1 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
3 * tests/completion-test.c: (main): Plug leaks reported by
5 * tests/convert-test.c: (test_iconv_state), (test_one_half):
7 * tests/patterntest.c: (test_compilation): Same
8 * tests/shell-test.c: (do_argv_test): Same.
9 * tests/unicode-caseconv.c: (main): Same.
10 * tests/uri-test.c: (run_to_uri_tests): Same.
13 2006-02-19 Martyn Russell <martyn@imendio.com>
15 * tests/asyncqueue-test.c:
17 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
18 _insert_sorted and _insert_sorted_with_data API.
20 2006-02-18 Matthias Clasen <mclasen@redhat.com>
22 * tests/gobject/Makefile.am: Add paramspec-test
24 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
26 * tests/gobject/gvalue-test.c: Add more tests.
28 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
30 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
32 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
34 * glib/gthreadpool.c: Fix deadlock when signalling the thread
35 which freed a thread pool (#331110, Chris Wilson).
37 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
39 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
40 posix_memalign() in case we actually are going to use it, because
41 we detected a compliant implementation (#328997).
43 * configure.in (enable_included_printf): don't include malloc.h when
44 testing for posix_memalign() funcitonality, since this may break the
45 test on some systems (#328997).
47 2006-02-14 Matthias Clasen <mclasen@redhat.com>
49 * glib/gstrfuncs.c: Improve docs.
51 2006-02-11 Matthias Clasen <mclasen@redhat.com>
53 * configure.in: Bump version
55 2006-02-10 Matthias Clasen <mclasen@redhat.com>
57 * === Released 2.9.6 ===
61 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
63 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
64 the type of varargs arguments. (bug #317679)
66 2006-01-31 Matthias Clasen <mclasen@redhat.com>
68 * autogen.sh: Touch README and INSTALL here to pacify
69 automake. (#329124, Kjartan Maraas, Tim Janik)
71 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
73 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
74 some missing cases where pointed out by Sebastian Wilhelmi.
76 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
78 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
80 2006-01-30 David Schleef <ds@schleef.org>
82 * glib/grand.c: Update URL
84 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
86 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
87 create container classes which are as flexible as a GValue is
89 2006-01-27 Matthias Clasen <mclasen@redhat.com>
91 * configure.in: Bump version
93 * === Released 2.9.5 ===
95 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
100 2006-01-26 Matthias Clasen <mclasen@redhat.com>
102 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
105 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
107 * fixed buglets reported by Jens Granseuer in #328254.
109 * configure.in: free the memory allocated in posix_memalign() tests.
111 * glib/gslice.c: spelling fixes.
113 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
115 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
116 slices, make sure g_mem_gc_friendly is properly initialized.
118 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
119 the first allocation. applied some branching optimizations.
121 * docs/macros.txt: reflected --enable-gc-friendly change and
122 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
124 * configure.in: changed --enable-gc-friendly=yes to define
125 ENABLE_GC_FRIENDLY_DEFAULT.
127 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
128 if (G_UNLIKELY (g_mem_gc_friendly)).
131 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
132 care of by g_slice_free1().
134 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
136 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
138 * glib/gslice.c: only use posix_memalign() if it's known to work,
139 revert to memalign() otherwise.
141 * configure.in: check for broken posix_memalign() implementations
144 2006-01-24 Matthias Clasen <mclasen@redhat.com>
146 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
147 on all little-endian systems. (#143380, Marc Moorcroft)
149 2006-01-23 Matthias Clasen <mclasen@redhat.com>
151 * configure.in: Change the shared libary extension for hpux-ia64
152 to so. (#328253, The Written Word)
154 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
156 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
157 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
160 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
162 * glib/gslice.c (slice_config_init): initialize GSlice config from
163 G_SLICE environment variable. we support G_SLICE=always-malloc
164 currently, which forces all g_slice_*() allocations to use the system
168 g_parse_debug_string(): added a note about not using g_malloc() here.
169 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
170 g_slice. contains only guesswork in the WIN32 branch.
