1 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
3 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
4 some missing cases where pointed out by Sebastian Wilhelmi.
6 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
8 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
10 2006-01-30 David Schleef <ds@schleef.org>
12 * glib/grand.c: Update URL
14 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
16 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
17 create container classes which are as flexible as a GValue is
19 2006-01-27 Matthias Clasen <mclasen@redhat.com>
21 * configure.in: Bump version
23 * === Released 2.9.5 ===
25 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
30 2006-01-26 Matthias Clasen <mclasen@redhat.com>
32 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
35 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
37 * fixed buglets reported by Jens Granseuer in #328254.
39 * configure.in: free the memory allocated in posix_memalign() tests.
41 * glib/gslice.c: spelling fixes.
43 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
45 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
46 slices, make sure g_mem_gc_friendly is properly initialized.
48 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
49 the first allocation. applied some branching optimizations.
51 * docs/macros.txt: reflected --enable-gc-friendly change and
52 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
54 * configure.in: changed --enable-gc-friendly=yes to define
55 ENABLE_GC_FRIENDLY_DEFAULT.
57 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
58 if (G_UNLIKELY (g_mem_gc_friendly)).
61 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
62 care of by g_slice_free1().
64 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
66 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
68 * glib/gslice.c: only use posix_memalign() if it's known to work,
69 revert to memalign() otherwise.
71 * configure.in: check for broken posix_memalign() implementations
74 2006-01-24 Matthias Clasen <mclasen@redhat.com>
76 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
77 on all little-endian systems. (#143380, Marc Moorcroft)
79 2006-01-23 Matthias Clasen <mclasen@redhat.com>
81 * configure.in: Change the shared libary extension for hpux-ia64
82 to so. (#328253, The Written Word)
84 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
86 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
87 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
90 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
92 * glib/gslice.c (slice_config_init): initialize GSlice config from
93 G_SLICE environment variable. we support G_SLICE=always-malloc
94 currently, which forces all g_slice_*() allocations to use the system
98 g_parse_debug_string(): added a note about not using g_malloc() here.
99 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
100 g_slice. contains only guesswork in the WIN32 branch.
102 2006-01-18 Matthias Clasen <mclasen@redhat.com>
106 * === Released 2.9.4 ===
110 2006-01-17 Matthias Clasen <mclasen@redhat.com>
112 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
113 call. (#327421, Morten Welinder)
115 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
117 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
118 G_LOCK. Use the unused_thread_queue lock instead. Change
119 g_thread_pool_thread_proxy such that threads only wait on
120 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
121 due to superfluous tasks. Global tasks wait at most for
122 max-idle-time milliseconds. Make sure, that no task is woken up
123 twice for the same event via a wakeup_serial. This fixes #324228.
125 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
126 invalid NULL into the thread pools. This as well fixes #327290.
128 2006-01-16 Matthias Clasen <mclasen@redhat.com>
130 * configure.in: Bump version
132 * === Released 2.9.3 ===
136 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
137 Rework the strict aliasing fix to not break C++, pointed
138 out by Murray Cumming.
140 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
141 Signal waiting threads, problem noticed by Christian Kellner.
143 2006-01-16 Matthias Clasen <mclasen@redhat.com>
145 Fix bug 326747, Alberto Ruiz:
147 * glib/gconvert.c (make_valid_utf8): Change this function to
148 replace unknown characters by the Unicode replacement character
149 instead of '?', and don't append "(invalid encoding)".
150 (g_filename_display_name, g_filename_display_basename): Document
151 how to determine if the filename was in an invalid encoding.
153 2006-01-14 Matthias Clasen <mclasen@redhat.com>
155 * glib/gtree.c: Replace the simple recursive implementation
156 by a nonrecursive, threaded implementation by Maurizio
159 2006-01-12 Matthias Clasen <mclasen@redhat.com>
161 * tests/asyncqueue-test.c (main): Fix the
162 assert to not always trigger. (#326558,
165 2006-01-11 Matthias Clasen <mclasen@redhat.com>
167 * tests/convert-test.c: Don't test items_read and
168 items_written in error cases where it is not set.
