1 2006-01-06 Matthias Clasen <mclasen@redhat.com>
3 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
4 that we have posix_memalign(). (#323937, Bogdan Nicula)
6 2006-01-05 Matthias Clasen <mclasen@redhat.com>
8 * configure.in: Bump version
10 * === Released 2.9.2 ===
12 2006-01-05 Hans Breuer <hans@breuer.org>
14 * glib/makefile.msc.in : added gslice.obj
16 2006-01-05 Matthias Clasen <mclasen@redhat.com>
20 * glib/gmain.c: Clarify the documentation of source ids
21 a bit. (#325874, Dan Williams)
23 * configure.in: Fix another strict aliasing problem.
25 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
28 2006-01-04 Manish Singh <yosh@gimp.org>
31 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
32 either, since they have the same issue as g_rmdir. (Related to
35 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
37 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
38 order when creating magazines, so we hand out chunks with ascending
41 2006-01-04 Abel Cheung <maddog@linuxhall.org>
43 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
45 2006-01-03 Martyn Russell <martyn@imendio.com>
47 * docs/reference/glib/glib-sections.txt:
49 * glib/gthreadpool.[ch]:
50 - Added new API g_thread_pool_get_idle_time() and
51 g_thread_pool_set_idle_time(). (#324228).
53 * tests/threadpool-test.c:
54 - Updated test case to do thread pool sorting, thread pool with
55 no sorting and a thread pool with idle thread timeouts.
57 2006-01-03 Matthias Clasen <mclasen@redhat.com>
59 * glib/gmain.h: Add new functions here, too.
61 * glib/glib.symbols: Add new functions.
63 2005-12-20 Michael Meeks <michael.meeks@novell.com>
65 * glib/gmain.c (g_main_context_is_owner): new method
66 to determine if the current thread is the owner of the
69 2006-01-02 Matthias Clasen <mclasen@redhat.com>
73 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
74 expanding to rmdir, since rmdir is not declared in a portable
75 system header we can include in gstdio.h. (#325249, Jani Monoses)
77 2006-01-01 Tor Lillqvist <tml@novell.com>
79 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
80 stdout and stderr to unbuffered. Otherwise the giochannel layer
81 will try to read from them regardless whether the
82 g_io_channel_win32_poll() call here has indicated
83 readability or not. (#325310)
85 2005-12-29 Matthias Clasen <mclasen@redhat.com>
87 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
88 in the docs. (#325273, Declan Naughton)
90 2005-12-27 Manish Singh <yosh@gimp.org>
92 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
93 make sure we really override things.
95 2005-12-27 Matthias Clasen <mclasen@redhat.com>
97 Fix #316221, Michal Benes, Stanislav Brabec;
99 * configure.in: Fix a strict aliasing problem in
100 g_static_mutex_get_mutex().
101 * glib/gthread.h: ...and in
102 g_static_mutex_get_mutex_impl_shortcut().
104 * glib/gdatasetprivate.h: Add a cast to silence compiler
105 warnings. (#321978, Andrew Paprocki)
107 Partial fix for bug #323937, Bogdan Nicula.
109 * configure.in: Check for malloc.h
110 * glib/gslice.c: Don't include malloc.h unconditionally.
112 2005-12-27 Manish Singh <yosh@gimp.org>
114 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
115 (#324950, Dan Yefimov)
117 2005-12-26 Matthias Clasen <mclasen@redhat.com>
119 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
122 2005-12-25 Matthias Clasen <mclasen@redhat.com>
124 * AUTHORS: Update my email
126 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
128 2005-12-21 Matthias Clasen <mclasen@redhat.com>
130 * README.in: Add some notes on when 'make check' may fail.
132 2005-12-20 Matthias Clasen <mclasen@redhat.com>
135 * glib/gthreadpool.h:
136 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
137 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
139 * tests/threadpool-test.c: Test this.
141 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
143 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
144 handle errors without depending on gmessage.c which might not be
145 setup when the error occours.
