tequilaOS/platform_bionic
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
platform_bionic/libc/include/pthread.h

244 lines
10 KiB
C
Raw Normal View History

auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
/*
* Copyright (C) 2008 The Android Open Source Project
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
* COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
* OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
* AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
* OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
* OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
#ifndef _PTHREAD_H_
#define _PTHREAD_H_
#include <limits.h>
Implement POSIX lockf. This has been requested a few times over the years. This is basically a very late rebase of https://android-review.googlesource.com/45470 which was abandoned years ago. One addition is that this version has _FILE_OFFSET_BITS=64 support. POSIX puts this in <unistd.h>. glibc also has it in <fcntl.h>. Bug: http://b/13077650 Change-Id: I5862b1dc326e326c01ad92438ecc1578d19ba739
2016-01-26 03:06:24 +01:00
#include <bits/pthread_types.h>
Move some pthread functions to signal.h. POSIX specifies that pthread_kill(3) and pthread_sigmask(3) are supposed to live in signal.h rather than pthread.h. Since signal.h now needs pthread_t and pthread_attr_t, I've moved those defintions into include/machine/pthread_types.h to keep the namespace clean. I also sorted some includes. The combination of these two things seems to have exploded into a cascade of missing includes, so this patch also cleans up all those. Change-Id: Icfa92a39432fe83f542a797e5a113289d7e4ad0c
2014-11-21 09:18:07 +01:00
#include <sched.h>
#include <sys/cdefs.h>
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
#include <sys/types.h>
Move some pthread functions to signal.h. POSIX specifies that pthread_kill(3) and pthread_sigmask(3) are supposed to live in signal.h rather than pthread.h. Since signal.h now needs pthread_t and pthread_attr_t, I've moved those defintions into include/machine/pthread_types.h to keep the namespace clean. I also sorted some includes. The combination of these two things seems to have exploded into a cascade of missing includes, so this patch also cleans up all those. Change-Id: Icfa92a39432fe83f542a797e5a113289d7e4ad0c
2014-11-21 09:18:07 +01:00
#include <time.h>
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
Make it possible to build bionic with -I instead of -isystem. Various things: * work around -Wnullability-completeness. * use C++ casts in C++ and C casts in C. * stop using attributes clang doesn't support (such as `warning`). * remove duplicate definitions of XATTR_CREATE and XATTR_REPLACE. Change-Id: I07649e46275b28a23ca477deea119fe843999533
2016-07-22 20:36:17 +02:00
__BEGIN_DECLS
#if defined(__clang__)
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wnullability-completeness"
#endif
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
enum {
PTHREAD_MUTEX_NORMAL = 0,
PTHREAD_MUTEX_RECURSIVE = 1,
PTHREAD_MUTEX_ERRORCHECK = 2,
PTHREAD_MUTEX_ERRORCHECK_NP = PTHREAD_MUTEX_ERRORCHECK,
PTHREAD_MUTEX_RECURSIVE_NP = PTHREAD_MUTEX_RECURSIVE,
PTHREAD_MUTEX_DEFAULT = PTHREAD_MUTEX_NORMAL
};
Hide internal of pthread_mutex_t. Bug: 19249079 Change-Id: Iffb79c8d861b698d474f212dc80c638fc2cf1620
2015-03-21 23:08:25 +01:00
#define PTHREAD_MUTEX_INITIALIZER { { ((PTHREAD_MUTEX_NORMAL & 3) << 14) } }
#define PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP { { ((PTHREAD_MUTEX_RECURSIVE & 3) << 14) } }
#define PTHREAD_ERRORCHECK_MUTEX_INITIALIZER_NP { { ((PTHREAD_MUTEX_ERRORCHECK & 3) << 14) } }
#define PTHREAD_COND_INITIALIZER { { 0 } }
