Bionic libs are part of the runtime APEX (com.android.runtime). In order
to be able to update the runtime APEX independetly from the platform, we
have to prevent things outside of the APEX from using bionic symbols
that are not guaranteed to be stable. Otherwise, platform could break
when a symbol is removed from the libs via the APEX update.
To achive this goal, this change adds stubs variant to the bionic libs.
With this, things outside of the runtime APEX (i.e. other APEXes and the
platform) are built with the stubs variants that provide only the
symbols that are guaranteed to be stable.
The set of symbols are basically the same as the symbols available to
NDK clients. However, there are a few additional symbols that are not
available for NDK but should be made available for platform components.
They are marked with "# apex" tag. Symbols with that tag are not exposed
to apps (via NDK stubs) or vendors (via LLNDK stubs).
Note that the stubs is a build-time only artifact. It is used just to
break the build when private symbols are used outside of the runtime
APEX. At runtime, the real library in the APEX is used.
Bug: 120266448
Test: m
Test: m bionic-unit-tests
Change-Id: I7b8d75830c81d7d7d54e2fca21a85b3417531b47
These modules have issues running with execute-only memory. Disable it
in them until we can resolve the issues.
Bug: 77958880
Test: No more XOM-related crashes in these binaries
Change-Id: Ie6c957731155566c2bbe7dbb7a91b9583d9aff93
This allows compiler optimize both calls to a single frinti.
Test: Builds, ran unit tests on arm64.
Change-Id: Ie59ac11a419281836a8051ecda26a5c7b72f3dea
This option removes mostly of the PLT calls for internal symbol, they
are routed to internal entrypoint instead. It should be safe for libm,
since symbol interposing is not supported.
For arm64 the remaining PLTs are __cxa_finalize, ldexp, __stack_chk_fail,
isnanf, memset, and isxdigit. All of them are expected (clang (r339432)
does not have optimization for isnan/isinf inline expansion and these
symbols are defined on libc.so).
Test: Builds, ran unit tests on arm and x86.
Change-Id: I1eb1a9f2b30f21788d9ad8639b4a0a15f6d0566a
This option allows inline calls of single intructions if target,
supports it. Currently it optimizes internal sqrt calls on
armeabi-v7a, arm64-v8a, and x86_64 (by removing internal calls
to sqrt).
Test: Builds, ran unit tests on arm and x86.
Change-Id: I8a7621a55aae82ff5b80e886c9ed49d480ef6263
This patch ues exp, exp2, log, log2, and pow from arm optimized
routines. For pow on x86_64, although slight slower it simplifies
the code required on both bionic and arm-optimized-routines (so
there is no need to select and export the symbol based on
architecture).
Performance-wise the improvements are:
x86_64 throughput latency
exp 1.16x 1.16x
log 1.08x 0.95x
exp2 1.27x 1.55x
log2 1.40x 1.43x
pow 0.77x 0.89x *
* I tried to check if AVX2/FMA but without success.
aarch64 throughput latency
exp 2.33x 2.16x
exp2 1.99x 1.50x
log 1.79x 1.43x
log2 2.15x 1.80x
pow 3.81x 3.07x
Test: ran bionic tests on static mode.
Change-Id: Ib16bf3280c5329fd257a3b3f0b6c4f2f3cb34deb
This reverts commit 253a830631 and moves
us forward to a revision that contains fixes for the problem with the
previous attempt.
This also makes sincos(3)/sincosf(3)/sincosl(3) available to `_BSD_SOURCE`
as well as `_GNU_SOURCE`.
The new FreeBSD libm code requires the FreeBSD `__CONCAT` macro, and all
our existing callers are FreeBSD too, so update that.
There's also an assumption that <complex.h> drags in <math.h> which isn't
true for us, so work around that with `-include` in the makefile. This
then causes clang to recognize a bug -- returning from a void function --
in our fake (LP32) sincosl(3), so fix that too.
Bug: http://b/111710419
Change-Id: I84703ad844f8afde6ec6b11604ab3c096ccb62c3
Test: ran tests
This includes an ld128 powl, plus the clog* and cpow* families.
Also teach the NOTICE generator to strip SPDX-License-Identifier lines.
Bug: N/A
Test: ran tests
Change-Id: Ic8289d1253666a19468a4088884cf7540f1ec66d
Global flag --pack-dyn-relocs=android is used with clang lld.
For b/24465209, we need to override that with --pack-dyn-relocs=none.
Bug: 80093890
Bug: 24465209
Test: build with USE_CLANG_LLD=true and run dlext.compat_elf_hash_and_relocation_tables
Change-Id: Ic3837446ff0deccf4e66425b39d3b062abcacc5e
See longer explanation in b/80093890.
Clang lld does not generate expected DT_REL and DT_RELA tags
with --hash-style=both and --pack-dyn-relocs=android.
I am not sure about the extent of b/24465209, so
I would rather not to use lld for these .so files for now.
Bug: 80093890
Bug: 24465209
Test: build with USE_CLANG_LLD=true and run dlext.compat_elf_hash_and_relocation_tables
Change-Id: I94e9fe8d687daeadb0182ae26596ca11d3f8bd9b
Libraries that are direct or indirect dependencies of modules installed
to recovery partition (e.g. toybox) are marked as recovery_available:
true. This allows a recovery variant of the lib is created when it is
depended by other recovery or recovery_available modules.
