1 2006-02-24 Matthias Clasen <mclasen@redhat.com>
3 * configure.in: Bump version
5 * === Released 2.10.0 ===
7 * README.in: Mention the slice allocator and GInitiallyUnowned.
11 * configure.in: Bump version to 2.10.0
13 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
15 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
18 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
20 * tests/completion-test.c: (main): Plug leaks reported by
22 * tests/convert-test.c: (test_iconv_state), (test_one_half):
24 * tests/patterntest.c: (test_compilation): Same
25 * tests/shell-test.c: (do_argv_test): Same.
26 * tests/unicode-caseconv.c: (main): Same.
27 * tests/uri-test.c: (run_to_uri_tests): Same.
30 2006-02-19 Martyn Russell <martyn@imendio.com>
32 * tests/asyncqueue-test.c:
34 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
35 _insert_sorted and _insert_sorted_with_data API.
37 2006-02-18 Matthias Clasen <mclasen@redhat.com>
39 * tests/gobject/Makefile.am: Add paramspec-test
41 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
43 * tests/gobject/gvalue-test.c: Add more tests.
45 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
47 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
49 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
51 * glib/gthreadpool.c: Fix deadlock when signalling the thread
52 which freed a thread pool (#331110, Chris Wilson).
54 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
56 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
57 posix_memalign() in case we actually are going to use it, because
58 we detected a compliant implementation (#328997).
60 * configure.in (enable_included_printf): don't include malloc.h when
61 testing for posix_memalign() funcitonality, since this may break the
62 test on some systems (#328997).
64 2006-02-14 Matthias Clasen <mclasen@redhat.com>
66 * glib/gstrfuncs.c: Improve docs.
68 2006-02-11 Matthias Clasen <mclasen@redhat.com>
70 * configure.in: Bump version
72 2006-02-10 Matthias Clasen <mclasen@redhat.com>
74 * === Released 2.9.6 ===
78 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
80 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
81 the type of varargs arguments. (bug #317679)
83 2006-01-31 Matthias Clasen <mclasen@redhat.com>
85 * autogen.sh: Touch README and INSTALL here to pacify
86 automake. (#329124, Kjartan Maraas, Tim Janik)
88 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
90 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
91 some missing cases where pointed out by Sebastian Wilhelmi.
93 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
95 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
97 2006-01-30 David Schleef <ds@schleef.org>
99 * glib/grand.c: Update URL
101 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
103 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
104 create container classes which are as flexible as a GValue is
106 2006-01-27 Matthias Clasen <mclasen@redhat.com>
108 * configure.in: Bump version
110 * === Released 2.9.5 ===
112 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
117 2006-01-26 Matthias Clasen <mclasen@redhat.com>
119 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
122 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
124 * fixed buglets reported by Jens Granseuer in #328254.
126 * configure.in: free the memory allocated in posix_memalign() tests.
128 * glib/gslice.c: spelling fixes.
130 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
132 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
133 slices, make sure g_mem_gc_friendly is properly initialized.
135 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
136 the first allocation. applied some branching optimizations.
138 * docs/macros.txt: reflected --enable-gc-friendly change and
139 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
141 * configure.in: changed --enable-gc-friendly=yes to define
142 ENABLE_GC_FRIENDLY_DEFAULT.
144 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
145 if (G_UNLIKELY (g_mem_gc_friendly)).
148 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
149 care of by g_slice_free1().
151 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
153 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
155 * glib/gslice.c: only use posix_memalign() if it's known to work,
156 revert to memalign() otherwise.
158 * configure.in: check for broken posix_memalign() implementations
161 2006-01-24 Matthias Clasen <mclasen@redhat.com>
163 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
164 on all little-endian systems. (#143380, Marc Moorcroft)
166 2006-01-23 Matthias Clasen <mclasen@redhat.com>
168 * configure.in: Change the shared libary extension for hpux-ia64
169 to so. (#328253, The Written Word)
171 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
173 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
174 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
177 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
179 * glib/gslice.c (slice_config_init): initialize GSlice config from
180 G_SLICE environment variable. we support G_SLICE=always-malloc
181 currently, which forces all g_slice_*() allocations to use the system
185 g_parse_debug_string(): added a note about not using g_malloc() here.
