1 2006-02-24 Matthias Clasen <mclasen@redhat.com>
3 * === Released 2.10.0 ===
5 * README.in: Mention the slice allocator and GInitiallyUnowned.
9 * configure.in: Bump version to 2.10.0
11 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
13 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
16 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
18 * tests/completion-test.c: (main): Plug leaks reported by
20 * tests/convert-test.c: (test_iconv_state), (test_one_half):
22 * tests/patterntest.c: (test_compilation): Same
23 * tests/shell-test.c: (do_argv_test): Same.
24 * tests/unicode-caseconv.c: (main): Same.
25 * tests/uri-test.c: (run_to_uri_tests): Same.
28 2006-02-19 Martyn Russell <martyn@imendio.com>
30 * tests/asyncqueue-test.c:
32 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
33 _insert_sorted and _insert_sorted_with_data API.
35 2006-02-18 Matthias Clasen <mclasen@redhat.com>
37 * tests/gobject/Makefile.am: Add paramspec-test
39 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
41 * tests/gobject/gvalue-test.c: Add more tests.
43 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
45 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
47 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
49 * glib/gthreadpool.c: Fix deadlock when signalling the thread
50 which freed a thread pool (#331110, Chris Wilson).
52 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
54 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
55 posix_memalign() in case we actually are going to use it, because
56 we detected a compliant implementation (#328997).
58 * configure.in (enable_included_printf): don't include malloc.h when
59 testing for posix_memalign() funcitonality, since this may break the
60 test on some systems (#328997).
62 2006-02-14 Matthias Clasen <mclasen@redhat.com>
64 * glib/gstrfuncs.c: Improve docs.
66 2006-02-11 Matthias Clasen <mclasen@redhat.com>
68 * configure.in: Bump version
70 2006-02-10 Matthias Clasen <mclasen@redhat.com>
72 * === Released 2.9.6 ===
76 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
78 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
79 the type of varargs arguments. (bug #317679)
81 2006-01-31 Matthias Clasen <mclasen@redhat.com>
83 * autogen.sh: Touch README and INSTALL here to pacify
84 automake. (#329124, Kjartan Maraas, Tim Janik)
86 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
88 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
89 some missing cases where pointed out by Sebastian Wilhelmi.
91 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
93 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
95 2006-01-30 David Schleef <ds@schleef.org>
97 * glib/grand.c: Update URL
99 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
101 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
102 create container classes which are as flexible as a GValue is
104 2006-01-27 Matthias Clasen <mclasen@redhat.com>
106 * configure.in: Bump version
108 * === Released 2.9.5 ===
110 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
115 2006-01-26 Matthias Clasen <mclasen@redhat.com>
117 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
120 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
122 * fixed buglets reported by Jens Granseuer in #328254.
124 * configure.in: free the memory allocated in posix_memalign() tests.
126 * glib/gslice.c: spelling fixes.
128 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
130 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
131 slices, make sure g_mem_gc_friendly is properly initialized.
133 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
134 the first allocation. applied some branching optimizations.
136 * docs/macros.txt: reflected --enable-gc-friendly change and
137 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
139 * configure.in: changed --enable-gc-friendly=yes to define
140 ENABLE_GC_FRIENDLY_DEFAULT.
142 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
143 if (G_UNLIKELY (g_mem_gc_friendly)).
146 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
147 care of by g_slice_free1().
149 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
151 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
153 * glib/gslice.c: only use posix_memalign() if it's known to work,
154 revert to memalign() otherwise.
156 * configure.in: check for broken posix_memalign() implementations
159 2006-01-24 Matthias Clasen <mclasen@redhat.com>
161 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
162 on all little-endian systems. (#143380, Marc Moorcroft)
164 2006-01-23 Matthias Clasen <mclasen@redhat.com>
166 * configure.in: Change the shared libary extension for hpux-ia64
167 to so. (#328253, The Written Word)
169 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
171 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
172 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
175 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
177 * glib/gslice.c (slice_config_init): initialize GSlice config from
178 G_SLICE environment variable. we support G_SLICE=always-malloc
179 currently, which forces all g_slice_*() allocations to use the system
183 g_parse_debug_string(): added a note about not using g_malloc() here.
