1 2006-01-16 Matthias Clasen <mclasen@redhat.com>
3 * === Released 2.9.3 ===
7 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
8 Rework the strict aliasing fix to not break C++, pointed
11 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
12 Signal waiting threads, problem noticed by Christian Kellner.
14 2006-01-16 Matthias Clasen <mclasen@redhat.com>
16 Fix bug 326747, Alberto Ruiz:
18 * glib/gconvert.c (make_valid_utf8): Change this function to
19 replace unknown characters by the Unicode replacement character
20 instead of '?', and don't append "(invalid encoding)".
21 (g_filename_display_name, g_filename_display_basename): Document
22 how to determine if the filename was in an invalid encoding.
24 2006-01-14 Matthias Clasen <mclasen@redhat.com>
26 * glib/gtree.c: Replace the simple recursive implementation
27 by a nonrecursive, threaded implementation by Maurizio
30 2006-01-12 Matthias Clasen <mclasen@redhat.com>
32 * tests/asyncqueue-test.c (main): Fix the
33 assert to not always trigger. (#326558,
36 2006-01-11 Matthias Clasen <mclasen@redhat.com>
38 * tests/convert-test.c: Don't test items_read and
39 items_written in error cases where it is not set.
41 * tests/asyncqueue-test.c: Set error to NULL.
43 2006-01-08 Matthias Clasen <mclasen@redhat.com>
45 * glib/gtree.c: Break some long lines.
47 * tests/tree-test.c: Add more tests.
49 * glib/gslice.c: Include config.h first, pointed out
52 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
54 * docs/reference/gobject/tmpl/gtype.sgml: explain that
55 G_TYPE_INSTANCE_GET_CLASS() does behave different during
58 2006-01-06 Matthias Clasen <mclasen@redhat.com>
60 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
61 that we have posix_memalign(). (#323937, Bogdan Nicula)
63 2006-01-05 Matthias Clasen <mclasen@redhat.com>
65 * configure.in: Bump version
67 * === Released 2.9.2 ===
69 2006-01-05 Hans Breuer <hans@breuer.org>
71 * glib/makefile.msc.in : added gslice.obj
73 2006-01-05 Matthias Clasen <mclasen@redhat.com>
77 * glib/gmain.c: Clarify the documentation of source ids
78 a bit. (#325874, Dan Williams)
80 * configure.in: Fix another strict aliasing problem.
82 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
85 2006-01-04 Manish Singh <yosh@gimp.org>
88 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
89 either, since they have the same issue as g_rmdir. (Related to
92 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
94 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
95 order when creating magazines, so we hand out chunks with ascending
98 2006-01-04 Abel Cheung <maddog@linuxhall.org>
100 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
102 2006-01-03 Martyn Russell <martyn@imendio.com>
104 * docs/reference/glib/glib-sections.txt:
106 * glib/gthreadpool.[ch]:
107 - Added new API g_thread_pool_get_idle_time() and
108 g_thread_pool_set_idle_time(). (#324228).
110 * tests/threadpool-test.c:
111 - Updated test case to do thread pool sorting, thread pool with
112 no sorting and a thread pool with idle thread timeouts.
114 2006-01-03 Matthias Clasen <mclasen@redhat.com>
116 * glib/gmain.h: Add new functions here, too.
118 * glib/glib.symbols: Add new functions.
120 2005-12-20 Michael Meeks <michael.meeks@novell.com>
122 * glib/gmain.c (g_main_context_is_owner): new method
123 to determine if the current thread is the owner of the
126 2006-01-02 Matthias Clasen <mclasen@redhat.com>
130 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
131 expanding to rmdir, since rmdir is not declared in a portable
132 system header we can include in gstdio.h. (#325249, Jani Monoses)
134 2006-01-01 Tor Lillqvist <tml@novell.com>
136 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
137 stdout and stderr to unbuffered. Otherwise the giochannel layer
138 will try to read from them regardless whether the
139 g_io_channel_win32_poll() call here has indicated
140 readability or not. (#325310)
142 2005-12-29 Matthias Clasen <mclasen@redhat.com>
144 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
145 in the docs. (#325273, Declan Naughton)
147 2005-12-27 Manish Singh <yosh@gimp.org>
149 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
150 make sure we really override things.