186 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
187 g_slice. contains only guesswork in the WIN32 branch.
189 2006-01-18 Matthias Clasen <mclasen@redhat.com>
193 * === Released 2.9.4 ===
197 2006-01-17 Matthias Clasen <mclasen@redhat.com>
199 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
200 call. (#327421, Morten Welinder)
202 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
204 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
205 G_LOCK. Use the unused_thread_queue lock instead. Change
206 g_thread_pool_thread_proxy such that threads only wait on
207 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
208 due to superfluous tasks. Global tasks wait at most for
209 max-idle-time milliseconds. Make sure, that no task is woken up
210 twice for the same event via a wakeup_serial. This fixes #324228.
212 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
213 invalid NULL into the thread pools. This as well fixes #327290.
215 2006-01-16 Matthias Clasen <mclasen@redhat.com>
217 * configure.in: Bump version
219 * === Released 2.9.3 ===
223 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
224 Rework the strict aliasing fix to not break C++, pointed
225 out by Murray Cumming.
227 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
228 Signal waiting threads, problem noticed by Christian Kellner.
230 2006-01-16 Matthias Clasen <mclasen@redhat.com>
232 Fix bug 326747, Alberto Ruiz:
234 * glib/gconvert.c (make_valid_utf8): Change this function to
235 replace unknown characters by the Unicode replacement character
236 instead of '?', and don't append "(invalid encoding)".
237 (g_filename_display_name, g_filename_display_basename): Document
238 how to determine if the filename was in an invalid encoding.
240 2006-01-14 Matthias Clasen <mclasen@redhat.com>
242 * glib/gtree.c: Replace the simple recursive implementation
243 by a nonrecursive, threaded implementation by Maurizio
246 2006-01-12 Matthias Clasen <mclasen@redhat.com>
248 * tests/asyncqueue-test.c (main): Fix the
249 assert to not always trigger. (#326558,
252 2006-01-11 Matthias Clasen <mclasen@redhat.com>
254 * tests/convert-test.c: Don't test items_read and
255 items_written in error cases where it is not set.
257 * tests/asyncqueue-test.c: Set error to NULL.
259 2006-01-08 Matthias Clasen <mclasen@redhat.com>
261 * glib/gtree.c: Break some long lines.
263 * tests/tree-test.c: Add more tests.
265 * glib/gslice.c: Include config.h first, pointed out
268 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
270 * docs/reference/gobject/tmpl/gtype.sgml: explain that
271 G_TYPE_INSTANCE_GET_CLASS() does behave different during
274 2006-01-06 Matthias Clasen <mclasen@redhat.com>
276 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
277 that we have posix_memalign(). (#323937, Bogdan Nicula)
279 2006-01-05 Matthias Clasen <mclasen@redhat.com>
281 * configure.in: Bump version
283 * === Released 2.9.2 ===
285 2006-01-05 Hans Breuer <hans@breuer.org>
287 * glib/makefile.msc.in : added gslice.obj
289 2006-01-05 Matthias Clasen <mclasen@redhat.com>
293 * glib/gmain.c: Clarify the documentation of source ids
294 a bit. (#325874, Dan Williams)
296 * configure.in: Fix another strict aliasing problem.
298 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
301 2006-01-04 Manish Singh <yosh@gimp.org>
304 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
305 either, since they have the same issue as g_rmdir. (Related to
308 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
310 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
311 order when creating magazines, so we hand out chunks with ascending
314 2006-01-04 Abel Cheung <maddog@linuxhall.org>
316 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
318 2006-01-03 Martyn Russell <martyn@imendio.com>
320 * docs/reference/glib/glib-sections.txt:
322 * glib/gthreadpool.[ch]:
323 - Added new API g_thread_pool_get_idle_time() and
324 g_thread_pool_set_idle_time(). (#324228).
326 * tests/threadpool-test.c:
327 - Updated test case to do thread pool sorting, thread pool with
328 no sorting and a thread pool with idle thread timeouts.
330 2006-01-03 Matthias Clasen <mclasen@redhat.com>
332 * glib/gmain.h: Add new functions here, too.
