1 2006-01-05 Matthias Clasen <mclasen@redhat.com>
3 * configure.in: Fix another strict aliasing problem.
5 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
8 2006-01-04 Manish Singh <yosh@gimp.org>
11 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
12 either, since they have the same issue as g_rmdir. (Related to
15 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
17 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
18 order when creating magazines, so we hand out chunks with ascending
21 2006-01-04 Abel Cheung <maddog@linuxhall.org>
23 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
25 2006-01-03 Martyn Russell <martyn@imendio.com>
27 * docs/reference/glib/glib-sections.txt:
29 * glib/gthreadpool.[ch]:
30 - Added new API g_thread_pool_get_idle_time() and
31 g_thread_pool_set_idle_time(). (#324228).
33 * tests/threadpool-test.c:
34 - Updated test case to do thread pool sorting, thread pool with
35 no sorting and a thread pool with idle thread timeouts.
37 2006-01-03 Matthias Clasen <mclasen@redhat.com>
39 * glib/gmain.h: Add new functions here, too.
41 * glib/glib.symbols: Add new functions.
43 2005-12-20 Michael Meeks <michael.meeks@novell.com>
45 * glib/gmain.c (g_main_context_is_owner): new method
46 to determine if the current thread is the owner of the
49 2006-01-02 Matthias Clasen <mclasen@redhat.com>
53 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
54 expanding to rmdir, since rmdir is not declared in a portable
55 system header we can include in gstdio.h. (#325249, Jani Monoses)
57 2006-01-01 Tor Lillqvist <tml@novell.com>
59 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
60 stdout and stderr to unbuffered. Otherwise the giochannel layer
61 will try to read from them regardless whether the
62 g_io_channel_win32_poll() call here has indicated
63 readability or not. (#325310)
65 2005-12-29 Matthias Clasen <mclasen@redhat.com>
67 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
68 in the docs. (#325273, Declan Naughton)
70 2005-12-27 Manish Singh <yosh@gimp.org>
72 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
73 make sure we really override things.
75 2005-12-27 Matthias Clasen <mclasen@redhat.com>
77 Fix #316221, Michal Benes, Stanislav Brabec;
79 * configure.in: Fix a strict aliasing problem in
80 g_static_mutex_get_mutex().
81 * glib/gthread.h: ...and in
82 g_static_mutex_get_mutex_impl_shortcut().
84 * glib/gdatasetprivate.h: Add a cast to silence compiler
85 warnings. (#321978, Andrew Paprocki)
87 Partial fix for bug #32937, Bogdan Nicula.
89 * configure.in: Check for malloc.h
90 * glib/gslice.c: Don't include malloc.h unconditionally.
92 2005-12-27 Manish Singh <yosh@gimp.org>
94 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
95 (#324950, Dan Yefimov)
97 2005-12-26 Matthias Clasen <mclasen@redhat.com>
99 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
102 2005-12-25 Matthias Clasen <mclasen@redhat.com>
104 * AUTHORS: Update my email
106 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
108 2005-12-21 Matthias Clasen <mclasen@redhat.com>
110 * README.in: Add some notes on when 'make check' may fail.
112 2005-12-20 Matthias Clasen <mclasen@redhat.com>
115 * glib/gthreadpool.h:
116 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
117 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
119 * tests/threadpool-test.c: Test this.
121 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
123 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
124 handle errors without depending on gmessage.c which might not be
125 setup when the error occours.
126 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
127 always freeing can be achieved by adjusting the working set time to
128 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
129 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
130 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
131 minimize wastage if small amounts of differently sized structrues are
132 allocated, this does come at a performance cost of roughly 5% though.
133 fixed up block alignment calculation, so it works for varying
134 block sizes. only use strerror() not g_strerror() since the latter
135 depends on working GQuark and GSlice.
136 mem_error(): implemented in terms of fprintf and vfprintf.
