platform/upstream/glib.git
18 years ago- Added new API g_thread_pool_get_idle_time() and
Martyn James Russell [Tue, 3 Jan 2006 15:09:52 +0000 (15:09 +0000)]
- Added new API g_thread_pool_get_idle_time() and

* docs/reference/glib/glib-sections.txt:
* glib/glib.symbols:
* glib/gthreadpool.[ch]:
- Added new API g_thread_pool_get_idle_time() and
g_thread_pool_set_idle_time(). (#324228).

* tests/threadpool-test.c:
- Updated test case to do thread pool sorting, thread pool with
no sorting and a thread pool with idle thread timeouts.

18 years agoAdd new functions
Matthias Clasen [Tue, 3 Jan 2006 14:52:34 +0000 (14:52 +0000)]
Add new functions

18 years agoSmall fixes
Matthias Clasen [Tue, 3 Jan 2006 12:01:05 +0000 (12:01 +0000)]
Small fixes

18 years agonew method to determine if the current thread is the owner of the context.
Michael Meeks [Tue, 3 Jan 2006 10:24:07 +0000 (10:24 +0000)]
new method to determine if the current thread is the owner of the context.

2005-12-20  Michael Meeks  <michael.meeks@novell.com>

* glib/gmain.c (g_main_context_is_owner): new method
to determine if the current thread is the owner of the
context.

18 years agoDon't provide g_rmdir() as a macro expanding to rmdir, since rmdir is not
Matthias Clasen [Tue, 3 Jan 2006 02:53:05 +0000 (02:53 +0000)]
Don't provide g_rmdir() as a macro expanding to rmdir, since rmdir is not

2006-01-02  Matthias Clasen  <mclasen@redhat.com>

* glib/glib.symbols:
* glib/gstdio.h:
* glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
expanding to rmdir, since rmdir is not declared in a portable
system header we can include in gstdio.h.  (#325249, Jani Monoses)

18 years agotypo
Stepan Kasal [Mon, 2 Jan 2006 13:50:15 +0000 (13:50 +0000)]
typo

18 years agoUpdated Norwegian bokmål translation. Same
Kjartan Maraas [Sun, 1 Jan 2006 23:33:07 +0000 (23:33 +0000)]
Updated Norwegian bokmål translation. Same

2006-01-02  Kjartan Maraas  <kmaraas@gnome.org>

* nb.po: Updated Norwegian bokmål translation.
* no.po: Same

18 years agoSet the GIOChannels for stdout and stderr to unbuffered. Otherwise the
Tor Lillqvist [Sun, 1 Jan 2006 11:15:56 +0000 (11:15 +0000)]
Set the GIOChannels for stdout and stderr to unbuffered. Otherwise the

2006-01-01  Tor Lillqvist  <tml@novell.com>

* glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
stdout and stderr to unbuffered. Otherwise the giochannel layer
will try to read from them regardless whether the
g_io_channel_win32_poll() call here has indicated
readability or not. (#325310)

18 years agoRemove the handler trash stack, which is now unused. (g_signal_init):
Matthias Clasen [Sat, 31 Dec 2005 02:44:53 +0000 (02:44 +0000)]
Remove the handler trash stack, which is now unused. (g_signal_init):

2005-12-30  Matthias Clasen  <mclasen@redhat.com>

* gsignal.c: Remove the handler trash stack, which
is now unused.
(g_signal_init): Remove the restriction that HandlerMatch
must be the same size as GList, and obsolete comments
referring to mem chunks.

18 years agoRemove the handler trash stack, which is now unused.
Matthias Clasen [Fri, 30 Dec 2005 07:32:05 +0000 (07:32 +0000)]
Remove the handler trash stack, which is now unused.

2005-12-30  Matthias Clasen  <mclasen@redhat.com>

* gsignal.c: Remove the handler trash stack, which
is now unused.

18 years agoFix a copy-and-paste error in the docs. (#325273, Declan Naughton)
Matthias Clasen [Fri, 30 Dec 2005 02:08:02 +0000 (02:08 +0000)]
Fix a copy-and-paste error in the docs. (#325273, Declan Naughton)

2005-12-29  Matthias Clasen  <mclasen@redhat.com>

* glib/gutils.c (glib_check_version): Fix a copy-and-paste error
in the docs.  (#325273, Declan Naughton)

18 years agoShut up CVS
Manish Singh [Tue, 27 Dec 2005 23:37:28 +0000 (23:37 +0000)]
Shut up CVS

18 years agouse LC_ALL instead of LC_COLLATE, to make sure we really override things.
Manish Singh [Tue, 27 Dec 2005 23:35:52 +0000 (23:35 +0000)]
use LC_ALL instead of LC_COLLATE, to make sure we really override things.

