This patch uses the new hardware feature macros for x86 to define
various compile-time macros used to make the C library use
SSE2 and/or SSSE3 optimized memory functions for target CPUs
that support these features.
Note that previously, we relied on the macros being defined by
build/core/combo/TARGET_linux-x86.mk, but this is no longer the
case.
Change-Id: Ieae5ff5284c0c839bc920953fb6b91d2f2633afc
this works by building a directed graph of acquired
pthread mutexes and making sure there are no loops in
that graph.
this feature is enabled with:
setprop debug.libc.pthread 1
when a potential deadlock is detected, a large warning is
output to the log with appropriate back traces.
currently disabled at compile-time. set PTHREAD_DEBUG_ENABLED=1
to enable.
Change-Id: I916eed2319599e8aaf8f229d3f18a8ddbec3aa8a
This patch provides several small optimizations to the
implementation of mutex locking and unlocking. Note that
a following patch will get rid of the global recursion
lock, and provide a few more aggressive changes, I
though it'd be simpler to split this change in two parts.
+ New behaviour: pthread_mutex_lock et al now detect
recursive mutex overflows and will return EAGAIN in
this case, as suggested by POSIX. Before, the counter
would just wrap to 0.
- Remove un-necessary reloads of the mutex value from memory
by storing it in a local variable (mvalue)
- Remove un-necessary reload of the mutex value by passing
the 'shared' local variable to _normal_lock / _normal_unlock
- Remove un-necessary reload of the mutex value by using a
new macro (MUTEX_VALUE_OWNER()) to compare the thread id
for recursive/errorcheck mutexes
- Use a common inlined function to increment the counter
of a recursive mutex. Also do not use the global
recursion lock in this case to speed it up.
Change-Id: I106934ec3a8718f8f852ef547f3f0e9d9435c816
This patch changes the implementation of pthread_once()
to avoid the use of a single global recursive mutex. This
should also slightly speed up the non-common case where
we have to call the init function, or wait for another
thread to finish the call.
Change-Id: I8a93f4386c56fb89b5d0eb716689c2ce43bdcad9
Fix dead loops in file ./bionic/libc/unistd/pathconf.c
Change-Id: I7a1e6bcd9879c96bacfd376b88a1f899793295c8
Author: Jin Wei <wei.a.jin@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
When forking of a new process in bionic, it is critical that it
does not allocate any memory according to the comment in
java_lang_ProcessManager.c:
"Note: We cannot malloc() or free() after this point!
A no-longer-running thread may be holding on to the heap lock, and
an attempt to malloc() or free() would result in deadlock."
However, as fork is using standard lib calls when tracing it a bit,
they might allocate memory, and thus causing the deadlock.
This is a rewrite so that the function cpuacct_add, that fork calls,
will use system calls instead of standard lib calls.
Signed-off-by: christian bejram <christian.bejram@stericsson.com>
Change-Id: Iff22ea6b424ce9f9bf0ac8e9c76593f689e0cc86
Pass kernel space sigset_t size to __rt_sigprocmask to workaround
the miss-match of NSIG/sigset_t definition between kernel and bionic.
Note: Patch originally from Google...
Change-Id: I4840fdc56d0b90d7ce2334250f04a84caffcba2a
Signed-off-by: Chenyang Du <chenyang.du@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Chars are signed for x86 -- correct the comparison semantics.
Change-Id: I2049e98eb063c0b4e83ea973d3fcae49c6817dde
Author: Liubov Dmitrieva <liubov.dmitrieva@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Fix the compile warning to let the libc.debug.malloc=10 works well
Due to unsuitable value comparison, which cause compiler optimize the
code of comparing two digits.
Change-Id: I0bedd596c9ca2ba308fb008da20ecb328d8548f5
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Author: liu chuansheng <chuansheng.liu@intel.com>
(1) in pthread_create:
If the one signal is received before esp is subtracted by 16 and
__thread_entry( ) is called, the stack will be cleared by kernel
when it tries to contruct the signal stack frame. That will cause
that __thread_entry will get a wrong tls pointer from the stack
which leads to the segment fault when trying to access tls content.
(2) in pthread_exit
After pthread_exit called system call unmap(), its stack will be
freed. If one signal is received at that time, there is no stack
available for it.
Fixed by subtracting the child's esp by 16 before the clone system
call and by blocking signal handling before pthread_exit is started.
Author: Jack Ren <jack.ren@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
When running the stress test of pthread create/destroy, a crash may
oocur in __get_tls(). That is caused by the race condition with __set_tls( ):
Author: Jack Ren <jack.ren@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Pull in an updated version of capabilities.h from the linux
kernel.
This file was generated using the following command:
cd bionic/libc/kernel/
./tools/clean_header.py -u ../../../external/kernel-headers/original/linux/capability.h
Change-Id: I43c8f014954f543858006f24e60a2e69955349da
The function must be named __atomic_cmpxchg, not __android_cmpxchg.
This typo broke existing prebuilt binaries (they couldn't be loaded
at runtime anymore).
Change-Id: I25ca7d18329817f0056e616a0409113269ad7b1f
We can't easily tell the protocol family of the secondary network,
so try both and trust that the carrier has configured dns servers
according to the protocols supported on its network.
bug:5468224
Change-Id: If4f017573d313a6ad8354574076de6d63d43b444
Use tgkill instead of tkill to implement pthread_kill.
This is safer in the event that the thread has already terminated
and its id has been reused by a different process.
Change-Id: Ied715e11d7eadeceead79f33db5e2b5722954ac9
Allow the kernel to choose a memory location to put the
thread stack, rather than hard coding 0x10000000
Change-Id: Ib1f37cf0273d4977e8d274fbdab9431ec1b7cb4f
__atomic_cmpxchg and other related atomic operations did not
provide memory barriers, which can be a problem for non-platform
code that links against them when it runs on multi-core devices.
This patch does two things to fix this:
- It modifies the existing implementation of the functions
that are exported by the C library to always provide
full memory barriers. We need to keep them exported by
the C library to prevent breaking existing application
machine code.
- It also modifies <sys/atomics.h> to only export
always-inlined versions of the functions, to ensure that
any application code compiled against the new header will
not rely on the platform version of the functions.
This ensure that said machine code will run properly on
all multi-core devices.
This is based on the GCC built-in sync primitives.
The end result should be only slightly slower than the
previous implementation.
Note that the platform code does not use these functions
at all. A previous patch completely removed their usage in
the pthread and libstdc++ code.
+ rename arch-arm/bionic/atomics_arm.S to futex_arm.S
+ rename arch-x86/bionic/atomics_x86.S to futex_x86.S
+ remove arch-x86/include/sys/atomics.h which already
provided inlined functions to the x86 platform.
Change-Id: I752a594475090cf37fa926bb38209c2175dda539
We're going to modify the __atomic_xxx implementation to provide
full memory barriers, to avoid problems for NDK machine code that
link to these functions.
First step is to remove their usage from our platform code.
We now use inlined versions of the same functions for a slight
performance boost.
+ remove obsolete atomics_x86.c (was never compiled)
NOTE: This improvement was benchmarked on various devices.
Comparing a pthread mutex lock + atomic increment + unlock
we get:
- ARMv7 emulator, running on a 2.4 GHz Xeon:
before: 396 ns after: 288 ns
- x86 emulator in KVM mode on same machine:
before: 27 ns after: 27 ns
- Google Nexus S, in ARMv7 mode (single-core):
before: 82 ns after: 76 ns
- Motorola Xoom, in ARMv7 mode (multi-core):
before: 121 ns after: 120 ns
The code has also been rebuilt in ARMv5TE mode for correctness.
Change-Id: Ic1dc72b173d59b2e7af901dd70d6a72fb2f64b17
For Honeycomb, we added proper file thread-safety for
all FILE* operations. However, we did implement that by
using an out-of-band hash table to map FILE* pointers
to phtread_mutex_t mutexes, because we couldn't change
the size of 'struct _sFILE' without breaking the ABI.
It turns out that our BSD-derived code already has
some support code to extend FILE* objects, so use it
instead. See libc/stdio/fileext.h
This patch gets rid of the hash table, and put the
mutex directly into the sFILE extension.