172 2006-01-18 Matthias Clasen <mclasen@redhat.com>
176 * === Released 2.9.4 ===
180 2006-01-17 Matthias Clasen <mclasen@redhat.com>
182 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
183 call. (#327421, Morten Welinder)
185 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
187 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
188 G_LOCK. Use the unused_thread_queue lock instead. Change
189 g_thread_pool_thread_proxy such that threads only wait on
190 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
191 due to superfluous tasks. Global tasks wait at most for
192 max-idle-time milliseconds. Make sure, that no task is woken up
193 twice for the same event via a wakeup_serial. This fixes #324228.
195 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
196 invalid NULL into the thread pools. This as well fixes #327290.
198 2006-01-16 Matthias Clasen <mclasen@redhat.com>
200 * configure.in: Bump version
202 * === Released 2.9.3 ===
206 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
207 Rework the strict aliasing fix to not break C++, pointed
208 out by Murray Cumming.
210 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
211 Signal waiting threads, problem noticed by Christian Kellner.
213 2006-01-16 Matthias Clasen <mclasen@redhat.com>
215 Fix bug 326747, Alberto Ruiz:
217 * glib/gconvert.c (make_valid_utf8): Change this function to
218 replace unknown characters by the Unicode replacement character
219 instead of '?', and don't append "(invalid encoding)".
220 (g_filename_display_name, g_filename_display_basename): Document
221 how to determine if the filename was in an invalid encoding.
223 2006-01-14 Matthias Clasen <mclasen@redhat.com>
225 * glib/gtree.c: Replace the simple recursive implementation
226 by a nonrecursive, threaded implementation by Maurizio
229 2006-01-12 Matthias Clasen <mclasen@redhat.com>
231 * tests/asyncqueue-test.c (main): Fix the
232 assert to not always trigger. (#326558,
235 2006-01-11 Matthias Clasen <mclasen@redhat.com>
237 * tests/convert-test.c: Don't test items_read and
238 items_written in error cases where it is not set.
240 * tests/asyncqueue-test.c: Set error to NULL.
242 2006-01-08 Matthias Clasen <mclasen@redhat.com>
244 * glib/gtree.c: Break some long lines.
246 * tests/tree-test.c: Add more tests.
248 * glib/gslice.c: Include config.h first, pointed out
251 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
253 * docs/reference/gobject/tmpl/gtype.sgml: explain that
254 G_TYPE_INSTANCE_GET_CLASS() does behave different during
257 2006-01-06 Matthias Clasen <mclasen@redhat.com>
259 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
260 that we have posix_memalign(). (#323937, Bogdan Nicula)
262 2006-01-05 Matthias Clasen <mclasen@redhat.com>
264 * configure.in: Bump version
266 * === Released 2.9.2 ===
268 2006-01-05 Hans Breuer <hans@breuer.org>
270 * glib/makefile.msc.in : added gslice.obj
272 2006-01-05 Matthias Clasen <mclasen@redhat.com>
276 * glib/gmain.c: Clarify the documentation of source ids
277 a bit. (#325874, Dan Williams)
279 * configure.in: Fix another strict aliasing problem.
281 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
284 2006-01-04 Manish Singh <yosh@gimp.org>
287 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
288 either, since they have the same issue as g_rmdir. (Related to
291 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
293 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
294 order when creating magazines, so we hand out chunks with ascending
297 2006-01-04 Abel Cheung <maddog@linuxhall.org>
299 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
301 2006-01-03 Martyn Russell <martyn@imendio.com>
303 * docs/reference/glib/glib-sections.txt:
305 * glib/gthreadpool.[ch]:
306 - Added new API g_thread_pool_get_idle_time() and
307 g_thread_pool_set_idle_time(). (#324228).
309 * tests/threadpool-test.c:
310 - Updated test case to do thread pool sorting, thread pool with
311 no sorting and a thread pool with idle thread timeouts.
313 2006-01-03 Matthias Clasen <mclasen@redhat.com>
315 * glib/gmain.h: Add new functions here, too.
317 * glib/glib.symbols: Add new functions.
