1 2006-01-24 Matthias Clasen <mclasen@redhat.com>
3 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
4 on all little-endian systems. (#143380, Marc Moorcroft)
6 2006-01-23 Matthias Clasen <mclasen@redhat.com>
8 * configure.in: Change the shared libary extension for hpux-ia64
9 to so. (#328253, The Written Word)
11 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
13 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
14 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
17 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
19 * glib/gslice.c (slice_config_init): initialize GSlice config from
20 G_SLICE environment variable. we support G_SLICE=always-malloc
21 currently, which forces all g_slice_*() allocations to use the system
25 g_parse_debug_string(): added a note about not using g_malloc() here.
26 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
27 g_slice. contains only guesswork in the WIN32 branch.
29 2006-01-18 Matthias Clasen <mclasen@redhat.com>
33 * === Released 2.9.4 ===
37 2006-01-17 Matthias Clasen <mclasen@redhat.com>
39 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
40 call. (#327421, Morten Welinder)
42 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
44 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
45 G_LOCK. Use the unused_thread_queue lock instead. Change
46 g_thread_pool_thread_proxy such that threads only wait on
47 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
48 due to superfluous tasks. Global tasks wait at most for
49 max-idle-time milliseconds. Make sure, that no task is woken up
50 twice for the same event via a wakeup_serial. This fixes #324228.
52 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
53 invalid NULL into the thread pools. This as well fixes #327290.
55 2006-01-16 Matthias Clasen <mclasen@redhat.com>
57 * configure.in: Bump version
59 * === Released 2.9.3 ===
63 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
64 Rework the strict aliasing fix to not break C++, pointed
65 out by Murray Cumming.
67 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
68 Signal waiting threads, problem noticed by Christian Kellner.
70 2006-01-16 Matthias Clasen <mclasen@redhat.com>
72 Fix bug 326747, Alberto Ruiz:
74 * glib/gconvert.c (make_valid_utf8): Change this function to
75 replace unknown characters by the Unicode replacement character
76 instead of '?', and don't append "(invalid encoding)".
77 (g_filename_display_name, g_filename_display_basename): Document
78 how to determine if the filename was in an invalid encoding.
80 2006-01-14 Matthias Clasen <mclasen@redhat.com>
82 * glib/gtree.c: Replace the simple recursive implementation
83 by a nonrecursive, threaded implementation by Maurizio
86 2006-01-12 Matthias Clasen <mclasen@redhat.com>
88 * tests/asyncqueue-test.c (main): Fix the
89 assert to not always trigger. (#326558,
92 2006-01-11 Matthias Clasen <mclasen@redhat.com>
94 * tests/convert-test.c: Don't test items_read and
95 items_written in error cases where it is not set.
97 * tests/asyncqueue-test.c: Set error to NULL.
99 2006-01-08 Matthias Clasen <mclasen@redhat.com>
101 * glib/gtree.c: Break some long lines.
103 * tests/tree-test.c: Add more tests.
105 * glib/gslice.c: Include config.h first, pointed out
108 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
110 * docs/reference/gobject/tmpl/gtype.sgml: explain that
111 G_TYPE_INSTANCE_GET_CLASS() does behave different during
114 2006-01-06 Matthias Clasen <mclasen@redhat.com>
116 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
117 that we have posix_memalign(). (#323937, Bogdan Nicula)
119 2006-01-05 Matthias Clasen <mclasen@redhat.com>
121 * configure.in: Bump version
123 * === Released 2.9.2 ===
125 2006-01-05 Hans Breuer <hans@breuer.org>
127 * glib/makefile.msc.in : added gslice.obj
129 2006-01-05 Matthias Clasen <mclasen@redhat.com>
133 * glib/gmain.c: Clarify the documentation of source ids
134 a bit. (#325874, Dan Williams)
136 * configure.in: Fix another strict aliasing problem.
138 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
141 2006-01-04 Manish Singh <yosh@gimp.org>
144 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
145 either, since they have the same issue as g_rmdir. (Related to
148 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
150 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
151 order when creating magazines, so we hand out chunks with ascending
154 2006-01-04 Abel Cheung <maddog@linuxhall.org>
156 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
158 2006-01-03 Martyn Russell <martyn@imendio.com>
160 * docs/reference/glib/glib-sections.txt:
162 * glib/gthreadpool.[ch]:
163 - Added new API g_thread_pool_get_idle_time() and
164 g_thread_pool_set_idle_time(). (#324228).
