1 2006-02-11 Matthias Clasen <mclasen@redhat.com>
3 * configure.in: Bump version
5 2006-02-10 Matthias Clasen <mclasen@redhat.com>
7 * === Released 2.9.6 ===
11 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
13 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
14 the type of varargs arguments. (bug #317679)
16 2006-01-31 Matthias Clasen <mclasen@redhat.com>
18 * autogen.sh: Touch README and INSTALL here to pacify
19 automake. (#329124, Kjartan Maraas, Tim Janik)
21 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
23 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
24 some missing cases where pointed out by Sebastian Wilhelmi.
26 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
28 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
30 2006-01-30 David Schleef <ds@schleef.org>
32 * glib/grand.c: Update URL
34 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
36 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
37 create container classes which are as flexible as a GValue is
39 2006-01-27 Matthias Clasen <mclasen@redhat.com>
41 * configure.in: Bump version
43 * === Released 2.9.5 ===
45 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
50 2006-01-26 Matthias Clasen <mclasen@redhat.com>
52 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
55 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
57 * fixed buglets reported by Jens Granseuer in #328254.
59 * configure.in: free the memory allocated in posix_memalign() tests.
61 * glib/gslice.c: spelling fixes.
63 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
65 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
66 slices, make sure g_mem_gc_friendly is properly initialized.
68 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
69 the first allocation. applied some branching optimizations.
71 * docs/macros.txt: reflected --enable-gc-friendly change and
72 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
74 * configure.in: changed --enable-gc-friendly=yes to define
75 ENABLE_GC_FRIENDLY_DEFAULT.
77 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
78 if (G_UNLIKELY (g_mem_gc_friendly)).
81 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
82 care of by g_slice_free1().
84 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
86 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
88 * glib/gslice.c: only use posix_memalign() if it's known to work,
89 revert to memalign() otherwise.
91 * configure.in: check for broken posix_memalign() implementations
94 2006-01-24 Matthias Clasen <mclasen@redhat.com>
96 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
97 on all little-endian systems. (#143380, Marc Moorcroft)
99 2006-01-23 Matthias Clasen <mclasen@redhat.com>
101 * configure.in: Change the shared libary extension for hpux-ia64
102 to so. (#328253, The Written Word)
104 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
106 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
107 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
110 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
112 * glib/gslice.c (slice_config_init): initialize GSlice config from
113 G_SLICE environment variable. we support G_SLICE=always-malloc
114 currently, which forces all g_slice_*() allocations to use the system
118 g_parse_debug_string(): added a note about not using g_malloc() here.
119 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
120 g_slice. contains only guesswork in the WIN32 branch.
122 2006-01-18 Matthias Clasen <mclasen@redhat.com>
126 * === Released 2.9.4 ===
130 2006-01-17 Matthias Clasen <mclasen@redhat.com>
132 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
133 call. (#327421, Morten Welinder)
135 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
137 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
138 G_LOCK. Use the unused_thread_queue lock instead. Change
139 g_thread_pool_thread_proxy such that threads only wait on
140 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
141 due to superfluous tasks. Global tasks wait at most for
142 max-idle-time milliseconds. Make sure, that no task is woken up
143 twice for the same event via a wakeup_serial. This fixes #324228.
145 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
146 invalid NULL into the thread pools. This as well fixes #327290.
148 2006-01-16 Matthias Clasen <mclasen@redhat.com>
150 * configure.in: Bump version
152 * === Released 2.9.3 ===
156 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
157 Rework the strict aliasing fix to not break C++, pointed
158 out by Murray Cumming.
160 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
161 Signal waiting threads, problem noticed by Christian Kellner.
163 2006-01-16 Matthias Clasen <mclasen@redhat.com>
165 Fix bug 326747, Alberto Ruiz:
167 * glib/gconvert.c (make_valid_utf8): Change this function to
168 replace unknown characters by the Unicode replacement character
169 instead of '?', and don't append "(invalid encoding)".
