1 2006-01-30 David Schleef <ds@schleef.org>
3 * glib/grand.c: Update URL
5 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
7 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
8 create container classes which are as flexible as a GValue is
10 2006-01-27 Matthias Clasen <mclasen@redhat.com>
12 * configure.in: Bump version
14 * === Released 2.9.5 ===
16 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
21 2006-01-26 Matthias Clasen <mclasen@redhat.com>
23 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
26 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
28 * fixed buglets reported by Jens Granseuer in #328254.
30 * configure.in: free the memory allocated in posix_memalign() tests.
32 * glib/gslice.c: spelling fixes.
34 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
36 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
37 slices, make sure g_mem_gc_friendly is properly initialized.
39 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
40 the first allocation. applied some branching optimizations.
42 * docs/macros.txt: reflected --enable-gc-friendly change and
43 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
45 * configure.in: changed --enable-gc-friendly=yes to define
46 ENABLE_GC_FRIENDLY_DEFAULT.
48 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
49 if (G_UNLIKELY (g_mem_gc_friendly)).
52 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
53 care of by g_slice_free1().
55 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
57 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
59 * glib/gslice.c: only use posix_memalign() if it's known to work,
60 revert to memalign() otherwise.
62 * configure.in: check for broken posix_memalign() implementations
65 2006-01-24 Matthias Clasen <mclasen@redhat.com>
67 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
68 on all little-endian systems. (#143380, Marc Moorcroft)
70 2006-01-23 Matthias Clasen <mclasen@redhat.com>
72 * configure.in: Change the shared libary extension for hpux-ia64
73 to so. (#328253, The Written Word)
75 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
77 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
78 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
81 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
83 * glib/gslice.c (slice_config_init): initialize GSlice config from
84 G_SLICE environment variable. we support G_SLICE=always-malloc
85 currently, which forces all g_slice_*() allocations to use the system
89 g_parse_debug_string(): added a note about not using g_malloc() here.
90 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
91 g_slice. contains only guesswork in the WIN32 branch.
93 2006-01-18 Matthias Clasen <mclasen@redhat.com>
97 * === Released 2.9.4 ===
101 2006-01-17 Matthias Clasen <mclasen@redhat.com>
103 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
104 call. (#327421, Morten Welinder)
106 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
108 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
109 G_LOCK. Use the unused_thread_queue lock instead. Change
110 g_thread_pool_thread_proxy such that threads only wait on
111 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
112 due to superfluous tasks. Global tasks wait at most for
113 max-idle-time milliseconds. Make sure, that no task is woken up
114 twice for the same event via a wakeup_serial. This fixes #324228.
116 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
117 invalid NULL into the thread pools. This as well fixes #327290.
119 2006-01-16 Matthias Clasen <mclasen@redhat.com>
121 * configure.in: Bump version
123 * === Released 2.9.3 ===
127 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
128 Rework the strict aliasing fix to not break C++, pointed
129 out by Murray Cumming.
131 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
132 Signal waiting threads, problem noticed by Christian Kellner.
134 2006-01-16 Matthias Clasen <mclasen@redhat.com>
136 Fix bug 326747, Alberto Ruiz:
138 * glib/gconvert.c (make_valid_utf8): Change this function to
139 replace unknown characters by the Unicode replacement character
140 instead of '?', and don't append "(invalid encoding)".
141 (g_filename_display_name, g_filename_display_basename): Document
142 how to determine if the filename was in an invalid encoding.
144 2006-01-14 Matthias Clasen <mclasen@redhat.com>
146 * glib/gtree.c: Replace the simple recursive implementation
147 by a nonrecursive, threaded implementation by Maurizio
150 2006-01-12 Matthias Clasen <mclasen@redhat.com>
152 * tests/asyncqueue-test.c (main): Fix the
153 assert to not always trigger. (#326558,
156 2006-01-11 Matthias Clasen <mclasen@redhat.com>
158 * tests/convert-test.c: Don't test items_read and
159 items_written in error cases where it is not set.
