timoreo/olm
1
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
olm/Makefile

318 lines
10 KiB
Makefile
Raw Permalink Normal View History

Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
#!/usr/bin/make -f
Android : improve the version number management (#46) Add functions to make the difference between the native and the java code version. Factor out the version management in the makefiles.
2017-03-02 13:01:32 +01:00
include common.mk
Specify the soname in the library and its name This provides users of this library the guarantee that the ABI will stay stable when MAJOR will reach 1, and will stay backwards compatible for the entire duration of the 1.x.y branch. It does require the maintainers to always update the version in the Makefile at every ABI change.
2016-07-04 23:38:04 +02:00
VERSION := $(MAJOR).$(MINOR).$(PATCH)
Add an install and an install-debug targets This makes the user able to use the familiar `make install` syntax, and allows overriding of the default directories using the DESTDIR and PREFIX variables, for example: make DESTDIR=packaging PREFIX=/usr install
2016-07-04 23:38:05 +02:00
PREFIX ?= /usr/local
Use a rule to make the build directories
2016-05-23 17:21:37 +02:00
BUILD_DIR := build
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
RELEASE_OPTIMIZE_FLAGS ?= -g -O3
DEBUG_OPTIMIZE_FLAGS ?= -g -O0
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
JS_OPTIMIZE_FLAGS ?= -O3
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
FUZZING_OPTIMIZE_FLAGS ?= -O3
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
CC = gcc
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
EMCC = emcc
Add a fuzzer for olm_group_decrypt
2016-05-26 14:25:34 +02:00
AFL_CC = afl-gcc
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
AFL_CXX = afl-g++
Support building as a static library Signed-off-by: Andreas Zwinkau <qznc@web.de>
2017-03-16 20:20:16 +01:00
AR = ar
Avoid ldconfig in the Makefile ... because OSX doesn't support it.
2016-09-05 12:56:53 +02:00
Build on mac
2018-09-20 12:31:09 +02:00
UNAME := $(shell uname)
ifeq ($(UNAME),Darwin)
SO := dylib
OLM_LDFLAGS :=
else
SO := so
OLM_LDFLAGS := -Wl,-soname,libolm.so.$(MAJOR) \
-Wl,--version-script,version_script.ver
endif
RELEASE_TARGET := $(BUILD_DIR)/libolm.$(SO).$(VERSION)
Support building as a static library Signed-off-by: Andreas Zwinkau <qznc@web.de>
2017-03-16 20:20:16 +01:00
STATIC_RELEASE_TARGET := $(BUILD_DIR)/libolm.a
Build on mac
2018-09-20 12:31:09 +02:00
DEBUG_TARGET := $(BUILD_DIR)/libolm_debug.$(SO).$(VERSION)
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
JS_WASM_TARGET := javascript/olm.js
JS_ASMJS_TARGET := javascript/olm_legacy.js
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
JS_EXPORTED_FUNCTIONS := javascript/exported_functions.json
WebAssembly support! Quite a lot going on in this PR: * Updates to support recent emscripten, switching to WASM which is now the default * Use emscripten's MODULARIZE option rather than wrapping it ourself, since doing so in pre-post js doesn't work anymore. * Most changes are moving the emscripten runtime functions to top-level calls rather than in the Module object. * Get rid of duplicated NULL_BYTE_PADDING_LENGTH * Fix ciphertext_length used without being declared * Fix things that caused the closure compiler to error, eg. using OLM_OPTIONS without a declaration. * Wait until module is inited to do OLM_ERROR = olm_error() The main BREAKING CHANGE here is that the module now needs to initialise asyncronously (because it has to load the wasm file). require()ing olm now gives a function which needs to be called to create an instance. The resulting object has a promise-like then() method that can be used to detect when the module is ready. (We could use MODULARIZE_INSTANCE to return the module directly as before, rather than the function, but then we don't get the .then() method).