146 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
147 always freeing can be achieved by adjusting the working set time to
148 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
149 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
150 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
151 minimize wastage if small amounts of differently sized structrues are
152 allocated, this does come at a performance cost of roughly 5% though.
153 fixed up block alignment calculation, so it works for varying
154 block sizes. only use strerror() not g_strerror() since the latter
155 depends on working GQuark and GSlice.
156 mem_error(): implemented in terms of fprintf and vfprintf.
158 * tests/slice-color.c: new program to test cache colorization effects.
160 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
161 G_SLICE_CONFIG_WORKING_SET_MSECS.
163 2005-12-17 Matthias Clasen <mclasen@redhat.com>
165 * glib/goption.c (parse_short_option): Set an error in all
166 failure cases. (#324332, Tim-Philipp Müller)
168 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
170 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
171 and g_atomic_pointer_get. Add g_atomic_int_set and
172 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
173 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
174 as functions (additionally to the macros in the header) for the
175 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
177 2005-12-16 Matthias Clasen <mclasen@redhat.com>
179 * glib/gmem.c (g_allocator_new): Don't return a pointer to
180 a const struct, since apps expect to be able to modify it.
181 (#324179, J. Ali Harlow)
183 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
185 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
186 for systems where the initialization of atomic variables requires a
187 write memory barrier.
189 2005-12-09 Matthias Clasen <mclasen@redhat.com>
191 * configure.in: Bump version
193 * === Released 2.9.1 ===
195 2005-12-09 Alexander Larsson <alexl@redhat.com>
199 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
202 2005-12-08 Matthias Clasen <mclasen@redhat.com>
206 2005-12-07 Martyn Russell <martyn@imendio.com>
208 * glib/gasyncqueue.c:
209 - Call g_queue_insert_sorted() instead of duplicating the code.
210 - Call g_queue_sort() instead of duplicating the code.
211 - Invert sort function results to make sure the same sort function
212 gives the same results across glist, gslist, gqueue and
215 * tests/asyncqueue-test.c:
216 - Updated the sort function to reflect the example in the
217 documentation for gasyncqueue.c.
219 2005-12-07 Martyn Russell <martyn@imendio.com>
221 * docs/reference/glib/glib-sections.txt:
222 * docs/reference/glib/tmpl/linked_lists_double.sgml:
223 * docs/reference/glib/tmpl/linked_lists_single.sgml:
226 - Added g_list_insert_sorted_with_data () and
227 g_slist_insert_sorted_with_data ().
228 - Removed the extra check in g_list_sort() and g_slist_sort() for
229 GCompareDataFunc vs. GCompareFunc.
231 2005-12-07 Tor Lillqvist <tml@novell.com>
233 * glib-gettextize.in: Look up prefix at run-time on Win32,
234 assuming the standard directory structure with glib-gettextize in
237 2005-12-06 Matthias Clasen <mclasen@redhat.com>
239 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
240 we need GTK+ 2.8 to compile against GLib 2.10.
242 2005-12-05 Manish Singh <yosh@gimp.org>
244 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
245 to override any user settings.
247 2005-12-05 Matthias Clasen <mclasen@redhat.com>
249 * configure.in: Revert an accidential commit
251 2005-12-05 Matthias Clasen <mclasen@redhat.com>
253 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
255 * tests/Makefile.am (TESTS_ENVIRONMENT):
256 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
257 MALLOC_CHECK_ and MALLOC_PERTURB_
259 * tests/run-collate-tests.sh: Run the collation tests explicitly
260 in en_US locale. (#320463)
262 * glib/gmem.h: Really deprecate GMemChunk.
269 * glib/gstrfuncs.c: Add versioned deprecation docs.
271 2005-12-05 Martyn Russell <martyn@imendio.com>
273 * docs/reference/glib/glib-sections.txt:
274 * glib/gasyncqueue.[ch]:
275 - Added support for sorting async queues by with _push_sorted(),
276 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
279 * tests/asyncqueue-test.c:
280 - Added test case for gasyncqueue.c
282 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
284 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
285 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
286 in bug 323178. moved time consuming logic from g_slice_free() out of
287 the inner loop, so g_slice_free_chain_with_offset() provides a real
288 performance benefit over g_slice_free1() now.