170 * tests/asyncqueue-test.c: Set error to NULL.
172 2006-01-08 Matthias Clasen <mclasen@redhat.com>
174 * glib/gtree.c: Break some long lines.
176 * tests/tree-test.c: Add more tests.
178 * glib/gslice.c: Include config.h first, pointed out
181 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
183 * docs/reference/gobject/tmpl/gtype.sgml: explain that
184 G_TYPE_INSTANCE_GET_CLASS() does behave different during
187 2006-01-06 Matthias Clasen <mclasen@redhat.com>
189 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
190 that we have posix_memalign(). (#323937, Bogdan Nicula)
192 2006-01-05 Matthias Clasen <mclasen@redhat.com>
194 * configure.in: Bump version
196 * === Released 2.9.2 ===
198 2006-01-05 Hans Breuer <hans@breuer.org>
200 * glib/makefile.msc.in : added gslice.obj
202 2006-01-05 Matthias Clasen <mclasen@redhat.com>
206 * glib/gmain.c: Clarify the documentation of source ids
207 a bit. (#325874, Dan Williams)
209 * configure.in: Fix another strict aliasing problem.
211 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
214 2006-01-04 Manish Singh <yosh@gimp.org>
217 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
218 either, since they have the same issue as g_rmdir. (Related to
221 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
223 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
224 order when creating magazines, so we hand out chunks with ascending
227 2006-01-04 Abel Cheung <maddog@linuxhall.org>
229 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
231 2006-01-03 Martyn Russell <martyn@imendio.com>
233 * docs/reference/glib/glib-sections.txt:
235 * glib/gthreadpool.[ch]:
236 - Added new API g_thread_pool_get_idle_time() and
237 g_thread_pool_set_idle_time(). (#324228).
239 * tests/threadpool-test.c:
240 - Updated test case to do thread pool sorting, thread pool with
241 no sorting and a thread pool with idle thread timeouts.
243 2006-01-03 Matthias Clasen <mclasen@redhat.com>
245 * glib/gmain.h: Add new functions here, too.
247 * glib/glib.symbols: Add new functions.
249 2005-12-20 Michael Meeks <michael.meeks@novell.com>
251 * glib/gmain.c (g_main_context_is_owner): new method
252 to determine if the current thread is the owner of the
255 2006-01-02 Matthias Clasen <mclasen@redhat.com>
259 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
260 expanding to rmdir, since rmdir is not declared in a portable
261 system header we can include in gstdio.h. (#325249, Jani Monoses)
263 2006-01-01 Tor Lillqvist <tml@novell.com>
265 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
266 stdout and stderr to unbuffered. Otherwise the giochannel layer
267 will try to read from them regardless whether the
268 g_io_channel_win32_poll() call here has indicated
269 readability or not. (#325310)
271 2005-12-29 Matthias Clasen <mclasen@redhat.com>
273 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
274 in the docs. (#325273, Declan Naughton)
276 2005-12-27 Manish Singh <yosh@gimp.org>
278 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
279 make sure we really override things.
281 2005-12-27 Matthias Clasen <mclasen@redhat.com>
283 Fix #316221, Michal Benes, Stanislav Brabec;
285 * configure.in: Fix a strict aliasing problem in
286 g_static_mutex_get_mutex().
287 * glib/gthread.h: ...and in
288 g_static_mutex_get_mutex_impl_shortcut().
290 * glib/gdatasetprivate.h: Add a cast to silence compiler
291 warnings. (#321978, Andrew Paprocki)
293 Partial fix for bug #323937, Bogdan Nicula.
295 * configure.in: Check for malloc.h
296 * glib/gslice.c: Don't include malloc.h unconditionally.
