1 2006-01-16 Matthias Clasen <mclasen@redhat.com>
3 * configure.in: Bump version
5 * === Released 2.9.3 ===
9 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
10 Rework the strict aliasing fix to not break C++, pointed
11 out by Murray Cumming.
13 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
14 Signal waiting threads, problem noticed by Christian Kellner.
16 2006-01-16 Matthias Clasen <mclasen@redhat.com>
18 Fix bug 326747, Alberto Ruiz:
20 * glib/gconvert.c (make_valid_utf8): Change this function to
21 replace unknown characters by the Unicode replacement character
22 instead of '?', and don't append "(invalid encoding)".
23 (g_filename_display_name, g_filename_display_basename): Document
24 how to determine if the filename was in an invalid encoding.
26 2006-01-14 Matthias Clasen <mclasen@redhat.com>
28 * glib/gtree.c: Replace the simple recursive implementation
29 by a nonrecursive, threaded implementation by Maurizio
32 2006-01-12 Matthias Clasen <mclasen@redhat.com>
34 * tests/asyncqueue-test.c (main): Fix the
35 assert to not always trigger. (#326558,
38 2006-01-11 Matthias Clasen <mclasen@redhat.com>
40 * tests/convert-test.c: Don't test items_read and
41 items_written in error cases where it is not set.
43 * tests/asyncqueue-test.c: Set error to NULL.
45 2006-01-08 Matthias Clasen <mclasen@redhat.com>
47 * glib/gtree.c: Break some long lines.
49 * tests/tree-test.c: Add more tests.
51 * glib/gslice.c: Include config.h first, pointed out
54 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
56 * docs/reference/gobject/tmpl/gtype.sgml: explain that
57 G_TYPE_INSTANCE_GET_CLASS() does behave different during
60 2006-01-06 Matthias Clasen <mclasen@redhat.com>
62 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
63 that we have posix_memalign(). (#323937, Bogdan Nicula)
65 2006-01-05 Matthias Clasen <mclasen@redhat.com>
67 * configure.in: Bump version
69 * === Released 2.9.2 ===
71 2006-01-05 Hans Breuer <hans@breuer.org>
73 * glib/makefile.msc.in : added gslice.obj
75 2006-01-05 Matthias Clasen <mclasen@redhat.com>
79 * glib/gmain.c: Clarify the documentation of source ids
80 a bit. (#325874, Dan Williams)
82 * configure.in: Fix another strict aliasing problem.
84 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
87 2006-01-04 Manish Singh <yosh@gimp.org>
90 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
91 either, since they have the same issue as g_rmdir. (Related to
94 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
96 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
97 order when creating magazines, so we hand out chunks with ascending
100 2006-01-04 Abel Cheung <maddog@linuxhall.org>
102 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
104 2006-01-03 Martyn Russell <martyn@imendio.com>
106 * docs/reference/glib/glib-sections.txt:
108 * glib/gthreadpool.[ch]:
109 - Added new API g_thread_pool_get_idle_time() and
110 g_thread_pool_set_idle_time(). (#324228).
112 * tests/threadpool-test.c:
113 - Updated test case to do thread pool sorting, thread pool with
114 no sorting and a thread pool with idle thread timeouts.
116 2006-01-03 Matthias Clasen <mclasen@redhat.com>
118 * glib/gmain.h: Add new functions here, too.
120 * glib/glib.symbols: Add new functions.
122 2005-12-20 Michael Meeks <michael.meeks@novell.com>
124 * glib/gmain.c (g_main_context_is_owner): new method
125 to determine if the current thread is the owner of the
128 2006-01-02 Matthias Clasen <mclasen@redhat.com>
132 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
133 expanding to rmdir, since rmdir is not declared in a portable
134 system header we can include in gstdio.h. (#325249, Jani Monoses)
136 2006-01-01 Tor Lillqvist <tml@novell.com>
138 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
139 stdout and stderr to unbuffered. Otherwise the giochannel layer
140 will try to read from them regardless whether the
141 g_io_channel_win32_poll() call here has indicated
142 readability or not. (#325310)
144 2005-12-29 Matthias Clasen <mclasen@redhat.com>
146 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
147 in the docs. (#325273, Declan Naughton)
149 2005-12-27 Manish Singh <yosh@gimp.org>
151 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
152 make sure we really override things.
