing.) Also don't clobber the existing CMAKE_ASM_FLAGS value. Change-Id: Ia04345f0dc2f9833ac13912d761cfab3479c10cd Reviewed-on: https://boringssl-review.googlesource.com/11083 CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org> Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com> Commit-Queue: David Benjamin <davidben@google.com>
2016-09-16 20:56:53 +01:00
# Clang's integerated assembler does not support debug symbols.
if(NOT CMAKE_CXX_COMPILER_ID MATCHES "Clang")
set(CMAKE_ASM_FLAGS "${CMAKE_ASM_FLAGS} -Wa,-g")
endif()
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
elseif(MSVC)
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
set(MSVC_DISABLED_WARNINGS_LIST
Do a cursory conversion of a few tests to GTest. For now, this is the laziest conversion possible. The intent is to just get the build setup ready so that we can get everything working in our consumers. The intended end state is: - The standalone build produces three test targets, one per library: {crypto,ssl,decrepit}_tests. - Each FOO_test is made up of: FOO/**/*_test.cc crypto/test/gtest_main.cc test_support - generate_build_files.py emits variables crypto_test_sources and ssl_test_sources. These variables are populated with FindCFiles, looking for *_test.cc. - The consuming file assembles those variables into the two test targets (plus decrepit) from there. This avoids having generate_build_files.py emit actual build rules. - Our standalone builders, Chromium, and Android just run the top-level test targets using whatever GTest-based reporting story they have. In transition, we start by converting one of two tests in each library to populate the three test targets. Those are added to all_tests.json and all_tests.go hacked to handle them transparently. This keeps our standalone builder working. generate_build_files.py, to start with, populates the new source lists manually and subtracts them out of the old machinery. We emit both for the time being. When this change rolls in, we'll write all the build glue needed to build the GTest-based tests and add it to consumers' continuous builders. Next, we'll subsume a file-based test and get the consumers working with that. (I.e. make sure the GTest targets can depend on a data file.) Once that's all done, we'll be sure all this will work. At that point, we start subsuming the remaining tests into the GTest targets and, asynchronously, rewriting tests to use GTest properly rather than cursory conversion here. When all non-GTest tests are gone, the old generate_build_files.py hooks will be removed, consumers updated to not depend on them, and standalone builders converted to not rely on all_tests.go, which can then be removed. (Unless bits end up being needed as a malloc test driver. I'm thinking we'll want to do something with --gtest_filter.) As part of this CL, I've bumped the CMake requirements (for target_include_directories) and added a few suppressions for warnings that GTest doesn't pass. BUG=129 Change-Id: I881b26b07a8739cc0b52dbb51a30956908e1b71a Reviewed-on: https://boringssl-review.googlesource.com/13232 Reviewed-by: Adam Langley <agl@google.com>
2017-01-20 00:05:47 +00:00
"C4061" # enumerator 'identifier' in switch of enum 'enumeration' is not
# explicitly handled by a case label
# Disable this because it flags even when there is a default.
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
"C4100" # 'exarg' : unreferenced formal parameter
"C4127" # conditional expression is constant
"C4200" # nonstandard extension used : zero-sized array in
# struct/union.
Add a helper function for parsing extensions blocks. TLS 1.3 adds a number of places with extensions blocks that don't easily fit into our ClientHello/EncryptedExtensions callbacks. Between HelloRetryRequest, ServerHello, draft 18 going nuts with Certificate, and NewSessionTicket when we do 0-RTT, this passes the "abstract things that are repeated three times" sniff test. For now, it rejects unknown extensions, but it will probably grow an allow_unknown parameter for NewSessionTicket. This involves disabling some MSVC warnings, but they're invalid as of C99 which we otherwise require. See https://connect.microsoft.com/VisualStudio/feedback/details/1230248/remove-c99-related-warnings-or-make-them-off-by-default Change-Id: Iea8bf8ab216270c081dd63e79aaad9ec73b3b550 Reviewed-on: https://boringssl-review.googlesource.com/12233 Reviewed-by: David Benjamin <davidben@google.com> Commit-Queue: David Benjamin <davidben@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-11-13 01:32:10 +00:00
"C4204" # nonstandard extension used: non-constant aggregate initializer
"C4221" # nonstandard extension used : 'identifier' : cannot be
# initialized using address of automatic variable
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
"C4242" # 'function' : conversion from 'int' to 'uint8_t',
# possible loss of data
"C4244" # 'function' : conversion from 'int' to 'uint8_t',
# possible loss of data
"C4245" # 'initializing' : conversion from 'long' to
# 'unsigned long', signed/unsigned mismatch
Fix standalone build on Win64. Win64 fires significantly more warnings than Win32. Also some recent changes made it grumpy. (We might want to reconsider enabling all of MSVC's warnings. Given the sorts of warnings some of these are, I'm not sure MSVC's version of -Wall -Werror is actually tenable. Plus, diverging from the Chromium build, especially before the bots are ready, is going to break pretty readily.) Change-Id: If3b8feccf910ceab4a233b0731e7624d7da46f87 Reviewed-on: https://boringssl-review.googlesource.com/3420 Reviewed-by: Adam Langley <agl@google.com>
2015-02-11 20:12:05 +00:00
"C4267" # conversion from 'size_t' to 'int', possible loss of data
"C4371" # layout of class may have changed from a previous version of the
# compiler due to better packing of member '...'
"C4388" # signed/unsigned mismatch
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
"C4296" # '>=' : expression is always true
"C4350" # behavior change: 'std::_Wrap_alloc...'
"C4365" # '=' : conversion from 'size_t' to 'int',
# signed/unsigned mismatch
"C4389" # '!=' : signed/unsigned mismatch
Start assuming MSVC 2015. BUG=43 Change-Id: I46ad1ca62b8921a03fae51f5d7bbe1c68fc0b170 Reviewed-on: https://boringssl-review.googlesource.com/7821 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com>
2016-05-02 17:57:01 +01:00
"C4464" # relative include path contains '..'