184 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
185 g_slice. contains only guesswork in the WIN32 branch.
187 2006-01-18 Matthias Clasen <mclasen@redhat.com>
191 * === Released 2.9.4 ===
195 2006-01-17 Matthias Clasen <mclasen@redhat.com>
197 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
198 call. (#327421, Morten Welinder)
200 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
202 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
203 G_LOCK. Use the unused_thread_queue lock instead. Change
204 g_thread_pool_thread_proxy such that threads only wait on
205 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
206 due to superfluous tasks. Global tasks wait at most for
207 max-idle-time milliseconds. Make sure, that no task is woken up
208 twice for the same event via a wakeup_serial. This fixes #324228.
210 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
211 invalid NULL into the thread pools. This as well fixes #327290.
213 2006-01-16 Matthias Clasen <mclasen@redhat.com>
215 * configure.in: Bump version
217 * === Released 2.9.3 ===
221 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
222 Rework the strict aliasing fix to not break C++, pointed
223 out by Murray Cumming.
225 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
226 Signal waiting threads, problem noticed by Christian Kellner.
228 2006-01-16 Matthias Clasen <mclasen@redhat.com>
230 Fix bug 326747, Alberto Ruiz:
232 * glib/gconvert.c (make_valid_utf8): Change this function to
233 replace unknown characters by the Unicode replacement character
234 instead of '?', and don't append "(invalid encoding)".
235 (g_filename_display_name, g_filename_display_basename): Document
236 how to determine if the filename was in an invalid encoding.
238 2006-01-14 Matthias Clasen <mclasen@redhat.com>
240 * glib/gtree.c: Replace the simple recursive implementation
241 by a nonrecursive, threaded implementation by Maurizio
244 2006-01-12 Matthias Clasen <mclasen@redhat.com>
246 * tests/asyncqueue-test.c (main): Fix the
247 assert to not always trigger. (#326558,
250 2006-01-11 Matthias Clasen <mclasen@redhat.com>
252 * tests/convert-test.c: Don't test items_read and
253 items_written in error cases where it is not set.
255 * tests/asyncqueue-test.c: Set error to NULL.
257 2006-01-08 Matthias Clasen <mclasen@redhat.com>
259 * glib/gtree.c: Break some long lines.
261 * tests/tree-test.c: Add more tests.
263 * glib/gslice.c: Include config.h first, pointed out
266 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
268 * docs/reference/gobject/tmpl/gtype.sgml: explain that
269 G_TYPE_INSTANCE_GET_CLASS() does behave different during
272 2006-01-06 Matthias Clasen <mclasen@redhat.com>
274 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
275 that we have posix_memalign(). (#323937, Bogdan Nicula)
277 2006-01-05 Matthias Clasen <mclasen@redhat.com>
279 * configure.in: Bump version
281 * === Released 2.9.2 ===
283 2006-01-05 Hans Breuer <hans@breuer.org>
285 * glib/makefile.msc.in : added gslice.obj
287 2006-01-05 Matthias Clasen <mclasen@redhat.com>
291 * glib/gmain.c: Clarify the documentation of source ids
292 a bit. (#325874, Dan Williams)
294 * configure.in: Fix another strict aliasing problem.
296 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
299 2006-01-04 Manish Singh <yosh@gimp.org>
302 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
303 either, since they have the same issue as g_rmdir. (Related to
306 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
308 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
309 order when creating magazines, so we hand out chunks with ascending
312 2006-01-04 Abel Cheung <maddog@linuxhall.org>
314 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
316 2006-01-03 Martyn Russell <martyn@imendio.com>
318 * docs/reference/glib/glib-sections.txt:
320 * glib/gthreadpool.[ch]:
321 - Added new API g_thread_pool_get_idle_time() and
322 g_thread_pool_set_idle_time(). (#324228).
324 * tests/threadpool-test.c:
325 - Updated test case to do thread pool sorting, thread pool with
326 no sorting and a thread pool with idle thread timeouts.
328 2006-01-03 Matthias Clasen <mclasen@redhat.com>
330 * glib/gmain.h: Add new functions here, too.