Change-Id: If1c3fe0a0a89da49c568e9a7560b7827737ff4d0
The old code didn't work because the kernel expects a 64-bit sigset_t
while the one provided by our ABI is only 32-bit. This is originally
due to the fact that the kernel headers themselves define sigset_t
as a 32-bit type when __KERNEL__ is not defined (apparently to cater
to libc5 or some similarly old C library).
We can't modify the size of sigset_t without breaking the NDK ABI,
so instead perform runtime translation during the call.
Change-Id: Ibfdc3cbceaff864af7a05ca193aa050047b4773f
There are three changes of note - most urgently, Cuba (America/Havana)
has extended summer time by two weeks, now to end on Nov 13, rather than
the (already past) Oct 30. Second, the Pridnestrovian Moldavian Republic
(Europe/Tiraspol) decided not to split from the rest of Moldova after
all, and consequently that zone has been removed (again) and reinstated
in the "backward" file as a link to Europe/Chisinau. And third, the
end date for Fiji's summer time this summer was moved forward from the
earlier planned Feb 26, to Jan 22.
Apart from that, Moldova (MD) returns to a single entry in zone.tab
(and the incorrect syntax that was in the 2011m version of that file
is so fixed - it would have been fixed in a different way had this
change not happened - that's the "missing" sccs version id).
Change-Id: I7a0fba88d1fc6face649648013aaf2b111c29d7f
Fixes for Europe/Tiraspol (Moldova) and all four Ukrainian zones.
Also show the MD5 of the downloaded data, for comparison against the MD5
given in the announcement mails. (There's a plan to move to proper signing,
but that's not implemented on their end yet.)
Change-Id: I845e6f125c0f54298abadc643adfeca2eff4827a
Modify the dynamic linker so that executables can be loaded
at locations other than 0x00000000.
Modify crtbegin* so that non-PIC compilant "thumb interwork
veneers" are not created by the linker.
Bug: 5323301
Change-Id: Iece0272e2b708c79034f302c20160e1fe9029588
Without this change strcmp size is zero (not set), and it gets
ignored by Valgrind. Changes to memcpy and atexit don't affect the
generated binary in any way.
Change-Id: I05818cb5951f75901dc8c0eef02807a2e83a9231
AI_ADDRCONFIG is currently implemented by trying to connect
to well-known addresses in order to see if IPv4 and/or IPv6
connectivity is available.
In some cases (e.g., walled gardens with no global
connectivity) both probes can fail. If this happens,
query for both IPv4 and IPv6 addresses instead of doing
nothing and failing the query.
Bug: 5284168
Change-Id: I4e3a69ea86fb6d839a6bd31236b98da81e5cbf45
Change Ife82a8d8 broke IPv6 on wifi. Change I4e3a69ea is
an alternate approach that does not require any framework
changes.
Bug: 5284168
Change-Id: Ib52614be3875a2ae2eaedd1be265f90e506eda41
Instead of checking for IPv4 or IPv6 connectivity, try using pid-specific
hints the framework has left for us.
bug:5284168
Change-Id: Id64d48db3900865a7d58ada8309870c63d6eab12
vfork() would not save the registers that the parent would expect
to have restored after execl() completed.
Specially that execl() would call execve() underneath, further messing
up the stack of the parent.
To avoid that, we fork() for now. Later we will revisit and cleanup
vfork()+execve() to actually have vfork() store all the register
that the parent expects to see, and not those left by execve().
In the original code, looking at the registers just before the call to
popen(), and after the call showed that r7 would get clobbered.
This would leave the caller with an invalid pointer, leading to all
kinds of data corruptions.
execve() is simpler that execl() in this case.
Bug: 5336252
Change-Id: I3bf718c0bb4c0439f6f2753f153cdea14175be9c
This patch ensure that __aeabi_f2uiz is embedded in our C library.
This is needed to avoid breaking certain applications when they are
loaded in ICS. It is likely that the issue is due to mis-linked
binaries generated with the stand-alone toolchain (the problem
should not exist if you use ndk-build), but this fix is easier
than asking all app developers to fix their custom build system.
If you want more technical details, read the comments inside
libgcc_compat.c
Change-Id: I59ac1fc781ecb70b90b5573c5a3c67560ca8f270
This patch changes the declaration of size_t on x86 targets
to test for the __ANDROID__ macro, instead of ANDROID
__ANDROID__ should be a builting toolchain macro, while ANDROID
is usually added manually during the build.
Testing against __ANDROID__ allows us to use the header when
using the NDK's standalone x86 toolchain.
This is related to http://code.google.com/p/android/issues/detail?id=19011
The bug was already fixed in the NDK platform headers, this simply updates
the C library one accordingly.
Change-Id: Ie038c4c8b37b7d24e2e4ae4d7a63371b69c9a51e
This happens too frequently. We should flush a per-interface cache
when it's dns server addrs changes.
Change-Id: I8a691c96ce9a775160ef55ddb8f755d649041583
* commit '6b6ebeca985fb3843b56b507ac4ac1be44080a9c':
enable support for large files (> 2G)
Enable functional DSO object destruction
x86: Enable -fstack-protector
Update X86 Bionic CRT files for unwind/exceptions
bionic, libthread_db x86 fixes
Updated gcc 4.4.3 IA toolchain doesn't require the .ctors list
Remove an extra register move.
Replace __atomic_XXX with GCC __sync_XXX intrinsics.
move some typedefs to procfs.h required by gdbserver build
use consistent guards for off_t and size_t defines for IA
Simplify variable typing for IA builds
sigsetmask.c was not processing the "mask" argument.
Add defines for CAIF support
Remove extra/unneeded copy of fenv.h
Use proper variable typing
Update ATOM string routines to latest
Fix undefined reference to dl_iterate_phdr for x86
Fix missing NL
ptrace.c Fix source file format to unix from dos
* commit '3a13102637c8be53edf28f96598ac11aaa3e14df':
enable support for large files (> 2G)
Enable functional DSO object destruction
x86: Enable -fstack-protector
Update X86 Bionic CRT files for unwind/exceptions
bionic, libthread_db x86 fixes
Updated gcc 4.4.3 IA toolchain doesn't require the .ctors list
Remove an extra register move.
Replace __atomic_XXX with GCC __sync_XXX intrinsics.
move some typedefs to procfs.h required by gdbserver build
use consistent guards for off_t and size_t defines for IA
Simplify variable typing for IA builds
sigsetmask.c was not processing the "mask" argument.
Add defines for CAIF support
Remove extra/unneeded copy of fenv.h
Use proper variable typing
Update ATOM string routines to latest
Fix undefined reference to dl_iterate_phdr for x86
Fix missing NL
ptrace.c Fix source file format to unix from dos
bionic has an i386 ifdef guard for the O_LARGEFILE flag. The reason for this
is not traceable in the git history, but it's most likely due to some old
toolchain or emulator problem.
Bug Id: 10035
Change-Id: I742ee6e50220276c2cb268d66f89b5655ef689fc
Signed-off-by: Dragos Tatulea <dragos.tatulea@intel.com>
Unfortunately, legacy .so files for ARM don't have a correct crtbegin file.
Consequently, we have to grandfather the old __dso_handle behaviour.
Add some ifdefs for ARM to allow it to use the old code until we can work
out a transition.
Change-Id: I6a28f368267d792c94e1d985d8344023bc632f6f
Author: H.J. Lu <hongjiu.lu@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Several updates to Bionic crt*.o files to implement Linux/x86
stack unwind/exception handling.
Mark __dso_handle as hidden.
Note: Requires complete rebuild due to CRT changes. Google will
need to regenerate and check-in all the prebuilt x86 CRT object
files and libraries which link to them.
