1 2006-02-24 Matthias Clasen <mclasen@redhat.com>
3 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
5 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
8 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
10 * tests/completion-test.c: (main): Plug leaks reported by
12 * tests/convert-test.c: (test_iconv_state), (test_one_half):
14 * tests/patterntest.c: (test_compilation): Same
15 * tests/shell-test.c: (do_argv_test): Same.
16 * tests/unicode-caseconv.c: (main): Same.
17 * tests/uri-test.c: (run_to_uri_tests): Same.
20 2006-02-19 Martyn Russell <martyn@imendio.com>
22 * tests/asyncqueue-test.c:
24 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
25 _insert_sorted and _insert_sorted_with_data API.
27 2006-02-18 Matthias Clasen <mclasen@redhat.com>
29 * tests/gobject/Makefile.am: Add paramspec-test
31 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
33 * tests/gobject/gvalue-test.c: Add more tests.
35 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
37 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
39 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
41 * glib/gthreadpool.c: Fix deadlock when signalling the thread
42 which freed a thread pool (#331110, Chris Wilson).
44 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
46 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
47 posix_memalign() in case we actually are going to use it, because
48 we detected a compliant implementation (#328997).
50 * configure.in (enable_included_printf): don't include malloc.h when
51 testing for posix_memalign() funcitonality, since this may break the
52 test on some systems (#328997).
54 2006-02-14 Matthias Clasen <mclasen@redhat.com>
56 * glib/gstrfuncs.c: Improve docs.
58 2006-02-11 Matthias Clasen <mclasen@redhat.com>
60 * configure.in: Bump version
62 2006-02-10 Matthias Clasen <mclasen@redhat.com>
64 * === Released 2.9.6 ===
68 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
70 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
71 the type of varargs arguments. (bug #317679)
73 2006-01-31 Matthias Clasen <mclasen@redhat.com>
75 * autogen.sh: Touch README and INSTALL here to pacify
76 automake. (#329124, Kjartan Maraas, Tim Janik)
78 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
80 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
81 some missing cases where pointed out by Sebastian Wilhelmi.
83 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
85 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
87 2006-01-30 David Schleef <ds@schleef.org>
89 * glib/grand.c: Update URL
91 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
93 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
94 create container classes which are as flexible as a GValue is
96 2006-01-27 Matthias Clasen <mclasen@redhat.com>
98 * configure.in: Bump version
100 * === Released 2.9.5 ===
102 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
107 2006-01-26 Matthias Clasen <mclasen@redhat.com>
109 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
112 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
114 * fixed buglets reported by Jens Granseuer in #328254.
116 * configure.in: free the memory allocated in posix_memalign() tests.
118 * glib/gslice.c: spelling fixes.
120 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
122 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
123 slices, make sure g_mem_gc_friendly is properly initialized.
125 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
126 the first allocation. applied some branching optimizations.
128 * docs/macros.txt: reflected --enable-gc-friendly change and
129 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
131 * configure.in: changed --enable-gc-friendly=yes to define
132 ENABLE_GC_FRIENDLY_DEFAULT.
134 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
135 if (G_UNLIKELY (g_mem_gc_friendly)).
138 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
139 care of by g_slice_free1().
141 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
143 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
145 * glib/gslice.c: only use posix_memalign() if it's known to work,
146 revert to memalign() otherwise.
148 * configure.in: check for broken posix_memalign() implementations
151 2006-01-24 Matthias Clasen <mclasen@redhat.com>
153 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
154 on all little-endian systems. (#143380, Marc Moorcroft)
156 2006-01-23 Matthias Clasen <mclasen@redhat.com>
158 * configure.in: Change the shared libary extension for hpux-ia64
159 to so. (#328253, The Written Word)
161 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
163 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
164 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
167 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
169 * glib/gslice.c (slice_config_init): initialize GSlice config from
170 G_SLICE environment variable. we support G_SLICE=always-malloc
171 currently, which forces all g_slice_*() allocations to use the system
175 g_parse_debug_string(): added a note about not using g_malloc() here.
176 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
177 g_slice. contains only guesswork in the WIN32 branch.