298 2005-12-27 Manish Singh <yosh@gimp.org>
300 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
301 (#324950, Dan Yefimov)
303 2005-12-26 Matthias Clasen <mclasen@redhat.com>
305 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
308 2005-12-25 Matthias Clasen <mclasen@redhat.com>
310 * AUTHORS: Update my email
312 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
314 2005-12-21 Matthias Clasen <mclasen@redhat.com>
316 * README.in: Add some notes on when 'make check' may fail.
318 2005-12-20 Matthias Clasen <mclasen@redhat.com>
321 * glib/gthreadpool.h:
322 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
323 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
325 * tests/threadpool-test.c: Test this.
327 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
329 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
330 handle errors without depending on gmessage.c which might not be
331 setup when the error occours.
332 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
333 always freeing can be achieved by adjusting the working set time to
334 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
335 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
336 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
337 minimize wastage if small amounts of differently sized structrues are
338 allocated, this does come at a performance cost of roughly 5% though.
339 fixed up block alignment calculation, so it works for varying
340 block sizes. only use strerror() not g_strerror() since the latter
341 depends on working GQuark and GSlice.
342 mem_error(): implemented in terms of fprintf and vfprintf.
344 * tests/slice-color.c: new program to test cache colorization effects.
346 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
347 G_SLICE_CONFIG_WORKING_SET_MSECS.
349 2005-12-17 Matthias Clasen <mclasen@redhat.com>
351 * glib/goption.c (parse_short_option): Set an error in all
352 failure cases. (#324332, Tim-Philipp Müller)
354 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
356 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
357 and g_atomic_pointer_get. Add g_atomic_int_set and
358 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
359 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
360 as functions (additionally to the macros in the header) for the
361 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
363 2005-12-16 Matthias Clasen <mclasen@redhat.com>
365 * glib/gmem.c (g_allocator_new): Don't return a pointer to
366 a const struct, since apps expect to be able to modify it.
367 (#324179, J. Ali Harlow)
369 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
371 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
372 for systems where the initialization of atomic variables requires a
373 write memory barrier.
375 2005-12-09 Matthias Clasen <mclasen@redhat.com>
377 * configure.in: Bump version
379 * === Released 2.9.1 ===
381 2005-12-09 Alexander Larsson <alexl@redhat.com>
385 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
388 2005-12-08 Matthias Clasen <mclasen@redhat.com>
392 2005-12-07 Martyn Russell <martyn@imendio.com>
394 * glib/gasyncqueue.c:
395 - Call g_queue_insert_sorted() instead of duplicating the code.
396 - Call g_queue_sort() instead of duplicating the code.
397 - Invert sort function results to make sure the same sort function
398 gives the same results across glist, gslist, gqueue and
401 * tests/asyncqueue-test.c:
402 - Updated the sort function to reflect the example in the
403 documentation for gasyncqueue.c.
405 2005-12-07 Martyn Russell <martyn@imendio.com>
407 * docs/reference/glib/glib-sections.txt:
408 * docs/reference/glib/tmpl/linked_lists_double.sgml:
409 * docs/reference/glib/tmpl/linked_lists_single.sgml:
412 - Added g_list_insert_sorted_with_data () and
413 g_slist_insert_sorted_with_data ().
414 - Removed the extra check in g_list_sort() and g_slist_sort() for
415 GCompareDataFunc vs. GCompareFunc.
417 2005-12-07 Tor Lillqvist <tml@novell.com>
419 * glib-gettextize.in: Look up prefix at run-time on Win32,
420 assuming the standard directory structure with glib-gettextize in
423 2005-12-06 Matthias Clasen <mclasen@redhat.com>
425 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
426 we need GTK+ 2.8 to compile against GLib 2.10.
428 2005-12-05 Manish Singh <yosh@gimp.org>
430 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
431 to override any user settings.