334 * glib/glib.symbols: Add new functions.
336 2005-12-20 Michael Meeks <michael.meeks@novell.com>
338 * glib/gmain.c (g_main_context_is_owner): new method
339 to determine if the current thread is the owner of the
342 2006-01-02 Matthias Clasen <mclasen@redhat.com>
346 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
347 expanding to rmdir, since rmdir is not declared in a portable
348 system header we can include in gstdio.h. (#325249, Jani Monoses)
350 2006-01-01 Tor Lillqvist <tml@novell.com>
352 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
353 stdout and stderr to unbuffered. Otherwise the giochannel layer
354 will try to read from them regardless whether the
355 g_io_channel_win32_poll() call here has indicated
356 readability or not. (#325310)
358 2005-12-29 Matthias Clasen <mclasen@redhat.com>
360 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
361 in the docs. (#325273, Declan Naughton)
363 2005-12-27 Manish Singh <yosh@gimp.org>
365 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
366 make sure we really override things.
368 2005-12-27 Matthias Clasen <mclasen@redhat.com>
370 Fix #316221, Michal Benes, Stanislav Brabec;
372 * configure.in: Fix a strict aliasing problem in
373 g_static_mutex_get_mutex().
374 * glib/gthread.h: ...and in
375 g_static_mutex_get_mutex_impl_shortcut().
377 * glib/gdatasetprivate.h: Add a cast to silence compiler
378 warnings. (#321978, Andrew Paprocki)
380 Partial fix for bug #323937, Bogdan Nicula.
382 * configure.in: Check for malloc.h
383 * glib/gslice.c: Don't include malloc.h unconditionally.
385 2005-12-27 Manish Singh <yosh@gimp.org>
387 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
388 (#324950, Dan Yefimov)
390 2005-12-26 Matthias Clasen <mclasen@redhat.com>
392 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
395 2005-12-25 Matthias Clasen <mclasen@redhat.com>
397 * AUTHORS: Update my email
399 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
401 2005-12-21 Matthias Clasen <mclasen@redhat.com>
403 * README.in: Add some notes on when 'make check' may fail.
405 2005-12-20 Matthias Clasen <mclasen@redhat.com>
408 * glib/gthreadpool.h:
409 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
410 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
412 * tests/threadpool-test.c: Test this.
414 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
416 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
417 handle errors without depending on gmessage.c which might not be
418 setup when the error occours.
419 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
420 always freeing can be achieved by adjusting the working set time to
421 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
422 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
423 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
424 minimize wastage if small amounts of differently sized structrues are
425 allocated, this does come at a performance cost of roughly 5% though.
426 fixed up block alignment calculation, so it works for varying
427 block sizes. only use strerror() not g_strerror() since the latter
428 depends on working GQuark and GSlice.
429 mem_error(): implemented in terms of fprintf and vfprintf.
431 * tests/slice-color.c: new program to test cache colorization effects.
433 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
434 G_SLICE_CONFIG_WORKING_SET_MSECS.
436 2005-12-17 Matthias Clasen <mclasen@redhat.com>
438 * glib/goption.c (parse_short_option): Set an error in all
439 failure cases. (#324332, Tim-Philipp Müller)
441 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
443 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
444 and g_atomic_pointer_get. Add g_atomic_int_set and
445 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
446 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
447 as functions (additionally to the macros in the header) for the
448 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
450 2005-12-16 Matthias Clasen <mclasen@redhat.com>
452 * glib/gmem.c (g_allocator_new): Don't return a pointer to
453 a const struct, since apps expect to be able to modify it.
454 (#324179, J. Ali Harlow)
456 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
458 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
459 for systems where the initialization of atomic variables requires a
460 write memory barrier.
462 2005-12-09 Matthias Clasen <mclasen@redhat.com>
464 * configure.in: Bump version
466 * === Released 2.9.1 ===
468 2005-12-09 Alexander Larsson <alexl@redhat.com>
472 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
475 2005-12-08 Matthias Clasen <mclasen@redhat.com>
479 2005-12-07 Martyn Russell <martyn@imendio.com>
481 * glib/gasyncqueue.c:
482 - Call g_queue_insert_sorted() instead of duplicating the code.