332 * glib/glib.symbols: Add new functions.
334 2005-12-20 Michael Meeks <michael.meeks@novell.com>
336 * glib/gmain.c (g_main_context_is_owner): new method
337 to determine if the current thread is the owner of the
340 2006-01-02 Matthias Clasen <mclasen@redhat.com>
344 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
345 expanding to rmdir, since rmdir is not declared in a portable
346 system header we can include in gstdio.h. (#325249, Jani Monoses)
348 2006-01-01 Tor Lillqvist <tml@novell.com>
350 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
351 stdout and stderr to unbuffered. Otherwise the giochannel layer
352 will try to read from them regardless whether the
353 g_io_channel_win32_poll() call here has indicated
354 readability or not. (#325310)
356 2005-12-29 Matthias Clasen <mclasen@redhat.com>
358 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
359 in the docs. (#325273, Declan Naughton)
361 2005-12-27 Manish Singh <yosh@gimp.org>
363 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
364 make sure we really override things.
366 2005-12-27 Matthias Clasen <mclasen@redhat.com>
368 Fix #316221, Michal Benes, Stanislav Brabec;
370 * configure.in: Fix a strict aliasing problem in
371 g_static_mutex_get_mutex().
372 * glib/gthread.h: ...and in
373 g_static_mutex_get_mutex_impl_shortcut().
375 * glib/gdatasetprivate.h: Add a cast to silence compiler
376 warnings. (#321978, Andrew Paprocki)
378 Partial fix for bug #323937, Bogdan Nicula.
380 * configure.in: Check for malloc.h
381 * glib/gslice.c: Don't include malloc.h unconditionally.
383 2005-12-27 Manish Singh <yosh@gimp.org>
385 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
386 (#324950, Dan Yefimov)
388 2005-12-26 Matthias Clasen <mclasen@redhat.com>
390 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
393 2005-12-25 Matthias Clasen <mclasen@redhat.com>
395 * AUTHORS: Update my email
397 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
399 2005-12-21 Matthias Clasen <mclasen@redhat.com>
401 * README.in: Add some notes on when 'make check' may fail.
403 2005-12-20 Matthias Clasen <mclasen@redhat.com>
406 * glib/gthreadpool.h:
407 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
408 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
410 * tests/threadpool-test.c: Test this.
412 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
414 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
415 handle errors without depending on gmessage.c which might not be
416 setup when the error occours.
417 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
418 always freeing can be achieved by adjusting the working set time to
419 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
420 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
421 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
422 minimize wastage if small amounts of differently sized structrues are
423 allocated, this does come at a performance cost of roughly 5% though.
424 fixed up block alignment calculation, so it works for varying
425 block sizes. only use strerror() not g_strerror() since the latter
426 depends on working GQuark and GSlice.
427 mem_error(): implemented in terms of fprintf and vfprintf.
429 * tests/slice-color.c: new program to test cache colorization effects.
431 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
432 G_SLICE_CONFIG_WORKING_SET_MSECS.
434 2005-12-17 Matthias Clasen <mclasen@redhat.com>
436 * glib/goption.c (parse_short_option): Set an error in all
437 failure cases. (#324332, Tim-Philipp Müller)
439 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
441 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
442 and g_atomic_pointer_get. Add g_atomic_int_set and
443 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
444 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
445 as functions (additionally to the macros in the header) for the
446 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
448 2005-12-16 Matthias Clasen <mclasen@redhat.com>
450 * glib/gmem.c (g_allocator_new): Don't return a pointer to
451 a const struct, since apps expect to be able to modify it.
452 (#324179, J. Ali Harlow)
454 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
456 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
457 for systems where the initialization of atomic variables requires a
458 write memory barrier.
460 2005-12-09 Matthias Clasen <mclasen@redhat.com>
462 * configure.in: Bump version
464 * === Released 2.9.1 ===
466 2005-12-09 Alexander Larsson <alexl@redhat.com>
470 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
473 2005-12-08 Matthias Clasen <mclasen@redhat.com>
477 2005-12-07 Martyn Russell <martyn@imendio.com>
479 * glib/gasyncqueue.c:
480 - Call g_queue_insert_sorted() instead of duplicating the code.