433 2005-12-05 Matthias Clasen <mclasen@redhat.com>
435 * configure.in: Revert an accidential commit
437 2005-12-05 Matthias Clasen <mclasen@redhat.com>
439 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
441 * tests/Makefile.am (TESTS_ENVIRONMENT):
442 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
443 MALLOC_CHECK_ and MALLOC_PERTURB_
445 * tests/run-collate-tests.sh: Run the collation tests explicitly
446 in en_US locale. (#320463)
448 * glib/gmem.h: Really deprecate GMemChunk.
455 * glib/gstrfuncs.c: Add versioned deprecation docs.
457 2005-12-05 Martyn Russell <martyn@imendio.com>
459 * docs/reference/glib/glib-sections.txt:
460 * glib/gasyncqueue.[ch]:
461 - Added support for sorting async queues by with _push_sorted(),
462 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
465 * tests/asyncqueue-test.c:
466 - Added test case for gasyncqueue.c
468 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
470 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
471 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
472 in bug 323178. moved time consuming logic from g_slice_free() out of
473 the inner loop, so g_slice_free_chain_with_offset() provides a real
474 performance benefit over g_slice_free1() now.
476 * glib/gslice.h: renamed g_slice_free_chain() to
477 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
478 a type-safe macro as suggested in bug 323178.
479 simplified the macro implementation of g_slice_free() and implemented
480 it in a type safe manner for all compliers as suggested by Morten
481 Welinder <mortenw@gnome.org>.
486 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
488 2005-12-05 Matthias Clasen <mclasen@redhat.com>
490 * glib/gasyncqueue.c: Add some docs.
492 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
494 * glib/gatomic.c: In the ia64 implementation, use
495 __sync builtin without _si or _di suffix. (#321229,
496 Stanislav Brabec, patch by Andreas Schwab)
498 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
500 * glib/gslice.h: Remove comma at the end of enum.
502 2005-12-04 Matthias Clasen <mclasen@redhat.com>
504 Handle multiple user names with the same UID better.
505 (#319535, Laszlo Peter)
507 * glib/gutils.c (g_get_any_init_do): When determining user
508 data, first look up $LOGNAME. If the UID doesn't match
509 getuid(), fall back to the current behaviour of looking
510 up the user data based on getuid().
512 2005-12-04 Matthias Clasen <mclasen@redhat.com>
514 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
515 to make gcc warn if a function result is ignored. (#145466,
516 Arjan van de Ven, Alex Larsson)
518 * glib/gmem.h: Add the new attribute to g_realloc and
521 2005-12-03 Matthias Clasen <mclasen@redhat.com>
524 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
527 * glib/gmappedfile.c: Make mapping of empty files
530 * glib/gfileutils.c: Don't fork a new process just to
531 fix the permissions of the created temp file. (#321318,
532 Alexis S. L. Carvalho)
534 2005-12-02 Matthias Clasen <mclasen@redhat.com>
536 * README.in: Add a note about Solaris threads.
539 * configure.in: Remove support for Solaris threads.
540 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
542 2005-12-02 Matthias Clasen <mclasen@redhat.com>
544 * configure.in: Define G_GUINT64_CONSTANT in analogy to
545 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
547 2005-12-02 Matthias Clasen <mclasen@redhat.com>
549 * glib/gslice.c: Win32 portability fixes and C99ism removal,
550 pointed out by Kazuki Iwamoto. (#323052)
552 2005-12-02 Matthias Clasen <mclasen@redhat.com>
554 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
555 the IA64 implementation.
557 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
559 * glib/gslice.c: some naming and type size fixups.
561 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
563 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
564 variant into something less verbose, but digestible for gcc-3.4.
566 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
568 * tests/slice-test.c: extended to perform the benchmarking on the old
569 memchunk code if 'O' is selected.
571 * tests/memchunks.c: new file which contains the old GLib mem chunks
572 implementation with prefix old_mem_chunk_.
574 * tests/Makefile.am: added memchunks.c
576 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
578 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
579 acceptable load distribution of the syscall.
