1 2006-01-16 Matthias Clasen <mclasen@redhat.com>
3 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
4 Rework the strict aliasing fix to not break C++, pointed
7 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
8 Signal waiting threads, problem noticed by Christian Kellner.
10 2006-01-16 Matthias Clasen <mclasen@redhat.com>
12 Fix bug 326747, Alberto Ruiz:
14 * glib/gconvert.c (make_valid_utf8): Change this function to
15 replace unknown characters by the Unicode replacement character
16 instead of '?', and don't append "(invalid encoding)".
17 (g_filename_display_name, g_filename_display_basename): Document
18 how to determine if the filename was in an invalid encoding.
20 2006-01-14 Matthias Clasen <mclasen@redhat.com>
22 * glib/gtree.c: Replace the simple recursive implementation
23 by a nonrecursive, threaded implementation by Maurizio
26 2006-01-12 Matthias Clasen <mclasen@redhat.com>
28 * tests/asyncqueue-test.c (main): Fix the
29 assert to not always trigger. (#326558,
32 2006-01-11 Matthias Clasen <mclasen@redhat.com>
34 * tests/convert-test.c: Don't test items_read and
35 items_written in error cases where it is not set.
37 * tests/asyncqueue-test.c: Set error to NULL.
39 2006-01-08 Matthias Clasen <mclasen@redhat.com>
41 * glib/gtree.c: Break some long lines.
43 * tests/tree-test.c: Add more tests.
45 * glib/gslice.c: Include config.h first, pointed out
48 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
50 * docs/reference/gobject/tmpl/gtype.sgml: explain that
51 G_TYPE_INSTANCE_GET_CLASS() does behave different during
54 2006-01-06 Matthias Clasen <mclasen@redhat.com>
56 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
57 that we have posix_memalign(). (#323937, Bogdan Nicula)
59 2006-01-05 Matthias Clasen <mclasen@redhat.com>
61 * configure.in: Bump version
63 * === Released 2.9.2 ===
65 2006-01-05 Hans Breuer <hans@breuer.org>
67 * glib/makefile.msc.in : added gslice.obj
69 2006-01-05 Matthias Clasen <mclasen@redhat.com>
73 * glib/gmain.c: Clarify the documentation of source ids
74 a bit. (#325874, Dan Williams)
76 * configure.in: Fix another strict aliasing problem.
78 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
81 2006-01-04 Manish Singh <yosh@gimp.org>
84 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
85 either, since they have the same issue as g_rmdir. (Related to
88 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
90 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
91 order when creating magazines, so we hand out chunks with ascending
94 2006-01-04 Abel Cheung <maddog@linuxhall.org>
96 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
98 2006-01-03 Martyn Russell <martyn@imendio.com>
100 * docs/reference/glib/glib-sections.txt:
102 * glib/gthreadpool.[ch]:
103 - Added new API g_thread_pool_get_idle_time() and
104 g_thread_pool_set_idle_time(). (#324228).
106 * tests/threadpool-test.c:
107 - Updated test case to do thread pool sorting, thread pool with
108 no sorting and a thread pool with idle thread timeouts.
110 2006-01-03 Matthias Clasen <mclasen@redhat.com>
112 * glib/gmain.h: Add new functions here, too.
114 * glib/glib.symbols: Add new functions.
116 2005-12-20 Michael Meeks <michael.meeks@novell.com>
118 * glib/gmain.c (g_main_context_is_owner): new method
119 to determine if the current thread is the owner of the
122 2006-01-02 Matthias Clasen <mclasen@redhat.com>
126 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
127 expanding to rmdir, since rmdir is not declared in a portable
128 system header we can include in gstdio.h. (#325249, Jani Monoses)
130 2006-01-01 Tor Lillqvist <tml@novell.com>
132 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
133 stdout and stderr to unbuffered. Otherwise the giochannel layer
134 will try to read from them regardless whether the
135 g_io_channel_win32_poll() call here has indicated
136 readability or not. (#325310)
138 2005-12-29 Matthias Clasen <mclasen@redhat.com>
140 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
141 in the docs. (#325273, Declan Naughton)
143 2005-12-27 Manish Singh <yosh@gimp.org>
145 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
146 make sure we really override things.