161 * tests/asyncqueue-test.c: Set error to NULL.
163 2006-01-08 Matthias Clasen <mclasen@redhat.com>
165 * glib/gtree.c: Break some long lines.
167 * tests/tree-test.c: Add more tests.
169 * glib/gslice.c: Include config.h first, pointed out
172 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
174 * docs/reference/gobject/tmpl/gtype.sgml: explain that
175 G_TYPE_INSTANCE_GET_CLASS() does behave different during
178 2006-01-06 Matthias Clasen <mclasen@redhat.com>
180 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
181 that we have posix_memalign(). (#323937, Bogdan Nicula)
183 2006-01-05 Matthias Clasen <mclasen@redhat.com>
185 * configure.in: Bump version
187 * === Released 2.9.2 ===
189 2006-01-05 Hans Breuer <hans@breuer.org>
191 * glib/makefile.msc.in : added gslice.obj
193 2006-01-05 Matthias Clasen <mclasen@redhat.com>
197 * glib/gmain.c: Clarify the documentation of source ids
198 a bit. (#325874, Dan Williams)
200 * configure.in: Fix another strict aliasing problem.
202 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
205 2006-01-04 Manish Singh <yosh@gimp.org>
208 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
209 either, since they have the same issue as g_rmdir. (Related to
212 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
214 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
215 order when creating magazines, so we hand out chunks with ascending
218 2006-01-04 Abel Cheung <maddog@linuxhall.org>
220 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
222 2006-01-03 Martyn Russell <martyn@imendio.com>
224 * docs/reference/glib/glib-sections.txt:
226 * glib/gthreadpool.[ch]:
227 - Added new API g_thread_pool_get_idle_time() and
228 g_thread_pool_set_idle_time(). (#324228).
230 * tests/threadpool-test.c:
231 - Updated test case to do thread pool sorting, thread pool with
232 no sorting and a thread pool with idle thread timeouts.
234 2006-01-03 Matthias Clasen <mclasen@redhat.com>
236 * glib/gmain.h: Add new functions here, too.
238 * glib/glib.symbols: Add new functions.
240 2005-12-20 Michael Meeks <michael.meeks@novell.com>
242 * glib/gmain.c (g_main_context_is_owner): new method
243 to determine if the current thread is the owner of the
246 2006-01-02 Matthias Clasen <mclasen@redhat.com>
250 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
251 expanding to rmdir, since rmdir is not declared in a portable
252 system header we can include in gstdio.h. (#325249, Jani Monoses)
254 2006-01-01 Tor Lillqvist <tml@novell.com>
256 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
257 stdout and stderr to unbuffered. Otherwise the giochannel layer
258 will try to read from them regardless whether the
259 g_io_channel_win32_poll() call here has indicated
260 readability or not. (#325310)
262 2005-12-29 Matthias Clasen <mclasen@redhat.com>
264 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
265 in the docs. (#325273, Declan Naughton)
267 2005-12-27 Manish Singh <yosh@gimp.org>
269 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
270 make sure we really override things.
272 2005-12-27 Matthias Clasen <mclasen@redhat.com>
274 Fix #316221, Michal Benes, Stanislav Brabec;
276 * configure.in: Fix a strict aliasing problem in
277 g_static_mutex_get_mutex().
278 * glib/gthread.h: ...and in
279 g_static_mutex_get_mutex_impl_shortcut().
281 * glib/gdatasetprivate.h: Add a cast to silence compiler
282 warnings. (#321978, Andrew Paprocki)
284 Partial fix for bug #323937, Bogdan Nicula.
286 * configure.in: Check for malloc.h
287 * glib/gslice.c: Don't include malloc.h unconditionally.