170 (g_filename_display_name, g_filename_display_basename): Document
171 how to determine if the filename was in an invalid encoding.
173 2006-01-14 Matthias Clasen <mclasen@redhat.com>
175 * glib/gtree.c: Replace the simple recursive implementation
176 by a nonrecursive, threaded implementation by Maurizio
179 2006-01-12 Matthias Clasen <mclasen@redhat.com>
181 * tests/asyncqueue-test.c (main): Fix the
182 assert to not always trigger. (#326558,
185 2006-01-11 Matthias Clasen <mclasen@redhat.com>
187 * tests/convert-test.c: Don't test items_read and
188 items_written in error cases where it is not set.
190 * tests/asyncqueue-test.c: Set error to NULL.
192 2006-01-08 Matthias Clasen <mclasen@redhat.com>
194 * glib/gtree.c: Break some long lines.
196 * tests/tree-test.c: Add more tests.
198 * glib/gslice.c: Include config.h first, pointed out
201 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
203 * docs/reference/gobject/tmpl/gtype.sgml: explain that
204 G_TYPE_INSTANCE_GET_CLASS() does behave different during
207 2006-01-06 Matthias Clasen <mclasen@redhat.com>
209 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
210 that we have posix_memalign(). (#323937, Bogdan Nicula)
212 2006-01-05 Matthias Clasen <mclasen@redhat.com>
214 * configure.in: Bump version
216 * === Released 2.9.2 ===
218 2006-01-05 Hans Breuer <hans@breuer.org>
220 * glib/makefile.msc.in : added gslice.obj
222 2006-01-05 Matthias Clasen <mclasen@redhat.com>
226 * glib/gmain.c: Clarify the documentation of source ids
227 a bit. (#325874, Dan Williams)
229 * configure.in: Fix another strict aliasing problem.
231 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
234 2006-01-04 Manish Singh <yosh@gimp.org>
237 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
238 either, since they have the same issue as g_rmdir. (Related to
241 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
243 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
244 order when creating magazines, so we hand out chunks with ascending
247 2006-01-04 Abel Cheung <maddog@linuxhall.org>
249 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
251 2006-01-03 Martyn Russell <martyn@imendio.com>
253 * docs/reference/glib/glib-sections.txt:
255 * glib/gthreadpool.[ch]:
256 - Added new API g_thread_pool_get_idle_time() and
257 g_thread_pool_set_idle_time(). (#324228).
259 * tests/threadpool-test.c:
260 - Updated test case to do thread pool sorting, thread pool with
261 no sorting and a thread pool with idle thread timeouts.
263 2006-01-03 Matthias Clasen <mclasen@redhat.com>
265 * glib/gmain.h: Add new functions here, too.
267 * glib/glib.symbols: Add new functions.
269 2005-12-20 Michael Meeks <michael.meeks@novell.com>
271 * glib/gmain.c (g_main_context_is_owner): new method
272 to determine if the current thread is the owner of the
275 2006-01-02 Matthias Clasen <mclasen@redhat.com>
279 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
280 expanding to rmdir, since rmdir is not declared in a portable
281 system header we can include in gstdio.h. (#325249, Jani Monoses)
283 2006-01-01 Tor Lillqvist <tml@novell.com>
285 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
286 stdout and stderr to unbuffered. Otherwise the giochannel layer
287 will try to read from them regardless whether the
288 g_io_channel_win32_poll() call here has indicated
289 readability or not. (#325310)
291 2005-12-29 Matthias Clasen <mclasen@redhat.com>
293 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
294 in the docs. (#325273, Declan Naughton)
296 2005-12-27 Manish Singh <yosh@gimp.org>
298 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
299 make sure we really override things.
301 2005-12-27 Matthias Clasen <mclasen@redhat.com>
303 Fix #316221, Michal Benes, Stanislav Brabec;
305 * configure.in: Fix a strict aliasing problem in
306 g_static_mutex_get_mutex().
307 * glib/gthread.h: ...and in
308 g_static_mutex_get_mutex_impl_shortcut().