148 2005-12-27 Matthias Clasen <mclasen@redhat.com>
150 Fix #316221, Michal Benes, Stanislav Brabec;
152 * configure.in: Fix a strict aliasing problem in
153 g_static_mutex_get_mutex().
154 * glib/gthread.h: ...and in
155 g_static_mutex_get_mutex_impl_shortcut().
157 * glib/gdatasetprivate.h: Add a cast to silence compiler
158 warnings. (#321978, Andrew Paprocki)
160 Partial fix for bug #323937, Bogdan Nicula.
162 * configure.in: Check for malloc.h
163 * glib/gslice.c: Don't include malloc.h unconditionally.
165 2005-12-27 Manish Singh <yosh@gimp.org>
167 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
168 (#324950, Dan Yefimov)
170 2005-12-26 Matthias Clasen <mclasen@redhat.com>
172 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
175 2005-12-25 Matthias Clasen <mclasen@redhat.com>
177 * AUTHORS: Update my email
179 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
181 2005-12-21 Matthias Clasen <mclasen@redhat.com>
183 * README.in: Add some notes on when 'make check' may fail.
185 2005-12-20 Matthias Clasen <mclasen@redhat.com>
188 * glib/gthreadpool.h:
189 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
190 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
192 * tests/threadpool-test.c: Test this.
194 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
196 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
197 handle errors without depending on gmessage.c which might not be
198 setup when the error occours.
199 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
200 always freeing can be achieved by adjusting the working set time to
201 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
202 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
203 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
204 minimize wastage if small amounts of differently sized structrues are
205 allocated, this does come at a performance cost of roughly 5% though.
206 fixed up block alignment calculation, so it works for varying
207 block sizes. only use strerror() not g_strerror() since the latter
208 depends on working GQuark and GSlice.
209 mem_error(): implemented in terms of fprintf and vfprintf.
211 * tests/slice-color.c: new program to test cache colorization effects.
213 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
214 G_SLICE_CONFIG_WORKING_SET_MSECS.
216 2005-12-17 Matthias Clasen <mclasen@redhat.com>
218 * glib/goption.c (parse_short_option): Set an error in all
219 failure cases. (#324332, Tim-Philipp Müller)
221 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
223 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
224 and g_atomic_pointer_get. Add g_atomic_int_set and
225 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
226 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
227 as functions (additionally to the macros in the header) for the
228 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
230 2005-12-16 Matthias Clasen <mclasen@redhat.com>
232 * glib/gmem.c (g_allocator_new): Don't return a pointer to
233 a const struct, since apps expect to be able to modify it.
234 (#324179, J. Ali Harlow)
236 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
238 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
239 for systems where the initialization of atomic variables requires a
240 write memory barrier.
242 2005-12-09 Matthias Clasen <mclasen@redhat.com>
244 * configure.in: Bump version
246 * === Released 2.9.1 ===
248 2005-12-09 Alexander Larsson <alexl@redhat.com>
252 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
255 2005-12-08 Matthias Clasen <mclasen@redhat.com>
259 2005-12-07 Martyn Russell <martyn@imendio.com>
261 * glib/gasyncqueue.c:
262 - Call g_queue_insert_sorted() instead of duplicating the code.
263 - Call g_queue_sort() instead of duplicating the code.
264 - Invert sort function results to make sure the same sort function
265 gives the same results across glist, gslist, gqueue and
268 * tests/asyncqueue-test.c:
269 - Updated the sort function to reflect the example in the
270 documentation for gasyncqueue.c.
272 2005-12-07 Martyn Russell <martyn@imendio.com>
274 * docs/reference/glib/glib-sections.txt:
275 * docs/reference/glib/tmpl/linked_lists_double.sgml:
276 * docs/reference/glib/tmpl/linked_lists_single.sgml:
279 - Added g_list_insert_sorted_with_data () and
280 g_slist_insert_sorted_with_data ().