2005-12-27  Manish Singh  <yosh@gimp.org>

        * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
        make sure we really override things.

18 years agoFix #316221, Michal Benes, Stanislav Brabec;
Matthias Clasen [Tue, 27 Dec 2005 19:50:02 +0000 (19:50 +0000)]
Fix #316221, Michal Benes, Stanislav Brabec;

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().

18 years agoAdd a cast to silence compiler warnings. (#321978, Andrew Paprocki)
Matthias Clasen [Tue, 27 Dec 2005 19:21:36 +0000 (19:21 +0000)]
Add a cast to silence compiler warnings. (#321978, Andrew Paprocki)

2005-12-27  Matthias Clasen  <mclasen@redhat.com>

        * glib/gdatasetprivate.h: Add a cast to silence compiler
        warnings.  (#321978, Andrew Paprocki)

18 years agoPartial fix for bug #32937, Bogdan Nicula.
Matthias Clasen [Tue, 27 Dec 2005 19:15:12 +0000 (19:15 +0000)]
Partial fix for bug #32937, Bogdan Nicula.

2005-12-27  Matthias Clasen  <mclasen@redhat.com>

        Partial fix for bug #32937, Bogdan Nicula.

        * configure.in: Check for malloc.h
        * glib/gslice.c: Don't include malloc.h unconditionally.

18 years agoexport LC_COLLATE so it takes. (#324950, Dan Yefimov)
Manish Singh [Tue, 27 Dec 2005 08:23:08 +0000 (08:23 +0000)]
export LC_COLLATE so it takes. (#324950, Dan Yefimov)

2005-12-27  Manish Singh  <yosh@gimp.org>

        * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
        (#324950, Dan Yefimov)

18 years agoOn Win32, include process.h (#325015, Kazuki Iwamoto)
Matthias Clasen [Mon, 26 Dec 2005 23:18:09 +0000 (23:18 +0000)]
On Win32, include process.h (#325015, Kazuki Iwamoto)

2005-12-26  Matthias Clasen  <mclasen@redhat.com>

* glib/gslice.c: On Win32, include process.h  (#325015, Kazuki
Iwamoto)

18 years agoUpdated Finnish translation
Ilkka Tuohela [Mon, 26 Dec 2005 09:02:48 +0000 (09:02 +0000)]
Updated Finnish translation

18 years agoUpdates
Matthias Clasen [Mon, 26 Dec 2005 05:23:08 +0000 (05:23 +0000)]
Updates

18 years agoFix C99isms. (#324950, Dan Yefimov)
Matthias Clasen [Mon, 26 Dec 2005 04:59:34 +0000 (04:59 +0000)]
Fix C99isms. (#324950, Dan Yefimov)

2005-12-25  Matthias Clasen  <mclasen@redhat.com>

        * tests/slice-test.c: Fix C99isms.  (#324950, Dan Yefimov)

18 years agoUpdated Canadian English translation.
Adam Weinberger [Sun, 25 Dec 2005 06:36:31 +0000 (06:36 +0000)]
Updated Canadian English translation.

2005-12-25  Adam Weinberger  <adamw@gnome.org>

* en_CA.po: Updated Canadian English translation.

18 years agoadditions
Matthias Clasen [Sat, 24 Dec 2005 07:49:51 +0000 (07:49 +0000)]
additions

18 years agoOne more
Matthias Clasen [Sat, 24 Dec 2005 07:41:19 +0000 (07:41 +0000)]
One more

18 years agoAdd new functions.
Matthias Clasen [Sat, 24 Dec 2005 07:38:26 +0000 (07:38 +0000)]
Add new functions.

2005-12-24  Matthias Clasen  <mclasen@redhat.com>

* gobject.symbols: Add new functions.

18 years agorenamed GUnowned to GInitiallyUnowned.
Tim Janik [Thu, 22 Dec 2005 17:58:21 +0000 (17:58 +0000)]
renamed GUnowned to GInitiallyUnowned.

Thu Dec 22 18:53:14 2005  Tim Janik  <timj@gtk.org>

        * gobject.[hc]: renamed GUnowned to GInitiallyUnowned.