Bug: 67916654
Bug: 64960723
Bug: 63673171
Bug: 29921292
Test: m -j
Change-Id: I59bf859a10a218af6591025a0fe7d1853e328405
We can turn on -Wall too, if we turn off const-unused-variable and
unused-variable.
Bug: N/A
Test: builds
Change-Id: I8d889ea0acd334fbacbf7e707bcc2d0be633e4d5
This is a squash revert of a4a4854 and 1b0f2b4
Revert "Remove a test for backwards compatibility we no longer support."
This reverts commit a4a485454a.
# This is the commit message #2:
Revert "Remove obsolete workaround."
This reverts commit 1b0f2b49d5.
Bug: 72143978
Bug: 24465209
Test: Tested failing case on sailfish, reverted back all CL's since
Test: 3471433 for b/24465209 and apps open
These are still needed for backwards compatibility with code built by old
versions of the NDK, but we don't need to pollute the headers with them.
Also lose the hand-written code for these. The compiler-generated code
is either the same or better, and no new code is calling these functions
anyway.
Bug: N/A
Test: ran tests
Change-Id: Ib01ad9805034433e0105aec882608cc8e6526f78
We can cut a lot of stuff out of the NDK's libandroid_support with this,
and reduce unnecessary relocations for all LP32 code. LP64 code should
be unaffected.
Bug: https://issuetracker.google.com/64450768
Bug: https://github.com/android-ndk/ndk/issues/507
Test: ran tests, plus manual readelf on the _test.o files
Change-Id: I3de6015921195304ea9c829ef31665cd34664066
Since ARM neon instructions were only used on armv7-a-neon architecture
variant, the default implementation for 32-bit armv8-a cores doesn't
use these advanced SIMD instructions. By using "neon" key in the
Android.bp, both arch variants (armv7-a-neon and armv8-a) are covered.
Bug: 65569003
Test: lunch aosp_arm64; emulator # on oc-mr1-dev; boot to home screen
Also checked if sqrt.o and floor.o are actually built and linked
in the resulted libm.a for both armv7-a-neon/cortex-a15 and
armv8-a/generic (2nd) arch/cpu variant.
Change-Id: I2084dbdb12e252b06bba5adc65adb59e97a99332
Some of this code is used in the NDK libandroid_support now, as a static
library, so just being HIDDEN in the ELF sense isn't sufficient.
Rename digittoint to __libm_digittoint so we don't trample anyone's toes.
Also remove imprecise_powl and imprecise_tgammal. It turns out (to my
surprise) that we don't even have ld128 implementations of powl and tgammal,
so even LP64 was just using the "fake_long_double.c" hack in effect. Since
that's the case, let's *actually* do that because then we're not polluting
with the internal names in addition to the aliases.
Bug: N/A
Test: readelf
Change-Id: I273cc8fdc7ce53f9b8dfd4ef7796e358fe901837
Explicitly disables the integer_overflow sanitizer in the libm library.
This change should not impact any other build targets.
Bug: 30969751
Test: Built with and without SANITIZE_TARGET=integer_overflow
Change-Id: I98d0312344086b01394f456d4b953463b5e6cb1a
Soong handles these automatically now.
Bug: 33241851
Test: Android-aosp_arm.mk is the same before/after
Test: build.ninja is the same before/after
Test: build-aosp_arm.ninja is the same before/after
Merged-In: Ibfe4cf5db732d9007b1cc12dbc523c427de1b23d
Change-Id: Ibfe4cf5db732d9007b1cc12dbc523c427de1b23d
The stub libraries are the same as the (equivalent version) NDK, but
they also contain symbols tagged with 'vndk'. Except there are none of
those in Bionic currently.
For headers, the LLNDK/VNDK isn't using a sysroot like the NDK. Nor is
it hardcoding an explicit source path list like the platform. Instead it
runs the bionic/libc/include directory through versioner like the NDK,
then exports those generated headers and the kernel headers from the
stub library like any other exported header. Except it uses -isystem
instead of -I due to export_headers_as_system.
Test: aosp_arm; m -j
Test: Enable BOARD_VNDK_VERSION on aosp_arm; m -j
Test: Inspect out/soong/build.ninja before/after (w/o vndk)
Change-Id: Ief58a73907a83053b408b1d4b62999cba470d61c
Disables debuggerd integration unless building for android.
Bug: 31559095
Test: Diff out/soong/build.ninja before/after, only change is moving
linker's libdebuggerd_client static lib to the beginning of the
list.
Test: lunch aosp_arm64-eng; mmma -j bionic
Change-Id: I62e725f7a9b98b7fe31637d0a835fd5846b0aff0
Just expose the ones that bionic historically leaked.
Also, many of the M_* constants in <math.h> are actually POSIX.
Change-Id: I6275df84c5866b872b71f1c8ed14e2aada12b793
* Remove --exclude-libs ldflags (e29e99c51)
* Mips strlen check was inverted
* ARM specific libc changes only apply to the shared library
* Reorder libjemalloc and malloc_common to match Make's ordering (for
better binary diffing)
Change-Id: Ib41bd47565b1e8fadd285daa7b337cff1edd0808