580 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
581 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
583 2005-12-01 Matthias Clasen <mclasen@redhat.com>
585 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
586 double-free base interface members.
588 * tests/Makefile.am: Remove duplicate variable.
590 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
592 * glib/gslice.[hc]: new slice allocator implementation.
594 * tests/slice-test.c: added random slice allocation test.
596 * glib/gthread.[hc]: removed newly added private thread mem API.
598 * glib/gthreadinit.h:
601 * glib/gmem.c: divided glib threading initialisation into three phases,
602 initialisation where private keys and messaging are not available (only
603 needed by gmem.c), initialisation without messaging but private keys
604 available (gslice.c, gmessage.c), and full fledged initialisers that
605 server the rest of glib. initialisation functions got renamed to reflect
606 the limitations of their corresponding phases.
608 * glib/gmem.c: removed memchunk code, defer allocations to
611 * glib/gmem.[hc]: removed g_slice_* skeletons.
613 * glib/glib.symbols: added g_slice_* symbols.
615 * configure.in: check for availability of posix_memalign(3), memalign(3)
618 * glib/Makefile.am: added gslice.[hc].
620 2005-12-01 Tor Lillqvist <tml@novell.com>
622 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
623 trailing slash(es) for non-root folders. stat() fails if non-root
624 folders are specified with trailing slashes. It's too much hassle
625 to demand that callers strip such slashes themselves, especially
626 as it is easy to get it wrong and strip the slash of a root
628 (g_rename): On NT-based Windows, use MoveFileEx() with
629 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
631 2005-11-28 Matthias Clasen <mclasen@redhat.com>
633 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
636 * configure.in: Check whether do { } while (0) works.
638 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
639 G_STMT_END if it works.
641 2005-11-28 Matthias Clasen <mclasen@redhat.com>
643 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
644 * glib/gnode.c (g_node_depth_traverse_level):
645 * glib/gmem.c (g_allocator_new):
646 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
647 (g_hash_table_foreach_remove_or_steal): Silence compiler
650 2005-11-27 Matthias Clasen <mclasen@redhat.com>
652 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
653 Don't read beyond len. (#322520, Christian Persch)
655 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
657 * glib/gdataset.c: access datalist flags via atomic pointer access
658 functions, instead of acquiring the dataset lock. this is faster and
659 also matches the atomic pointer readouts in gdatalistprivate.h and
660 direct pointer modifications required by gobject.c.
662 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
664 * glib/gdataset.c: streamlined initialization code somewhat,
665 removed GData node cache.
667 2005-11-22 Matthias Clasen <mclasen@redhat.com>
669 * glib/ghash.c (g_hash_table_ref):
670 (g_hash_table_unref): Mark these as new API.
672 Avoid double locking in g_intern_string (#322133,
675 * glib/gdataset.c (g_quark_from_string_internal): New
676 internal function which factors out common parts of
677 g_quark_from[_static]_string.
678 (g_quark_from_string, g_quark_from_static_string):
679 Use g_quark_from_string_internal.
680 (g_intern_string, g_intern_static_string): Use
681 g_quark_from_string_internal, and only take the
683 (g_quark_new): Don't store the strings shifted by -1
684 in the g_quarks array.
685 (g_quark_to_string): Adapt to the previous change.
687 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
691 g_hash_table_new_full(): create hash tables with a ref count of 1.
692 g_hash_table_ref(): atomically ref_count+=1
693 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
694 when refcount reaches 0.
695 g_hash_table_destroy(): just destroy keys and values, unref by 1.
696 g_hash_table_insert():
697 g_hash_table_replace(): assert ref_count>0.
700 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
703 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
705 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
706 variables. (#321984, Andrew Paprocki)
708 2005-11-18 Matthias Clasen <mclasen@redhat.com>
710 * configure.in: Bump version
712 2005-11-17 Matthias Clasen <mclasen@redhat.com>
714 * === Released 2.9.0 ===
716 * glib/glib.symbols: Updates
718 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
719 hints to the docs. (Paolo Borelli)
723 * glib/gthread.c (g_thread_foreach): Mark as new api.