290 * glib/gslice.h: renamed g_slice_free_chain() to
291 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
292 a type-safe macro as suggested in bug 323178.
293 simplified the macro implementation of g_slice_free() and implemented
294 it in a type safe manner for all compliers as suggested by Morten
295 Welinder <mortenw@gnome.org>.
300 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
302 2005-12-05 Matthias Clasen <mclasen@redhat.com>
304 * glib/gasyncqueue.c: Add some docs.
306 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
308 * glib/gatomic.c: In the ia64 implementation, use
309 __sync builtin without _si or _di suffix. (#321229,
310 Stanislav Brabec, patch by Andreas Schwab)
312 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
314 * glib/gslice.h: Remove comma at the end of enum.
316 2005-12-04 Matthias Clasen <mclasen@redhat.com>
318 Handle multiple user names with the same UID better.
319 (#319535, Laszlo Peter)
321 * glib/gutils.c (g_get_any_init_do): When determining user
322 data, first look up $LOGNAME. If the UID doesn't match
323 getuid(), fall back to the current behaviour of looking
324 up the user data based on getuid().
326 2005-12-04 Matthias Clasen <mclasen@redhat.com>
328 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
329 to make gcc warn if a function result is ignored. (#145466,
330 Arjan van de Ven, Alex Larsson)
332 * glib/gmem.h: Add the new attribute to g_realloc and
335 2005-12-03 Matthias Clasen <mclasen@redhat.com>
338 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
341 * glib/gmappedfile.c: Make mapping of empty files
344 * glib/gfileutils.c: Don't fork a new process just to
345 fix the permissions of the created temp file. (#321318,
346 Alexis S. L. Carvalho)
348 2005-12-02 Matthias Clasen <mclasen@redhat.com>
350 * README.in: Add a note about Solaris threads.
353 * configure.in: Remove support for Solaris threads.
354 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
356 2005-12-02 Matthias Clasen <mclasen@redhat.com>
358 * configure.in: Define G_GUINT64_CONSTANT in analogy to
359 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
361 2005-12-02 Matthias Clasen <mclasen@redhat.com>
363 * glib/gslice.c: Win32 portability fixes and C99ism removal,
364 pointed out by Kazuki Iwamoto. (#323052)
366 2005-12-02 Matthias Clasen <mclasen@redhat.com>
368 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
369 the IA64 implementation.
371 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
373 * glib/gslice.c: some naming and type size fixups.
375 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
377 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
378 variant into something less verbose, but digestible for gcc-3.4.
380 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
382 * tests/slice-test.c: extended to perform the benchmarking on the old
383 memchunk code if 'O' is selected.
385 * tests/memchunks.c: new file which contains the old GLib mem chunks
386 implementation with prefix old_mem_chunk_.
388 * tests/Makefile.am: added memchunks.c
390 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
392 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
393 acceptable load distribution of the syscall.
394 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
395 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
397 2005-12-01 Matthias Clasen <mclasen@redhat.com>
399 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
400 double-free base interface members.
402 * tests/Makefile.am: Remove duplicate variable.
404 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
406 * glib/gslice.[hc]: new slice allocator implementation.
408 * tests/slice-test.c: added random slice allocation test.
410 * glib/gthread.[hc]: removed newly added private thread mem API.
412 * glib/gthreadinit.h:
415 * glib/gmem.c: divided glib threading initialisation into three phases,
416 initialisation where private keys and messaging are not available (only
417 needed by gmem.c), initialisation without messaging but private keys
418 available (gslice.c, gmessage.c), and full fledged initialisers that
419 server the rest of glib. initialisation functions got renamed to reflect
420 the limitations of their corresponding phases.