319 2005-12-20 Michael Meeks <michael.meeks@novell.com>
321 * glib/gmain.c (g_main_context_is_owner): new method
322 to determine if the current thread is the owner of the
325 2006-01-02 Matthias Clasen <mclasen@redhat.com>
329 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
330 expanding to rmdir, since rmdir is not declared in a portable
331 system header we can include in gstdio.h. (#325249, Jani Monoses)
333 2006-01-01 Tor Lillqvist <tml@novell.com>
335 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
336 stdout and stderr to unbuffered. Otherwise the giochannel layer
337 will try to read from them regardless whether the
338 g_io_channel_win32_poll() call here has indicated
339 readability or not. (#325310)
341 2005-12-29 Matthias Clasen <mclasen@redhat.com>
343 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
344 in the docs. (#325273, Declan Naughton)
346 2005-12-27 Manish Singh <yosh@gimp.org>
348 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
349 make sure we really override things.
351 2005-12-27 Matthias Clasen <mclasen@redhat.com>
353 Fix #316221, Michal Benes, Stanislav Brabec;
355 * configure.in: Fix a strict aliasing problem in
356 g_static_mutex_get_mutex().
357 * glib/gthread.h: ...and in
358 g_static_mutex_get_mutex_impl_shortcut().
360 * glib/gdatasetprivate.h: Add a cast to silence compiler
361 warnings. (#321978, Andrew Paprocki)
363 Partial fix for bug #323937, Bogdan Nicula.
365 * configure.in: Check for malloc.h
366 * glib/gslice.c: Don't include malloc.h unconditionally.
368 2005-12-27 Manish Singh <yosh@gimp.org>
370 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
371 (#324950, Dan Yefimov)
373 2005-12-26 Matthias Clasen <mclasen@redhat.com>
375 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
378 2005-12-25 Matthias Clasen <mclasen@redhat.com>
380 * AUTHORS: Update my email
382 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
384 2005-12-21 Matthias Clasen <mclasen@redhat.com>
386 * README.in: Add some notes on when 'make check' may fail.
388 2005-12-20 Matthias Clasen <mclasen@redhat.com>
391 * glib/gthreadpool.h:
392 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
393 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
395 * tests/threadpool-test.c: Test this.
397 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
399 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
400 handle errors without depending on gmessage.c which might not be
401 setup when the error occours.
402 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
403 always freeing can be achieved by adjusting the working set time to
404 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
405 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
406 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
407 minimize wastage if small amounts of differently sized structrues are
408 allocated, this does come at a performance cost of roughly 5% though.
409 fixed up block alignment calculation, so it works for varying
410 block sizes. only use strerror() not g_strerror() since the latter
411 depends on working GQuark and GSlice.
412 mem_error(): implemented in terms of fprintf and vfprintf.
414 * tests/slice-color.c: new program to test cache colorization effects.
416 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
417 G_SLICE_CONFIG_WORKING_SET_MSECS.
419 2005-12-17 Matthias Clasen <mclasen@redhat.com>
421 * glib/goption.c (parse_short_option): Set an error in all
422 failure cases. (#324332, Tim-Philipp Müller)
424 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
426 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
427 and g_atomic_pointer_get. Add g_atomic_int_set and
428 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
429 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
430 as functions (additionally to the macros in the header) for the
431 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
433 2005-12-16 Matthias Clasen <mclasen@redhat.com>
435 * glib/gmem.c (g_allocator_new): Don't return a pointer to
436 a const struct, since apps expect to be able to modify it.
437 (#324179, J. Ali Harlow)
439 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
441 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
442 for systems where the initialization of atomic variables requires a
443 write memory barrier.
445 2005-12-09 Matthias Clasen <mclasen@redhat.com>
447 * configure.in: Bump version
449 * === Released 2.9.1 ===
451 2005-12-09 Alexander Larsson <alexl@redhat.com>
455 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
458 2005-12-08 Matthias Clasen <mclasen@redhat.com>
462 2005-12-07 Martyn Russell <martyn@imendio.com>
464 * glib/gasyncqueue.c:
465 - Call g_queue_insert_sorted() instead of duplicating the code.
466 - Call g_queue_sort() instead of duplicating the code.