281 - Removed the extra check in g_list_sort() and g_slist_sort() for
282 GCompareDataFunc vs. GCompareFunc.
284 2005-12-07 Tor Lillqvist <tml@novell.com>
286 * glib-gettextize.in: Look up prefix at run-time on Win32,
287 assuming the standard directory structure with glib-gettextize in
290 2005-12-06 Matthias Clasen <mclasen@redhat.com>
292 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
293 we need GTK+ 2.8 to compile against GLib 2.10.
295 2005-12-05 Manish Singh <yosh@gimp.org>
297 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
298 to override any user settings.
300 2005-12-05 Matthias Clasen <mclasen@redhat.com>
302 * configure.in: Revert an accidential commit
304 2005-12-05 Matthias Clasen <mclasen@redhat.com>
306 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
308 * tests/Makefile.am (TESTS_ENVIRONMENT):
309 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
310 MALLOC_CHECK_ and MALLOC_PERTURB_
312 * tests/run-collate-tests.sh: Run the collation tests explicitly
313 in en_US locale. (#320463)
315 * glib/gmem.h: Really deprecate GMemChunk.
322 * glib/gstrfuncs.c: Add versioned deprecation docs.
324 2005-12-05 Martyn Russell <martyn@imendio.com>
326 * docs/reference/glib/glib-sections.txt:
327 * glib/gasyncqueue.[ch]:
328 - Added support for sorting async queues by with _push_sorted(),
329 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
332 * tests/asyncqueue-test.c:
333 - Added test case for gasyncqueue.c
335 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
337 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
338 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
339 in bug 323178. moved time consuming logic from g_slice_free() out of
340 the inner loop, so g_slice_free_chain_with_offset() provides a real
341 performance benefit over g_slice_free1() now.
343 * glib/gslice.h: renamed g_slice_free_chain() to
344 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
345 a type-safe macro as suggested in bug 323178.
346 simplified the macro implementation of g_slice_free() and implemented
347 it in a type safe manner for all compliers as suggested by Morten
348 Welinder <mortenw@gnome.org>.
353 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
355 2005-12-05 Matthias Clasen <mclasen@redhat.com>
357 * glib/gasyncqueue.c: Add some docs.
359 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
361 * glib/gatomic.c: In the ia64 implementation, use
362 __sync builtin without _si or _di suffix. (#321229,
363 Stanislav Brabec, patch by Andreas Schwab)
365 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
367 * glib/gslice.h: Remove comma at the end of enum.
369 2005-12-04 Matthias Clasen <mclasen@redhat.com>
371 Handle multiple user names with the same UID better.
372 (#319535, Laszlo Peter)
374 * glib/gutils.c (g_get_any_init_do): When determining user
375 data, first look up $LOGNAME. If the UID doesn't match
376 getuid(), fall back to the current behaviour of looking
377 up the user data based on getuid().
379 2005-12-04 Matthias Clasen <mclasen@redhat.com>
381 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
382 to make gcc warn if a function result is ignored. (#145466,
383 Arjan van de Ven, Alex Larsson)
385 * glib/gmem.h: Add the new attribute to g_realloc and
388 2005-12-03 Matthias Clasen <mclasen@redhat.com>
391 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
394 * glib/gmappedfile.c: Make mapping of empty files
397 * glib/gfileutils.c: Don't fork a new process just to
398 fix the permissions of the created temp file. (#321318,
399 Alexis S. L. Carvalho)
401 2005-12-02 Matthias Clasen <mclasen@redhat.com>
403 * README.in: Add a note about Solaris threads.
406 * configure.in: Remove support for Solaris threads.
407 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
409 2005-12-02 Matthias Clasen <mclasen@redhat.com>
411 * configure.in: Define G_GUINT64_CONSTANT in analogy to
412 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
414 2005-12-02 Matthias Clasen <mclasen@redhat.com>
416 * glib/gslice.c: Win32 portability fixes and C99ism removal,
417 pointed out by Kazuki Iwamoto. (#323052)
419 2005-12-02 Matthias Clasen <mclasen@redhat.com>
421 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
422 the IA64 implementation.
424 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
426 * glib/gslice.c: some naming and type size fixups.
428 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
430 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
431 variant into something less verbose, but digestible for gcc-3.4.
433 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
435 * tests/slice-test.c: extended to perform the benchmarking on the old
436 memchunk code if 'O' is selected.
