Its target arch is goldmont, but without xsaves support.
It can build images that run on old Intel/AMD CPUs as well as new CPUs.
Bug: 314243939
Test: Build an image with TARGET_ARCH_VARIANT:=goldmont-without-xsaves
Change-Id: I9c4bfd3928511f5b4a075f9a9337db6e1270a765
These makefiles are actually architecture-independent (as far as I can
tell, the Darwin one doesn't have anything arch-specific and there's
just one of them anyway), so let's remove the architecture name from the
names of these files to make it easier to support new host architectures.
Bug: 294584591
Change-Id: Ic4cb0bf393c53ac9593afb7c2c6a6e07fbe6079e
Add barebones files to support building for a riscv64 target.
Bug: 250918230
Test: lunch aosp_riscv64-userdebug && m ALLOW_MISSING_DEPENDENCIES=true nothing
Change-Id: I61e6ecf6706b7682c77b07e6593d78c7d1eaef39
llvm-ar bundled in the next llvm update will stop accepting the legacy
single-dash argument format.
Bug: 219872355
Test: presubmit
Change-Id: Iede68e4b13c3a01c176d26f335f96879c5c72c12
Soong will use this to turn on universal binary support (X86_64 + Arm64
in the same binary).
Bug: 203607969
Test: m sdk-repo-platform-tools sdk-repo-build-tools on Mac
Change-Id: I04612136a42e85f4add95202ce20e741d9aaa302
This empty file is required to be present.
Test: fvp_mini-eng* builds and boots on FVP
*: with TARGET_ARCH_VARIANT := armv8-a-branchprot
Change-Id: I689033f16663822e263f45ab5e46c9f120c9dca0
This patch introduces the .mk file required to support build for the new
armv8-2a-dotprod Arm64 arch variant. The file just needs to be present
but does not require to contain anything.
Test: test-art-target on Pixel 4.
Change-Id: I4b6a1e306f9819cf31cd556351bd3e8cf08e2537
Previously, HOST_CROSS_OS/ARCH were fixed to windows/x86. This change
makes the setting configuration and adds the support for new OS/ARCH
combo: linux_bionic/arm64.
linux_bionic is the Linux-based host target that uses Bionic (instead of
glibc) as libc. Previously, it supported only x86_64 and the x86_64
target was NOT configured via Make, but directly via editing
soong.variables file. Now, the support for arm64 is being added in the
Soong side and this change makes it possible to configure the target via
Make.
The new HOST_CROSS_OS/ARCH combo will be used for building the host-side
tools (adb, crosvm, etc.) for running Cuttlefish natively on Linux/ARM
hosts.
Bug: 159685774
Test: HOST_CROSS_OS=linux_bionic HOST_CROSS_ARCH=arm64 m nothing
Change-Id: I6b8ed8f7e26908749bbe778fbdcc34cfbde68179
The P flag is supported as of llvm r354044, add back the flag to reduce
the size of intermediate archieve files. This does not affect the final
binaries.
Test: build
Bug: 71618641
Change-Id: I4021aec783ba341768893392b73b976c1fa5bdf7
JVM occasionally (0.5% time) crashes on entering a method with SIGSEGV.
It might be due to excessive load on the host machine. Reduce the number
of the compiler threads to 6, and do not start all the GC threads
immediately. If JVM crashes, copy full diagnostics to stderr.
Bug: 132766811
Test: treehugger
Change-Id: Ic1b09e0f2cbcf4c22214eaaeae464899b2b99ef7
This reverts commit 6c77a6baa6.
Reason for revert: aog/974065 fixes the problem that caused the revert
Change-Id: I935f78762b23ac63a79a9529515ff4ef394d5c3c
This reverts commit 09fd82ea89.
Reason for revert: This change set the value of JAVA make variable to a string containing "foo bar". When soong_ui is asked to evaluate such string, it drops the quotes, causing bash to treat 'foo' and 'bar' as separate command line arguments.
Change-Id: I81e07c4cd333751d185dc50c49d0638f6511727e
It was previously a warning to set an incorrect ARCH_VARIANT
when CPU_VARIANT was set to a cpu for which we knew the arch.
Make it an error, as there are no offenders in the tree.
Re-structure the code to make it a little clearer when you get
a default and when you get an error.