422 * glib/gmem.c: removed memchunk code, defer allocations to
425 * glib/gmem.[hc]: removed g_slice_* skeletons.
427 * glib/glib.symbols: added g_slice_* symbols.
429 * configure.in: check for availability of posix_memalign(3), memalign(3)
432 * glib/Makefile.am: added gslice.[hc].
434 2005-12-01 Tor Lillqvist <tml@novell.com>
436 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
437 trailing slash(es) for non-root folders. stat() fails if non-root
438 folders are specified with trailing slashes. It's too much hassle
439 to demand that callers strip such slashes themselves, especially
440 as it is easy to get it wrong and strip the slash of a root
442 (g_rename): On NT-based Windows, use MoveFileEx() with
443 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
445 2005-11-28 Matthias Clasen <mclasen@redhat.com>
447 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
450 * configure.in: Check whether do { } while (0) works.
452 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
453 G_STMT_END if it works.
455 2005-11-28 Matthias Clasen <mclasen@redhat.com>
457 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
458 * glib/gnode.c (g_node_depth_traverse_level):
459 * glib/gmem.c (g_allocator_new):
460 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
461 (g_hash_table_foreach_remove_or_steal): Silence compiler
464 2005-11-27 Matthias Clasen <mclasen@redhat.com>
466 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
467 Don't read beyond len. (#322520, Christian Persch)
469 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
471 * glib/gdataset.c: access datalist flags via atomic pointer access
472 functions, instead of acquiring the dataset lock. this is faster and
473 also matches the atomic pointer readouts in gdatalistprivate.h and
474 direct pointer modifications required by gobject.c.
476 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
478 * glib/gdataset.c: streamlined initialization code somewhat,
479 removed GData node cache.
481 2005-11-22 Matthias Clasen <mclasen@redhat.com>
483 * glib/ghash.c (g_hash_table_ref):
484 (g_hash_table_unref): Mark these as new API.
486 Avoid double locking in g_intern_string (#322133,
489 * glib/gdataset.c (g_quark_from_string_internal): New
490 internal function which factors out common parts of
491 g_quark_from[_static]_string.
492 (g_quark_from_string, g_quark_from_static_string):
493 Use g_quark_from_string_internal.
494 (g_intern_string, g_intern_static_string): Use
495 g_quark_from_string_internal, and only take the
497 (g_quark_new): Don't store the strings shifted by -1
498 in the g_quarks array.
499 (g_quark_to_string): Adapt to the previous change.
501 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
505 g_hash_table_new_full(): create hash tables with a ref count of 1.
506 g_hash_table_ref(): atomically ref_count+=1
507 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
508 when refcount reaches 0.
509 g_hash_table_destroy(): just destroy keys and values, unref by 1.
510 g_hash_table_insert():
511 g_hash_table_replace(): assert ref_count>0.
514 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
517 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
519 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
520 variables. (#321984, Andrew Paprocki)
522 2005-11-18 Matthias Clasen <mclasen@redhat.com>
524 * configure.in: Bump version
526 2005-11-17 Matthias Clasen <mclasen@redhat.com>
528 * === Released 2.9.0 ===
530 * glib/glib.symbols: Updates
532 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
533 hints to the docs. (Paolo Borelli)
537 * glib/gthread.c (g_thread_foreach): Mark as new api.
539 * README.in: Updates.
543 * glib/gdate.c (g_date_set_time_t):
544 (g_date_set_time_val): Add functions to set a date from
545 a time_t and from a GTimeVal.
546 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
549 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
550 and g_utf8_offset_to_pointer.
552 * glib/gutf8.c (g_utf8_pointer_to_offset)
553 (g_utf8_offset_to_pointer): Handle negative offsets, and use
554 "stutter stepping" for going backwards. (#320638, Larry
559 * glib/gthread.c: const correctness fixes, found
560 by Arjan van de Ven and gcc.
562 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
564 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
565 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
567 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
569 * glib/guniprop.c: Use bit hacks instead when checking a general
570 category value against multiple values.