Change-Id: I01c14c19ef7569c9ffed89495cae2a459972c04b
Orig-Change-Id: I191ce1afb0dee7bbc28dcd9694e5919226dbd070
Signed-off-by: Bruce J Beare <bruce.j.beare@intel.com>
Signed-off-by: Mark D Horn <mark.d.horn@intel.com>
Author: Lu, Hongjiu <hongjiu.lu@intel.com>
Orig-Change-Id: I3be997f5f1f6a894a3c200d4f325cf3bfd428c66
Author: James Rose <james.rose@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Orig-Change-Id: Ia840a19a45257128eccdcf25d105f500f2d90741
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
Orig-Change-Id: I32d2e02814f1965ebdcecb804e947ff303dbe230
Author: Mark Gross <mark.gross@intel.com>
Signed-off-by: Bruce Beare <brucex.j.beare@intel.com>
libcutils/mspace.c includes libc/bionic/dlmalloc.c, we need to
take care of the fact that any internal C library function cannot
be used from it.
Change-Id: I0bc81ae090b7ac2d464f26b97fc6b94a08cdad9c
Our dlmalloc implementation currently calls abort() when it detects
that the heap is corrupted, or that an invalid pointer is passed to
one of its functions.
The only way to detect this is because abort() will force-fully
crash the current program with a magic fault address of '0xdeadbaad'.
However, this is not really well documented, and a frequent topic
on the android-ndk forum (among others).
This change makes our dlmalloc code dump a simple message to the
log just before the abort() call (and hence before the stack trace)
to better help identify the problem.
Change-Id: Iebf7eb7fe26463ecadfaca8f247d237edb441e3c
bionic has an i386 ifdef guard for the O_LARGEFILE flag. The reason for this
is not traceable in the git history, but it's most likely due to some old
toolchain or emulator problem.
Bug Id: 10035
Change-Id: I742ee6e50220276c2cb268d66f89b5655ef689fc
Signed-off-by: Dragos Tatulea <dragos.tatulea@intel.com>
This includes the Egyptian and Russian DST changes.
(Cherry pick of dd0ffc738e3f956cf4f0ae409cda22595a33f08b.)
Change-Id: Ibf4b6a1f004c997a0b8df66e78f8867425d15dbe
The latest merge conflict resolution worked only for ARM, but not for x86
Both ARM and x86 builds have been tested with this fix.
Change-Id: I4ec1f1b0ea32fa86e7e2515b221133473d6550a3
NOTE: Care has been taken to not change the compiler flags for ARM builds.
In particular, when building crtbegin_so.so, the -fPIC flag is only
used for x86 (previous reports indicate that it breaks the ARM
build at runtime for obscure reasons).
Quote from Linux Programmer's Manual:
"If t is non-NULL, the return value is also stored in the memory
pointed to by t."
Change-Id: I8cb66b67e5f34c536ce2f0db76a6dc337c42ea3f
Signed-off-by: Jim Huang <jserv@0xlab.org>
Unfortunately, legacy .so files for ARM don't have a correct crtbegin file.
Consequently, we have to grandfather the old __dso_handle behaviour.
Add some ifdefs for ARM to allow it to use the old code until we can work
out a transition.
Change-Id: I6a28f368267d792c94e1d985d8344023bc632f6f
Author: H.J. Lu <hongjiu.lu@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>
The spec says "A value in the range between -4095 to -1 indicates an error" (not -129).
This was pointed out in the comment in 22039.
Change-Id: I11b7c45015a9e4ccf09aed5364a889437eab6ab8
Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
Initial commit for dns cache per interface.
Added a type that holds a reference to a
cache and name of associated interface,
address of interface, name server(s)
associated with an interface etc.
New functions to set default interface,
address of name servers etc.
Change-Id: Ie991bc5592fd998409853d8bf77d7fe69035dac5
Make netinet/ip.h have IPTOS_MINCOST which matches what
kernel/common/linux/ip.h has to say.
Fixup gethostbyaddr() 1st arg to match what man has to say.
Change-Id: Iff9647d4a46ea88b1fc32163eb69bb9b27cdf370
These functions were already declared in <pwd.h>, but hadn't been implemented
yet.
git cherry-pick --no-commit 081504af74
Change-Id: I316acf4cffb9f2c6788e8e342aa620f9a00886d5
This change fixes the prctl() declaration to match GLibc, and allow us
to build gcc-4.6 for Android (among other things). Note that this does
not change the source and binary interfaces.
Change-Id: I3900c957f0da8b3548cbe9da8c41d50099dc19d6
Several updates to Bionic crt*.o files to implement Linux/x86
stack unwind/exception handling.
Mark __dso_handle as hidden.
Note: Requires complete rebuild due to CRT changes. Google will
need to regenerate and check-in all the prebuilt x86 CRT object
files and libraries which link to them.
Change-Id: I191ce1afb0dee7bbc28dcd9694e5919226dbd070
Signed-off-by: Bruce J Beare <bruce.j.beare@intel.com>
Signed-off-by: Mark D Horn <mark.d.horn@intel.com>
Author: Lu, Hongjiu <hongjiu.lu@intel.com>
A suggestion how to make a smarter delete function when the cache
is full. First look through the entire cache and remove all entries
which have expired. If none use the old solution and just remove
the last entry in the MRU list.
Change-Id: I5f997ab35290a55dc6e1ddf37d725759edf83d36
Make netinet/ip.h have IPTOS_MINCOST which matches what
kernel/common/linux/ip.h has to say.
Fixup gethostbyaddr() 1st arg to match what man has to say.
Change-Id: Iff9647d4a46ea88b1fc32163eb69bb9b27cdf370
NOTE: This is a back-port from the internal HC branch.
This patch fixes a leak that occurs when creating a new
thread-specific DNS resolver state object.
Essentially, each thread that calls gethostbyname() or getaddrinfo()
at least once will leak a small memory block. Another leak happens
anytime these functions are called after a change of the network
settings.
The leak is insignificant and hard to notice on typical programs.
However, netd tends to create one new thread for each DNS request
it processes, and quickly grows in size after a > 20 hours.
The same problem is seen in other system processes that tend to
create one thread per request too.
The leak occured becasue res_ninit() was called twice when creating
a new thread-specific DNS resolver state in _res_get_thread().
This function could not properly reset an existing thread and was
leaking a memory block.
The patch does two things:
- First, it fixes res_ninit() to prevent any leakage when resetting
the state of a given res_state instance.
- Second, it modifies the _res_get_thread() implementation to
make it more explicit, and avoid calling res_ninit() twice
in a row on first-time creation.
Fix for Bug 4089945, and Bug 4090857
Change-Id: Icde1d4d1dfb9383efdbf38d0658ba915be77942e
scanf()'s man page suggests support for %Ld, (and ioux).
Implement this so that 3rd party code will work correctly.
Change-Id: Idce9d266071cb688ca71429395a2d9edf6813595
The gethostbyaddr code in system/netd now expects a string address
from inet_ntop, not raw bytes, in order to properly pass addresses
containing null and probably spaces and newlines characeters as well.
Bug: 4344448
Change-Id: I8ec0eab94d5b1d38e9269ba3afb2825e946f8df3
The gethostbyaddr code in system/netd now expects a string address
from inet_ntop, not raw bytes, in order to properly pass addresses
containing null and probably spaces and newlines characeters as well.
Bug: 4344448
git cherry-pick -e c13fa640ee
The gethostbyaddr code in system/netd now expects a string address
from inet_ntop, not raw bytes, in order to properly pass addresses
containing null and probably spaces and newlines characeters as well.
Bug: 4344448
Change-Id: I935abbbe522d96b64a5f975c7937e3aed3f7b335
Reference results of the experiments on TI OMAP3430 at 600 MHz
$ bench_strcmp -N "strcmp_1k" -s 1k -I 200
[original C code]
prc thr usecs/call samples errors cnt/samp size
strcmp_1k 1 1 10.38000 102 0 15000 1024
[ARM optimized code]
prc thr usecs/call samples errors cnt/samp size
strcmp_1k 1 1 3.08840 88 0 15000 1024
The work was derived from ARM Ltd, contributed to newlib, and reworked
for Android by Linaro.
Change-Id: Ib0d5755e1eb9adb07d80ef0252f57a5c4c57a425
Signed-off-by: Jim Huang <jserv@0xlab.org>
According to Intel ABI spec, there is no need to save %eax, %ecx, %edx
on the stack. Worse, popping %eax will wipe out the return value.
Change-Id: Ida536c3b98174b8deef88f8f3b9352eaaeb7c0c0
The kernel has supported this syscall for quite some time now,
but bionic did not. Now that there is a need for it, let's
add it to bionic.