138 * tests/slice-color.c: new program to test cache colorization effects.
140 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
141 G_SLICE_CONFIG_WORKING_SET_MSECS.
143 2005-12-17 Matthias Clasen <mclasen@redhat.com>
145 * glib/goption.c (parse_short_option): Set an error in all
146 failure cases. (#324332, Tim-Philipp Müller)
148 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
150 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
151 and g_atomic_pointer_get. Add g_atomic_int_set and
152 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
153 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
154 as functions (additionally to the macros in the header) for the
155 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
157 2005-12-16 Matthias Clasen <mclasen@redhat.com>
159 * glib/gmem.c (g_allocator_new): Don't return a pointer to
160 a const struct, since apps expect to be able to modify it.
161 (#324179, J. Ali Harlow)
163 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
165 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
166 for systems where the initialization of atomic variables requires a
167 write memory barrier.
169 2005-12-09 Matthias Clasen <mclasen@redhat.com>
171 * configure.in: Bump version
173 * === Released 2.9.1 ===
175 2005-12-09 Alexander Larsson <alexl@redhat.com>
179 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
182 2005-12-08 Matthias Clasen <mclasen@redhat.com>
186 2005-12-07 Martyn Russell <martyn@imendio.com>
188 * glib/gasyncqueue.c:
189 - Call g_queue_insert_sorted() instead of duplicating the code.
190 - Call g_queue_sort() instead of duplicating the code.
191 - Invert sort function results to make sure the same sort function
192 gives the same results across glist, gslist, gqueue and
195 * tests/asyncqueue-test.c:
196 - Updated the sort function to reflect the example in the
197 documentation for gasyncqueue.c.
199 2005-12-07 Martyn Russell <martyn@imendio.com>
201 * docs/reference/glib/glib-sections.txt:
202 * docs/reference/glib/tmpl/linked_lists_double.sgml:
203 * docs/reference/glib/tmpl/linked_lists_single.sgml:
206 - Added g_list_insert_sorted_with_data () and
207 g_slist_insert_sorted_with_data ().
208 - Removed the extra check in g_list_sort() and g_slist_sort() for
209 GCompareDataFunc vs. GCompareFunc.
211 2005-12-07 Tor Lillqvist <tml@novell.com>
213 * glib-gettextize.in: Look up prefix at run-time on Win32,
214 assuming the standard directory structure with glib-gettextize in
217 2005-12-06 Matthias Clasen <mclasen@redhat.com>
219 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
220 we need GTK+ 2.8 to compile against GLib 2.10.
222 2005-12-05 Manish Singh <yosh@gimp.org>
224 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
225 to override any user settings.
227 2005-12-05 Matthias Clasen <mclasen@redhat.com>
229 * configure.in: Revert an accidential commit
231 2005-12-05 Matthias Clasen <mclasen@redhat.com>
233 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
235 * tests/Makefile.am (TESTS_ENVIRONMENT):
236 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
237 MALLOC_CHECK_ and MALLOC_PERTURB_
239 * tests/run-collate-tests.sh: Run the collation tests explicitly
240 in en_US locale. (#320463)
242 * glib/gmem.h: Really deprecate GMemChunk.
249 * glib/gstrfuncs.c: Add versioned deprecation docs.
251 2005-12-05 Martyn Russell <martyn@imendio.com>
253 * docs/reference/glib/glib-sections.txt:
254 * glib/gasyncqueue.[ch]:
255 - Added support for sorting async queues by with _push_sorted(),
256 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
259 * tests/asyncqueue-test.c:
260 - Added test case for gasyncqueue.c
262 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
264 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
265 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
266 in bug 323178. moved time consuming logic from g_slice_free() out of
267 the inner loop, so g_slice_free_chain_with_offset() provides a real
268 performance benefit over g_slice_free1() now.
270 * glib/gslice.h: renamed g_slice_free_chain() to
271 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
272 a type-safe macro as suggested in bug 323178.