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
"C4510" # 'argument' : default constructor could not be generated
"C4512" # 'argument' : assignment operator could not be generated
"C4514" # 'function': unreferenced inline function has been removed
"C4548" # expression before comma has no effect; expected expression with
# side-effect" caused by FD_* macros.
"C4610" # struct 'argument' can never be instantiated - user defined
# constructor required.
Start assuming MSVC 2015. BUG=43 Change-Id: I46ad1ca62b8921a03fae51f5d7bbe1c68fc0b170 Reviewed-on: https://boringssl-review.googlesource.com/7821 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com>
2016-05-02 17:57:01 +01:00
"C4623" # default constructor was implicitly defined as deleted
Fix standalone build on Win64. Win64 fires significantly more warnings than Win32. Also some recent changes made it grumpy. (We might want to reconsider enabling all of MSVC's warnings. Given the sorts of warnings some of these are, I'm not sure MSVC's version of -Wall -Werror is actually tenable. Plus, diverging from the Chromium build, especially before the bots are ready, is going to break pretty readily.) Change-Id: If3b8feccf910ceab4a233b0731e7624d7da46f87 Reviewed-on: https://boringssl-review.googlesource.com/3420 Reviewed-by: Adam Langley <agl@google.com>
2015-02-11 20:12:05 +00:00
"C4625" # copy constructor could not be generated because a base class
# copy constructor is inaccessible or deleted
"C4626" # assignment operator could not be generated because a base class
# assignment operator is inaccessible or deleted
Do a cursory conversion of a few tests to GTest. For now, this is the laziest conversion possible. The intent is to just get the build setup ready so that we can get everything working in our consumers. The intended end state is: - The standalone build produces three test targets, one per library: {crypto,ssl,decrepit}_tests. - Each FOO_test is made up of: FOO/**/*_test.cc crypto/test/gtest_main.cc test_support - generate_build_files.py emits variables crypto_test_sources and ssl_test_sources. These variables are populated with FindCFiles, looking for *_test.cc. - The consuming file assembles those variables into the two test targets (plus decrepit) from there. This avoids having generate_build_files.py emit actual build rules. - Our standalone builders, Chromium, and Android just run the top-level test targets using whatever GTest-based reporting story they have. In transition, we start by converting one of two tests in each library to populate the three test targets. Those are added to all_tests.json and all_tests.go hacked to handle them transparently. This keeps our standalone builder working. generate_build_files.py, to start with, populates the new source lists manually and subtracts them out of the old machinery. We emit both for the time being. When this change rolls in, we'll write all the build glue needed to build the GTest-based tests and add it to consumers' continuous builders. Next, we'll subsume a file-based test and get the consumers working with that. (I.e. make sure the GTest targets can depend on a data file.) Once that's all done, we'll be sure all this will work. At that point, we start subsuming the remaining tests into the GTest targets and, asynchronously, rewriting tests to use GTest properly rather than cursory conversion here. When all non-GTest tests are gone, the old generate_build_files.py hooks will be removed, consumers updated to not depend on them, and standalone builders converted to not rely on all_tests.go, which can then be removed. (Unless bits end up being needed as a malloc test driver. I'm thinking we'll want to do something with --gtest_filter.) As part of this CL, I've bumped the CMake requirements (for target_include_directories) and added a few suppressions for warnings that GTest doesn't pass. BUG=129 Change-Id: I881b26b07a8739cc0b52dbb51a30956908e1b71a Reviewed-on: https://boringssl-review.googlesource.com/13232 Reviewed-by: Adam Langley <agl@google.com>
2017-01-20 00:05:47 +00:00
"C4668" # 'symbol' is not defined as a preprocessor macro, replacing with
# '0' for 'directives'
# Disable this because GTest uses it everywhere.
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
"C4706" # assignment within conditional expression
"C4710" # 'function': function not inlined
"C4711" # function 'function' selected for inline expansion
"C4800" # 'int' : forcing value to bool 'true' or 'false'
# (performance warning)
"C4820" # 'bytes' bytes padding added after construct 'member_name'
Do a cursory conversion of a few tests to GTest. For now, this is the laziest conversion possible. The intent is to just get the build setup ready so that we can get everything working in our consumers. The intended end state is: - The standalone build produces three test targets, one per library: {crypto,ssl,decrepit}_tests. - Each FOO_test is made up of: FOO/**/*_test.cc crypto/test/gtest_main.cc test_support - generate_build_files.py emits variables crypto_test_sources and ssl_test_sources. These variables are populated with FindCFiles, looking for *_test.cc. - The consuming file assembles those variables into the two test targets (plus decrepit) from there. This avoids having generate_build_files.py emit actual build rules. - Our standalone builders, Chromium, and Android just run the top-level test targets using whatever GTest-based reporting story they have. In transition, we start by converting one of two tests in each library to populate the three test targets. Those are added to all_tests.json and all_tests.go hacked to handle them transparently. This keeps our standalone builder working. generate_build_files.py, to start with, populates the new source lists manually and subtracts them out of the old machinery. We emit both for the time being. When this change rolls in, we'll write all the build glue needed to build the GTest-based tests and add it to consumers' continuous builders. Next, we'll subsume a file-based test and get the consumers working with that. (I.e. make sure the GTest targets can depend on a data file.) Once that's all done, we'll be sure all this will work. At that point, we start subsuming the remaining tests into the GTest targets and, asynchronously, rewriting tests to use GTest properly rather than cursory conversion here. When all non-GTest tests are gone, the old generate_build_files.py hooks will be removed, consumers updated to not depend on them, and standalone builders converted to not rely on all_tests.go, which can then be removed. (Unless bits end up being needed as a malloc test driver. I'm thinking we'll want to do something with --gtest_filter.) As part of this CL, I've bumped the CMake requirements (for target_include_directories) and added a few suppressions for warnings that GTest doesn't pass. BUG=129 Change-Id: I881b26b07a8739cc0b52dbb51a30956908e1b71a Reviewed-on: https://boringssl-review.googlesource.com/13232 Reviewed-by: Adam Langley <agl@google.com>
2017-01-20 00:05:47 +00:00
"C5026" # move constructor was implicitly defined as deleted