166 * tests/threadpool-test.c:
167 - Updated test case to do thread pool sorting, thread pool with
168 no sorting and a thread pool with idle thread timeouts.
170 2006-01-03 Matthias Clasen <mclasen@redhat.com>
172 * glib/gmain.h: Add new functions here, too.
174 * glib/glib.symbols: Add new functions.
176 2005-12-20 Michael Meeks <michael.meeks@novell.com>
178 * glib/gmain.c (g_main_context_is_owner): new method
179 to determine if the current thread is the owner of the
182 2006-01-02 Matthias Clasen <mclasen@redhat.com>
186 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
187 expanding to rmdir, since rmdir is not declared in a portable
188 system header we can include in gstdio.h. (#325249, Jani Monoses)
190 2006-01-01 Tor Lillqvist <tml@novell.com>
192 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
193 stdout and stderr to unbuffered. Otherwise the giochannel layer
194 will try to read from them regardless whether the
195 g_io_channel_win32_poll() call here has indicated
196 readability or not. (#325310)
198 2005-12-29 Matthias Clasen <mclasen@redhat.com>
200 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
201 in the docs. (#325273, Declan Naughton)
203 2005-12-27 Manish Singh <yosh@gimp.org>
205 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
206 make sure we really override things.
208 2005-12-27 Matthias Clasen <mclasen@redhat.com>
210 Fix #316221, Michal Benes, Stanislav Brabec;
212 * configure.in: Fix a strict aliasing problem in
213 g_static_mutex_get_mutex().
214 * glib/gthread.h: ...and in
215 g_static_mutex_get_mutex_impl_shortcut().
217 * glib/gdatasetprivate.h: Add a cast to silence compiler
218 warnings. (#321978, Andrew Paprocki)
220 Partial fix for bug #323937, Bogdan Nicula.
222 * configure.in: Check for malloc.h
223 * glib/gslice.c: Don't include malloc.h unconditionally.
225 2005-12-27 Manish Singh <yosh@gimp.org>
227 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
228 (#324950, Dan Yefimov)
230 2005-12-26 Matthias Clasen <mclasen@redhat.com>
232 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
235 2005-12-25 Matthias Clasen <mclasen@redhat.com>
237 * AUTHORS: Update my email
239 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
241 2005-12-21 Matthias Clasen <mclasen@redhat.com>
243 * README.in: Add some notes on when 'make check' may fail.
245 2005-12-20 Matthias Clasen <mclasen@redhat.com>
248 * glib/gthreadpool.h:
249 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
250 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
252 * tests/threadpool-test.c: Test this.
254 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
256 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
257 handle errors without depending on gmessage.c which might not be
258 setup when the error occours.
259 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
260 always freeing can be achieved by adjusting the working set time to
261 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
262 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
263 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
264 minimize wastage if small amounts of differently sized structrues are
265 allocated, this does come at a performance cost of roughly 5% though.
266 fixed up block alignment calculation, so it works for varying
267 block sizes. only use strerror() not g_strerror() since the latter
268 depends on working GQuark and GSlice.
269 mem_error(): implemented in terms of fprintf and vfprintf.
271 * tests/slice-color.c: new program to test cache colorization effects.
273 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
274 G_SLICE_CONFIG_WORKING_SET_MSECS.
276 2005-12-17 Matthias Clasen <mclasen@redhat.com>
278 * glib/goption.c (parse_short_option): Set an error in all
279 failure cases. (#324332, Tim-Philipp Müller)
281 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
283 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
284 and g_atomic_pointer_get. Add g_atomic_int_set and
285 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
286 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
287 as functions (additionally to the macros in the header) for the
288 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
290 2005-12-16 Matthias Clasen <mclasen@redhat.com>
292 * glib/gmem.c (g_allocator_new): Don't return a pointer to
293 a const struct, since apps expect to be able to modify it.
294 (#324179, J. Ali Harlow)
296 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
298 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
299 for systems where the initialization of atomic variables requires a
300 write memory barrier.