725 * README.in: Updates.
729 * glib/gdate.c (g_date_set_time_t):
730 (g_date_set_time_val): Add functions to set a date from
731 a time_t and from a GTimeVal.
732 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
735 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
736 and g_utf8_offset_to_pointer.
738 * glib/gutf8.c (g_utf8_pointer_to_offset)
739 (g_utf8_offset_to_pointer): Handle negative offsets, and use
740 "stutter stepping" for going backwards. (#320638, Larry
745 * glib/gthread.c: const correctness fixes, found
746 by Arjan van de Ven and gcc.
748 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
750 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
751 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
753 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
755 * glib/guniprop.c: Use bit hacks instead when checking a general
756 category value against multiple values.
758 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
760 2005-11-10 Simos Xenitellis <simos@gnome.org>
762 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
764 2005-11-05 Tor Lillqvist <tml@novell.com>
766 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
767 (#320688) Instead set error and return failure.
769 2005-11-07 Matthias Clasen <mclasen@redhat.com>
771 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
772 while holding the context lock. (#320886, Andy Wingo)
774 2005-11-07 Billy Biggs <vektor@dumbterm.net>
776 * tests/refcount/Makefile.am: Add a missing reference to libglib
777 in the LDADD for this test.
779 2005-11-07 Matthias Clasen <mclasen@redhat.com>
781 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
782 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
783 GLib 2.9. (#320755, Luca Ferretti)
785 2005-11-04 Matthias Clasen <mclasen@redhat.com>
787 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
788 use a faster lookup table for g_unichar_get_mirror_char().
789 * glib/gmirroringtable.h: The new table.
790 * glib/gunichartables.h: Remove bidi_mirroring_table.
791 * glib/gen-unicode-tables.pl: Don't generate the mirroring
793 * glib/glib-mirroring-tab/*: Sources for the program
794 which generated gmirroringtable.h.
796 * glib/glist.c: Avoid some code duplication.
798 * glib/gscanner.h: Include gdataset.h. (#320322)
800 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
801 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
802 to help find critical warnings in applications. (#320017,
805 2005-11-02 Tor Lillqvist <tml@novell.com>
807 * glib/glib.symbols: Remove large amount of trailing whitespace
808 from one line. Remove #ifdef G_UNDEFINED from around
809 g_hash_table_new and g_hash_table_new_full symbols, what was the
810 purpose of that? Add the g_slice functions.
812 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
814 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
815 added g_slice_*() API to allocate and cache small bits of memory.
816 an actuall allocator implementation for g_slice_*() is still pending.
818 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
819 changed GRealThread list to use in-structure *next; fields instead
820 of GSList, in order for thread iteration to not depenend on g_slice_*()
822 _g_thread_mem_private_get():
823 _g_thread_mem_private_set(): added accessors for private memory,
824 needed because the ordinary GPrivate implementation relies on GArray
825 and GSList and therefore indirectly on working g_slice_*() allocations.
828 g_thread_foreach(): new public API function to loop over all existing threads.
842 removed GAllocator and free list usages and accompanying locks.
843 use g_slice_*() API to allocate and cache small bits of memory.
845 * glib/ghook.h: removed GMemChunk field from public API.
848 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
850 * glib/gnode.h: deprecate allocator API.
852 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
853 be used for poll rec lists.
855 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
856 g_relation_destroy(): free all tuples from the all_tuples hash table,
857 this effectively maintains the life time track keeping of tuples.
858 g_relation_delete_tuple(): free tuples which are removed from the
859 all_tuples hash table. this fixes a temporary leak that was present
860 in the memchunk code until the destruction of the relation.