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Hide content of pthread_rwlock_t in pthread_rwlock_internal_t. Bug: 19249079 Change-Id: Ifbe634c716b6793bef897ec5134b55eb44c6b8d5
2015-03-13 22:22:05 +01:00
#define PTHREAD_RWLOCK_INITIALIZER { { 0 } }
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Provide writer preference option in rwlock. Previous implementation of rwlock contains four atomic variables, which is hard to maintain and change. So I make following changes in this CL: 1. Add pending flags in rwlock.state, so we don't need to synchronize between different atomic variables. Using compare_and_swap operations on rwlock.state is enough for all state change. 2. Add pending_lock to protect readers/writers waiting and wake up operations. As waiting/wakeup is not performance critical, using a lock is easier to maintain. 3. Add writer preference option. 4. Add unit tests for rwlock. Bug: 19109156 Change-Id: Idcaa58d695ea401d64445610b465ac5cff23ec7c
2015-03-17 22:22:09 +01:00
enum {
PTHREAD_RWLOCK_PREFER_READER_NP = 0,
PTHREAD_RWLOCK_PREFER_WRITER_NONRECURSIVE_NP = 1,
};
Reserve space in pthread structures for future extensibility. Make our structures equal in size with glibc structures. This should give us plenty of space to implement any missing feature. Bug: 13278744 Bug: 12875898 (cherry picked from commit cf83fd77caa3128ecb6ac935e342cb0ea258b9ee) Change-Id: I76968d31024eb51bc73887687e5ac492eb02a27f
2014-04-23 18:35:04 +02:00
#define PTHREAD_ONCE_INIT 0
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
Implement pthread barrier. Bug: 24341262 Change-Id: I5472549e5d7545c1c3f0bef78235f545557b9630
2015-11-06 07:06:09 +01:00
#define PTHREAD_BARRIER_SERIAL_THREAD -1
Increase minimum thread size for 64 bit. Bug: 14468519 Change-Id: I642a5515acf5fbed46e77e176f6b5dd3b7c7bd45
2014-05-15 02:58:29 +02:00
#if defined(__LP64__)
#define PTHREAD_STACK_MIN (4 * PAGE_SIZE)
#else
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
#define PTHREAD_STACK_MIN (2 * PAGE_SIZE)
Increase minimum thread size for 64 bit. Bug: 14468519 Change-Id: I642a5515acf5fbed46e77e176f6b5dd3b7c7bd45
2014-05-15 02:58:29 +02:00
#endif
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
#define PTHREAD_CREATE_DETACHED 0x00000001
#define PTHREAD_CREATE_JOINABLE 0x00000000
#define PTHREAD_PROCESS_PRIVATE 0
#define PTHREAD_PROCESS_SHARED 1
#define PTHREAD_SCOPE_SYSTEM 0
#define PTHREAD_SCOPE_PROCESS 1
Update header versions for NDK platform fixes. Bug: http://b/28178111 Change-Id: Icd638673b409aa43a91490f77c6b4d79c9ea20d9
2016-05-27 20:14:16 +02:00
int pthread_atfork(void (*)(void), void (*)(void), void (*)(void)) __INTRODUCED_IN(12);
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_attr_destroy(pthread_attr_t* _Nonnull);
int pthread_attr_getdetachstate(const pthread_attr_t* _Nonnull, int* _Nonnull);
int pthread_attr_getguardsize(const pthread_attr_t* _Nonnull, size_t* _Nonnull);
int pthread_attr_getschedparam(const pthread_attr_t* _Nonnull, struct sched_param* _Nonnull);
int pthread_attr_getschedpolicy(const pthread_attr_t* _Nonnull, int* _Nonnull);
int pthread_attr_getscope(const pthread_attr_t* _Nonnull, int* _Nonnull);
int pthread_attr_getstack(const pthread_attr_t* _Nonnull, void** _Nonnull, size_t* _Nonnull);
int pthread_attr_getstacksize(const pthread_attr_t* _Nonnull, size_t* _Nonnull);
int pthread_attr_init(pthread_attr_t* _Nonnull);
int pthread_attr_setdetachstate(pthread_attr_t* _Nonnull, int);
int pthread_attr_setguardsize(pthread_attr_t* _Nonnull, size_t);
int pthread_attr_setschedparam(pthread_attr_t* _Nonnull, const struct sched_param* _Nonnull);
int pthread_attr_setschedpolicy(pthread_attr_t* _Nonnull, int);
int pthread_attr_setscope(pthread_attr_t* _Nonnull, int);
