1 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
3 * glib/gslice.c: only use posix_memalign() if it's known to work,
4 revert to memalign() otherwise.
6 * configure.in: check for broken posix_memalign() implementations
9 2006-01-24 Matthias Clasen <mclasen@redhat.com>
11 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
12 on all little-endian systems. (#143380, Marc Moorcroft)
14 2006-01-23 Matthias Clasen <mclasen@redhat.com>
16 * configure.in: Change the shared libary extension for hpux-ia64
17 to so. (#328253, The Written Word)
19 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
21 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
22 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
25 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
27 * glib/gslice.c (slice_config_init): initialize GSlice config from
28 G_SLICE environment variable. we support G_SLICE=always-malloc
29 currently, which forces all g_slice_*() allocations to use the system
33 g_parse_debug_string(): added a note about not using g_malloc() here.
34 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
35 g_slice. contains only guesswork in the WIN32 branch.
37 2006-01-18 Matthias Clasen <mclasen@redhat.com>
41 * === Released 2.9.4 ===
45 2006-01-17 Matthias Clasen <mclasen@redhat.com>
47 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
48 call. (#327421, Morten Welinder)
50 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
52 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
53 G_LOCK. Use the unused_thread_queue lock instead. Change
54 g_thread_pool_thread_proxy such that threads only wait on
55 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
56 due to superfluous tasks. Global tasks wait at most for
57 max-idle-time milliseconds. Make sure, that no task is woken up
58 twice for the same event via a wakeup_serial. This fixes #324228.
60 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
61 invalid NULL into the thread pools. This as well fixes #327290.
63 2006-01-16 Matthias Clasen <mclasen@redhat.com>
65 * configure.in: Bump version
67 * === Released 2.9.3 ===
71 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
72 Rework the strict aliasing fix to not break C++, pointed
73 out by Murray Cumming.
75 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
76 Signal waiting threads, problem noticed by Christian Kellner.
78 2006-01-16 Matthias Clasen <mclasen@redhat.com>
80 Fix bug 326747, Alberto Ruiz:
82 * glib/gconvert.c (make_valid_utf8): Change this function to
83 replace unknown characters by the Unicode replacement character
84 instead of '?', and don't append "(invalid encoding)".
85 (g_filename_display_name, g_filename_display_basename): Document
86 how to determine if the filename was in an invalid encoding.
88 2006-01-14 Matthias Clasen <mclasen@redhat.com>
90 * glib/gtree.c: Replace the simple recursive implementation
91 by a nonrecursive, threaded implementation by Maurizio
94 2006-01-12 Matthias Clasen <mclasen@redhat.com>
96 * tests/asyncqueue-test.c (main): Fix the
97 assert to not always trigger. (#326558,
100 2006-01-11 Matthias Clasen <mclasen@redhat.com>
102 * tests/convert-test.c: Don't test items_read and
103 items_written in error cases where it is not set.
105 * tests/asyncqueue-test.c: Set error to NULL.
107 2006-01-08 Matthias Clasen <mclasen@redhat.com>
109 * glib/gtree.c: Break some long lines.
111 * tests/tree-test.c: Add more tests.
113 * glib/gslice.c: Include config.h first, pointed out
116 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
118 * docs/reference/gobject/tmpl/gtype.sgml: explain that
119 G_TYPE_INSTANCE_GET_CLASS() does behave different during
122 2006-01-06 Matthias Clasen <mclasen@redhat.com>
124 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
125 that we have posix_memalign(). (#323937, Bogdan Nicula)
127 2006-01-05 Matthias Clasen <mclasen@redhat.com>
129 * configure.in: Bump version
131 * === Released 2.9.2 ===
133 2006-01-05 Hans Breuer <hans@breuer.org>
135 * glib/makefile.msc.in : added gslice.obj
137 2006-01-05 Matthias Clasen <mclasen@redhat.com>
141 * glib/gmain.c: Clarify the documentation of source ids
142 a bit. (#325874, Dan Williams)
144 * configure.in: Fix another strict aliasing problem.
146 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
149 2006-01-04 Manish Singh <yosh@gimp.org>
152 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
153 either, since they have the same issue as g_rmdir. (Related to
156 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
158 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
159 order when creating magazines, so we hand out chunks with ascending
162 2006-01-04 Abel Cheung <maddog@linuxhall.org>
164 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
166 2006-01-03 Martyn Russell <martyn@imendio.com>
168 * docs/reference/glib/glib-sections.txt:
170 * glib/gthreadpool.[ch]:
171 - Added new API g_thread_pool_get_idle_time() and
172 g_thread_pool_set_idle_time(). (#324228).