302 2005-12-09 Matthias Clasen <mclasen@redhat.com>
304 * configure.in: Bump version
306 * === Released 2.9.1 ===
308 2005-12-09 Alexander Larsson <alexl@redhat.com>
312 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
315 2005-12-08 Matthias Clasen <mclasen@redhat.com>
319 2005-12-07 Martyn Russell <martyn@imendio.com>
321 * glib/gasyncqueue.c:
322 - Call g_queue_insert_sorted() instead of duplicating the code.
323 - Call g_queue_sort() instead of duplicating the code.
324 - Invert sort function results to make sure the same sort function
325 gives the same results across glist, gslist, gqueue and
328 * tests/asyncqueue-test.c:
329 - Updated the sort function to reflect the example in the
330 documentation for gasyncqueue.c.
332 2005-12-07 Martyn Russell <martyn@imendio.com>
334 * docs/reference/glib/glib-sections.txt:
335 * docs/reference/glib/tmpl/linked_lists_double.sgml:
336 * docs/reference/glib/tmpl/linked_lists_single.sgml:
339 - Added g_list_insert_sorted_with_data () and
340 g_slist_insert_sorted_with_data ().
341 - Removed the extra check in g_list_sort() and g_slist_sort() for
342 GCompareDataFunc vs. GCompareFunc.
344 2005-12-07 Tor Lillqvist <tml@novell.com>
346 * glib-gettextize.in: Look up prefix at run-time on Win32,
347 assuming the standard directory structure with glib-gettextize in
350 2005-12-06 Matthias Clasen <mclasen@redhat.com>
352 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
353 we need GTK+ 2.8 to compile against GLib 2.10.
355 2005-12-05 Manish Singh <yosh@gimp.org>
357 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
358 to override any user settings.
360 2005-12-05 Matthias Clasen <mclasen@redhat.com>
362 * configure.in: Revert an accidential commit
364 2005-12-05 Matthias Clasen <mclasen@redhat.com>
366 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
368 * tests/Makefile.am (TESTS_ENVIRONMENT):
369 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
370 MALLOC_CHECK_ and MALLOC_PERTURB_
372 * tests/run-collate-tests.sh: Run the collation tests explicitly
373 in en_US locale. (#320463)
375 * glib/gmem.h: Really deprecate GMemChunk.
382 * glib/gstrfuncs.c: Add versioned deprecation docs.
384 2005-12-05 Martyn Russell <martyn@imendio.com>
386 * docs/reference/glib/glib-sections.txt:
387 * glib/gasyncqueue.[ch]:
388 - Added support for sorting async queues by with _push_sorted(),
389 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
392 * tests/asyncqueue-test.c:
393 - Added test case for gasyncqueue.c
395 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
397 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
398 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
399 in bug 323178. moved time consuming logic from g_slice_free() out of
400 the inner loop, so g_slice_free_chain_with_offset() provides a real
401 performance benefit over g_slice_free1() now.
403 * glib/gslice.h: renamed g_slice_free_chain() to
404 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
405 a type-safe macro as suggested in bug 323178.
406 simplified the macro implementation of g_slice_free() and implemented
407 it in a type safe manner for all compliers as suggested by Morten
408 Welinder <mortenw@gnome.org>.
413 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
415 2005-12-05 Matthias Clasen <mclasen@redhat.com>
417 * glib/gasyncqueue.c: Add some docs.
419 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
421 * glib/gatomic.c: In the ia64 implementation, use
422 __sync builtin without _si or _di suffix. (#321229,
423 Stanislav Brabec, patch by Andreas Schwab)
425 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
427 * glib/gslice.h: Remove comma at the end of enum.
429 2005-12-04 Matthias Clasen <mclasen@redhat.com>
431 Handle multiple user names with the same UID better.
432 (#319535, Laszlo Peter)
434 * glib/gutils.c (g_get_any_init_do): When determining user
435 data, first look up $LOGNAME. If the UID doesn't match
436 getuid(), fall back to the current behaviour of looking
437 up the user data based on getuid().
439 2005-12-04 Matthias Clasen <mclasen@redhat.com>
441 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
442 to make gcc warn if a function result is ignored. (#145466,
443 Arjan van de Ven, Alex Larsson)
445 * glib/gmem.h: Add the new attribute to g_realloc and
448 2005-12-03 Matthias Clasen <mclasen@redhat.com>
451 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
454 * glib/gmappedfile.c: Make mapping of empty files
457 * glib/gfileutils.c: Don't fork a new process just to
458 fix the permissions of the created temp file. (#321318,
459 Alexis S. L. Carvalho)
461 2005-12-02 Matthias Clasen <mclasen@redhat.com>
463 * README.in: Add a note about Solaris threads.
