Previously we only expanded product_MODULES with LOCAL_REQUIRED_MODULES,
but not modules introduced by LOCAL_SHARED_LIBRARIES; Later we did a further
shared libary expansion in vendor_module_check.mk.
It couldn't track C in the following case:
A : B, by LOCAL_SHARED_LIBRARIES; B : C, by LOCAL_REQUIRED_MODULES.
With this change, we transformed the LOCAL_SHARED_LIBRARIES dependencies
into LOCAL_REQUIRED_MODULES dependencies before doing the required
module expansion and the loophole is closed.
All module names are now expanded to product_MODULES now and it makes
vendor_module_check.mk simpler.
Change-Id: I8835a478d2ce0ce10601a8449f446f07b01c2b7f
Previously the RS cpp files are generated by the timestamp rule. Though
we have the generated RS cpp files depend on the timestamp file, we
don't have a build recipe. In such case gmake does some "optimization"
that it skip recompiling the generated cpp files, because it assumes the
generated cpp files are already up to date even if the rs files have
been updated.
Bug: 15313144
Change-Id: Ie69ecd2c788057d3619f9c7d2a125d44c4a534a1
This fixed issue that gnumake skip updating the cpp file that includes
the generated header file when the .proto file gets updated.
For example:
Say a.cc includes b.pb.h, since b.pb.h is just byproduct of the rule
that generates b.pb.cc, and though we have dependency "b.pb.h :
b.pb.cc", but we don't have build recipe for that rule.
Gmake stupidly thinks that b.pb.h must not be updated in that case so
it skips all targets that depends on b.pb.h!
With the dumy build recipe, gmake now doesn't skip the depedent targets.
Bug: 13009798
Change-Id: I39adc09b7656bdd023f578fb8933667944fd974c
This change basically ported our target multilib to the host side.
It supports 2 host build modes: x86 and x86_64 multilib build.
For now you need to set "BUILD_HOST_64bit=true" to switch to x86_64
multilib build. Later we'll default to x86_64 build and have a flag
to force 32-bit only build, which may be needed by SDK build.
In host module definition, like in target ones, you can use the
following
LOCAL variables to set up multilib configuration:
LOCAL_MULTILIB: can be "both", "first", "32" or "64".
It also supports the same set of arch or 32-vs-64 specific LOCAL
variables.
By default, it builds only for the first arch.
To keep path compatibility, in x86_64 build files are still output to
out/host/linux-x86; Both 32-bit and 64-bit executables are in
out/host/linux-86/bin;
In x86_64 build 32-bit shared libraries are installed to
out/host/linux-x86/lib32
and 64-bit shared libraries are installed to out/host/linux-x86/lib;
32-bit object files are output to out/host/linux-x86/obj32 and 64-bit
object files
are output to out/host/linux-x86/obj.
Bug: 13751317
Change-Id: I6044f83b7db369a33e05209e8c588eb6dc83409f
prebuilts/ndk/current/platforms/android-19/arch-x86_64/usr/lib
is renamed to usr/lib64 to be more consistent with rest of
lib paths in x86_64 toolchain, which is multilib
See https://android-review.googlesource.com/#/c/92441/
Change-Id: I4e59245505d0fa87ae3608e81e715ccfcecc5ec8
Some executables will need to be built for both 32-bit and 64-bit.
For tests, it will be convienient to keep the name of the executable
the same, but install them in a different location. Add
LOCAL_MODULE_PATH_32 and LOCAL_MODULE_PATH_64 to allow a module
to specify different paths for 32-bit and 64-bit executables.
Change-Id: I3be830e899c6d485fe55c25c66b20b3fe64c795e
Previously we have only one set of include/lib paths for
LOCAL_NDK_STL_VARIANT:=gnustl_static regardless of GCC
version, which is wrong because each GCC version
come with its own libstdc++.
Change-Id: I2a01c2120b6948aedce00e2f8d08dfc6932126dd
Previously the installed shared library dependency doesn't include
modules introduced by LOCAL_SHARED_LIBRARIES_<arch>.
This change fix the problem.
It also cleans up use of the shared library variable.
Bug: 13528787
Change-Id: Id8d807cc57f0ec4a71f18b64545d91191efad8fb
So a library can export the proto's include path that can be used with
both archs in multilib build.
Change-Id: Ia0f92f0b40e39dc3fa426c69c52139a0a8f04077
This makes sure copy_headers.mk only be included onces, no matter
it's for the 1st arch or the 2nd arch.
Change-Id: I80a558fbdb52861f176bd27a21c302069a5cc3ce
The LOCAL_*_$(TARGET_ARCH) variables don't make sense for host
modules, only append use them for target modules.
Also complete the list of LOCAL_*_arch and LOCAL_*_32/64 to be
consistent.
Change-Id: I00c83e5c4e08ed9a844f9f99a79ce4bcc3f0bf11
1. Following the setup of gcc in build/core/combo/,
we added the [HOST|TARGET]_<arch>.mk clang config files,
and load only the configs needed by the current product.