862 2005-10-29 Matthias Clasen <mclasen@redhat.com>
864 * tests/convert-test.c: Add some tests for conversions between
865 UTF-8, UCS-4 and UTF-16.
867 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
868 of len == -1, noticed by Morten Welinder.
870 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
872 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
874 2005-10-26 Matthias Clasen <mclasen@redhat.com>
876 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
877 in the error case, and add some documentation. (#319806, Morten
880 2005-10-19 Manish Singh <yosh@gimp.org>
882 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
883 config.h symbol automatically. Fixes bug #313731.
885 2005-10-19 Tor Lillqvist <tml@novell.com>
887 * glib/gwin32.c (g_win32_get_package_installation_directory):
888 Return a g_strdup()ed copy of the value stored in the hash table,
889 so that it can be g_free()d without leaving a dangling pointer in
890 the hash table. (#319232)
892 2005-10-06 Matthias Clasen <mclasen@redhat.com>
894 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
895 all-zero sequences correctly. (#317930, Sebastien Bacher)
897 2005-10-05 Matthias Clasen <mclasen@redhat.com>
899 * glib/gmarkup.c (g_markup_escape_text): Doc addition
901 * Makefile.am: Add Collation tests.
903 * tests/collate/*: Inputs and expected outputs for collation tests.
905 * tests/run-collate-tests.sh: Script to run collation tests.
907 * tests/unicode-collate.c (main): Rework slightly to make
908 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
910 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
912 * docs/reference/glib/tmpl/unicode.sgml:
913 * glib/gen-unicode-tables.pl:
915 * glib/gunichartables.h:
917 * tests/casefold.txt:
918 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
919 GUnicodeBreakType types. That may break some applications, like
922 2005-09-26 Matthias Clasen <mclasen@redhat.com>
924 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
925 docs. (#314393, Matthew F. Barnes)
928 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
929 they are already declared in glib.h. This doesn't break documented
930 use of gprintf.h, but should probably be pointed out in the
931 release notes for 2.10. (#314232, Behdad Esfahbod)
933 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
935 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
936 complexity optimization, based on a patch from Matthias Clasen.
938 * tests/patterntest.c: more tests, mostly from matthias.
940 2005-09-20 Matthias Clasen <mclasen@redhat.com>
942 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
943 (#316703, Mark Drago)
945 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
947 * glib/gpattern.c: applied a patch from matthias which checks on the
948 upper bound of GPatternSpec length to optimize matches.
951 * tests/patterntest.c: added more match cases.
953 2005-09-16 Tor Lillqvist <tml@novell.com>
955 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
956 warning to doc comment that these functions should not be used on
957 encodings like CP932.
959 2005-09-14 Matthias Clasen <mclasen@redhat.com>
961 * tests/keyfile-test.c: Add a test for grup names of length 1.
963 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
964 of length 1. (#316309)
966 2005-09-12 Matthias Clasen <mclasen@redhat.com>
968 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
970 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
972 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
973 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
976 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
978 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
979 (child_watch_helper_thread): Remove some dead code. Closes
982 2005-09-07 Tor Lillqvist <tml@novell.com>
984 * glib/Makefile.am: Create also a console version of the
985 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
986 It's otherwise identical to gspawn-win32-helper.exe, except marked
987 as a console application (linked without the -mwindows option).
989 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
990 Drop the dont_wait parameter. Its truth value correlated 100% with
991 the NULLness of the exit_status parameter anyway, so it's enough
992 to check whether exit_status is NULL. Invert the sense of the
993 dont_return_handle parameter and rename it to do_return_handle, to
994 make the code easier to read by avoiding double negations.
996 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
997 to do_spawn_with_pipes() accordingly.
999 (do_spawn_with_pipes): If we have a console, use the console
1000 version of the helper program, otherwise use the GUI one. This
1001 avoids extra console windows opening up in some situations. (In
1002 case a console application uses the GUI gspawn-win32-helper.exe to
1003 spawn another console application we would get a separate console
1004 for the spawned console application).