174 * tests/threadpool-test.c:
175 - Updated test case to do thread pool sorting, thread pool with
176 no sorting and a thread pool with idle thread timeouts.
178 2006-01-03 Matthias Clasen <mclasen@redhat.com>
180 * glib/gmain.h: Add new functions here, too.
182 * glib/glib.symbols: Add new functions.
184 2005-12-20 Michael Meeks <michael.meeks@novell.com>
186 * glib/gmain.c (g_main_context_is_owner): new method
187 to determine if the current thread is the owner of the
190 2006-01-02 Matthias Clasen <mclasen@redhat.com>
194 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
195 expanding to rmdir, since rmdir is not declared in a portable
196 system header we can include in gstdio.h. (#325249, Jani Monoses)
198 2006-01-01 Tor Lillqvist <tml@novell.com>
200 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
201 stdout and stderr to unbuffered. Otherwise the giochannel layer
202 will try to read from them regardless whether the
203 g_io_channel_win32_poll() call here has indicated
204 readability or not. (#325310)
206 2005-12-29 Matthias Clasen <mclasen@redhat.com>
208 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
209 in the docs. (#325273, Declan Naughton)
211 2005-12-27 Manish Singh <yosh@gimp.org>
213 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
214 make sure we really override things.
216 2005-12-27 Matthias Clasen <mclasen@redhat.com>
218 Fix #316221, Michal Benes, Stanislav Brabec;
220 * configure.in: Fix a strict aliasing problem in
221 g_static_mutex_get_mutex().
222 * glib/gthread.h: ...and in
223 g_static_mutex_get_mutex_impl_shortcut().
225 * glib/gdatasetprivate.h: Add a cast to silence compiler
226 warnings. (#321978, Andrew Paprocki)
228 Partial fix for bug #323937, Bogdan Nicula.
230 * configure.in: Check for malloc.h
231 * glib/gslice.c: Don't include malloc.h unconditionally.
233 2005-12-27 Manish Singh <yosh@gimp.org>
235 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
236 (#324950, Dan Yefimov)
238 2005-12-26 Matthias Clasen <mclasen@redhat.com>
240 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
243 2005-12-25 Matthias Clasen <mclasen@redhat.com>
245 * AUTHORS: Update my email
247 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
249 2005-12-21 Matthias Clasen <mclasen@redhat.com>
251 * README.in: Add some notes on when 'make check' may fail.
253 2005-12-20 Matthias Clasen <mclasen@redhat.com>
256 * glib/gthreadpool.h:
257 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
258 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
260 * tests/threadpool-test.c: Test this.
262 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
264 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
265 handle errors without depending on gmessage.c which might not be
266 setup when the error occours.
267 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
268 always freeing can be achieved by adjusting the working set time to
269 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
270 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
271 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
272 minimize wastage if small amounts of differently sized structrues are
273 allocated, this does come at a performance cost of roughly 5% though.
274 fixed up block alignment calculation, so it works for varying
275 block sizes. only use strerror() not g_strerror() since the latter
276 depends on working GQuark and GSlice.
277 mem_error(): implemented in terms of fprintf and vfprintf.
279 * tests/slice-color.c: new program to test cache colorization effects.
281 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
282 G_SLICE_CONFIG_WORKING_SET_MSECS.
284 2005-12-17 Matthias Clasen <mclasen@redhat.com>
286 * glib/goption.c (parse_short_option): Set an error in all
287 failure cases. (#324332, Tim-Philipp Müller)
289 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
291 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
292 and g_atomic_pointer_get. Add g_atomic_int_set and
293 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
294 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
295 as functions (additionally to the macros in the header) for the
296 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
298 2005-12-16 Matthias Clasen <mclasen@redhat.com>
300 * glib/gmem.c (g_allocator_new): Don't return a pointer to
301 a const struct, since apps expect to be able to modify it.
302 (#324179, J. Ali Harlow)
304 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
306 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
307 for systems where the initialization of atomic variables requires a
308 write memory barrier.