Start assuming MSVC 2015. BUG=43 Change-Id: I46ad1ca62b8921a03fae51f5d7bbe1c68fc0b170 Reviewed-on: https://boringssl-review.googlesource.com/7821 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com>
2016-05-02 17:57:01 +01:00
"C5027" # move assignment operator was implicitly defined as deleted
)
set(MSVC_LEVEL4_WARNINGS_LIST
# See https://connect.microsoft.com/VisualStudio/feedback/details/1217660/warning-c4265-when-using-functional-header
"C4265" # class has virtual functions, but destructor is not virtual
)
Enable more warnings & treat warnings as errors on Windows. Change-Id: I2bf0144aaa8b670ff00b8e8dfe36bd4d237b9a8a Reviewed-on: https://boringssl-review.googlesource.com/3140 Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 00:20:02 +00:00
string(REPLACE "C" " -wd" MSVC_DISABLED_WARNINGS_STR
${MSVC_DISABLED_WARNINGS_LIST})
Start assuming MSVC 2015. BUG=43 Change-Id: I46ad1ca62b8921a03fae51f5d7bbe1c68fc0b170 Reviewed-on: https://boringssl-review.googlesource.com/7821 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com>
2016-05-02 17:57:01 +01:00
string(REPLACE "C" " -w4" MSVC_LEVEL4_WARNINGS_STR
${MSVC_LEVEL4_WARNINGS_LIST})
Fix build when using Visual Studio 2015 Update 1. Many of the compatibility issues are described at https://msdn.microsoft.com/en-us/library/mt612856.aspx. The macros that suppressed warnings on a per-function basis no longer work in Update 1, so replace them with #pragmas. Update 1 warns when |size_t| arguments to |printf| are casted, so stop doing that casting. Unfortunately, this requires an ugly hack to continue working in MSVC 2013 as MSVC 2013 doesn't support "%zu". Finally, Update 1 has new warnings, some of which need to be suppressed. --- Updated by davidben to give up on suppressing warnings in crypto/x509 and crypto/x509v3 as those directories aren't changed much from upstream. In each of these cases, upstream opted just blindly initialize the variable, so do the same. Also switch C4265 to level 4, per Microsoft's recommendation and work around a bug in limits.h that happens to get fixed by Google include order style. (limits.h is sensitive to whether corecrt.h, pulled in by stddef.h and some other headers, is included before it. The reason it affected just one file is we often put the file's header first, which means base.h is pulling in stddef.h. Relying on this is ugly, but it's no worse than what everything else is doing and this doesn't seem worth making something as tame as limits.h so messy to use.) Change-Id: I02d1f935356899f424d3525d03eca401bfa3e6cd Reviewed-on: https://boringssl-review.googlesource.com/7480 Reviewed-by: David Benjamin <davidben@google.com>
2016-01-18 08:21:42 +00:00
set(CMAKE_C_FLAGS "-Wall -WX ${MSVC_DISABLED_WARNINGS_STR} ${MSVC_LEVEL4_WARNINGS_STR}")
set(CMAKE_CXX_FLAGS "-Wall -WX ${MSVC_DISABLED_WARNINGS_STR} ${MSVC_LEVEL4_WARNINGS_STR}")
Build debug symbols for assembly. We build with them for C files, so include it for the assembly files too. Change-Id: I1c595e18314b245330a6f047d0a97a031d55b115 Reviewed-on: https://boringssl-review.googlesource.com/9150 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com> Commit-Queue: David Benjamin <davidben@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-08-05 00:09:45 +01:00
set(CMAKE_ASM_NASM_FLAGS "-g cv8")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
add_definitions(-D_HAS_EXCEPTIONS=0)
Define WIN32_LEAN_AND_MEAN in the build system, not in the code. This avoids a conflict with the Chromium build system, which defines WIN32_LEAN_AND_MEAN with a different value. BUG=crbug.com/453196 Change-Id: Ia15ec7c20325c1961af4f32e5208266e5f846f35 Reviewed-on: https://boringssl-review.googlesource.com/3150 Reviewed-by: David Benjamin <davidben@chromium.org> Reviewed-by: Adam Langley <agl@google.com>
2015-01-29 04:34:47 +00:00
add_definitions(-DWIN32_LEAN_AND_MEAN)
Define NOMINMAX across the entire project. This fixes the standalone build on Windows and matches Chromium. Change-Id: I194f53e0a610c5ae9cef53c826b22f7bded5f357 Reviewed-on: https://boringssl-review.googlesource.com/4201 Reviewed-by: Adam Langley <agl@google.com>
2015-04-02 18:20:01 +01:00
add_definitions(-DNOMINMAX)
Use non-deprecated methods on windows. Use of strdup, close, lseek, read, and write prevent linking statically againt libcmt.lib. Change-Id: I04f7876ec0f03f29f000bbcc6b2ccdec844452d2 Reviewed-on: https://boringssl-review.googlesource.com/8010 Reviewed-by: David Benjamin <davidben@google.com>
2016-05-19 16:49:59 +01:00
add_definitions(-D_CRT_SECURE_NO_WARNINGS) # Allow use of fopen
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
endif()
Bump the -Wshadow cutoff to 4.8 rather than 4.6. Per https://gcc.gnu.org/gcc-4.8/changes.html, -Wshadow was made less trigger-happy starting 4.8. Change-Id: I023919dea3497e2dd27679a6f0c4a114c0b30337 Reviewed-on: https://boringssl-review.googlesource.com/3730 Reviewed-by: Adam Langley <agl@google.com>
2015-02-28 05:00:44 +00:00
if((CMAKE_COMPILER_IS_GNUCXX AND CMAKE_C_COMPILER_VERSION VERSION_GREATER "4.7.99") OR
Improve compatibility with CMake 3.0 and later CMake 3.0 changed the identifier for Apple-supplied Clang to AppleClang. CMake 3.1 changed the behavior of variable expansion in quoted strings and complains with warning CMP0054 twice without these changes. BUG=crbug.com/451610 Change-Id: I0f1514ec302cf5f1b5cfc2c5a1c71c9e20d5e855 Reviewed-on: https://boringssl-review.googlesource.com/3011 Reviewed-by: Adam Langley <agl@google.com>
2015-01-24 01:25:44 +00:00
CMAKE_CXX_COMPILER_ID MATCHES "Clang")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wshadow")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wshadow")
Only use -Wshadow with GCC > 4.6.0. Prior to 4.6.0, -Wshadow would cause GCC to warn when variables shadowed global functions. Since libc defines a number of functions with common names, this is a problem. Also, without this change, we'll keep breaking on older versions of GCC because we won't be testing with them. OpenBSD, specifically is reported to have a problem: https://boringssl-review.googlesource.com/#/c/2900/ (Note the test should really be >= 4.6.0, but CMake doesn't have a VERSION_GREATEROREQUAL.) Change-Id: I1aedda01ab629e138c8781e4319bfaaed0b236b0 Reviewed-on: https://boringssl-review.googlesource.com/2952 Reviewed-by: Adam Langley <agl@google.com>
2015-01-16 23:20:54 +00:00
endif()
Use one C99-style for loop. Switch one for loop to the new spelling as a canary. All our compilers seem to support it fine, except GCC needs to be told to build with -std=c99. (And, upon doing so, it'll require _XOPEN_SOURCE=700 for pthread_rwlock_t.) We'll let this sit for a bit until it's gotten into downstreams without issue and then open the floodgates. BUG=47 Change-Id: I1c69d4b2df8206e0b55f30aa59b5874d82fca893 Reviewed-on: https://boringssl-review.googlesource.com/8235 Reviewed-by: Adam Langley <agl@google.com>
2016-06-09 21:22:26 +01:00
if(CMAKE_COMPILER_IS_GNUCXX)
if ((CMAKE_C_COMPILER_VERSION VERSION_GREATER "4.8.99") OR
CMAKE_CXX_COMPILER_ID MATCHES "Clang")
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=c11")
else()
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=c99")
endif()
endif()