467 - Invert sort function results to make sure the same sort function
468 gives the same results across glist, gslist, gqueue and
471 * tests/asyncqueue-test.c:
472 - Updated the sort function to reflect the example in the
473 documentation for gasyncqueue.c.
475 2005-12-07 Martyn Russell <martyn@imendio.com>
477 * docs/reference/glib/glib-sections.txt:
478 * docs/reference/glib/tmpl/linked_lists_double.sgml:
479 * docs/reference/glib/tmpl/linked_lists_single.sgml:
482 - Added g_list_insert_sorted_with_data () and
483 g_slist_insert_sorted_with_data ().
484 - Removed the extra check in g_list_sort() and g_slist_sort() for
485 GCompareDataFunc vs. GCompareFunc.
487 2005-12-07 Tor Lillqvist <tml@novell.com>
489 * glib-gettextize.in: Look up prefix at run-time on Win32,
490 assuming the standard directory structure with glib-gettextize in
493 2005-12-06 Matthias Clasen <mclasen@redhat.com>
495 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
496 we need GTK+ 2.8 to compile against GLib 2.10.
498 2005-12-05 Manish Singh <yosh@gimp.org>
500 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
501 to override any user settings.
503 2005-12-05 Matthias Clasen <mclasen@redhat.com>
505 * configure.in: Revert an accidential commit
507 2005-12-05 Matthias Clasen <mclasen@redhat.com>
509 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
511 * tests/Makefile.am (TESTS_ENVIRONMENT):
512 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
513 MALLOC_CHECK_ and MALLOC_PERTURB_
515 * tests/run-collate-tests.sh: Run the collation tests explicitly
516 in en_US locale. (#320463)
518 * glib/gmem.h: Really deprecate GMemChunk.
525 * glib/gstrfuncs.c: Add versioned deprecation docs.
527 2005-12-05 Martyn Russell <martyn@imendio.com>
529 * docs/reference/glib/glib-sections.txt:
530 * glib/gasyncqueue.[ch]:
531 - Added support for sorting async queues by with _push_sorted(),
532 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
535 * tests/asyncqueue-test.c:
536 - Added test case for gasyncqueue.c
538 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
540 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
541 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
542 in bug 323178. moved time consuming logic from g_slice_free() out of
543 the inner loop, so g_slice_free_chain_with_offset() provides a real
544 performance benefit over g_slice_free1() now.
546 * glib/gslice.h: renamed g_slice_free_chain() to
547 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
548 a type-safe macro as suggested in bug 323178.
549 simplified the macro implementation of g_slice_free() and implemented
550 it in a type safe manner for all compliers as suggested by Morten
551 Welinder <mortenw@gnome.org>.
556 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
558 2005-12-05 Matthias Clasen <mclasen@redhat.com>
560 * glib/gasyncqueue.c: Add some docs.
562 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
564 * glib/gatomic.c: In the ia64 implementation, use
565 __sync builtin without _si or _di suffix. (#321229,
566 Stanislav Brabec, patch by Andreas Schwab)
568 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
570 * glib/gslice.h: Remove comma at the end of enum.
572 2005-12-04 Matthias Clasen <mclasen@redhat.com>
574 Handle multiple user names with the same UID better.
575 (#319535, Laszlo Peter)
577 * glib/gutils.c (g_get_any_init_do): When determining user
578 data, first look up $LOGNAME. If the UID doesn't match
579 getuid(), fall back to the current behaviour of looking
580 up the user data based on getuid().
582 2005-12-04 Matthias Clasen <mclasen@redhat.com>
584 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
585 to make gcc warn if a function result is ignored. (#145466,
586 Arjan van de Ven, Alex Larsson)
588 * glib/gmem.h: Add the new attribute to g_realloc and
591 2005-12-03 Matthias Clasen <mclasen@redhat.com>
594 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
597 * glib/gmappedfile.c: Make mapping of empty files
600 * glib/gfileutils.c: Don't fork a new process just to
601 fix the permissions of the created temp file. (#321318,
602 Alexis S. L. Carvalho)
604 2005-12-02 Matthias Clasen <mclasen@redhat.com>
606 * README.in: Add a note about Solaris threads.
609 * configure.in: Remove support for Solaris threads.