466 * configure.in: Remove support for Solaris threads.
467 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
469 2005-12-02 Matthias Clasen <mclasen@redhat.com>
471 * configure.in: Define G_GUINT64_CONSTANT in analogy to
472 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
474 2005-12-02 Matthias Clasen <mclasen@redhat.com>
476 * glib/gslice.c: Win32 portability fixes and C99ism removal,
477 pointed out by Kazuki Iwamoto. (#323052)
479 2005-12-02 Matthias Clasen <mclasen@redhat.com>
481 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
482 the IA64 implementation.
484 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
486 * glib/gslice.c: some naming and type size fixups.
488 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
490 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
491 variant into something less verbose, but digestible for gcc-3.4.
493 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
495 * tests/slice-test.c: extended to perform the benchmarking on the old
496 memchunk code if 'O' is selected.
498 * tests/memchunks.c: new file which contains the old GLib mem chunks
499 implementation with prefix old_mem_chunk_.
501 * tests/Makefile.am: added memchunks.c
503 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
505 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
506 acceptable load distribution of the syscall.
507 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
508 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
510 2005-12-01 Matthias Clasen <mclasen@redhat.com>
512 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
513 double-free base interface members.
515 * tests/Makefile.am: Remove duplicate variable.
517 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
519 * glib/gslice.[hc]: new slice allocator implementation.
521 * tests/slice-test.c: added random slice allocation test.
523 * glib/gthread.[hc]: removed newly added private thread mem API.
525 * glib/gthreadinit.h:
528 * glib/gmem.c: divided glib threading initialisation into three phases,
529 initialisation where private keys and messaging are not available (only
530 needed by gmem.c), initialisation without messaging but private keys
531 available (gslice.c, gmessage.c), and full fledged initialisers that
532 server the rest of glib. initialisation functions got renamed to reflect
533 the limitations of their corresponding phases.
535 * glib/gmem.c: removed memchunk code, defer allocations to
538 * glib/gmem.[hc]: removed g_slice_* skeletons.
540 * glib/glib.symbols: added g_slice_* symbols.
542 * configure.in: check for availability of posix_memalign(3), memalign(3)
545 * glib/Makefile.am: added gslice.[hc].
547 2005-12-01 Tor Lillqvist <tml@novell.com>
549 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
550 trailing slash(es) for non-root folders. stat() fails if non-root
551 folders are specified with trailing slashes. It's too much hassle
552 to demand that callers strip such slashes themselves, especially
553 as it is easy to get it wrong and strip the slash of a root
555 (g_rename): On NT-based Windows, use MoveFileEx() with
556 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
558 2005-11-28 Matthias Clasen <mclasen@redhat.com>
560 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
563 * configure.in: Check whether do { } while (0) works.
565 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
566 G_STMT_END if it works.
568 2005-11-28 Matthias Clasen <mclasen@redhat.com>
570 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
571 * glib/gnode.c (g_node_depth_traverse_level):
572 * glib/gmem.c (g_allocator_new):
573 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
574 (g_hash_table_foreach_remove_or_steal): Silence compiler
577 2005-11-27 Matthias Clasen <mclasen@redhat.com>
579 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
580 Don't read beyond len. (#322520, Christian Persch)
582 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
584 * glib/gdataset.c: access datalist flags via atomic pointer access
585 functions, instead of acquiring the dataset lock. this is faster and
586 also matches the atomic pointer readouts in gdatalistprivate.h and
587 direct pointer modifications required by gobject.c.
589 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
591 * glib/gdataset.c: streamlined initialization code somewhat,
592 removed GData node cache.
594 2005-11-22 Matthias Clasen <mclasen@redhat.com>
596 * glib/ghash.c (g_hash_table_ref):
597 (g_hash_table_unref): Mark these as new API.
599 Avoid double locking in g_intern_string (#322133,
602 * glib/gdataset.c (g_quark_from_string_internal): New
603 internal function which factors out common parts of
604 g_quark_from[_static]_string.
605 (g_quark_from_string, g_quark_from_static_string):
606 Use g_quark_from_string_internal.
607 (g_intern_string, g_intern_static_string): Use
608 g_quark_from_string_internal, and only take the
610 (g_quark_new): Don't store the strings shifted by -1
611 in the g_quarks array.