# pthread_rwlock_t requires a feature flag.
if(NOT WIN32)
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -D_XOPEN_SOURCE=700")
Add infrastructure for reference counts. OpenSSL has traditionally done reference counting with |int|s and the |CRYPTO_add| function. Unless a special callback is installed (rare), this is implemented by doing the reference count operations under a lock. This change adds infrastructure for handling reference counts and uses atomic operations when C11 support is available. Change-Id: Ia023ce432319efd00f77a7340da27d16ee4b63c3 Reviewed-on: https://boringssl-review.googlesource.com/4771 Reviewed-by: Adam Langley <agl@google.com>
2015-05-15 20:01:29 +01:00
endif()
Add four, basic fuzz tests. This change adds fuzzing tests for: ∙ Certificate parsing ∙ Private key parsing ∙ ClientHello parsing ∙ Server first flow (ServerHello, Certificate, etc) parsing. Change-Id: I5f53282263eaaff69b1a03c819cca73750433653 Reviewed-on: https://boringssl-review.googlesource.com/6460 Reviewed-by: Adam Langley <agl@google.com>
2015-11-09 21:57:26 +00:00
if(FUZZ)
Fix CMake negation and make error fatal Not sure if version-dependant, but with CMake 3.6 "!" doesn't work. Change-Id: I5f91234b27e340142b479b602d4102134ad55ccb Reviewed-on: https://boringssl-review.googlesource.com/11381 Reviewed-by: Adam Langley <agl@google.com>
2016-09-28 22:14:01 +01:00
if(NOT CMAKE_CXX_COMPILER_ID MATCHES "Clang")
message(FATAL_ERROR "You need to build with Clang for fuzzing to work")
Add four, basic fuzz tests. This change adds fuzzing tests for: ∙ Certificate parsing ∙ Private key parsing ∙ ClientHello parsing ∙ Server first flow (ServerHello, Certificate, etc) parsing. Change-Id: I5f53282263eaaff69b1a03c819cca73750433653 Reviewed-on: https://boringssl-review.googlesource.com/6460 Reviewed-by: Adam Langley <agl@google.com>
2015-11-09 21:57:26 +00:00
endif()
Add corpora for fuzzers with fuzzer mode disabled. Fuzzer mode explores the handshake, but at the cost of losing coverage on the record layer. Add a separate build flag and client/server corpora for this mode. Note this requires tweaks in consumers' fuzzer build definitions. BUG=111 Change-Id: I1026dc7301645e165a761068a1daad6eedc9271e Reviewed-on: https://boringssl-review.googlesource.com/12108 Commit-Queue: David Benjamin <davidben@google.com> Reviewed-by: Adam Langley <agl@google.com>
2016-11-04 22:59:33 +00:00
add_definitions(-DBORINGSSL_UNSAFE_DETERMINISTIC_MODE)
set(RUNNER_ARGS "-deterministic")
if(NOT NO_FUZZER_MODE)
add_definitions(-DBORINGSSL_UNSAFE_FUZZER_MODE)
set(RUNNER_ARGS ${RUNNER_ARGS} "-fuzzer" "-shim-config" "fuzzer_mode.json")
endif()
Add a deterministic PRNG for fuzzing. If running the stack through a fuzzer, we would like execution to be completely deterministic. This is gated on a BORINGSSL_UNSAFE_FUZZER_MODE #ifdef. For now, this just uses the zero ChaCha20 key and a global counter. As needed, we can extend this to a thread-local counter and a separate ChaCha20 stream and counter per input length. Change-Id: Ic6c9d8a25e70d68e5dc6804e2c234faf48e51395 Reviewed-on: https://boringssl-review.googlesource.com/7286 Reviewed-by: Adam Langley <agl@google.com>
2016-03-02 03:57:32 +00:00
Add `8bit-counters` option for fuzzing. This enables coverage counters[1] when fuzzing. [1] http://clang.llvm.org/docs/SanitizerCoverage.html#coverage-counters Change-Id: I33fca02d0406b75ac1f7598f41fe4c2ce43538d1 Reviewed-on: https://boringssl-review.googlesource.com/7274 Reviewed-by: David Benjamin <davidben@google.com>
2016-03-03 17:50:10 +00:00
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fsanitize=address -fsanitize-coverage=edge,indirect-calls,8bit-counters")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsanitize=address -fsanitize-coverage=edge,indirect-calls,8bit-counters")
Add four, basic fuzz tests. This change adds fuzzing tests for: ∙ Certificate parsing ∙ Private key parsing ∙ ClientHello parsing ∙ Server first flow (ServerHello, Certificate, etc) parsing. Change-Id: I5f53282263eaaff69b1a03c819cca73750433653 Reviewed-on: https://boringssl-review.googlesource.com/6460 Reviewed-by: Adam Langley <agl@google.com>
2015-11-09 21:57:26 +00:00
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_C_FLAGS} -fsanitize=address")
Update and fix fuzzing instructions. It's easier to put libFuzzer.a into the source directory than to install it globally. Change-Id: I4dc7b56f81c7aa0371475c68d23368b025186505 Reviewed-on: https://boringssl-review.googlesource.com/6461 Reviewed-by: Adam Langley <agl@google.com>
2015-11-10 23:00:51 +00:00
link_directories(.)