610 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
612 2005-12-02 Matthias Clasen <mclasen@redhat.com>
614 * configure.in: Define G_GUINT64_CONSTANT in analogy to
615 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
617 2005-12-02 Matthias Clasen <mclasen@redhat.com>
619 * glib/gslice.c: Win32 portability fixes and C99ism removal,
620 pointed out by Kazuki Iwamoto. (#323052)
622 2005-12-02 Matthias Clasen <mclasen@redhat.com>
624 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
625 the IA64 implementation.
627 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
629 * glib/gslice.c: some naming and type size fixups.
631 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
633 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
634 variant into something less verbose, but digestible for gcc-3.4.
636 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
638 * tests/slice-test.c: extended to perform the benchmarking on the old
639 memchunk code if 'O' is selected.
641 * tests/memchunks.c: new file which contains the old GLib mem chunks
642 implementation with prefix old_mem_chunk_.
644 * tests/Makefile.am: added memchunks.c
646 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
648 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
649 acceptable load distribution of the syscall.
650 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
651 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
653 2005-12-01 Matthias Clasen <mclasen@redhat.com>
655 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
656 double-free base interface members.
658 * tests/Makefile.am: Remove duplicate variable.
660 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
662 * glib/gslice.[hc]: new slice allocator implementation.
664 * tests/slice-test.c: added random slice allocation test.
666 * glib/gthread.[hc]: removed newly added private thread mem API.
668 * glib/gthreadinit.h:
671 * glib/gmem.c: divided glib threading initialisation into three phases,
672 initialisation where private keys and messaging are not available (only
673 needed by gmem.c), initialisation without messaging but private keys
674 available (gslice.c, gmessage.c), and full fledged initialisers that
675 server the rest of glib. initialisation functions got renamed to reflect
676 the limitations of their corresponding phases.
678 * glib/gmem.c: removed memchunk code, defer allocations to
681 * glib/gmem.[hc]: removed g_slice_* skeletons.
683 * glib/glib.symbols: added g_slice_* symbols.
685 * configure.in: check for availability of posix_memalign(3), memalign(3)
688 * glib/Makefile.am: added gslice.[hc].
690 2005-12-01 Tor Lillqvist <tml@novell.com>
692 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
693 trailing slash(es) for non-root folders. stat() fails if non-root
694 folders are specified with trailing slashes. It's too much hassle
695 to demand that callers strip such slashes themselves, especially
696 as it is easy to get it wrong and strip the slash of a root
698 (g_rename): On NT-based Windows, use MoveFileEx() with
699 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
701 2005-11-28 Matthias Clasen <mclasen@redhat.com>
703 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
706 * configure.in: Check whether do { } while (0) works.
708 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
709 G_STMT_END if it works.
711 2005-11-28 Matthias Clasen <mclasen@redhat.com>
713 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
714 * glib/gnode.c (g_node_depth_traverse_level):
715 * glib/gmem.c (g_allocator_new):
716 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
717 (g_hash_table_foreach_remove_or_steal): Silence compiler
720 2005-11-27 Matthias Clasen <mclasen@redhat.com>
722 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
723 Don't read beyond len. (#322520, Christian Persch)
725 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
727 * glib/gdataset.c: access datalist flags via atomic pointer access
728 functions, instead of acquiring the dataset lock. this is faster and
729 also matches the atomic pointer readouts in gdatalistprivate.h and
730 direct pointer modifications required by gobject.c.
732 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
734 * glib/gdataset.c: streamlined initialization code somewhat,
735 removed GData node cache.
737 2005-11-22 Matthias Clasen <mclasen@redhat.com>
739 * glib/ghash.c (g_hash_table_ref):
740 (g_hash_table_unref): Mark these as new API.
742 Avoid double locking in g_intern_string (#322133,
745 * glib/gdataset.c (g_quark_from_string_internal): New
746 internal function which factors out common parts of
747 g_quark_from[_static]_string.
748 (g_quark_from_string, g_quark_from_static_string):
749 Use g_quark_from_string_internal.
750 (g_intern_string, g_intern_static_string): Use
751 g_quark_from_string_internal, and only take the
753 (g_quark_new): Don't store the strings shifted by -1
754 in the g_quarks array.