179 2006-01-18 Matthias Clasen <mclasen@redhat.com>
183 * === Released 2.9.4 ===
187 2006-01-17 Matthias Clasen <mclasen@redhat.com>
189 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
190 call. (#327421, Morten Welinder)
192 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
194 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
195 G_LOCK. Use the unused_thread_queue lock instead. Change
196 g_thread_pool_thread_proxy such that threads only wait on
197 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
198 due to superfluous tasks. Global tasks wait at most for
199 max-idle-time milliseconds. Make sure, that no task is woken up
200 twice for the same event via a wakeup_serial. This fixes #324228.
202 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
203 invalid NULL into the thread pools. This as well fixes #327290.
205 2006-01-16 Matthias Clasen <mclasen@redhat.com>
207 * configure.in: Bump version
209 * === Released 2.9.3 ===
213 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
214 Rework the strict aliasing fix to not break C++, pointed
215 out by Murray Cumming.
217 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
218 Signal waiting threads, problem noticed by Christian Kellner.
220 2006-01-16 Matthias Clasen <mclasen@redhat.com>
222 Fix bug 326747, Alberto Ruiz:
224 * glib/gconvert.c (make_valid_utf8): Change this function to
225 replace unknown characters by the Unicode replacement character
226 instead of '?', and don't append "(invalid encoding)".
227 (g_filename_display_name, g_filename_display_basename): Document
228 how to determine if the filename was in an invalid encoding.
230 2006-01-14 Matthias Clasen <mclasen@redhat.com>
232 * glib/gtree.c: Replace the simple recursive implementation
233 by a nonrecursive, threaded implementation by Maurizio
236 2006-01-12 Matthias Clasen <mclasen@redhat.com>
238 * tests/asyncqueue-test.c (main): Fix the
239 assert to not always trigger. (#326558,
242 2006-01-11 Matthias Clasen <mclasen@redhat.com>
244 * tests/convert-test.c: Don't test items_read and
245 items_written in error cases where it is not set.
247 * tests/asyncqueue-test.c: Set error to NULL.
249 2006-01-08 Matthias Clasen <mclasen@redhat.com>
251 * glib/gtree.c: Break some long lines.
253 * tests/tree-test.c: Add more tests.
255 * glib/gslice.c: Include config.h first, pointed out
258 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
260 * docs/reference/gobject/tmpl/gtype.sgml: explain that
261 G_TYPE_INSTANCE_GET_CLASS() does behave different during
264 2006-01-06 Matthias Clasen <mclasen@redhat.com>
266 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
267 that we have posix_memalign(). (#323937, Bogdan Nicula)
269 2006-01-05 Matthias Clasen <mclasen@redhat.com>
271 * configure.in: Bump version
273 * === Released 2.9.2 ===
275 2006-01-05 Hans Breuer <hans@breuer.org>
277 * glib/makefile.msc.in : added gslice.obj
279 2006-01-05 Matthias Clasen <mclasen@redhat.com>
283 * glib/gmain.c: Clarify the documentation of source ids
284 a bit. (#325874, Dan Williams)
286 * configure.in: Fix another strict aliasing problem.
288 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
291 2006-01-04 Manish Singh <yosh@gimp.org>
294 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
295 either, since they have the same issue as g_rmdir. (Related to
298 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
300 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
301 order when creating magazines, so we hand out chunks with ascending
304 2006-01-04 Abel Cheung <maddog@linuxhall.org>
306 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
308 2006-01-03 Martyn Russell <martyn@imendio.com>
310 * docs/reference/glib/glib-sections.txt:
312 * glib/gthreadpool.[ch]:
313 - Added new API g_thread_pool_get_idle_time() and
314 g_thread_pool_set_idle_time(). (#324228).
316 * tests/threadpool-test.c:
317 - Updated test case to do thread pool sorting, thread pool with
318 no sorting and a thread pool with idle thread timeouts.
320 2006-01-03 Matthias Clasen <mclasen@redhat.com>
322 * glib/gmain.h: Add new functions here, too.
324 * glib/glib.symbols: Add new functions.