310 2005-12-09 Matthias Clasen <mclasen@redhat.com>
312 * configure.in: Bump version
314 * === Released 2.9.1 ===
316 2005-12-09 Alexander Larsson <alexl@redhat.com>
320 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
323 2005-12-08 Matthias Clasen <mclasen@redhat.com>
327 2005-12-07 Martyn Russell <martyn@imendio.com>
329 * glib/gasyncqueue.c:
330 - Call g_queue_insert_sorted() instead of duplicating the code.
331 - Call g_queue_sort() instead of duplicating the code.
332 - Invert sort function results to make sure the same sort function
333 gives the same results across glist, gslist, gqueue and
336 * tests/asyncqueue-test.c:
337 - Updated the sort function to reflect the example in the
338 documentation for gasyncqueue.c.
340 2005-12-07 Martyn Russell <martyn@imendio.com>
342 * docs/reference/glib/glib-sections.txt:
343 * docs/reference/glib/tmpl/linked_lists_double.sgml:
344 * docs/reference/glib/tmpl/linked_lists_single.sgml:
347 - Added g_list_insert_sorted_with_data () and
348 g_slist_insert_sorted_with_data ().
349 - Removed the extra check in g_list_sort() and g_slist_sort() for
350 GCompareDataFunc vs. GCompareFunc.
352 2005-12-07 Tor Lillqvist <tml@novell.com>
354 * glib-gettextize.in: Look up prefix at run-time on Win32,
355 assuming the standard directory structure with glib-gettextize in
358 2005-12-06 Matthias Clasen <mclasen@redhat.com>
360 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
361 we need GTK+ 2.8 to compile against GLib 2.10.
363 2005-12-05 Manish Singh <yosh@gimp.org>
365 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
366 to override any user settings.
368 2005-12-05 Matthias Clasen <mclasen@redhat.com>
370 * configure.in: Revert an accidential commit
372 2005-12-05 Matthias Clasen <mclasen@redhat.com>
374 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
376 * tests/Makefile.am (TESTS_ENVIRONMENT):
377 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
378 MALLOC_CHECK_ and MALLOC_PERTURB_
380 * tests/run-collate-tests.sh: Run the collation tests explicitly
381 in en_US locale. (#320463)
383 * glib/gmem.h: Really deprecate GMemChunk.
390 * glib/gstrfuncs.c: Add versioned deprecation docs.
392 2005-12-05 Martyn Russell <martyn@imendio.com>
394 * docs/reference/glib/glib-sections.txt:
395 * glib/gasyncqueue.[ch]:
396 - Added support for sorting async queues by with _push_sorted(),
397 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
400 * tests/asyncqueue-test.c:
401 - Added test case for gasyncqueue.c
403 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
405 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
406 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
407 in bug 323178. moved time consuming logic from g_slice_free() out of
408 the inner loop, so g_slice_free_chain_with_offset() provides a real
409 performance benefit over g_slice_free1() now.
411 * glib/gslice.h: renamed g_slice_free_chain() to
412 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
413 a type-safe macro as suggested in bug 323178.
414 simplified the macro implementation of g_slice_free() and implemented
415 it in a type safe manner for all compliers as suggested by Morten
416 Welinder <mortenw@gnome.org>.
421 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
423 2005-12-05 Matthias Clasen <mclasen@redhat.com>
425 * glib/gasyncqueue.c: Add some docs.
427 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
429 * glib/gatomic.c: In the ia64 implementation, use
430 __sync builtin without _si or _di suffix. (#321229,
431 Stanislav Brabec, patch by Andreas Schwab)
433 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
435 * glib/gslice.h: Remove comma at the end of enum.
437 2005-12-04 Matthias Clasen <mclasen@redhat.com>
439 Handle multiple user names with the same UID better.
440 (#319535, Laszlo Peter)
442 * glib/gutils.c (g_get_any_init_do): When determining user
443 data, first look up $LOGNAME. If the UID doesn't match
444 getuid(), fall back to the current behaviour of looking
445 up the user data based on getuid().