2018-09-21 17:01:51 +02:00
JS_EXTRA_EXPORTED_RUNTIME_METHODS := ALLOC_STACK
Make OLM_OPTIONS work again The closure compiler was just renaming the variable so it never would have picked them up. Make it an extern so it knows what to do.
2018-09-21 17:35:17 +02:00
JS_EXTERNS := javascript/externs.js
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
initial implementation of short authentication string generation
2019-01-22 05:21:41 +01:00
PUBLIC_HEADERS := include/olm/olm.h include/olm/outbound_group_session.h include/olm/inbound_group_session.h include/olm/pk.h include/olm/sas.h
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
SOURCES := $(wildcard src/*.cpp) $(wildcard src/*.c) \
lib/crypto-algorithms/sha256.c \
lib/crypto-algorithms/aes.c \
lib/curve25519-donna/curve25519-donna.c
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
FUZZER_SOURCES := $(wildcard fuzzers/fuzz_*.cpp) $(wildcard fuzzers/fuzz_*.c)
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
TEST_SOURCES := $(wildcard tests/test_*.cpp) $(wildcard tests/test_*.c)
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
OBJECTS := $(patsubst %.c,%.o,$(patsubst %.cpp,%.o,$(SOURCES)))
RELEASE_OBJECTS := $(addprefix $(BUILD_DIR)/release/,$(OBJECTS))
DEBUG_OBJECTS := $(addprefix $(BUILD_DIR)/debug/,$(OBJECTS))
FUZZER_OBJECTS := $(addprefix $(BUILD_DIR)/fuzzers/objects/,$(OBJECTS))
FUZZER_BINARIES := $(addprefix $(BUILD_DIR)/,$(basename $(FUZZER_SOURCES)))
FUZZER_DEBUG_BINARIES := $(patsubst $(BUILD_DIR)/fuzzers/fuzz_%,$(BUILD_DIR)/fuzzers/debug_%,$(FUZZER_BINARIES))
TEST_BINARIES := $(patsubst tests/%,$(BUILD_DIR)/tests/%,$(basename $(TEST_SOURCES)))
JS_OBJECTS := $(addprefix $(BUILD_DIR)/javascript/,$(OBJECTS))
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
# pre & post are the js-pre/js-post options to emcc.
# They are injected inside the modularised code and
# processed by the optimiser.
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
JS_PRE := $(wildcard javascript/*pre.js)
Javascript bindings for group sessions
2016-05-25 15:42:49 +02:00
JS_POST := javascript/olm_outbound_group_session.js \
javascript/olm_inbound_group_session.js \
add initial implementation of basic private key encryption functionality
2018-06-11 23:48:45 +02:00
javascript/olm_pk.js \
initial implementation of short authentication string generation
2019-01-22 05:21:41 +01:00
javascript/olm_sas.js \
Javascript bindings for group sessions
2016-05-25 15:42:49 +02:00
javascript/olm_post.js
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
# The prefix & suffix are just added onto the start & end
# of what comes out emcc, so are outside of the modularised
# code and not seen by the opimiser.
JS_PREFIX := javascript/olm_prefix.js
JS_SUFFIX := javascript/olm_suffix.js
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
DOCS := tracing/README.html \
First stab at a megolm spec
2016-09-22 12:04:19 +02:00
docs/megolm.html \
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
docs/olm.html \
Add a document on signing keys
2016-10-27 12:53:55 +02:00
docs/signing.html \
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
README.html \
CHANGELOG.html
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
Fix Ed25519 keypair generation Ed25519 private keys, it turns out, have 64 bytes, not 32. We were previously generating only 32 bytes (which is all that is required to generate the public key), and then using the public key as the upper 32 bytes when generating the per-message session key. This meant that everything appeared to work, but the security of the private key was severely compromised. By way of fixes: * Use the correct algorithm for generating the Ed25519 private key, and store all 512 bits of it. * Update the account pickle format and refuse to load the old format (since we should consider it compromised). * Bump the library version, and add a function to retrieve the library version, so that applications can verify that they are linked against a fixed version of the library. * Remove the curve25519_{sign, verify} functions which were unused and of dubious quality.