481 - Call g_queue_sort() instead of duplicating the code.
482 - Invert sort function results to make sure the same sort function
483 gives the same results across glist, gslist, gqueue and
486 * tests/asyncqueue-test.c:
487 - Updated the sort function to reflect the example in the
488 documentation for gasyncqueue.c.
490 2005-12-07 Martyn Russell <martyn@imendio.com>
492 * docs/reference/glib/glib-sections.txt:
493 * docs/reference/glib/tmpl/linked_lists_double.sgml:
494 * docs/reference/glib/tmpl/linked_lists_single.sgml:
497 - Added g_list_insert_sorted_with_data () and
498 g_slist_insert_sorted_with_data ().
499 - Removed the extra check in g_list_sort() and g_slist_sort() for
500 GCompareDataFunc vs. GCompareFunc.
502 2005-12-07 Tor Lillqvist <tml@novell.com>
504 * glib-gettextize.in: Look up prefix at run-time on Win32,
505 assuming the standard directory structure with glib-gettextize in
508 2005-12-06 Matthias Clasen <mclasen@redhat.com>
510 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
511 we need GTK+ 2.8 to compile against GLib 2.10.
513 2005-12-05 Manish Singh <yosh@gimp.org>
515 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
516 to override any user settings.
518 2005-12-05 Matthias Clasen <mclasen@redhat.com>
520 * configure.in: Revert an accidential commit
522 2005-12-05 Matthias Clasen <mclasen@redhat.com>
524 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
526 * tests/Makefile.am (TESTS_ENVIRONMENT):
527 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
528 MALLOC_CHECK_ and MALLOC_PERTURB_
530 * tests/run-collate-tests.sh: Run the collation tests explicitly
531 in en_US locale. (#320463)
533 * glib/gmem.h: Really deprecate GMemChunk.
540 * glib/gstrfuncs.c: Add versioned deprecation docs.
542 2005-12-05 Martyn Russell <martyn@imendio.com>
544 * docs/reference/glib/glib-sections.txt:
545 * glib/gasyncqueue.[ch]:
546 - Added support for sorting async queues by with _push_sorted(),
547 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
550 * tests/asyncqueue-test.c:
551 - Added test case for gasyncqueue.c
553 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
555 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
556 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
557 in bug 323178. moved time consuming logic from g_slice_free() out of
558 the inner loop, so g_slice_free_chain_with_offset() provides a real
559 performance benefit over g_slice_free1() now.
561 * glib/gslice.h: renamed g_slice_free_chain() to
562 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
563 a type-safe macro as suggested in bug 323178.
564 simplified the macro implementation of g_slice_free() and implemented
565 it in a type safe manner for all compliers as suggested by Morten
566 Welinder <mortenw@gnome.org>.
571 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
573 2005-12-05 Matthias Clasen <mclasen@redhat.com>
575 * glib/gasyncqueue.c: Add some docs.
577 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
579 * glib/gatomic.c: In the ia64 implementation, use
580 __sync builtin without _si or _di suffix. (#321229,
581 Stanislav Brabec, patch by Andreas Schwab)
583 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
585 * glib/gslice.h: Remove comma at the end of enum.
587 2005-12-04 Matthias Clasen <mclasen@redhat.com>
589 Handle multiple user names with the same UID better.
590 (#319535, Laszlo Peter)
592 * glib/gutils.c (g_get_any_init_do): When determining user
593 data, first look up $LOGNAME. If the UID doesn't match
594 getuid(), fall back to the current behaviour of looking
595 up the user data based on getuid().
597 2005-12-04 Matthias Clasen <mclasen@redhat.com>
599 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
600 to make gcc warn if a function result is ignored. (#145466,
601 Arjan van de Ven, Alex Larsson)
603 * glib/gmem.h: Add the new attribute to g_realloc and
606 2005-12-03 Matthias Clasen <mclasen@redhat.com>
609 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
612 * glib/gmappedfile.c: Make mapping of empty files
615 * glib/gfileutils.c: Don't fork a new process just to
616 fix the permissions of the created temp file. (#321318,
617 Alexis S. L. Carvalho)
619 2005-12-02 Matthias Clasen <mclasen@redhat.com>
621 * README.in: Add a note about Solaris threads.
