CVS log for src/lib/libc/gen/pthread_atfork.c

Up to [cvs.NetBSD.org] / src / lib / libc / gen

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Catch up with all the lint warnings since exit on warning was disabled.
Disable 'missing header declaration' and 'nested extern' warnings for now.

pthread_atfork(3): Block signals during the call to pthread_atfork.

This doesn't affect the calls to the atfork handlers -- it only
protects access to the lists of handlers from interruption by a
signal, in case the signal handler calls fork(2).

fix various typos in comments, documentation and messages.

Hook up proper fork lock handling for malloc:
- lock all relevant mutexes just before fork
- unlock all mutexes just after fork in the parent
- full reinit non-spinlocks in the child
This is not using the normal pthread_atfork interface to ensure order of
operation, malloc is used as implementation detail too often.

Pull up following revision(s) (requested by chs in ticket #907):

	libexec/ld.elf_so/rtld.c: revision 1.205
	libexec/ld.elf_so/rtld.h: revision 1.140
	libexec/ld.elf_so/symbols.map: revision 1.3
	libexec/ld.elf_so/symbols.map: revision 1.4
	lib/libc/gen/pthread_atfork.c: revision 1.13
	lib/libc/gen/pthread_atfork.c: revision 1.14
	libexec/ld.elf_so/rtld.h: revision 1.139
	libexec/ld.elf_so/rtld.c: revision 1.204

Introduce intermediate locking for fork, so that the dynamic linker is
in a consistent state. This most importantly avoids races between dlopen
and friends and fork, potentially resulting in dead locks in the child
when it itself tries to acquire locks.

Rename __atomic_fork to __locked_fork and give it &errno as argument.
rtld and libc use different storage, so the initial version would
incorrectly report the failure reason for fork().

There is still a small race condition inside ld.elf_so as it doesn't use
thread-safe errno internally, but that's a more contained internal
issue.

Sync with HEAD

Rename __atomic_fork to __locked_fork and give it &errno as argument.
rtld and libc use different storage, so the initial version would
incorrectly report the failure reason for fork().

There is still a small race condition inside ld.elf_so as it doesn't use
thread-safe errno internally, but that's a more contained internal
issue.

Introduce intermediate locking for fork, so that the dynamic linker is
in a consistent state. This most importantly avoids races between dlopen
and friends and fork, potentially resulting in dead locks in the child
when it itself tries to acquire locks.

Revert previous

'git grep' breaks now.

Switch atform allocations from malloc()+free() to mmap()+munmap()

This avoid bootstrapping malloc too early when libc+libpthread are not
ready. It is called through pthread__init() -> _pthread_atfork().

This also helps LLVM Leak Sanitizer to pacify false positive reports.

Fix non _REENTRANT build.

sync with head

Use C89 definitions.
Remove use of __P

sync with head.

Remove clause 3 and 4 from TNF licenses

file pthread_atfork.c was added on branch christos-time_t on 2008-04-28 20:23:00 +0000

Remove clause 3 and 4 from TNF licenses

sync with head.

sync with HEAD

Keep one builtin slot so that pthread initialization doesn't need to use
malloc().

sync with HEAD

fix indentation.

compat core reorg.

Pull up revision 1.4 (requested by lukem in ticket #394):
Ensure that we don't try to free random memory when another malloc fails,
by initializing newprepare, newparent, and newchild to NULL.
Detected by gcc -Wuninitialized.

Ensure that we don't try to free random memory when another malloc fails,
by initializing newprepare, newparent, and newchild to NULL.
Detected by gcc -Wuninitialized.

minor KNF, making it easier to find where fork() is implemented

Oops, make the atfork queues static so they don't pollute the
namespace.

Implement pthread_atfork() (in libc, because the required threadlib
stub behavior is exactly the same as the usual behavior).