483 - Call g_queue_sort() instead of duplicating the code.
484 - Invert sort function results to make sure the same sort function
485 gives the same results across glist, gslist, gqueue and
488 * tests/asyncqueue-test.c:
489 - Updated the sort function to reflect the example in the
490 documentation for gasyncqueue.c.
492 2005-12-07 Martyn Russell <martyn@imendio.com>
494 * docs/reference/glib/glib-sections.txt:
495 * docs/reference/glib/tmpl/linked_lists_double.sgml:
496 * docs/reference/glib/tmpl/linked_lists_single.sgml:
499 - Added g_list_insert_sorted_with_data () and
500 g_slist_insert_sorted_with_data ().
501 - Removed the extra check in g_list_sort() and g_slist_sort() for
502 GCompareDataFunc vs. GCompareFunc.
504 2005-12-07 Tor Lillqvist <tml@novell.com>
506 * glib-gettextize.in: Look up prefix at run-time on Win32,
507 assuming the standard directory structure with glib-gettextize in
510 2005-12-06 Matthias Clasen <mclasen@redhat.com>
512 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
513 we need GTK+ 2.8 to compile against GLib 2.10.
515 2005-12-05 Manish Singh <yosh@gimp.org>
517 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
518 to override any user settings.
520 2005-12-05 Matthias Clasen <mclasen@redhat.com>
522 * configure.in: Revert an accidential commit
524 2005-12-05 Matthias Clasen <mclasen@redhat.com>
526 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
528 * tests/Makefile.am (TESTS_ENVIRONMENT):
529 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
530 MALLOC_CHECK_ and MALLOC_PERTURB_
532 * tests/run-collate-tests.sh: Run the collation tests explicitly
533 in en_US locale. (#320463)
535 * glib/gmem.h: Really deprecate GMemChunk.
542 * glib/gstrfuncs.c: Add versioned deprecation docs.
544 2005-12-05 Martyn Russell <martyn@imendio.com>
546 * docs/reference/glib/glib-sections.txt:
547 * glib/gasyncqueue.[ch]:
548 - Added support for sorting async queues by with _push_sorted(),
549 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
552 * tests/asyncqueue-test.c:
553 - Added test case for gasyncqueue.c
555 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
557 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
558 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
559 in bug 323178. moved time consuming logic from g_slice_free() out of
560 the inner loop, so g_slice_free_chain_with_offset() provides a real
561 performance benefit over g_slice_free1() now.
563 * glib/gslice.h: renamed g_slice_free_chain() to
564 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
565 a type-safe macro as suggested in bug 323178.
566 simplified the macro implementation of g_slice_free() and implemented
567 it in a type safe manner for all compliers as suggested by Morten
568 Welinder <mortenw@gnome.org>.
573 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
575 2005-12-05 Matthias Clasen <mclasen@redhat.com>
577 * glib/gasyncqueue.c: Add some docs.
579 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
581 * glib/gatomic.c: In the ia64 implementation, use
582 __sync builtin without _si or _di suffix. (#321229,
583 Stanislav Brabec, patch by Andreas Schwab)
585 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
587 * glib/gslice.h: Remove comma at the end of enum.
589 2005-12-04 Matthias Clasen <mclasen@redhat.com>
591 Handle multiple user names with the same UID better.
592 (#319535, Laszlo Peter)
594 * glib/gutils.c (g_get_any_init_do): When determining user
595 data, first look up $LOGNAME. If the UID doesn't match
596 getuid(), fall back to the current behaviour of looking
597 up the user data based on getuid().
599 2005-12-04 Matthias Clasen <mclasen@redhat.com>
601 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
602 to make gcc warn if a function result is ignored. (#145466,
603 Arjan van de Ven, Alex Larsson)
605 * glib/gmem.h: Add the new attribute to g_realloc and
608 2005-12-03 Matthias Clasen <mclasen@redhat.com>
611 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
614 * glib/gmappedfile.c: Make mapping of empty files
617 * glib/gfileutils.c: Don't fork a new process just to
618 fix the permissions of the created temp file. (#321318,
619 Alexis S. L. Carvalho)
621 2005-12-02 Matthias Clasen <mclasen@redhat.com>
623 * README.in: Add a note about Solaris threads.