310 * glib/gdatasetprivate.h: Add a cast to silence compiler
311 warnings. (#321978, Andrew Paprocki)
313 Partial fix for bug #323937, Bogdan Nicula.
315 * configure.in: Check for malloc.h
316 * glib/gslice.c: Don't include malloc.h unconditionally.
318 2005-12-27 Manish Singh <yosh@gimp.org>
320 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
321 (#324950, Dan Yefimov)
323 2005-12-26 Matthias Clasen <mclasen@redhat.com>
325 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
328 2005-12-25 Matthias Clasen <mclasen@redhat.com>
330 * AUTHORS: Update my email
332 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
334 2005-12-21 Matthias Clasen <mclasen@redhat.com>
336 * README.in: Add some notes on when 'make check' may fail.
338 2005-12-20 Matthias Clasen <mclasen@redhat.com>
341 * glib/gthreadpool.h:
342 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
343 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
345 * tests/threadpool-test.c: Test this.
347 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
349 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
350 handle errors without depending on gmessage.c which might not be
351 setup when the error occours.
352 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
353 always freeing can be achieved by adjusting the working set time to
354 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
355 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
356 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
357 minimize wastage if small amounts of differently sized structrues are
358 allocated, this does come at a performance cost of roughly 5% though.
359 fixed up block alignment calculation, so it works for varying
360 block sizes. only use strerror() not g_strerror() since the latter
361 depends on working GQuark and GSlice.
362 mem_error(): implemented in terms of fprintf and vfprintf.
364 * tests/slice-color.c: new program to test cache colorization effects.
366 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
367 G_SLICE_CONFIG_WORKING_SET_MSECS.
369 2005-12-17 Matthias Clasen <mclasen@redhat.com>
371 * glib/goption.c (parse_short_option): Set an error in all
372 failure cases. (#324332, Tim-Philipp Müller)
374 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
376 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
377 and g_atomic_pointer_get. Add g_atomic_int_set and
378 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
379 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
380 as functions (additionally to the macros in the header) for the
381 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
383 2005-12-16 Matthias Clasen <mclasen@redhat.com>
385 * glib/gmem.c (g_allocator_new): Don't return a pointer to
386 a const struct, since apps expect to be able to modify it.
387 (#324179, J. Ali Harlow)
389 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
391 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
392 for systems where the initialization of atomic variables requires a
393 write memory barrier.
395 2005-12-09 Matthias Clasen <mclasen@redhat.com>
397 * configure.in: Bump version
399 * === Released 2.9.1 ===
401 2005-12-09 Alexander Larsson <alexl@redhat.com>
405 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
408 2005-12-08 Matthias Clasen <mclasen@redhat.com>
412 2005-12-07 Martyn Russell <martyn@imendio.com>
414 * glib/gasyncqueue.c:
415 - Call g_queue_insert_sorted() instead of duplicating the code.
416 - Call g_queue_sort() instead of duplicating the code.
417 - Invert sort function results to make sure the same sort function
418 gives the same results across glist, gslist, gqueue and
421 * tests/asyncqueue-test.c:
422 - Updated the sort function to reflect the example in the
423 documentation for gasyncqueue.c.
425 2005-12-07 Martyn Russell <martyn@imendio.com>
427 * docs/reference/glib/glib-sections.txt:
428 * docs/reference/glib/tmpl/linked_lists_double.sgml:
429 * docs/reference/glib/tmpl/linked_lists_single.sgml:
432 - Added g_list_insert_sorted_with_data () and
433 g_slist_insert_sorted_with_data ().
434 - Removed the extra check in g_list_sort() and g_slist_sort() for
435 GCompareDataFunc vs. GCompareFunc.
437 2005-12-07 Tor Lillqvist <tml@novell.com>
439 * glib-gettextize.in: Look up prefix at run-time on Win32,
440 assuming the standard directory structure with glib-gettextize in
443 2005-12-06 Matthias Clasen <mclasen@redhat.com>
445 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
446 we need GTK+ 2.8 to compile against GLib 2.10.