2. Added support for the 2nd arch.
Change-Id: I2a383418a9688a050b39492f8e489d40eeeb5f2d
Support the following new variables based on whether the current multilib
target is 32 bit or 64 bit:
LOCAL_CFLAGS_32
LOCAL_CFLAGS_64
LOCAL_LDFLAGS_32
LOCAL_LDFLAGS_64
LOCAL_ASFLAGS_32
LOCAL_ASFLAGS_64
LOCAL_C_INCLUDES_32
LOCAL_C_INCLUDES_64
Change-Id: Ia868d56dff114be301bf8297eec768675f186927
To ease the transition between toolchains, allow a target to specify
a list of cflags that the toolchain does not support. These will be
filtered out of the cflags provided by the module.
Add TARGET_GLOBAL_UNSUPPORTED_CFLAGS := -fstack-protector for the
aarch64 toolchain, it does not yet suport -fstack-protector.
Change-Id: I168d0c6f131326fad305ec86fad46e6a3e03295a
Allow modules to generate source into $OUT/gen, which will then
be copied into $OUT/obj and $OUT/obj_$(TARGET_2ND_ARCH) as
necessary. This allows a single build rule invocation that includes
generated source to build for the first and second architectures.
Modules will need to change calls to local-intermediates-dir into
local-generated-sources-dir.
Change-Id: I62504bad9454b3d9fde7b84ab9f0a487a2ecf0bf
Set arm_objects_mode and normal_objects_mode when building a
module for arm when it is the 2nd arch.
Change-Id: I5f7df519b6e1dde6cbf92d106681f07a58e1f1f2
With those variables, you can set up different values for TARGET_ARCH
and TARGET_2ND_ARCH.
Also fixed a couple of variables.
Bug: 11654773
Change-Id: I4c7684a562cd5877d18f67d4f848b8df07d0103b
Conflicts:
core/base_rules.mk
The rules for the 2nd arch are set up in the second inclusion
of static_library_internal.mk.
libfoo of the 2nd arch will be built into
$(PRODUCT_OUT)/obj_$(TARGET_2ND_ARCH)/libfoo_intermediates/libfoo.a.
Bug: 11654773
Change-Id: I1d92733968fc442e9225b4df5bd1b551a81d89f7
The new option WITH_STATIC_ANALYZER=1 instructs build system to
run static analyzer via "clang --analyze" on a successful build.
If analyzer finds any issue, instruction to open report is displayed.
See http://clang-analyzer.llvm.org/scan-build.html for details.
WITH_STATIC_ANALYZER trumps WITH_SYNTAX_CHECK if both exist.
Project use lots of GCC extensions (eg. nested function) not supported
by clang may opt out by adding LOCAL_NO_STATIC_ANALYZER:=true
Change-Id: I9970560560bd52ce5f0fd7129c3488629627c735
The new option WITH_SYNTAX_CHECK=1 instructs build system to invoke
"clang -fsyntax-only" to utilize clang's better diagnostics before calling
LOCAL_CC/LOCAL_CXX for code generation. The compilation time is slightly
longer, and the generated object file should be the same as w/o WITH_SYNTAX_CHECK
Project use lots of GCC extensions (eg. nested function) not supported
by clang may opt out by adding LOCAL_NO_SYNTAX_CHECK:=true
Change-Id: I5689586788ef049bd967364f71f31f1e359bd121
The issues:
- The size increase from utilizing FDO is quite large while
utilizing runtime profiles in build.
- By default, FDO is utilized globally if the target arch variant
profiles exist.
- Not all modules can show statistical significance in
performance comparison, yet still suffer the size increase.
The solution:
- Only enable FDO locally with LOCAL_FDO_SUPPORT
for modules which may benefit enough to justify the size
tradeoff.
Solution notes:
- I've noted statistical significance in libwebcore and libskia
thus far from utilizing FDO.
- Analysis included sunspider, drawcanvas benchmarks, as
well as gooda analysis on both arm and x86
- To support runtime profile generation in modules which have
LOCAL_FDO_SUPPORT specified,
BUILD_FDO_INSTRUMENTATION is still used. Otherwise,
if the target arch variant profiles exist, FDO is utilized for
specified modules.
Change-Id: I7e95266943ff47c7d82b02e6200fd09911d0bb57
Rule-generated .o files (in gen_o_objects) were being given a dependency
on everything in LOCAL_GENERATED_SOURCES (except for other .o files);
unfortunately this can still create cycles in cases where there are
explicit dependencies between entries in LOCAL_GENERATED_SOURCES.
Instead, make handling of generated .o files consistent with other
generated files (which don't automatically get any dependencies on other
generated files) by excluding them from the target side of the rule.
Change-Id: I3fb5652dc3d85012c179a03b81887d16a85ab3bf
Now you can have a board config variable BOARD_HAL_STATIC_LIBRARIES,
which is a list of board-specific HAL static library names with pattern
"lib<library_name>.<board_specific_suffix>". LOCAL_HAL_STATIC_LIBRARIES
is a list of "lib<library_name>" and any matched
BOARD_HAL_STATIC_LIBRARIES will be added to the LOCAL_STATIC_LIBRARIES;
if no match is found, lib<library_name>.default will be used.
Bug: 10262105
Change-Id: Ic89d8d417d1dd65a227e4187a157fd3b77c4af34