152 2005-12-27 Matthias Clasen <mclasen@redhat.com>
154 Fix #316221, Michal Benes, Stanislav Brabec;
156 * configure.in: Fix a strict aliasing problem in
157 g_static_mutex_get_mutex().
158 * glib/gthread.h: ...and in
159 g_static_mutex_get_mutex_impl_shortcut().
161 * glib/gdatasetprivate.h: Add a cast to silence compiler
162 warnings. (#321978, Andrew Paprocki)
164 Partial fix for bug #323937, Bogdan Nicula.
166 * configure.in: Check for malloc.h
167 * glib/gslice.c: Don't include malloc.h unconditionally.
169 2005-12-27 Manish Singh <yosh@gimp.org>
171 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
172 (#324950, Dan Yefimov)
174 2005-12-26 Matthias Clasen <mclasen@redhat.com>
176 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
179 2005-12-25 Matthias Clasen <mclasen@redhat.com>
181 * AUTHORS: Update my email
183 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
185 2005-12-21 Matthias Clasen <mclasen@redhat.com>
187 * README.in: Add some notes on when 'make check' may fail.
189 2005-12-20 Matthias Clasen <mclasen@redhat.com>
192 * glib/gthreadpool.h:
193 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
194 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
196 * tests/threadpool-test.c: Test this.
198 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
200 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
201 handle errors without depending on gmessage.c which might not be
202 setup when the error occours.
203 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
204 always freeing can be achieved by adjusting the working set time to
205 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
206 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
207 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
208 minimize wastage if small amounts of differently sized structrues are
209 allocated, this does come at a performance cost of roughly 5% though.
210 fixed up block alignment calculation, so it works for varying
211 block sizes. only use strerror() not g_strerror() since the latter
212 depends on working GQuark and GSlice.
213 mem_error(): implemented in terms of fprintf and vfprintf.
215 * tests/slice-color.c: new program to test cache colorization effects.
217 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
218 G_SLICE_CONFIG_WORKING_SET_MSECS.
220 2005-12-17 Matthias Clasen <mclasen@redhat.com>
222 * glib/goption.c (parse_short_option): Set an error in all
223 failure cases. (#324332, Tim-Philipp Müller)
225 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
227 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
228 and g_atomic_pointer_get. Add g_atomic_int_set and
229 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
230 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
231 as functions (additionally to the macros in the header) for the
232 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
234 2005-12-16 Matthias Clasen <mclasen@redhat.com>
236 * glib/gmem.c (g_allocator_new): Don't return a pointer to
237 a const struct, since apps expect to be able to modify it.
238 (#324179, J. Ali Harlow)
240 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
242 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
243 for systems where the initialization of atomic variables requires a
244 write memory barrier.
246 2005-12-09 Matthias Clasen <mclasen@redhat.com>
248 * configure.in: Bump version
250 * === Released 2.9.1 ===
252 2005-12-09 Alexander Larsson <alexl@redhat.com>
256 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
259 2005-12-08 Matthias Clasen <mclasen@redhat.com>
263 2005-12-07 Martyn Russell <martyn@imendio.com>
265 * glib/gasyncqueue.c:
266 - Call g_queue_insert_sorted() instead of duplicating the code.
267 - Call g_queue_sort() instead of duplicating the code.
268 - Invert sort function results to make sure the same sort function
269 gives the same results across glist, gslist, gqueue and
272 * tests/asyncqueue-test.c:
273 - Updated the sort function to reflect the example in the
274 documentation for gasyncqueue.c.
276 2005-12-07 Martyn Russell <martyn@imendio.com>
278 * docs/reference/glib/glib-sections.txt:
279 * docs/reference/glib/tmpl/linked_lists_double.sgml:
280 * docs/reference/glib/tmpl/linked_lists_single.sgml:
283 - Added g_list_insert_sorted_with_data () and
284 g_slist_insert_sorted_with_data ().