2016-09-01 14:35:23 +02:00
CPPFLAGS += -Iinclude -Ilib \
-DOLMLIB_VERSION_MAJOR=$(MAJOR) -DOLMLIB_VERSION_MINOR=$(MINOR) \
-DOLMLIB_VERSION_PATCH=$(PATCH)
Build the crypto libs as C99 Some of the crypto libs rely on UINT64_C, which in glibc 2.17 and earlier was not defined for C++ code (see https://sourceware.org/bugzilla/show_bug.cgi?id=15366).
2016-05-24 15:36:07 +02:00
# we rely on <stdint.h>, which was introduced in C99
CFLAGS += -Wall -Werror -std=c99 -fPIC
Fix warnings and set -Werror
2016-05-18 19:03:28 +02:00
CXXFLAGS += -Wall -Werror -std=c++11 -fPIC
LDFLAGS += -Wall -Werror
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
WebAssembly support! Quite a lot going on in this PR: * Updates to support recent emscripten, switching to WASM which is now the default * Use emscripten's MODULARIZE option rather than wrapping it ourself, since doing so in pre-post js doesn't work anymore. * Most changes are moving the emscripten runtime functions to top-level calls rather than in the Module object. * Get rid of duplicated NULL_BYTE_PADDING_LENGTH * Fix ciphertext_length used without being declared * Fix things that caused the closure compiler to error, eg. using OLM_OPTIONS without a declaration. * Wait until module is inited to do OLM_ERROR = olm_error() The main BREAKING CHANGE here is that the module now needs to initialise asyncronously (because it has to load the wasm file). require()ing olm now gives a function which needs to be called to create an instance. The resulting object has a promise-like then() method that can be used to detect when the module is ready. (We could use MODULARIZE_INSTANCE to return the module directly as before, rather than the function, but then we don't get the .then() method).
2018-09-21 17:01:51 +02:00
EMCCFLAGS = --closure 1 --memory-init-file 0 -s NO_FILESYSTEM=1 -s INVOKE_RUN=0 -s MODULARIZE=1
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
# Olm generally doesn't need a lot of memory to encrypt / decrypt its usual
# payloads (ie. Matrix messages), but we do need about 128K of heap to encrypt
# a 64K event (enough to store the ciphertext and the plaintext, bearing in
# mind that the plaintext can only be 48K because base64). We also have about
# 36K of statics. So let's have 256K of memory.
# (This can't be changed by the app with wasm since it's baked into the wasm).
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
# (emscripten also mandates at least 16MB of memory for asm.js now, so
# we don't use this for the legacy build.)
EMCCFLAGS_WASM += -s TOTAL_STACK=65536 -s TOTAL_MEMORY=262144
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
EMCCFLAGS_ASMJS += -s WASM=0
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
EMCC.c = $(EMCC) $(CFLAGS) $(CPPFLAGS) -c
EMCC.cc = $(EMCC) $(CXXFLAGS) $(CPPFLAGS) -c
EMCC_LINK = $(EMCC) $(LDFLAGS) $(EMCCFLAGS)
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
AFL.c = $(AFL_CC) $(CFLAGS) $(CPPFLAGS) -c
AFL.cc = $(AFL_CXX) $(CXXFLAGS) $(CPPFLAGS) -c
AFL_LINK.c = $(AFL_CC) $(LDFLAGS) $(CFLAGS) $(CPPFLAGS)
AFL_LINK.cc = $(AFL_CXX) $(LDFLAGS) $(CXXFLAGS) $(CPPFLAGS)
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
# generate .d files when compiling
CPPFLAGS += -MMD
### per-target variables
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(RELEASE_OBJECTS): CFLAGS += $(RELEASE_OPTIMIZE_FLAGS)
$(RELEASE_OBJECTS): CXXFLAGS += $(RELEASE_OPTIMIZE_FLAGS)
$(RELEASE_TARGET): LDFLAGS += $(RELEASE_OPTIMIZE_FLAGS)
$(DEBUG_OBJECTS): CFLAGS += $(DEBUG_OPTIMIZE_FLAGS)
$(DEBUG_OBJECTS): CXXFLAGS += $(DEBUG_OPTIMIZE_FLAGS)
$(DEBUG_TARGET): LDFLAGS += $(DEBUG_OPTIMIZE_FLAGS)
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
$(TEST_BINARIES): CPPFLAGS += -Itests/include