572 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
574 2005-11-10 Simos Xenitellis <simos@gnome.org>
576 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
578 2005-11-05 Tor Lillqvist <tml@novell.com>
580 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
581 (#320688) Instead set error and return failure.
583 2005-11-07 Matthias Clasen <mclasen@redhat.com>
585 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
586 while holding the context lock. (#320886, Andy Wingo)
588 2005-11-07 Billy Biggs <vektor@dumbterm.net>
590 * tests/refcount/Makefile.am: Add a missing reference to libglib
591 in the LDADD for this test.
593 2005-11-07 Matthias Clasen <mclasen@redhat.com>
595 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
596 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
597 GLib 2.9. (#320755, Luca Ferretti)
599 2005-11-04 Matthias Clasen <mclasen@redhat.com>
601 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
602 use a faster lookup table for g_unichar_get_mirror_char().
603 * glib/gmirroringtable.h: The new table.
604 * glib/gunichartables.h: Remove bidi_mirroring_table.
605 * glib/gen-unicode-tables.pl: Don't generate the mirroring
607 * glib/glib-mirroring-tab/*: Sources for the program
608 which generated gmirroringtable.h.
610 * glib/glist.c: Avoid some code duplication.
612 * glib/gscanner.h: Include gdataset.h. (#320322)
614 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
615 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
616 to help find critical warnings in applications. (#320017,
619 2005-11-02 Tor Lillqvist <tml@novell.com>
621 * glib/glib.symbols: Remove large amount of trailing whitespace
622 from one line. Remove #ifdef G_UNDEFINED from around
623 g_hash_table_new and g_hash_table_new_full symbols, what was the
624 purpose of that? Add the g_slice functions.
626 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
628 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
629 added g_slice_*() API to allocate and cache small bits of memory.
630 an actuall allocator implementation for g_slice_*() is still pending.
632 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
633 changed GRealThread list to use in-structure *next; fields instead
634 of GSList, in order for thread iteration to not depenend on g_slice_*()
636 _g_thread_mem_private_get():
637 _g_thread_mem_private_set(): added accessors for private memory,
638 needed because the ordinary GPrivate implementation relies on GArray
639 and GSList and therefore indirectly on working g_slice_*() allocations.
642 g_thread_foreach(): new public API function to loop over all existing threads.
656 removed GAllocator and free list usages and accompanying locks.
657 use g_slice_*() API to allocate and cache small bits of memory.
659 * glib/ghook.h: removed GMemChunk field from public API.
662 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
664 * glib/gnode.h: deprecate allocator API.
666 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
667 be used for poll rec lists.
669 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
670 g_relation_destroy(): free all tuples from the all_tuples hash table,
671 this effectively maintains the life time track keeping of tuples.
672 g_relation_delete_tuple(): free tuples which are removed from the
673 all_tuples hash table. this fixes a temporary leak that was present
674 in the memchunk code until the destruction of the relation.
676 2005-10-29 Matthias Clasen <mclasen@redhat.com>
678 * tests/convert-test.c: Add some tests for conversions between
679 UTF-8, UCS-4 and UTF-16.
681 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
682 of len == -1, noticed by Morten Welinder.
684 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
686 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
688 2005-10-26 Matthias Clasen <mclasen@redhat.com>
690 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
691 in the error case, and add some documentation. (#319806, Morten
694 2005-10-19 Manish Singh <yosh@gimp.org>
696 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
697 config.h symbol automatically. Fixes bug #313731.
699 2005-10-19 Tor Lillqvist <tml@novell.com>
701 * glib/gwin32.c (g_win32_get_package_installation_directory):
702 Return a g_strdup()ed copy of the value stored in the hash table,
703 so that it can be g_free()d without leaving a dangling pointer in
704 the hash table. (#319232)
706 2005-10-06 Matthias Clasen <mclasen@redhat.com>
708 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
709 all-zero sequences correctly. (#317930, Sebastien Bacher)
711 2005-10-05 Matthias Clasen <mclasen@redhat.com>
713 * glib/gmarkup.c (g_markup_escape_text): Doc addition
715 * Makefile.am: Add Collation tests.