154 2005-12-27 Matthias Clasen <mclasen@redhat.com>
156 Fix #316221, Michal Benes, Stanislav Brabec;
158 * configure.in: Fix a strict aliasing problem in
159 g_static_mutex_get_mutex().
160 * glib/gthread.h: ...and in
161 g_static_mutex_get_mutex_impl_shortcut().
163 * glib/gdatasetprivate.h: Add a cast to silence compiler
164 warnings. (#321978, Andrew Paprocki)
166 Partial fix for bug #323937, Bogdan Nicula.
168 * configure.in: Check for malloc.h
169 * glib/gslice.c: Don't include malloc.h unconditionally.
171 2005-12-27 Manish Singh <yosh@gimp.org>
173 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
174 (#324950, Dan Yefimov)
176 2005-12-26 Matthias Clasen <mclasen@redhat.com>
178 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
181 2005-12-25 Matthias Clasen <mclasen@redhat.com>
183 * AUTHORS: Update my email
185 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
187 2005-12-21 Matthias Clasen <mclasen@redhat.com>
189 * README.in: Add some notes on when 'make check' may fail.
191 2005-12-20 Matthias Clasen <mclasen@redhat.com>
194 * glib/gthreadpool.h:
195 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
196 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
198 * tests/threadpool-test.c: Test this.
200 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
202 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
203 handle errors without depending on gmessage.c which might not be
204 setup when the error occours.
205 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
206 always freeing can be achieved by adjusting the working set time to
207 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
208 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
209 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
210 minimize wastage if small amounts of differently sized structrues are
211 allocated, this does come at a performance cost of roughly 5% though.
212 fixed up block alignment calculation, so it works for varying
213 block sizes. only use strerror() not g_strerror() since the latter
214 depends on working GQuark and GSlice.
215 mem_error(): implemented in terms of fprintf and vfprintf.
217 * tests/slice-color.c: new program to test cache colorization effects.
219 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
220 G_SLICE_CONFIG_WORKING_SET_MSECS.
222 2005-12-17 Matthias Clasen <mclasen@redhat.com>
224 * glib/goption.c (parse_short_option): Set an error in all
225 failure cases. (#324332, Tim-Philipp Müller)
227 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
229 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
230 and g_atomic_pointer_get. Add g_atomic_int_set and
231 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
232 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
233 as functions (additionally to the macros in the header) for the
234 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
236 2005-12-16 Matthias Clasen <mclasen@redhat.com>
238 * glib/gmem.c (g_allocator_new): Don't return a pointer to
239 a const struct, since apps expect to be able to modify it.
240 (#324179, J. Ali Harlow)
242 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
244 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
245 for systems where the initialization of atomic variables requires a
246 write memory barrier.
248 2005-12-09 Matthias Clasen <mclasen@redhat.com>
250 * configure.in: Bump version
252 * === Released 2.9.1 ===
254 2005-12-09 Alexander Larsson <alexl@redhat.com>
258 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
261 2005-12-08 Matthias Clasen <mclasen@redhat.com>
265 2005-12-07 Martyn Russell <martyn@imendio.com>
267 * glib/gasyncqueue.c:
268 - Call g_queue_insert_sorted() instead of duplicating the code.
269 - Call g_queue_sort() instead of duplicating the code.
270 - Invert sort function results to make sure the same sort function
271 gives the same results across glist, gslist, gqueue and
274 * tests/asyncqueue-test.c:
275 - Updated the sort function to reflect the example in the
276 documentation for gasyncqueue.c.
278 2005-12-07 Martyn Russell <martyn@imendio.com>
280 * docs/reference/glib/glib-sections.txt:
281 * docs/reference/glib/tmpl/linked_lists_double.sgml:
282 * docs/reference/glib/tmpl/linked_lists_single.sgml:
285 - Added g_list_insert_sorted_with_data () and
286 g_slist_insert_sorted_with_data ().