Build the tests with -g think this got lost when mark added separate release/debug builds
2016-05-24 17:28:14 +02:00
$(TEST_BINARIES): LDFLAGS += $(DEBUG_OPTIMIZE_FLAGS) -L$(BUILD_DIR)
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
$(FUZZER_OBJECTS): CFLAGS += $(FUZZER_OPTIMIZE_FLAGS)
$(FUZZER_OBJECTS): CXXFLAGS += $(FUZZER_OPTIMIZE_FLAGS)
$(FUZZER_BINARIES): CPPFLAGS += -Ifuzzers/include
$(FUZZER_BINARIES): LDFLAGS += $(FUZZER_OPTIMIZE_FLAGS) -L$(BUILD_DIR)
$(FUZZER_DEBUG_BINARIES): CPPFLAGS += -Ifuzzers/include
$(FUZZER_DEBUG_BINARIES): LDFLAGS += $(DEBUG_OPTIMIZE_FLAGS)
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
$(JS_OBJECTS): CFLAGS += $(JS_OPTIMIZE_FLAGS)
$(JS_OBJECTS): CXXFLAGS += $(JS_OPTIMIZE_FLAGS)
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
$(JS_WASM_TARGET): LDFLAGS += $(JS_OPTIMIZE_FLAGS)
$(JS_ASMJS_TARGET): LDFLAGS += $(JS_OPTIMIZE_FLAGS)
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
### Fix to make mkdir work on windows and linux
ifeq ($(shell echo "check_quotes"),"check_quotes")
WINDOWS := yes
else
WINDOWS := no
endif
ifeq ($(WINDOWS),yes)
mkdir = mkdir $(subst /,\,$(1)) > nul 2>&1 || (exit 0)
else
mkdir = mkdir -p $(1)
endif
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
### top-level targets
Fix Ed25519 keypair generation Ed25519 private keys, it turns out, have 64 bytes, not 32. We were previously generating only 32 bytes (which is all that is required to generate the public key), and then using the public key as the upper 32 bytes when generating the per-message session key. This meant that everything appeared to work, but the security of the private key was severely compromised. By way of fixes: * Use the correct algorithm for generating the Ed25519 private key, and store all 512 bits of it. * Update the account pickle format and refuse to load the old format (since we should consider it compromised). * Bump the library version, and add a function to retrieve the library version, so that applications can verify that they are linked against a fixed version of the library. * Remove the curve25519_{sign, verify} functions which were unused and of dubious quality.
2016-09-01 14:35:23 +02:00
lib: $(RELEASE_TARGET)
Add an 'all' target to the Makefile So that we can build everything together.
2016-05-20 16:28:54 +02:00
.PHONY: lib
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(RELEASE_TARGET): $(RELEASE_OBJECTS)
$(CXX) $(LDFLAGS) --shared -fPIC \
Build on mac
2018-09-20 12:31:09 +02:00
$(OLM_LDFLAGS) \
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(OUTPUT_OPTION) $(RELEASE_OBJECTS)
Build on mac
2018-09-20 12:31:09 +02:00
ln -sf libolm.$(SO).$(VERSION) $(BUILD_DIR)/libolm.$(SO).$(MAJOR)
Merge branch 'master' into poljar
2018-10-12 23:02:51 +02:00
ln -sf libolm.$(SO).$(VERSION) $(BUILD_DIR)/libolm.$(SO)
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
Fix Ed25519 keypair generation Ed25519 private keys, it turns out, have 64 bytes, not 32. We were previously generating only 32 bytes (which is all that is required to generate the public key), and then using the public key as the upper 32 bytes when generating the per-message session key. This meant that everything appeared to work, but the security of the private key was severely compromised. By way of fixes: * Use the correct algorithm for generating the Ed25519 private key, and store all 512 bits of it. * Update the account pickle format and refuse to load the old format (since we should consider it compromised). * Bump the library version, and add a function to retrieve the library version, so that applications can verify that they are linked against a fixed version of the library. * Remove the curve25519_{sign, verify} functions which were unused and of dubious quality.