18 years agoimplemented G_TYPE_GTPYE. applied patch from matthias which implements
Tim Janik [Thu, 22 Dec 2005 15:07:03 +0000 (15:07 +0000)]
implemented G_TYPE_GTPYE. applied patch from matthias which implements

Thu Dec 22 14:59:24 2005  Tim Janik  <timj@imendio.com>

        * gvaluetypes.[hc]: implemented G_TYPE_GTPYE. applied patch
        from matthias which implements GType accessors for GValue.

        * gparamspecs.[hc]: applied patch from matthias which
        implements G_TYPE_PARAM_GTYPE.

        * gobject.[hc]:
        GUnowned: introduced a new object type that has an initially
        floating reference.
        g_object_compat_control(): allow setting of a floating flag handler.

18 years agoUpdates
Matthias Clasen [Wed, 21 Dec 2005 06:33:37 +0000 (06:33 +0000)]
Updates

18 years agoNew function to sort tasks pushed into a threadpool. (#324479, Martyn
Matthias Clasen [Wed, 21 Dec 2005 04:45:56 +0000 (04:45 +0000)]
New function to sort tasks pushed into a threadpool. (#324479, Martyn

2005-12-20  Matthias Clasen  <mclasen@redhat.com>

* glib/glib.symbols:
* glib/gthreadpool.h:
* glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
to sort tasks pushed into a threadpool.  (#324479, Martyn Russell)

* tests/threadpool-test.c: Test this.

18 years agoadded mem_error() and mem_assert() to test and handle errors without
Tim Janik [Tue, 20 Dec 2005 17:35:02 +0000 (17:35 +0000)]
added mem_error() and mem_assert() to test and handle errors without

Tue Dec 20 18:14:14 2005  Tim Janik  <timj@imendio.com>

        * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
        handle  errors without depending on gmessage.c which might not be
        setup when the error occours.
        removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
        always freeing can be achieved by adjusting the working set time to
        0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
        added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
        (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
        minimize wastage if small amounts of differently sized structrues are
        allocated, this does come at a performance cost of roughly 5% though.
        fixed up block alignment calculation, so it works for varying
        block sizes. only use strerror() not g_strerror() since the latter
        depends on working GQuark and GSlice.
        mem_error(): implemented in terms of fprintf and vfprintf.

        * tests/slice-color.c: new program to test cache colorization effects.

        * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
        G_SLICE_CONFIG_WORKING_SET_MSECS.

18 years agoUpdates
Matthias Clasen [Tue, 20 Dec 2005 04:44:25 +0000 (04:44 +0000)]
Updates

18 years agoDocument G_TYPE_HASH_TABLE.
Matthias Clasen [Mon, 19 Dec 2005 21:31:49 +0000 (21:31 +0000)]
Document G_TYPE_HASH_TABLE.

2005-12-19  Matthias Clasen  <mclasen@redhat.com>

* gobject/tmpl/gboxed.sgml: Document G_TYPE_HASH_TABLE.

18 years agoAdd g_list_free1
Matthias Clasen [Mon, 19 Dec 2005 21:22:29 +0000 (21:22 +0000)]
Add g_list_free1

* glib/glib-sections.txt: Add g_list_free1

* glib/glib-overrides.txt: Remove G_THREADS_IMPL_SOLARIS

18 years agoUpdated French translation.
Christophe Merlet [Mon, 19 Dec 2005 18:38:44 +0000 (18:38 +0000)]
Updated French translation.

18 years ago*** empty log message ***
Ignacio Casal Quinteiro [Sun, 18 Dec 2005 12:16:43 +0000 (12:16 +0000)]
*** empty log message ***

18 years agoSet an error in all failure cases. (#324332, Tim-Philipp Müller)
Matthias Clasen [Sun, 18 Dec 2005 02:46:27 +0000 (02:46 +0000)]
Set an error in all failure cases. (#324332, Tim-Philipp Müller)

2005-12-17  Matthias Clasen  <mclasen@redhat.com>

        * glib/goption.c (parse_short_option): Set an error in all
        failure cases.  (#324332, Tim-Philipp Müller)

18 years agoFix memory barrier position in g_atomic_int_get and g_atomic_pointer_get.
Sebastian Wilhelmi [Sat, 17 Dec 2005 12:20:50 +0000 (12:20 +0000)]
Fix memory barrier position in g_atomic_int_get and g_atomic_pointer_get.

2005-12-17  Sebastian Wilhelmi  <seppi@seppi.de>

* glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
and g_atomic_pointer_get. Add g_atomic_int_set and
g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
as functions (additionally to the macros in the header) for the
!G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.