626 * configure.in: Remove support for Solaris threads.
627 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
629 2005-12-02 Matthias Clasen <mclasen@redhat.com>
631 * configure.in: Define G_GUINT64_CONSTANT in analogy to
632 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
634 2005-12-02 Matthias Clasen <mclasen@redhat.com>
636 * glib/gslice.c: Win32 portability fixes and C99ism removal,
637 pointed out by Kazuki Iwamoto. (#323052)
639 2005-12-02 Matthias Clasen <mclasen@redhat.com>
641 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
642 the IA64 implementation.
644 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
646 * glib/gslice.c: some naming and type size fixups.
648 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
650 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
651 variant into something less verbose, but digestible for gcc-3.4.
653 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
655 * tests/slice-test.c: extended to perform the benchmarking on the old
656 memchunk code if 'O' is selected.
658 * tests/memchunks.c: new file which contains the old GLib mem chunks
659 implementation with prefix old_mem_chunk_.
661 * tests/Makefile.am: added memchunks.c
663 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
665 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
666 acceptable load distribution of the syscall.
667 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
668 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
670 2005-12-01 Matthias Clasen <mclasen@redhat.com>
672 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
673 double-free base interface members.
675 * tests/Makefile.am: Remove duplicate variable.
677 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
679 * glib/gslice.[hc]: new slice allocator implementation.
681 * tests/slice-test.c: added random slice allocation test.
683 * glib/gthread.[hc]: removed newly added private thread mem API.
685 * glib/gthreadinit.h:
688 * glib/gmem.c: divided glib threading initialisation into three phases,
689 initialisation where private keys and messaging are not available (only
690 needed by gmem.c), initialisation without messaging but private keys
691 available (gslice.c, gmessage.c), and full fledged initialisers that
692 server the rest of glib. initialisation functions got renamed to reflect
693 the limitations of their corresponding phases.
695 * glib/gmem.c: removed memchunk code, defer allocations to
698 * glib/gmem.[hc]: removed g_slice_* skeletons.
700 * glib/glib.symbols: added g_slice_* symbols.
702 * configure.in: check for availability of posix_memalign(3), memalign(3)
705 * glib/Makefile.am: added gslice.[hc].
707 2005-12-01 Tor Lillqvist <tml@novell.com>
709 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
710 trailing slash(es) for non-root folders. stat() fails if non-root
711 folders are specified with trailing slashes. It's too much hassle
712 to demand that callers strip such slashes themselves, especially
713 as it is easy to get it wrong and strip the slash of a root
715 (g_rename): On NT-based Windows, use MoveFileEx() with
716 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
718 2005-11-28 Matthias Clasen <mclasen@redhat.com>
720 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
723 * configure.in: Check whether do { } while (0) works.
725 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
726 G_STMT_END if it works.
728 2005-11-28 Matthias Clasen <mclasen@redhat.com>
730 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
731 * glib/gnode.c (g_node_depth_traverse_level):
732 * glib/gmem.c (g_allocator_new):
733 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
734 (g_hash_table_foreach_remove_or_steal): Silence compiler
737 2005-11-27 Matthias Clasen <mclasen@redhat.com>
739 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
740 Don't read beyond len. (#322520, Christian Persch)
742 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
744 * glib/gdataset.c: access datalist flags via atomic pointer access
745 functions, instead of acquiring the dataset lock. this is faster and
746 also matches the atomic pointer readouts in gdatalistprivate.h and
747 direct pointer modifications required by gobject.c.
749 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
751 * glib/gdataset.c: streamlined initialization code somewhat,
752 removed GData node cache.
754 2005-11-22 Matthias Clasen <mclasen@redhat.com>
756 * glib/ghash.c (g_hash_table_ref):
757 (g_hash_table_unref): Mark these as new API.
759 Avoid double locking in g_intern_string (#322133,
762 * glib/gdataset.c (g_quark_from_string_internal): New
763 internal function which factors out common parts of
764 g_quark_from[_static]_string.