326 2005-12-20 Michael Meeks <michael.meeks@novell.com>
328 * glib/gmain.c (g_main_context_is_owner): new method
329 to determine if the current thread is the owner of the
332 2006-01-02 Matthias Clasen <mclasen@redhat.com>
336 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
337 expanding to rmdir, since rmdir is not declared in a portable
338 system header we can include in gstdio.h. (#325249, Jani Monoses)
340 2006-01-01 Tor Lillqvist <tml@novell.com>
342 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
343 stdout and stderr to unbuffered. Otherwise the giochannel layer
344 will try to read from them regardless whether the
345 g_io_channel_win32_poll() call here has indicated
346 readability or not. (#325310)
348 2005-12-29 Matthias Clasen <mclasen@redhat.com>
350 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
351 in the docs. (#325273, Declan Naughton)
353 2005-12-27 Manish Singh <yosh@gimp.org>
355 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
356 make sure we really override things.
358 2005-12-27 Matthias Clasen <mclasen@redhat.com>
360 Fix #316221, Michal Benes, Stanislav Brabec;
362 * configure.in: Fix a strict aliasing problem in
363 g_static_mutex_get_mutex().
364 * glib/gthread.h: ...and in
365 g_static_mutex_get_mutex_impl_shortcut().
367 * glib/gdatasetprivate.h: Add a cast to silence compiler
368 warnings. (#321978, Andrew Paprocki)
370 Partial fix for bug #323937, Bogdan Nicula.
372 * configure.in: Check for malloc.h
373 * glib/gslice.c: Don't include malloc.h unconditionally.
375 2005-12-27 Manish Singh <yosh@gimp.org>
377 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
378 (#324950, Dan Yefimov)
380 2005-12-26 Matthias Clasen <mclasen@redhat.com>
382 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
385 2005-12-25 Matthias Clasen <mclasen@redhat.com>
387 * AUTHORS: Update my email
389 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
391 2005-12-21 Matthias Clasen <mclasen@redhat.com>
393 * README.in: Add some notes on when 'make check' may fail.
395 2005-12-20 Matthias Clasen <mclasen@redhat.com>
398 * glib/gthreadpool.h:
399 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
400 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
402 * tests/threadpool-test.c: Test this.
404 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
406 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
407 handle errors without depending on gmessage.c which might not be
408 setup when the error occours.
409 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
410 always freeing can be achieved by adjusting the working set time to
411 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
412 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
413 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
414 minimize wastage if small amounts of differently sized structrues are
415 allocated, this does come at a performance cost of roughly 5% though.
416 fixed up block alignment calculation, so it works for varying
417 block sizes. only use strerror() not g_strerror() since the latter
418 depends on working GQuark and GSlice.
419 mem_error(): implemented in terms of fprintf and vfprintf.
421 * tests/slice-color.c: new program to test cache colorization effects.
423 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
424 G_SLICE_CONFIG_WORKING_SET_MSECS.
426 2005-12-17 Matthias Clasen <mclasen@redhat.com>
428 * glib/goption.c (parse_short_option): Set an error in all
429 failure cases. (#324332, Tim-Philipp Müller)
431 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
433 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
434 and g_atomic_pointer_get. Add g_atomic_int_set and
435 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
436 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
437 as functions (additionally to the macros in the header) for the
438 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
440 2005-12-16 Matthias Clasen <mclasen@redhat.com>
442 * glib/gmem.c (g_allocator_new): Don't return a pointer to
443 a const struct, since apps expect to be able to modify it.
444 (#324179, J. Ali Harlow)
446 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
448 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
449 for systems where the initialization of atomic variables requires a
450 write memory barrier.
452 2005-12-09 Matthias Clasen <mclasen@redhat.com>
454 * configure.in: Bump version
456 * === Released 2.9.1 ===
458 2005-12-09 Alexander Larsson <alexl@redhat.com>
462 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
465 2005-12-08 Matthias Clasen <mclasen@redhat.com>
469 2005-12-07 Martyn Russell <martyn@imendio.com>
471 * glib/gasyncqueue.c:
472 - Call g_queue_insert_sorted() instead of duplicating the code.
473 - Call g_queue_sort() instead of duplicating the code.