448 2005-12-05 Manish Singh <yosh@gimp.org>
450 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
451 to override any user settings.
453 2005-12-05 Matthias Clasen <mclasen@redhat.com>
455 * configure.in: Revert an accidential commit
457 2005-12-05 Matthias Clasen <mclasen@redhat.com>
459 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
461 * tests/Makefile.am (TESTS_ENVIRONMENT):
462 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
463 MALLOC_CHECK_ and MALLOC_PERTURB_
465 * tests/run-collate-tests.sh: Run the collation tests explicitly
466 in en_US locale. (#320463)
468 * glib/gmem.h: Really deprecate GMemChunk.
475 * glib/gstrfuncs.c: Add versioned deprecation docs.
477 2005-12-05 Martyn Russell <martyn@imendio.com>
479 * docs/reference/glib/glib-sections.txt:
480 * glib/gasyncqueue.[ch]:
481 - Added support for sorting async queues by with _push_sorted(),
482 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
485 * tests/asyncqueue-test.c:
486 - Added test case for gasyncqueue.c
488 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
490 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
491 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
492 in bug 323178. moved time consuming logic from g_slice_free() out of
493 the inner loop, so g_slice_free_chain_with_offset() provides a real
494 performance benefit over g_slice_free1() now.
496 * glib/gslice.h: renamed g_slice_free_chain() to
497 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
498 a type-safe macro as suggested in bug 323178.
499 simplified the macro implementation of g_slice_free() and implemented
500 it in a type safe manner for all compliers as suggested by Morten
501 Welinder <mortenw@gnome.org>.
506 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
508 2005-12-05 Matthias Clasen <mclasen@redhat.com>
510 * glib/gasyncqueue.c: Add some docs.
512 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
514 * glib/gatomic.c: In the ia64 implementation, use
515 __sync builtin without _si or _di suffix. (#321229,
516 Stanislav Brabec, patch by Andreas Schwab)
518 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
520 * glib/gslice.h: Remove comma at the end of enum.
522 2005-12-04 Matthias Clasen <mclasen@redhat.com>
524 Handle multiple user names with the same UID better.
525 (#319535, Laszlo Peter)
527 * glib/gutils.c (g_get_any_init_do): When determining user
528 data, first look up $LOGNAME. If the UID doesn't match
529 getuid(), fall back to the current behaviour of looking
530 up the user data based on getuid().
532 2005-12-04 Matthias Clasen <mclasen@redhat.com>
534 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
535 to make gcc warn if a function result is ignored. (#145466,
536 Arjan van de Ven, Alex Larsson)
538 * glib/gmem.h: Add the new attribute to g_realloc and
541 2005-12-03 Matthias Clasen <mclasen@redhat.com>
544 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
547 * glib/gmappedfile.c: Make mapping of empty files
550 * glib/gfileutils.c: Don't fork a new process just to
551 fix the permissions of the created temp file. (#321318,
552 Alexis S. L. Carvalho)
554 2005-12-02 Matthias Clasen <mclasen@redhat.com>
556 * README.in: Add a note about Solaris threads.
559 * configure.in: Remove support for Solaris threads.
560 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
562 2005-12-02 Matthias Clasen <mclasen@redhat.com>
564 * configure.in: Define G_GUINT64_CONSTANT in analogy to
565 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
567 2005-12-02 Matthias Clasen <mclasen@redhat.com>
569 * glib/gslice.c: Win32 portability fixes and C99ism removal,
570 pointed out by Kazuki Iwamoto. (#323052)
572 2005-12-02 Matthias Clasen <mclasen@redhat.com>
574 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
575 the IA64 implementation.
577 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
579 * glib/gslice.c: some naming and type size fixups.
581 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
583 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
584 variant into something less verbose, but digestible for gcc-3.4.
586 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
588 * tests/slice-test.c: extended to perform the benchmarking on the old
589 memchunk code if 'O' is selected.