287 - Removed the extra check in g_list_sort() and g_slist_sort() for
288 GCompareDataFunc vs. GCompareFunc.
290 2005-12-07 Tor Lillqvist <tml@novell.com>
292 * glib-gettextize.in: Look up prefix at run-time on Win32,
293 assuming the standard directory structure with glib-gettextize in
296 2005-12-06 Matthias Clasen <mclasen@redhat.com>
298 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
299 we need GTK+ 2.8 to compile against GLib 2.10.
301 2005-12-05 Manish Singh <yosh@gimp.org>
303 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
304 to override any user settings.
306 2005-12-05 Matthias Clasen <mclasen@redhat.com>
308 * configure.in: Revert an accidential commit
310 2005-12-05 Matthias Clasen <mclasen@redhat.com>
312 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
314 * tests/Makefile.am (TESTS_ENVIRONMENT):
315 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
316 MALLOC_CHECK_ and MALLOC_PERTURB_
318 * tests/run-collate-tests.sh: Run the collation tests explicitly
319 in en_US locale. (#320463)
321 * glib/gmem.h: Really deprecate GMemChunk.
328 * glib/gstrfuncs.c: Add versioned deprecation docs.
330 2005-12-05 Martyn Russell <martyn@imendio.com>
332 * docs/reference/glib/glib-sections.txt:
333 * glib/gasyncqueue.[ch]:
334 - Added support for sorting async queues by with _push_sorted(),
335 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
338 * tests/asyncqueue-test.c:
339 - Added test case for gasyncqueue.c
341 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
343 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
344 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
345 in bug 323178. moved time consuming logic from g_slice_free() out of
346 the inner loop, so g_slice_free_chain_with_offset() provides a real
347 performance benefit over g_slice_free1() now.
349 * glib/gslice.h: renamed g_slice_free_chain() to
350 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
351 a type-safe macro as suggested in bug 323178.
352 simplified the macro implementation of g_slice_free() and implemented
353 it in a type safe manner for all compliers as suggested by Morten
354 Welinder <mortenw@gnome.org>.
359 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
361 2005-12-05 Matthias Clasen <mclasen@redhat.com>
363 * glib/gasyncqueue.c: Add some docs.
365 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
367 * glib/gatomic.c: In the ia64 implementation, use
368 __sync builtin without _si or _di suffix. (#321229,
369 Stanislav Brabec, patch by Andreas Schwab)
371 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
373 * glib/gslice.h: Remove comma at the end of enum.
375 2005-12-04 Matthias Clasen <mclasen@redhat.com>
377 Handle multiple user names with the same UID better.
378 (#319535, Laszlo Peter)
380 * glib/gutils.c (g_get_any_init_do): When determining user
381 data, first look up $LOGNAME. If the UID doesn't match
382 getuid(), fall back to the current behaviour of looking
383 up the user data based on getuid().
385 2005-12-04 Matthias Clasen <mclasen@redhat.com>
387 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
388 to make gcc warn if a function result is ignored. (#145466,
389 Arjan van de Ven, Alex Larsson)
391 * glib/gmem.h: Add the new attribute to g_realloc and
394 2005-12-03 Matthias Clasen <mclasen@redhat.com>
397 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
400 * glib/gmappedfile.c: Make mapping of empty files
403 * glib/gfileutils.c: Don't fork a new process just to
404 fix the permissions of the created temp file. (#321318,
405 Alexis S. L. Carvalho)
407 2005-12-02 Matthias Clasen <mclasen@redhat.com>
409 * README.in: Add a note about Solaris threads.
412 * configure.in: Remove support for Solaris threads.
413 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
415 2005-12-02 Matthias Clasen <mclasen@redhat.com>
417 * configure.in: Define G_GUINT64_CONSTANT in analogy to
418 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
420 2005-12-02 Matthias Clasen <mclasen@redhat.com>
422 * glib/gslice.c: Win32 portability fixes and C99ism removal,
423 pointed out by Kazuki Iwamoto. (#323052)
425 2005-12-02 Matthias Clasen <mclasen@redhat.com>
427 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
428 the IA64 implementation.
430 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
432 * glib/gslice.c: some naming and type size fixups.
434 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
436 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
437 variant into something less verbose, but digestible for gcc-3.4.