447 2005-12-04 Matthias Clasen <mclasen@redhat.com>
449 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
450 to make gcc warn if a function result is ignored. (#145466,
451 Arjan van de Ven, Alex Larsson)
453 * glib/gmem.h: Add the new attribute to g_realloc and
456 2005-12-03 Matthias Clasen <mclasen@redhat.com>
459 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
462 * glib/gmappedfile.c: Make mapping of empty files
465 * glib/gfileutils.c: Don't fork a new process just to
466 fix the permissions of the created temp file. (#321318,
467 Alexis S. L. Carvalho)
469 2005-12-02 Matthias Clasen <mclasen@redhat.com>
471 * README.in: Add a note about Solaris threads.
474 * configure.in: Remove support for Solaris threads.
475 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
477 2005-12-02 Matthias Clasen <mclasen@redhat.com>
479 * configure.in: Define G_GUINT64_CONSTANT in analogy to
480 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
482 2005-12-02 Matthias Clasen <mclasen@redhat.com>
484 * glib/gslice.c: Win32 portability fixes and C99ism removal,
485 pointed out by Kazuki Iwamoto. (#323052)
487 2005-12-02 Matthias Clasen <mclasen@redhat.com>
489 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
490 the IA64 implementation.
492 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
494 * glib/gslice.c: some naming and type size fixups.
496 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
498 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
499 variant into something less verbose, but digestible for gcc-3.4.
501 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
503 * tests/slice-test.c: extended to perform the benchmarking on the old
504 memchunk code if 'O' is selected.
506 * tests/memchunks.c: new file which contains the old GLib mem chunks
507 implementation with prefix old_mem_chunk_.
509 * tests/Makefile.am: added memchunks.c
511 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
513 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
514 acceptable load distribution of the syscall.
515 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
516 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
518 2005-12-01 Matthias Clasen <mclasen@redhat.com>
520 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
521 double-free base interface members.
523 * tests/Makefile.am: Remove duplicate variable.
525 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
527 * glib/gslice.[hc]: new slice allocator implementation.
529 * tests/slice-test.c: added random slice allocation test.
531 * glib/gthread.[hc]: removed newly added private thread mem API.
533 * glib/gthreadinit.h:
536 * glib/gmem.c: divided glib threading initialisation into three phases,
537 initialisation where private keys and messaging are not available (only
538 needed by gmem.c), initialisation without messaging but private keys
539 available (gslice.c, gmessage.c), and full fledged initialisers that
540 server the rest of glib. initialisation functions got renamed to reflect
541 the limitations of their corresponding phases.
543 * glib/gmem.c: removed memchunk code, defer allocations to
546 * glib/gmem.[hc]: removed g_slice_* skeletons.
548 * glib/glib.symbols: added g_slice_* symbols.
550 * configure.in: check for availability of posix_memalign(3), memalign(3)
553 * glib/Makefile.am: added gslice.[hc].
555 2005-12-01 Tor Lillqvist <tml@novell.com>
557 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
558 trailing slash(es) for non-root folders. stat() fails if non-root
559 folders are specified with trailing slashes. It's too much hassle
560 to demand that callers strip such slashes themselves, especially
561 as it is easy to get it wrong and strip the slash of a root
563 (g_rename): On NT-based Windows, use MoveFileEx() with
564 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
566 2005-11-28 Matthias Clasen <mclasen@redhat.com>
568 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
571 * configure.in: Check whether do { } while (0) works.
573 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
574 G_STMT_END if it works.
576 2005-11-28 Matthias Clasen <mclasen@redhat.com>
578 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
579 * glib/gnode.c (g_node_depth_traverse_level):
580 * glib/gmem.c (g_allocator_new):
581 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
582 (g_hash_table_foreach_remove_or_steal): Silence compiler
585 2005-11-27 Matthias Clasen <mclasen@redhat.com>
587 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
588 Don't read beyond len. (#322520, Christian Persch)
590 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
592 * glib/gdataset.c: access datalist flags via atomic pointer access
593 functions, instead of acquiring the dataset lock. this is faster and
594 also matches the atomic pointer readouts in gdatalistprivate.h and
595 direct pointer modifications required by gobject.c.
597 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
599 * glib/gdataset.c: streamlined initialization code somewhat,
600 removed GData node cache.
602 2005-11-22 Matthias Clasen <mclasen@redhat.com>
604 * glib/ghash.c (g_hash_table_ref):
605 (g_hash_table_unref): Mark these as new API.
607 Avoid double locking in g_intern_string (#322133,
610 * glib/gdataset.c (g_quark_from_string_internal): New
611 internal function which factors out common parts of
612 g_quark_from[_static]_string.