765 (g_quark_from_string, g_quark_from_static_string):
766 Use g_quark_from_string_internal.
767 (g_intern_string, g_intern_static_string): Use
768 g_quark_from_string_internal, and only take the
770 (g_quark_new): Don't store the strings shifted by -1
771 in the g_quarks array.
772 (g_quark_to_string): Adapt to the previous change.
774 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
778 g_hash_table_new_full(): create hash tables with a ref count of 1.
779 g_hash_table_ref(): atomically ref_count+=1
780 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
781 when refcount reaches 0.
782 g_hash_table_destroy(): just destroy keys and values, unref by 1.
783 g_hash_table_insert():
784 g_hash_table_replace(): assert ref_count>0.
787 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
790 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
792 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
793 variables. (#321984, Andrew Paprocki)
795 2005-11-18 Matthias Clasen <mclasen@redhat.com>
797 * configure.in: Bump version
799 2005-11-17 Matthias Clasen <mclasen@redhat.com>
801 * === Released 2.9.0 ===
803 * glib/glib.symbols: Updates
805 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
806 hints to the docs. (Paolo Borelli)
810 * glib/gthread.c (g_thread_foreach): Mark as new api.
812 * README.in: Updates.
816 * glib/gdate.c (g_date_set_time_t):
817 (g_date_set_time_val): Add functions to set a date from
818 a time_t and from a GTimeVal.
819 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
822 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
823 and g_utf8_offset_to_pointer.
825 * glib/gutf8.c (g_utf8_pointer_to_offset)
826 (g_utf8_offset_to_pointer): Handle negative offsets, and use
827 "stutter stepping" for going backwards. (#320638, Larry
832 * glib/gthread.c: const correctness fixes, found
833 by Arjan van de Ven and gcc.
835 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
837 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
838 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
840 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
842 * glib/guniprop.c: Use bit hacks instead when checking a general
843 category value against multiple values.
845 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
847 2005-11-10 Simos Xenitellis <simos@gnome.org>
849 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
851 2005-11-05 Tor Lillqvist <tml@novell.com>
853 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
854 (#320688) Instead set error and return failure.
856 2005-11-07 Matthias Clasen <mclasen@redhat.com>
858 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
859 while holding the context lock. (#320886, Andy Wingo)
861 2005-11-07 Billy Biggs <vektor@dumbterm.net>
863 * tests/refcount/Makefile.am: Add a missing reference to libglib
864 in the LDADD for this test.
866 2005-11-07 Matthias Clasen <mclasen@redhat.com>
868 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
869 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
870 GLib 2.9. (#320755, Luca Ferretti)
872 2005-11-04 Matthias Clasen <mclasen@redhat.com>
874 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
875 use a faster lookup table for g_unichar_get_mirror_char().
876 * glib/gmirroringtable.h: The new table.
877 * glib/gunichartables.h: Remove bidi_mirroring_table.
878 * glib/gen-unicode-tables.pl: Don't generate the mirroring
880 * glib/glib-mirroring-tab/*: Sources for the program
881 which generated gmirroringtable.h.
883 * glib/glist.c: Avoid some code duplication.
885 * glib/gscanner.h: Include gdataset.h. (#320322)
887 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
888 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
889 to help find critical warnings in applications. (#320017,
892 2005-11-02 Tor Lillqvist <tml@novell.com>
894 * glib/glib.symbols: Remove large amount of trailing whitespace
895 from one line. Remove #ifdef G_UNDEFINED from around
896 g_hash_table_new and g_hash_table_new_full symbols, what was the
897 purpose of that? Add the g_slice functions.
899 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
901 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
902 added g_slice_*() API to allocate and cache small bits of memory.
903 an actuall allocator implementation for g_slice_*() is still pending.
905 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
906 changed GRealThread list to use in-structure *next; fields instead
907 of GSList, in order for thread iteration to not depenend on g_slice_*()
909 _g_thread_mem_private_get():
910 _g_thread_mem_private_set(): added accessors for private memory,
911 needed because the ordinary GPrivate implementation relies on GArray
912 and GSList and therefore indirectly on working g_slice_*() allocations.
915 g_thread_foreach(): new public API function to loop over all existing threads.
929 removed GAllocator and free list usages and accompanying locks.