439 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
441 * tests/slice-test.c: extended to perform the benchmarking on the old
442 memchunk code if 'O' is selected.
444 * tests/memchunks.c: new file which contains the old GLib mem chunks
445 implementation with prefix old_mem_chunk_.
447 * tests/Makefile.am: added memchunks.c
449 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
451 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
452 acceptable load distribution of the syscall.
453 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
454 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
456 2005-12-01 Matthias Clasen <mclasen@redhat.com>
458 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
459 double-free base interface members.
461 * tests/Makefile.am: Remove duplicate variable.
463 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
465 * glib/gslice.[hc]: new slice allocator implementation.
467 * tests/slice-test.c: added random slice allocation test.
469 * glib/gthread.[hc]: removed newly added private thread mem API.
471 * glib/gthreadinit.h:
474 * glib/gmem.c: divided glib threading initialisation into three phases,
475 initialisation where private keys and messaging are not available (only
476 needed by gmem.c), initialisation without messaging but private keys
477 available (gslice.c, gmessage.c), and full fledged initialisers that
478 server the rest of glib. initialisation functions got renamed to reflect
479 the limitations of their corresponding phases.
481 * glib/gmem.c: removed memchunk code, defer allocations to
484 * glib/gmem.[hc]: removed g_slice_* skeletons.
486 * glib/glib.symbols: added g_slice_* symbols.
488 * configure.in: check for availability of posix_memalign(3), memalign(3)
491 * glib/Makefile.am: added gslice.[hc].
493 2005-12-01 Tor Lillqvist <tml@novell.com>
495 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
496 trailing slash(es) for non-root folders. stat() fails if non-root
497 folders are specified with trailing slashes. It's too much hassle
498 to demand that callers strip such slashes themselves, especially
499 as it is easy to get it wrong and strip the slash of a root
501 (g_rename): On NT-based Windows, use MoveFileEx() with
502 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
504 2005-11-28 Matthias Clasen <mclasen@redhat.com>
506 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
509 * configure.in: Check whether do { } while (0) works.
511 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
512 G_STMT_END if it works.
514 2005-11-28 Matthias Clasen <mclasen@redhat.com>
516 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
517 * glib/gnode.c (g_node_depth_traverse_level):
518 * glib/gmem.c (g_allocator_new):
519 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
520 (g_hash_table_foreach_remove_or_steal): Silence compiler
523 2005-11-27 Matthias Clasen <mclasen@redhat.com>
525 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
526 Don't read beyond len. (#322520, Christian Persch)
528 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
530 * glib/gdataset.c: access datalist flags via atomic pointer access
531 functions, instead of acquiring the dataset lock. this is faster and
532 also matches the atomic pointer readouts in gdatalistprivate.h and
533 direct pointer modifications required by gobject.c.
535 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
537 * glib/gdataset.c: streamlined initialization code somewhat,
538 removed GData node cache.
540 2005-11-22 Matthias Clasen <mclasen@redhat.com>
542 * glib/ghash.c (g_hash_table_ref):
543 (g_hash_table_unref): Mark these as new API.
545 Avoid double locking in g_intern_string (#322133,
548 * glib/gdataset.c (g_quark_from_string_internal): New
549 internal function which factors out common parts of
550 g_quark_from[_static]_string.
551 (g_quark_from_string, g_quark_from_static_string):
552 Use g_quark_from_string_internal.
553 (g_intern_string, g_intern_static_string): Use
554 g_quark_from_string_internal, and only take the
556 (g_quark_new): Don't store the strings shifted by -1
557 in the g_quarks array.
558 (g_quark_to_string): Adapt to the previous change.
560 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
564 g_hash_table_new_full(): create hash tables with a ref count of 1.
565 g_hash_table_ref(): atomically ref_count+=1
566 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
567 when refcount reaches 0.
568 g_hash_table_destroy(): just destroy keys and values, unref by 1.
569 g_hash_table_insert():
570 g_hash_table_replace(): assert ref_count>0.