612 (g_quark_to_string): Adapt to the previous change.
614 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
618 g_hash_table_new_full(): create hash tables with a ref count of 1.
619 g_hash_table_ref(): atomically ref_count+=1
620 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
621 when refcount reaches 0.
622 g_hash_table_destroy(): just destroy keys and values, unref by 1.
623 g_hash_table_insert():
624 g_hash_table_replace(): assert ref_count>0.
627 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
630 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
632 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
633 variables. (#321984, Andrew Paprocki)
635 2005-11-18 Matthias Clasen <mclasen@redhat.com>
637 * configure.in: Bump version
639 2005-11-17 Matthias Clasen <mclasen@redhat.com>
641 * === Released 2.9.0 ===
643 * glib/glib.symbols: Updates
645 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
646 hints to the docs. (Paolo Borelli)
650 * glib/gthread.c (g_thread_foreach): Mark as new api.
652 * README.in: Updates.
656 * glib/gdate.c (g_date_set_time_t):
657 (g_date_set_time_val): Add functions to set a date from
658 a time_t and from a GTimeVal.
659 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
662 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
663 and g_utf8_offset_to_pointer.
665 * glib/gutf8.c (g_utf8_pointer_to_offset)
666 (g_utf8_offset_to_pointer): Handle negative offsets, and use
667 "stutter stepping" for going backwards. (#320638, Larry
672 * glib/gthread.c: const correctness fixes, found
673 by Arjan van de Ven and gcc.
675 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
677 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
678 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
680 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
682 * glib/guniprop.c: Use bit hacks instead when checking a general
683 category value against multiple values.
685 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
687 2005-11-10 Simos Xenitellis <simos@gnome.org>
689 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
691 2005-11-05 Tor Lillqvist <tml@novell.com>
693 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
694 (#320688) Instead set error and return failure.
696 2005-11-07 Matthias Clasen <mclasen@redhat.com>
698 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
699 while holding the context lock. (#320886, Andy Wingo)
701 2005-11-07 Billy Biggs <vektor@dumbterm.net>
703 * tests/refcount/Makefile.am: Add a missing reference to libglib
704 in the LDADD for this test.
706 2005-11-07 Matthias Clasen <mclasen@redhat.com>
708 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
709 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
710 GLib 2.9. (#320755, Luca Ferretti)
712 2005-11-04 Matthias Clasen <mclasen@redhat.com>
714 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
715 use a faster lookup table for g_unichar_get_mirror_char().
716 * glib/gmirroringtable.h: The new table.
717 * glib/gunichartables.h: Remove bidi_mirroring_table.
718 * glib/gen-unicode-tables.pl: Don't generate the mirroring
720 * glib/glib-mirroring-tab/*: Sources for the program
721 which generated gmirroringtable.h.
723 * glib/glist.c: Avoid some code duplication.
725 * glib/gscanner.h: Include gdataset.h. (#320322)
727 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
728 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
729 to help find critical warnings in applications. (#320017,
732 2005-11-02 Tor Lillqvist <tml@novell.com>
734 * glib/glib.symbols: Remove large amount of trailing whitespace
735 from one line. Remove #ifdef G_UNDEFINED from around
736 g_hash_table_new and g_hash_table_new_full symbols, what was the
737 purpose of that? Add the g_slice functions.
739 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
741 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
742 added g_slice_*() API to allocate and cache small bits of memory.
743 an actuall allocator implementation for g_slice_*() is still pending.
745 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
746 changed GRealThread list to use in-structure *next; fields instead
747 of GSList, in order for thread iteration to not depenend on g_slice_*()
749 _g_thread_mem_private_get():
750 _g_thread_mem_private_set(): added accessors for private memory,
751 needed because the ordinary GPrivate implementation relies on GArray
752 and GSList and therefore indirectly on working g_slice_*() allocations.
755 g_thread_foreach(): new public API function to loop over all existing threads.
769 removed GAllocator and free list usages and accompanying locks.
770 use g_slice_*() API to allocate and cache small bits of memory.
772 * glib/ghook.h: removed GMemChunk field from public API.
775 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
777 * glib/gnode.h: deprecate allocator API.
779 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
780 be used for poll rec lists.