474 - Invert sort function results to make sure the same sort function
475 gives the same results across glist, gslist, gqueue and
478 * tests/asyncqueue-test.c:
479 - Updated the sort function to reflect the example in the
480 documentation for gasyncqueue.c.
482 2005-12-07 Martyn Russell <martyn@imendio.com>
484 * docs/reference/glib/glib-sections.txt:
485 * docs/reference/glib/tmpl/linked_lists_double.sgml:
486 * docs/reference/glib/tmpl/linked_lists_single.sgml:
489 - Added g_list_insert_sorted_with_data () and
490 g_slist_insert_sorted_with_data ().
491 - Removed the extra check in g_list_sort() and g_slist_sort() for
492 GCompareDataFunc vs. GCompareFunc.
494 2005-12-07 Tor Lillqvist <tml@novell.com>
496 * glib-gettextize.in: Look up prefix at run-time on Win32,
497 assuming the standard directory structure with glib-gettextize in
500 2005-12-06 Matthias Clasen <mclasen@redhat.com>
502 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
503 we need GTK+ 2.8 to compile against GLib 2.10.
505 2005-12-05 Manish Singh <yosh@gimp.org>
507 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
508 to override any user settings.
510 2005-12-05 Matthias Clasen <mclasen@redhat.com>
512 * configure.in: Revert an accidential commit
514 2005-12-05 Matthias Clasen <mclasen@redhat.com>
516 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
518 * tests/Makefile.am (TESTS_ENVIRONMENT):
519 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
520 MALLOC_CHECK_ and MALLOC_PERTURB_
522 * tests/run-collate-tests.sh: Run the collation tests explicitly
523 in en_US locale. (#320463)
525 * glib/gmem.h: Really deprecate GMemChunk.
532 * glib/gstrfuncs.c: Add versioned deprecation docs.
534 2005-12-05 Martyn Russell <martyn@imendio.com>
536 * docs/reference/glib/glib-sections.txt:
537 * glib/gasyncqueue.[ch]:
538 - Added support for sorting async queues by with _push_sorted(),
539 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
542 * tests/asyncqueue-test.c:
543 - Added test case for gasyncqueue.c
545 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
547 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
548 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
549 in bug 323178. moved time consuming logic from g_slice_free() out of
550 the inner loop, so g_slice_free_chain_with_offset() provides a real
551 performance benefit over g_slice_free1() now.
553 * glib/gslice.h: renamed g_slice_free_chain() to
554 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
555 a type-safe macro as suggested in bug 323178.
556 simplified the macro implementation of g_slice_free() and implemented
557 it in a type safe manner for all compliers as suggested by Morten
558 Welinder <mortenw@gnome.org>.
563 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
565 2005-12-05 Matthias Clasen <mclasen@redhat.com>
567 * glib/gasyncqueue.c: Add some docs.
569 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
571 * glib/gatomic.c: In the ia64 implementation, use
572 __sync builtin without _si or _di suffix. (#321229,
573 Stanislav Brabec, patch by Andreas Schwab)
575 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
577 * glib/gslice.h: Remove comma at the end of enum.
579 2005-12-04 Matthias Clasen <mclasen@redhat.com>
581 Handle multiple user names with the same UID better.
582 (#319535, Laszlo Peter)
584 * glib/gutils.c (g_get_any_init_do): When determining user
585 data, first look up $LOGNAME. If the UID doesn't match
586 getuid(), fall back to the current behaviour of looking
587 up the user data based on getuid().
589 2005-12-04 Matthias Clasen <mclasen@redhat.com>
591 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
592 to make gcc warn if a function result is ignored. (#145466,
593 Arjan van de Ven, Alex Larsson)
595 * glib/gmem.h: Add the new attribute to g_realloc and
598 2005-12-03 Matthias Clasen <mclasen@redhat.com>
601 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
604 * glib/gmappedfile.c: Make mapping of empty files
607 * glib/gfileutils.c: Don't fork a new process just to
608 fix the permissions of the created temp file. (#321318,
609 Alexis S. L. Carvalho)
611 2005-12-02 Matthias Clasen <mclasen@redhat.com>
613 * README.in: Add a note about Solaris threads.