573 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
576 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
578 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
579 variables. (#321984, Andrew Paprocki)
581 2005-11-18 Matthias Clasen <mclasen@redhat.com>
583 * configure.in: Bump version
585 2005-11-17 Matthias Clasen <mclasen@redhat.com>
587 * === Released 2.9.0 ===
589 * glib/glib.symbols: Updates
591 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
592 hints to the docs. (Paolo Borelli)
596 * glib/gthread.c (g_thread_foreach): Mark as new api.
598 * README.in: Updates.
602 * glib/gdate.c (g_date_set_time_t):
603 (g_date_set_time_val): Add functions to set a date from
604 a time_t and from a GTimeVal.
605 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
608 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
609 and g_utf8_offset_to_pointer.
611 * glib/gutf8.c (g_utf8_pointer_to_offset)
612 (g_utf8_offset_to_pointer): Handle negative offsets, and use
613 "stutter stepping" for going backwards. (#320638, Larry
618 * glib/gthread.c: const correctness fixes, found
619 by Arjan van de Ven and gcc.
621 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
623 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
624 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
626 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
628 * glib/guniprop.c: Use bit hacks instead when checking a general
629 category value against multiple values.
631 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
633 2005-11-10 Simos Xenitellis <simos@gnome.org>
635 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
637 2005-11-05 Tor Lillqvist <tml@novell.com>
639 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
640 (#320688) Instead set error and return failure.
642 2005-11-07 Matthias Clasen <mclasen@redhat.com>
644 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
645 while holding the context lock. (#320886, Andy Wingo)
647 2005-11-07 Billy Biggs <vektor@dumbterm.net>
649 * tests/refcount/Makefile.am: Add a missing reference to libglib
650 in the LDADD for this test.
652 2005-11-07 Matthias Clasen <mclasen@redhat.com>
654 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
655 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
656 GLib 2.9. (#320755, Luca Ferretti)
658 2005-11-04 Matthias Clasen <mclasen@redhat.com>
660 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
661 use a faster lookup table for g_unichar_get_mirror_char().
662 * glib/gmirroringtable.h: The new table.
663 * glib/gunichartables.h: Remove bidi_mirroring_table.
664 * glib/gen-unicode-tables.pl: Don't generate the mirroring
666 * glib/glib-mirroring-tab/*: Sources for the program
667 which generated gmirroringtable.h.
669 * glib/glist.c: Avoid some code duplication.
671 * glib/gscanner.h: Include gdataset.h. (#320322)
673 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
674 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
675 to help find critical warnings in applications. (#320017,
678 2005-11-02 Tor Lillqvist <tml@novell.com>
680 * glib/glib.symbols: Remove large amount of trailing whitespace
681 from one line. Remove #ifdef G_UNDEFINED from around
682 g_hash_table_new and g_hash_table_new_full symbols, what was the
683 purpose of that? Add the g_slice functions.
685 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
687 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
688 added g_slice_*() API to allocate and cache small bits of memory.
689 an actuall allocator implementation for g_slice_*() is still pending.
691 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
692 changed GRealThread list to use in-structure *next; fields instead
693 of GSList, in order for thread iteration to not depenend on g_slice_*()
695 _g_thread_mem_private_get():
696 _g_thread_mem_private_set(): added accessors for private memory,
697 needed because the ordinary GPrivate implementation relies on GArray
698 and GSList and therefore indirectly on working g_slice_*() allocations.
701 g_thread_foreach(): new public API function to loop over all existing threads.
715 removed GAllocator and free list usages and accompanying locks.
716 use g_slice_*() API to allocate and cache small bits of memory.
718 * glib/ghook.h: removed GMemChunk field from public API.
721 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
723 * glib/gnode.h: deprecate allocator API.
725 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
726 be used for poll rec lists.
728 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
729 g_relation_destroy(): free all tuples from the all_tuples hash table,
730 this effectively maintains the life time track keeping of tuples.
731 g_relation_delete_tuple(): free tuples which are removed from the
732 all_tuples hash table. this fixes a temporary leak that was present
733 in the memchunk code until the destruction of the relation.
735 2005-10-29 Matthias Clasen <mclasen@redhat.com>
737 * tests/convert-test.c: Add some tests for conversions between
738 UTF-8, UCS-4 and UTF-16.
740 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
741 of len == -1, noticed by Morten Welinder.