289 2005-12-27 Manish Singh <yosh@gimp.org>
291 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
292 (#324950, Dan Yefimov)
294 2005-12-26 Matthias Clasen <mclasen@redhat.com>
296 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
299 2005-12-25 Matthias Clasen <mclasen@redhat.com>
301 * AUTHORS: Update my email
303 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
305 2005-12-21 Matthias Clasen <mclasen@redhat.com>
307 * README.in: Add some notes on when 'make check' may fail.
309 2005-12-20 Matthias Clasen <mclasen@redhat.com>
312 * glib/gthreadpool.h:
313 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
314 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
316 * tests/threadpool-test.c: Test this.
318 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
320 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
321 handle errors without depending on gmessage.c which might not be
322 setup when the error occours.
323 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
324 always freeing can be achieved by adjusting the working set time to
325 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
326 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
327 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
328 minimize wastage if small amounts of differently sized structrues are
329 allocated, this does come at a performance cost of roughly 5% though.
330 fixed up block alignment calculation, so it works for varying
331 block sizes. only use strerror() not g_strerror() since the latter
332 depends on working GQuark and GSlice.
333 mem_error(): implemented in terms of fprintf and vfprintf.
335 * tests/slice-color.c: new program to test cache colorization effects.
337 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
338 G_SLICE_CONFIG_WORKING_SET_MSECS.
340 2005-12-17 Matthias Clasen <mclasen@redhat.com>
342 * glib/goption.c (parse_short_option): Set an error in all
343 failure cases. (#324332, Tim-Philipp Müller)
345 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
347 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
348 and g_atomic_pointer_get. Add g_atomic_int_set and
349 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
350 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
351 as functions (additionally to the macros in the header) for the
352 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
354 2005-12-16 Matthias Clasen <mclasen@redhat.com>
356 * glib/gmem.c (g_allocator_new): Don't return a pointer to
357 a const struct, since apps expect to be able to modify it.
358 (#324179, J. Ali Harlow)
360 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
362 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
363 for systems where the initialization of atomic variables requires a
364 write memory barrier.
366 2005-12-09 Matthias Clasen <mclasen@redhat.com>
368 * configure.in: Bump version
370 * === Released 2.9.1 ===
372 2005-12-09 Alexander Larsson <alexl@redhat.com>
376 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
379 2005-12-08 Matthias Clasen <mclasen@redhat.com>
383 2005-12-07 Martyn Russell <martyn@imendio.com>
385 * glib/gasyncqueue.c:
386 - Call g_queue_insert_sorted() instead of duplicating the code.
387 - Call g_queue_sort() instead of duplicating the code.
388 - Invert sort function results to make sure the same sort function
389 gives the same results across glist, gslist, gqueue and
392 * tests/asyncqueue-test.c:
393 - Updated the sort function to reflect the example in the
394 documentation for gasyncqueue.c.
396 2005-12-07 Martyn Russell <martyn@imendio.com>
398 * docs/reference/glib/glib-sections.txt:
399 * docs/reference/glib/tmpl/linked_lists_double.sgml:
400 * docs/reference/glib/tmpl/linked_lists_single.sgml:
403 - Added g_list_insert_sorted_with_data () and
404 g_slist_insert_sorted_with_data ().
405 - Removed the extra check in g_list_sort() and g_slist_sort() for
406 GCompareDataFunc vs. GCompareFunc.
408 2005-12-07 Tor Lillqvist <tml@novell.com>
410 * glib-gettextize.in: Look up prefix at run-time on Win32,
411 assuming the standard directory structure with glib-gettextize in
414 2005-12-06 Matthias Clasen <mclasen@redhat.com>
416 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
417 we need GTK+ 2.8 to compile against GLib 2.10.
419 2005-12-05 Manish Singh <yosh@gimp.org>
421 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
422 to override any user settings.