2016-09-01 14:35:23 +02:00
debug: $(DEBUG_TARGET)
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
.PHONY: debug
$(DEBUG_TARGET): $(DEBUG_OBJECTS)
Use a version script to restrict symbols in the .so We want to stop aes_* and sha_* functions leaking out of our .so, as well as internal _olm_* symbols. This also means we need to link the unit tests against the objects. Possibly we should distinguish between unit tests and integration tests.
2016-05-20 13:01:18 +02:00
$(CXX) $(LDFLAGS) --shared -fPIC \
Build on mac
2018-09-20 12:31:09 +02:00
$(OLM_LDFLAGS) \
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(OUTPUT_OPTION) $(DEBUG_OBJECTS)
Build on mac
2018-09-20 12:31:09 +02:00
ln -sf libolm_debug.$(SO).$(VERSION) $(BUILD_DIR)/libolm_debug.$(SO).$(MAJOR)
Specify the soname in the library and its name This provides users of this library the guarantee that the ABI will stay stable when MAJOR will reach 1, and will stay backwards compatible for the entire duration of the 1.x.y branch. It does require the maintainers to always update the version in the Makefile at every ABI change.
2016-07-04 23:38:04 +02:00
Support building as a static library Signed-off-by: Andreas Zwinkau <qznc@web.de>
2017-03-16 20:20:16 +01:00
static: $(STATIC_RELEASE_TARGET)
.PHONY: static
$(STATIC_RELEASE_TARGET): $(RELEASE_OBJECTS)
Makefile: fix tab/space substitution.
2017-03-29 11:44:54 +02:00
$(AR) rcs $@ $^
Support building as a static library Signed-off-by: Andreas Zwinkau <qznc@web.de>
2017-03-16 20:20:16 +01:00
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
js: $(JS_WASM_TARGET) $(JS_ASMJS_TARGET)
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
.PHONY: js
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
# Note that the output file we give to emcc determines the name of the
# wasm file baked into the js, hence messing around outputting to olm.js
# and then renaming it.
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
$(JS_WASM_TARGET): $(JS_OBJECTS) $(JS_PRE) $(JS_POST) $(JS_EXPORTED_FUNCTIONS) $(JS_PREFIX) $(JS_SUFFIX)
Make OLM_OPTIONS work again The closure compiler was just renaming the variable so it never would have picked them up. Make it an extern so it knows what to do.
2018-09-21 17:35:17 +02:00
EMCC_CLOSURE_ARGS="--externs $(JS_EXTERNS)" $(EMCC_LINK) \
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
$(EMCCFLAGS_WASM) \
$(foreach f,$(JS_PRE),--pre-js $(f)) \
$(foreach f,$(JS_POST),--post-js $(f)) \
-s "EXPORTED_FUNCTIONS=@$(JS_EXPORTED_FUNCTIONS)" \
-s "EXTRA_EXPORTED_RUNTIME_METHODS=$(JS_EXTRA_EXPORTED_RUNTIME_METHODS)" \
$(JS_OBJECTS) -o $@
mv $@ javascript/olmtmp.js
cat $(JS_PREFIX) javascript/olmtmp.js $(JS_SUFFIX) > $@
rm javascript/olmtmp.js
$(JS_ASMJS_TARGET): $(JS_OBJECTS) $(JS_PRE) $(JS_POST) $(JS_EXPORTED_FUNCTIONS) $(JS_PREFIX) $(JS_SUFFIX)
Make OLM_OPTIONS work again The closure compiler was just renaming the variable so it never would have picked them up. Make it an extern so it knows what to do.
