Previously we call __sinit() lazily. But it is likely to cause data
races like in https://android-review.googlesource.com/#/c/183237/. So
we prefer to call __sinit() explicitly at libc initialization.
Bug: 25392375
Change-Id: I181ea7a4b2e4c7350b45f2e6c86886ea023e80b8
Currently we use __thread variable to store thread_local_dtors,
which makes tsan test fork_atexit.cc hang. The problem is as below:
The main thread creates a worker thread, the worker thread calls
pthread_exit() -> __cxa_thread_finalize() -> __emutls_get_address()
-> pthread_once(emutls_init) -> emutls_init().
Then the main thread calls fork(), the child process cals
exit() -> __cxa_thread_finalize() -> __emutls_get_address()
-> pthread_once(emutls_init).
So the child process is waiting for pthread_once(emutls_init)
to finish which will never occur.
It might be the test's fault because POSIX standard says if a
multi-threaded process calls fork(), the new process may only
execute async-signal-safe operations until exec functions are
called. And exit() is not async-signal-safe. But we can make
bionic more reliable by not using __thread in
__cxa_thread_finalize().
Bug: 25392375
Change-Id: Ife403dd7379dad8ddf1859c348c1c0adea07afb3
After ToolchainCflags were added to soong crt no longer needs to try to
recreate the bare minimum cflags for compiling for each architecture.
Also always use GCC to match crt.mk
Change-Id: I01a833ab70d989033c84f072e3660d060189688c
The anonymous namespace is introduced to
handle cases when linker can not find the
caller. This usually happens when caller
code was not loaded by dynamic linker;
for example mono-generated code.
Bug: http://b/25844435
Bug: http://b/22548808
Change-Id: I9e5b1d23c1c75bc78548d68e79216a6a943a33cf
Use blueprint's new anonymous embedded struct feature to move the arch
variant properties down a level, replacing arch.cortex_a9.srcs with
arch.arm.cortex_a9.srcs, while still supporting top-level properties
like arch.arm.srcs.
Change-Id: Ib41c80e3549440d5efdfd293a15cffa3f51a0fe4
It is reported by tsan that funlockfile() can unlock an unlocked mutex.
It happens when printf() is called before fopen() or other stdio stuff.
As FLOCKFILE(fp) is called before __sinit(), _stdio_handles_locking is false,
and _FLOCK(fp) will not be locked. But then cantwrite(fp) in __vfprintf()
calls__sinit(), which makes _stdio_handles_locking become true, and
FUNLOCKFILE(fp) unlocks _FLOCK(fp).
Change _stdio_handles_locking into _caller_handles_locking,
so __sinit() won't change its value. Add test due to my previous fault.
Bug: 25392375
Change-Id: I483e3c3cdb28da65e62f1fd9615bf58c5403b4dd
* Many processes, including adbd, failed to start in mips/mips64
emulator when libc.so was compiled by clang.
BUG: 25291096
Change-Id: If3434ebdca4a3a6bf6102b120ee838a7ab66cd74
Currently is_private_anonymous is calculated as true if _either_
MAP_PRIVATE or MAP_ANONYMOUS is set, which is a mistake.
According to Documentation/vm/ksm.txt, "KSM only merges anonymous
(private) pages, never pagecache (file) pages". MAP_PRIVATE can
still be set on file cache pages so in order to not redundantly
set MADV_MERGEABLE on pages that are not fitted for it, both
MAP_PRIVATE and MAP_ANONYMOUS should be set.
Along with this fix, add an extra check that the mapped page is
not a stack page before setting MADV_MERGEABLE for it. Stack pages
change too quickly and always end up in KSM 'page_volatile' list.
Change-Id: If4954142852f17cc61f02985ea1cb625a7f3dec6
If tsan is used, the following callchain can happen:
__libc_preinit() -> __libc_init_globals() ->
__libc_init_vdso() -> strcmp() -> __tsan_init()
-> sysconf(_SC_PAGE_SIZE) -> getauxval().
But __libc_auxv is initialized in __libc_init_common(),
after __libc_init_globals(). One simple way to fix
this is to initialize __libc_auxv at __libc_init_globals().
Bug: 25392375
Change-Id: I3893b1f567d5f3b7a8c881c0c1b8234b06b7751b
This removes another way to obtain objects larger than PTRDIFF_MAX. The
only known remaining hole is now jemalloc's merging of virtual memory
spans.