616 * configure.in: Remove support for Solaris threads.
617 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
619 2005-12-02 Matthias Clasen <mclasen@redhat.com>
621 * configure.in: Define G_GUINT64_CONSTANT in analogy to
622 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
624 2005-12-02 Matthias Clasen <mclasen@redhat.com>
626 * glib/gslice.c: Win32 portability fixes and C99ism removal,
627 pointed out by Kazuki Iwamoto. (#323052)
629 2005-12-02 Matthias Clasen <mclasen@redhat.com>
631 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
632 the IA64 implementation.
634 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
636 * glib/gslice.c: some naming and type size fixups.
638 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
640 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
641 variant into something less verbose, but digestible for gcc-3.4.
643 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
645 * tests/slice-test.c: extended to perform the benchmarking on the old
646 memchunk code if 'O' is selected.
648 * tests/memchunks.c: new file which contains the old GLib mem chunks
649 implementation with prefix old_mem_chunk_.
651 * tests/Makefile.am: added memchunks.c
653 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
655 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
656 acceptable load distribution of the syscall.
657 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
658 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
660 2005-12-01 Matthias Clasen <mclasen@redhat.com>
662 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
663 double-free base interface members.
665 * tests/Makefile.am: Remove duplicate variable.
667 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
669 * glib/gslice.[hc]: new slice allocator implementation.
671 * tests/slice-test.c: added random slice allocation test.
673 * glib/gthread.[hc]: removed newly added private thread mem API.
675 * glib/gthreadinit.h:
678 * glib/gmem.c: divided glib threading initialisation into three phases,
679 initialisation where private keys and messaging are not available (only
680 needed by gmem.c), initialisation without messaging but private keys
681 available (gslice.c, gmessage.c), and full fledged initialisers that
682 server the rest of glib. initialisation functions got renamed to reflect
683 the limitations of their corresponding phases.
685 * glib/gmem.c: removed memchunk code, defer allocations to
688 * glib/gmem.[hc]: removed g_slice_* skeletons.
690 * glib/glib.symbols: added g_slice_* symbols.
692 * configure.in: check for availability of posix_memalign(3), memalign(3)
695 * glib/Makefile.am: added gslice.[hc].
697 2005-12-01 Tor Lillqvist <tml@novell.com>
699 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
700 trailing slash(es) for non-root folders. stat() fails if non-root
701 folders are specified with trailing slashes. It's too much hassle
702 to demand that callers strip such slashes themselves, especially
703 as it is easy to get it wrong and strip the slash of a root
705 (g_rename): On NT-based Windows, use MoveFileEx() with
706 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
708 2005-11-28 Matthias Clasen <mclasen@redhat.com>
710 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
713 * configure.in: Check whether do { } while (0) works.
715 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
716 G_STMT_END if it works.
718 2005-11-28 Matthias Clasen <mclasen@redhat.com>
720 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
721 * glib/gnode.c (g_node_depth_traverse_level):
722 * glib/gmem.c (g_allocator_new):
723 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
724 (g_hash_table_foreach_remove_or_steal): Silence compiler
727 2005-11-27 Matthias Clasen <mclasen@redhat.com>
729 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
730 Don't read beyond len. (#322520, Christian Persch)
732 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
734 * glib/gdataset.c: access datalist flags via atomic pointer access
735 functions, instead of acquiring the dataset lock. this is faster and
736 also matches the atomic pointer readouts in gdatalistprivate.h and
737 direct pointer modifications required by gobject.c.
739 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
741 * glib/gdataset.c: streamlined initialization code somewhat,
742 removed GData node cache.
744 2005-11-22 Matthias Clasen <mclasen@redhat.com>
746 * glib/ghash.c (g_hash_table_ref):
747 (g_hash_table_unref): Mark these as new API.
749 Avoid double locking in g_intern_string (#322133,
752 * glib/gdataset.c (g_quark_from_string_internal): New
753 internal function which factors out common parts of
754 g_quark_from[_static]_string.
755 (g_quark_from_string, g_quark_from_static_string):
756 Use g_quark_from_string_internal.
757 (g_intern_string, g_intern_static_string): Use
758 g_quark_from_string_internal, and only take the
760 (g_quark_new): Don't store the strings shifted by -1
761 in the g_quarks array.