755 (g_quark_to_string): Adapt to the previous change.
757 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
761 g_hash_table_new_full(): create hash tables with a ref count of 1.
762 g_hash_table_ref(): atomically ref_count+=1
763 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
764 when refcount reaches 0.
765 g_hash_table_destroy(): just destroy keys and values, unref by 1.
766 g_hash_table_insert():
767 g_hash_table_replace(): assert ref_count>0.
770 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
773 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
775 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
776 variables. (#321984, Andrew Paprocki)
778 2005-11-18 Matthias Clasen <mclasen@redhat.com>
780 * configure.in: Bump version
782 2005-11-17 Matthias Clasen <mclasen@redhat.com>
784 * === Released 2.9.0 ===
786 * glib/glib.symbols: Updates
788 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
789 hints to the docs. (Paolo Borelli)
793 * glib/gthread.c (g_thread_foreach): Mark as new api.
795 * README.in: Updates.
799 * glib/gdate.c (g_date_set_time_t):
800 (g_date_set_time_val): Add functions to set a date from
801 a time_t and from a GTimeVal.
802 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
805 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
806 and g_utf8_offset_to_pointer.
808 * glib/gutf8.c (g_utf8_pointer_to_offset)
809 (g_utf8_offset_to_pointer): Handle negative offsets, and use
810 "stutter stepping" for going backwards. (#320638, Larry
815 * glib/gthread.c: const correctness fixes, found
816 by Arjan van de Ven and gcc.
818 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
820 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
821 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
823 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
825 * glib/guniprop.c: Use bit hacks instead when checking a general
826 category value against multiple values.
828 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
830 2005-11-10 Simos Xenitellis <simos@gnome.org>
832 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
834 2005-11-05 Tor Lillqvist <tml@novell.com>
836 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
837 (#320688) Instead set error and return failure.
839 2005-11-07 Matthias Clasen <mclasen@redhat.com>
841 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
842 while holding the context lock. (#320886, Andy Wingo)
844 2005-11-07 Billy Biggs <vektor@dumbterm.net>
846 * tests/refcount/Makefile.am: Add a missing reference to libglib
847 in the LDADD for this test.
849 2005-11-07 Matthias Clasen <mclasen@redhat.com>
851 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
852 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
853 GLib 2.9. (#320755, Luca Ferretti)
855 2005-11-04 Matthias Clasen <mclasen@redhat.com>
857 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
858 use a faster lookup table for g_unichar_get_mirror_char().
859 * glib/gmirroringtable.h: The new table.
860 * glib/gunichartables.h: Remove bidi_mirroring_table.
861 * glib/gen-unicode-tables.pl: Don't generate the mirroring
863 * glib/glib-mirroring-tab/*: Sources for the program
864 which generated gmirroringtable.h.
866 * glib/glist.c: Avoid some code duplication.
868 * glib/gscanner.h: Include gdataset.h. (#320322)
870 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
871 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
872 to help find critical warnings in applications. (#320017,
875 2005-11-02 Tor Lillqvist <tml@novell.com>
877 * glib/glib.symbols: Remove large amount of trailing whitespace
878 from one line. Remove #ifdef G_UNDEFINED from around
879 g_hash_table_new and g_hash_table_new_full symbols, what was the
880 purpose of that? Add the g_slice functions.
882 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
884 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
885 added g_slice_*() API to allocate and cache small bits of memory.
886 an actuall allocator implementation for g_slice_*() is still pending.
888 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
889 changed GRealThread list to use in-structure *next; fields instead
890 of GSList, in order for thread iteration to not depenend on g_slice_*()
892 _g_thread_mem_private_get():
893 _g_thread_mem_private_set(): added accessors for private memory,
894 needed because the ordinary GPrivate implementation relies on GArray
895 and GSList and therefore indirectly on working g_slice_*() allocations.
898 g_thread_foreach(): new public API function to loop over all existing threads.
912 removed GAllocator and free list usages and accompanying locks.
913 use g_slice_*() API to allocate and cache small bits of memory.