273 simplified the macro implementation of g_slice_free() and implemented
274 it in a type safe manner for all compliers as suggested by Morten
275 Welinder <mortenw@gnome.org>.
280 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
282 2005-12-05 Matthias Clasen <mclasen@redhat.com>
284 * glib/gasyncqueue.c: Add some docs.
286 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
288 * glib/gatomic.c: In the ia64 implementation, use
289 __sync builtin without _si or _di suffix. (#321229,
290 Stanislav Brabec, patch by Andreas Schwab)
292 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
294 * glib/gslice.h: Remove comma at the end of enum.
296 2005-12-04 Matthias Clasen <mclasen@redhat.com>
298 Handle multiple user names with the same UID better.
299 (#319535, Laszlo Peter)
301 * glib/gutils.c (g_get_any_init_do): When determining user
302 data, first look up $LOGNAME. If the UID doesn't match
303 getuid(), fall back to the current behaviour of looking
304 up the user data based on getuid().
306 2005-12-04 Matthias Clasen <mclasen@redhat.com>
308 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
309 to make gcc warn if a function result is ignored. (#145466,
310 Arjan van de Ven, Alex Larsson)
312 * glib/gmem.h: Add the new attribute to g_realloc and
315 2005-12-03 Matthias Clasen <mclasen@redhat.com>
318 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
321 * glib/gmappedfile.c: Make mapping of empty files
324 * glib/gfileutils.c: Don't fork a new process just to
325 fix the permissions of the created temp file. (#321318,
326 Alexis S. L. Carvalho)
328 2005-12-02 Matthias Clasen <mclasen@redhat.com>
330 * README.in: Add a note about Solaris threads.
333 * configure.in: Remove support for Solaris threads.
334 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
336 2005-12-02 Matthias Clasen <mclasen@redhat.com>
338 * configure.in: Define G_GUINT64_CONSTANT in analogy to
339 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
341 2005-12-02 Matthias Clasen <mclasen@redhat.com>
343 * glib/gslice.c: Win32 portability fixes and C99ism removal,
344 pointed out by Kazuki Iwamoto. (#323052)
346 2005-12-02 Matthias Clasen <mclasen@redhat.com>
348 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
349 the IA64 implementation.
351 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
353 * glib/gslice.c: some naming and type size fixups.
355 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
357 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
358 variant into something less verbose, but digestible for gcc-3.4.
360 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
362 * tests/slice-test.c: extended to perform the benchmarking on the old
363 memchunk code if 'O' is selected.
365 * tests/memchunks.c: new file which contains the old GLib mem chunks
366 implementation with prefix old_mem_chunk_.
368 * tests/Makefile.am: added memchunks.c
370 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
372 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
373 acceptable load distribution of the syscall.
374 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
375 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
377 2005-12-01 Matthias Clasen <mclasen@redhat.com>
379 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
380 double-free base interface members.
382 * tests/Makefile.am: Remove duplicate variable.
384 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
386 * glib/gslice.[hc]: new slice allocator implementation.
388 * tests/slice-test.c: added random slice allocation test.
390 * glib/gthread.[hc]: removed newly added private thread mem API.
392 * glib/gthreadinit.h:
395 * glib/gmem.c: divided glib threading initialisation into three phases,
396 initialisation where private keys and messaging are not available (only
397 needed by gmem.c), initialisation without messaging but private keys
398 available (gslice.c, gmessage.c), and full fledged initialisers that
399 server the rest of glib. initialisation functions got renamed to reflect
400 the limitations of their corresponding phases.
402 * glib/gmem.c: removed memchunk code, defer allocations to
405 * glib/gmem.[hc]: removed g_slice_* skeletons.
407 * glib/glib.symbols: added g_slice_* symbols.
409 * configure.in: check for availability of posix_memalign(3), memalign(3)
412 * glib/Makefile.am: added gslice.[hc].
