Both aidl and clang/gcc were putting their dependencies in the same
place. Move aidl's dependencies to a file ending with .aidl.P rather
than the compiler's .P.
While here, inform kati that we have these special dep files.
Bug:26409006
Test: Rebuild, note both files being generated
Change-Id: I29d2eea822235d60713c2059f3a314e475eb5aa3
Transform ../ to dotdot/ for C++ generateds from .aidl source files.
This forces us to use one layer of indirection to calculate the build
rules for .aidl files, since we can no longer use a pattern rule.
This was tested by modifying system/tools/aidl's Android.mk to refer to
its .aidl files by going up two directories and then repeating the
directories again. When I print the build rules with $(info) I see that
dotdot/ appears in appropriate places (C++ paths, but not .aidl paths).
Bug: 26407018
Test: Described above.
Change-Id: I397c9d10408c0c66d8b5a247a1f34eb4bf4f74ce
This effectively changes the default instruction set of assembly files
from arm to thumb in order to match the default for C/C++.
Change-Id: I8684f144a1195b53b3e0fdd04cacf77f6a131c7e
- For .l/.y source files, generate .c files; for .ll/.yy source files,
generate c++ files.
- Simplified the rules by adding the generated sources to
my_generated_sources.
- Simplified generated header file naming by always using .h extension
with bison's "--defines=" option.
- Removed the unnecesarry conditional inclusion to the generated
headers. Bison already automatically generates such things.
Bug: 26492989
Change-Id: I9ab6dc149c258f7642bc36c3fa32f90ff7ee51a4
- For .l/.y source files, generate .c files; for .ll/.yy source files,
generate c++ files.
- Simplified the rules by adding the generated sources to
my_generated_sources.
- Simplified generated header file naming by always using .h extension
with bison's "--defines=" option.
- Removed the unnecesarry conditional inclusion to the generated
headers. Bison already automatically generates such things.
Bug: 26492989
Change-Id: I9ab6dc149c258f7642bc36c3fa32f90ff7ee51a4
When USE_CLANG_PLATFORM_BUILD is not set, default will be clang/llvm.
USE_CLANG_PLATFORM_BUILD=false can be used to select gcc as default.
Bug: 23163853
Bug: 26102335
Change-Id: I434176732fa4a382be9d8d8642a1c705b023cf84
Host binaries may be run during the build process and the internal
implementation of the shared libraries makes a difference for the build
result. This change makes sure host tools get re-linked and re-run when
any of its dependency libraries gets updated.
DEX2OAT is such a host tool. We also changed DEX2OAT as full dependency
of dex-preoptimization, so we rebuild the odex files if DEX2OAT itself,
or any dependency libraries changed.
Bug: 24597504
Change-Id: Idf0d9be82ccebd826d9c5b405a39cff437e0af29
This was previously working because for some reason prebuilts/ndk had
a tangled mess of hand assembled symlinks that pointed lib -> lib64
for the multilib architectures.
Change-Id: I294d67f58f2008b1a53790cf676f5223df449cbc
When USE_CLANG_PLATFORM_BUILD is not set, default will be clang/llvm.
USE_CLANG_PLATFORM_BUILD=false can be used to select gcc as default.
BUG: 23163853
BUG: 26102335
Change-Id: I00604c2aef4849e8c3505b2c4002eb1c46cd1fd1
Error out if there is a file listed in LOCAL_SRC_FILES_EXCLUDE but not in
LOCAL_SRC_FILES. This should catch typos or other mistakes that would
otherwise be missed.
Change-Id: Iaddf575a6ce35238998ac47b59591a7d05fbcd0d
There is currently an intentional incremental rebuild issue with
import_includes. export_includes might get updated with an identical
version, but we don't want to force everything downstream of it to
rebuild.
When BUILDING_WITH_NINJA==true, only update export_includes if it
changes, and use .KATI_RESTAT to only run downstream rules if it
changes. import_includes will only be updated if one of the
export_includes files is updated, so object files can have a normal
dependency on import_includes instead of an order-only dependency.
All downstream object files will now be recompiled if their imported
include paths change.
Bug: 25910568
Change-Id: I626f3b24ac02ac1309049cf1ce66cfe8ec816513
The export_includes file for a library needs to express a dependency on
all generated exported headers. For aidl generated headers, express a
dependency on the .cpp file instead, since the generator promises to
generate this file last. Unfortunately, the C++ headers generated from
a .aidl file depend on the contents of the file.
Change-Id: I9402b364e4538b502c0958ac8c7bd72cb0add724
It is common for developers to generate/compile AIDL in a static
library, then link that library into an executable. When doing this,
developers need to export the generated headers.