930 use g_slice_*() API to allocate and cache small bits of memory.
932 * glib/ghook.h: removed GMemChunk field from public API.
935 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
937 * glib/gnode.h: deprecate allocator API.
939 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
940 be used for poll rec lists.
942 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
943 g_relation_destroy(): free all tuples from the all_tuples hash table,
944 this effectively maintains the life time track keeping of tuples.
945 g_relation_delete_tuple(): free tuples which are removed from the
946 all_tuples hash table. this fixes a temporary leak that was present
947 in the memchunk code until the destruction of the relation.
949 2005-10-29 Matthias Clasen <mclasen@redhat.com>
951 * tests/convert-test.c: Add some tests for conversions between
952 UTF-8, UCS-4 and UTF-16.
954 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
955 of len == -1, noticed by Morten Welinder.
957 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
959 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
961 2005-10-26 Matthias Clasen <mclasen@redhat.com>
963 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
964 in the error case, and add some documentation. (#319806, Morten
967 2005-10-19 Manish Singh <yosh@gimp.org>
969 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
970 config.h symbol automatically. Fixes bug #313731.
972 2005-10-19 Tor Lillqvist <tml@novell.com>
974 * glib/gwin32.c (g_win32_get_package_installation_directory):
975 Return a g_strdup()ed copy of the value stored in the hash table,
976 so that it can be g_free()d without leaving a dangling pointer in
977 the hash table. (#319232)
979 2005-10-06 Matthias Clasen <mclasen@redhat.com>
981 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
982 all-zero sequences correctly. (#317930, Sebastien Bacher)
984 2005-10-05 Matthias Clasen <mclasen@redhat.com>
986 * glib/gmarkup.c (g_markup_escape_text): Doc addition
988 * Makefile.am: Add Collation tests.
990 * tests/collate/*: Inputs and expected outputs for collation tests.
992 * tests/run-collate-tests.sh: Script to run collation tests.
994 * tests/unicode-collate.c (main): Rework slightly to make
995 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
997 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
999 * docs/reference/glib/tmpl/unicode.sgml:
1000 * glib/gen-unicode-tables.pl:
1002 * glib/gunichartables.h:
1004 * tests/casefold.txt:
1005 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1006 GUnicodeBreakType types. That may break some applications, like
1009 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1011 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1012 docs. (#314393, Matthew F. Barnes)
1014 * glib/glib.symbols:
1015 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1016 they are already declared in glib.h. This doesn't break documented
1017 use of gprintf.h, but should probably be pointed out in the
1018 release notes for 2.10. (#314232, Behdad Esfahbod)
1020 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1022 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1023 complexity optimization, based on a patch from Matthias Clasen.
1025 * tests/patterntest.c: more tests, mostly from matthias.
1027 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1029 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1030 (#316703, Mark Drago)
1032 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1034 * glib/gpattern.c: applied a patch from matthias which checks on the
1035 upper bound of GPatternSpec length to optimize matches.
1038 * tests/patterntest.c: added more match cases.
1040 2005-09-16 Tor Lillqvist <tml@novell.com>
1042 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1043 warning to doc comment that these functions should not be used on
1044 encodings like CP932.
1046 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1048 * tests/keyfile-test.c: Add a test for grup names of length 1.
1050 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1051 of length 1. (#316309)
1053 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1055 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1057 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1059 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1060 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1063 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1065 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1066 (child_watch_helper_thread): Remove some dead code. Closes
1069 2005-09-07 Tor Lillqvist <tml@novell.com>
1071 * glib/Makefile.am: Create also a console version of the
1072 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1073 It's otherwise identical to gspawn-win32-helper.exe, except marked
1074 as a console application (linked without the -mwindows option).
1076 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1077 Drop the dont_wait parameter. Its truth value correlated 100% with
1078 the NULLness of the exit_status parameter anyway, so it's enough
1079 to check whether exit_status is NULL. Invert the sense of the
1080 dont_return_handle parameter and rename it to do_return_handle, to
1081 make the code easier to read by avoiding double negations.
1083 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1084 to do_spawn_with_pipes() accordingly.