414 2005-12-01 Tor Lillqvist <tml@novell.com>
416 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
417 trailing slash(es) for non-root folders. stat() fails if non-root
418 folders are specified with trailing slashes. It's too much hassle
419 to demand that callers strip such slashes themselves, especially
420 as it is easy to get it wrong and strip the slash of a root
422 (g_rename): On NT-based Windows, use MoveFileEx() with
423 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
425 2005-11-28 Matthias Clasen <mclasen@redhat.com>
427 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
430 * configure.in: Check whether do { } while (0) works.
432 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
433 G_STMT_END if it works.
435 2005-11-28 Matthias Clasen <mclasen@redhat.com>
437 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
438 * glib/gnode.c (g_node_depth_traverse_level):
439 * glib/gmem.c (g_allocator_new):
440 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
441 (g_hash_table_foreach_remove_or_steal): Silence compiler
444 2005-11-27 Matthias Clasen <mclasen@redhat.com>
446 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
447 Don't read beyond len. (#322520, Christian Persch)
449 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
451 * glib/gdataset.c: access datalist flags via atomic pointer access
452 functions, instead of acquiring the dataset lock. this is faster and
453 also matches the atomic pointer readouts in gdatalistprivate.h and
454 direct pointer modifications required by gobject.c.
456 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
458 * glib/gdataset.c: streamlined initialization code somewhat,
459 removed GData node cache.
461 2005-11-22 Matthias Clasen <mclasen@redhat.com>
463 * glib/ghash.c (g_hash_table_ref):
464 (g_hash_table_unref): Mark these as new API.
466 Avoid double locking in g_intern_string (#322133,
469 * glib/gdataset.c (g_quark_from_string_internal): New
470 internal function which factors out common parts of
471 g_quark_from[_static]_string.
472 (g_quark_from_string, g_quark_from_static_string):
473 Use g_quark_from_string_internal.
474 (g_intern_string, g_intern_static_string): Use
475 g_quark_from_string_internal, and only take the
477 (g_quark_new): Don't store the strings shifted by -1
478 in the g_quarks array.
479 (g_quark_to_string): Adapt to the previous change.
481 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
485 g_hash_table_new_full(): create hash tables with a ref count of 1.
486 g_hash_table_ref(): atomically ref_count+=1
487 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
488 when refcount reaches 0.
489 g_hash_table_destroy(): just destroy keys and values, unref by 1.
490 g_hash_table_insert():
491 g_hash_table_replace(): assert ref_count>0.
494 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
497 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
499 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
500 variables. (#321984, Andrew Paprocki)
502 2005-11-18 Matthias Clasen <mclasen@redhat.com>
504 * configure.in: Bump version
506 2005-11-17 Matthias Clasen <mclasen@redhat.com>
508 * === Released 2.9.0 ===
510 * glib/glib.symbols: Updates
512 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
513 hints to the docs. (Paolo Borelli)
517 * glib/gthread.c (g_thread_foreach): Mark as new api.
519 * README.in: Updates.
523 * glib/gdate.c (g_date_set_time_t):
524 (g_date_set_time_val): Add functions to set a date from
525 a time_t and from a GTimeVal.
526 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
529 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
530 and g_utf8_offset_to_pointer.
532 * glib/gutf8.c (g_utf8_pointer_to_offset)
533 (g_utf8_offset_to_pointer): Handle negative offsets, and use
534 "stutter stepping" for going backwards. (#320638, Larry
539 * glib/gthread.c: const correctness fixes, found
540 by Arjan van de Ven and gcc.
542 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
544 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
545 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
547 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
549 * glib/guniprop.c: Use bit hacks instead when checking a general
550 category value against multiple values.
552 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
554 2005-11-10 Simos Xenitellis <simos@gnome.org>
556 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
558 2005-11-05 Tor Lillqvist <tml@novell.com>
560 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
561 (#320688) Instead set error and return failure.