int pthread_attr_setstack(pthread_attr_t* _Nonnull, void*, size_t);
int pthread_attr_setstacksize(pthread_attr_t* _Nonnull, size_t);
int pthread_condattr_destroy(pthread_condattr_t* _Nonnull);
int pthread_condattr_getclock(const pthread_condattr_t* _Nonnull, clockid_t* _Nonnull)
Add versioning information to symbols. Bug: http://b/28178111 Change-Id: I46bf95accd819f4521afb1173d8badcc5e9df31c
2016-04-29 21:00:55 +02:00
__INTRODUCED_IN(21);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_condattr_getpshared(const pthread_condattr_t* _Nonnull, int* _Nonnull);
int pthread_condattr_init(pthread_condattr_t* _Nonnull);
int pthread_condattr_setclock(pthread_condattr_t* _Nonnull, clockid_t) __INTRODUCED_IN(21);
int pthread_condattr_setpshared(pthread_condattr_t* _Nonnull, int);
int pthread_cond_broadcast(pthread_cond_t* _Nonnull);
int pthread_cond_destroy(pthread_cond_t* _Nonnull);
int pthread_cond_init(pthread_cond_t* _Nonnull, const pthread_condattr_t*);
int pthread_cond_signal(pthread_cond_t* _Nonnull);
int pthread_cond_timedwait(pthread_cond_t* _Nonnull, pthread_mutex_t* _Nonnull,
const struct timespec* _Nonnull);
int pthread_cond_wait(pthread_cond_t* _Nonnull, pthread_mutex_t* _Nonnull);
int pthread_create(pthread_t* _Nonnull, pthread_attr_t const*,
void* (* _Nonnull start_routine)(void*), void*);
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
int pthread_detach(pthread_t);
void pthread_exit(void*) __noreturn;
int pthread_equal(pthread_t, pthread_t);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_getattr_np(pthread_t, pthread_attr_t* _Nonnull);
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_getcpuclockid(pthread_t, clockid_t* _Nonnull);
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_getschedparam(pthread_t, int* _Nonnull, struct sched_param* _Nonnull);
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
void* pthread_getspecific(pthread_key_t);
Add versioning information to symbols. Bug: http://b/28178111 Change-Id: I46bf95accd819f4521afb1173d8badcc5e9df31c
2016-04-29 21:00:55 +02:00
pid_t pthread_gettid_np(pthread_t) __INTRODUCED_IN(21);
Add pthread_gettid_np and re-expose __get_thread for LP32. A lot of third-party code calls the private __get_thread symbol, often as part of a backport of bionic's pthread_rwlock implementation. Hopefully this will go away for LP64 (since you're guaranteed the real implementation there), but there are still APIs that take a tid and no way to convert between a pthread_t and a tid. pthread_gettid_np is a public API for that. To aid the transition, make __get_thread available again for LP32. (cherry-pick of 27efc48814b8153c55cbcd0af5d9add824816e69.) Bug: 14079438 Change-Id: I43fabc7f1918250d31d4665ffa4ca352d0dbeac1
2014-09-12 01:11:43 +02:00
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
int pthread_join(pthread_t, void**);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_key_create(pthread_key_t* _Nonnull, void (*)(void*));
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
int pthread_key_delete(pthread_key_t);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_mutexattr_destroy(pthread_mutexattr_t* _Nonnull);
int pthread_mutexattr_getpshared(const pthread_mutexattr_t* _Nonnull, int* _Nonnull);
int pthread_mutexattr_gettype(const pthread_mutexattr_t* _Nonnull, int* _Nonnull);
int pthread_mutexattr_init(pthread_mutexattr_t* _Nonnull);
int pthread_mutexattr_setpshared(pthread_mutexattr_t* _Nonnull, int);
int pthread_mutexattr_settype(pthread_mutexattr_t* _Nonnull, int);
int pthread_mutex_destroy(pthread_mutex_t* _Nonnull);
int pthread_mutex_init(pthread_mutex_t* _Nonnull, const pthread_mutexattr_t*);
int pthread_mutex_lock(pthread_mutex_t* _Nonnull);