915 * glib/ghook.h: removed GMemChunk field from public API.
918 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
920 * glib/gnode.h: deprecate allocator API.
922 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
923 be used for poll rec lists.
925 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
926 g_relation_destroy(): free all tuples from the all_tuples hash table,
927 this effectively maintains the life time track keeping of tuples.
928 g_relation_delete_tuple(): free tuples which are removed from the
929 all_tuples hash table. this fixes a temporary leak that was present
930 in the memchunk code until the destruction of the relation.
932 2005-10-29 Matthias Clasen <mclasen@redhat.com>
934 * tests/convert-test.c: Add some tests for conversions between
935 UTF-8, UCS-4 and UTF-16.
937 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
938 of len == -1, noticed by Morten Welinder.
940 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
942 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
944 2005-10-26 Matthias Clasen <mclasen@redhat.com>
946 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
947 in the error case, and add some documentation. (#319806, Morten
950 2005-10-19 Manish Singh <yosh@gimp.org>
952 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
953 config.h symbol automatically. Fixes bug #313731.
955 2005-10-19 Tor Lillqvist <tml@novell.com>
957 * glib/gwin32.c (g_win32_get_package_installation_directory):
958 Return a g_strdup()ed copy of the value stored in the hash table,
959 so that it can be g_free()d without leaving a dangling pointer in
960 the hash table. (#319232)
962 2005-10-06 Matthias Clasen <mclasen@redhat.com>
964 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
965 all-zero sequences correctly. (#317930, Sebastien Bacher)
967 2005-10-05 Matthias Clasen <mclasen@redhat.com>
969 * glib/gmarkup.c (g_markup_escape_text): Doc addition
971 * Makefile.am: Add Collation tests.
973 * tests/collate/*: Inputs and expected outputs for collation tests.
975 * tests/run-collate-tests.sh: Script to run collation tests.
977 * tests/unicode-collate.c (main): Rework slightly to make
978 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
980 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
982 * docs/reference/glib/tmpl/unicode.sgml:
983 * glib/gen-unicode-tables.pl:
985 * glib/gunichartables.h:
987 * tests/casefold.txt:
988 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
989 GUnicodeBreakType types. That may break some applications, like
992 2005-09-26 Matthias Clasen <mclasen@redhat.com>
994 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
995 docs. (#314393, Matthew F. Barnes)
998 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
999 they are already declared in glib.h. This doesn't break documented
1000 use of gprintf.h, but should probably be pointed out in the
1001 release notes for 2.10. (#314232, Behdad Esfahbod)
1003 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1005 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1006 complexity optimization, based on a patch from Matthias Clasen.
1008 * tests/patterntest.c: more tests, mostly from matthias.
1010 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1012 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1013 (#316703, Mark Drago)
1015 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1017 * glib/gpattern.c: applied a patch from matthias which checks on the
1018 upper bound of GPatternSpec length to optimize matches.
1021 * tests/patterntest.c: added more match cases.
1023 2005-09-16 Tor Lillqvist <tml@novell.com>
1025 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1026 warning to doc comment that these functions should not be used on
1027 encodings like CP932.
1029 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1031 * tests/keyfile-test.c: Add a test for grup names of length 1.
1033 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1034 of length 1. (#316309)
1036 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1038 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1040 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1042 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1043 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1046 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1048 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1049 (child_watch_helper_thread): Remove some dead code. Closes
1052 2005-09-07 Tor Lillqvist <tml@novell.com>
1054 * glib/Makefile.am: Create also a console version of the
1055 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1056 It's otherwise identical to gspawn-win32-helper.exe, except marked
1057 as a console application (linked without the -mwindows option).
1059 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1060 Drop the dont_wait parameter. Its truth value correlated 100% with
1061 the NULLness of the exit_status parameter anyway, so it's enough
1062 to check whether exit_status is NULL. Invert the sense of the
1063 dont_return_handle parameter and rename it to do_return_handle, to
1064 make the code easier to read by avoiding double negations.
1066 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1067 to do_spawn_with_pipes() accordingly.