717 * tests/collate/*: Inputs and expected outputs for collation tests.
719 * tests/run-collate-tests.sh: Script to run collation tests.
721 * tests/unicode-collate.c (main): Rework slightly to make
722 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
724 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
726 * docs/reference/glib/tmpl/unicode.sgml:
727 * glib/gen-unicode-tables.pl:
729 * glib/gunichartables.h:
731 * tests/casefold.txt:
732 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
733 GUnicodeBreakType types. That may break some applications, like
736 2005-09-26 Matthias Clasen <mclasen@redhat.com>
738 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
739 docs. (#314393, Matthew F. Barnes)
742 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
743 they are already declared in glib.h. This doesn't break documented
744 use of gprintf.h, but should probably be pointed out in the
745 release notes for 2.10. (#314232, Behdad Esfahbod)
747 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
749 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
750 complexity optimization, based on a patch from Matthias Clasen.
752 * tests/patterntest.c: more tests, mostly from matthias.
754 2005-09-20 Matthias Clasen <mclasen@redhat.com>
756 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
757 (#316703, Mark Drago)
759 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
761 * glib/gpattern.c: applied a patch from matthias which checks on the
762 upper bound of GPatternSpec length to optimize matches.
765 * tests/patterntest.c: added more match cases.
767 2005-09-16 Tor Lillqvist <tml@novell.com>
769 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
770 warning to doc comment that these functions should not be used on
771 encodings like CP932.
773 2005-09-14 Matthias Clasen <mclasen@redhat.com>
775 * tests/keyfile-test.c: Add a test for grup names of length 1.
777 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
778 of length 1. (#316309)
780 2005-09-12 Matthias Clasen <mclasen@redhat.com>
782 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
784 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
786 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
787 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
790 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
792 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
793 (child_watch_helper_thread): Remove some dead code. Closes
796 2005-09-07 Tor Lillqvist <tml@novell.com>
798 * glib/Makefile.am: Create also a console version of the
799 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
800 It's otherwise identical to gspawn-win32-helper.exe, except marked
801 as a console application (linked without the -mwindows option).
803 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
804 Drop the dont_wait parameter. Its truth value correlated 100% with
805 the NULLness of the exit_status parameter anyway, so it's enough
806 to check whether exit_status is NULL. Invert the sense of the
807 dont_return_handle parameter and rename it to do_return_handle, to
808 make the code easier to read by avoiding double negations.
810 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
811 to do_spawn_with_pipes() accordingly.
813 (do_spawn_with_pipes): If we have a console, use the console
814 version of the helper program, otherwise use the GUI one. This
815 avoids extra console windows opening up in some situations. (In
816 case a console application uses the GUI gspawn-win32-helper.exe to
817 spawn another console application we would get a separate console
818 for the spawned console application).
820 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
822 2005-09-05 Matthias Clasen <mclasen@redhat.com>
824 * glib/gmappedfile.c (g_mapped_file_new): Report an error
825 if the file is too large. (#315275, Kjartan Maraas)
827 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
828 of read() is signed. (#315273, Kjartan Maraas)
830 2005-08-31 Tor Lillqvist <tml@novell.com>
832 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
833 the EXE and each DLL have their own atexit function chains.
835 #define g_atexit as atexit instead. This means it has a
836 better chance of doing what the caller wants. For instance,
837 gtkhtml calls g_atexit() registering a function in gtkhtml
838 itself. This caused a crash when g_atexit() was implemented as a
839 function in the GLib DLL. The gtkhtml DLL was already unloaded by
840 the time the GLib DLL got unloaded.
842 * glib/gutils.c: #undef the #define mentioned above, to also get a
843 real g_atexit() into the DLL for backward compatibility. Document
844 the Windows behaviour of g_atexit(), and document the varying ways
845 atexit() can behave in the context of dynamically loaded modules
848 2005-08-31 Matthias Clasen <mclasen@redhat.com>
852 * glib/gdataset.c: Add string interning functions.