591 * tests/memchunks.c: new file which contains the old GLib mem chunks
592 implementation with prefix old_mem_chunk_.
594 * tests/Makefile.am: added memchunks.c
596 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
598 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
599 acceptable load distribution of the syscall.
600 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
601 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
603 2005-12-01 Matthias Clasen <mclasen@redhat.com>
605 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
606 double-free base interface members.
608 * tests/Makefile.am: Remove duplicate variable.
610 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
612 * glib/gslice.[hc]: new slice allocator implementation.
614 * tests/slice-test.c: added random slice allocation test.
616 * glib/gthread.[hc]: removed newly added private thread mem API.
618 * glib/gthreadinit.h:
621 * glib/gmem.c: divided glib threading initialisation into three phases,
622 initialisation where private keys and messaging are not available (only
623 needed by gmem.c), initialisation without messaging but private keys
624 available (gslice.c, gmessage.c), and full fledged initialisers that
625 server the rest of glib. initialisation functions got renamed to reflect
626 the limitations of their corresponding phases.
628 * glib/gmem.c: removed memchunk code, defer allocations to
631 * glib/gmem.[hc]: removed g_slice_* skeletons.
633 * glib/glib.symbols: added g_slice_* symbols.
635 * configure.in: check for availability of posix_memalign(3), memalign(3)
638 * glib/Makefile.am: added gslice.[hc].
640 2005-12-01 Tor Lillqvist <tml@novell.com>
642 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
643 trailing slash(es) for non-root folders. stat() fails if non-root
644 folders are specified with trailing slashes. It's too much hassle
645 to demand that callers strip such slashes themselves, especially
646 as it is easy to get it wrong and strip the slash of a root
648 (g_rename): On NT-based Windows, use MoveFileEx() with
649 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
651 2005-11-28 Matthias Clasen <mclasen@redhat.com>
653 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
656 * configure.in: Check whether do { } while (0) works.
658 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
659 G_STMT_END if it works.
661 2005-11-28 Matthias Clasen <mclasen@redhat.com>
663 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
664 * glib/gnode.c (g_node_depth_traverse_level):
665 * glib/gmem.c (g_allocator_new):
666 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
667 (g_hash_table_foreach_remove_or_steal): Silence compiler
670 2005-11-27 Matthias Clasen <mclasen@redhat.com>
672 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
673 Don't read beyond len. (#322520, Christian Persch)
675 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
677 * glib/gdataset.c: access datalist flags via atomic pointer access
678 functions, instead of acquiring the dataset lock. this is faster and
679 also matches the atomic pointer readouts in gdatalistprivate.h and
680 direct pointer modifications required by gobject.c.
682 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
684 * glib/gdataset.c: streamlined initialization code somewhat,
685 removed GData node cache.
687 2005-11-22 Matthias Clasen <mclasen@redhat.com>
689 * glib/ghash.c (g_hash_table_ref):
690 (g_hash_table_unref): Mark these as new API.
692 Avoid double locking in g_intern_string (#322133,
695 * glib/gdataset.c (g_quark_from_string_internal): New
696 internal function which factors out common parts of
697 g_quark_from[_static]_string.
698 (g_quark_from_string, g_quark_from_static_string):
699 Use g_quark_from_string_internal.
700 (g_intern_string, g_intern_static_string): Use
701 g_quark_from_string_internal, and only take the
703 (g_quark_new): Don't store the strings shifted by -1
704 in the g_quarks array.
705 (g_quark_to_string): Adapt to the previous change.
707 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
711 g_hash_table_new_full(): create hash tables with a ref count of 1.
712 g_hash_table_ref(): atomically ref_count+=1
713 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
714 when refcount reaches 0.
715 g_hash_table_destroy(): just destroy keys and values, unref by 1.
716 g_hash_table_insert():
717 g_hash_table_replace(): assert ref_count>0.