Add four, basic fuzz tests. This change adds fuzzing tests for: ∙ Certificate parsing ∙ Private key parsing ∙ ClientHello parsing ∙ Server first flow (ServerHello, Certificate, etc) parsing. Change-Id: I5f53282263eaaff69b1a03c819cca73750433653 Reviewed-on: https://boringssl-review.googlesource.com/6460 Reviewed-by: Adam Langley <agl@google.com>
2015-11-09 21:57:26 +00:00
endif()
Add visibility rules. This change marks public symbols as dynamically exported. This means that it becomes viable to build a shared library of libcrypto and libssl with -fvisibility=hidden. On Windows, one not only needs to mark functions for export in a component, but also for import when using them from a different component. Because of this we have to build with |BORINGSSL_IMPLEMENTATION| defined when building the code. Other components, when including our headers, won't have that defined and then the |OPENSSL_EXPORT| tag becomes an import tag instead. See the #defines in base.h In the asm code, symbols are now hidden by default and those that need to be exported are wrapped by a C function. In order to support Chromium, a couple of libssl functions were moved to ssl.h from ssl_locl.h: ssl_get_new_session and ssl_update_cache. Change-Id: Ib4b76e2f1983ee066e7806c24721e8626d08a261 Reviewed-on: https://boringssl-review.googlesource.com/1350 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 00:02:14 +01:00
add_definitions(-DBORINGSSL_IMPLEMENTATION)
Honor the standard BUILD_SHARED_LIBS cmake variable. The variable switches the default type for add_library from STATIC to SHARED. We can condition additional stuff on that for convenience. (tabtest still doesn't build.) BoringSSL as any kind of stable system shared library is still very much unsupported, but this is probably handy for making sure we don't forget all those pesky OPENSSL_EXPORTs. Change-Id: I66ab80bcddbf3724e03e85384141fdf4f4acbc2e Reviewed-on: https://boringssl-review.googlesource.com/3092 Reviewed-by: Adam Langley <agl@google.com>
2015-01-28 05:50:21 +00:00
if (BUILD_SHARED_LIBS)
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
add_definitions(-DBORINGSSL_SHARED_LIBRARY)
# Enable position-independent code globally. This is needed because
# some library targets are OBJECT libraries.
set(CMAKE_POSITION_INDEPENDENT_CODE TRUE)
Honor the standard BUILD_SHARED_LIBS cmake variable. The variable switches the default type for add_library from STATIC to SHARED. We can condition additional stuff on that for convenience. (tabtest still doesn't build.) BoringSSL as any kind of stable system shared library is still very much unsupported, but this is probably handy for making sure we don't forget all those pesky OPENSSL_EXPORTs. Change-Id: I66ab80bcddbf3724e03e85384141fdf4f4acbc2e Reviewed-on: https://boringssl-review.googlesource.com/3092 Reviewed-by: Adam Langley <agl@google.com>
2015-01-28 05:50:21 +00:00
endif()
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
if (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "x86_64")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(ARCH "x86_64")
Fix CPU architecture detection on BSDs. CMake calls "uname" in order to detect the CPU architecture, so $(CMAKE_SYSTEM_PROCESSOR) varies from platform to platform. This changes adds support for "i386" and "amd64" values, which are used by BSDs for the x86 family of CPUs. Change-Id: I532ce787a9ac06220c92a6d8c78ad5a55d8c40bf Signed-off-by: Piotr Sikora <piotr@cloudflare.com> Reviewed-on: https://boringssl-review.googlesource.com/1360 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 11:09:49 +01:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "amd64")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(ARCH "x86_64")
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "AMD64")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
# cmake reports AMD64 on Windows, but we might be building for 32-bit.
if (CMAKE_CL_64)
set(ARCH "x86_64")
else()
set(ARCH "x86")
endif()
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "x86")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(ARCH "x86")
Fix CPU architecture detection on BSDs. CMake calls "uname" in order to detect the CPU architecture, so $(CMAKE_SYSTEM_PROCESSOR) varies from platform to platform. This changes adds support for "i386" and "amd64" values, which are used by BSDs for the x86 family of CPUs. Change-Id: I532ce787a9ac06220c92a6d8c78ad5a55d8c40bf Signed-off-by: Piotr Sikora <piotr@cloudflare.com> Reviewed-on: https://boringssl-review.googlesource.com/1360 Reviewed-by: Adam Langley <agl@google.com>
2014-07-31 11:09:49 +01:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "i386")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(ARCH "x86")
Recognise "i686" as an x86 CPU. Building under Ubuntu x86 fails with "Unknown processor:i686". This adds i686 to CMakeLists.txt, fixing the build. Change-Id: Ic1c7ce452c019b8a2a875d64a707f640d86c7e31
2014-08-11 21:27:44 +01:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "i686")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(ARCH "x86")
More flexible detection of arm processors to fix cmake errors on armv6l and armv7l devices Change-Id: Iee7653f620c56136df75908830f06a1d5c597609 Reviewed-on: https://boringssl-review.googlesource.com/12240 Reviewed-by: Adam Langley <agl@google.com> Commit-Queue: Adam Langley <agl@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-11-13 00:04:20 +00:00
elseif (${CMAKE_SYSTEM_PROCESSOR} MATCHES "^arm*")
Add support for building with the Android NDK. Previously I've been using the Linaro toolchains and just building static binaries. However, the Linaro toolchains have a broken pthread_rwlock_wrlock—it does nothing and then unlocking corrupts the lock. Building with the Android NDK avoids this. These build instructions depend on https://github.com/taka-no-me/android-cmake which people will need to clone into util/ if they want to use the Android NDK. Change-Id: Ic64919f9399af2a57e8df4fb4b3400865ddb2427 Reviewed-on: https://boringssl-review.googlesource.com/4600 Reviewed-by: Adam Langley <agl@google.com>
2015-04-29 01:46:58 +01:00
set(ARCH "arm")
aarch64 support. This is an initial cut at aarch64 support. I have only qemu to test it however—hopefully hardware will be coming soon. This also affects 32-bit ARM in that aarch64 chips can run 32-bit code and we would like to be able to take advantage of the crypto operations even in 32-bit mode. AES and GHASH should Just Work in this case: the -armx.pl files can be built for either 32- or 64-bit mode based on the flavour argument given to the Perl script. SHA-1 and SHA-256 don't work like this however because they've never support for multiple implementations, thus BoringSSL built for 32-bit won't use the SHA instructions on an aarch64 chip. No dedicated ChaCha20 or Poly1305 support yet. Change-Id: Ib275bc4894a365c8ec7c42f4e91af6dba3bd686c Reviewed-on: https://boringssl-review.googlesource.com/2801 Reviewed-by: Adam Langley <agl@google.com>
2015-01-09 23:44:37 +00:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "aarch64")
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
set(ARCH "aarch64")
Allow building on MIPS. Change-Id: I596e77fa5a7b97b405f52c44e9e36a8226196997 Reviewed-on: https://boringssl-review.googlesource.com/11900 Reviewed-by: Adam Langley <agl@google.com>
2016-10-28 14:05:32 +01:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "mips")
