Do a substantial rework of the GMutex and GCond APIs.
- remove all of the macro indirection hackery which is no longer needed
since we dropped support for switchable thread implementations
- expose the structure types and add G_MUTEX_INIT and G_COND_INIT
static initialiser macros
- add g_mutex_init() and g_mutex_clear() for use when embedding GMutex
into another structure type and do the same for GCond as well
- avoid infinite recursion hazards by ensuring that neither GCond or
GMutex ever calls back into any other part of GLib
- substantially rework the Windows implementation of GCond and GMutex
to use the SRWLock and CONDITION_VARIABLE APIs present on Windows
2008/Vista and later, emulating these APIs on XP
g_thread_gettime() is an undocumented public function pointer that
points to a function that returns the monotonic time in nanoseconds.
g_get_monotonic_time() does the same in microseconds, so it can be used
instead.
GLib had one internal user in GFileMonitor that only cared about
millisecond accuracy; it has been ported to g_get_monotonic_time().
G_THREADS_ENABLED still exists, but is always defined. It is still
possible to use libglib without threads, but gobject (and everything
above it) is now guaranteed to be using threads (as, in fact, it was
before, since it was accidentally impossible to compile with
--disable-threads).
https://bugzilla.gnome.org/show_bug.cgi?id=616754
Since gobject now initialized threads unconditionally if threads are
available it makes no sense to have each thread operation
conditionally check if threads are enables, so allow this to be avoided.
https://bugzilla.gnome.org/show_bug.cgi?id=606775
GLib users buildable with gcc 4.4. Patch by Jakub Jelinek.
* glib/gatomic.[hc]: Add G_GNUC_MAY_ALIAS to pointer arguments,
fix macro versions to only operate on objects of the same size.
* glib/gdataset.c:
* glib/gthread.[hc]:
* glib/gdatasetprivate.h: Remove unnecessary casts in
g_atomic_pointer_get calls.
svn path=/trunk/; revision=7875
* configure.in:
* glib/gthread.h: Revert previous patch as it doesn't improve the
situation and results in other warnings.
svn path=/trunk/; revision=7064
2008-05-05 Michael Natterer <mitch@imendio.com>
* glib/glib.h: #define __GLIB_H_INSIDE__ around including
everything.
* glib/*.h: check for that define instead of __G_LIB_H__ if
G_DISABLE_SINGLE_INCLUDES is defined.
* glib/gdatasetprivate.h: #include <glib.h> instead of
<glib/gdataset.h>
svn path=/trunk/; revision=6875
2008-03-14 Michael Natterer <mitch@imendio.com>
* glib/*.h: make it possible to disable single-file includes by
defining G_DISABLE_SINGLE_INCLUDES when building against GLib.
Approved by Tim Janik.
* glib/glib.h: include <glib/gslice.h>.
* glib/gi18n.h
* glib/gi18n-lib.h
* glib/gprintf.h: include <glib.h> so the above works when these
files are included without including <glib.h> first.
svn path=/trunk/; revision=6713
* configure.in:
* glib/gthread.h: Prevent the compiler from warning about breaking
strict-aliasing rules when using gcc 4.3 and G_LOCK on C sources.
svn path=/trunk/; revision=6691
Tue Aug 14 02:06:10 2007 Tim Janik <timj@imendio.com>
* glib/gthread.c (g_once_init_enter_impl): prevent race covered
by g_once_init_enter(), by checking for previous initializations
before entering initialisation branch.
* tests/onceinit.c: added multi-thread/multi-initializer stress test
using unoptimized g_once_init_enter_impl().
svn path=/trunk/; revision=5701
Fri Jul 13 01:01:46 2007 Tim Janik <timj@imendio.com>
* glib/gthread.[hc]: more atomic ops pointer cast fixes. this time it'll
work with atomic op macros *and* atomic op functions.
svn path=/trunk/; revision=5638
Tue Jul 10 12:24:35 2007 Tim Janik <timj@imendio.com>
* glib/gthread.[hc]: implemented g_once_init_enter(),
g_once_init_enter_impl() and g_once_init_leave(), based on a patch by
Antoine Tremblay, fixes#65041.
adapted exported inline function mechanism from gutils.[hc] for inlining
g_once_init_enter_impl() in gthread.[hc].
svn path=/trunk/; revision=5616
2007-01-16 Matthias Clasen <mclasen@redhat.com>
* glib/gthread.h:
* glib/gthread.c:
* glib/glib.symbols: Revert an accidental ABI break by
moving gettime out of the GThreadFunctions struct and making
it a separate variable. (#397139, Joe Marcus Clarke)
* gthread/*.c: Adapt.
svn path=/trunk/; revision=5279
2007-01-07 Matthias Clasen <mclasen@redhat.com>
Don't link glib against libpthread. (#393812)
* configure.in: Link gthread against librt, not glib itself.