782 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
783 g_relation_destroy(): free all tuples from the all_tuples hash table,
784 this effectively maintains the life time track keeping of tuples.
785 g_relation_delete_tuple(): free tuples which are removed from the
786 all_tuples hash table. this fixes a temporary leak that was present
787 in the memchunk code until the destruction of the relation.
789 2005-10-29 Matthias Clasen <mclasen@redhat.com>
791 * tests/convert-test.c: Add some tests for conversions between
792 UTF-8, UCS-4 and UTF-16.
794 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
795 of len == -1, noticed by Morten Welinder.
797 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
799 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
801 2005-10-26 Matthias Clasen <mclasen@redhat.com>
803 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
804 in the error case, and add some documentation. (#319806, Morten
807 2005-10-19 Manish Singh <yosh@gimp.org>
809 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
810 config.h symbol automatically. Fixes bug #313731.
812 2005-10-19 Tor Lillqvist <tml@novell.com>
814 * glib/gwin32.c (g_win32_get_package_installation_directory):
815 Return a g_strdup()ed copy of the value stored in the hash table,
816 so that it can be g_free()d without leaving a dangling pointer in
817 the hash table. (#319232)
819 2005-10-06 Matthias Clasen <mclasen@redhat.com>
821 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
822 all-zero sequences correctly. (#317930, Sebastien Bacher)
824 2005-10-05 Matthias Clasen <mclasen@redhat.com>
826 * glib/gmarkup.c (g_markup_escape_text): Doc addition
828 * Makefile.am: Add Collation tests.
830 * tests/collate/*: Inputs and expected outputs for collation tests.
832 * tests/run-collate-tests.sh: Script to run collation tests.
834 * tests/unicode-collate.c (main): Rework slightly to make
835 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
837 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
839 * docs/reference/glib/tmpl/unicode.sgml:
840 * glib/gen-unicode-tables.pl:
842 * glib/gunichartables.h:
844 * tests/casefold.txt:
845 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
846 GUnicodeBreakType types. That may break some applications, like
849 2005-09-26 Matthias Clasen <mclasen@redhat.com>
851 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
852 docs. (#314393, Matthew F. Barnes)
855 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
856 they are already declared in glib.h. This doesn't break documented
857 use of gprintf.h, but should probably be pointed out in the
858 release notes for 2.10. (#314232, Behdad Esfahbod)
860 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
862 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
863 complexity optimization, based on a patch from Matthias Clasen.
865 * tests/patterntest.c: more tests, mostly from matthias.
867 2005-09-20 Matthias Clasen <mclasen@redhat.com>
869 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
870 (#316703, Mark Drago)
872 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
874 * glib/gpattern.c: applied a patch from matthias which checks on the
875 upper bound of GPatternSpec length to optimize matches.
878 * tests/patterntest.c: added more match cases.
880 2005-09-16 Tor Lillqvist <tml@novell.com>
882 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
883 warning to doc comment that these functions should not be used on
884 encodings like CP932.
886 2005-09-14 Matthias Clasen <mclasen@redhat.com>
888 * tests/keyfile-test.c: Add a test for grup names of length 1.
890 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
891 of length 1. (#316309)
893 2005-09-12 Matthias Clasen <mclasen@redhat.com>
895 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
897 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
899 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
900 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
903 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
905 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
906 (child_watch_helper_thread): Remove some dead code. Closes
909 2005-09-07 Tor Lillqvist <tml@novell.com>
911 * glib/Makefile.am: Create also a console version of the
912 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
913 It's otherwise identical to gspawn-win32-helper.exe, except marked
914 as a console application (linked without the -mwindows option).
916 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
917 Drop the dont_wait parameter. Its truth value correlated 100% with
918 the NULLness of the exit_status parameter anyway, so it's enough
919 to check whether exit_status is NULL. Invert the sense of the
920 dont_return_handle parameter and rename it to do_return_handle, to
921 make the code easier to read by avoiding double negations.
923 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
924 to do_spawn_with_pipes() accordingly.
926 (do_spawn_with_pipes): If we have a console, use the console
927 version of the helper program, otherwise use the GUI one. This
928 avoids extra console windows opening up in some situations. (In
929 case a console application uses the GUI gspawn-win32-helper.exe to
930 spawn another console application we would get a separate console
931 for the spawned console application).
