This implementation simply iterates over each static TLS module and
copies its initialization image into a new thread's static TLS block.
Bug: http://b/78026329
Test: bionic unit tests
Change-Id: Ib7edb665271a07010bc68e306feb5df422f2f9e6
Replace reserve_tcb with reserve_exe_segment_and_tcb, which lays out both
the TCB and the executable's TLS segment, accounting for the difference in
layout between variant 1 and variant 2 targets.
The function isn't actually called with a non-null TlsSegment* yet.
Bug: http://b/78026329
Test: bionic unit tests
Change-Id: Ibd6238577423a7d0451f36da7e64912046959796
* changes:
Add a __bionic_get_tls_segment function
Factor out ScopedRWLock into its own header
Build the linker with -D_USING_LIBCXX
Provide a stub aeabi.read_tp on other archs
Remove TLS_SLOT_TSAN(8)
This reverts commit 220f51e566.
The internal modules that were using extra symbols are all fixed.
Bug: 120266448
Test: m ndk_translation_all in cf_x86_phone
Change-Id: I561b16de1c320d2624e7cf8e6211e0c70edc823d
The function searches for a TLS segment in a ElfXX_Phdr table.
Bug: http://b/78026329
Test: bionic unit tests
Change-Id: I221b13420d1a2da33fc2174b7dd256589f6ecfdb
This macro is necessary for using stdatomic.h and <atomic> in the same
libc++-based program. <atomic> is used implicitly by other STL headers,
and stdatomic.h is used (or will be used) in headers shared with libc.
libc only has access to stdatomic.h because it can't use an STL.
Bug: http://b/78026329
Test: bionic unit tests
Change-Id: I89b128df08b478287ef4e2867319e74dbcc30d4e
Using "#if defined(...)" inside the test body seems to be the convention.
Bug: none
Test: bionic-unit-tests
Change-Id: Icb348a1184e630c5e199129bbe2090fd1e490fa5
As of the switch to clang-r346389c, it has been replaced with
TLS_SLOT_SANITIZER(6). lld reserves 8 words beyond the TP on arm/arm64, so
Bionic can't use anything beyond 7.
The DTV and bionic_tls slots on x86 haven't been part of a release yet,
and they should be strictly internal to Bionic anyway, so shift them down.
Bug: http://b/78026329
Test: bionic unit tests
Change-Id: Ifb3b2d8d85efe1417ee9a10b657b665ec6f2fd3d
This includes one manual change:
In the file bionic/libc/kernel/uapi/linux/in.h, the macro IN_BADCLASS
was not definied correctly. Change the macro from:
#define IN_BADCLASS(a) ((((long int) (a)) == 0xffffffff)
to:
#define IN_BADCLASS(a) (((long int) (a)) == (long int)0xffffffff)
This change is being pushed to the upstream kernels.
Test: Builds and boots.
Change-Id: Ia304773a9dc6789b34d9769d73742384d6afb571
Merged-In: Ia304773a9dc6789b34d9769d73742384d6afb571
(cherry picked from commit 967fb01cce)
By sorting symbols by size, small symbols are grouped together and we
usually have less dirty pages at runtime. On cuttlefish, this results
in 20KB less dirty pages just after libc is loaded.
Bug: 112073665
Test: Build libc and check symbol ordering.
Test: Compare runtime private dirty memory usage on cuttlefish.
Change-Id: Ic8fa996f81adb5a8cbc4b97817d2b94ef0697a2a
For ELF TLS "local-exec" accesses, the static linker assumes that an
executable's TLS segment is located at a statically-known offset from the
thread pointer (i.e. "variant 1" for ARM and "variant 2" for x86).
Because these layouts are incompatible, Bionic generally needs to allocate
its TLS slots differently between different architectures.
To allow per-architecture TLS slots:
- Replace the TLS_SLOT_xxx enumerators with macros. New ARM slots are
generally negative, while new x86 slots are generally positive.
- Define a bionic_tcb struct that provides two things:
- a void* raw_slots_storage[BIONIC_TLS_SLOTS] field
- an inline accessor function: void*& tls_slot(size_t tpindex);
For ELF TLS, it's necessary to allocate a temporary TCB (i.e. TLS slots),
because the runtime linker doesn't know how large the static TLS area is
until after it has loaded all of the initial solibs.
To accommodate Golang, it's necessary to allocate the pthread keys at a
fixed, small, positive offset from the thread pointer.
This CL moves the pthread keys into bionic_tls, then allocates a single
mapping per thread that looks like so:
- stack guard
- stack [omitted for main thread and with pthread_attr_setstack]
- static TLS:
- bionic_tcb [exec TLS will either precede or succeed the TCB]
- bionic_tls [prefixed by the pthread keys]
- [solib TLS segments will be placed here]
- guard page
As before, if the new mapping includes a stack, the pthread_internal_t
is allocated on it.
At startup, Bionic allocates a temporary bionic_tcb object on the stack,
then allocates a temporary bionic_tls object using mmap. This mmap is
delayed because the linker can't currently call async_safe_fatal() before
relocating itself.