285 - Removed the extra check in g_list_sort() and g_slist_sort() for
286 GCompareDataFunc vs. GCompareFunc.
288 2005-12-07 Tor Lillqvist <tml@novell.com>
290 * glib-gettextize.in: Look up prefix at run-time on Win32,
291 assuming the standard directory structure with glib-gettextize in
294 2005-12-06 Matthias Clasen <mclasen@redhat.com>
296 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
297 we need GTK+ 2.8 to compile against GLib 2.10.
299 2005-12-05 Manish Singh <yosh@gimp.org>
301 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
302 to override any user settings.
304 2005-12-05 Matthias Clasen <mclasen@redhat.com>
306 * configure.in: Revert an accidential commit
308 2005-12-05 Matthias Clasen <mclasen@redhat.com>
310 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
312 * tests/Makefile.am (TESTS_ENVIRONMENT):
313 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
314 MALLOC_CHECK_ and MALLOC_PERTURB_
316 * tests/run-collate-tests.sh: Run the collation tests explicitly
317 in en_US locale. (#320463)
319 * glib/gmem.h: Really deprecate GMemChunk.
326 * glib/gstrfuncs.c: Add versioned deprecation docs.
328 2005-12-05 Martyn Russell <martyn@imendio.com>
330 * docs/reference/glib/glib-sections.txt:
331 * glib/gasyncqueue.[ch]:
332 - Added support for sorting async queues by with _push_sorted(),
333 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
336 * tests/asyncqueue-test.c:
337 - Added test case for gasyncqueue.c
339 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
341 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
342 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
343 in bug 323178. moved time consuming logic from g_slice_free() out of
344 the inner loop, so g_slice_free_chain_with_offset() provides a real
345 performance benefit over g_slice_free1() now.
347 * glib/gslice.h: renamed g_slice_free_chain() to
348 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
349 a type-safe macro as suggested in bug 323178.
350 simplified the macro implementation of g_slice_free() and implemented
351 it in a type safe manner for all compliers as suggested by Morten
352 Welinder <mortenw@gnome.org>.
357 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
359 2005-12-05 Matthias Clasen <mclasen@redhat.com>
361 * glib/gasyncqueue.c: Add some docs.
363 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
365 * glib/gatomic.c: In the ia64 implementation, use
366 __sync builtin without _si or _di suffix. (#321229,
367 Stanislav Brabec, patch by Andreas Schwab)
369 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
371 * glib/gslice.h: Remove comma at the end of enum.
373 2005-12-04 Matthias Clasen <mclasen@redhat.com>
375 Handle multiple user names with the same UID better.
376 (#319535, Laszlo Peter)
378 * glib/gutils.c (g_get_any_init_do): When determining user
379 data, first look up $LOGNAME. If the UID doesn't match
380 getuid(), fall back to the current behaviour of looking
381 up the user data based on getuid().
383 2005-12-04 Matthias Clasen <mclasen@redhat.com>
385 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
386 to make gcc warn if a function result is ignored. (#145466,
387 Arjan van de Ven, Alex Larsson)
389 * glib/gmem.h: Add the new attribute to g_realloc and
392 2005-12-03 Matthias Clasen <mclasen@redhat.com>
395 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
398 * glib/gmappedfile.c: Make mapping of empty files
401 * glib/gfileutils.c: Don't fork a new process just to
402 fix the permissions of the created temp file. (#321318,
403 Alexis S. L. Carvalho)
405 2005-12-02 Matthias Clasen <mclasen@redhat.com>
407 * README.in: Add a note about Solaris threads.
410 * configure.in: Remove support for Solaris threads.
411 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
413 2005-12-02 Matthias Clasen <mclasen@redhat.com>
415 * configure.in: Define G_GUINT64_CONSTANT in analogy to
416 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
418 2005-12-02 Matthias Clasen <mclasen@redhat.com>
420 * glib/gslice.c: Win32 portability fixes and C99ism removal,
421 pointed out by Kazuki Iwamoto. (#323052)
423 2005-12-02 Matthias Clasen <mclasen@redhat.com>
425 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
426 the IA64 implementation.
428 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
430 * glib/gslice.c: some naming and type size fixups.
432 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
434 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
435 variant into something less verbose, but digestible for gcc-3.4.
437 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
439 * tests/slice-test.c: extended to perform the benchmarking on the old
440 memchunk code if 'O' is selected.
