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
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
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
Overwrite TARGET_(2ND_)ARCH_VARIANT as armv8-a if
TARGET_(2ND_)CPU_VARIANT is set to some known armv8-a core like
cortex-a53, cortex-a73, kryo, denver64 etc.
For clang, -march is ignored if -mcpu is set to specific core so this
change doen't impact the objects built for armv7-a-neon/some_armv8_core
since it's treated as armv8-a/some_armv8_core.
Bug: 62895439
Test: Built modified aosp_arm64 with armv8-a as its TARGET_2ND_CPU_ARCH
and generic as its TARGET_2ND_CPU_VARIANT successfully.
"lunch hikey-userdebug; make" and verify TARGET_2ND_ARCH_VARIANT
is overwritten.
Change-Id: If4260cf397783b5f56c4fd432615f4676429a1d3
These are all either recently unused due to the removal of CFLAGS/etc,
or have been exported by Soong and are no longer necessary.
Change-Id: I5930d43fda21acc8202b3d8ea010fbefb6ae4cf1
-Wl,--fix-cortex-a8 is a workaround for an Erratum in Cortex-A8
processors. It slightly increases code size and decreases performance,
and there's no point in using it on non-A8 CPUs.
Instead of forcing it unconditionally, use it when targeting
Cortex-A8 or generic armv7-a (which might or might not be A8).
Change-Id: Ifa59765d380445237edccfe5440a67b3ba1e459a
Signed-off-by: Bernhard Rosenkränzer <Bernhard.Rosenkranzer@linaro.org>
LPAE indicates better instructions can be used when atomicity guarantees are
needed. However, LPAE's presence isn't advertised by clang/GCC. We fake an
ARM feature to advertise its presence on architectures where it is.
Also, add a TODO documenting that cortex-a15 is not the correct CPU variant
for krait.
Change-Id: I02a1248025c32d94eca0bc8a249dc524f1ac9c36
Users of ARCH_ARM_* defines don't care about first vs. second arch,
set ARCH_ARM_* regardless of which arch is arm.
Change-Id: I2ae83ec5c3f839ff91a0e352c95d76ec2cbd5dc5
This is the first step to build 32-bit libraries in a 64-bit product.
It will work like this:
1) In the product's BoardConfig.mk, define:
TARGET_2ND_ARCH, TARGET_2ND_ARCH_VARIANT, TARGET_2ND_CPU_VARIANT.
The build system uses those variables to set up an additional compiler
environment for the second arch.
2) When parsing Android.mks, the build system sets up rules to build a
module for both the 1st arch and the 2nd arch, unless it's explicitly
asked to skip so.
Android.mk will be adapted if there is additional rule of generating
source files.
The build system will accept arch-specific LOCAL_ variables, such as
LOCAL_CFLAGS_arm, LOCAL_CFLAGS_armv7-a-neon, LOCAL_CFLAGS_cortex-a15,
LOCAL_CFLAGS_aarch64 etc. Modules use such variables to set up build for
various archs at the same time.
3) Install binary of the 2nd arch by adding "<module_name>:32" to
PRODUCT_PACKAGES. All 2nd-arch libraries linked in by "<module_name>:32"
will be installed automatically.
Bug: 11654773
Change-Id: I2df63cd5463a07bf5358bee2a109f8fb9590fe30
Conflicts:
core/combo/TARGET_linux-arm.mk
Everything that is based on ARMv7a should have a cp15 TLS register.
Enable it by default so it's not accidentally missed in newer board
configurations. In fact, this could be enabled for ARMv6 as well, but we
currently don't distinguish between ARMv5 and ARMv6 in the build system.
This can still be disabled by setting it to "false" in the board
configuration, but this shouldn't ever be needed.
Change-Id: Ic2918f32899c8bcfa482f92c98f5a192fa318470
This allows TARGET_ARCH_VARIANT to be set by the vendor before we choose the
architecture in core/combo/select.mk.
Also add a primitive armv7-a.mk for turning on hardware floating point.