Change-Id: Ifcef3e46f1438d79435b600c4e6063857ab16903
This patch fixes a leak that occurs when creating a new
thread-specific DNS resolver state object.
Essentially, each thread that calls gethostbyname() or getaddrinfo()
at least once will leak a small memory block. Another leak happens
anytime these functions are called after a change of the network
settings.
The leak is insignificant and hard to notice on typical programs.
However, netd tends to create one new thread for each DNS request
it processes, and quickly grows in size after a > 20 hours.
The same problem is seen in other system processes that tend to
create one thread per request too.
The leak occured becasue res_ninit() was called twice when creating
a new thread-specific DNS resolver state in _res_get_thread().
This function could not properly reset an existing thread and was
leaking a memory block.
The patch does two things:
- First, it fixes res_ninit() to prevent any leakage when resetting
the state of a given res_state instance.
- Second, it modifies the _res_get_thread() implementation to
make it more explicit, and avoid calling res_ninit() twice
in a row on first-time creation.
Fix for Bug 4089945, and Bug 4090857
Change-Id: Ie4831a8dbe82be8f07fce5ddd1d36bf95994f836
* commit '38e53adaddb474d41d1b6282eec38b71b8447d5d':
Use sigemptyset to initialise sigset_t values. sigset_t may not be implemented as an integral type.
* commit 'c59b248241b18492d20e112230a907c14400275b':
Use sigemptyset to initialise sigset_t values. sigset_t may not be implemented as an integral type.
* commit '5a6f593b6197b96f980d9d45f9a981ef8705caf4':
Use sigemptyset to initialise sigset_t values. sigset_t may not be implemented as an integral type.
Backport I110b653a58f3
All the other property stuff is already here. Property setting was
only in libcutils previously to leverage a utility function / constant
or two.
Unfortunately in the process of fixing a race condition we would've
had to do break abstraction boundaries and put some libc-internal
details into libcutils so instead of that we'll just move this
into bionic.
Along with Iee1ca9b7, this now passes:
$ adb shell am instrument -w -e class android.os.SystemPropertiesTest \
com.android.frameworks.coretests.systemproperties/android.test.InstrumentationTestRunner
Bug: 3511230
Change-Id: I1b588db3344169621e1279ecc0b660cf4e1015d7
Backport I110b653a58f3
All the other property stuff is already here. Property setting was
only in libcutils previously to leverage a utility function / constant
or two.
Unfortunately in the process of fixing a race condition we would've
had to do break abstraction boundaries and put some libc-internal
details into libcutils so instead of that we'll just move this
into bionic.
Along with Iee1ca9b7, this now passes:
$ adb shell am instrument -w -e class android.os.SystemPropertiesTest \
com.android.frameworks.coretests.systemproperties/android.test.InstrumentationTestRunner
Bug: 3511230
Change-Id: I1b588db3344169621e1279ecc0b660cf4e1015d7
All the other property stuff is already here. Property setting was
only in libcutils previously to leverage a utility function / constant
or two.
Unfortunately in the process of fixing a race condition we would've
had to do break abstraction boundaries and put some libc-internal
details into libcutils so instead of that we'll just move this
into bionic.
Along with Iee1ca9b7, this now passes:
$ adb shell am instrument -w -e class android.os.SystemPropertiesTest \
com.android.frameworks.coretests.systemproperties/android.test.InstrumentationTestRunner
Bug: 3511230
Change-Id: I110b653a58f312fbe069dca59892a877ae9bc911
sigset_t may not be implemented as an integral type.
Change-Id: I3f5ec86d6383685979d8d23ef508b6c609569b38
Signed-off-by: Raghu Gandham <raghu@mips.com>
* commit '8d1f2e6ffc7230e4000e5352e5798095a1f6d428':
libc: <stdint.h>: Don't make __STD_CONSTANT_MACROS available if only __STD_LIMIT_MACROS is defined.
* commit '638608b11afbd255ddbd15e739f0f638791a4305':
libc: <stdint.h>: Don't make __STD_CONSTANT_MACROS available if only __STD_LIMIT_MACROS is defined.
* commit 'da8ea213abb8bec08da37622179061630bd8e2f8':
libc: <stdint.h>: Don't make __STD_CONSTANT_MACROS available if only __STD_LIMIT_MACROS is defined.
This patch prevents the definition of various macros when <stdint.h> is
included from C++. The ISO C99 standard mentions that when this header
is included from a C++ source file, limit and constant related macros
should only be defined when asked explicitely by defining
__STDC_LIMIT_MACROS and __STD_CONSTANT_MACROS, respectively.
The <stdint.h> lacked the proper #ifdef .. #endif blocks for the
following macros:
INTPTR_MIN, INTPTR_MAX, UINTPTR_MAX, PTRDIFF_MIN, PTRDIFF_MAX
INTMAX_MIN, INTMAX_MAX, UINTMAX_MAX,
INPTR_C, UINTPR_C, PTRDIFF_C, INTMAX_C, UINTMAX_C
This is intended to fix http://code.google.com/p/android/issues/detail?id=14380
after we copy this file to development/ndk/platforms/android-3/include/
Change-Id: Ia77e0822edfaaf568ea599d7de673b310eeeaa4a
* commit '81d79f9fb38d9ae8a5711f5774aa5a1516f6cc26':
libc: Fix PTHREAD_RWLOCK_INITIALIZER
Proxy getnameinfo through netd
Updated gcc 4.4.3 IA toolchain doesn't require the .ctors list
Convert cname lenght before use
* commit '962dcb22218a1a6d4ebd05e4fc4a69875d037234':
libc: Fix PTHREAD_RWLOCK_INITIALIZER
Proxy getnameinfo through netd
Updated gcc 4.4.3 IA toolchain doesn't require the .ctors list
Convert cname lenght before use
* commit 'fed58049d50083e4c960d89f4bbd516a181c2d18':
libc: Fix PTHREAD_RWLOCK_INITIALIZER
Proxy getnameinfo through netd
Updated gcc 4.4.3 IA toolchain doesn't require the .ctors list
Convert cname lenght before use
The length of the cname is sent in big-endian
order. Thus, it has to be converted before used
in android_getaddrinfo_proxy
Change-Id: I1a0cc12780c47f7493fcf06f690515829f88c01e
This also incorporates a bunch of changes to the previous script; this one
requires no setup, can be run from anywhere, and leaves no droppings.
Change-Id: I38f299f03e33950d2a64e9336f4ba7cb3c5cf6f0
Used the new gensyscalls.py script to update the ARM syscalls with the
BEGIN(x) and END(x) macros to give size information for the code of the
syscall. Useful for valgrind.
Change-Id: I8c481c8928401ac110fd19b087f7d67e4db39326
Add size information to gensyscalls output for ARM syscalls with new
BEGIN(x) and END(x) macros in arch-arm/include/machine/asm.h
Change-Id: I791406f8b17abcb83b70a6d15a65a527de15d3f5
Add a macro to annotate function end and start using both ENTRY and END
for each function. This allows valgrind (and presumably other debugging
tools) to use the debug symbols to trace the functions.
Change-Id: I5f09cef8e22fb356eb6f5cee952b031e567599b6
The length of the cname is sent in big-endian
order. Thus, it has to be converted before used
in android_getaddrinfo_proxy
Change-Id: I1a0cc12780c47f7493fcf06f690515829f88c01e
Fix the handle locking in stdio to use flockfile/funlockfile
internally when and where required. Macros in <stdio.h> are updated
to automatically call the underlying functions when the process is
threaded to obtain the necessary locking. A private mutex is added
to protect __sglue, the internal list of FILE handles, and another
to protect the one-time initialization. Some routines in libc that
use getc() change to use getc_unlocked() as they're either protected
by their own lock or aren't thread-safe routines anyway.
Based on OpenBSD change by guenther@openbsd.orghttp://www.mail-archive.com/source-changes@cvs.openbsd.org/msg01015.html
Bug: 3446659
Change-Id: Ie82116e358c541718d6709ec45ca6796be5a007b
Use system property ro.net.dns_cache_size to set
the size of the cache. If the system property
is not set the default cache size is defined by
CONFIG_MAX_ENTRIES.