933 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
935 2005-09-05 Matthias Clasen <mclasen@redhat.com>
937 * glib/gmappedfile.c (g_mapped_file_new): Report an error
938 if the file is too large. (#315275, Kjartan Maraas)
940 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
941 of read() is signed. (#315273, Kjartan Maraas)
943 2005-08-31 Tor Lillqvist <tml@novell.com>
945 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
946 the EXE and each DLL have their own atexit function chains.
948 #define g_atexit as atexit instead. This means it has a
949 better chance of doing what the caller wants. For instance,
950 gtkhtml calls g_atexit() registering a function in gtkhtml
951 itself. This caused a crash when g_atexit() was implemented as a
952 function in the GLib DLL. The gtkhtml DLL was already unloaded by
953 the time the GLib DLL got unloaded.
955 * glib/gutils.c: #undef the #define mentioned above, to also get a
956 real g_atexit() into the DLL for backward compatibility. Document
957 the Windows behaviour of g_atexit(), and document the varying ways
958 atexit() can behave in the context of dynamically loaded modules
961 2005-08-31 Matthias Clasen <mclasen@redhat.com>
965 * glib/gdataset.c: Add string interning functions.
967 2005-08-28 Matthias Clasen <mclasen@redhat.com>
969 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
972 2005-08-26 Matthias Clasen <mclasen@redhat.com>
974 * configure.in: Bump version to 2.9.0
976 2005-08-25 Tor Lillqvist <tml@novell.com>
978 Make also the g_spawn*() functions take parameters in the GLib
979 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
980 API or ABI. Like the other GLib API that was earlier changed to
981 use UTF-8 on Windows, the names of the functions that take UTF-8
982 have _utf8 suffixes added by using preprocessor macros in the
983 header file. The old names are kept for functions with the old
984 behaviour, taking parameters in the system codepage, for DLL ABI
987 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
988 g_spawn*() functions.
990 * glib/gspawn-win32.c: Use wide-char API on NT-based
991 Windows. Convert parameters from UTF-8 to wide chars (NT) or
992 system codepage (Win9x) and call the C library _wspawn*() or
993 spawn*() functions respectvely. Add DLL ABI stability versions
994 that take parameters in the system codepage.
996 * glib/gspawn-win32-helper.c: On NT-based Windows use the
997 wide-char versions of argv and envp, and use wide-char API to
998 change directory and spawn the program to run. Remove the verbose
999 debugging output, it was too complex to modify for the wide-char
1000 features. (Just add temporary debugging printouts if needed, no
1001 need to have them permanently in the source.)
1003 * glib/gspawn.c: Corresponding documentation updates.
1005 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1006 symbols as PRIVATE, add the new _utf8-suffixed ones.
1008 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1010 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1011 guint64 otherwise and that can produce warnings about comparison
1012 between signed and unsigned.
1014 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1016 * glib/gutils.c: Fix the crt_externs.h include.
1018 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1020 * NEWS: Fix spelling of my first name.
1022 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1026 * === Released 2.8.1 ===
1030 2005-08-20 Hans Breuer <hans@breuer.org>
1032 * glib/makefile.msc.in : link with ws2_32.lib
1034 2005-08-18 Tor Lillqvist <tml@novell.com>
1036 * configure.in: Check for <sys/wait.h>
1038 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1040 2005-08-18 Ross Burton <ross@burtonini.com>
1043 Optimise single-character insertions.
1048 * tests/string-test.c:
1049 Add tests for new optimisation, and fix a leak.
1051 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1053 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1055 * glib/gutils.c: On Darwin, include crt-externs.h and
1056 define environ using _NSGetEnviron(). (#313731)
1058 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1060 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1061 (g_get_any_init_do): ... to this new function.
1062 (g_get_any_init): Declare as inline.
1063 (g_get_any_init_locked): New inline function, does the locking.
1064 Make use of these two throughout the code.
1066 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1068 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1069 the child process and then simply return. This makes
1070 The "S" option work as documented in g_on_error_query().
1071 (#313125, Matthew F. Barnes)
1073 * glib/gunicode.h: Update the link to Unicode category
1074 values. (#313369, Behnam Esfahbod)
1076 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1077 a little. (#311727, Tristan van Berkom)
1079 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1080 check work on ia64 too, where some symbols ended up
1081 in yet another section.