743 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
745 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
747 2005-10-26 Matthias Clasen <mclasen@redhat.com>
749 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
750 in the error case, and add some documentation. (#319806, Morten
753 2005-10-19 Manish Singh <yosh@gimp.org>
755 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
756 config.h symbol automatically. Fixes bug #313731.
758 2005-10-19 Tor Lillqvist <tml@novell.com>
760 * glib/gwin32.c (g_win32_get_package_installation_directory):
761 Return a g_strdup()ed copy of the value stored in the hash table,
762 so that it can be g_free()d without leaving a dangling pointer in
763 the hash table. (#319232)
765 2005-10-06 Matthias Clasen <mclasen@redhat.com>
767 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
768 all-zero sequences correctly. (#317930, Sebastien Bacher)
770 2005-10-05 Matthias Clasen <mclasen@redhat.com>
772 * glib/gmarkup.c (g_markup_escape_text): Doc addition
774 * Makefile.am: Add Collation tests.
776 * tests/collate/*: Inputs and expected outputs for collation tests.
778 * tests/run-collate-tests.sh: Script to run collation tests.
780 * tests/unicode-collate.c (main): Rework slightly to make
781 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
783 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
785 * docs/reference/glib/tmpl/unicode.sgml:
786 * glib/gen-unicode-tables.pl:
788 * glib/gunichartables.h:
790 * tests/casefold.txt:
791 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
792 GUnicodeBreakType types. That may break some applications, like
795 2005-09-26 Matthias Clasen <mclasen@redhat.com>
797 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
798 docs. (#314393, Matthew F. Barnes)
801 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
802 they are already declared in glib.h. This doesn't break documented
803 use of gprintf.h, but should probably be pointed out in the
804 release notes for 2.10. (#314232, Behdad Esfahbod)
806 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
808 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
809 complexity optimization, based on a patch from Matthias Clasen.
811 * tests/patterntest.c: more tests, mostly from matthias.
813 2005-09-20 Matthias Clasen <mclasen@redhat.com>
815 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
816 (#316703, Mark Drago)
818 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
820 * glib/gpattern.c: applied a patch from matthias which checks on the
821 upper bound of GPatternSpec length to optimize matches.
824 * tests/patterntest.c: added more match cases.
826 2005-09-16 Tor Lillqvist <tml@novell.com>
828 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
829 warning to doc comment that these functions should not be used on
830 encodings like CP932.
832 2005-09-14 Matthias Clasen <mclasen@redhat.com>
834 * tests/keyfile-test.c: Add a test for grup names of length 1.
836 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
837 of length 1. (#316309)
839 2005-09-12 Matthias Clasen <mclasen@redhat.com>
841 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
843 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
845 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
846 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
849 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
851 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
852 (child_watch_helper_thread): Remove some dead code. Closes
855 2005-09-07 Tor Lillqvist <tml@novell.com>
857 * glib/Makefile.am: Create also a console version of the
858 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
859 It's otherwise identical to gspawn-win32-helper.exe, except marked
860 as a console application (linked without the -mwindows option).
862 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
863 Drop the dont_wait parameter. Its truth value correlated 100% with
864 the NULLness of the exit_status parameter anyway, so it's enough
865 to check whether exit_status is NULL. Invert the sense of the
866 dont_return_handle parameter and rename it to do_return_handle, to
867 make the code easier to read by avoiding double negations.
869 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
870 to do_spawn_with_pipes() accordingly.
872 (do_spawn_with_pipes): If we have a console, use the console
873 version of the helper program, otherwise use the GUI one. This
874 avoids extra console windows opening up in some situations. (In
875 case a console application uses the GUI gspawn-win32-helper.exe to
876 spawn another console application we would get a separate console
877 for the spawned console application).
879 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
881 2005-09-05 Matthias Clasen <mclasen@redhat.com>
883 * glib/gmappedfile.c (g_mapped_file_new): Report an error
884 if the file is too large. (#315275, Kjartan Maraas)
886 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
887 of read() is signed. (#315273, Kjartan Maraas)
889 2005-08-31 Tor Lillqvist <tml@novell.com>
891 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
892 the EXE and each DLL have their own atexit function chains.