The number of entries in the hash table will be equal
to the number of max entries allowed in the cache.
Change-Id: I77d69d7c178937fa483d0b40512483ad29232d28
Added a memmove() variant for Dalvik's System.arraycopy()
implementation. It guarantees 16-bit or 32-bit atomicity depending
on the alignment of the arguments.
Bug 3398352
(cherry-pick from master)
Change-Id: I5e64de1c76971ba234bf2dd05477ba390e0412a4
Added a memmove() variant for Dalvik's System.arraycopy()
implementation. It guarantees 16-bit or 32-bit atomicity depending
on the alignment of the arguments.
Bug 3398352
Change-Id: Ie7bd246305ef0ff8290513663327c5b81680368d
Make the scripts use external/kernel-headers/original by default.
clean_header.py: Document -k<path>, add -d<path>
find_headers.py: Make kernel config files optional
update_all.py: Allow setting the path to kernel headers on the command-line
update_all.py: Better formatting of output on ttys
update_all.py: Automatically perform "git add/rm" on affected files.
SYSCALLS.TXT: Fix typo in __socketcall definition.
checksyscalls.py: Add support for superH architecture in the checks.
gensyscalls.py: Automatically perform "git add/rm" on affected files.
cpp.py: Fixed a bug that prevented certain type definitions to
be kept in the generated clean header (e.g.
struct ethtool_drvinfo in <linux/ethtool.h>)
All scripts will use the content of external/kernel-headers/original by default now.
The generated code removes all empty lines and trailing whitespace. This is useful
to ensure a unified output even if we change the parser again in the future.
The top-level disclaimer has been edited with update instructions to regenerate
the headers when needed.
Also, a warning is now inserted every 8th line in the final output:
/* WARNING: DO NOT EDIT, AUTO-GENERATED CODE - SEE TOP FOR INSTRUCTIONS */
Changes under kernel/arch-arm and kernel/arch-x86 should correspond to whitespace
differences and additionnal struct definitions that were missed by the previous
parser implementation.
Change-Id: Icd1c056bacd766759f3e9b7bb5d63a246f3d656a
WARNING: If you run these script, do not submit the result to gerrit for now.
It seems there are discrepancies between the content of original headers
and those currently commited under bionic/libc/kernel/.
(This problem is the main motivation to insert the warning repeatedly).
Current list of issues:
- Missing SuperH headers (i.e. external/kernel-headers/original/asm-sh)
On ARM there is currently no assembler optimized memmove in libc.
There is however a more optimized bcopy which copies long instead
of bytes where possible. This almost doubles the performance in
best case.
Change-Id: I1f1cd27529443358047c385730deaf938ce4e642
* changes:
Add icmp6.h and ip6.h files from current NetBSD libc.
Include the Linux ipv6.h header from netinet/in.h, to get things like in6_pktinfo.
Add IN6ADDR_LOOPBACK_INIT macro to <netinet/in6.h>, as the kernel headers do not provide it anymore.
Update IPv6 kernel headers with new 2.6.37 headers, cleaned by update_all.py. (It wanted to do many other changes, most of them cosmetic, but I only included the IPv6 changes.)
Fix syntax error in defaults.py.
Use the the TTL of the answer as the time a query
shall remain in the resolver cache.
Added some debugging support as well, i.e.
parse answer and print a la dig.
Change-Id: I724d3392245032592f1912f3ca7a81a8987ebbac
This patch moves the definition of struct user_vfp from <asm/ptrace.h>
to <asm/user.h> for the ARM platform, in order to match the original
kernel headers.
See patches 94519 and 94520 for a description of the problem and a
fix for debuggerd that uses the structure.
This prevented us from properly re-running the kernel headers cleanup
scripts.
Change-Id: I43718a59ae49b9d89fc2d1c8affb65973fb2919c
getaddrinfo only asks DNS for IPv6 addresses if the system
has IPv6 connectivity, but always asks for IPv4 addresses.
Don't ask for IPv4 addresses if there is no IPv4
connectivity.
Change-Id: Iefe9fcb006fabe60b4b11dd4653a7c4a406506f4
to gingerbread.
Implement RFC3484 policy table changes from draft-ietf-6man-rfc3484-revise-01.
The changes in a nutshell:
- Handle v4-mapped as different from v4-compat (this was probably
an existing bug in our code).
- Add policy entries for ULA, above most everything else.
- Put v4-compat, old-style IPv6 site-local and 6bone addresses
way down in the preference table.
The rest is just shuffling numbers around (no actual changes to
priority).
to gingerbread.
Add a new #define IN6_IS_ADDR_ULA, for testing for Universal Local IPv6 Unicast addresses (ULAs). These replace the old site-local IPv6 addresses.
to gingerbread.
Don't treat private IPv4 addresses as being in a non-global scope. The effect of this change is essentially to prefer NATed IPv4 over 6to4.
The changes in a nutshell:
- Handle v4-mapped as different from v4-compat (this was probably
an existing bug in our code).
- Add policy entries for ULA, above most everything else.
- Put v4-compat, old-style IPv6 site-local and 6bone addresses
way down in the preference table.
The rest is just shuffling numbers around (no actual changes to
priority).
Change-Id: Ieb111e0edd705992bc9d066befaaee95a8c1a5df
It adds two functions sigsetjmp and siglongjmp
that fix compiling errors in bluetooth and mksh.
Copy directly from sigsetjmp.S of OpenBSD.
Change-Id: I4696f82ee6f85d1c93cbdd3c9e40f4917d50f3a6
The source file is not part of the C library build, so the Android.mk
is unaffected. In other words, this source file was never compiled.
Change-Id: Idec3d5b6ec30dc9ee38296d12dc6e522997df29a
After this change, SIGRTMAX will be set to 64 (instead of 32 currently).
Note that this doesn't change the fact that our sigset_t is still defined
as a 32-bit unsigned integer, so most functions that deal with this type
won't support real-time signals though.
Change-Id: Ie1e2f97d646f1664f05a0ac9cac4a43278c3cfa8
Plus a hack in cpu-features.h to avoid errors like this:
bionic/libc/arch-arm/bionic/strcpy.S: Assembler messages:
bionic/libc/arch-arm/bionic/strcpy.S:42: Error: bad instruction `pld(r1, #0)'
(The alternative would probably have been to pull in the build change
that defines __ASSEMBLY__, but I don't want to do any more than necessary.)
Change-Id: Idad9665eb8cf5ef6a093c85d3098d11a37f5ac6a
Add missing declarations:
INET_ADDRSTRLEN
IN6_IS_ADDR_MC_NODELOCAL
IN6_IS_ADDR_MC_GLOBAL
It has been reported that these prevent compilation of the Boost libraries
with the NDK. The corresponding patch has already been performed under
development/ndk/platforms/android-3/include.
Change-Id: I4ac514973daf3c06a8ef5538d7df79142a98e562
This patch improves the handling of 64-bit parameters in syscalls on ARM.
The ARM EABI mandates that 64-bit quantities be passed in even/odd register
pairs, which requires special treatment.
This allows us to simplify our implementations of pread() and pwrite()
and remove the C stubs for pread64() and pwrite64().
Also add ftruncate64() to <unistd.h>
Change-Id: I407e2fd223ba0093dd2d0b04c6152fadfc9ce3ef
Bug 3107933
The second declaration came from an incorrect change in AOSP.
The eventfd stubs are not affected by this, they are already correct.
Change-Id: Icfc7612a68fc37a48dde2687137960a5d1dbc534
asm() conflicts with userland code compiled with -std=c99, the userland
libc should only use __asm__() instead. Therefore, this transformation
has to be applied to all exported headers.
This only changes arch-arm/asm/byteorder.h.