2018-09-21 17:35:17 +02:00
EMCC_CLOSURE_ARGS="--externs $(JS_EXTERNS)" $(EMCC_LINK) \
Dual-build wasm and asm.js olm
2018-10-04 21:09:54 +02:00
$(EMCCFLAGS_ASMJS) \
Javascript bindings for group sessions
2016-05-25 15:42:49 +02:00
$(foreach f,$(JS_PRE),--pre-js $(f)) \
$(foreach f,$(JS_POST),--post-js $(f)) \
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
-s "EXPORTED_FUNCTIONS=@$(JS_EXPORTED_FUNCTIONS)" \
WebAssembly support! Quite a lot going on in this PR: * Updates to support recent emscripten, switching to WASM which is now the default * Use emscripten's MODULARIZE option rather than wrapping it ourself, since doing so in pre-post js doesn't work anymore. * Most changes are moving the emscripten runtime functions to top-level calls rather than in the Module object. * Get rid of duplicated NULL_BYTE_PADDING_LENGTH * Fix ciphertext_length used without being declared * Fix things that caused the closure compiler to error, eg. using OLM_OPTIONS without a declaration. * Wait until module is inited to do OLM_ERROR = olm_error() The main BREAKING CHANGE here is that the module now needs to initialise asyncronously (because it has to load the wasm file). require()ing olm now gives a function which needs to be called to create an instance. The resulting object has a promise-like then() method that can be used to detect when the module is ready. (We could use MODULARIZE_INSTANCE to return the module directly as before, rather than the function, but then we don't get the .then() method).
2018-09-21 17:01:51 +02:00
-s "EXTRA_EXPORTED_RUNTIME_METHODS=$(JS_EXTRA_EXPORTED_RUNTIME_METHODS)" \
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
$(JS_OBJECTS) -o $@
Another day, another interface Change the interface again, hopefully this time a bit more normal. Now we wrap the emscripten module completely and just expose the high level objects. The olm library export is now imported as normal (ie. returns a module rather than a function returning a module) but has an `init` method which *must* be called. This returns a promise which resolves when the module is ready. It also rejects if the module failed to set up, unlike before (and unlike the promise-not-a-promise that emscripten returns). Generally catch failures to init the module.
2018-09-25 18:13:29 +02:00
mv $@ javascript/olmtmp.js
cat $(JS_PREFIX) javascript/olmtmp.js $(JS_SUFFIX) > $@
rm javascript/olmtmp.js
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
build_tests: $(TEST_BINARIES)
test: build_tests
for i in $(TEST_BINARIES); do \
echo $$i; \
Use a version script to restrict symbols in the .so We want to stop aes_* and sha_* functions leaking out of our .so, as well as internal _olm_* symbols. This also means we need to link the unit tests against the objects. Possibly we should distinguish between unit tests and integration tests.
2016-05-20 13:01:18 +02:00
$$i || exit $$?; \
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
done
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
fuzzers: $(FUZZER_BINARIES) $(FUZZER_DEBUG_BINARIES)
.PHONY: fuzzers
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
$(JS_EXPORTED_FUNCTIONS): $(PUBLIC_HEADERS)
Replace the impenetrable line of perl with python Mostly because the standard emscripten docker image does not have libjson-perl, but python always comes with json. But also because it was impenetrable.