854 2005-08-28 Matthias Clasen <mclasen@redhat.com>
856 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
859 2005-08-26 Matthias Clasen <mclasen@redhat.com>
861 * configure.in: Bump version to 2.9.0
863 2005-08-25 Tor Lillqvist <tml@novell.com>
865 Make also the g_spawn*() functions take parameters in the GLib
866 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
867 API or ABI. Like the other GLib API that was earlier changed to
868 use UTF-8 on Windows, the names of the functions that take UTF-8
869 have _utf8 suffixes added by using preprocessor macros in the
870 header file. The old names are kept for functions with the old
871 behaviour, taking parameters in the system codepage, for DLL ABI
874 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
875 g_spawn*() functions.
877 * glib/gspawn-win32.c: Use wide-char API on NT-based
878 Windows. Convert parameters from UTF-8 to wide chars (NT) or
879 system codepage (Win9x) and call the C library _wspawn*() or
880 spawn*() functions respectvely. Add DLL ABI stability versions
881 that take parameters in the system codepage.
883 * glib/gspawn-win32-helper.c: On NT-based Windows use the
884 wide-char versions of argv and envp, and use wide-char API to
885 change directory and spawn the program to run. Remove the verbose
886 debugging output, it was too complex to modify for the wide-char
887 features. (Just add temporary debugging printouts if needed, no
888 need to have them permanently in the source.)
890 * glib/gspawn.c: Corresponding documentation updates.
892 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
893 symbols as PRIVATE, add the new _utf8-suffixed ones.
895 2005-08-24 Stepan Kasal <kasal@ucw.cz>
897 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
898 guint64 otherwise and that can produce warnings about comparison
899 between signed and unsigned.
901 2005-08-23 Matthias Clasen <mclasen@redhat.com>
903 * glib/gutils.c: Fix the crt_externs.h include.
905 2005-08-23 Stepan Kasal <kasal@ucw.cz>
907 * NEWS: Fix spelling of my first name.
909 2005-08-23 Matthias Clasen <mclasen@redhat.com>
913 * === Released 2.8.1 ===
917 2005-08-20 Hans Breuer <hans@breuer.org>
919 * glib/makefile.msc.in : link with ws2_32.lib
921 2005-08-18 Tor Lillqvist <tml@novell.com>
923 * configure.in: Check for <sys/wait.h>
925 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
927 2005-08-18 Ross Burton <ross@burtonini.com>
930 Optimise single-character insertions.
935 * tests/string-test.c:
936 Add tests for new optimisation, and fix a leak.
938 2005-08-17 Matthias Clasen <mclasen@redhat.com>
940 * configure.in: Check for crt_externs.h and _NSGetEnviron.
942 * glib/gutils.c: On Darwin, include crt-externs.h and
943 define environ using _NSGetEnviron(). (#313731)
945 2005-08-16 Stepan Kasal <kasal@ucw.cz>
947 * glib/gutils.c (g_get_any_init): Move the body of the big if...
948 (g_get_any_init_do): ... to this new function.
949 (g_get_any_init): Declare as inline.
950 (g_get_any_init_locked): New inline function, does the locking.
951 Make use of these two throughout the code.
953 2005-08-15 Matthias Clasen <mclasen@redhat.com>
955 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
956 the child process and then simply return. This makes
957 The "S" option work as documented in g_on_error_query().
958 (#313125, Matthew F. Barnes)
960 * glib/gunicode.h: Update the link to Unicode category
961 values. (#313369, Behnam Esfahbod)
963 * glib/gqueue.c (g_queue_find_custom): Clarify docs
964 a little. (#311727, Tristan van Berkom)
966 * glib/abicheck.sh, gobject/abicheck.sh: Make the
967 check work on ia64 too, where some symbols ended up
968 in yet another section.