442 * tests/memchunks.c: new file which contains the old GLib mem chunks
443 implementation with prefix old_mem_chunk_.
445 * tests/Makefile.am: added memchunks.c
447 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
449 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
450 acceptable load distribution of the syscall.
451 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
452 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
454 2005-12-01 Matthias Clasen <mclasen@redhat.com>
456 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
457 double-free base interface members.
459 * tests/Makefile.am: Remove duplicate variable.
461 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
463 * glib/gslice.[hc]: new slice allocator implementation.
465 * tests/slice-test.c: added random slice allocation test.
467 * glib/gthread.[hc]: removed newly added private thread mem API.
469 * glib/gthreadinit.h:
472 * glib/gmem.c: divided glib threading initialisation into three phases,
473 initialisation where private keys and messaging are not available (only
474 needed by gmem.c), initialisation without messaging but private keys
475 available (gslice.c, gmessage.c), and full fledged initialisers that
476 server the rest of glib. initialisation functions got renamed to reflect
477 the limitations of their corresponding phases.
479 * glib/gmem.c: removed memchunk code, defer allocations to
482 * glib/gmem.[hc]: removed g_slice_* skeletons.
484 * glib/glib.symbols: added g_slice_* symbols.
486 * configure.in: check for availability of posix_memalign(3), memalign(3)
489 * glib/Makefile.am: added gslice.[hc].
491 2005-12-01 Tor Lillqvist <tml@novell.com>
493 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
494 trailing slash(es) for non-root folders. stat() fails if non-root
495 folders are specified with trailing slashes. It's too much hassle
496 to demand that callers strip such slashes themselves, especially
497 as it is easy to get it wrong and strip the slash of a root
499 (g_rename): On NT-based Windows, use MoveFileEx() with
500 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
502 2005-11-28 Matthias Clasen <mclasen@redhat.com>
504 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
507 * configure.in: Check whether do { } while (0) works.
509 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
510 G_STMT_END if it works.
512 2005-11-28 Matthias Clasen <mclasen@redhat.com>
514 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
515 * glib/gnode.c (g_node_depth_traverse_level):
516 * glib/gmem.c (g_allocator_new):
517 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
518 (g_hash_table_foreach_remove_or_steal): Silence compiler
521 2005-11-27 Matthias Clasen <mclasen@redhat.com>
523 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
524 Don't read beyond len. (#322520, Christian Persch)
526 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
528 * glib/gdataset.c: access datalist flags via atomic pointer access
529 functions, instead of acquiring the dataset lock. this is faster and
530 also matches the atomic pointer readouts in gdatalistprivate.h and
531 direct pointer modifications required by gobject.c.
533 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
535 * glib/gdataset.c: streamlined initialization code somewhat,
536 removed GData node cache.
538 2005-11-22 Matthias Clasen <mclasen@redhat.com>
540 * glib/ghash.c (g_hash_table_ref):
541 (g_hash_table_unref): Mark these as new API.
543 Avoid double locking in g_intern_string (#322133,
546 * glib/gdataset.c (g_quark_from_string_internal): New
547 internal function which factors out common parts of
548 g_quark_from[_static]_string.
549 (g_quark_from_string, g_quark_from_static_string):
550 Use g_quark_from_string_internal.
551 (g_intern_string, g_intern_static_string): Use
552 g_quark_from_string_internal, and only take the
554 (g_quark_new): Don't store the strings shifted by -1
555 in the g_quarks array.
556 (g_quark_to_string): Adapt to the previous change.
558 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
562 g_hash_table_new_full(): create hash tables with a ref count of 1.
563 g_hash_table_ref(): atomically ref_count+=1
564 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
565 when refcount reaches 0.
566 g_hash_table_destroy(): just destroy keys and values, unref by 1.
567 g_hash_table_insert():
568 g_hash_table_replace(): assert ref_count>0.
571 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
574 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
576 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
577 variables. (#321984, Andrew Paprocki)
579 2005-11-18 Matthias Clasen <mclasen@redhat.com>
581 * configure.in: Bump version
583 2005-11-17 Matthias Clasen <mclasen@redhat.com>
585 * === Released 2.9.0 ===
587 * glib/glib.symbols: Updates
589 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
590 hints to the docs. (Paolo Borelli)
594 * glib/gthread.c (g_thread_foreach): Mark as new api.