563 2005-11-07 Matthias Clasen <mclasen@redhat.com>
565 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
566 while holding the context lock. (#320886, Andy Wingo)
568 2005-11-07 Billy Biggs <vektor@dumbterm.net>
570 * tests/refcount/Makefile.am: Add a missing reference to libglib
571 in the LDADD for this test.
573 2005-11-07 Matthias Clasen <mclasen@redhat.com>
575 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
576 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
577 GLib 2.9. (#320755, Luca Ferretti)
579 2005-11-04 Matthias Clasen <mclasen@redhat.com>
581 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
582 use a faster lookup table for g_unichar_get_mirror_char().
583 * glib/gmirroringtable.h: The new table.
584 * glib/gunichartables.h: Remove bidi_mirroring_table.
585 * glib/gen-unicode-tables.pl: Don't generate the mirroring
587 * glib/glib-mirroring-tab/*: Sources for the program
588 which generated gmirroringtable.h.
590 * glib/glist.c: Avoid some code duplication.
592 * glib/gscanner.h: Include gdataset.h. (#320322)
594 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
595 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
596 to help find critical warnings in applications. (#320017,
599 2005-11-02 Tor Lillqvist <tml@novell.com>
601 * glib/glib.symbols: Remove large amount of trailing whitespace
602 from one line. Remove #ifdef G_UNDEFINED from around
603 g_hash_table_new and g_hash_table_new_full symbols, what was the
604 purpose of that? Add the g_slice functions.
606 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
608 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
609 added g_slice_*() API to allocate and cache small bits of memory.
610 an actuall allocator implementation for g_slice_*() is still pending.
612 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
613 changed GRealThread list to use in-structure *next; fields instead
614 of GSList, in order for thread iteration to not depenend on g_slice_*()
616 _g_thread_mem_private_get():
617 _g_thread_mem_private_set(): added accessors for private memory,
618 needed because the ordinary GPrivate implementation relies on GArray
619 and GSList and therefore indirectly on working g_slice_*() allocations.
622 g_thread_foreach(): new public API function to loop over all existing threads.
636 removed GAllocator and free list usages and accompanying locks.
637 use g_slice_*() API to allocate and cache small bits of memory.
639 * glib/ghook.h: removed GMemChunk field from public API.
642 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
644 * glib/gnode.h: deprecate allocator API.
646 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
647 be used for poll rec lists.
649 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
650 g_relation_destroy(): free all tuples from the all_tuples hash table,
651 this effectively maintains the life time track keeping of tuples.
652 g_relation_delete_tuple(): free tuples which are removed from the
653 all_tuples hash table. this fixes a temporary leak that was present
654 in the memchunk code until the destruction of the relation.
656 2005-10-29 Matthias Clasen <mclasen@redhat.com>
658 * tests/convert-test.c: Add some tests for conversions between
659 UTF-8, UCS-4 and UTF-16.
661 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
662 of len == -1, noticed by Morten Welinder.
664 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
666 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
668 2005-10-26 Matthias Clasen <mclasen@redhat.com>
670 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
671 in the error case, and add some documentation. (#319806, Morten
674 2005-10-19 Manish Singh <yosh@gimp.org>
676 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
677 config.h symbol automatically. Fixes bug #313731.
679 2005-10-19 Tor Lillqvist <tml@novell.com>
681 * glib/gwin32.c (g_win32_get_package_installation_directory):
682 Return a g_strdup()ed copy of the value stored in the hash table,
683 so that it can be g_free()d without leaving a dangling pointer in
684 the hash table. (#319232)
686 2005-10-06 Matthias Clasen <mclasen@redhat.com>
688 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
689 all-zero sequences correctly. (#317930, Sebastien Bacher)
691 2005-10-05 Matthias Clasen <mclasen@redhat.com>
693 * glib/gmarkup.c (g_markup_escape_text): Doc addition
695 * Makefile.am: Add Collation tests.