18 years agoDon't return a pointer to a const struct, since apps expect to be able to
Matthias Clasen [Fri, 16 Dec 2005 21:13:18 +0000 (21:13 +0000)]
Don't return a pointer to a const struct, since apps expect to be able to

2005-12-16  Matthias Clasen  <mclasen@redhat.com>

        *  glib/gmem.c (g_allocator_new): Don't return a pointer to
        a const struct, since apps expect to be able to modify it.
        (#324179, J. Ali Harlow)

18 years agofixed conditional symbols definitions
Tim Janik [Thu, 15 Dec 2005 10:01:02 +0000 (10:01 +0000)]
fixed conditional symbols definitions

18 years agoDocument new atomic operations.
Matthias Clasen [Thu, 15 Dec 2005 04:37:32 +0000 (04:37 +0000)]
Document new atomic operations.

2005-12-14  Matthias Clasen  <mclasen@redhat.com>

* glib/glib-sections.txt:
* glib/tmpl/atomic_operations.sgml: Document new atomic
operations.

18 years agoadded g_atomic_pointer_set() and g_atomic_int_set()
Tim Janik [Thu, 15 Dec 2005 02:23:15 +0000 (02:23 +0000)]
added g_atomic_pointer_set() and g_atomic_int_set()

Tue Dec 13 10:13:32 2005  Tim Janik  <timj@imendio.com>

* glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()

18 years agoadded g_atomic_pointer_set() and g_atomic_int_set() for systems where the
Tim Janik [Tue, 13 Dec 2005 09:15:52 +0000 (09:15 +0000)]
added g_atomic_pointer_set() and g_atomic_int_set() for systems where the

Tue Dec 13 10:13:32 2005  Tim Janik  <timj@imendio.com>

        * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
        for systems where the initialization of atomic variables requires a
        write memory barrier.

18 years agocorrected floating reference documentation.
Tim Janik [Mon, 12 Dec 2005 14:32:27 +0000 (14:32 +0000)]
corrected floating reference documentation.

Mon Dec 12 15:31:41 2005  Tim Janik  <timj@imendio.com>

        * gobject/tmpl/objects.sgml: corrected floating reference documentation.

18 years agoBump version
Matthias Clasen [Fri, 9 Dec 2005 20:00:06 +0000 (20:00 +0000)]
Bump version

18 years ago2.9.1 GLIB_2_9_1
Matthias Clasen [Fri, 9 Dec 2005 18:05:25 +0000 (18:05 +0000)]
2.9.1

18 years agofix type inconsistency pointed out by gmorten
Tim Janik [Fri, 9 Dec 2005 16:34:26 +0000 (16:34 +0000)]
fix type inconsistency pointed out by gmorten

18 years agoUse G_GNUC_WARN_UNUSED_RESULT on list functions that return the whole
Alexander Larsson [Fri, 9 Dec 2005 13:58:57 +0000 (13:58 +0000)]
Use G_GNUC_WARN_UNUSED_RESULT on list functions that return the whole

2005-12-09  Alexander Larsson  <alexl@redhat.com>

* glib/glist.h:
* glib/gslist.h:
Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
the whole list.

18 years agoUpdates
Matthias Clasen [Thu, 8 Dec 2005 05:00:35 +0000 (05:00 +0000)]
Updates

2005-12-08  Matthias Clasen  <mclasen@redhat.com>

* NEWS: Updates

18 years agoDocument n_preallocs as ignored.
Matthias Clasen [Wed, 7 Dec 2005 15:57:28 +0000 (15:57 +0000)]
Document n_preallocs as ignored.

2005-12-07  Matthias Clasen  <mclasen@redhat.com>

* gobject/tmpl/gparamspec.sgml:
* gobject/tmpl/gtype.sgml: Document n_preallocs as
ignored.

18 years agoremoved commit leftover
Tim Janik [Wed, 7 Dec 2005 15:48:37 +0000 (15:48 +0000)]
removed commit leftover

18 years ago- Call g_queue_insert_sorted() instead of duplicating the code. - Call
Martyn James Russell [Wed, 7 Dec 2005 12:09:44 +0000 (12:09 +0000)]
- Call g_queue_insert_sorted() instead of duplicating the code. - Call

* glib/gasyncqueue.c:
- Call g_queue_insert_sorted() instead of duplicating the code.
- Call g_queue_sort() instead of duplicating the code.
- Invert sort function results to make sure the same sort function
gives the same results across glist, gslist, gqueue and
gasyncqueue.