Bug: 25779424
Test: a refactoring of the aidl Android.mk shows this works
Change-Id: I4f7d471a601d2a683cb5a9da5e02e3fab576c26a
When a shared object is rebuilt, all dependent libraries and
executables are rebuilt. Such rebuild is unnecessary when there
is no interface change. With this patch, .toc files will be
generated for all .so files. The rule which generates .toc files
has ninja's restat=1 and .toc files are not changed ninja won't
rebuild dependent targets.
Performance:
$ m && touch bionic/libc/stdio/stdio.c && time m
Before: 1m03s (2563 targets)
After: 21s (90 targets)
Bug: 24597504
Change-Id: Ia5dd950273d143f4e99eee8bef7478f1a94cd138
LOCAL_SRC_FILES_EXCLUDE will be used to filter files out of
LOCAL_SRC_FILES. A common usage will be to use
LOCAL_SRC_FILES_EXCLUDE_<arch> to remove a source file that will be
replaced with an arch-optimized version.
Change-Id: I75cc6114c47fb784bab65cae8f618c4f395f07bb
Ninja has an implicit dependency on the command being run, and kati will
regenerate the ninja manifest if any read makefile changes, so there is no
need to have dependencies on makefiles.
This won't catch all the cases where LOCAL_ADDITIONAL_DEPENDENCIES contains
a .mk file, because a few users of LOCAL_ADDITIONAL_DEPENDENCIES don't
include base_rules.mk, but it will fix the most common ones.
Bug: 23566977
Change-Id: I66de882421376303ab7233c8ce7274548f6b2199
* commit '782b98eaa1c02d935b338f7317fef139067291bb':
Revert "Default to hiding libgcc symbols in each object."
Revert "Don't apply --exclude-libs for the host."
The Mac linker doesn't support this flag, and we don't actually need
it there anyway because we link dynamically to the system's compiler
runtime lib.
Bug: http://b/24166967
Change-Id: I62a926ed39d9fc487638e0c1a172762503dd633e
libchrome uses .mm (Objective-C++) files to bridge C++ code with
OS X Frameworks. This adds support for compiling .mm to .o by just
using the existing C++ support.
Bug: 24168923
Change-Id: Ia65357e2e2584dfffcb6796e214fe6b27635c3a6
shamu checkbuilds set USE_CLANG_PLATFORM_BUILD, which shouldn't apply to
modules built for windows. Also fix some flags that were being set
improperly.
Bug: 23566667
Change-Id: Id4c5b7cc59966328483d90f2b7be3f35e439ecee
Instead of using recursive make to change the HOST_OS when building the
windows SDK under linux, add the concept of cross-building to another
host os.
Bug: 23566667
Change-Id: I6dc525b601b6251d458d197c30bf4660d7485502
So that we can support building both linux and windows binaries at the
same time on a linux host. This replaces the ifeq($(HOST_OS),...) checks
in Android.mk files.
Bug: 23566667
Change-Id: I693e11984e36d55bb6f09fa0d49bc485463e16fb
Work around gyp's inability to handle compound extensions by expecting
a similar looking simple extension for XML files definine DBus
interfaces. We'll need to rename these sources in the places we're
using them already.
Bug: 23380180
Change-Id: Ieb2050f3ef05456cd70de65c3e128d57a6a508f8
Enable daemons exposing an interface over DBus to easily
build client libraries. Now daemons can write rules like:
include $(CLEAR_VARS)
LOCAL_MODULE := libdbus-binding-example-client
LOCAL_DBUS_PROXY_PREFIX := dbus-example-example
LOCAL_SRC_FILES := \
dbus_bindings/org.chromium.Example.Manager.dbus.xml \
dbus_bindings/dbus-service-config.json
include $(BUILD_SHARED_LIBRARY)
to expose a client library.
While here, add support for generating independent adaptor header
files on a per interface basis.
Bug: 22608897
Change-Id: I011f9afc234811c31e445898321c2731c482fa77
Apparently -w will disable all warnings on GCC regardless of ordering
(clang will still respect ordering so warnings that are enabled after
-w are still respected). This is insane. Strip -w from the cflags.
Anyone that wants this flag should be turning off the specific
warnings (or just fix them), not disabling all warnings.
Change-Id: I2ba065637dfdc192921da4d9adbdc63b728c166f
This also drops the NDK default back to C++98 (or C++11 for code using
libc++). The platform NDK build should match the normal NDK build.
Bug: http://b/23043421
Change-Id: I3a336767ce271e84f4dfdebdadb3a98e5689def9
Its presence requires #include directives to contain the build target
name, which is problematic because these directives can live in headers
that are shared by multiple build targets. Furthermore, having
LOCAL_MODULE in the generated header path is redundant because the
target directory is already private to the current build target (e.g.
.../<target_name>_intermediates/...).