424 2005-12-05 Matthias Clasen <mclasen@redhat.com>
426 * configure.in: Revert an accidential commit
428 2005-12-05 Matthias Clasen <mclasen@redhat.com>
430 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
432 * tests/Makefile.am (TESTS_ENVIRONMENT):
433 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
434 MALLOC_CHECK_ and MALLOC_PERTURB_
436 * tests/run-collate-tests.sh: Run the collation tests explicitly
437 in en_US locale. (#320463)
439 * glib/gmem.h: Really deprecate GMemChunk.
446 * glib/gstrfuncs.c: Add versioned deprecation docs.
448 2005-12-05 Martyn Russell <martyn@imendio.com>
450 * docs/reference/glib/glib-sections.txt:
451 * glib/gasyncqueue.[ch]:
452 - Added support for sorting async queues by with _push_sorted(),
453 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
456 * tests/asyncqueue-test.c:
457 - Added test case for gasyncqueue.c
459 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
461 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
462 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
463 in bug 323178. moved time consuming logic from g_slice_free() out of
464 the inner loop, so g_slice_free_chain_with_offset() provides a real
465 performance benefit over g_slice_free1() now.
467 * glib/gslice.h: renamed g_slice_free_chain() to
468 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
469 a type-safe macro as suggested in bug 323178.
470 simplified the macro implementation of g_slice_free() and implemented
471 it in a type safe manner for all compliers as suggested by Morten
472 Welinder <mortenw@gnome.org>.
477 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
479 2005-12-05 Matthias Clasen <mclasen@redhat.com>
481 * glib/gasyncqueue.c: Add some docs.
483 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
485 * glib/gatomic.c: In the ia64 implementation, use
486 __sync builtin without _si or _di suffix. (#321229,
487 Stanislav Brabec, patch by Andreas Schwab)
489 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
491 * glib/gslice.h: Remove comma at the end of enum.
493 2005-12-04 Matthias Clasen <mclasen@redhat.com>
495 Handle multiple user names with the same UID better.
496 (#319535, Laszlo Peter)
498 * glib/gutils.c (g_get_any_init_do): When determining user
499 data, first look up $LOGNAME. If the UID doesn't match
500 getuid(), fall back to the current behaviour of looking
501 up the user data based on getuid().
503 2005-12-04 Matthias Clasen <mclasen@redhat.com>
505 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
506 to make gcc warn if a function result is ignored. (#145466,
507 Arjan van de Ven, Alex Larsson)
509 * glib/gmem.h: Add the new attribute to g_realloc and
512 2005-12-03 Matthias Clasen <mclasen@redhat.com>
515 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
518 * glib/gmappedfile.c: Make mapping of empty files
521 * glib/gfileutils.c: Don't fork a new process just to
522 fix the permissions of the created temp file. (#321318,
523 Alexis S. L. Carvalho)
525 2005-12-02 Matthias Clasen <mclasen@redhat.com>
527 * README.in: Add a note about Solaris threads.
530 * configure.in: Remove support for Solaris threads.
531 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
533 2005-12-02 Matthias Clasen <mclasen@redhat.com>
535 * configure.in: Define G_GUINT64_CONSTANT in analogy to
536 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
538 2005-12-02 Matthias Clasen <mclasen@redhat.com>
540 * glib/gslice.c: Win32 portability fixes and C99ism removal,
541 pointed out by Kazuki Iwamoto. (#323052)
543 2005-12-02 Matthias Clasen <mclasen@redhat.com>
545 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
546 the IA64 implementation.
548 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
550 * glib/gslice.c: some naming and type size fixups.
552 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
554 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
555 variant into something less verbose, but digestible for gcc-3.4.
557 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
559 * tests/slice-test.c: extended to perform the benchmarking on the old
560 memchunk code if 'O' is selected.
562 * tests/memchunks.c: new file which contains the old GLib mem chunks
563 implementation with prefix old_mem_chunk_.
565 * tests/Makefile.am: added memchunks.c
567 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
569 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
570 acceptable load distribution of the syscall.