* tests/asyncqueue-test.c:
- Updated the sort function to reflect the example in the
documentation for gasyncqueue.c.

18 years agoMissed the ChangeLog
Martyn James Russell [Wed, 7 Dec 2005 11:36:48 +0000 (11:36 +0000)]
Missed the ChangeLog

18 years ago- Added g_list_insert_sorted_with_data () and
Martyn James Russell [Wed, 7 Dec 2005 11:35:27 +0000 (11:35 +0000)]
- Added g_list_insert_sorted_with_data () and

* docs/reference/glib/glib-sections.txt:
* docs/reference/glib/tmpl/linked_lists_double.sgml:
* docs/reference/glib/tmpl/linked_lists_single.sgml:
* glib/glist.[ch]:
* glib/gslist.[ch]:
- Added g_list_insert_sorted_with_data () and
g_slist_insert_sorted_with_data ().
- Removed the extra check in g_list_sort() and g_slist_sort() for
GCompareDataFunc vs. GCompareFunc.

18 years agoLook up prefix at run-time on Win32, assuming the standard directory
Tor Lillqvist [Wed, 7 Dec 2005 03:57:14 +0000 (03:57 +0000)]
Look up prefix at run-time on Win32, assuming the standard directory

2005-12-07  Tor Lillqvist  <tml@novell.com>

* glib-gettextize.in: Look up prefix at run-time on Win32,
assuming the standard directory structure with glib-gettextize in
$prefix/bin.

18 years agoExempt GTK+ from the mem chunk deprecation, since we need GTK+ 2.8 to
Matthias Clasen [Tue, 6 Dec 2005 14:16:27 +0000 (14:16 +0000)]
Exempt GTK+ from the mem chunk deprecation, since we need GTK+ 2.8 to

2005-12-06  Matthias Clasen  <mclasen@redhat.com>

* glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
we need GTK+ 2.8 to compile against GLib 2.10.

18 years agoset LC_COLLATE, not LANG, to be sure to override any user settings.
Manish Singh [Mon, 5 Dec 2005 21:45:06 +0000 (21:45 +0000)]
set LC_COLLATE, not LANG, to be sure to override any user settings.

2005-12-05  Manish Singh  <yosh@gimp.org>

        * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
        to override any user settings.

18 years agoRevert an accidental commit
Matthias Clasen [Mon, 5 Dec 2005 21:09:20 +0000 (21:09 +0000)]
Revert an accidental commit

18 years agoDocumentation updates. (#323291, Morten Welinder)
Matthias Clasen [Mon, 5 Dec 2005 19:09:00 +0000 (19:09 +0000)]
Documentation updates. (#323291, Morten Welinder)

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* glib/gutf8.c: Documentation updates.  (#323291, Morten Welinder)

18 years agoUse G_FLAGS_CLASS_TYPE. (#323273, Kalle Vahlman)
Matthias Clasen [Mon, 5 Dec 2005 18:01:51 +0000 (18:01 +0000)]
Use G_FLAGS_CLASS_TYPE. (#323273, Kalle Vahlman)

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* genums.h (G_FLAGS_CLASS_TYPE_NAME): Use G_FLAGS_CLASS_TYPE.
(#323273, Kalle Vahlman)

18 years agobroken commit
Matthias Clasen [Mon, 5 Dec 2005 16:39:34 +0000 (16:39 +0000)]
broken commit

18 years agoSet MALLOC_CHECK_ and MALLOC_PERTURB_
Matthias Clasen [Mon, 5 Dec 2005 16:35:38 +0000 (16:35 +0000)]
Set MALLOC_CHECK_ and MALLOC_PERTURB_

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* tests/Makefile.am (TESTS_ENVIRONMENT):
* tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
MALLOC_CHECK_ and MALLOC_PERTURB_

* tests/run-collate-tests.sh: Run the collation tests explicitly
in en_US locale. (#320463)

18 years agoUpdates
Matthias Clasen [Mon, 5 Dec 2005 16:04:28 +0000 (16:04 +0000)]
Updates

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* glib/glib-sections.txt: Updates

* glib/tmpl/memory_slices.sgml: Fix a small formatting
problem.