Bug: 22608897
Change-Id: I059f71a1231e80f89c99441794a4491f2685036f
With this patch, we can now write Android makefiles like:
include $(CLEAR_VARS)
LOCAL_MODULE := dbus-binding-example
LOCAL_SRC_FILES := main.cpp \
dbus-service-config.json \
org.example.Daemon.Command.dbus.xml \
org.example.Daemon.Manager.dbus.xml
include $(BUILD_EXECUTABLE)
This will cause header files defining native DBus interfaces
to be generated. These can be included from main.cpp to
easily expose object oriented interface over DBus.
Bug: 22608897
Change-Id: Ic4304ac8de77de74d6955ed17789e5477be9a53e
- Don't overwrite [TARGET|HOST]_[CC|CXX] with the [CC|CXX]_WRAPPER prefix,
so that we can disable the wrapper per module.
- Disable ccache on a module when FDO is enabled.
Bug: 22612634
Change-Id: Ibc04a4742d589955066c7eceb43a0da9a2b893bc
(cherry-pick from commit c671a7cf5c)
- Don't overwrite [TARGET|HOST]_[CC|CXX] with the [CC|CXX]_WRAPPER prefix,
so that we can disable the wrapper per module.
- Disable ccache on a module when FDO is enabled.
Bug: 22612634
Change-Id: Ibc04a4742d589955066c7eceb43a0da9a2b893bc
Another change in bionic/linker adds linker_asan/linker_asan64 that
know where to find ASan shared libraries.
Also, include linker_asan to the required packages list when building
for ASan.
Change-Id: I8ebe7c0091bbeb0c135708a891d33d9844373d37
Clang is really aggressive at optimizing a handful of cases (read:
clang will ruin your day some if you write bad code). Fortunately, it
also emits a warning when it's about to do this.
To prevent anyone from suffering from these optimizations, make these
warnings errors and make them impossible to disable.
Change-Id: I5e10bb0fc2ca23190017da716b3b84635577a0bd
Clang will sometimes generate this call (dex2oat with ubsan is one
known case), and it doesn't exist in libgcc.
Change-Id: I2eb68e2a326eb0407dca03b5870077eeebca1c0a
"LOCAL_FDO_SUPPORT := always" enables FDO without user specifying
"BUILD_FDO_OPTIMIZE := true", i.e. it turns on FDO for a
module in any build configuration.
Change-Id: I05d8db2edb2b3f5db073fa14d5bf1083a04571c0
(cherry picked from commit 45d0143ab1)
There will be two version of the the nanopb-c library,
libnanopb-c-2.8.0 which doesn't support automatic malloc
and libnanopb-c-2.8.0-enable_malloc which does.
There will be two version of the the nanopb-c library,
libnanopb-c-2.8.0 which doesn't support automatic malloc
and libnanopb-c-2.8.0-enable_malloc which does.
Set LOCAL_PROTO_OPTIMIZE_TYPE=nanopb-c which doesn't support
malloc and set it to nanopb-c-enable_malloc which does.
For client code details see nanopb-api:
http://koti.kapsi.fi/jpa/nanopb/docs/reference.html
Change-Id: If238412463aabb5e1d556dfc9c464bcaf9e3333a
Previously when a file in LOCAL_SRC_FILES starts with "../", the object
file may escape out of the module's intermediate directory, because we
insert the source file's path (but not with LOCAL_PATH) to the object
file's path. Even worse when two object files escape to the same destination
and cause conflict.
This change fixes the issue by removing the "../" inside the object
files' paths. To do that, we have to set up the compilation rules for
those files one by one, instead of using the one-for-all static
pattern rules.
Bug: 19641115
Change-Id: I19f3c48ece3244fa14acb2caa609deea710840d3
- Removed unnecessary dependency of
"$(my_symlink) : $(LOCAL_INSTALLED_MODULE)"
We can generate symlink to nonexistent file.
Actually in multilib build $(LOCAL_INSTALLED_MODULE) points to file
that may not be the target file of the symlink and leads to always
obsolete $(my_symlink) in the above dependnecy.
- Touch by-product in the dummy rule, to make sure the by-product is
newer than the main-product.
Change-Id: I2f0e0cc197c49f920fa1f6794083b21cdc333c20
Note that this doesn't play nicely with acov out of the box. Clang
apparently generates .gcno files that aren't compatible with gcov-4.8.
This can be solved by installing gcc-4.6 and invoking lcov with
`--gcov-tool /usr/bin/gcov-4.6`.
http://stackoverflow.com/questions/17758126/clang-code-coverage-invalid-output
Change-Id: I79547e1c579fa79db47ff07d5e90c42cedbd5cbb
Don't remember why I didn't enable this for the host when I made the
first pass, but it works just fine.
Change-Id: I0892c0bc353bf8b60b432ba9f69f97281177d41d
So the build system regenerates import_includes when you modify
Android.mk to add a new dependency library.