int pthread_mutex_timedlock(pthread_mutex_t* _Nonnull, const struct timespec* _Nonnull)
Add versioning information to symbols. Bug: http://b/28178111 Change-Id: I46bf95accd819f4521afb1173d8badcc5e9df31c
2016-04-29 21:00:55 +02:00
__INTRODUCED_IN(21);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_mutex_trylock(pthread_mutex_t* _Nonnull);
int pthread_mutex_unlock(pthread_mutex_t* _Nonnull);
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_once(pthread_once_t* _Nonnull, void (* _Nonnull init_routine)(void));
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_rwlockattr_init(pthread_rwlockattr_t* _Nonnull);
int pthread_rwlockattr_destroy(pthread_rwlockattr_t* _Nonnull);
int pthread_rwlockattr_getpshared(const pthread_rwlockattr_t* _Nonnull, int* _Nonnull);
int pthread_rwlockattr_setpshared(pthread_rwlockattr_t* _Nonnull, int);
int pthread_rwlockattr_getkind_np(const pthread_rwlockattr_t* _Nonnull, int* _Nonnull)
Add versioning information to symbols. Bug: http://b/28178111 Change-Id: I46bf95accd819f4521afb1173d8badcc5e9df31c
2016-04-29 21:00:55 +02:00
__INTRODUCED_IN(23);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_rwlockattr_setkind_np(pthread_rwlockattr_t* _Nonnull, int) __INTRODUCED_IN(23);
int pthread_rwlock_destroy(pthread_rwlock_t* _Nonnull);
int pthread_rwlock_init(pthread_rwlock_t* _Nonnull, const pthread_rwlockattr_t*);
int pthread_rwlock_rdlock(pthread_rwlock_t* _Nonnull);
int pthread_rwlock_timedrdlock(pthread_rwlock_t* _Nonnull, const struct timespec* _Nonnull);
int pthread_rwlock_timedwrlock(pthread_rwlock_t* _Nonnull, const struct timespec* _Nonnull);
int pthread_rwlock_tryrdlock(pthread_rwlock_t* _Nonnull);
int pthread_rwlock_trywrlock(pthread_rwlock_t* _Nonnull);
int pthread_rwlock_unlock(pthread_rwlock_t* _Nonnull);
int pthread_rwlock_wrlock(pthread_rwlock_t* _Nonnull);
int pthread_barrierattr_init(pthread_barrierattr_t* _Nonnull attr) __INTRODUCED_IN(24);
int pthread_barrierattr_destroy(pthread_barrierattr_t* _Nonnull attr) __INTRODUCED_IN(24);
Fix pthread_barrierattr_getpshared prototype. As http://pubs.opengroup.org/onlinepubs/9699919799/functions/pthread_barrierattr_getpshared.html is my witness, there's a missing `const` here. Change-Id: Ifd68188fb952460b64d144bdc21a53bc40ed82a7
2016-08-12 08:18:13 +02:00
int pthread_barrierattr_getpshared(const pthread_barrierattr_t* _Nonnull attr,
int* _Nonnull pshared) __INTRODUCED_IN(24);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_barrierattr_setpshared(pthread_barrierattr_t* _Nonnull attr, int pshared)
Add versioning information to symbols. Bug: http://b/28178111 Change-Id: I46bf95accd819f4521afb1173d8badcc5e9df31c
2016-04-29 21:00:55 +02:00
__INTRODUCED_IN(24);
Implement pthread barrier. Bug: 24341262 Change-Id: I5472549e5d7545c1c3f0bef78235f545557b9630
2015-11-06 07:06:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_barrier_init(pthread_barrier_t* _Nonnull, const pthread_barrierattr_t*, unsigned)
Add versioning information to symbols. Bug: http://b/28178111 Change-Id: I46bf95accd819f4521afb1173d8badcc5e9df31c
2016-04-29 21:00:55 +02:00
__INTRODUCED_IN(24);
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_barrier_destroy(pthread_barrier_t* _Nonnull) __INTRODUCED_IN(24);
int pthread_barrier_wait(pthread_barrier_t* _Nonnull) __INTRODUCED_IN(24);
Implement pthread barrier. Bug: 24341262 Change-Id: I5472549e5d7545c1c3f0bef78235f545557b9630
2015-11-06 07:06:09 +01:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_spin_destroy(pthread_spinlock_t* _Nonnull) __INTRODUCED_IN(24);
int pthread_spin_init(pthread_spinlock_t* _Nonnull, int) __INTRODUCED_IN(24);
int pthread_spin_lock(pthread_spinlock_t* _Nonnull) __INTRODUCED_IN(24);