* glib/gthread.h:
* glib/gthread.c: Add a new thread function, gettime.
* glib/gtimer.c: Use gettime instead of directly working with
the various system interfaces.
* gthread/gthread-impl.c:
* gthread/gthread-posix.c:
* gthread/gthread-win32.c: Implement gettime.
svn path=/trunk/; revision=5227
2006-05-10 Sebastian Wilhelmi <wilhelmi@google.com>
* glib/gthread.h, gthread/gthread-impl.c: Make the magic and
location arguments to the error-checking-mutex functions const and
do not write to them, as we might not own them. Clean up the
error-checking-mutex code quite a bit. (#335198, Chris Wilson)
2006-01-16 Matthias Clasen <mclasen@redhat.com>
* glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
Rework the strict aliasing fix to not break C++, pointed
out by Murray Cumming.
2005-12-27 Matthias Clasen <mclasen@redhat.com>
Fix#316221, Michal Benes, Stanislav Brabec;
* configure.in: Fix a strict aliasing problem in
g_static_mutex_get_mutex().
* glib/gthread.h: ...and in
g_static_mutex_get_mutex_impl_shortcut().
Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
* glib/gslice.[hc]: new slice allocator implementation.
* tests/slice-test.c: added random slice allocation test.
* glib/gthread.[hc]: removed newly added private thread mem API.
* glib/gthreadinit.h:
* glib/gmessages.c:
* glib/gthread.c:
* glib/gmem.c: divided glib threading initialisation into three phases,
initialisation where private keys and messaging are not available (only
needed by gmem.c), initialisation without messaging but private keys
available (gslice.c, gmessage.c), and full fledged initialisers that
server the rest of glib. initialisation functions got renamed to reflect
the limitations of their corresponding phases.
* glib/gmem.c: removed memchunk code, defer allocations to
g_slice_* instead.
* glib/gmem.[hc]: removed g_slice_* skeletons.
* glib/glib.symbols: added g_slice_* symbols.
* configure.in: check for availability of posix_memalign(3), memalign(3)
and valloc(3).
* glib/Makefile.am: added gslice.[hc].
Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
* glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
added g_slice_*() API to allocate and cache small bits of memory.
an actuall allocator implementation for g_slice_*() is still pending.
* glib/gthread.[hc]: changes from a patch by Matthias Clasen.
changed GRealThread list to use in-structure *next; fields instead
of GSList, in order for thread iteration to not depenend on g_slice_*()
indirectly.
_g_thread_mem_private_get():
_g_thread_mem_private_set(): added accessors for private memory,
needed because the ordinary GPrivate implementation relies on GArray
and GSList and therefore indirectly on working g_slice_*() allocations.
* glib/gthread.[hc]:
g_thread_foreach(): new public API function to loop over all existing threads.
* glib/gdataset.c:
* glib/gstring.c:
* glib/gcache.c:
* glib/garray.c:
* glib/gqueue.c:
* glib/gslist.c:
* glib/glist.c:
* glib/ghash.c:
* glib/gtree.c:
* glib/ghook.c:
* glib/gmain.c:
* glib/gnode.c:
removed GAllocator and free list usages and accompanying locks.
use g_slice_*() API to allocate and cache small bits of memory.
* glib/ghook.h: removed GMemChunk field from public API.
* glib/gslist.h:
* glib/glist.h: deprecate allocator API, provide _free1() for consistency.
* glib/gnode.h: deprecate allocator API.
* glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
be used for poll rec lists.
* glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
g_relation_destroy(): free all tuples from the all_tuples hash table,
this effectively maintains the life time track keeping of tuples.
g_relation_delete_tuple(): free tuples which are removed from the
all_tuples hash table. this fixes a temporary leak that was present
in the memchunk code until the destruction of the relation.
2004-02-29 Sebastian Wilhelmi <seppi@seppi.de>
* configure.in, glib/gatomic.c, glib/gatomic.h: Moved the
assembler functions from gatomic.h to gatomic.c, which makes for
better maintainability. Also use gint instead of gint32 to be able
to use reference counting for ABI-fixed structures with
gint/guint.
* glib/gthread.h: Adapted accordingly.
* tests/atomic-test.c: Updated to test for G_MAXINT and G_MININT.
2004-02-26 Sebastian Wilhelmi <seppi@seppi.de>
* glib/gatomic.h: empty G_ATOMIC_MEMORY_BARRIER() definition for
!G_THREADS_ENABLED
* glib/gthread.h: Define g_once and
g_static_mutex_get_mutex_impl_shortcut in terms of
G_MEMORY_BARRIER, if defined and if we can inline.
* configure.in: Remove double checked locking feature check.