18 years agofix a formatting problem
Matthias Clasen [Mon, 5 Dec 2005 15:54:58 +0000 (15:54 +0000)]
fix a formatting problem

18 years agoDeprecate mem chunks
Matthias Clasen [Mon, 5 Dec 2005 15:54:44 +0000 (15:54 +0000)]
Deprecate mem chunks

18 years agoAdd versioned deprecation information
Matthias Clasen [Mon, 5 Dec 2005 15:38:54 +0000 (15:38 +0000)]
Add versioned deprecation information

18 years ago- Added support for sorting async queues by with _push_sorted(),
Martyn James Russell [Mon, 5 Dec 2005 15:08:45 +0000 (15:08 +0000)]
- Added support for sorting async queues by with _push_sorted(),

* docs/reference/glib/glib-sections.txt:
* glib/gasyncqueue.[ch]:
- Added support for sorting async queues by with _push_sorted(),
_push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).

* tests/Makefile.am:
* tests/asyncqueue-test.c:
- Added test case for gasyncqueue.c

18 years agoimplement chain walking for arbitrary ->next pointer offsets in
Tim Janik [Mon, 5 Dec 2005 15:01:27 +0000 (15:01 +0000)]
implement chain walking for arbitrary ->next pointer offsets in

Mon Dec  5 15:53:20 2005  Tim Janik  <timj@imendio.com>

        * glib/gslice.c: implement chain walking for arbitrary ->next pointer
        offsets in g_slice_free_chain_with_offset() based on a patch by behdad
        in bug 323178. moved time consuming logic from g_slice_free() out of
        the inner loop, so g_slice_free_chain_with_offset() provides a real
        performance benefit over g_slice_free1() now.

        * glib/gslice.h: renamed g_slice_free_chain() to
        g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
        a type-safe macro as suggested in bug 323178.
        simplified the macro implementation of g_slice_free() and implemented
        it in a type safe manner for all compliers as suggested by Morten
        Welinder <mortenw@gnome.org>.

        * glib/gmain.c:
        * glib/glist.c:
        * glib/gslist.c:
        * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/

18 years agoAdd some docs.
Matthias Clasen [Mon, 5 Dec 2005 13:50:39 +0000 (13:50 +0000)]
Add some docs.

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* glib/gasyncqueue.c: Add some docs.

18 years agoFix compiler warnings.
Matthias Clasen [Mon, 5 Dec 2005 13:37:55 +0000 (13:37 +0000)]
Fix compiler warnings.

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* tests/libmoduletestplugin_a.c: Fix compiler warnings.

* glib/gatomic.c: In the ia64 implementation, use
__sync builtin without _si or _di suffix.  (#321229,
Stanislav Brabec, patch by Andreas Schwab)

18 years agoUpdates
Matthias Clasen [Mon, 5 Dec 2005 06:01:10 +0000 (06:01 +0000)]
Updates

2005-12-05  Matthias Clasen  <mclasen@redhat.com>

* gobject/tmpl/generic_values.sgml:
* glib/tmpl/trees-nary.sgml:
* glib/tmpl/linked_lists_double.sgml:
* glib/tmpl/linked_lists_single.sgml:
* glib/tmpl/memory_chunks.sgml:
* glib/tmpl/allocators.sgml:
* glib/tmpl/macros_misc.sgml: Updates

* glib/Makefile.am: Ignore gmirroringtable.h

18 years agoUpdate for latest header additions
Matthias Clasen [Mon, 5 Dec 2005 05:03:11 +0000 (05:03 +0000)]
Update for latest header additions

18 years agoRemove comma at the end of enum.
Behdad Esfahbod [Mon, 5 Dec 2005 00:45:55 +0000 (00:45 +0000)]
Remove comma at the end of enum.

2005-12-04  Behdad Esfahbod  <behdad@gnome.org>

        * glib/gslice.h: Remove comma at the end of enum.

18 years agoHandle multiple user names with the same UID better. (#319535, Laszlo
Matthias Clasen [Sun, 4 Dec 2005 07:00:50 +0000 (07:00 +0000)]
Handle multiple user names with the same UID better. (#319535, Laszlo

2005-12-04  Matthias Clasen  <mclasen@redhat.com>

Handle multiple user names with the same UID better.
(#319535, Laszlo Peter)

* glib/gutils.c (g_get_any_init_do): When determining user
data, first look up $LOGNAME. If the UID doesn't match
getuid(), fall back to the current behaviour of looking
up the user data based on getuid().

18 years agoAdd a macro to make gcc warn if a function result is ignored. (#145466,
Matthias Clasen [Sun, 4 Dec 2005 06:41:58 +0000 (06:41 +0000)]
Add a macro to make gcc warn if a function result is ignored. (#145466,

2005-12-04  Matthias Clasen  <mclasen@redhat.com>

* glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
to make gcc warn if a function result is ignored.  (#145466,
Arjan van de Ven, Alex Larsson)

* glib/gmem.h: Add the new attribute to g_realloc and
g_try_realloc.