438 * tests/memchunks.c: new file which contains the old GLib mem chunks
439 implementation with prefix old_mem_chunk_.
441 * tests/Makefile.am: added memchunks.c
443 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
445 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
446 acceptable load distribution of the syscall.
447 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
448 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
450 2005-12-01 Matthias Clasen <mclasen@redhat.com>
452 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
453 double-free base interface members.
455 * tests/Makefile.am: Remove duplicate variable.
457 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
459 * glib/gslice.[hc]: new slice allocator implementation.
461 * tests/slice-test.c: added random slice allocation test.
463 * glib/gthread.[hc]: removed newly added private thread mem API.
465 * glib/gthreadinit.h:
468 * glib/gmem.c: divided glib threading initialisation into three phases,
469 initialisation where private keys and messaging are not available (only
470 needed by gmem.c), initialisation without messaging but private keys
471 available (gslice.c, gmessage.c), and full fledged initialisers that
472 server the rest of glib. initialisation functions got renamed to reflect
473 the limitations of their corresponding phases.
475 * glib/gmem.c: removed memchunk code, defer allocations to
478 * glib/gmem.[hc]: removed g_slice_* skeletons.
480 * glib/glib.symbols: added g_slice_* symbols.
482 * configure.in: check for availability of posix_memalign(3), memalign(3)
485 * glib/Makefile.am: added gslice.[hc].
487 2005-12-01 Tor Lillqvist <tml@novell.com>
489 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
490 trailing slash(es) for non-root folders. stat() fails if non-root
491 folders are specified with trailing slashes. It's too much hassle
492 to demand that callers strip such slashes themselves, especially
493 as it is easy to get it wrong and strip the slash of a root
495 (g_rename): On NT-based Windows, use MoveFileEx() with
496 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
498 2005-11-28 Matthias Clasen <mclasen@redhat.com>
500 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
503 * configure.in: Check whether do { } while (0) works.
505 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
506 G_STMT_END if it works.
508 2005-11-28 Matthias Clasen <mclasen@redhat.com>
510 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
511 * glib/gnode.c (g_node_depth_traverse_level):
512 * glib/gmem.c (g_allocator_new):
513 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
514 (g_hash_table_foreach_remove_or_steal): Silence compiler
517 2005-11-27 Matthias Clasen <mclasen@redhat.com>
519 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
520 Don't read beyond len. (#322520, Christian Persch)
522 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
524 * glib/gdataset.c: access datalist flags via atomic pointer access
525 functions, instead of acquiring the dataset lock. this is faster and
526 also matches the atomic pointer readouts in gdatalistprivate.h and
527 direct pointer modifications required by gobject.c.
529 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
531 * glib/gdataset.c: streamlined initialization code somewhat,
532 removed GData node cache.
534 2005-11-22 Matthias Clasen <mclasen@redhat.com>
536 * glib/ghash.c (g_hash_table_ref):
537 (g_hash_table_unref): Mark these as new API.
539 Avoid double locking in g_intern_string (#322133,
542 * glib/gdataset.c (g_quark_from_string_internal): New
543 internal function which factors out common parts of
544 g_quark_from[_static]_string.
545 (g_quark_from_string, g_quark_from_static_string):
546 Use g_quark_from_string_internal.
547 (g_intern_string, g_intern_static_string): Use
548 g_quark_from_string_internal, and only take the
550 (g_quark_new): Don't store the strings shifted by -1
551 in the g_quarks array.
552 (g_quark_to_string): Adapt to the previous change.
554 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
558 g_hash_table_new_full(): create hash tables with a ref count of 1.
559 g_hash_table_ref(): atomically ref_count+=1
560 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
561 when refcount reaches 0.
562 g_hash_table_destroy(): just destroy keys and values, unref by 1.
563 g_hash_table_insert():
564 g_hash_table_replace(): assert ref_count>0.
567 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
570 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
572 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
573 variables. (#321984, Andrew Paprocki)
575 2005-11-18 Matthias Clasen <mclasen@redhat.com>
577 * configure.in: Bump version
579 2005-11-17 Matthias Clasen <mclasen@redhat.com>
581 * === Released 2.9.0 ===
583 * glib/glib.symbols: Updates
585 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
586 hints to the docs. (Paolo Borelli)
590 * glib/gthread.c (g_thread_foreach): Mark as new api.