Change-Id: I1cf88c37201c3a91668d387293a18885c316d53c
If compiling userland code with -std=c99, the current header produces an
error. The content of this header originally is a kernel internal header,
where asm() is acceptable. In a header visible to userland, this should be
__asm__ instead.
Change-Id: I4d3188dd96f7836148ca89f5053d0389dd459d6e
This adds the cpu_set_t type definition and related functions
and macros used to handle CPU thread affinity.
sched_getcpu()
sched_setaffinity()
sched_getaffinity()
Change-Id: If382ecafde8926341a88478062b275553645065b
Add a new header that defines __ANDROID_API__ as a constant integer
corresponding to the current native API level. This header is included
by <sys/cdefs.h>
Change-Id: Ib4774e247b395991266245815a24292afc919848
NOTE: This header is already provided by the NDK's platform headers.
for levels 3 to 9.
The mkdtemp() function is implemented in libc/stdio/mktemp.c but not
exposed in stdlib.h. This change adds the prototype to stdlib.h.
Change-Id: I5a98650c665d2e45b2cf6ed3382742f7bdc7c88a
The above prototypes are missing from libc/include/wchar.h but
the functions are present (although with a limited implementation)
in libc.
Change-Id: I1b7c6accfc59ff0f56f3f41a9d3c075e935ac54a
64-bit pread() and pwrite() is needed for ZipFileRO to be able to read
ludicrously large ZIP files just in case someone is crazy enough to do
it.
Also fix a license header that was apparently mangled.
Change-Id: I6819ef8b36e46b63b40749c95717b1ecf9307033
Driver has changed to support transferring > 4 gigabyte files
Change-Id: I9ddca13e8e906a403fe84c9af27ec1fab565be77
Signed-off-by: Mike Lockwood <lockwood@android.com>
Will also need to do gethostinfo, but that's probably about it.
It was cleaner to do it at this level, rather than speaking in terms
of DNS packets.
Change-Id: I047cc459979ffb0170a3eb0d432a7e827fb71c26
Originally, there are _rand48 (in libc/bionic/_rand48.c) and __rand48
(in libc/stdlib/_rand48.c) implemented in bionic. Besides the naming,
the functionality is identical. This patch removes the duplicated
_rand48. Also, drand48 and erand48 are modified accordingly.
Change-Id: Ie5761a0a97f45df8538222a77edacb7c3e0125d7
This change allows an executable to call its destructor functions
(declared with __attribute__((destructor))) to be properly called
when it normally exits.
Note that this is different from calling the destructors of a shared
library when it is unloaded with dlclose() or through program exit,
which are already supported.
Bug: 3106500
Change-Id: I1412ef5407f13b613fc6cb6103e0a691dbee4b1a
Merge commit 'ade2a92991d1c7edc32c4a300bd83d622fa1567d' into gingerbread-plus-aosp
* commit 'ade2a92991d1c7edc32c4a300bd83d622fa1567d':
bionic: stubs: Make internal symbol static
Although header libc/stdio/local.h declares the macros and private
variables of stdio, there are several internal symbols exposed
unexpectedly.
Change-Id: Ie7a07f85b70322fb9cd05b3c8e1bcc416061eb4b
Merge commit 'ff7d94530db60cf4fe4a4c287ee8821ebdf6263a'
* commit 'ff7d94530db60cf4fe4a4c287ee8821ebdf6263a':
Hide the symbol of helper function __libc_android_abort
Merge commit '1aeeeae166920f871c1e4ecd960bb92dcaef0896'
* commit '1aeeeae166920f871c1e4ecd960bb92dcaef0896':
libc: tag missing functions in system headers.
Merge commit 'f67e5211e045af1b12f646448a5a35f96ba5e8f1' into gingerbread-plus-aosp
* commit 'f67e5211e045af1b12f646448a5a35f96ba5e8f1':
Hide the symbol of helper function __libc_android_abort
Merge commit 'b8d2233e8abc30f33be7d7acc7f3ef6e6aabebac' into gingerbread-plus-aosp
* commit 'b8d2233e8abc30f33be7d7acc7f3ef6e6aabebac':
libc: tag missing functions in system headers.
This matches recent changes in the NDK header.
We enclose missing functions in #if 0 .. #endif blocks
with a clear "MISSING" in comments in order to locate
them later.
Change-Id: I87b3a62e777897e75c9243360fb0a82bcc53d9fb
This elimates the makefile warning at the beginning of each build:
libc_common_intermediates/arch-arm/syscalls/eventfd.o' given more than once in the same rule.
Change-Id: I25cc0c373d55b5036dd17811a595c097fd6ca69a
Merge commit '5b81b918173b4bf446c1a85240c094e4dd77231f' into gingerbread-plus-aosp
* commit '5b81b918173b4bf446c1a85240c094e4dd77231f':
libc: optimize memmove() with memcpy() if possible.
The get_malloc_leak_info() currently asssumes that the totalMemory out parameter
was pre-initialized to zero before the routine is called. If it is not then the
accumulated totalMemory value will be incorrect. It is likely that many callers
will simply allocate totalMemory on the stack with no initialization and assume
that get_malloc_leak_info will set the proper value.
As an example, the caller in frameworks/base/core/jni/android_os_Debug.cpp
calls get_malloc_leak_info() with the address uninitiazed stack variable
for totalMemory. It is probably best to fix this in get_malloc_leak_info.
Change-Id: I84c927c3781419585794726115b7d34d8fdd24ae
The change explicitly isolates the assembly-only macros in header
<machine/cpu-features.h> in order to prevent mis-inclusion in C/C++
source files.
Change-Id: I0258e87c5ac3fd24944fb227290ac3b9cac4bfba
1. Make the feature test work by excluding known-deficient processors, so
we don't have to maintain a complete list of all the processors that support
REV and REV16.
2. Don't abuse 'register' to get an effect similar to GCC's +l constraint,
but which was unnecessarily restrictive.
3. Fix __swap64md so _x isn't clobbered, breaking 64-bit swaps.
4. Make <byteswap.h> (which declars bswap_16 and friends) use <endian.h>
rather than <sys/endian.h>, so we get the machine-dependent implementations.
(git cherry-pick of 3cf53d1a78)
Merge commit '9aac38249b579282ae8eaa630de3deaf97c4edff'
* commit '9aac38249b579282ae8eaa630de3deaf97c4edff':
libc: remove C++ comments from public headers.
This reverts commit 80fba9a2fe,
which caused the system to not boot anymore, aborting with:
"java.lang.RuntimeException: Missing static main on com.android.server.SystemServer".
Change-Id: I745e0a23c728cccf5f95a3c7642d544478a4e57e
Merge commit '307aadb8473744c3d946add8b3e4947ada5481aa' into gingerbread-plus-aosp
* commit '307aadb8473744c3d946add8b3e4947ada5481aa':
libc: Add missing C++ guards to <pathconf.h>
Merge commit 'bd8d987b3c3aa6d9d00cede2cb091f00bdb42204' into gingerbread-plus-aosp
* commit 'bd8d987b3c3aa6d9d00cede2cb091f00bdb42204':
libc: remove C++ comments from public headers.
Merge commit '8120a8df848e6dec31ef5f00bc0b41ed95ce1310'
* commit '8120a8df848e6dec31ef5f00bc0b41ed95ce1310':
Fixes for the ARM-specific bswap_16, bswap_32, and bswap_64.
Merge commit '3cf53d1a7814e1520df09d24b009c16f4f27db0d' into gingerbread-plus-aosp
* commit '3cf53d1a7814e1520df09d24b009c16f4f27db0d':
Fixes for the ARM-specific bswap_16, bswap_32, and bswap_64.
In order not to conflict with the symbols defined in file
libc/netbsd/getaddrinfo.c, this patch makes the internal/helper
functions static.
Change-Id: I0f85599e0b4ce0a637d005ff1680e1805dec4380
Ideally __libc_android_abort would be static, but it could not be
because gcc would not allow calling a static function from an asm
statement. Instead, using GCC visibility is work around.
Change-Id: Ifff6b9957ca3f0fc03c75c3e42582a48d43cefa2
1. Make the feature test work by excluding known-deficient processors, so
we don't have to maintain a complete list of all the processors that support
REV and REV16.
2. Don't abuse 'register' to get an effect similar to GCC's +l constraint,
but which was unnecessarily restrictive.
3. Fix __swap64md so _x isn't clobbered, breaking 64-bit swaps.
4. Make <byteswap.h> (which declars bswap_16 and friends) use <endian.h>
rather than <sys/endian.h>, so we get the machine-dependent implementations.