894 #define g_atexit as atexit instead. This means it has a
895 better chance of doing what the caller wants. For instance,
896 gtkhtml calls g_atexit() registering a function in gtkhtml
897 itself. This caused a crash when g_atexit() was implemented as a
898 function in the GLib DLL. The gtkhtml DLL was already unloaded by
899 the time the GLib DLL got unloaded.
901 * glib/gutils.c: #undef the #define mentioned above, to also get a
902 real g_atexit() into the DLL for backward compatibility. Document
903 the Windows behaviour of g_atexit(), and document the varying ways
904 atexit() can behave in the context of dynamically loaded modules
907 2005-08-31 Matthias Clasen <mclasen@redhat.com>
911 * glib/gdataset.c: Add string interning functions.
913 2005-08-28 Matthias Clasen <mclasen@redhat.com>
915 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
918 2005-08-26 Matthias Clasen <mclasen@redhat.com>
920 * configure.in: Bump version to 2.9.0
922 2005-08-25 Tor Lillqvist <tml@novell.com>
924 Make also the g_spawn*() functions take parameters in the GLib
925 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
926 API or ABI. Like the other GLib API that was earlier changed to
927 use UTF-8 on Windows, the names of the functions that take UTF-8
928 have _utf8 suffixes added by using preprocessor macros in the
929 header file. The old names are kept for functions with the old
930 behaviour, taking parameters in the system codepage, for DLL ABI
933 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
934 g_spawn*() functions.
936 * glib/gspawn-win32.c: Use wide-char API on NT-based
937 Windows. Convert parameters from UTF-8 to wide chars (NT) or
938 system codepage (Win9x) and call the C library _wspawn*() or
939 spawn*() functions respectvely. Add DLL ABI stability versions
940 that take parameters in the system codepage.
942 * glib/gspawn-win32-helper.c: On NT-based Windows use the
943 wide-char versions of argv and envp, and use wide-char API to
944 change directory and spawn the program to run. Remove the verbose
945 debugging output, it was too complex to modify for the wide-char
946 features. (Just add temporary debugging printouts if needed, no
947 need to have them permanently in the source.)
949 * glib/gspawn.c: Corresponding documentation updates.
951 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
952 symbols as PRIVATE, add the new _utf8-suffixed ones.
954 2005-08-24 Stepan Kasal <kasal@ucw.cz>
956 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
957 guint64 otherwise and that can produce warnings about comparison
958 between signed and unsigned.
960 2005-08-23 Matthias Clasen <mclasen@redhat.com>
962 * glib/gutils.c: Fix the crt_externs.h include.
964 2005-08-23 Stepan Kasal <kasal@ucw.cz>
966 * NEWS: Fix spelling of my first name.
968 2005-08-23 Matthias Clasen <mclasen@redhat.com>
972 * === Released 2.8.1 ===
976 2005-08-20 Hans Breuer <hans@breuer.org>
978 * glib/makefile.msc.in : link with ws2_32.lib
980 2005-08-18 Tor Lillqvist <tml@novell.com>
982 * configure.in: Check for <sys/wait.h>
984 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
986 2005-08-18 Ross Burton <ross@burtonini.com>
989 Optimise single-character insertions.
994 * tests/string-test.c:
995 Add tests for new optimisation, and fix a leak.
997 2005-08-17 Matthias Clasen <mclasen@redhat.com>
999 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1001 * glib/gutils.c: On Darwin, include crt-externs.h and
1002 define environ using _NSGetEnviron(). (#313731)
1004 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1006 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1007 (g_get_any_init_do): ... to this new function.
1008 (g_get_any_init): Declare as inline.
1009 (g_get_any_init_locked): New inline function, does the locking.
1010 Make use of these two throughout the code.
1012 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1014 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1015 the child process and then simply return. This makes
1016 The "S" option work as documented in g_on_error_query().
1017 (#313125, Matthew F. Barnes)
1019 * glib/gunicode.h: Update the link to Unicode category
1020 values. (#313369, Behnam Esfahbod)
1022 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1023 a little. (#311727, Tristan van Berkom)
1025 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1026 check work on ia64 too, where some symbols ended up
1027 in yet another section.