592 * README.in: Updates.
596 * glib/gdate.c (g_date_set_time_t):
597 (g_date_set_time_val): Add functions to set a date from
598 a time_t and from a GTimeVal.
599 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
602 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
603 and g_utf8_offset_to_pointer.
605 * glib/gutf8.c (g_utf8_pointer_to_offset)
606 (g_utf8_offset_to_pointer): Handle negative offsets, and use
607 "stutter stepping" for going backwards. (#320638, Larry
612 * glib/gthread.c: const correctness fixes, found
613 by Arjan van de Ven and gcc.
615 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
617 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
618 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
620 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
622 * glib/guniprop.c: Use bit hacks instead when checking a general
623 category value against multiple values.
625 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
627 2005-11-10 Simos Xenitellis <simos@gnome.org>
629 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
631 2005-11-05 Tor Lillqvist <tml@novell.com>
633 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
634 (#320688) Instead set error and return failure.
636 2005-11-07 Matthias Clasen <mclasen@redhat.com>
638 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
639 while holding the context lock. (#320886, Andy Wingo)
641 2005-11-07 Billy Biggs <vektor@dumbterm.net>
643 * tests/refcount/Makefile.am: Add a missing reference to libglib
644 in the LDADD for this test.
646 2005-11-07 Matthias Clasen <mclasen@redhat.com>
648 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
649 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
650 GLib 2.9. (#320755, Luca Ferretti)
652 2005-11-04 Matthias Clasen <mclasen@redhat.com>
654 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
655 use a faster lookup table for g_unichar_get_mirror_char().
656 * glib/gmirroringtable.h: The new table.
657 * glib/gunichartables.h: Remove bidi_mirroring_table.
658 * glib/gen-unicode-tables.pl: Don't generate the mirroring
660 * glib/glib-mirroring-tab/*: Sources for the program
661 which generated gmirroringtable.h.
663 * glib/glist.c: Avoid some code duplication.
665 * glib/gscanner.h: Include gdataset.h. (#320322)
667 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
668 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
669 to help find critical warnings in applications. (#320017,
672 2005-11-02 Tor Lillqvist <tml@novell.com>
674 * glib/glib.symbols: Remove large amount of trailing whitespace
675 from one line. Remove #ifdef G_UNDEFINED from around
676 g_hash_table_new and g_hash_table_new_full symbols, what was the
677 purpose of that? Add the g_slice functions.
679 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
681 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
682 added g_slice_*() API to allocate and cache small bits of memory.
683 an actuall allocator implementation for g_slice_*() is still pending.
685 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
686 changed GRealThread list to use in-structure *next; fields instead
687 of GSList, in order for thread iteration to not depenend on g_slice_*()
689 _g_thread_mem_private_get():
690 _g_thread_mem_private_set(): added accessors for private memory,
691 needed because the ordinary GPrivate implementation relies on GArray
692 and GSList and therefore indirectly on working g_slice_*() allocations.
695 g_thread_foreach(): new public API function to loop over all existing threads.
709 removed GAllocator and free list usages and accompanying locks.
710 use g_slice_*() API to allocate and cache small bits of memory.
712 * glib/ghook.h: removed GMemChunk field from public API.
715 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
717 * glib/gnode.h: deprecate allocator API.
719 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
720 be used for poll rec lists.
722 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
723 g_relation_destroy(): free all tuples from the all_tuples hash table,
724 this effectively maintains the life time track keeping of tuples.
725 g_relation_delete_tuple(): free tuples which are removed from the
726 all_tuples hash table. this fixes a temporary leak that was present
727 in the memchunk code until the destruction of the relation.
729 2005-10-29 Matthias Clasen <mclasen@redhat.com>
731 * tests/convert-test.c: Add some tests for conversions between
732 UTF-8, UCS-4 and UTF-16.
734 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
735 of len == -1, noticed by Morten Welinder.