613 (g_quark_from_string, g_quark_from_static_string):
614 Use g_quark_from_string_internal.
615 (g_intern_string, g_intern_static_string): Use
616 g_quark_from_string_internal, and only take the
618 (g_quark_new): Don't store the strings shifted by -1
619 in the g_quarks array.
620 (g_quark_to_string): Adapt to the previous change.
622 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
626 g_hash_table_new_full(): create hash tables with a ref count of 1.
627 g_hash_table_ref(): atomically ref_count+=1
628 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
629 when refcount reaches 0.
630 g_hash_table_destroy(): just destroy keys and values, unref by 1.
631 g_hash_table_insert():
632 g_hash_table_replace(): assert ref_count>0.
635 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
638 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
640 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
641 variables. (#321984, Andrew Paprocki)
643 2005-11-18 Matthias Clasen <mclasen@redhat.com>
645 * configure.in: Bump version
647 2005-11-17 Matthias Clasen <mclasen@redhat.com>
649 * === Released 2.9.0 ===
651 * glib/glib.symbols: Updates
653 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
654 hints to the docs. (Paolo Borelli)
658 * glib/gthread.c (g_thread_foreach): Mark as new api.
660 * README.in: Updates.
664 * glib/gdate.c (g_date_set_time_t):
665 (g_date_set_time_val): Add functions to set a date from
666 a time_t and from a GTimeVal.
667 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
670 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
671 and g_utf8_offset_to_pointer.
673 * glib/gutf8.c (g_utf8_pointer_to_offset)
674 (g_utf8_offset_to_pointer): Handle negative offsets, and use
675 "stutter stepping" for going backwards. (#320638, Larry
680 * glib/gthread.c: const correctness fixes, found
681 by Arjan van de Ven and gcc.
683 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
685 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
686 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
688 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
690 * glib/guniprop.c: Use bit hacks instead when checking a general
691 category value against multiple values.
693 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
695 2005-11-10 Simos Xenitellis <simos@gnome.org>
697 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
699 2005-11-05 Tor Lillqvist <tml@novell.com>
701 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
702 (#320688) Instead set error and return failure.
704 2005-11-07 Matthias Clasen <mclasen@redhat.com>
706 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
707 while holding the context lock. (#320886, Andy Wingo)
709 2005-11-07 Billy Biggs <vektor@dumbterm.net>
711 * tests/refcount/Makefile.am: Add a missing reference to libglib
712 in the LDADD for this test.
714 2005-11-07 Matthias Clasen <mclasen@redhat.com>
716 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
717 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
718 GLib 2.9. (#320755, Luca Ferretti)
720 2005-11-04 Matthias Clasen <mclasen@redhat.com>
722 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
723 use a faster lookup table for g_unichar_get_mirror_char().
724 * glib/gmirroringtable.h: The new table.
725 * glib/gunichartables.h: Remove bidi_mirroring_table.
726 * glib/gen-unicode-tables.pl: Don't generate the mirroring
728 * glib/glib-mirroring-tab/*: Sources for the program
729 which generated gmirroringtable.h.
731 * glib/glist.c: Avoid some code duplication.
733 * glib/gscanner.h: Include gdataset.h. (#320322)
735 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
736 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
737 to help find critical warnings in applications. (#320017,
740 2005-11-02 Tor Lillqvist <tml@novell.com>
742 * glib/glib.symbols: Remove large amount of trailing whitespace
743 from one line. Remove #ifdef G_UNDEFINED from around
744 g_hash_table_new and g_hash_table_new_full symbols, what was the
745 purpose of that? Add the g_slice functions.
747 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
749 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
750 added g_slice_*() API to allocate and cache small bits of memory.
751 an actuall allocator implementation for g_slice_*() is still pending.
753 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
754 changed GRealThread list to use in-structure *next; fields instead
755 of GSList, in order for thread iteration to not depenend on g_slice_*()
757 _g_thread_mem_private_get():
758 _g_thread_mem_private_set(): added accessors for private memory,
759 needed because the ordinary GPrivate implementation relies on GArray
760 and GSList and therefore indirectly on working g_slice_*() allocations.
763 g_thread_foreach(): new public API function to loop over all existing threads.
777 removed GAllocator and free list usages and accompanying locks.
778 use g_slice_*() API to allocate and cache small bits of memory.