2003-07-09 Matthias Clasen <maclas@gmx.de>
Support for one-time initialization functions. (#69668, Sebastian Wilhelmi)
* configure.in: Check whether double checked locking is safe, define g_once() in
glibconfig.h accordingly.
* glib/gthread.h: Add GOnce, GOnceStatus, G_ONCE_INIT and g_once_impl.
* glib/gthread.c (g_once_impl): Fallback implementation using a mutex if double checked
locking is unsafe.
* tests/thread-test.c: Add tests for g_once().
Sat Dec 14 21:10:57 2002 Owen Taylor <otaylor@redhat.com>
* glib/gthread.h: Mark the contents of the strucures
in this file /*< private >*/
* glib/gthread.[ch]: Rename the 'write' field of the
structure to 'have_writer' to avoid any possible
conflict with system headers. (#90549, Morten Welinder)
2002-02-04 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* glib/gthread.c (g_thread_create_full): Delay allocation until
after all g_return_val_if_fail ().
* glib/gthread.h: Make depth member guint for cosmetic reasons.
* glib/gthread.c: (g_static_rec_mutex_unlock_full): depth should
be unsigned. All that spotted by Jrgen Viksell
<jorgen.viksell@telia.com>
2001-09-19 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* glib/gthread.h (GThreadFunctions): Add thread_equal function to
allow for platform defined function to compare two threads.
* glib/gthread.c: Use g_thread_functions_for_glib_use.thread_equal
when non-NULL instead of ==.
* gthread/gthread-posix.c: Add g_thread_equal_posix_impl and
add to the function vector g_thread_functions_for_glib_use_default.
* gthread/gthread-solaris.c, gthread/gthread-win32.c: Add NULL
as equal function, as on those two platforms you don't need an
equal function.
2001-08-30 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* glib/gthread.h, glib/gthread.c: Add 'want_to_read' to
GStaticRWLock to avoid calling g_cond_broadcast, when no one is
waiting.
Tue Jun 26 11:43:46 2001 Owen Taylor <otaylor@redhat.com>
* configure.in Makefile.am *.[ch] glib/*.[ch] glib/Makefile.am:
Move glib library into a subdirectory, make all GLib include
files include as <glib/glist.h>
* tests/testglib.c tests/testgdate.c tests/testgdateparser.c
tests/timeloop.c tests/timeloop-basic.c: Move all tests into
the tests/ subdirectory.
2001-05-18 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread.c, gthread.h: Renamed g_thread_create to
g_thread_create_full and added macro g_thread_create, which omits
'stack_size', 'bound' and 'priority' parameters. Also removed
'bound' from GThread struct.
* gthreadpool.h, gthreadpool.c: Adapted GThreadPool to the above
changes. GThreadPool lost the 'priority' and 'bound'
members. g_thread_pool_new the 'stack_size', 'bound' and
'priority' parameters.
* tests/mainloop-test.c, tests/thread-test.c,
tests/threadpool-test.c: Adapted to the above changes.
2001-05-09 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread.c, gthread.h: Renamed 'value' and 'arg' to 'data' and
'thread_func' to 'func' to make it more consistent with the rest
of GLib.
* gthreadpool.c, gthreadpool.h: Moved 'stack_size' from the public
members of GThreadPool to the private ones. Renamed 'thread_func'
to 'func' as above. Moved up 'user_data' in g_thead_pool_new
argument list and in GThreadPool struct.
2001-05-08 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gmain.c, gthread.c, gthread.h: Moved func and arg members from
GRealThread to GThread, such that they can be accessed by the
user.
* gthread.c, gthread.h: Due to popular demand (Tim being the
populus here ;-) threads now have a 'return value', which is
returned by g_thread_join and is either the return of the topmost
thread function or the value given to g_thread_exit.
* gthreadpool.c, tests/mainloop-test.c, tests/thread-test.c:
Adapted to the above change.
2001-04-02 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gmain.c: Use the new GRealThread member "context" instead of a
GStaticPrivate to store the thread specific main loop context.
* gthread.c: Added "context" member to GRealThread and updated
g_thread_create, g_thread_self and g_thread_cleanup accordingly.
* gthread.c, gthread.h: Removed the functions
g_static_private_(get|set)_for_thread and adapted
g_static_private_(get|set) and g_static_private_free accordingly.
2001-02-23 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread.h (struct _GThread): Change the order to match the order
in g_thread_create().
* gthread.c (g_static_rec_mutex_lock_full): Also do the right
thing (behave like 'depth' calls to g_static_rec_mutex_lock) for a
mutex, that is already locked.
* glib/tmpl/threads.sgml: Big update. Almost ready.
* glib/tmpl/async_queues.sgml: Typo.