18 years agoDeprecate g_cache_value_foreach. (#322956, Nicolas Caniart)
Matthias Clasen [Sat, 3 Dec 2005 06:53:31 +0000 (06:53 +0000)]
Deprecate g_cache_value_foreach. (#322956, Nicolas Caniart)

2005-12-03  Matthias Clasen  <mclasen@redhat.com>

* glib/glib.symbols:
* glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
Nicolas Caniart)

18 years agoMake mapping of empty files work. (#321530)
Matthias Clasen [Sat, 3 Dec 2005 06:44:02 +0000 (06:44 +0000)]
Make mapping of empty files work. (#321530)

2005-12-03  Matthias Clasen  <mclasen@redhat.com>

        * glib/gmappedfile.c: Make mapping of empty files
        work.  (#321530)

18 years agoAdd more api
Matthias Clasen [Sat, 3 Dec 2005 06:36:01 +0000 (06:36 +0000)]
Add more api

18 years agoDocument the slice allocator
Matthias Clasen [Sat, 3 Dec 2005 06:29:12 +0000 (06:29 +0000)]
Document the slice allocator

18 years agoDon't fork a new process just to fix the permissions of the created temp
Matthias Clasen [Sat, 3 Dec 2005 06:10:52 +0000 (06:10 +0000)]
Don't fork a new process just to fix the permissions of the created temp

2005-12-03  Matthias Clasen  <mclasen@redhat.com>

* glib/gfileutils.c: Don't fork a new process just to
fix the permissions of the created temp file.  (#321318,
Alexis S. L. Carvalho)

18 years agoAdd a note about Solaris threads.
Matthias Clasen [Fri, 2 Dec 2005 21:40:43 +0000 (21:40 +0000)]
Add a note about Solaris threads.

2005-12-02  Matthias Clasen  <mclasen@redhat.com>

* README.in: Add a note about Solaris threads.

18 years agoRemove support for Solaris threads. (#136971, Sebastian Wilhelmi, patch by
Matthias Clasen [Fri, 2 Dec 2005 21:37:25 +0000 (21:37 +0000)]
Remove support for Solaris threads. (#136971, Sebastian Wilhelmi, patch by

2005-12-02  Matthias Clasen  <mclasen@redhat.com>

* glib/gspawn.c:
* configure.in: Remove support for Solaris threads.
(#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)

18 years agoBroken commit
Matthias Clasen [Fri, 2 Dec 2005 21:26:23 +0000 (21:26 +0000)]
Broken commit

18 years agoDefine G_GUINT64_CONSTANT in analogy to G_GINT64_CONSTANT. (#322568,
Matthias Clasen [Fri, 2 Dec 2005 21:25:45 +0000 (21:25 +0000)]
Define G_GUINT64_CONSTANT in analogy to G_GINT64_CONSTANT. (#322568,

2005-12-02  Matthias Clasen  <mclasen@redhat.com>

* configure.in: Define G_GUINT64_CONSTANT in analogy to
G_GINT64_CONSTANT.  (#322568, Andrew Paprocki)

18 years agoPortability fixes, C99ism removal
Matthias Clasen [Fri, 2 Dec 2005 20:11:28 +0000 (20:11 +0000)]
Portability fixes, C99ism removal

18 years agoAdd a missing volatile for the IA64 implementation.
Matthias Clasen [Fri, 2 Dec 2005 16:57:24 +0000 (16:57 +0000)]
Add a missing volatile for the IA64 implementation.

2005-12-02  Matthias Clasen  <mclasen@redhat.com>

* glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
the IA64 implementation.

18 years agofix uint/pointer casts
Tim Janik [Fri, 2 Dec 2005 15:28:56 +0000 (15:28 +0000)]
fix uint/pointer casts

18 years agosome naming and type size fixups.
Tim Janik [Fri, 2 Dec 2005 15:18:39 +0000 (15:18 +0000)]
some naming and type size fixups.

Fri Dec  2 16:18:09 2005  Tim Janik  <timj@imendio.com>

        * glib/gslice.c: some naming and type size fixups.

18 years agoreworked GCC-specific type-safe macro variant into something less verbose,
Tim Janik [Fri, 2 Dec 2005 12:11:34 +0000 (12:11 +0000)]
reworked GCC-specific type-safe macro variant into something less verbose,

Fri Dec  2 13:08:58 2005  Tim Janik  <timj@gtk.org>

        * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
        variant into something less verbose, but digestible for gcc-3.4.