762 (g_quark_to_string): Adapt to the previous change.
764 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
768 g_hash_table_new_full(): create hash tables with a ref count of 1.
769 g_hash_table_ref(): atomically ref_count+=1
770 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
771 when refcount reaches 0.
772 g_hash_table_destroy(): just destroy keys and values, unref by 1.
773 g_hash_table_insert():
774 g_hash_table_replace(): assert ref_count>0.
777 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
780 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
782 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
783 variables. (#321984, Andrew Paprocki)
785 2005-11-18 Matthias Clasen <mclasen@redhat.com>
787 * configure.in: Bump version
789 2005-11-17 Matthias Clasen <mclasen@redhat.com>
791 * === Released 2.9.0 ===
793 * glib/glib.symbols: Updates
795 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
796 hints to the docs. (Paolo Borelli)
800 * glib/gthread.c (g_thread_foreach): Mark as new api.
802 * README.in: Updates.
806 * glib/gdate.c (g_date_set_time_t):
807 (g_date_set_time_val): Add functions to set a date from
808 a time_t and from a GTimeVal.
809 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
812 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
813 and g_utf8_offset_to_pointer.
815 * glib/gutf8.c (g_utf8_pointer_to_offset)
816 (g_utf8_offset_to_pointer): Handle negative offsets, and use
817 "stutter stepping" for going backwards. (#320638, Larry
822 * glib/gthread.c: const correctness fixes, found
823 by Arjan van de Ven and gcc.
825 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
827 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
828 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
830 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
832 * glib/guniprop.c: Use bit hacks instead when checking a general
833 category value against multiple values.
835 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
837 2005-11-10 Simos Xenitellis <simos@gnome.org>
839 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
841 2005-11-05 Tor Lillqvist <tml@novell.com>
843 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
844 (#320688) Instead set error and return failure.
846 2005-11-07 Matthias Clasen <mclasen@redhat.com>
848 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
849 while holding the context lock. (#320886, Andy Wingo)
851 2005-11-07 Billy Biggs <vektor@dumbterm.net>
853 * tests/refcount/Makefile.am: Add a missing reference to libglib
854 in the LDADD for this test.
856 2005-11-07 Matthias Clasen <mclasen@redhat.com>
858 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
859 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
860 GLib 2.9. (#320755, Luca Ferretti)
862 2005-11-04 Matthias Clasen <mclasen@redhat.com>
864 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
865 use a faster lookup table for g_unichar_get_mirror_char().
866 * glib/gmirroringtable.h: The new table.
867 * glib/gunichartables.h: Remove bidi_mirroring_table.
868 * glib/gen-unicode-tables.pl: Don't generate the mirroring
870 * glib/glib-mirroring-tab/*: Sources for the program
871 which generated gmirroringtable.h.
873 * glib/glist.c: Avoid some code duplication.
875 * glib/gscanner.h: Include gdataset.h. (#320322)
877 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
878 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
879 to help find critical warnings in applications. (#320017,
882 2005-11-02 Tor Lillqvist <tml@novell.com>
884 * glib/glib.symbols: Remove large amount of trailing whitespace
885 from one line. Remove #ifdef G_UNDEFINED from around
886 g_hash_table_new and g_hash_table_new_full symbols, what was the
887 purpose of that? Add the g_slice functions.
889 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
891 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
892 added g_slice_*() API to allocate and cache small bits of memory.
893 an actuall allocator implementation for g_slice_*() is still pending.
895 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
896 changed GRealThread list to use in-structure *next; fields instead
897 of GSList, in order for thread iteration to not depenend on g_slice_*()
899 _g_thread_mem_private_get():
900 _g_thread_mem_private_set(): added accessors for private memory,
901 needed because the ordinary GPrivate implementation relies on GArray
902 and GSList and therefore indirectly on working g_slice_*() allocations.
905 g_thread_foreach(): new public API function to loop over all existing threads.
919 removed GAllocator and free list usages and accompanying locks.
920 use g_slice_*() API to allocate and cache small bits of memory.