720 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
723 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
725 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
726 variables. (#321984, Andrew Paprocki)
728 2005-11-18 Matthias Clasen <mclasen@redhat.com>
730 * configure.in: Bump version
732 2005-11-17 Matthias Clasen <mclasen@redhat.com>
734 * === Released 2.9.0 ===
736 * glib/glib.symbols: Updates
738 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
739 hints to the docs. (Paolo Borelli)
743 * glib/gthread.c (g_thread_foreach): Mark as new api.
745 * README.in: Updates.
749 * glib/gdate.c (g_date_set_time_t):
750 (g_date_set_time_val): Add functions to set a date from
751 a time_t and from a GTimeVal.
752 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
755 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
756 and g_utf8_offset_to_pointer.
758 * glib/gutf8.c (g_utf8_pointer_to_offset)
759 (g_utf8_offset_to_pointer): Handle negative offsets, and use
760 "stutter stepping" for going backwards. (#320638, Larry
765 * glib/gthread.c: const correctness fixes, found
766 by Arjan van de Ven and gcc.
768 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
770 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
771 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
773 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
775 * glib/guniprop.c: Use bit hacks instead when checking a general
776 category value against multiple values.
778 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
780 2005-11-10 Simos Xenitellis <simos@gnome.org>
782 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
784 2005-11-05 Tor Lillqvist <tml@novell.com>
786 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
787 (#320688) Instead set error and return failure.
789 2005-11-07 Matthias Clasen <mclasen@redhat.com>
791 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
792 while holding the context lock. (#320886, Andy Wingo)
794 2005-11-07 Billy Biggs <vektor@dumbterm.net>
796 * tests/refcount/Makefile.am: Add a missing reference to libglib
797 in the LDADD for this test.
799 2005-11-07 Matthias Clasen <mclasen@redhat.com>
801 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
802 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
803 GLib 2.9. (#320755, Luca Ferretti)
805 2005-11-04 Matthias Clasen <mclasen@redhat.com>
807 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
808 use a faster lookup table for g_unichar_get_mirror_char().
809 * glib/gmirroringtable.h: The new table.
810 * glib/gunichartables.h: Remove bidi_mirroring_table.
811 * glib/gen-unicode-tables.pl: Don't generate the mirroring
813 * glib/glib-mirroring-tab/*: Sources for the program
814 which generated gmirroringtable.h.
816 * glib/glist.c: Avoid some code duplication.
818 * glib/gscanner.h: Include gdataset.h. (#320322)
820 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
821 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
822 to help find critical warnings in applications. (#320017,
825 2005-11-02 Tor Lillqvist <tml@novell.com>
827 * glib/glib.symbols: Remove large amount of trailing whitespace
828 from one line. Remove #ifdef G_UNDEFINED from around
829 g_hash_table_new and g_hash_table_new_full symbols, what was the
830 purpose of that? Add the g_slice functions.
832 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
834 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
835 added g_slice_*() API to allocate and cache small bits of memory.
836 an actuall allocator implementation for g_slice_*() is still pending.
838 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
839 changed GRealThread list to use in-structure *next; fields instead
840 of GSList, in order for thread iteration to not depenend on g_slice_*()
842 _g_thread_mem_private_get():
843 _g_thread_mem_private_set(): added accessors for private memory,
844 needed because the ordinary GPrivate implementation relies on GArray
845 and GSList and therefore indirectly on working g_slice_*() allocations.
848 g_thread_foreach(): new public API function to loop over all existing threads.
862 removed GAllocator and free list usages and accompanying locks.
863 use g_slice_*() API to allocate and cache small bits of memory.
865 * glib/ghook.h: removed GMemChunk field from public API.
868 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
870 * glib/gnode.h: deprecate allocator API.
872 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
873 be used for poll rec lists.
875 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
876 g_relation_destroy(): free all tuples from the all_tuples hash table,
877 this effectively maintains the life time track keeping of tuples.
878 g_relation_delete_tuple(): free tuples which are removed from the
879 all_tuples hash table. this fixes a temporary leak that was present
880 in the memchunk code until the destruction of the relation.