624 * configure.in: Remove support for Solaris threads.
625 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
627 2005-12-02 Matthias Clasen <mclasen@redhat.com>
629 * configure.in: Define G_GUINT64_CONSTANT in analogy to
630 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
632 2005-12-02 Matthias Clasen <mclasen@redhat.com>
634 * glib/gslice.c: Win32 portability fixes and C99ism removal,
635 pointed out by Kazuki Iwamoto. (#323052)
637 2005-12-02 Matthias Clasen <mclasen@redhat.com>
639 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
640 the IA64 implementation.
642 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
644 * glib/gslice.c: some naming and type size fixups.
646 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
648 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
649 variant into something less verbose, but digestible for gcc-3.4.
651 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
653 * tests/slice-test.c: extended to perform the benchmarking on the old
654 memchunk code if 'O' is selected.
656 * tests/memchunks.c: new file which contains the old GLib mem chunks
657 implementation with prefix old_mem_chunk_.
659 * tests/Makefile.am: added memchunks.c
661 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
663 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
664 acceptable load distribution of the syscall.
665 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
666 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
668 2005-12-01 Matthias Clasen <mclasen@redhat.com>
670 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
671 double-free base interface members.
673 * tests/Makefile.am: Remove duplicate variable.
675 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
677 * glib/gslice.[hc]: new slice allocator implementation.
679 * tests/slice-test.c: added random slice allocation test.
681 * glib/gthread.[hc]: removed newly added private thread mem API.
683 * glib/gthreadinit.h:
686 * glib/gmem.c: divided glib threading initialisation into three phases,
687 initialisation where private keys and messaging are not available (only
688 needed by gmem.c), initialisation without messaging but private keys
689 available (gslice.c, gmessage.c), and full fledged initialisers that
690 server the rest of glib. initialisation functions got renamed to reflect
691 the limitations of their corresponding phases.
693 * glib/gmem.c: removed memchunk code, defer allocations to
696 * glib/gmem.[hc]: removed g_slice_* skeletons.
698 * glib/glib.symbols: added g_slice_* symbols.
700 * configure.in: check for availability of posix_memalign(3), memalign(3)
703 * glib/Makefile.am: added gslice.[hc].
705 2005-12-01 Tor Lillqvist <tml@novell.com>
707 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
708 trailing slash(es) for non-root folders. stat() fails if non-root
709 folders are specified with trailing slashes. It's too much hassle
710 to demand that callers strip such slashes themselves, especially
711 as it is easy to get it wrong and strip the slash of a root
713 (g_rename): On NT-based Windows, use MoveFileEx() with
714 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
716 2005-11-28 Matthias Clasen <mclasen@redhat.com>
718 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
721 * configure.in: Check whether do { } while (0) works.
723 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
724 G_STMT_END if it works.
726 2005-11-28 Matthias Clasen <mclasen@redhat.com>
728 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
729 * glib/gnode.c (g_node_depth_traverse_level):
730 * glib/gmem.c (g_allocator_new):
731 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
732 (g_hash_table_foreach_remove_or_steal): Silence compiler
735 2005-11-27 Matthias Clasen <mclasen@redhat.com>
737 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
738 Don't read beyond len. (#322520, Christian Persch)
740 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
742 * glib/gdataset.c: access datalist flags via atomic pointer access
743 functions, instead of acquiring the dataset lock. this is faster and
744 also matches the atomic pointer readouts in gdatalistprivate.h and
745 direct pointer modifications required by gobject.c.
747 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
749 * glib/gdataset.c: streamlined initialization code somewhat,
750 removed GData node cache.
752 2005-11-22 Matthias Clasen <mclasen@redhat.com>
754 * glib/ghash.c (g_hash_table_ref):
755 (g_hash_table_unref): Mark these as new API.
757 Avoid double locking in g_intern_string (#322133,
760 * glib/gdataset.c (g_quark_from_string_internal): New
761 internal function which factors out common parts of
762 g_quark_from[_static]_string.