697 * tests/collate/*: Inputs and expected outputs for collation tests.
699 * tests/run-collate-tests.sh: Script to run collation tests.
701 * tests/unicode-collate.c (main): Rework slightly to make
702 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
704 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
706 * docs/reference/glib/tmpl/unicode.sgml:
707 * glib/gen-unicode-tables.pl:
709 * glib/gunichartables.h:
711 * tests/casefold.txt:
712 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
713 GUnicodeBreakType types. That may break some applications, like
716 2005-09-26 Matthias Clasen <mclasen@redhat.com>
718 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
719 docs. (#314393, Matthew F. Barnes)
722 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
723 they are already declared in glib.h. This doesn't break documented
724 use of gprintf.h, but should probably be pointed out in the
725 release notes for 2.10. (#314232, Behdad Esfahbod)
727 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
729 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
730 complexity optimization, based on a patch from Matthias Clasen.
732 * tests/patterntest.c: more tests, mostly from matthias.
734 2005-09-20 Matthias Clasen <mclasen@redhat.com>
736 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
737 (#316703, Mark Drago)
739 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
741 * glib/gpattern.c: applied a patch from matthias which checks on the
742 upper bound of GPatternSpec length to optimize matches.
745 * tests/patterntest.c: added more match cases.
747 2005-09-16 Tor Lillqvist <tml@novell.com>
749 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
750 warning to doc comment that these functions should not be used on
751 encodings like CP932.
753 2005-09-14 Matthias Clasen <mclasen@redhat.com>
755 * tests/keyfile-test.c: Add a test for grup names of length 1.
757 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
758 of length 1. (#316309)
760 2005-09-12 Matthias Clasen <mclasen@redhat.com>
762 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
764 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
766 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
767 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
770 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
772 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
773 (child_watch_helper_thread): Remove some dead code. Closes
776 2005-09-07 Tor Lillqvist <tml@novell.com>
778 * glib/Makefile.am: Create also a console version of the
779 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
780 It's otherwise identical to gspawn-win32-helper.exe, except marked
781 as a console application (linked without the -mwindows option).
783 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
784 Drop the dont_wait parameter. Its truth value correlated 100% with
785 the NULLness of the exit_status parameter anyway, so it's enough
786 to check whether exit_status is NULL. Invert the sense of the
787 dont_return_handle parameter and rename it to do_return_handle, to
788 make the code easier to read by avoiding double negations.
790 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
791 to do_spawn_with_pipes() accordingly.
793 (do_spawn_with_pipes): If we have a console, use the console
794 version of the helper program, otherwise use the GUI one. This
795 avoids extra console windows opening up in some situations. (In
796 case a console application uses the GUI gspawn-win32-helper.exe to
797 spawn another console application we would get a separate console
798 for the spawned console application).
800 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
802 2005-09-05 Matthias Clasen <mclasen@redhat.com>
804 * glib/gmappedfile.c (g_mapped_file_new): Report an error
805 if the file is too large. (#315275, Kjartan Maraas)
807 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
808 of read() is signed. (#315273, Kjartan Maraas)
810 2005-08-31 Tor Lillqvist <tml@novell.com>
812 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
813 the EXE and each DLL have their own atexit function chains.
815 #define g_atexit as atexit instead. This means it has a
816 better chance of doing what the caller wants. For instance,
817 gtkhtml calls g_atexit() registering a function in gtkhtml
818 itself. This caused a crash when g_atexit() was implemented as a
819 function in the GLib DLL. The gtkhtml DLL was already unloaded by
820 the time the GLib DLL got unloaded.
822 * glib/gutils.c: #undef the #define mentioned above, to also get a
823 real g_atexit() into the DLL for backward compatibility. Document
824 the Windows behaviour of g_atexit(), and document the varying ways
825 atexit() can behave in the context of dynamically loaded modules
828 2005-08-31 Matthias Clasen <mclasen@redhat.com>
832 * glib/gdataset.c: Add string interning functions.