# Just to avoid the “unknown processor” error.
Add ppc64le vector implementation of SHA-1. This change contains a C implementation of SHA-1 for POWER using AltiVec. It is almost as fast as the scalar-only assembly implementation for POWER/POWERPC family in OpenSSL but it is easier to maintain and it allows error checking with tools like ASAN. This is tested only for ppc64le. It may nor may not work for other platforms in the POWER/POWERPC familiy. Before: SHA-1 @ 16 bytes: ~30 MB/s SHA-1 @ 8K: ~140 MB/s After: SHA-1 @ 16 bytes: ~70 MB/s SHA-1 @ 8K: ~480 MB/s Change-Id: I790352e86d9c0cc4e1e57d11c5a0aa5b0780ca6b Reviewed-on: https://boringssl-review.googlesource.com/12203 Reviewed-by: Adam Langley <agl@google.com> Commit-Queue: Adam Langley <agl@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-11-12 01:18:53 +00:00
elseif (${CMAKE_SYSTEM_PROCESSOR} STREQUAL "ppc64le")
set(ARCH "ppc64le")
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
else()
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
message(FATAL_ERROR "Unknown processor:" ${CMAKE_SYSTEM_PROCESSOR})
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
endif()
Add support for building with the Android NDK. Previously I've been using the Linaro toolchains and just building static binaries. However, the Linaro toolchains have a broken pthread_rwlock_wrlock—it does nothing and then unlocking corrupts the lock. Building with the Android NDK avoids this. These build instructions depend on https://github.com/taka-no-me/android-cmake which people will need to clone into util/ if they want to use the Android NDK. Change-Id: Ic64919f9399af2a57e8df4fb4b3400865ddb2427 Reviewed-on: https://boringssl-review.googlesource.com/4600 Reviewed-by: Adam Langley <agl@google.com>
2015-04-29 01:46:58 +01:00
if (ANDROID AND ${ARCH} STREQUAL "arm")
# The Android-NDK CMake files somehow fail to set the -march flag for
# assembly files. Without this flag, the compiler believes that it's
# building for ARMv5.
Allow compilation for armv6 Use when giving -DANDROID_ABI="armeabi-v6 with VFP" to android-cmake Change-Id: Ifb053bcf4788e34fb54e20c27d4e519416ca9a11 Reviewed-on: https://boringssl-review.googlesource.com/4945 Reviewed-by: Adam Langley <agl@google.com>
2015-05-29 14:06:45 +01:00
set(CMAKE_ASM_FLAGS "${CMAKE_ASM_FLAGS} -march=${CMAKE_SYSTEM_PROCESSOR}")
Add support for building with the Android NDK. Previously I've been using the Linaro toolchains and just building static binaries. However, the Linaro toolchains have a broken pthread_rwlock_wrlock—it does nothing and then unlocking corrupts the lock. Building with the Android NDK avoids this. These build instructions depend on https://github.com/taka-no-me/android-cmake which people will need to clone into util/ if they want to use the Android NDK. Change-Id: Ic64919f9399af2a57e8df4fb4b3400865ddb2427 Reviewed-on: https://boringssl-review.googlesource.com/4600 Reviewed-by: Adam Langley <agl@google.com>
2015-04-29 01:46:58 +01:00
endif()
Make build work on OS X with older cmake versions. `uname -p` is still i386 on OS X for some reason, which causes cmake 2.8 to set CMAKE_SYSTEM_PROCESSOR to i386, making the build think it's doing a 32-bit build. However, since the system is almost completely 64-bit these days, clang defaults to producing 64-bit object files unless told otherwise. As a result, the produced .o files are all 64-bit except for the .o files from assembly, and then linking fails. Fix this by forcing ARCH to 64-bit on OS X. This matches the default behavior of cmake 3.0, where CMAKE_SYSTEM_PROCESSOR is x86_64. Change-Id: I7a2abc4cef84dfbaf205852a9d7b647e83dd249f Reviewed-on: https://boringssl-review.googlesource.com/2330 Reviewed-by: Adam Langley <agl@google.com> Reviewed-by: Piotr Sikora <piotr@cloudflare.com>
2014-11-18 20:14:46 +00:00
if (${ARCH} STREQUAL "x86" AND APPLE)
Change CMakeLists.txt to two-space indent. find -name CMakeLists.txt -type f | xargs sed -e 's/\t/ /g' -i Change-Id: I01636b1849c00ba918f48828252492d99b0403ac