1086 (do_spawn_with_pipes): If we have a console, use the console
1087 version of the helper program, otherwise use the GUI one. This
1088 avoids extra console windows opening up in some situations. (In
1089 case a console application uses the GUI gspawn-win32-helper.exe to
1090 spawn another console application we would get a separate console
1091 for the spawned console application).
1093 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1095 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1097 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1098 if the file is too large. (#315275, Kjartan Maraas)
1100 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1101 of read() is signed. (#315273, Kjartan Maraas)
1103 2005-08-31 Tor Lillqvist <tml@novell.com>
1105 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1106 the EXE and each DLL have their own atexit function chains.
1108 #define g_atexit as atexit instead. This means it has a
1109 better chance of doing what the caller wants. For instance,
1110 gtkhtml calls g_atexit() registering a function in gtkhtml
1111 itself. This caused a crash when g_atexit() was implemented as a
1112 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1113 the time the GLib DLL got unloaded.
1115 * glib/gutils.c: #undef the #define mentioned above, to also get a
1116 real g_atexit() into the DLL for backward compatibility. Document
1117 the Windows behaviour of g_atexit(), and document the varying ways
1118 atexit() can behave in the context of dynamically loaded modules
1121 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1123 * glib/glib.symbols:
1125 * glib/gdataset.c: Add string interning functions.
1127 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1129 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1132 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1134 * configure.in: Bump version to 2.9.0
1136 2005-08-25 Tor Lillqvist <tml@novell.com>
1138 Make also the g_spawn*() functions take parameters in the GLib
1139 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1140 API or ABI. Like the other GLib API that was earlier changed to
1141 use UTF-8 on Windows, the names of the functions that take UTF-8
1142 have _utf8 suffixes added by using preprocessor macros in the
1143 header file. The old names are kept for functions with the old
1144 behaviour, taking parameters in the system codepage, for DLL ABI
1147 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1148 g_spawn*() functions.
1150 * glib/gspawn-win32.c: Use wide-char API on NT-based
1151 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1152 system codepage (Win9x) and call the C library _wspawn*() or
1153 spawn*() functions respectvely. Add DLL ABI stability versions
1154 that take parameters in the system codepage.
1156 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1157 wide-char versions of argv and envp, and use wide-char API to
1158 change directory and spawn the program to run. Remove the verbose
1159 debugging output, it was too complex to modify for the wide-char
1160 features. (Just add temporary debugging printouts if needed, no
1161 need to have them permanently in the source.)
1163 * glib/gspawn.c: Corresponding documentation updates.
1165 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1166 symbols as PRIVATE, add the new _utf8-suffixed ones.
1168 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1170 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1171 guint64 otherwise and that can produce warnings about comparison
1172 between signed and unsigned.
1174 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1176 * glib/gutils.c: Fix the crt_externs.h include.
1178 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1180 * NEWS: Fix spelling of my first name.
1182 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1186 * === Released 2.8.1 ===
1190 2005-08-20 Hans Breuer <hans@breuer.org>
1192 * glib/makefile.msc.in : link with ws2_32.lib
1194 2005-08-18 Tor Lillqvist <tml@novell.com>
1196 * configure.in: Check for <sys/wait.h>
1198 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1200 2005-08-18 Ross Burton <ross@burtonini.com>
1203 Optimise single-character insertions.
1208 * tests/string-test.c:
1209 Add tests for new optimisation, and fix a leak.
1211 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1213 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1215 * glib/gutils.c: On Darwin, include crt-externs.h and
1216 define environ using _NSGetEnviron(). (#313731)
1218 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1220 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1221 (g_get_any_init_do): ... to this new function.
1222 (g_get_any_init): Declare as inline.
1223 (g_get_any_init_locked): New inline function, does the locking.
1224 Make use of these two throughout the code.
1226 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1228 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1229 the child process and then simply return. This makes
1230 The "S" option work as documented in g_on_error_query().
1231 (#313125, Matthew F. Barnes)
1233 * glib/gunicode.h: Update the link to Unicode category
1234 values. (#313369, Behnam Esfahbod)
1236 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1237 a little. (#311727, Tristan van Berkom)
1239 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1240 check work on ia64 too, where some symbols ended up
1241 in yet another section.