1006 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1008 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1010 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1011 if the file is too large. (#315275, Kjartan Maraas)
1013 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1014 of read() is signed. (#315273, Kjartan Maraas)
1016 2005-08-31 Tor Lillqvist <tml@novell.com>
1018 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1019 the EXE and each DLL have their own atexit function chains.
1021 #define g_atexit as atexit instead. This means it has a
1022 better chance of doing what the caller wants. For instance,
1023 gtkhtml calls g_atexit() registering a function in gtkhtml
1024 itself. This caused a crash when g_atexit() was implemented as a
1025 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1026 the time the GLib DLL got unloaded.
1028 * glib/gutils.c: #undef the #define mentioned above, to also get a
1029 real g_atexit() into the DLL for backward compatibility. Document
1030 the Windows behaviour of g_atexit(), and document the varying ways
1031 atexit() can behave in the context of dynamically loaded modules
1034 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1036 * glib/glib.symbols:
1038 * glib/gdataset.c: Add string interning functions.
1040 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1042 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1045 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1047 * configure.in: Bump version to 2.9.0
1049 2005-08-25 Tor Lillqvist <tml@novell.com>
1051 Make also the g_spawn*() functions take parameters in the GLib
1052 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1053 API or ABI. Like the other GLib API that was earlier changed to
1054 use UTF-8 on Windows, the names of the functions that take UTF-8
1055 have _utf8 suffixes added by using preprocessor macros in the
1056 header file. The old names are kept for functions with the old
1057 behaviour, taking parameters in the system codepage, for DLL ABI
1060 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1061 g_spawn*() functions.
1063 * glib/gspawn-win32.c: Use wide-char API on NT-based
1064 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1065 system codepage (Win9x) and call the C library _wspawn*() or
1066 spawn*() functions respectvely. Add DLL ABI stability versions
1067 that take parameters in the system codepage.
1069 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1070 wide-char versions of argv and envp, and use wide-char API to
1071 change directory and spawn the program to run. Remove the verbose
1072 debugging output, it was too complex to modify for the wide-char
1073 features. (Just add temporary debugging printouts if needed, no
1074 need to have them permanently in the source.)
1076 * glib/gspawn.c: Corresponding documentation updates.
1078 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1079 symbols as PRIVATE, add the new _utf8-suffixed ones.
1081 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1083 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1084 guint64 otherwise and that can produce warnings about comparison
1085 between signed and unsigned.
1087 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1089 * glib/gutils.c: Fix the crt_externs.h include.
1091 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1093 * NEWS: Fix spelling of my first name.
1095 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1099 * === Released 2.8.1 ===
1103 2005-08-20 Hans Breuer <hans@breuer.org>
1105 * glib/makefile.msc.in : link with ws2_32.lib
1107 2005-08-18 Tor Lillqvist <tml@novell.com>
1109 * configure.in: Check for <sys/wait.h>
1111 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1113 2005-08-18 Ross Burton <ross@burtonini.com>
1116 Optimise single-character insertions.
1121 * tests/string-test.c:
1122 Add tests for new optimisation, and fix a leak.
1124 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1126 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1128 * glib/gutils.c: On Darwin, include crt-externs.h and
1129 define environ using _NSGetEnviron(). (#313731)
1131 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1133 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1134 (g_get_any_init_do): ... to this new function.
1135 (g_get_any_init): Declare as inline.
1136 (g_get_any_init_locked): New inline function, does the locking.
1137 Make use of these two throughout the code.
1139 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1141 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1142 the child process and then simply return. This makes
1143 The "S" option work as documented in g_on_error_query().
1144 (#313125, Matthew F. Barnes)
1146 * glib/gunicode.h: Update the link to Unicode category
1147 values. (#313369, Behnam Esfahbod)
1149 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1150 a little. (#311727, Tristan van Berkom)
1152 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1153 check work on ia64 too, where some symbols ended up
1154 in yet another section.