2015-01-29 00:37:10 +00:00
# With CMake 2.8.x, ${CMAKE_SYSTEM_PROCESSOR} evalutes to i386 on OS X,
# but clang defaults to 64-bit builds on OS X unless otherwise told.
# Set ARCH to x86_64 so clang and CMake agree. This is fixed in CMake 3.
set(ARCH "x86_64")
Make build work on OS X with older cmake versions. `uname -p` is still i386 on OS X for some reason, which causes cmake 2.8 to set CMAKE_SYSTEM_PROCESSOR to i386, making the build think it's doing a 32-bit build. However, since the system is almost completely 64-bit these days, clang defaults to producing 64-bit object files unless told otherwise. As a result, the produced .o files are all 64-bit except for the .o files from assembly, and then linking fails. Fix this by forcing ARCH to 64-bit on OS X. This matches the default behavior of cmake 3.0, where CMAKE_SYSTEM_PROCESSOR is x86_64. Change-Id: I7a2abc4cef84dfbaf205852a9d7b647e83dd249f Reviewed-on: https://boringssl-review.googlesource.com/2330 Reviewed-by: Adam Langley <agl@google.com> Reviewed-by: Piotr Sikora <piotr@cloudflare.com>
2014-11-18 20:14:46 +00:00
endif()
Add CMake support for ASAN and MSAN. This only works with Clang, and MSAN seems to have a false-positive for me in libstdc++, but it can be helpful to test with these Change-Id: I068edabcda69c9239ee4f0247f5d8f873dea77bb Reviewed-on: https://boringssl-review.googlesource.com/12940 Reviewed-by: David Benjamin <davidben@google.com> Commit-Queue: David Benjamin <davidben@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-12-16 18:48:23 +00:00
if (MSAN)
if(NOT CMAKE_CXX_COMPILER_ID MATCHES "Clang")
message(FATAL_ERROR "Cannot enable MSAN unless using Clang")
endif()
if (ASAN)
message(FATAL_ERROR "ASAN and MSAN are mutually exclusive")
endif()
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fsanitize=memory -fsanitize-memory-track-origins -fno-omit-frame-pointer")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsanitize=memory -fsanitize-memory-track-origins -fno-omit-frame-pointer")
set(OPENSSL_NO_ASM "1")
endif()
if (ASAN)
if(NOT CMAKE_CXX_COMPILER_ID MATCHES "Clang")
message(FATAL_ERROR "Cannot enable ASAN unless using Clang")
endif()
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fsanitize=address -fno-omit-frame-pointer")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsanitize=address -fno-omit-frame-pointer")
set(OPENSSL_NO_ASM "1")
endif()
Add a GCOV option to CMakeLists.txt. Get us a little closer to productionizing the coverage generation, which will require taking all the logic out of the coverage script. Change-Id: If410cc198a888ee87a84b1c2d532322682d3c44e Reviewed-on: https://boringssl-review.googlesource.com/13043 Reviewed-by: Steven Valdez <svaldez@google.com> Reviewed-by: David Benjamin <davidben@google.com> Commit-Queue: David Benjamin <davidben@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-12-27 17:15:56 +00:00
if (GCOV)
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fprofile-arcs -ftest-coverage")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fprofile-arcs -ftest-coverage")
endif()
Fix OPENSSL_NO_ASM build and promote to a cmake build flag. Just changing preprocessor definitions doesn't exclude the assembly files. BUG=484327 Change-Id: I98453d291a6afb8dc8beb179f931c74301a7f434 Reviewed-on: https://boringssl-review.googlesource.com/4610 Reviewed-by: Adam Langley <agl@google.com>
2015-05-04 20:16:57 +01:00
if (OPENSSL_NO_ASM)
add_definitions(-DOPENSSL_NO_ASM)
set(ARCH "generic")
endif()
Do a cursory conversion of a few tests to GTest. For now, this is the laziest conversion possible. The intent is to just get the build setup ready so that we can get everything working in our consumers. The intended end state is: - The standalone build produces three test targets, one per library: {crypto,ssl,decrepit}_tests. - Each FOO_test is made up of: FOO/**/*_test.cc crypto/test/gtest_main.cc test_support - generate_build_files.py emits variables crypto_test_sources and ssl_test_sources. These variables are populated with FindCFiles, looking for *_test.cc. - The consuming file assembles those variables into the two test targets (plus decrepit) from there. This avoids having generate_build_files.py emit actual build rules. - Our standalone builders, Chromium, and Android just run the top-level test targets using whatever GTest-based reporting story they have. In transition, we start by converting one of two tests in each library to populate the three test targets. Those are added to all_tests.json and all_tests.go hacked to handle them transparently. This keeps our standalone builder working. generate_build_files.py, to start with, populates the new source lists manually and subtracts them out of the old machinery. We emit both for the time being. When this change rolls in, we'll write all the build glue needed to build the GTest-based tests and add it to consumers' continuous builders. Next, we'll subsume a file-based test and get the consumers working with that. (I.e. make sure the GTest targets can depend on a data file.) Once that's all done, we'll be sure all this will work. At that point, we start subsuming the remaining tests into the GTest targets and, asynchronously, rewriting tests to use GTest properly rather than cursory conversion here. When all non-GTest tests are gone, the old generate_build_files.py hooks will be removed, consumers updated to not depend on them, and standalone builders converted to not rely on all_tests.go, which can then be removed. (Unless bits end up being needed as a malloc test driver. I'm thinking we'll want to do something with --gtest_filter.) As part of this CL, I've bumped the CMake requirements (for target_include_directories) and added a few suppressions for warnings that GTest doesn't pass. BUG=129 Change-Id: I881b26b07a8739cc0b52dbb51a30956908e1b71a Reviewed-on: https://boringssl-review.googlesource.com/13232 Reviewed-by: Adam Langley <agl@google.com>
2017-01-20 00:05:47 +00:00
# Add minimal googletest targets. The provided one has many side-effects, and
# googletest has a very straightforward build.
add_library(gtest third_party/googletest/src/gtest-all.cc)
target_include_directories(gtest PRIVATE third_party/googletest)
if(CMAKE_COMPILER_IS_GNUCXX OR CMAKE_CXX_COMPILER_ID MATCHES "Clang")
# TODO(davidben): Make googletest pass -Wmissing-declarations.
set_target_properties(gtest PROPERTIES COMPILE_FLAGS "-Wno-missing-declarations")
endif()
include_directories(third_party/googletest/include)
Add a run_tests target to run all tests. It's very annoying having to remember the right incant every time I want to switch around between my build, build-release, build-asan, etc., output directories. Unfortunately, this target is pretty unfriendly without CMake 3.2+ (and Ninja 1.5+). This combination gives a USES_TERMINAL flag to add_custom_target which uses Ninja's "console" pool, otherwise the output buffering gets in the way. Ubuntu LTS is still on an older CMake, so do a version check in the meantime. CMake also has its own test mechanism (CTest), but this doesn't use it. It seems to prefer knowing what all the tests are and then tries to do its own output management and parallelizing and such. We already have our own runners. all_tests.go could actually be converted tidily, but generate_build_files.py also needs to read it, and runner.go has very specific needs. Naming the target ninja -C build test would be nice, but CTest squats that name and CMake grumps when you use a reserved name, so I've gone with run_tests. Change-Id: Ibd20ebd50febe1b4e91bb19921f3bbbd9fbcf66c Reviewed-on: https://boringssl-review.googlesource.com/6270 Reviewed-by: Adam Langley <alangley@gmail.com>
2015-10-15 02:34:40 +01:00
# Declare a dummy target to build all unit tests. Test targets should inject
# themselves as dependencies next to the target definition.
add_custom_target(all_tests)
Inital import. Initial fork from f2d678e6e89b6508147086610e985d4e8416e867 (1.0.2 beta). (This change contains substantial changes from the original and effectively starts a new history.)