571 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
572 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
574 2005-12-01 Matthias Clasen <mclasen@redhat.com>
576 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
577 double-free base interface members.
579 * tests/Makefile.am: Remove duplicate variable.
581 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
583 * glib/gslice.[hc]: new slice allocator implementation.
585 * tests/slice-test.c: added random slice allocation test.
587 * glib/gthread.[hc]: removed newly added private thread mem API.
589 * glib/gthreadinit.h:
592 * glib/gmem.c: divided glib threading initialisation into three phases,
593 initialisation where private keys and messaging are not available (only
594 needed by gmem.c), initialisation without messaging but private keys
595 available (gslice.c, gmessage.c), and full fledged initialisers that
596 server the rest of glib. initialisation functions got renamed to reflect
597 the limitations of their corresponding phases.
599 * glib/gmem.c: removed memchunk code, defer allocations to
602 * glib/gmem.[hc]: removed g_slice_* skeletons.
604 * glib/glib.symbols: added g_slice_* symbols.
606 * configure.in: check for availability of posix_memalign(3), memalign(3)
609 * glib/Makefile.am: added gslice.[hc].
611 2005-12-01 Tor Lillqvist <tml@novell.com>
613 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
614 trailing slash(es) for non-root folders. stat() fails if non-root
615 folders are specified with trailing slashes. It's too much hassle
616 to demand that callers strip such slashes themselves, especially
617 as it is easy to get it wrong and strip the slash of a root
619 (g_rename): On NT-based Windows, use MoveFileEx() with
620 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
622 2005-11-28 Matthias Clasen <mclasen@redhat.com>
624 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
627 * configure.in: Check whether do { } while (0) works.
629 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
630 G_STMT_END if it works.
632 2005-11-28 Matthias Clasen <mclasen@redhat.com>
634 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
635 * glib/gnode.c (g_node_depth_traverse_level):
636 * glib/gmem.c (g_allocator_new):
637 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
638 (g_hash_table_foreach_remove_or_steal): Silence compiler
641 2005-11-27 Matthias Clasen <mclasen@redhat.com>
643 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
644 Don't read beyond len. (#322520, Christian Persch)
646 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
648 * glib/gdataset.c: access datalist flags via atomic pointer access
649 functions, instead of acquiring the dataset lock. this is faster and
650 also matches the atomic pointer readouts in gdatalistprivate.h and
651 direct pointer modifications required by gobject.c.
653 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
655 * glib/gdataset.c: streamlined initialization code somewhat,
656 removed GData node cache.
658 2005-11-22 Matthias Clasen <mclasen@redhat.com>
660 * glib/ghash.c (g_hash_table_ref):
661 (g_hash_table_unref): Mark these as new API.
663 Avoid double locking in g_intern_string (#322133,
666 * glib/gdataset.c (g_quark_from_string_internal): New
667 internal function which factors out common parts of
668 g_quark_from[_static]_string.
669 (g_quark_from_string, g_quark_from_static_string):
670 Use g_quark_from_string_internal.
671 (g_intern_string, g_intern_static_string): Use
672 g_quark_from_string_internal, and only take the
674 (g_quark_new): Don't store the strings shifted by -1
675 in the g_quarks array.
676 (g_quark_to_string): Adapt to the previous change.
678 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
682 g_hash_table_new_full(): create hash tables with a ref count of 1.
683 g_hash_table_ref(): atomically ref_count+=1
684 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
685 when refcount reaches 0.
686 g_hash_table_destroy(): just destroy keys and values, unref by 1.
687 g_hash_table_insert():
688 g_hash_table_replace(): assert ref_count>0.
691 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
694 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
696 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
697 variables. (#321984, Andrew Paprocki)
699 2005-11-18 Matthias Clasen <mclasen@redhat.com>
701 * configure.in: Bump version
703 2005-11-17 Matthias Clasen <mclasen@redhat.com>
705 * === Released 2.9.0 ===
707 * glib/glib.symbols: Updates
709 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
710 hints to the docs. (Paolo Borelli)
714 * glib/gthread.c (g_thread_foreach): Mark as new api.