int pthread_spin_trylock(pthread_spinlock_t* _Nonnull) __INTRODUCED_IN(24);
int pthread_spin_unlock(pthread_spinlock_t* _Nonnull) __INTRODUCED_IN(24);
Implement pthread spin. In order to run tsan unit tests, we need to support pthread spin APIs. Bug: 18623621 Bug: 25392375 Change-Id: Icbb4a74e72e467824b3715982a01600031868e29
2015-11-18 01:03:18 +01:00
Rename __pure, __pure2, and __purefunc. These names were pretty misleading (aka "backwards"), so switch to the same obvious names glibc uses. Test: build. Change-Id: Ia98c9dbbccd0820386116562347654e84669034a
2016-09-28 21:29:52 +02:00
pthread_t pthread_self(void) __attribute_const__;
added pthread_cond_timedwait_relative_np()
2009-07-14 00:00:46 +02:00
Add pthread_getname_np. Also guard both these GNU extensions with _GNU_SOURCE. Also improve the tests to test each case on both the current thread and another thread, since the code paths are totally different. Bug: http://b/27810459 Change-Id: I72b05bca5c5b6ca8ba4585b8edfb716a1c252f92
2016-03-23 19:20:47 +01:00
#if defined(__USE_GNU)
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_getname_np(pthread_t, char* _Nonnull, size_t) __INTRODUCED_IN_FUTURE;
Add pthread_getname_np. Also guard both these GNU extensions with _GNU_SOURCE. Also improve the tests to test each case on both the current thread and another thread, since the code paths are totally different. Bug: http://b/27810459 Change-Id: I72b05bca5c5b6ca8ba4585b8edfb716a1c252f92
2016-03-23 19:20:47 +01:00
#endif
Remove the __USE_GNU guard from pthread_setname_np again. Caused build breakage, but not spotted by TreeHugger: external/libnfc-nxp/Linux_x86/phDal4Nfc.c:737:5: error: implicit declaration of function 'pthread_setname_np' is invalid in C99 [-Werror,-Wimplicit-function-declaration] pthread_setname_np(pthread_self(), "reader"); ^ Nice file name. Change-Id: I102e55718babf4e4f2f2c64066c6a83ed3ac86d3
2016-03-24 04:26:35 +01:00
/* TODO: this should be __USE_GNU too. */
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_setname_np(pthread_t, const char* _Nonnull);
Add pthread_rwlock_t implementation to the C library (DO NOT MERGE) Change-Id: I756d8c26afc37cd7b71117ddbaa02a2cb40fdecb
2010-05-12 01:39:22 +02:00
Use clang's nullability instead of nonnull. http://clang.llvm.org/docs/AttributeReference.html#nonnull _Nonnull is similar to the nonnull attribute in that it will instruct compilers to warn the user if it can prove that a null argument is being passed. Unlike the nonnull attribute, this annotation indicated that a value *should not* be null, not that it *cannot* be null, or even that the behavior is undefined. The important distinction is that the optimizer will perform surprising optimizations like the following: void foo(void*) __attribute__(nonnull, 1); int bar(int* p) { foo(p); // The following null check will be elided because nonnull // attribute means that, since we call foo with p, p can be // assumed to not be null. Thus this will crash if we are called // with a null pointer. if (src != NULL) { return *p; } return 0; } int main() { return bar(NULL); } Note that by doing this we are no longer attaching any sort of attribute for GCC (GCC doesn't support attaching nonnull directly to a parameter, only to the function and naming the arguments positionally). This means we won't be getting a warning for this case from GCC any more. People that listen to warnings tend to use clang anyway, and we're quickly moving toward that as the default, so this seems to be an acceptable tradeoff. Change-Id: Ie05fe7cec2f19a082c1defb303f82bcf9241b88d
2015-08-14 01:58:50 +02:00
int pthread_setschedparam(pthread_t, int, const struct sched_param* _Nonnull);