2018-10-03 17:06:15 +02:00
./exports.py $^ > $@.tmp
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
mv $@.tmp $@
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
all: test js lib debug doc
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
.PHONY: all
Add a Makefile rule to install the headers Signed-off-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr>
2016-09-21 19:08:05 +02:00
install-headers: $(PUBLIC_HEADERS)
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
test -d $(DESTDIR)$(PREFIX)/include/olm || $(call mkdir,$(DESTDIR)$(PREFIX)/include/olm)
Add a Makefile rule to install the headers Signed-off-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr>
2016-09-21 19:08:05 +02:00
install -Dm644 $(PUBLIC_HEADERS) $(DESTDIR)$(PREFIX)/include/olm/
Makefile: install-headers is phony
2016-09-23 13:22:34 +02:00
.PHONY: install-headers
Add a Makefile rule to install the headers Signed-off-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr>
2016-09-21 19:08:05 +02:00
install-debug: debug install-headers
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
test -d $(DESTDIR)$(PREFIX)/lib || $(call mkdir,$(DESTDIR)$(PREFIX)/lib)
Build on mac
2018-09-20 12:31:09 +02:00
install -Dm755 $(DEBUG_TARGET) $(DESTDIR)$(PREFIX)/lib/libolm_debug.$(SO).$(VERSION)
Makefile: Overwrite symlinks in install and install-debug Without this, it's not possible to run the install or install-debug rules multiple times. Signed-off-by: Trygve Aaberge <trygveaa@gmail.com>
2019-08-07 21:22:09 +02:00
ln -sf libolm_debug.$(SO).$(VERSION) $(DESTDIR)$(PREFIX)/lib/libolm_debug.$(SO).$(MAJOR)
ln -sf libolm_debug.$(SO).$(VERSION) $(DESTDIR)$(PREFIX)/lib/libolm_debug.$(SO)
Add an install and an install-debug targets This makes the user able to use the familiar `make install` syntax, and allows overriding of the default directories using the DESTDIR and PREFIX variables, for example: make DESTDIR=packaging PREFIX=/usr install
2016-07-04 23:38:05 +02:00
.PHONY: install-debug
Add a Makefile rule to install the headers Signed-off-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr>
2016-09-21 19:08:05 +02:00
install: lib install-headers
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
test -d $(DESTDIR)$(PREFIX)/lib || $(call mkdir,$(DESTDIR)$(PREFIX)/lib)
Build on mac
2018-09-20 12:31:09 +02:00
install -Dm755 $(RELEASE_TARGET) $(DESTDIR)$(PREFIX)/lib/libolm.$(SO).$(VERSION)
Makefile: Overwrite symlinks in install and install-debug Without this, it's not possible to run the install or install-debug rules multiple times. Signed-off-by: Trygve Aaberge <trygveaa@gmail.com>
2019-08-07 21:22:09 +02:00
ln -sf libolm.$(SO).$(VERSION) $(DESTDIR)$(PREFIX)/lib/libolm.$(SO).$(MAJOR)
ln -sf libolm.$(SO).$(VERSION) $(DESTDIR)$(PREFIX)/lib/libolm.$(SO)
Add an install and an install-debug targets This makes the user able to use the familiar `make install` syntax, and allows overriding of the default directories using the DESTDIR and PREFIX variables, for example: make DESTDIR=packaging PREFIX=/usr install
2016-07-04 23:38:05 +02:00
.PHONY: install
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
clean:;
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
rm -rf $(BUILD_DIR) $(DOCS)
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
.PHONY: clean
Add an 'all' target to the Makefile So that we can build everything together.
2016-05-20 16:28:54 +02:00
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
doc: $(DOCS)
.PHONY: doc
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
### rules for building objects
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/release/%.o: %.c
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(COMPILE.c) $(OUTPUT_OPTION) $<
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/release/%.o: %.cpp
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(COMPILE.cc) $(OUTPUT_OPTION) $<
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/debug/%.o: %.c
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
$(COMPILE.c) $(OUTPUT_OPTION) $<
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/debug/%.o: %.cpp
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
$(COMPILE.cc) $(OUTPUT_OPTION) $<
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/javascript/%.o: %.c
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
$(EMCC.c) $(OUTPUT_OPTION) $<
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/javascript/%.o: %.cpp
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Add support for building the JS wrappers to the Makefile Now that we have C and C++, we need to split the compile and link steps (because we need different flags for the C and C++ files), so this is easier with a Makefile.