763 (g_quark_from_string, g_quark_from_static_string):
764 Use g_quark_from_string_internal.
765 (g_intern_string, g_intern_static_string): Use
766 g_quark_from_string_internal, and only take the
768 (g_quark_new): Don't store the strings shifted by -1
769 in the g_quarks array.
770 (g_quark_to_string): Adapt to the previous change.
772 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
776 g_hash_table_new_full(): create hash tables with a ref count of 1.
777 g_hash_table_ref(): atomically ref_count+=1
778 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
779 when refcount reaches 0.
780 g_hash_table_destroy(): just destroy keys and values, unref by 1.
781 g_hash_table_insert():
782 g_hash_table_replace(): assert ref_count>0.
785 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
788 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
790 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
791 variables. (#321984, Andrew Paprocki)
793 2005-11-18 Matthias Clasen <mclasen@redhat.com>
795 * configure.in: Bump version
797 2005-11-17 Matthias Clasen <mclasen@redhat.com>
799 * === Released 2.9.0 ===
801 * glib/glib.symbols: Updates
803 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
804 hints to the docs. (Paolo Borelli)
808 * glib/gthread.c (g_thread_foreach): Mark as new api.
810 * README.in: Updates.
814 * glib/gdate.c (g_date_set_time_t):
815 (g_date_set_time_val): Add functions to set a date from
816 a time_t and from a GTimeVal.
817 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
820 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
821 and g_utf8_offset_to_pointer.
823 * glib/gutf8.c (g_utf8_pointer_to_offset)
824 (g_utf8_offset_to_pointer): Handle negative offsets, and use
825 "stutter stepping" for going backwards. (#320638, Larry
830 * glib/gthread.c: const correctness fixes, found
831 by Arjan van de Ven and gcc.
833 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
835 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
836 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
838 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
840 * glib/guniprop.c: Use bit hacks instead when checking a general
841 category value against multiple values.
843 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
845 2005-11-10 Simos Xenitellis <simos@gnome.org>
847 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
849 2005-11-05 Tor Lillqvist <tml@novell.com>
851 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
852 (#320688) Instead set error and return failure.
854 2005-11-07 Matthias Clasen <mclasen@redhat.com>
856 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
857 while holding the context lock. (#320886, Andy Wingo)
859 2005-11-07 Billy Biggs <vektor@dumbterm.net>
861 * tests/refcount/Makefile.am: Add a missing reference to libglib
862 in the LDADD for this test.
864 2005-11-07 Matthias Clasen <mclasen@redhat.com>
866 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
867 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
868 GLib 2.9. (#320755, Luca Ferretti)
870 2005-11-04 Matthias Clasen <mclasen@redhat.com>
872 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
873 use a faster lookup table for g_unichar_get_mirror_char().
874 * glib/gmirroringtable.h: The new table.
875 * glib/gunichartables.h: Remove bidi_mirroring_table.
876 * glib/gen-unicode-tables.pl: Don't generate the mirroring
878 * glib/glib-mirroring-tab/*: Sources for the program
879 which generated gmirroringtable.h.
881 * glib/glist.c: Avoid some code duplication.
883 * glib/gscanner.h: Include gdataset.h. (#320322)
885 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
886 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
887 to help find critical warnings in applications. (#320017,
890 2005-11-02 Tor Lillqvist <tml@novell.com>
892 * glib/glib.symbols: Remove large amount of trailing whitespace
893 from one line. Remove #ifdef G_UNDEFINED from around
894 g_hash_table_new and g_hash_table_new_full symbols, what was the
895 purpose of that? Add the g_slice functions.
897 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
899 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
900 added g_slice_*() API to allocate and cache small bits of memory.
901 an actuall allocator implementation for g_slice_*() is still pending.
903 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
904 changed GRealThread list to use in-structure *next; fields instead
905 of GSList, in order for thread iteration to not depenend on g_slice_*()
907 _g_thread_mem_private_get():
908 _g_thread_mem_private_set(): added accessors for private memory,
909 needed because the ordinary GPrivate implementation relies on GArray
910 and GSList and therefore indirectly on working g_slice_*() allocations.
913 g_thread_foreach(): new public API function to loop over all existing threads.
927 removed GAllocator and free list usages and accompanying locks.