596 * README.in: Updates.
600 * glib/gdate.c (g_date_set_time_t):
601 (g_date_set_time_val): Add functions to set a date from
602 a time_t and from a GTimeVal.
603 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
606 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
607 and g_utf8_offset_to_pointer.
609 * glib/gutf8.c (g_utf8_pointer_to_offset)
610 (g_utf8_offset_to_pointer): Handle negative offsets, and use
611 "stutter stepping" for going backwards. (#320638, Larry
616 * glib/gthread.c: const correctness fixes, found
617 by Arjan van de Ven and gcc.
619 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
621 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
622 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
624 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
626 * glib/guniprop.c: Use bit hacks instead when checking a general
627 category value against multiple values.
629 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
631 2005-11-10 Simos Xenitellis <simos@gnome.org>
633 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
635 2005-11-05 Tor Lillqvist <tml@novell.com>
637 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
638 (#320688) Instead set error and return failure.
640 2005-11-07 Matthias Clasen <mclasen@redhat.com>
642 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
643 while holding the context lock. (#320886, Andy Wingo)
645 2005-11-07 Billy Biggs <vektor@dumbterm.net>
647 * tests/refcount/Makefile.am: Add a missing reference to libglib
648 in the LDADD for this test.
650 2005-11-07 Matthias Clasen <mclasen@redhat.com>
652 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
653 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
654 GLib 2.9. (#320755, Luca Ferretti)
656 2005-11-04 Matthias Clasen <mclasen@redhat.com>
658 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
659 use a faster lookup table for g_unichar_get_mirror_char().
660 * glib/gmirroringtable.h: The new table.
661 * glib/gunichartables.h: Remove bidi_mirroring_table.
662 * glib/gen-unicode-tables.pl: Don't generate the mirroring
664 * glib/glib-mirroring-tab/*: Sources for the program
665 which generated gmirroringtable.h.
667 * glib/glist.c: Avoid some code duplication.
669 * glib/gscanner.h: Include gdataset.h. (#320322)
671 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
672 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
673 to help find critical warnings in applications. (#320017,
676 2005-11-02 Tor Lillqvist <tml@novell.com>
678 * glib/glib.symbols: Remove large amount of trailing whitespace
679 from one line. Remove #ifdef G_UNDEFINED from around
680 g_hash_table_new and g_hash_table_new_full symbols, what was the
681 purpose of that? Add the g_slice functions.
683 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
685 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
686 added g_slice_*() API to allocate and cache small bits of memory.
687 an actuall allocator implementation for g_slice_*() is still pending.
689 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
690 changed GRealThread list to use in-structure *next; fields instead
691 of GSList, in order for thread iteration to not depenend on g_slice_*()
693 _g_thread_mem_private_get():
694 _g_thread_mem_private_set(): added accessors for private memory,
695 needed because the ordinary GPrivate implementation relies on GArray
696 and GSList and therefore indirectly on working g_slice_*() allocations.
699 g_thread_foreach(): new public API function to loop over all existing threads.
713 removed GAllocator and free list usages and accompanying locks.
714 use g_slice_*() API to allocate and cache small bits of memory.
716 * glib/ghook.h: removed GMemChunk field from public API.
719 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
721 * glib/gnode.h: deprecate allocator API.
723 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
724 be used for poll rec lists.
726 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
727 g_relation_destroy(): free all tuples from the all_tuples hash table,
728 this effectively maintains the life time track keeping of tuples.
729 g_relation_delete_tuple(): free tuples which are removed from the
730 all_tuples hash table. this fixes a temporary leak that was present
731 in the memchunk code until the destruction of the relation.
733 2005-10-29 Matthias Clasen <mclasen@redhat.com>
735 * tests/convert-test.c: Add some tests for conversions between
736 UTF-8, UCS-4 and UTF-16.
738 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
739 of len == -1, noticed by Morten Welinder.