1069 (do_spawn_with_pipes): If we have a console, use the console
1070 version of the helper program, otherwise use the GUI one. This
1071 avoids extra console windows opening up in some situations. (In
1072 case a console application uses the GUI gspawn-win32-helper.exe to
1073 spawn another console application we would get a separate console
1074 for the spawned console application).
1076 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1078 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1080 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1081 if the file is too large. (#315275, Kjartan Maraas)
1083 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1084 of read() is signed. (#315273, Kjartan Maraas)
1086 2005-08-31 Tor Lillqvist <tml@novell.com>
1088 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1089 the EXE and each DLL have their own atexit function chains.
1091 #define g_atexit as atexit instead. This means it has a
1092 better chance of doing what the caller wants. For instance,
1093 gtkhtml calls g_atexit() registering a function in gtkhtml
1094 itself. This caused a crash when g_atexit() was implemented as a
1095 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1096 the time the GLib DLL got unloaded.
1098 * glib/gutils.c: #undef the #define mentioned above, to also get a
1099 real g_atexit() into the DLL for backward compatibility. Document
1100 the Windows behaviour of g_atexit(), and document the varying ways
1101 atexit() can behave in the context of dynamically loaded modules
1104 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1106 * glib/glib.symbols:
1108 * glib/gdataset.c: Add string interning functions.
1110 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1112 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1115 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1117 * configure.in: Bump version to 2.9.0
1119 2005-08-25 Tor Lillqvist <tml@novell.com>
1121 Make also the g_spawn*() functions take parameters in the GLib
1122 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1123 API or ABI. Like the other GLib API that was earlier changed to
1124 use UTF-8 on Windows, the names of the functions that take UTF-8
1125 have _utf8 suffixes added by using preprocessor macros in the
1126 header file. The old names are kept for functions with the old
1127 behaviour, taking parameters in the system codepage, for DLL ABI
1130 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1131 g_spawn*() functions.
1133 * glib/gspawn-win32.c: Use wide-char API on NT-based
1134 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1135 system codepage (Win9x) and call the C library _wspawn*() or
1136 spawn*() functions respectvely. Add DLL ABI stability versions
1137 that take parameters in the system codepage.
1139 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1140 wide-char versions of argv and envp, and use wide-char API to
1141 change directory and spawn the program to run. Remove the verbose
1142 debugging output, it was too complex to modify for the wide-char
1143 features. (Just add temporary debugging printouts if needed, no
1144 need to have them permanently in the source.)
1146 * glib/gspawn.c: Corresponding documentation updates.
1148 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1149 symbols as PRIVATE, add the new _utf8-suffixed ones.
1151 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1153 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1154 guint64 otherwise and that can produce warnings about comparison
1155 between signed and unsigned.
1157 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1159 * glib/gutils.c: Fix the crt_externs.h include.
1161 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1163 * NEWS: Fix spelling of my first name.
1165 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1169 * === Released 2.8.1 ===
1173 2005-08-20 Hans Breuer <hans@breuer.org>
1175 * glib/makefile.msc.in : link with ws2_32.lib
1177 2005-08-18 Tor Lillqvist <tml@novell.com>
1179 * configure.in: Check for <sys/wait.h>
1181 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1183 2005-08-18 Ross Burton <ross@burtonini.com>
1186 Optimise single-character insertions.
1191 * tests/string-test.c:
1192 Add tests for new optimisation, and fix a leak.
1194 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1196 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1198 * glib/gutils.c: On Darwin, include crt-externs.h and
1199 define environ using _NSGetEnviron(). (#313731)
1201 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1203 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1204 (g_get_any_init_do): ... to this new function.
1205 (g_get_any_init): Declare as inline.
1206 (g_get_any_init_locked): New inline function, does the locking.
1207 Make use of these two throughout the code.
1209 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1211 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1212 the child process and then simply return. This makes
1213 The "S" option work as documented in g_on_error_query().
1214 (#313125, Matthew F. Barnes)
1216 * glib/gunicode.h: Update the link to Unicode category
1217 values. (#313369, Behnam Esfahbod)
1219 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1220 a little. (#311727, Tristan van Berkom)
1222 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1223 check work on ia64 too, where some symbols ended up
1224 in yet another section.