Technically this could be wrapped in an __LP64__ ifndef since it can't
occur on 64-bit due to the 1:1 split. It doesn't really matter either
way.
Change-Id: Iab2af242b775bc98a59421994d87aca0433215bd
In order to run tsan unit tests, we need to support pthread spin APIs.
Bug: 18623621
Bug: 25392375
Change-Id: Icbb4a74e72e467824b3715982a01600031868e29
If calling pthread_mutex_trylock from pthread_mutex_destroy, tsan
warns about an attempt to destroy a locked mutex.
Bug: 25392375
Change-Id: I5feee20e7a0d0915adad24da874ec1ccce241381
We should fall back to GMT if neither the environment variable nor the
system property is set. This is the case if you wipe a WiFi-only device,
because we currently only take the time zone from cell networks.
Bug: http://b/24773112
Change-Id: I90d236d4d492b6562d75021bd312030b91c1e298
Upstream tzcode said "On platforms with tm_zone, strftime.c now assumes it
is not NULL". Which is fine for any struct tm generated by tzcode, but not
necessarily true of a struct tm constructed by arbitrary code. In particular,
Netflix on Nexus Player was failing to start because they format "%Z" with
a struct tm whose tm_zone is null (the other fields are valid, but, yeah,
that's probably not intentional).
glibc takes a null tm_zone to mean "the current time zone", so let's do that
too. (Historically Android would use the empty string, and POSIX doesn't
clarify which of this is the appropriate behavior when tm_zone is null.)
Bug: http://b/25170306
Change-Id: Idbf68bfe90d143aca7dada8607742905188b1d33
When __cxa_thread_atexit_impl.cpp is compiled with clang on arm64,
Android cannot boot up.
BUG: 25662915
Change-Id: If997b544f43e956172ce605d86fe147d42fd39e0
New 3.8 clang/llvm can compile TLS code now.
* For x86_64, still disable clang due to f128 bug.
* For b/25643775, arm and arm64, disable clang in unit tests.
* Fix thread_local_test.cpp to compile with clang and
limit gcc workaround only to arm and aarch64.
BUG: 25643775
Change-Id: Iecd006bf1fc417dbcce2c63343a59c4bf1fa77ea
The current comment implies that we only strip sensitive
environment variables on executing a setuid program. This is
true but incomplete. The AT_SECURE flag is set whenever a
security transition occurs, such as executing a setuid program,
SELinux security transition, executing a file with file capabilities,
etc...
Fixup the comments.
Change-Id: I30a73992adfde14d6e5f642b3a1ead2ee56726be
The mremap definition was incorrect (unsigned long instead of int) and
it was missing the optional new_address parameter.
Change-Id: Ib9d0675aaa098c21617cedc9b2b8cf267be3aec4
Including glibc's <libgen.h> will result in the user getting the POSIX
version of basename always, regardless of when it is included relative
to <string.h>. Prior to this patch, our implementation would result in
the one that's included first winning.
Bug: http://b/25459151
Change-Id: Id4aaf1670dad317d6bbc05763a84ee87596e8e59
The original comment implied that Android 32 bit was the only one
using a bad time_t, but it turns out 32 bit glibc has the same time_t
as Android. Update the comment to reflect this reality.
Change-Id: I72cdd2c677a6521f7b43d4695a52fd8fdc3305bd
_signal was static in 64 bit, and hidden on 32 bit. There is no
reason to have this distinction, so make it hidden in all cases.
Change-Id: I09d5d93ac8cab4fe14dc7bdfeb25aa46a3b7413d
Android doesn't allow SysV IPC, so this only serves two purposes. It's
good in that it helps random code (especially code that doesn't actually
need this header) to compile anyway. But it's bad for the same reason:
it's misleading for configure scripts that test for the header but not
for the corresponding functions.
Bug: http://b/19340456
Change-Id: I4a5f319d9aecf46fc512fc94bb4f9e3f6ea64e71
These don't work, aren't thread-safe, aren't in POSIX (or our header
files), and are only used by one app (whose developers I've contacted).
But the presence of these symbols causes configure to be confused, which
is a pain for Brillo.
Bug: http://b/24812426
Change-Id: I7fa6ef82864d5563929d9b8a7f8fcacb30b26d45