741 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
743 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
745 2005-10-26 Matthias Clasen <mclasen@redhat.com>
747 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
748 in the error case, and add some documentation. (#319806, Morten
751 2005-10-19 Manish Singh <yosh@gimp.org>
753 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
754 config.h symbol automatically. Fixes bug #313731.
756 2005-10-19 Tor Lillqvist <tml@novell.com>
758 * glib/gwin32.c (g_win32_get_package_installation_directory):
759 Return a g_strdup()ed copy of the value stored in the hash table,
760 so that it can be g_free()d without leaving a dangling pointer in
761 the hash table. (#319232)
763 2005-10-06 Matthias Clasen <mclasen@redhat.com>
765 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
766 all-zero sequences correctly. (#317930, Sebastien Bacher)
768 2005-10-05 Matthias Clasen <mclasen@redhat.com>
770 * glib/gmarkup.c (g_markup_escape_text): Doc addition
772 * Makefile.am: Add Collation tests.
774 * tests/collate/*: Inputs and expected outputs for collation tests.
776 * tests/run-collate-tests.sh: Script to run collation tests.
778 * tests/unicode-collate.c (main): Rework slightly to make
779 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
781 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
783 * docs/reference/glib/tmpl/unicode.sgml:
784 * glib/gen-unicode-tables.pl:
786 * glib/gunichartables.h:
788 * tests/casefold.txt:
789 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
790 GUnicodeBreakType types. That may break some applications, like
793 2005-09-26 Matthias Clasen <mclasen@redhat.com>
795 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
796 docs. (#314393, Matthew F. Barnes)
799 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
800 they are already declared in glib.h. This doesn't break documented
801 use of gprintf.h, but should probably be pointed out in the
802 release notes for 2.10. (#314232, Behdad Esfahbod)
804 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
806 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
807 complexity optimization, based on a patch from Matthias Clasen.
809 * tests/patterntest.c: more tests, mostly from matthias.
811 2005-09-20 Matthias Clasen <mclasen@redhat.com>
813 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
814 (#316703, Mark Drago)
816 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
818 * glib/gpattern.c: applied a patch from matthias which checks on the
819 upper bound of GPatternSpec length to optimize matches.
822 * tests/patterntest.c: added more match cases.
824 2005-09-16 Tor Lillqvist <tml@novell.com>
826 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
827 warning to doc comment that these functions should not be used on
828 encodings like CP932.
830 2005-09-14 Matthias Clasen <mclasen@redhat.com>
832 * tests/keyfile-test.c: Add a test for grup names of length 1.
834 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
835 of length 1. (#316309)
837 2005-09-12 Matthias Clasen <mclasen@redhat.com>
839 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
841 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
843 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
844 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
847 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
849 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
850 (child_watch_helper_thread): Remove some dead code. Closes
853 2005-09-07 Tor Lillqvist <tml@novell.com>
855 * glib/Makefile.am: Create also a console version of the
856 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
857 It's otherwise identical to gspawn-win32-helper.exe, except marked
858 as a console application (linked without the -mwindows option).
860 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
861 Drop the dont_wait parameter. Its truth value correlated 100% with
862 the NULLness of the exit_status parameter anyway, so it's enough
863 to check whether exit_status is NULL. Invert the sense of the
864 dont_return_handle parameter and rename it to do_return_handle, to
865 make the code easier to read by avoiding double negations.
867 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
868 to do_spawn_with_pipes() accordingly.
870 (do_spawn_with_pipes): If we have a console, use the console
871 version of the helper program, otherwise use the GUI one. This
872 avoids extra console windows opening up in some situations. (In
873 case a console application uses the GUI gspawn-win32-helper.exe to
874 spawn another console application we would get a separate console
875 for the spawned console application).
877 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
879 2005-09-05 Matthias Clasen <mclasen@redhat.com>
881 * glib/gmappedfile.c (g_mapped_file_new): Report an error
882 if the file is too large. (#315275, Kjartan Maraas)
884 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
885 of read() is signed. (#315273, Kjartan Maraas)
887 2005-08-31 Tor Lillqvist <tml@novell.com>
889 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
890 the EXE and each DLL have their own atexit function chains.