Change-Id: I6a38fad7a9fbe394aff141489617eb3883e1e944
ARMv6 ISA has several instructions to handle data in different byte order.
For endian conversion (byte swapping) of single data words, it might be a
good idea to use the REV/REV16 instruction simply.
(git cherry-pick -e aa35095517)
Change-Id: Id6cca2e5e0a978f5abbee6d48876dfa8fa198f6a
The implementation was using a double-checked locking approach that
could break on SMP.
In addition to the barriers I also switched to a volatile pointer. I
don't think this will matter unless gcc can conclude that _normal_lock
can't affect *once_control, but I figured it was better to be safe.
(It seems to have no impact whatsoever on the generated code.)
Bug 3022795.
Change-Id: Ib91da25d57ff5bee4288526e39d457153ef6aacd
Merge commit 'b9e49ad56e5776ace7c6eab2e997d5b7acb16792' into gingerbread-plus-aosp
* commit 'b9e49ad56e5776ace7c6eab2e997d5b7acb16792':
Use a recursive lock for pthread_once. DO NOT MERGE.
Merge commit 'ef3644d110bcb44010672c0ad0bc7c96fd130c8f'
* commit 'ef3644d110bcb44010672c0ad0bc7c96fd130c8f':
Use ARMv6 instruction for handling byte order
Merge commit '312be567a03aaf851707a268807ee666b12f8c74' into gingerbread-plus-aosp
* commit '312be567a03aaf851707a268807ee666b12f8c74':
Use ARMv6 instruction for handling byte order
Merge commit '6bf08d131f4109088b7fa1e53196513636c49e9b'
* commit '6bf08d131f4109088b7fa1e53196513636c49e9b':
Move the atom optimization flags into TARGET_linux-x86.mk.
Merge commit '3558834c565bbfe96dfcf233373c3814b4655347' into gingerbread-plus-aosp
* commit '3558834c565bbfe96dfcf233373c3814b4655347':
Move the atom optimization flags into TARGET_linux-x86.mk.
Merge commit '3a1bbee36b773862e88c6f1895b607c0cd81b499' into gingerbread-plus-aosp
* commit '3a1bbee36b773862e88c6f1895b607c0cd81b499':
Add eventfd system call support
This adds an explicit memory barrier to condition variable signaling.
It's a little murky as to whether it's strictly required, but it seems
like a wise thing to do.
Change-Id: Id0faa542d61e4b8ffa775e4adf68e4d7471f4fb7
Return a valid pointer (not NULL) when the character "c" is at the end of "src".
Change-Id: Iab0b677943f2c8a9fbb255c44689f5d6dc3535d7
Example:
memccpy(dest, "xzy", 'y', 3) should return dest+3 rather than null.
Merge commit '3fda8700bd042fbb7ac37e2bbe7496df8a087244'
* commit '3fda8700bd042fbb7ac37e2bbe7496df8a087244':
Make timezones available to the host build.
Merge commit 'cc6b18291075a7960ad0d638d2047f41be1297e8' into gingerbread-plus-aosp
* commit 'cc6b18291075a7960ad0d638d2047f41be1297e8':
Make timezones available to the host build.
ARMv6 ISA has several instructions to handle data in different byte order.
For endian conversion (byte swapping) of single data words, it might be a
good idea to use the REV/REV16 instruction simply.
Change-Id: Ic4a5ed6254e082763e54aa70d428f59a0088636e
Merge commit '994b68f5c7d637d2945c06260fe35b266c10379d'
* commit '994b68f5c7d637d2945c06260fe35b266c10379d':
Fix __get_tls() in static C library to use kernel helpers.
Previously, the malloc leak checking code would crash in qsort()
if null entries existed in its bookkeeping table. This change
makes the comparison function detect null entries and sort them
to the end safely.
Change-Id: I88244a7df1e289dd9d7992ce29606d505bd63079
Merge commit 'a685faf3e3de181ca6c795b0062432c74073d024'
* commit 'a685faf3e3de181ca6c795b0062432c74073d024':
Update time zone data to the new version 2010k.
Merge commit 'cd7014fd0634f1337fc3857f1c61e0d39443c4be' into gingerbread-plus-aosp
* commit 'cd7014fd0634f1337fc3857f1c61e0d39443c4be':
Update time zone data to the new version 2010k.
Merge commit '90e83bcc1c41a4d6d310b2bd5cb747a53a88fff3'
* commit '90e83bcc1c41a4d6d310b2bd5cb747a53a88fff3':
Fix undefined reference to dl_iterate_phdr for x86
Merge commit '519763265ec0b634bd9c264a0aca034882458ecc' into gingerbread-plus-aosp
* commit '519763265ec0b634bd9c264a0aca034882458ecc':
libc: Fix sem_post() implementation to wake up all waiting threads.
This also allows us to optimize the case where we increment an
uncontended semaphore (no need to call futex_wake() then).
Change-Id: Iad48efe8551dc66dc89d3e3f18c001e5a6c1939f
With this patch _and_ an upcoming build/ patch, the destruction
of static C++ objects contained in shared libraries will happen
properly when dlclose() is called.
Note that this change introduces crtbegin_so.S and crtend_so.S which
are currently ignored by the build system.
+ move definition of __dso_handle to the right place
(before that, all shared libraries used the __dso_handle
global variable from the C library).
Note that we keep a 'weak' __dso_handle in aeabi.c to avoid
breaking the build until the next patch to build/core/combo/
appears. We will be able to remove that later.
+ move bionic/aeabi.c to arch-arm/bionic/ (its proper location)
NOTE: The NDK will need to be modified to enable this feature in
the shared libraries that are generated through it.
Change-Id: I99cd801375bbaef0581175893d1aa0943211b9bc
Add ip6t_get_target() to kernel_known_generic_statics in
libc/kernel/tools/defaults.py to be able to build ip6tables.
Change-Id: Iadb885db3faa85b2d0070dc2e0ac493af6e62bb6
Merge commit 'fa5755ae980b6f4ac5df37b263ba20f472db31ca' into gingerbread-plus-aosp
* commit 'fa5755ae980b6f4ac5df37b263ba20f472db31ca':
Add clean headers for 7x30 video codec.
Merge commit '50ace4fec5e8cb5afcbc656a4556fa528adfd760' into gingerbread-plus-aosp
* commit '50ace4fec5e8cb5afcbc656a4556fa528adfd760':
Remove compiler warnings when building Bionic.
Merge commit 'e1a739ec38a39fea9e66038dfc7cea364cb6a256'
* commit 'e1a739ec38a39fea9e66038dfc7cea364cb6a256':
Allow static C++ destructors to be properly called on dlclose().
Also add missing declarations to misc. functions.
Fix clearerr() implementation (previous was broken).
Handle feature test macros like _POSIX_C_SOURCE properly.
Change-Id: Icdc973a6b9d550a166fc2545f727ea837fe800c4
Merge commit '6a9b888d7c4b246f6f66360789c72b754ff85021' into gingerbread-plus-aosp
* commit '6a9b888d7c4b246f6f66360789c72b754ff85021':
Allow static C++ destructors to be properly called on dlclose().
With this patch, _and_ an upcoming build/ patch, the destruction
of static C++ objects contained in shared libraries will happen
properly when dlclose() is called.
Note that this change introduces crtbegin_so.S and crtend_so.S which
are currently ignored by the build system.
+ move definition of __dso_handle to the right place
(before that, all shared libraries used the __dso_handle
global variable from the C library).
Note that we keep a 'weak' __dso_handle in aeabi.c to avoid
breaking the build until the next patch to build/core/combo/
appears. We will be able to remove that later.
+ move bionic/aeabi.c to arch-arm/bionic/ (its proper location)
Change-Id: Ie771aa204e3acbdf02fd30ebd4150373a1398f39
NOTE: The NDK will need to be modified to enable this feature in
the shared libraries that are generated through it.
When calling sysconf with _SC_NPROCESSORS_ONLN, the value one (1) was
returned on systems with two or more cores, since '/proc/stat' was
incorrectly parsed.
The function line_parser_getc (LineParser* p) read 128 characters of
input for each invocation.