737 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
739 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
741 2005-10-26 Matthias Clasen <mclasen@redhat.com>
743 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
744 in the error case, and add some documentation. (#319806, Morten
747 2005-10-19 Manish Singh <yosh@gimp.org>
749 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
750 config.h symbol automatically. Fixes bug #313731.
752 2005-10-19 Tor Lillqvist <tml@novell.com>
754 * glib/gwin32.c (g_win32_get_package_installation_directory):
755 Return a g_strdup()ed copy of the value stored in the hash table,
756 so that it can be g_free()d without leaving a dangling pointer in
757 the hash table. (#319232)
759 2005-10-06 Matthias Clasen <mclasen@redhat.com>
761 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
762 all-zero sequences correctly. (#317930, Sebastien Bacher)
764 2005-10-05 Matthias Clasen <mclasen@redhat.com>
766 * glib/gmarkup.c (g_markup_escape_text): Doc addition
768 * Makefile.am: Add Collation tests.
770 * tests/collate/*: Inputs and expected outputs for collation tests.
772 * tests/run-collate-tests.sh: Script to run collation tests.
774 * tests/unicode-collate.c (main): Rework slightly to make
775 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
777 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
779 * docs/reference/glib/tmpl/unicode.sgml:
780 * glib/gen-unicode-tables.pl:
782 * glib/gunichartables.h:
784 * tests/casefold.txt:
785 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
786 GUnicodeBreakType types. That may break some applications, like
789 2005-09-26 Matthias Clasen <mclasen@redhat.com>
791 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
792 docs. (#314393, Matthew F. Barnes)
795 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
796 they are already declared in glib.h. This doesn't break documented
797 use of gprintf.h, but should probably be pointed out in the
798 release notes for 2.10. (#314232, Behdad Esfahbod)
800 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
802 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
803 complexity optimization, based on a patch from Matthias Clasen.
805 * tests/patterntest.c: more tests, mostly from matthias.
807 2005-09-20 Matthias Clasen <mclasen@redhat.com>
809 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
810 (#316703, Mark Drago)
812 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
814 * glib/gpattern.c: applied a patch from matthias which checks on the
815 upper bound of GPatternSpec length to optimize matches.
818 * tests/patterntest.c: added more match cases.
820 2005-09-16 Tor Lillqvist <tml@novell.com>
822 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
823 warning to doc comment that these functions should not be used on
824 encodings like CP932.
826 2005-09-14 Matthias Clasen <mclasen@redhat.com>
828 * tests/keyfile-test.c: Add a test for grup names of length 1.
830 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
831 of length 1. (#316309)
833 2005-09-12 Matthias Clasen <mclasen@redhat.com>
835 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
837 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
839 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
840 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
843 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
845 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
846 (child_watch_helper_thread): Remove some dead code. Closes
849 2005-09-07 Tor Lillqvist <tml@novell.com>
851 * glib/Makefile.am: Create also a console version of the
852 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
853 It's otherwise identical to gspawn-win32-helper.exe, except marked
854 as a console application (linked without the -mwindows option).
856 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
857 Drop the dont_wait parameter. Its truth value correlated 100% with
858 the NULLness of the exit_status parameter anyway, so it's enough
859 to check whether exit_status is NULL. Invert the sense of the
860 dont_return_handle parameter and rename it to do_return_handle, to
861 make the code easier to read by avoiding double negations.
863 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
864 to do_spawn_with_pipes() accordingly.
866 (do_spawn_with_pipes): If we have a console, use the console
867 version of the helper program, otherwise use the GUI one. This
868 avoids extra console windows opening up in some situations. (In
869 case a console application uses the GUI gspawn-win32-helper.exe to
870 spawn another console application we would get a separate console
871 for the spawned console application).
873 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
875 2005-09-05 Matthias Clasen <mclasen@redhat.com>
877 * glib/gmappedfile.c (g_mapped_file_new): Report an error
878 if the file is too large. (#315275, Kjartan Maraas)
880 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
881 of read() is signed. (#315273, Kjartan Maraas)
883 2005-08-31 Tor Lillqvist <tml@novell.com>
885 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
886 the EXE and each DLL have their own atexit function chains.