892 #define g_atexit as atexit instead. This means it has a
893 better chance of doing what the caller wants. For instance,
894 gtkhtml calls g_atexit() registering a function in gtkhtml
895 itself. This caused a crash when g_atexit() was implemented as a
896 function in the GLib DLL. The gtkhtml DLL was already unloaded by
897 the time the GLib DLL got unloaded.
899 * glib/gutils.c: #undef the #define mentioned above, to also get a
900 real g_atexit() into the DLL for backward compatibility. Document
901 the Windows behaviour of g_atexit(), and document the varying ways
902 atexit() can behave in the context of dynamically loaded modules
905 2005-08-31 Matthias Clasen <mclasen@redhat.com>
909 * glib/gdataset.c: Add string interning functions.
911 2005-08-28 Matthias Clasen <mclasen@redhat.com>
913 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
916 2005-08-26 Matthias Clasen <mclasen@redhat.com>
918 * configure.in: Bump version to 2.9.0
920 2005-08-25 Tor Lillqvist <tml@novell.com>
922 Make also the g_spawn*() functions take parameters in the GLib
923 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
924 API or ABI. Like the other GLib API that was earlier changed to
925 use UTF-8 on Windows, the names of the functions that take UTF-8
926 have _utf8 suffixes added by using preprocessor macros in the
927 header file. The old names are kept for functions with the old
928 behaviour, taking parameters in the system codepage, for DLL ABI
931 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
932 g_spawn*() functions.
934 * glib/gspawn-win32.c: Use wide-char API on NT-based
935 Windows. Convert parameters from UTF-8 to wide chars (NT) or
936 system codepage (Win9x) and call the C library _wspawn*() or
937 spawn*() functions respectvely. Add DLL ABI stability versions
938 that take parameters in the system codepage.
940 * glib/gspawn-win32-helper.c: On NT-based Windows use the
941 wide-char versions of argv and envp, and use wide-char API to
942 change directory and spawn the program to run. Remove the verbose
943 debugging output, it was too complex to modify for the wide-char
944 features. (Just add temporary debugging printouts if needed, no
945 need to have them permanently in the source.)
947 * glib/gspawn.c: Corresponding documentation updates.
949 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
950 symbols as PRIVATE, add the new _utf8-suffixed ones.
952 2005-08-24 Stepan Kasal <kasal@ucw.cz>
954 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
955 guint64 otherwise and that can produce warnings about comparison
956 between signed and unsigned.
958 2005-08-23 Matthias Clasen <mclasen@redhat.com>
960 * glib/gutils.c: Fix the crt_externs.h include.
962 2005-08-23 Stepan Kasal <kasal@ucw.cz>
964 * NEWS: Fix spelling of my first name.
966 2005-08-23 Matthias Clasen <mclasen@redhat.com>
970 * === Released 2.8.1 ===
974 2005-08-20 Hans Breuer <hans@breuer.org>
976 * glib/makefile.msc.in : link with ws2_32.lib
978 2005-08-18 Tor Lillqvist <tml@novell.com>
980 * configure.in: Check for <sys/wait.h>
982 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
984 2005-08-18 Ross Burton <ross@burtonini.com>
987 Optimise single-character insertions.
992 * tests/string-test.c:
993 Add tests for new optimisation, and fix a leak.
995 2005-08-17 Matthias Clasen <mclasen@redhat.com>
997 * configure.in: Check for crt_externs.h and _NSGetEnviron.
999 * glib/gutils.c: On Darwin, include crt-externs.h and
1000 define environ using _NSGetEnviron(). (#313731)
1002 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1004 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1005 (g_get_any_init_do): ... to this new function.
1006 (g_get_any_init): Declare as inline.
1007 (g_get_any_init_locked): New inline function, does the locking.
1008 Make use of these two throughout the code.
1010 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1012 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1013 the child process and then simply return. This makes
1014 The "S" option work as documented in g_on_error_query().
1015 (#313125, Matthew F. Barnes)
1017 * glib/gunicode.h: Update the link to Unicode category
1018 values. (#313369, Behnam Esfahbod)
1020 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1021 a little. (#311727, Tristan van Berkom)
1023 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1024 check work on ia64 too, where some symbols ended up
1025 in yet another section.