2014-06-20 20:00:00 +01:00
add_subdirectory(crypto)
add_subdirectory(ssl)
add_subdirectory(ssl/test)
add_subdirectory(tool)
Add decrepit, initially containing CAST and Blowfish. decrepit will contain algorithms that we really wish didn't exist any longer. It won't be built by default in Chromium etc, but the code will exist for crummy code that still needs it. Change-Id: Ic307f5f0a69efe9e0a5fd54052f49d219e90dcdd
2015-02-03 18:45:12 +00:00
add_subdirectory(decrepit)
Add a run_tests target to run all tests. It's very annoying having to remember the right incant every time I want to switch around between my build, build-release, build-asan, etc., output directories. Unfortunately, this target is pretty unfriendly without CMake 3.2+ (and Ninja 1.5+). This combination gives a USES_TERMINAL flag to add_custom_target which uses Ninja's "console" pool, otherwise the output buffering gets in the way. Ubuntu LTS is still on an older CMake, so do a version check in the meantime. CMake also has its own test mechanism (CTest), but this doesn't use it. It seems to prefer knowing what all the tests are and then tries to do its own output management and parallelizing and such. We already have our own runners. all_tests.go could actually be converted tidily, but generate_build_files.py also needs to read it, and runner.go has very specific needs. Naming the target ninja -C build test would be nice, but CTest squats that name and CMake grumps when you use a reserved name, so I've gone with run_tests. Change-Id: Ibd20ebd50febe1b4e91bb19921f3bbbd9fbcf66c Reviewed-on: https://boringssl-review.googlesource.com/6270 Reviewed-by: Adam Langley <alangley@gmail.com>
2015-10-15 02:34:40 +01:00
Add four, basic fuzz tests. This change adds fuzzing tests for: ∙ Certificate parsing ∙ Private key parsing ∙ ClientHello parsing ∙ Server first flow (ServerHello, Certificate, etc) parsing. Change-Id: I5f53282263eaaff69b1a03c819cca73750433653 Reviewed-on: https://boringssl-review.googlesource.com/6460 Reviewed-by: Adam Langley <agl@google.com>
2015-11-09 21:57:26 +00:00
if(FUZZ)
add_subdirectory(fuzz)
endif()
Add a run_tests target to run all tests. It's very annoying having to remember the right incant every time I want to switch around between my build, build-release, build-asan, etc., output directories. Unfortunately, this target is pretty unfriendly without CMake 3.2+ (and Ninja 1.5+). This combination gives a USES_TERMINAL flag to add_custom_target which uses Ninja's "console" pool, otherwise the output buffering gets in the way. Ubuntu LTS is still on an older CMake, so do a version check in the meantime. CMake also has its own test mechanism (CTest), but this doesn't use it. It seems to prefer knowing what all the tests are and then tries to do its own output management and parallelizing and such. We already have our own runners. all_tests.go could actually be converted tidily, but generate_build_files.py also needs to read it, and runner.go has very specific needs. Naming the target ninja -C build test would be nice, but CTest squats that name and CMake grumps when you use a reserved name, so I've gone with run_tests. Change-Id: Ibd20ebd50febe1b4e91bb19921f3bbbd9fbcf66c Reviewed-on: https://boringssl-review.googlesource.com/6270 Reviewed-by: Adam Langley <alangley@gmail.com>
2015-10-15 02:34:40 +01:00
if (NOT ${CMAKE_VERSION} VERSION_LESS "3.2")
# USES_TERMINAL is only available in CMake 3.2 or later.
set(MAYBE_USES_TERMINAL USES_TERMINAL)
endif()
add_custom_target(
run_tests
COMMAND ${GO_EXECUTABLE} run util/all_tests.go -build-dir
${CMAKE_BINARY_DIR}
Fix run_tests target. The COMMANDs will “not necessarily composed into a stateful shell or batch script” so the change directory needs to be linked with the command that needs it. This fixes “make run_tests”. Change-Id: I364530fe1331aba7fa9899616916f610981c2c95 Reviewed-on: https://boringssl-review.googlesource.com/11263 Reviewed-by: Adam Langley <agl@google.com> Commit-Queue: Adam Langley <agl@google.com> CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
2016-09-24 00:43:17 +01:00
COMMAND cd ssl/test/runner &&
${GO_EXECUTABLE} test -shim-path $<TARGET_FILE:bssl_shim>
${RUNNER_ARGS}
Add a run_tests target to run all tests. It's very annoying having to remember the right incant every time I want to switch around between my build, build-release, build-asan, etc., output directories. Unfortunately, this target is pretty unfriendly without CMake 3.2+ (and Ninja 1.5+). This combination gives a USES_TERMINAL flag to add_custom_target which uses Ninja's "console" pool, otherwise the output buffering gets in the way. Ubuntu LTS is still on an older CMake, so do a version check in the meantime. CMake also has its own test mechanism (CTest), but this doesn't use it. It seems to prefer knowing what all the tests are and then tries to do its own output management and parallelizing and such. We already have our own runners. all_tests.go could actually be converted tidily, but generate_build_files.py also needs to read it, and runner.go has very specific needs. Naming the target ninja -C build test would be nice, but CTest squats that name and CMake grumps when you use a reserved name, so I've gone with run_tests. Change-Id: Ibd20ebd50febe1b4e91bb19921f3bbbd9fbcf66c Reviewed-on: https://boringssl-review.googlesource.com/6270 Reviewed-by: Adam Langley <alangley@gmail.com>
2015-10-15 02:34:40 +01:00
WORKING_DIRECTORY ${CMAKE_SOURCE_DIR}
DEPENDS all_tests bssl_shim
${MAYBE_USES_TERMINAL})
Reference in New Issue Copy Permalink