888 #define g_atexit as atexit instead. This means it has a
889 better chance of doing what the caller wants. For instance,
890 gtkhtml calls g_atexit() registering a function in gtkhtml
891 itself. This caused a crash when g_atexit() was implemented as a
892 function in the GLib DLL. The gtkhtml DLL was already unloaded by
893 the time the GLib DLL got unloaded.
895 * glib/gutils.c: #undef the #define mentioned above, to also get a
896 real g_atexit() into the DLL for backward compatibility. Document
897 the Windows behaviour of g_atexit(), and document the varying ways
898 atexit() can behave in the context of dynamically loaded modules
901 2005-08-31 Matthias Clasen <mclasen@redhat.com>
905 * glib/gdataset.c: Add string interning functions.
907 2005-08-28 Matthias Clasen <mclasen@redhat.com>
909 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
912 2005-08-26 Matthias Clasen <mclasen@redhat.com>
914 * configure.in: Bump version to 2.9.0
916 2005-08-25 Tor Lillqvist <tml@novell.com>
918 Make also the g_spawn*() functions take parameters in the GLib
919 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
920 API or ABI. Like the other GLib API that was earlier changed to
921 use UTF-8 on Windows, the names of the functions that take UTF-8
922 have _utf8 suffixes added by using preprocessor macros in the
923 header file. The old names are kept for functions with the old
924 behaviour, taking parameters in the system codepage, for DLL ABI
927 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
928 g_spawn*() functions.
930 * glib/gspawn-win32.c: Use wide-char API on NT-based
931 Windows. Convert parameters from UTF-8 to wide chars (NT) or
932 system codepage (Win9x) and call the C library _wspawn*() or
933 spawn*() functions respectvely. Add DLL ABI stability versions
934 that take parameters in the system codepage.
936 * glib/gspawn-win32-helper.c: On NT-based Windows use the
937 wide-char versions of argv and envp, and use wide-char API to
938 change directory and spawn the program to run. Remove the verbose
939 debugging output, it was too complex to modify for the wide-char
940 features. (Just add temporary debugging printouts if needed, no
941 need to have them permanently in the source.)
943 * glib/gspawn.c: Corresponding documentation updates.
945 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
946 symbols as PRIVATE, add the new _utf8-suffixed ones.
948 2005-08-24 Stepan Kasal <kasal@ucw.cz>
950 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
951 guint64 otherwise and that can produce warnings about comparison
952 between signed and unsigned.
954 2005-08-23 Matthias Clasen <mclasen@redhat.com>
956 * glib/gutils.c: Fix the crt_externs.h include.
958 2005-08-23 Stepan Kasal <kasal@ucw.cz>
960 * NEWS: Fix spelling of my first name.
962 2005-08-23 Matthias Clasen <mclasen@redhat.com>
966 * === Released 2.8.1 ===
970 2005-08-20 Hans Breuer <hans@breuer.org>
972 * glib/makefile.msc.in : link with ws2_32.lib
974 2005-08-18 Tor Lillqvist <tml@novell.com>
976 * configure.in: Check for <sys/wait.h>
978 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
980 2005-08-18 Ross Burton <ross@burtonini.com>
983 Optimise single-character insertions.
988 * tests/string-test.c:
989 Add tests for new optimisation, and fix a leak.
991 2005-08-17 Matthias Clasen <mclasen@redhat.com>
993 * configure.in: Check for crt_externs.h and _NSGetEnviron.
995 * glib/gutils.c: On Darwin, include crt-externs.h and
996 define environ using _NSGetEnviron(). (#313731)
998 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1000 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1001 (g_get_any_init_do): ... to this new function.
1002 (g_get_any_init): Declare as inline.
1003 (g_get_any_init_locked): New inline function, does the locking.
1004 Make use of these two throughout the code.
1006 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1008 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1009 the child process and then simply return. This makes
1010 The "S" option work as documented in g_on_error_query().
1011 (#313125, Matthew F. Barnes)
1013 * glib/gunicode.h: Update the link to Unicode category
1014 values. (#313369, Behnam Esfahbod)
1016 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1017 a little. (#311727, Tristan van Berkom)
1019 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1020 check work on ia64 too, where some symbols ended up
1021 in yet another section.