Add pthread_rwlock_t implementation to the C library (DO NOT MERGE) Change-Id: I756d8c26afc37cd7b71117ddbaa02a2cb40fdecb
2010-05-12 01:39:22 +02:00
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
int pthread_setspecific(pthread_key_t, const void*);
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
typedef void (*__pthread_cleanup_func_t)(void*);
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
typedef struct __pthread_cleanup_t {
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
struct __pthread_cleanup_t* __cleanup_prev;
__pthread_cleanup_func_t __cleanup_routine;
void* __cleanup_arg;
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
} __pthread_cleanup_t;
Make it possible to build bionic with -I instead of -isystem. Various things: * work around -Wnullability-completeness. * use C++ casts in C++ and C casts in C. * stop using attributes clang doesn't support (such as `warning`). * remove duplicate definitions of XATTR_CREATE and XATTR_REPLACE. Change-Id: I07649e46275b28a23ca477deea119fe843999533
2016-07-22 20:36:17 +02:00
void __pthread_cleanup_push(__pthread_cleanup_t* c, __pthread_cleanup_func_t, void*);
void __pthread_cleanup_pop(__pthread_cleanup_t*, int);
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
/* Believe or not, the definitions of pthread_cleanup_push and
* pthread_cleanup_pop below are correct. Posix states that these
* can be implemented as macros that might introduce opening and
* closing braces, and that using setjmp/longjmp/return/break/continue
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
* between them results in undefined behavior.
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
*/
#define pthread_cleanup_push(routine, arg) \
do { \
__pthread_cleanup_t __cleanup; \
__pthread_cleanup_push( &__cleanup, (routine), (arg) ); \
#define pthread_cleanup_pop(execute) \
__pthread_cleanup_pop( &__cleanup, (execute)); \
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
} while (0); \
Make it possible to build bionic with -I instead of -isystem. Various things: * work around -Wnullability-completeness. * use C++ casts in C++ and C casts in C. * stop using attributes clang doesn't support (such as `warning`). * remove duplicate definitions of XATTR_CREATE and XATTR_REPLACE. Change-Id: I07649e46275b28a23ca477deea119fe843999533
2016-07-22 20:36:17 +02:00
#if defined(__clang__)
#pragma clang diagnostic pop
#endif
<pthread.h> fixes and pthread cleanup. <pthread.h> was missing nonnull attributes, noreturn on pthread_exit, and had incorrect cv qualifiers for several standard functions. I've also marked the non-standard stuff (where I count glibc rather than POSIX as "standard") so we can revisit this cruft for LP64 and try to ensure we're compatible with glibc. I've also broken out the pthread_cond* functions into a new file. I've made the remaining pthread files (plus ptrace) part of the bionic code and fixed all the warnings. I've added a few more smoke tests for chunks of untested pthread functionality. We no longer need the libc_static_common_src_files hack for any of the pthread implementation because we long since stripped out the rest of the armv5 support, and this hack was just to ensure that __get_tls in libc.a went via the kernel if necessary. This patch also finishes the job of breaking up the pthread.c monolith, and adds a handful of new tests. Change-Id: Idc0ae7f5d8aa65989598acd4c01a874fe21582c7
2013-10-30 22:40:09 +01:00
__END_DECLS
auto import from //depot/cupcake/@135843
2009-03-04 04:28:35 +01:00
libc: remove C++ comments from public headers. Change-Id: I4af84f912062cd2ff34711c25122fb323f20c032
2010-09-26 23:00:36 +02:00
#endif /* _PTHREAD_H_ */
Reference in a new issue Copy permalink