2016-05-20 10:36:50 +02:00
$(EMCC.cc) $(OUTPUT_OPTION) $<
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(BUILD_DIR)/tests/%: tests/%.c $(DEBUG_OBJECTS)
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Don't use $^ cause that includes the header files
2016-05-23 17:23:57 +02:00
$(LINK.c) $< $(DEBUG_OBJECTS) $(LOADLIBES) $(LDLIBS) -o $@
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
$(BUILD_DIR)/tests/%: tests/%.cpp $(DEBUG_OBJECTS)
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Don't use $^ cause that includes the header files
2016-05-23 17:23:57 +02:00
$(LINK.cc) $< $(DEBUG_OBJECTS) $(LOADLIBES) $(LDLIBS) -o $@
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/fuzzers/objects/%.o: %.c
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
$(AFL.c) $(OUTPUT_OPTION) $<
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
$(BUILD_DIR)/fuzzers/objects/%.o: %.cpp
Update Makefile to support building on Windows The code for this change is taken from https://stackoverflow.com/a/30225575/4929236 This patch is Signed-Off-By: Marcel Radzio <marcel@radzio-sh.de>
2018-04-12 11:51:16 +02:00
$(call mkdir,$(dir $@))
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
$(AFL.cc) $(OUTPUT_OPTION) $<
$(BUILD_DIR)/fuzzers/fuzz_%: fuzzers/fuzz_%.c $(FUZZER_OBJECTS)
$(AFL_LINK.c) $< $(FUZZER_OBJECTS) $(LOADLIBES) $(LDLIBS) -o $@
$(BUILD_DIR)/fuzzers/fuzz_%: fuzzers/fuzz_%.cpp $(FUZZER_OBJECTS)
$(AFL_LINK.cc) $< $(FUZZER_OBJECTS) $(LOADLIBES) $(LDLIBS) -o $@
$(BUILD_DIR)/fuzzers/debug_%: fuzzers/fuzz_%.c $(DEBUG_OBJECTS)
$(LINK.c) $< $(DEBUG_OBJECTS) $(LOADLIBES) $(LDLIBS) -o $@
$(BUILD_DIR)/fuzzers/debug_%: fuzzers/fuzz_%.cpp $(DEBUG_OBJECTS)
$(LINK.cc) $< $(DEBUG_OBJECTS) $(LOADLIBES) $(LDLIBS) -o $@
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
Makefile rules to build the rst into html
2016-07-11 15:50:49 +02:00
%.html: %.rst
rst2html $< $@
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
### dependencies
Put the .o files in separate directories so we can have both release and debug versions
2016-05-23 16:58:18 +02:00
-include $(RELEASE_OBJECTS:.o=.d)
-include $(DEBUG_OBJECTS:.o=.d)
Compile some of the crypto libs directly sha256.c and aes.c contain conflicting declarations, so we need to compile them as separate units. This requires a bit more Makefile-shuffling; the build directory now includes 'src' or 'lib' as appropriate, and we just mkdir -p before each compilation.
2016-05-25 18:08:44 +02:00
-include $(JS_OBJECTS:.o=.d)
Add a makefile Replace the python scripts for building the shared lib and tests with a Makefile, which makes it easier to handle a mix of C and C++.
2016-05-13 12:36:41 +02:00
-include $(TEST_BINARIES:=.d)
Add support for building fuzzers using american fuzzy lop Builds fuzzers using http://lcamtuf.coredump.cx/afl/
2016-05-23 18:32:24 +02:00
-include $(FUZZER_OBJECTS:.o=.d)
-include $(FUZZER_BINARIES:=.d)
-include $(FUZZER_DEBUG_BINARIES:=.d)
Reference in a new issue Copy permalink