922 * glib/ghook.h: removed GMemChunk field from public API.
925 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
927 * glib/gnode.h: deprecate allocator API.
929 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
930 be used for poll rec lists.
932 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
933 g_relation_destroy(): free all tuples from the all_tuples hash table,
934 this effectively maintains the life time track keeping of tuples.
935 g_relation_delete_tuple(): free tuples which are removed from the
936 all_tuples hash table. this fixes a temporary leak that was present
937 in the memchunk code until the destruction of the relation.
939 2005-10-29 Matthias Clasen <mclasen@redhat.com>
941 * tests/convert-test.c: Add some tests for conversions between
942 UTF-8, UCS-4 and UTF-16.
944 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
945 of len == -1, noticed by Morten Welinder.
947 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
949 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
951 2005-10-26 Matthias Clasen <mclasen@redhat.com>
953 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
954 in the error case, and add some documentation. (#319806, Morten
957 2005-10-19 Manish Singh <yosh@gimp.org>
959 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
960 config.h symbol automatically. Fixes bug #313731.
962 2005-10-19 Tor Lillqvist <tml@novell.com>
964 * glib/gwin32.c (g_win32_get_package_installation_directory):
965 Return a g_strdup()ed copy of the value stored in the hash table,
966 so that it can be g_free()d without leaving a dangling pointer in
967 the hash table. (#319232)
969 2005-10-06 Matthias Clasen <mclasen@redhat.com>
971 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
972 all-zero sequences correctly. (#317930, Sebastien Bacher)
974 2005-10-05 Matthias Clasen <mclasen@redhat.com>
976 * glib/gmarkup.c (g_markup_escape_text): Doc addition
978 * Makefile.am: Add Collation tests.
980 * tests/collate/*: Inputs and expected outputs for collation tests.
982 * tests/run-collate-tests.sh: Script to run collation tests.
984 * tests/unicode-collate.c (main): Rework slightly to make
985 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
987 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
989 * docs/reference/glib/tmpl/unicode.sgml:
990 * glib/gen-unicode-tables.pl:
992 * glib/gunichartables.h:
994 * tests/casefold.txt:
995 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
996 GUnicodeBreakType types. That may break some applications, like
999 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1001 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1002 docs. (#314393, Matthew F. Barnes)
1004 * glib/glib.symbols:
1005 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1006 they are already declared in glib.h. This doesn't break documented
1007 use of gprintf.h, but should probably be pointed out in the
1008 release notes for 2.10. (#314232, Behdad Esfahbod)
1010 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1012 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1013 complexity optimization, based on a patch from Matthias Clasen.
1015 * tests/patterntest.c: more tests, mostly from matthias.
1017 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1019 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1020 (#316703, Mark Drago)
1022 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1024 * glib/gpattern.c: applied a patch from matthias which checks on the
1025 upper bound of GPatternSpec length to optimize matches.
1028 * tests/patterntest.c: added more match cases.
1030 2005-09-16 Tor Lillqvist <tml@novell.com>
1032 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1033 warning to doc comment that these functions should not be used on
1034 encodings like CP932.
1036 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1038 * tests/keyfile-test.c: Add a test for grup names of length 1.
1040 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1041 of length 1. (#316309)
1043 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1045 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1047 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1049 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1050 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1053 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1055 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1056 (child_watch_helper_thread): Remove some dead code. Closes
1059 2005-09-07 Tor Lillqvist <tml@novell.com>
1061 * glib/Makefile.am: Create also a console version of the
1062 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1063 It's otherwise identical to gspawn-win32-helper.exe, except marked
1064 as a console application (linked without the -mwindows option).
1066 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1067 Drop the dont_wait parameter. Its truth value correlated 100% with
1068 the NULLness of the exit_status parameter anyway, so it's enough
1069 to check whether exit_status is NULL. Invert the sense of the
1070 dont_return_handle parameter and rename it to do_return_handle, to
1071 make the code easier to read by avoiding double negations.
1073 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1074 to do_spawn_with_pipes() accordingly.