780 * glib/ghook.h: removed GMemChunk field from public API.
783 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
785 * glib/gnode.h: deprecate allocator API.
787 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
788 be used for poll rec lists.
790 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
791 g_relation_destroy(): free all tuples from the all_tuples hash table,
792 this effectively maintains the life time track keeping of tuples.
793 g_relation_delete_tuple(): free tuples which are removed from the
794 all_tuples hash table. this fixes a temporary leak that was present
795 in the memchunk code until the destruction of the relation.
797 2005-10-29 Matthias Clasen <mclasen@redhat.com>
799 * tests/convert-test.c: Add some tests for conversions between
800 UTF-8, UCS-4 and UTF-16.
802 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
803 of len == -1, noticed by Morten Welinder.
805 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
807 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
809 2005-10-26 Matthias Clasen <mclasen@redhat.com>
811 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
812 in the error case, and add some documentation. (#319806, Morten
815 2005-10-19 Manish Singh <yosh@gimp.org>
817 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
818 config.h symbol automatically. Fixes bug #313731.
820 2005-10-19 Tor Lillqvist <tml@novell.com>
822 * glib/gwin32.c (g_win32_get_package_installation_directory):
823 Return a g_strdup()ed copy of the value stored in the hash table,
824 so that it can be g_free()d without leaving a dangling pointer in
825 the hash table. (#319232)
827 2005-10-06 Matthias Clasen <mclasen@redhat.com>
829 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
830 all-zero sequences correctly. (#317930, Sebastien Bacher)
832 2005-10-05 Matthias Clasen <mclasen@redhat.com>
834 * glib/gmarkup.c (g_markup_escape_text): Doc addition
836 * Makefile.am: Add Collation tests.
838 * tests/collate/*: Inputs and expected outputs for collation tests.
840 * tests/run-collate-tests.sh: Script to run collation tests.
842 * tests/unicode-collate.c (main): Rework slightly to make
843 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
845 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
847 * docs/reference/glib/tmpl/unicode.sgml:
848 * glib/gen-unicode-tables.pl:
850 * glib/gunichartables.h:
852 * tests/casefold.txt:
853 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
854 GUnicodeBreakType types. That may break some applications, like
857 2005-09-26 Matthias Clasen <mclasen@redhat.com>
859 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
860 docs. (#314393, Matthew F. Barnes)
863 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
864 they are already declared in glib.h. This doesn't break documented
865 use of gprintf.h, but should probably be pointed out in the
866 release notes for 2.10. (#314232, Behdad Esfahbod)
868 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
870 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
871 complexity optimization, based on a patch from Matthias Clasen.
873 * tests/patterntest.c: more tests, mostly from matthias.
875 2005-09-20 Matthias Clasen <mclasen@redhat.com>
877 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
878 (#316703, Mark Drago)
880 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
882 * glib/gpattern.c: applied a patch from matthias which checks on the
883 upper bound of GPatternSpec length to optimize matches.
886 * tests/patterntest.c: added more match cases.
888 2005-09-16 Tor Lillqvist <tml@novell.com>
890 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
891 warning to doc comment that these functions should not be used on
892 encodings like CP932.
894 2005-09-14 Matthias Clasen <mclasen@redhat.com>
896 * tests/keyfile-test.c: Add a test for grup names of length 1.
898 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
899 of length 1. (#316309)
901 2005-09-12 Matthias Clasen <mclasen@redhat.com>
903 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
905 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
907 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
908 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
911 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
913 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
914 (child_watch_helper_thread): Remove some dead code. Closes
917 2005-09-07 Tor Lillqvist <tml@novell.com>
919 * glib/Makefile.am: Create also a console version of the
920 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
921 It's otherwise identical to gspawn-win32-helper.exe, except marked
922 as a console application (linked without the -mwindows option).
924 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
925 Drop the dont_wait parameter. Its truth value correlated 100% with
926 the NULLness of the exit_status parameter anyway, so it's enough
927 to check whether exit_status is NULL. Invert the sense of the
928 dont_return_handle parameter and rename it to do_return_handle, to
929 make the code easier to read by avoiding double negations.
931 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
932 to do_spawn_with_pipes() accordingly.
934 (do_spawn_with_pipes): If we have a console, use the console
935 version of the helper program, otherwise use the GUI one. This
936 avoids extra console windows opening up in some situations. (In
937 case a console application uses the GUI gspawn-win32-helper.exe to
938 spawn another console application we would get a separate console
939 for the spawned console application).