882 2005-10-29 Matthias Clasen <mclasen@redhat.com>
884 * tests/convert-test.c: Add some tests for conversions between
885 UTF-8, UCS-4 and UTF-16.
887 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
888 of len == -1, noticed by Morten Welinder.
890 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
892 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
894 2005-10-26 Matthias Clasen <mclasen@redhat.com>
896 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
897 in the error case, and add some documentation. (#319806, Morten
900 2005-10-19 Manish Singh <yosh@gimp.org>
902 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
903 config.h symbol automatically. Fixes bug #313731.
905 2005-10-19 Tor Lillqvist <tml@novell.com>
907 * glib/gwin32.c (g_win32_get_package_installation_directory):
908 Return a g_strdup()ed copy of the value stored in the hash table,
909 so that it can be g_free()d without leaving a dangling pointer in
910 the hash table. (#319232)
912 2005-10-06 Matthias Clasen <mclasen@redhat.com>
914 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
915 all-zero sequences correctly. (#317930, Sebastien Bacher)
917 2005-10-05 Matthias Clasen <mclasen@redhat.com>
919 * glib/gmarkup.c (g_markup_escape_text): Doc addition
921 * Makefile.am: Add Collation tests.
923 * tests/collate/*: Inputs and expected outputs for collation tests.
925 * tests/run-collate-tests.sh: Script to run collation tests.
927 * tests/unicode-collate.c (main): Rework slightly to make
928 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
930 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
932 * docs/reference/glib/tmpl/unicode.sgml:
933 * glib/gen-unicode-tables.pl:
935 * glib/gunichartables.h:
937 * tests/casefold.txt:
938 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
939 GUnicodeBreakType types. That may break some applications, like
942 2005-09-26 Matthias Clasen <mclasen@redhat.com>
944 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
945 docs. (#314393, Matthew F. Barnes)
948 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
949 they are already declared in glib.h. This doesn't break documented
950 use of gprintf.h, but should probably be pointed out in the
951 release notes for 2.10. (#314232, Behdad Esfahbod)
953 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
955 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
956 complexity optimization, based on a patch from Matthias Clasen.
958 * tests/patterntest.c: more tests, mostly from matthias.
960 2005-09-20 Matthias Clasen <mclasen@redhat.com>
962 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
963 (#316703, Mark Drago)
965 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
967 * glib/gpattern.c: applied a patch from matthias which checks on the
968 upper bound of GPatternSpec length to optimize matches.
971 * tests/patterntest.c: added more match cases.
973 2005-09-16 Tor Lillqvist <tml@novell.com>
975 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
976 warning to doc comment that these functions should not be used on
977 encodings like CP932.
979 2005-09-14 Matthias Clasen <mclasen@redhat.com>
981 * tests/keyfile-test.c: Add a test for grup names of length 1.
983 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
984 of length 1. (#316309)
986 2005-09-12 Matthias Clasen <mclasen@redhat.com>
988 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
990 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
992 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
993 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
996 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
998 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
999 (child_watch_helper_thread): Remove some dead code. Closes
1002 2005-09-07 Tor Lillqvist <tml@novell.com>
1004 * glib/Makefile.am: Create also a console version of the
1005 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1006 It's otherwise identical to gspawn-win32-helper.exe, except marked
1007 as a console application (linked without the -mwindows option).
1009 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1010 Drop the dont_wait parameter. Its truth value correlated 100% with
1011 the NULLness of the exit_status parameter anyway, so it's enough
1012 to check whether exit_status is NULL. Invert the sense of the
1013 dont_return_handle parameter and rename it to do_return_handle, to
1014 make the code easier to read by avoiding double negations.
1016 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1017 to do_spawn_with_pipes() accordingly.
1019 (do_spawn_with_pipes): If we have a console, use the console
1020 version of the helper program, otherwise use the GUI one. This
1021 avoids extra console windows opening up in some situations. (In
1022 case a console application uses the GUI gspawn-win32-helper.exe to
1023 spawn another console application we would get a separate console
1024 for the spawned console application).