716 * README.in: Updates.
720 * glib/gdate.c (g_date_set_time_t):
721 (g_date_set_time_val): Add functions to set a date from
722 a time_t and from a GTimeVal.
723 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
726 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
727 and g_utf8_offset_to_pointer.
729 * glib/gutf8.c (g_utf8_pointer_to_offset)
730 (g_utf8_offset_to_pointer): Handle negative offsets, and use
731 "stutter stepping" for going backwards. (#320638, Larry
736 * glib/gthread.c: const correctness fixes, found
737 by Arjan van de Ven and gcc.
739 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
741 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
742 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
744 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
746 * glib/guniprop.c: Use bit hacks instead when checking a general
747 category value against multiple values.
749 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
751 2005-11-10 Simos Xenitellis <simos@gnome.org>
753 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
755 2005-11-05 Tor Lillqvist <tml@novell.com>
757 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
758 (#320688) Instead set error and return failure.
760 2005-11-07 Matthias Clasen <mclasen@redhat.com>
762 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
763 while holding the context lock. (#320886, Andy Wingo)
765 2005-11-07 Billy Biggs <vektor@dumbterm.net>
767 * tests/refcount/Makefile.am: Add a missing reference to libglib
768 in the LDADD for this test.
770 2005-11-07 Matthias Clasen <mclasen@redhat.com>
772 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
773 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
774 GLib 2.9. (#320755, Luca Ferretti)
776 2005-11-04 Matthias Clasen <mclasen@redhat.com>
778 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
779 use a faster lookup table for g_unichar_get_mirror_char().
780 * glib/gmirroringtable.h: The new table.
781 * glib/gunichartables.h: Remove bidi_mirroring_table.
782 * glib/gen-unicode-tables.pl: Don't generate the mirroring
784 * glib/glib-mirroring-tab/*: Sources for the program
785 which generated gmirroringtable.h.
787 * glib/glist.c: Avoid some code duplication.
789 * glib/gscanner.h: Include gdataset.h. (#320322)
791 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
792 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
793 to help find critical warnings in applications. (#320017,
796 2005-11-02 Tor Lillqvist <tml@novell.com>
798 * glib/glib.symbols: Remove large amount of trailing whitespace
799 from one line. Remove #ifdef G_UNDEFINED from around
800 g_hash_table_new and g_hash_table_new_full symbols, what was the
801 purpose of that? Add the g_slice functions.
803 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
805 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
806 added g_slice_*() API to allocate and cache small bits of memory.
807 an actuall allocator implementation for g_slice_*() is still pending.
809 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
810 changed GRealThread list to use in-structure *next; fields instead
811 of GSList, in order for thread iteration to not depenend on g_slice_*()
813 _g_thread_mem_private_get():
814 _g_thread_mem_private_set(): added accessors for private memory,
815 needed because the ordinary GPrivate implementation relies on GArray
816 and GSList and therefore indirectly on working g_slice_*() allocations.
819 g_thread_foreach(): new public API function to loop over all existing threads.
833 removed GAllocator and free list usages and accompanying locks.
834 use g_slice_*() API to allocate and cache small bits of memory.
836 * glib/ghook.h: removed GMemChunk field from public API.
839 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
841 * glib/gnode.h: deprecate allocator API.
843 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
844 be used for poll rec lists.
846 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
847 g_relation_destroy(): free all tuples from the all_tuples hash table,
848 this effectively maintains the life time track keeping of tuples.
849 g_relation_delete_tuple(): free tuples which are removed from the
850 all_tuples hash table. this fixes a temporary leak that was present
851 in the memchunk code until the destruction of the relation.