Change-Id: Ic92b097b659bb68a9065e1d66da59e0dc7e2836a
The proto handling will modify the set of dependent libraries, but
this was not actually accounted for in dependency handling because
dependencies had already been established.
Change-Id: Iba1582f3c9eeeada19569e4b5358b6ec4168fccc
We had discussed the idea of making all host tools default to using
ASAN. Even if we don't make it the default, this makes it easy for the
user to switch all host binaries over.
Change-Id: I64a5c741b1b4e9aefed3a6be8dcd4f386e06b29c
Pass -fsanitize=address instead of manually specifying asan libraries
and other linker flags.
Note that we enable LOCAL_ALLOW_UNDEFINED_SYMBOLS by default for host
builds because ASAN only links symbols in the final executable, so
there will _always_ be undefined symbols in intermediate libraries.
Bug: 18208352
Change-Id: Ief55ab296e94974560eeb10507ec8d90f0025d5c
This will be necessary to support -std=gnu99 mode for clang 3.6, which
defaults to C11 mode (unlike prior releases that use C99).
Change-Id: Iea84582f9f12ba76b988463cbc0a20bd61042538
This feature is now available in AOSP, but not for any shipped
release. We don't have an API version for the release that this will
be available in yet, so for now the check is commented out.
Bug: 18395015
Change-Id: I247233d047ed5a7564d6602d47c9ad962313c8dc
The previous position of libgcc.a/libatomic.a on the link line causes
the linker to prefer satisfying dependencies from these libraries from
other libraries that might include them, rather than from libgcc.a (or
libatomic.a) itself. This imposes an ABI requirement that those
intermediate shared libraries _always_ export those symbols, which is
undesirable.
Change-Id: Ib593236b475d3e98356b2b1be6f96cee2b67378f
This should obviate much of the need for cleanspecs, and also make it
unnecessary to continue adding LOCAL_ADDITIONAL_DEPENDENCIES for this
sort of thing all over the tree.
Change-Id: I97aa8fd280ae868a5f6364f8b7bf3c2fe235d6ce
The NDK protobuf library depends on the final target linking stlport
(since it is a static library). Since the platform stlport is going
away, we need to use a separate version of the protobuf library that
is compiled for the platform against libc++.
Note that this should be the case for _all_ libraries built with the
NDK. If a library needs to be used by both an NDK built final target
and a platform built final target, there should be both an NDK and
platform version of the library.
Bug: 15193147
Change-Id: I0ead61c2d1cd9d0248b304ab7d8682dedd6e8366
"LOCAL_FDO_SUPPORT := always" enables FDO without user specifying
"BUILD_FDO_OPTIMIZE := true", i.e. it turns on FDO for a
module in any build configuration.
Change-Id: I05d8db2edb2b3f5db073fa14d5bf1083a04571c0
Only sort the list of shared libraries used for naming dependencies,
not the order they are actually linked in. The order in which shared
libraries appear to the linker affects which symbols get used if there
is a multiply defined symbol.
Also link system shared libraries _after_ user provided libraries,
since a user will want their functions to override the system's if
they exist.
Change-Id: I071059d940d40a648d69d90e0699073ef520138a
If a module is explicitly depending on a versioned protolib, we strip
the dependency and log a warning so the unneeded dependency can be
removed.
Change-Id: I949d32fb5126f1c05e2a6ed48f6636a4a9b15a48
Because LOCAL_CXX_STL modifies a module's required shared libaries,
we need this for also prebuilt shared libraries and executables.
Change-Id: I418c26143999a613c40aadf990f131b123e0ac3d
my_compiler_dependencies was never assigned to, but the way it was
included in the rules prevented the user from being able to use | in
LOCAL_ADDITIONAL_DEPENDENCIES. Since it is unneeded, just remove it.
Change-Id: I74bb59e81b97756296060eea5b7a42909be50130
These aren't needed now that we only use the compiler/headers that exist in
the prebuilts/clang directory.
Change-Id: I9978efb10815e92577d45629db324e0a5094f880
To enable building with coverage, the environment variable
NATIVE_COVERAGE must be set to true.
Set `LOCAL_NATIVE_COVERAGE := true` to generate coverage information for
a given component.
This is currently not supported for clang (b/17574078, b/17583330).
If static library A is included in a binary B (dynamic or static
executable, or shared library), and A is built with coverage
information, B is required to link with libgcov.a. Since the make does
not offer a good way to track this dependency, link libgcov.a even if
LOCAL_NATIVE_COVERAGE is not set (but still guarded by NATIVE_COVERAGE).
This ensures that all of the libgcov dependencies will always be
resolved, and causes no change in the resulting binary if coverage is
not used.
Bug: 10134489
Change-Id: Id5a19f2c215e4be80e6eae27ecc19b582f2f6813
Christopher Wiley