The proper and probably aimed for behavior is to read 128 characters
at the first call, then for each subsequent call only return the next
buffered character until a new read is needed and only then read
another 128 characters.
Due to a flipped comparison between the two variables in_len and
in_pos that track the number of bytes of data read into the input
buffer and how much of it has been parsed, a new group of 128
characters were read at almost every call to line_parser_getc,
overwriting the still unhandled bytes from the previous call to
read. This caused the lines to be read to be sampled more than parsed.
Change-Id: I93eec3c8c9b9f19ef798748579d0977111b5c0bb
Signed-off-by: Christian Bejram <christian.bejram@stericsson.com>
Merge commit '6a09cfd9f916e3a60de707ff0806cdeb143d77a4' into gingerbread-plus-aosp
* commit '6a09cfd9f916e3a60de707ff0806cdeb143d77a4':
wchar.h: improve wchar_t support in Bionic
We simply copy the stuff we need from cutils headers.
A future patch will change cutils to include the private <bionic_atomic_inline.h>
Change-Id: Ib6fd9a03bc9e337ce867bd606dc94c2b4438480a
bionic/libc/include/../include/stdlib.h: In function 'int grantpt(int)':
bionic/libc/include/../include/stdlib.h:138:23: warning: parameter '__fd' set but not used [-Wunused-but-set-parameter]
By adding __attribute((unused)) to __fd, the warning is gone, and this
attribute is compatible with gcc-4.4.0, gcc-4.3.1 gcc-4.2.1.
There is no any side effect.
Change-Id: I385f0f4da1013ffd1499e391eac9123aafe1f7a5
Update ARM atomic ops to use LDREX/STREX. Stripped out #if 0 chunk.
Insert explicit memory barriers in pthread and semaphore code.
For bug 2721865.
Change-Id: I0f153b797753a655702d8be41679273d1d5d6ae7
bionic/libc/include/../include/stdlib.h: In function 'int grantpt(int)':
bionic/libc/include/../include/stdlib.h:138:23: warning: parameter '__fd' set but not used [-Wunused-but-set-parameter]
By adding __attribute((unused)) to __fd, the warning is gone, and this
attribute is compatible with gcc-4.4.0, gcc-4.3.1 gcc-4.2.1.
There is no any side effect.
Change-Id: Id9d2b0aeefdd9c37132da0f00c48387f7d897e83
Added an underscore to _ARM_HAVE_LDREX_STREX to make it match the others.
Added __ARM_HAVE_DMB and __ARM_HAVE_LDREXD when appropriate.
Fixed some typos.
Change-Id: I2f55febcff4aeb7de572a514fb2cd2f820dca27c
... so that each cloned process at the kernel level can be named
independently. Tools like 'top' can display the CPU/memory statistics
for each process's thread if "Show Threads" mode is on.
With this function in place, we can convert dalvik/Thread.c setThreadName()
function over this function. This feature ought to be provided by the
underlying C library and not coded directly in Dalvik.
Change-Id: Ifa997665dbaa114e0b126f8c667708be9a4137fd
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
GDB looks for specific opcode sequences when trying to recognize a stack
frame as a signal trampoline. The sequences it looks for happen to be those
created when SA_RESTORER is set, since glibc always sets a restorer. This
patch does the same here, so that the trampolines can be correctly identified.
Change-Id: I0ac574a68818cb24d939c3527f3aaeb04b853d04
... by removing extraneous NULL check, as free() already does it.
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
Change-Id: I0445f35c7ad0a049a0e4aee1fbe002ed2f13b94b
... by removing unneeded NULL check, as free() already does it.
By the way, we don't need to set a stack variable back to NULL.
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
Change-Id: Id90eb8f042b5c922c5ff139b11ff8366fb404566
... by removing unneeded NULL check, as free() already does it.
By the way, we don't need to set a stack variable back to NULL.
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
Change-Id: Id1f72e872f73366dddcea4abc75885a3d9a318c6
... by checking most probable condition first (elements do differ)
Change-Id: I424eab9c32a6d9eb82b686ca04025ec8c9097035
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
Typo assigned prefixlen1 twice instead of to the two different variables
for comparison and difference computation.
Change-Id: I6631b8269ca6aae264c8d7d414127b756838df96
Now that the system properly uses shared condvars when needed, we
can enable the use of private futexes for them too.
Change-Id: Icf8351fc0a2309f764cba45c65bc3af047720cdf
This does not change the implementation of conditional variables
since we're waiting for other system components to properly use
pthread_condattr_init/setpshared before that.
Also remove an obsolete x86 source file.
Change-Id: Ia3e3fbac35b87a534fb04d4381c3c66b975bc8f7
Note that this does not change the implementation of conditional variables
which still use shared futexes, independent on the flags being selected.
This will be fixed in a later patch, once our system is modified to use
pthread_condattr_setpshared(attr, PTHREAD_PROCESS_SHARED) properly.
Change-Id: I935de50964cd41f97a13dbfd6626d3407b0406c3
dladdr() is a GNU extension function, which allows the caller to retrieve
symbol information for a specified memory address. It is useful for things
like generating backtrace information at runtime.
Change-Id: I3a1def1a6c9c666d93e1e97b7d260dfa5b9b79a9
Private futexes are a recent kernel addition: faster futexes that cannot be
shared between processes. This patch uses them by default, unless the PROCESS_SHARED
attribute flag is used when creating a mutex and/or conditional variable.
Also introduces pthread_condattr_init/destroy/setpshared/getpshared.
Change-Id: I3a0e2116f467072b046524cb5babc00e41057a53
To avoid introducing NULL checks everytime a Bigint computation
is performed, introduce a special value (BIGINT_INVALID) and only
check for it when absolutely needed (which means when the code
needs to access the Bigint's internal structure fields).
Change-Id: Ie3954523b0985b6a8cb62340802d8dcf63cbf16f
The problem was that the 'defdname' field of res_state structure
was not properly initialized in __res_vinit(). This field is used
to store the default domain name, which is normally build from
calling gethostname() (see line 549 of res_init.c).
Unfortunately, in the typical Android case, gethostname() returns
an error (the hostname is configured) and a random stack string is
used later to build the DNS search list (see lines 556+ in res_init.c)
For the sake of illustration, let's say the search list is set to
a random value like 'xWLK'.
The end result is that when trying to result an unknown domain name
(e.g. 'www.ptn'), the query fails then the resolver tries to make a
new query with the DNS search list path(s) appended (e.g. 'www.ptn.xWLK').
The patch simply initializes 'defdname' to an empty string to avoid
this when the net.dns.search system property is not set.
Also contains whitespace/formatting fixes
Any of the setuid functions now updates /acct/uid/ with its own tid
before changing users. This is so we can properly account for cpu time
per uid.
Change-Id: I34186cf4d5228cac8439e582a9e26c01ef3011e4
Signed-off-by: Mike Chan <mike@android.com>
Java changes required not to mess up the ordering from bionic will arrive in a
later commit.) In particular, this will give us more correct behavior when on a
6to4 network, in that IPv4 will usually be preferred over 6to4.
Most of RFC 3484 is implemented -- what's not is rule 3 (avoid deprecated
addresses), 4 (prefer home addresses) and 7 (prefer native transport) as they
require low-level access to the kernel routing table via netlink. (glibc also
started out this way, and these rules are primarily useful in pretty obscure
circumstances, so we should be fine for the time being.)
Also, rule 9 (use longest matching prefix) has been modified so it does not try
to sort IPv4 addresses; given current IPv4 addressing practice these rules are
pretty much meaningless. Finally, I've added support for Teredo as a separate
label, with slightly lower preference than 6to4. (Vista puts the preference
below IPv4 by default. glibc puts the preference together with non-tunneled
IPv6.)
Note that this patch removes support for the "sortlist" directive in
resolv.conf; I've never seen it in actual use, it's irrelevant for Android
(since we don't use resolv.conf anyway), and it's not clear how it would be
implemented alongside RFC 3484.
On ARM EABI, 64-bit function parameters must be aligned
to an even/odd register pair. The weird way these stubs
were written (using separate lo/hi parameters) prevented
this alignment from being enforced by the compiler.