941 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
943 2005-09-05 Matthias Clasen <mclasen@redhat.com>
945 * glib/gmappedfile.c (g_mapped_file_new): Report an error
946 if the file is too large. (#315275, Kjartan Maraas)
948 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
949 of read() is signed. (#315273, Kjartan Maraas)
951 2005-08-31 Tor Lillqvist <tml@novell.com>
953 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
954 the EXE and each DLL have their own atexit function chains.
956 #define g_atexit as atexit instead. This means it has a
957 better chance of doing what the caller wants. For instance,
958 gtkhtml calls g_atexit() registering a function in gtkhtml
959 itself. This caused a crash when g_atexit() was implemented as a
960 function in the GLib DLL. The gtkhtml DLL was already unloaded by
961 the time the GLib DLL got unloaded.
963 * glib/gutils.c: #undef the #define mentioned above, to also get a
964 real g_atexit() into the DLL for backward compatibility. Document
965 the Windows behaviour of g_atexit(), and document the varying ways
966 atexit() can behave in the context of dynamically loaded modules
969 2005-08-31 Matthias Clasen <mclasen@redhat.com>
973 * glib/gdataset.c: Add string interning functions.
975 2005-08-28 Matthias Clasen <mclasen@redhat.com>
977 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
980 2005-08-26 Matthias Clasen <mclasen@redhat.com>
982 * configure.in: Bump version to 2.9.0
984 2005-08-25 Tor Lillqvist <tml@novell.com>
986 Make also the g_spawn*() functions take parameters in the GLib
987 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
988 API or ABI. Like the other GLib API that was earlier changed to
989 use UTF-8 on Windows, the names of the functions that take UTF-8
990 have _utf8 suffixes added by using preprocessor macros in the
991 header file. The old names are kept for functions with the old
992 behaviour, taking parameters in the system codepage, for DLL ABI
995 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
996 g_spawn*() functions.
998 * glib/gspawn-win32.c: Use wide-char API on NT-based
999 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1000 system codepage (Win9x) and call the C library _wspawn*() or
1001 spawn*() functions respectvely. Add DLL ABI stability versions
1002 that take parameters in the system codepage.
1004 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1005 wide-char versions of argv and envp, and use wide-char API to
1006 change directory and spawn the program to run. Remove the verbose
1007 debugging output, it was too complex to modify for the wide-char
1008 features. (Just add temporary debugging printouts if needed, no
1009 need to have them permanently in the source.)
1011 * glib/gspawn.c: Corresponding documentation updates.
1013 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1014 symbols as PRIVATE, add the new _utf8-suffixed ones.
1016 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1018 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1019 guint64 otherwise and that can produce warnings about comparison
1020 between signed and unsigned.
1022 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1024 * glib/gutils.c: Fix the crt_externs.h include.
1026 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1028 * NEWS: Fix spelling of my first name.
1030 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1034 * === Released 2.8.1 ===
1038 2005-08-20 Hans Breuer <hans@breuer.org>
1040 * glib/makefile.msc.in : link with ws2_32.lib
1042 2005-08-18 Tor Lillqvist <tml@novell.com>
1044 * configure.in: Check for <sys/wait.h>
1046 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1048 2005-08-18 Ross Burton <ross@burtonini.com>
1051 Optimise single-character insertions.
1056 * tests/string-test.c:
1057 Add tests for new optimisation, and fix a leak.
1059 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1061 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1063 * glib/gutils.c: On Darwin, include crt-externs.h and
1064 define environ using _NSGetEnviron(). (#313731)
1066 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1068 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1069 (g_get_any_init_do): ... to this new function.
1070 (g_get_any_init): Declare as inline.
1071 (g_get_any_init_locked): New inline function, does the locking.
1072 Make use of these two throughout the code.
1074 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1076 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1077 the child process and then simply return. This makes
1078 The "S" option work as documented in g_on_error_query().
1079 (#313125, Matthew F. Barnes)
1081 * glib/gunicode.h: Update the link to Unicode category
1082 values. (#313369, Behnam Esfahbod)
1084 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1085 a little. (#311727, Tristan van Berkom)
1087 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1088 check work on ia64 too, where some symbols ended up
1089 in yet another section.