18 years agoextended to perform the benchmarking on the old memchunk code if 'O' is
Tim Janik [Fri, 2 Dec 2005 12:10:49 +0000 (12:10 +0000)]
extended to perform the benchmarking on the old memchunk code if 'O' is

Fri Dec  2 10:55:07 2005  Tim Janik  <timj@imendio.com>

        * tests/slice-test.c: extended to perform the benchmarking on the old
        memchunk code if 'O' is selected.

        * tests/memchunks.c: new file which contains the old GLib mem chunks
        implementation with prefix old_mem_chunk_.

        * tests/Makefile.am: added memchunks.c

18 years agoextended to perform the benchmarking on the old memchunk code if 'O' is
Tim Janik [Fri, 2 Dec 2005 09:57:06 +0000 (09:57 +0000)]
extended to perform the benchmarking on the old memchunk code if 'O' is

Fri Dec  2 10:55:07 2005  Tim Janik  <timj@imendio.com>

        * tests/slice-test.c: extended to perform the benchmarking on the old
        memchunk code if 'O' is selected.

        * tests/memchunks.c: new file which contains the old GLib mem chunks
        implementation with prefix old_mem_chunk_.

        * tests/Makefile.am: added memchunks.c

18 years agoimproved gettimeofday() timer resolution by maintaining acceptable load
Tim Janik [Thu, 1 Dec 2005 23:21:49 +0000 (23:21 +0000)]
improved gettimeofday() timer resolution by maintaining acceptable load

Fri Dec  2 00:16:59 2005  Tim Janik  <timj@imendio.com>

        * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
        acceptable load distribution of the syscall.
        (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
        sizes to approximately page_size/2.5 to avoid excessive startup allocations.

18 years agoimproved gettimeofday() timer resolution by maintaining acceptable load
Tim Janik [Thu, 1 Dec 2005 23:20:22 +0000 (23:20 +0000)]
improved gettimeofday() timer resolution by maintaining acceptable load

Fri Dec  2 00:16:59 2005  Tim Janik  <timj@imendio.com>

        * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
        acceptable load distribution of the syscall.
        (slab_allocator_free_chunk): reduce minimum/single-thread magazine sizes
        to approximately page_size/2.5 to avoid excessive startup allocations.

18 years agoMake tests work with slice allocator
Matthias Clasen [Thu, 1 Dec 2005 18:06:17 +0000 (18:06 +0000)]
Make tests work with slice allocator

18 years agoRemove a duplicate _LDADD
Matthias Clasen [Thu, 1 Dec 2005 16:43:56 +0000 (16:43 +0000)]
Remove a duplicate _LDADD

18 years agonew slice allocator implementation.
Tim Janik [Thu, 1 Dec 2005 16:34:33 +0000 (16:34 +0000)]
new slice allocator implementation.

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].

18 years agouse g_slic_* for all object allocations.
Tim Janik [Thu, 1 Dec 2005 16:33:51 +0000 (16:33 +0000)]
use g_slic_* for all object allocations.

Thu Dec  1 12:53:26 2005  Tim Janik  <timj@imendio.com>

        * gtype.c: use g_slic_* for all object allocations.

18 years agoexport g_hash_table_get_type
Tim Janik [Thu, 1 Dec 2005 16:32:11 +0000 (16:32 +0000)]
export g_hash_table_get_type

18 years agoUpdated Tagalog translation by Eric Pareja <xenos@upm.edu.ph>.
Christian Rose [Thu, 1 Dec 2005 10:43:58 +0000 (10:43 +0000)]
Updated Tagalog translation by Eric Pareja <xenos@upm.edu.ph>.

2005-12-01  Christian Rose  <menthos@menthos.com>

* tl.po: Updated Tagalog translation by
Eric Pareja <xenos@upm.edu.ph>.

18 years agoDon't call g_path_skip_root() unless the path is absolute.
Tor Lillqvist [Thu, 1 Dec 2005 04:32:52 +0000 (04:32 +0000)]
Don't call g_path_skip_root() unless the path is absolute.

2005-12-01  Tor Lillqvist  <tml@novell.com>

* glib/gstdio.c (g_stat): Don't call g_path_skip_root() unless
the path is absolute.

18 years agog_rename() doc comment update.
Tor Lillqvist [Thu, 1 Dec 2005 04:06:45 +0000 (04:06 +0000)]
g_rename() doc comment update.