928 use g_slice_*() API to allocate and cache small bits of memory.
930 * glib/ghook.h: removed GMemChunk field from public API.
933 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
935 * glib/gnode.h: deprecate allocator API.
937 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
938 be used for poll rec lists.
940 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
941 g_relation_destroy(): free all tuples from the all_tuples hash table,
942 this effectively maintains the life time track keeping of tuples.
943 g_relation_delete_tuple(): free tuples which are removed from the
944 all_tuples hash table. this fixes a temporary leak that was present
945 in the memchunk code until the destruction of the relation.
947 2005-10-29 Matthias Clasen <mclasen@redhat.com>
949 * tests/convert-test.c: Add some tests for conversions between
950 UTF-8, UCS-4 and UTF-16.
952 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
953 of len == -1, noticed by Morten Welinder.
955 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
957 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
959 2005-10-26 Matthias Clasen <mclasen@redhat.com>
961 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
962 in the error case, and add some documentation. (#319806, Morten
965 2005-10-19 Manish Singh <yosh@gimp.org>
967 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
968 config.h symbol automatically. Fixes bug #313731.
970 2005-10-19 Tor Lillqvist <tml@novell.com>
972 * glib/gwin32.c (g_win32_get_package_installation_directory):
973 Return a g_strdup()ed copy of the value stored in the hash table,
974 so that it can be g_free()d without leaving a dangling pointer in
975 the hash table. (#319232)
977 2005-10-06 Matthias Clasen <mclasen@redhat.com>
979 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
980 all-zero sequences correctly. (#317930, Sebastien Bacher)
982 2005-10-05 Matthias Clasen <mclasen@redhat.com>
984 * glib/gmarkup.c (g_markup_escape_text): Doc addition
986 * Makefile.am: Add Collation tests.
988 * tests/collate/*: Inputs and expected outputs for collation tests.
990 * tests/run-collate-tests.sh: Script to run collation tests.
992 * tests/unicode-collate.c (main): Rework slightly to make
993 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
995 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
997 * docs/reference/glib/tmpl/unicode.sgml:
998 * glib/gen-unicode-tables.pl:
1000 * glib/gunichartables.h:
1002 * tests/casefold.txt:
1003 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1004 GUnicodeBreakType types. That may break some applications, like
1007 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1009 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1010 docs. (#314393, Matthew F. Barnes)
1012 * glib/glib.symbols:
1013 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1014 they are already declared in glib.h. This doesn't break documented
1015 use of gprintf.h, but should probably be pointed out in the
1016 release notes for 2.10. (#314232, Behdad Esfahbod)
1018 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1020 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1021 complexity optimization, based on a patch from Matthias Clasen.
1023 * tests/patterntest.c: more tests, mostly from matthias.
1025 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1027 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1028 (#316703, Mark Drago)
1030 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1032 * glib/gpattern.c: applied a patch from matthias which checks on the
1033 upper bound of GPatternSpec length to optimize matches.
1036 * tests/patterntest.c: added more match cases.
1038 2005-09-16 Tor Lillqvist <tml@novell.com>
1040 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1041 warning to doc comment that these functions should not be used on
1042 encodings like CP932.
1044 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1046 * tests/keyfile-test.c: Add a test for grup names of length 1.
1048 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1049 of length 1. (#316309)
1051 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1053 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1055 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1057 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1058 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1061 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1063 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1064 (child_watch_helper_thread): Remove some dead code. Closes
1067 2005-09-07 Tor Lillqvist <tml@novell.com>
1069 * glib/Makefile.am: Create also a console version of the
1070 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1071 It's otherwise identical to gspawn-win32-helper.exe, except marked
1072 as a console application (linked without the -mwindows option).
1074 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1075 Drop the dont_wait parameter. Its truth value correlated 100% with
1076 the NULLness of the exit_status parameter anyway, so it's enough
1077 to check whether exit_status is NULL. Invert the sense of the
1078 dont_return_handle parameter and rename it to do_return_handle, to
1079 make the code easier to read by avoiding double negations.
1081 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1082 to do_spawn_with_pipes() accordingly.