Later, Bionic allocates a stack-less thread mapping for the main thread,
and copies slots from the temporary TCB to the new TCB.
(See *::copy_from_bootstrap methods.)
Bug: http://b/78026329
Test: bionic unit tests
Test: verify that a Golang app still works
Test: verify that a Golang app crashes if bionic_{tls,tcb} are swapped
Merged-In: I6543063752f4ec8ef6dc9c7f2a06ce2a18fc5af3
Change-Id: I6543063752f4ec8ef6dc9c7f2a06ce2a18fc5af3
(cherry picked from commit 1e660b70da)
This change makes it easier to move the location of the pthread keys
(e.g. into the bionic_tls struct).
Bug: http://b/78026329
Test: bionic unit tests
Test: disassembly of libc.so doesn't change
Merged-In: Ib75d9dab8726de96856af91ec3daa2c5cdbc2178
Change-Id: Ib75d9dab8726de96856af91ec3daa2c5cdbc2178
(cherry picked from commit ecad24fad9)
The only 2019 android copyright is coming from a test file, which is
not included in our normal notice.
Test: Can upload.
Change-Id: Ie22332f927b99c28eb71c1daf5615adfef8c5d11
The current version has these bugs:
- Adding a semicolon after a function results in the removal of structures
following the function.
- Function like macros get removed on accident rather than on purpose.
- It removes extern "C" { completely, which might not be a bug, but doesn't
seem right.
I couldn't easily fix any of these problems because the code depends heavily
on the header being correct.
New unit tests added for the function to cover all of these cases.
A follow-on CL will include the updated headers.
Bug: 112290385
Test: Passes all new unit tests.
Test: When run on the current kernel headers, the generated headers are
Test: nearly the same, missing data is being added.
Change-Id: Ib22a5f2e78873544e8a9d54e385af1156b2a72bb
Track movement of ICU .dat file into the runtime module
This reverts commit b9e7b699a2.
This is a reapply of commit 68c6bd8b43
Bug: 120853401
Bug: 119293618
Test: build / boot
Change-Id: I69e0dde2756da64ac0b17f1726145dccacf5961f
PR_SET_VMA takes mmap_sem, which can cause contention and reduce
performance any time many threads are created at the same time,
like app startup.
Test: camera launch performance
Bug: 122471935
Change-Id: If7fa7ad99654c01d503f694976fd92bfd30d2afd
By default, Clang uses this arm32 function to read the thread pointer,
either for ELF TLS or via __builtin_thread_pointer(). It's probably better
to inline the cp15 access using -mtp=cp15, but that's not the default yet.
See https://reviews.llvm.org/D34878?id=114573.
Bug: http://b/78026329
Test: bionic unit tests
Change-Id: I93b8926075f0b2cea8df9ef518d54f2820a8ff5b
The "if defined(__ANDROID__)" else branch is there to support
bionic when running on host environments.
The code now checks the
${ANDROID_RUNTIME_ROOT}/etc/tz/tzdata path as well.
This is similar to the current on-device expectations.
Eventually, we may remove the /system / ANDROID_ROOT cases.
Bug: 119293618
Test: build only
Change-Id: I8310f7b5a1c14567a648673970c636c7c84ff818
When parsing a dir.${section} line in ld.config.txt, the linker uses
realpath() to resolve each directory path. If realpath() fails, the linker
uses the non-resolved path instead. Previously, it issued a warning to
stderr and logcat, but realpath() is expected to fail on /postinstall when
running a binary in /data/local/tmp, so downgrade the warning to an INFO()
notice.
Bug: http://b/120996057
Test: copy ping to /data/local/tmp, run it, verify no errors
Test: bionic unit tests
Change-Id: I2104084e12ea18840230997408559299bf3bddac
Following additional symbols are temporarily exposed from the runtime APEX
to satisfy the dependencies from other non-AOSP modules. This will be
reverted when the modules are built with the boostrap bionic libraries.
Bug: 120266448
Test: m
Change-Id: I39ac40a40545723ea2aeb185b9cd293d511f802f
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
Having a constexpr constructor should guarantee that the static
`globals` variable in __libc_shared_globals is initialized statically
(as opposed to dynamically), which is important because
__libc_shared_globals is called very early (before the linker has
relocated itself). With the constructor, though, the fields can safely
have in-line default initializers.
Bug: none
Test: bionic unit tests
Change-Id: Icde821557369625734a4d85d7ff55428bad5c247
We could cause illegal memory access due to not checking the scanddir
return value.
Test: run linux-bionic dex2oat to build boot.oat
Bug: 119332362
Change-Id: Ic5824e0373b9b99d8f892b4fcf22e742c6f792da
* Add explicit to conversion constructors/operators
Bug: 28341362
Test: make with WITH_TIDY=1 DEFAULT_GLOBAL_TIDY_CHECKS=-*,google-explicit-constructor
Change-Id: Id1ad0327c1b8c6f094bcbb3ae599bc1f716b3f2f
The previous test would fail to crash if the first byte of the canary was
already NUL (1 of 256 runs assuming uniform canary distribution).
Bug: http://b/121156651
Test: bionic unit tests