853 2005-10-29 Matthias Clasen <mclasen@redhat.com>
855 * tests/convert-test.c: Add some tests for conversions between
856 UTF-8, UCS-4 and UTF-16.
858 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
859 of len == -1, noticed by Morten Welinder.
861 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
863 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
865 2005-10-26 Matthias Clasen <mclasen@redhat.com>
867 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
868 in the error case, and add some documentation. (#319806, Morten
871 2005-10-19 Manish Singh <yosh@gimp.org>
873 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
874 config.h symbol automatically. Fixes bug #313731.
876 2005-10-19 Tor Lillqvist <tml@novell.com>
878 * glib/gwin32.c (g_win32_get_package_installation_directory):
879 Return a g_strdup()ed copy of the value stored in the hash table,
880 so that it can be g_free()d without leaving a dangling pointer in
881 the hash table. (#319232)
883 2005-10-06 Matthias Clasen <mclasen@redhat.com>
885 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
886 all-zero sequences correctly. (#317930, Sebastien Bacher)
888 2005-10-05 Matthias Clasen <mclasen@redhat.com>
890 * glib/gmarkup.c (g_markup_escape_text): Doc addition
892 * Makefile.am: Add Collation tests.
894 * tests/collate/*: Inputs and expected outputs for collation tests.
896 * tests/run-collate-tests.sh: Script to run collation tests.
898 * tests/unicode-collate.c (main): Rework slightly to make
899 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
901 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
903 * docs/reference/glib/tmpl/unicode.sgml:
904 * glib/gen-unicode-tables.pl:
906 * glib/gunichartables.h:
908 * tests/casefold.txt:
909 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
910 GUnicodeBreakType types. That may break some applications, like
913 2005-09-26 Matthias Clasen <mclasen@redhat.com>
915 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
916 docs. (#314393, Matthew F. Barnes)
919 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
920 they are already declared in glib.h. This doesn't break documented
921 use of gprintf.h, but should probably be pointed out in the
922 release notes for 2.10. (#314232, Behdad Esfahbod)
924 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
926 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
927 complexity optimization, based on a patch from Matthias Clasen.
929 * tests/patterntest.c: more tests, mostly from matthias.
931 2005-09-20 Matthias Clasen <mclasen@redhat.com>
933 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
934 (#316703, Mark Drago)
936 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
938 * glib/gpattern.c: applied a patch from matthias which checks on the
939 upper bound of GPatternSpec length to optimize matches.
942 * tests/patterntest.c: added more match cases.
944 2005-09-16 Tor Lillqvist <tml@novell.com>
946 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
947 warning to doc comment that these functions should not be used on
948 encodings like CP932.
950 2005-09-14 Matthias Clasen <mclasen@redhat.com>
952 * tests/keyfile-test.c: Add a test for grup names of length 1.
954 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
955 of length 1. (#316309)
957 2005-09-12 Matthias Clasen <mclasen@redhat.com>
959 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
961 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
963 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
964 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
967 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
969 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
970 (child_watch_helper_thread): Remove some dead code. Closes
973 2005-09-07 Tor Lillqvist <tml@novell.com>
975 * glib/Makefile.am: Create also a console version of the
976 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
977 It's otherwise identical to gspawn-win32-helper.exe, except marked
978 as a console application (linked without the -mwindows option).
980 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
981 Drop the dont_wait parameter. Its truth value correlated 100% with
982 the NULLness of the exit_status parameter anyway, so it's enough
983 to check whether exit_status is NULL. Invert the sense of the
984 dont_return_handle parameter and rename it to do_return_handle, to
985 make the code easier to read by avoiding double negations.
987 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
988 to do_spawn_with_pipes() accordingly.
990 (do_spawn_with_pipes): If we have a console, use the console
991 version of the helper program, otherwise use the GUI one. This
992 avoids extra console windows opening up in some situations. (In
993 case a console application uses the GUI gspawn-win32-helper.exe to
994 spawn another console application we would get a separate console
995 for the spawned console application).