1084 (do_spawn_with_pipes): If we have a console, use the console
1085 version of the helper program, otherwise use the GUI one. This
1086 avoids extra console windows opening up in some situations. (In
1087 case a console application uses the GUI gspawn-win32-helper.exe to
1088 spawn another console application we would get a separate console
1089 for the spawned console application).
1091 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1093 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1095 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1096 if the file is too large. (#315275, Kjartan Maraas)
1098 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1099 of read() is signed. (#315273, Kjartan Maraas)
1101 2005-08-31 Tor Lillqvist <tml@novell.com>
1103 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1104 the EXE and each DLL have their own atexit function chains.
1106 #define g_atexit as atexit instead. This means it has a
1107 better chance of doing what the caller wants. For instance,
1108 gtkhtml calls g_atexit() registering a function in gtkhtml
1109 itself. This caused a crash when g_atexit() was implemented as a
1110 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1111 the time the GLib DLL got unloaded.
1113 * glib/gutils.c: #undef the #define mentioned above, to also get a
1114 real g_atexit() into the DLL for backward compatibility. Document
1115 the Windows behaviour of g_atexit(), and document the varying ways
1116 atexit() can behave in the context of dynamically loaded modules
1119 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1121 * glib/glib.symbols:
1123 * glib/gdataset.c: Add string interning functions.
1125 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1127 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1130 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1132 * configure.in: Bump version to 2.9.0
1134 2005-08-25 Tor Lillqvist <tml@novell.com>
1136 Make also the g_spawn*() functions take parameters in the GLib
1137 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1138 API or ABI. Like the other GLib API that was earlier changed to
1139 use UTF-8 on Windows, the names of the functions that take UTF-8
1140 have _utf8 suffixes added by using preprocessor macros in the
1141 header file. The old names are kept for functions with the old
1142 behaviour, taking parameters in the system codepage, for DLL ABI
1145 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1146 g_spawn*() functions.
1148 * glib/gspawn-win32.c: Use wide-char API on NT-based
1149 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1150 system codepage (Win9x) and call the C library _wspawn*() or
1151 spawn*() functions respectvely. Add DLL ABI stability versions
1152 that take parameters in the system codepage.
1154 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1155 wide-char versions of argv and envp, and use wide-char API to
1156 change directory and spawn the program to run. Remove the verbose
1157 debugging output, it was too complex to modify for the wide-char
1158 features. (Just add temporary debugging printouts if needed, no
1159 need to have them permanently in the source.)
1161 * glib/gspawn.c: Corresponding documentation updates.
1163 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1164 symbols as PRIVATE, add the new _utf8-suffixed ones.
1166 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1168 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1169 guint64 otherwise and that can produce warnings about comparison
1170 between signed and unsigned.
1172 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1174 * glib/gutils.c: Fix the crt_externs.h include.
1176 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1178 * NEWS: Fix spelling of my first name.
1180 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1184 * === Released 2.8.1 ===
1188 2005-08-20 Hans Breuer <hans@breuer.org>
1190 * glib/makefile.msc.in : link with ws2_32.lib
1192 2005-08-18 Tor Lillqvist <tml@novell.com>
1194 * configure.in: Check for <sys/wait.h>
1196 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1198 2005-08-18 Ross Burton <ross@burtonini.com>
1201 Optimise single-character insertions.
1206 * tests/string-test.c:
1207 Add tests for new optimisation, and fix a leak.
1209 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1211 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1213 * glib/gutils.c: On Darwin, include crt-externs.h and
1214 define environ using _NSGetEnviron(). (#313731)
1216 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1218 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1219 (g_get_any_init_do): ... to this new function.
1220 (g_get_any_init): Declare as inline.
1221 (g_get_any_init_locked): New inline function, does the locking.
1222 Make use of these two throughout the code.
1224 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1226 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1227 the child process and then simply return. This makes
1228 The "S" option work as documented in g_on_error_query().
1229 (#313125, Matthew F. Barnes)
1231 * glib/gunicode.h: Update the link to Unicode category
1232 values. (#313369, Behnam Esfahbod)
1234 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1235 a little. (#311727, Tristan van Berkom)
1237 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1238 check work on ia64 too, where some symbols ended up
1239 in yet another section.