Test: build_test
Change-Id: I5821a8162a59e8c2aed2370068f9eb656ced9d99
Stop using armv7-a without neon for 32-bit unbundled apps, and update
generic_x86_arm to match the cuttlefish configuration that includes
neon.
Test: treehugger
Change-Id: Ieb6a2106655803a8ca609907c12168e628ee1b85
Newer cores are implementing armv8-2a ISAs.
Enabling 2nd arch. variant to support for new type of cores.
Test: set TARGET_2ND_ARCH_VARIANT := armv8-2a, build without warnings and not ignore armv8-2a
BUG: 118414869
Change-Id: I1cd64ab0ad9b253ec3d109ebd1dbc7882011ce77
Mac OSX 10.14 Mojave deprecates 32-bit executables, and the 10.14 SDK
does not successfully link 32-bit files.
Test: Install Xcode 10 beta, `m checkbuild host`
Change-Id: Ifad471fb88ead4bda870f49837d937fb5742f3f6
This properly sets -format=gnu for all non-Darwin targets, because
llvm-ar is cross-platform (but defaults to the host environment).
Bug: http://b/71618641
Test: m checkbuild
Change-Id: Ib5d1c112a4ac792a11d1a0721ae6b0a5f9a42727
This architecture only existed for unbundled use, but even the NDK is
removing support in their r17 release, so just remove support for it.
Test: build/soong/build_test.bash -only-config
Test: check buildserver configs, I don't see anything still using armv5
Change-Id: Ic183b510c9ada94438bd4cc2b9362fa438a29ced
llvm-ar is required for platform LTO build.
With GNU ar, we used crsPD flags, however 'P' flag is not supported by
llvm-ar. Since none of the platform archive files are built using third
party AR, the 'P' flag should not matter to us. However we've been
relying on one side-effect of 'P', where 'ar foo.a some/path/to/obj.o'
when foo.a already has a file named 'obj.o' will be an append operation,
regardless whether the existing one is the same (ar has no way of
telling the difference). We workaround this by always appending ('q'
flag) for llvm-ar. This may result in larger intermediate archive files
(more duplication) but will not affect the final build result.
Bug: 71618641
Test: m checkbuild
Change-Id: Id96a244cfe49cecfba08cc868dd18934ecb4ff23
The Java PATH fixup is now handled within soong_ui, along with the
values of ANDROID_JAVA_HOME based on OVERRIDE_ANDROID_JAVA_HOME /
EXPERIMENTAL_USE_OPENJDK9.
Mark PATH as deprecated, so that any reads/writes will cause warnings.
This will be switched to obsolete once it's verified that there are no
more users.
Using PATH within Kati means that we've got to rebuild the ninja files
whenever your PATH changes, which is not ideal, especially since some of
the envsetup functions can change your PATH. In most cases you only need
to use PATH within the bash portions of the build rules ($${PATH}), which
isn't treated as a make variable, so won't produce an error.
I'm also planning on replacing PATH in a future change with our own
directory that has placeholders for everything in your PATH. This will
let us remove tools that shouldn't be used from the build.
Test: m nothing
Test: build/soong/build_test.bash on AOSP and internal master
Change-Id: I18d8d19cfba313ff9176345bf73ac34e8dbebfbb
Default to ANDROID_COMPILE_WITH_JACK=false when TARGET_BUILD_APPS
is set.
Test: m -j TARGET_BUILD_APPS=Gallery2
Change-Id: Ibd2bdac6c7e2c4e2c6107fd35bd049fc8cb8dafe
Use TARGET_JAVAC and HOST_JAVAC exported from soong.
Test: m -j checkbuild
Merged-In: Ic094174149ea304c0b06ddc023417acf3602d5a3
Change-Id: Ic094174149ea304c0b06ddc023417acf3602d5a3
(cherry picked from commit 204dd14a2c)
This reverts commit 7de79cbeaf.
Fixed CTS build issue by adding -march=armv7-a in cts/tests/tests/os/jni/Android.mk since the library built here needs/uses instruction (swp) obsolete in ARMv8.
Bug: 64964151
Test: lunch aosp_arm64-userdebug; make -j cts
lunch aosp_x86-userdebug; make -j cts
lunch aosp_sailfish-userdebug; make -j cts
lunch aosp_bullhead-userdebug; make -j cts
Change-Id: Ic3a121600da8e2a9a2d5176b9680cd18ce457959