997 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
999 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1001 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1002 if the file is too large. (#315275, Kjartan Maraas)
1004 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1005 of read() is signed. (#315273, Kjartan Maraas)
1007 2005-08-31 Tor Lillqvist <tml@novell.com>
1009 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1010 the EXE and each DLL have their own atexit function chains.
1012 #define g_atexit as atexit instead. This means it has a
1013 better chance of doing what the caller wants. For instance,
1014 gtkhtml calls g_atexit() registering a function in gtkhtml
1015 itself. This caused a crash when g_atexit() was implemented as a
1016 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1017 the time the GLib DLL got unloaded.
1019 * glib/gutils.c: #undef the #define mentioned above, to also get a
1020 real g_atexit() into the DLL for backward compatibility. Document
1021 the Windows behaviour of g_atexit(), and document the varying ways
1022 atexit() can behave in the context of dynamically loaded modules
1025 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1027 * glib/glib.symbols:
1029 * glib/gdataset.c: Add string interning functions.
1031 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1033 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1036 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1038 * configure.in: Bump version to 2.9.0
1040 2005-08-25 Tor Lillqvist <tml@novell.com>
1042 Make also the g_spawn*() functions take parameters in the GLib
1043 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1044 API or ABI. Like the other GLib API that was earlier changed to
1045 use UTF-8 on Windows, the names of the functions that take UTF-8
1046 have _utf8 suffixes added by using preprocessor macros in the
1047 header file. The old names are kept for functions with the old
1048 behaviour, taking parameters in the system codepage, for DLL ABI
1051 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1052 g_spawn*() functions.
1054 * glib/gspawn-win32.c: Use wide-char API on NT-based
1055 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1056 system codepage (Win9x) and call the C library _wspawn*() or
1057 spawn*() functions respectvely. Add DLL ABI stability versions
1058 that take parameters in the system codepage.
1060 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1061 wide-char versions of argv and envp, and use wide-char API to
1062 change directory and spawn the program to run. Remove the verbose
1063 debugging output, it was too complex to modify for the wide-char
1064 features. (Just add temporary debugging printouts if needed, no
1065 need to have them permanently in the source.)
1067 * glib/gspawn.c: Corresponding documentation updates.
1069 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1070 symbols as PRIVATE, add the new _utf8-suffixed ones.
1072 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1074 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1075 guint64 otherwise and that can produce warnings about comparison
1076 between signed and unsigned.
1078 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1080 * glib/gutils.c: Fix the crt_externs.h include.
1082 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1084 * NEWS: Fix spelling of my first name.
1086 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1090 * === Released 2.8.1 ===
1094 2005-08-20 Hans Breuer <hans@breuer.org>
1096 * glib/makefile.msc.in : link with ws2_32.lib
1098 2005-08-18 Tor Lillqvist <tml@novell.com>
1100 * configure.in: Check for <sys/wait.h>
1102 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1104 2005-08-18 Ross Burton <ross@burtonini.com>
1107 Optimise single-character insertions.
1112 * tests/string-test.c:
1113 Add tests for new optimisation, and fix a leak.
1115 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1117 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1119 * glib/gutils.c: On Darwin, include crt-externs.h and
1120 define environ using _NSGetEnviron(). (#313731)
1122 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1124 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1125 (g_get_any_init_do): ... to this new function.
1126 (g_get_any_init): Declare as inline.
1127 (g_get_any_init_locked): New inline function, does the locking.
1128 Make use of these two throughout the code.
1130 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1132 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1133 the child process and then simply return. This makes
1134 The "S" option work as documented in g_on_error_query().
1135 (#313125, Matthew F. Barnes)
1137 * glib/gunicode.h: Update the link to Unicode category
1138 values. (#313369, Behnam Esfahbod)
1140 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1141 a little. (#311727, Tristan van Berkom)
1143 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1144 check work on ia64 too, where some symbols ended up
1145 in yet another section.