834 2005-08-28 Matthias Clasen <mclasen@redhat.com>
836 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
839 2005-08-26 Matthias Clasen <mclasen@redhat.com>
841 * configure.in: Bump version to 2.9.0
843 2005-08-25 Tor Lillqvist <tml@novell.com>
845 Make also the g_spawn*() functions take parameters in the GLib
846 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
847 API or ABI. Like the other GLib API that was earlier changed to
848 use UTF-8 on Windows, the names of the functions that take UTF-8
849 have _utf8 suffixes added by using preprocessor macros in the
850 header file. The old names are kept for functions with the old
851 behaviour, taking parameters in the system codepage, for DLL ABI
854 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
855 g_spawn*() functions.
857 * glib/gspawn-win32.c: Use wide-char API on NT-based
858 Windows. Convert parameters from UTF-8 to wide chars (NT) or
859 system codepage (Win9x) and call the C library _wspawn*() or
860 spawn*() functions respectvely. Add DLL ABI stability versions
861 that take parameters in the system codepage.
863 * glib/gspawn-win32-helper.c: On NT-based Windows use the
864 wide-char versions of argv and envp, and use wide-char API to
865 change directory and spawn the program to run. Remove the verbose
866 debugging output, it was too complex to modify for the wide-char
867 features. (Just add temporary debugging printouts if needed, no
868 need to have them permanently in the source.)
870 * glib/gspawn.c: Corresponding documentation updates.
872 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
873 symbols as PRIVATE, add the new _utf8-suffixed ones.
875 2005-08-24 Stepan Kasal <kasal@ucw.cz>
877 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
878 guint64 otherwise and that can produce warnings about comparison
879 between signed and unsigned.
881 2005-08-23 Matthias Clasen <mclasen@redhat.com>
883 * glib/gutils.c: Fix the crt_externs.h include.
885 2005-08-23 Stepan Kasal <kasal@ucw.cz>
887 * NEWS: Fix spelling of my first name.
889 2005-08-23 Matthias Clasen <mclasen@redhat.com>
893 * === Released 2.8.1 ===
897 2005-08-20 Hans Breuer <hans@breuer.org>
899 * glib/makefile.msc.in : link with ws2_32.lib
901 2005-08-18 Tor Lillqvist <tml@novell.com>
903 * configure.in: Check for <sys/wait.h>
905 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
907 2005-08-18 Ross Burton <ross@burtonini.com>
910 Optimise single-character insertions.
915 * tests/string-test.c:
916 Add tests for new optimisation, and fix a leak.
918 2005-08-17 Matthias Clasen <mclasen@redhat.com>
920 * configure.in: Check for crt_externs.h and _NSGetEnviron.
922 * glib/gutils.c: On Darwin, include crt-externs.h and
923 define environ using _NSGetEnviron(). (#313731)
925 2005-08-16 Stepan Kasal <kasal@ucw.cz>
927 * glib/gutils.c (g_get_any_init): Move the body of the big if...
928 (g_get_any_init_do): ... to this new function.
929 (g_get_any_init): Declare as inline.
930 (g_get_any_init_locked): New inline function, does the locking.
931 Make use of these two throughout the code.
933 2005-08-15 Matthias Clasen <mclasen@redhat.com>
935 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
936 the child process and then simply return. This makes
937 The "S" option work as documented in g_on_error_query().
938 (#313125, Matthew F. Barnes)
940 * glib/gunicode.h: Update the link to Unicode category
941 values. (#313369, Behnam Esfahbod)
943 * glib/gqueue.c (g_queue_find_custom): Clarify docs
944 a little. (#311727, Tristan van Berkom)
946 * glib/abicheck.sh, gobject/abicheck.sh: Make the
947 check work on ia64 too, where some symbols ended up
948 in yet another section.