1026 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1028 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1030 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1031 if the file is too large. (#315275, Kjartan Maraas)
1033 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1034 of read() is signed. (#315273, Kjartan Maraas)
1036 2005-08-31 Tor Lillqvist <tml@novell.com>
1038 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1039 the EXE and each DLL have their own atexit function chains.
1041 #define g_atexit as atexit instead. This means it has a
1042 better chance of doing what the caller wants. For instance,
1043 gtkhtml calls g_atexit() registering a function in gtkhtml
1044 itself. This caused a crash when g_atexit() was implemented as a
1045 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1046 the time the GLib DLL got unloaded.
1048 * glib/gutils.c: #undef the #define mentioned above, to also get a
1049 real g_atexit() into the DLL for backward compatibility. Document
1050 the Windows behaviour of g_atexit(), and document the varying ways
1051 atexit() can behave in the context of dynamically loaded modules
1054 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1056 * glib/glib.symbols:
1058 * glib/gdataset.c: Add string interning functions.
1060 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1062 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1065 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1067 * configure.in: Bump version to 2.9.0
1069 2005-08-25 Tor Lillqvist <tml@novell.com>
1071 Make also the g_spawn*() functions take parameters in the GLib
1072 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1073 API or ABI. Like the other GLib API that was earlier changed to
1074 use UTF-8 on Windows, the names of the functions that take UTF-8
1075 have _utf8 suffixes added by using preprocessor macros in the
1076 header file. The old names are kept for functions with the old
1077 behaviour, taking parameters in the system codepage, for DLL ABI
1080 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1081 g_spawn*() functions.
1083 * glib/gspawn-win32.c: Use wide-char API on NT-based
1084 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1085 system codepage (Win9x) and call the C library _wspawn*() or
1086 spawn*() functions respectvely. Add DLL ABI stability versions
1087 that take parameters in the system codepage.
1089 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1090 wide-char versions of argv and envp, and use wide-char API to
1091 change directory and spawn the program to run. Remove the verbose
1092 debugging output, it was too complex to modify for the wide-char
1093 features. (Just add temporary debugging printouts if needed, no
1094 need to have them permanently in the source.)
1096 * glib/gspawn.c: Corresponding documentation updates.
1098 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1099 symbols as PRIVATE, add the new _utf8-suffixed ones.
1101 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1103 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1104 guint64 otherwise and that can produce warnings about comparison
1105 between signed and unsigned.
1107 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1109 * glib/gutils.c: Fix the crt_externs.h include.
1111 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1113 * NEWS: Fix spelling of my first name.
1115 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1119 * === Released 2.8.1 ===
1123 2005-08-20 Hans Breuer <hans@breuer.org>
1125 * glib/makefile.msc.in : link with ws2_32.lib
1127 2005-08-18 Tor Lillqvist <tml@novell.com>
1129 * configure.in: Check for <sys/wait.h>
1131 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1133 2005-08-18 Ross Burton <ross@burtonini.com>
1136 Optimise single-character insertions.
1141 * tests/string-test.c:
1142 Add tests for new optimisation, and fix a leak.
1144 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1146 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1148 * glib/gutils.c: On Darwin, include crt-externs.h and
1149 define environ using _NSGetEnviron(). (#313731)
1151 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1153 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1154 (g_get_any_init_do): ... to this new function.
1155 (g_get_any_init): Declare as inline.
1156 (g_get_any_init_locked): New inline function, does the locking.
1157 Make use of these two throughout the code.
1159 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1161 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1162 the child process and then simply return. This makes
1163 The "S" option work as documented in g_on_error_query().
1164 (#313125, Matthew F. Barnes)
1166 * glib/gunicode.h: Update the link to Unicode category
1167 values. (#313369, Behnam Esfahbod)
1169 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1170 a little. (#311727, Tristan van Berkom)
1172 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1173 check work on ia64 too, where some symbols ended up
1174 in yet another section.