1076 (do_spawn_with_pipes): If we have a console, use the console
1077 version of the helper program, otherwise use the GUI one. This
1078 avoids extra console windows opening up in some situations. (In
1079 case a console application uses the GUI gspawn-win32-helper.exe to
1080 spawn another console application we would get a separate console
1081 for the spawned console application).
1083 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1085 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1087 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1088 if the file is too large. (#315275, Kjartan Maraas)
1090 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1091 of read() is signed. (#315273, Kjartan Maraas)
1093 2005-08-31 Tor Lillqvist <tml@novell.com>
1095 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1096 the EXE and each DLL have their own atexit function chains.
1098 #define g_atexit as atexit instead. This means it has a
1099 better chance of doing what the caller wants. For instance,
1100 gtkhtml calls g_atexit() registering a function in gtkhtml
1101 itself. This caused a crash when g_atexit() was implemented as a
1102 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1103 the time the GLib DLL got unloaded.
1105 * glib/gutils.c: #undef the #define mentioned above, to also get a
1106 real g_atexit() into the DLL for backward compatibility. Document
1107 the Windows behaviour of g_atexit(), and document the varying ways
1108 atexit() can behave in the context of dynamically loaded modules
1111 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1113 * glib/glib.symbols:
1115 * glib/gdataset.c: Add string interning functions.
1117 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1119 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1122 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1124 * configure.in: Bump version to 2.9.0
1126 2005-08-25 Tor Lillqvist <tml@novell.com>
1128 Make also the g_spawn*() functions take parameters in the GLib
1129 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1130 API or ABI. Like the other GLib API that was earlier changed to
1131 use UTF-8 on Windows, the names of the functions that take UTF-8
1132 have _utf8 suffixes added by using preprocessor macros in the
1133 header file. The old names are kept for functions with the old
1134 behaviour, taking parameters in the system codepage, for DLL ABI
1137 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1138 g_spawn*() functions.
1140 * glib/gspawn-win32.c: Use wide-char API on NT-based
1141 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1142 system codepage (Win9x) and call the C library _wspawn*() or
1143 spawn*() functions respectvely. Add DLL ABI stability versions
1144 that take parameters in the system codepage.
1146 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1147 wide-char versions of argv and envp, and use wide-char API to
1148 change directory and spawn the program to run. Remove the verbose
1149 debugging output, it was too complex to modify for the wide-char
1150 features. (Just add temporary debugging printouts if needed, no
1151 need to have them permanently in the source.)
1153 * glib/gspawn.c: Corresponding documentation updates.
1155 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1156 symbols as PRIVATE, add the new _utf8-suffixed ones.
1158 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1160 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1161 guint64 otherwise and that can produce warnings about comparison
1162 between signed and unsigned.
1164 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1166 * glib/gutils.c: Fix the crt_externs.h include.
1168 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1170 * NEWS: Fix spelling of my first name.
1172 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1176 * === Released 2.8.1 ===
1180 2005-08-20 Hans Breuer <hans@breuer.org>
1182 * glib/makefile.msc.in : link with ws2_32.lib
1184 2005-08-18 Tor Lillqvist <tml@novell.com>
1186 * configure.in: Check for <sys/wait.h>
1188 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1190 2005-08-18 Ross Burton <ross@burtonini.com>
1193 Optimise single-character insertions.
1198 * tests/string-test.c:
1199 Add tests for new optimisation, and fix a leak.
1201 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1203 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1205 * glib/gutils.c: On Darwin, include crt-externs.h and
1206 define environ using _NSGetEnviron(). (#313731)
1208 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1210 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1211 (g_get_any_init_do): ... to this new function.
1212 (g_get_any_init): Declare as inline.
1213 (g_get_any_init_locked): New inline function, does the locking.
1214 Make use of these two throughout the code.
1216 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1218 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1219 the child process and then simply return. This makes
1220 The "S" option work as documented in g_on_error_query().
1221 (#313125, Matthew F. Barnes)
1223 * glib/gunicode.h: Update the link to Unicode category
1224 values. (#313369, Behnam Esfahbod)
1226 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1227 a little. (#311727, Tristan van Berkom)
1229 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1230 check work on ia64 too, where some symbols ended up
1231 in yet another section.