* glib/glib-sections.txt: Added g_static_rec_mutex_init,
g_static_rec_mutex_free, g_static_rw_lock_init,
g_static_private_init and g_static_private_free.
* glib/glib-overrides.txt: Added g_thread_yield and g_thread_exit.
2001-02-21 Tor Lillqvist <tml@iki.fi>
* gutils.c (g_path_is_absolute): Match also UNC paths on Win32.
* gthread.h: Include gtypes.h to be sure to get GLIB_VAR
definition, remove definition from here.
* gunicode.h: Remove duplicate GLIB_VAR definition from here, too.
* gutils.c: (Win32) Use USERPROFILE as home dir if present. On
Win2k, HOMEDRIVE and HOMEPATH aren't reliable.
* Makefile.am (INCLUDES): Add -DGLIB_COMPILATION.
* makefile.mingw.in: Remove install target, Windows isn't Unix.
* testgdate.c
* testgdateparser.c
* testglib.c: Undefine GLIB_COMPILATION.
* glib.def: Updates.
* .cvsignore
* */.cvsignore: Ignore also .obj, .dll, .lib and .exe files.
2001-02-13 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread.c, gthread.h: Added functions g_static_rec_mutex_init,
g_static_rec_mutex_free, g_static_private_init,
g_static_private_free, g_static_rw_lock_init to allow the usage of
all those types dynamically. (g_static_rw_lock_free already
existed). Aditionally freed static_private indeces are reused
now. Untill now the array would just grow if you would use more
and more static_private. That required adding a slist of all
running threads, which could potentially be of good use later. It
is not exported however. Renamed a LOCK and small indentation
fixes.
* tests/thread-test.c: Test the new static_private index freing
and reusing feature.
2001-02-01 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread.c, gthread.h: Added g_static_mutex_init to allow
initialization of a GStaticMutex, that can not be initialized with
G_STATIC_MUTEX_INIT, for example in allocated structures.
* glib/tmpl/threads.sgml, glib/glib-sections.txt: Added
documentation for g_static_mutex_init().
WCVS: ----------------------------------------------------------------------
2001-01-30 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread/gthread-impl.c (g_thread_init_with_errorcheck_mutexes): Call
g_thread_impl_init(), as g_thread_init won't call it.
* glib/glib-sections.txt, glib/tmpl/threads.sgml: Added
g_static_mutex_free
* gthread.c, gthread.h: Added g_static_mutex_free to allow using
GStaticMutexes with limited lifetime without leaking.
* configure.in: GStaticMutex doesn't have to provide extra space
for debugging information for G_ERRORCHECK_MUTEXES, as then the
non-default implementation (runtime_mutex) is used anyway.
2001-01-30 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread.h (g_cond_wait): Added debug information for
g_mutex_free calls, if G_ERRORCHECK_MUTEXES is defined.
* gthread/gthread-impl.c (g_mutex_free_errorcheck_impl): Fixed it for
real. Sorry for this mess. It looked like a real obvious fix, so I
didn't check. Bad boy. Added some casts to quiet the compiler.
Tue Dec 5 12:23:04 2000 Owen Taylor <otaylor@redhat.com>
* gmain.[hc]: Major change in API for creating sources
to handle multiple main loops (GMainContext *).
GSources are now exposed as GSource * and implemented
with structure derivation.
* giochannel.[ch]: Changed vtable for GIOChannel to correspond
to the new mainloop API, add g_io_channel_create_watch().
* gtypes.h: Move GTimeVal here.
* gthread.h: Remove gmain.h include to avoid circularity.
* giounix.c: Update for new GMain API.
* giowin32.c: Update for new GMain API. (No check for
proper compilation or working.)
* timeloop.c timeloop-basic.c: A benchmarking program for
the main loop comparing the main loop against a
hand-written (timeloop-basic.c) variant.
* tests/mainloop-test.c: New torture test of mainloop.
* docs/Changes-2.0.txt: Started. Added text about
changes to GMain.
* gmain.c (g_main_add_poll_unlocked): Initial fd->revents
to zero. (#8482, Benjamin Kahn)
2000-11-28 Sebastian Wilhelmi <wilhelmi@ira.uka.de>
* gthread-impl.c: Revamped errorcheck mutexes and added errorcheck
cond_wait() and cond_timed_wait() funtions. This makes he whole
thing work. Now we only show the location of the locking/unlocking
for -DG_ERRORCHECK_MUTEXES and not the name of the mutex.
* gthread.c: Set the thread data before locking the mutex, because
the locking call might use g_thread_self ().
* gthread.h: Do only show the location of the locking/unlocking
for -DG_ERRORCHECK_MUTEXES and not the name of the mutex. Add the
errorcheck capability for g_cond_wait and g_cond_timed_wait as
well.
