1 2006-01-31 Matthias Clasen <mclasen@redhat.com>
3 * autogen.sh: Touch README and INSTALL here to pacify
4 automake. (#329124, Kjartan Maraas, Tim Janik)
6 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
8 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
9 some missing cases where pointed out by Sebastian Wilhelmi.
11 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
13 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
15 2006-01-30 David Schleef <ds@schleef.org>
17 * glib/grand.c: Update URL
19 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
21 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
22 create container classes which are as flexible as a GValue is
24 2006-01-27 Matthias Clasen <mclasen@redhat.com>
26 * configure.in: Bump version
28 * === Released 2.9.5 ===
30 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
35 2006-01-26 Matthias Clasen <mclasen@redhat.com>
37 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
40 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
42 * fixed buglets reported by Jens Granseuer in #328254.
44 * configure.in: free the memory allocated in posix_memalign() tests.
46 * glib/gslice.c: spelling fixes.
48 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
50 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
51 slices, make sure g_mem_gc_friendly is properly initialized.
53 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
54 the first allocation. applied some branching optimizations.
56 * docs/macros.txt: reflected --enable-gc-friendly change and
57 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
59 * configure.in: changed --enable-gc-friendly=yes to define
60 ENABLE_GC_FRIENDLY_DEFAULT.
62 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
63 if (G_UNLIKELY (g_mem_gc_friendly)).
66 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
67 care of by g_slice_free1().
69 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
71 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
73 * glib/gslice.c: only use posix_memalign() if it's known to work,
74 revert to memalign() otherwise.
76 * configure.in: check for broken posix_memalign() implementations
79 2006-01-24 Matthias Clasen <mclasen@redhat.com>
81 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
82 on all little-endian systems. (#143380, Marc Moorcroft)
84 2006-01-23 Matthias Clasen <mclasen@redhat.com>
86 * configure.in: Change the shared libary extension for hpux-ia64
87 to so. (#328253, The Written Word)
89 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
91 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
92 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
95 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
97 * glib/gslice.c (slice_config_init): initialize GSlice config from
98 G_SLICE environment variable. we support G_SLICE=always-malloc
99 currently, which forces all g_slice_*() allocations to use the system
103 g_parse_debug_string(): added a note about not using g_malloc() here.
104 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
105 g_slice. contains only guesswork in the WIN32 branch.
107 2006-01-18 Matthias Clasen <mclasen@redhat.com>
111 * === Released 2.9.4 ===
115 2006-01-17 Matthias Clasen <mclasen@redhat.com>
117 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
118 call. (#327421, Morten Welinder)
120 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
122 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
123 G_LOCK. Use the unused_thread_queue lock instead. Change
124 g_thread_pool_thread_proxy such that threads only wait on
125 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
126 due to superfluous tasks. Global tasks wait at most for
127 max-idle-time milliseconds. Make sure, that no task is woken up
128 twice for the same event via a wakeup_serial. This fixes #324228.
130 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
131 invalid NULL into the thread pools. This as well fixes #327290.
133 2006-01-16 Matthias Clasen <mclasen@redhat.com>
135 * configure.in: Bump version
137 * === Released 2.9.3 ===
141 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
142 Rework the strict aliasing fix to not break C++, pointed
143 out by Murray Cumming.
145 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
146 Signal waiting threads, problem noticed by Christian Kellner.
148 2006-01-16 Matthias Clasen <mclasen@redhat.com>
150 Fix bug 326747, Alberto Ruiz:
152 * glib/gconvert.c (make_valid_utf8): Change this function to
153 replace unknown characters by the Unicode replacement character
154 instead of '?', and don't append "(invalid encoding)".
155 (g_filename_display_name, g_filename_display_basename): Document
156 how to determine if the filename was in an invalid encoding.
158 2006-01-14 Matthias Clasen <mclasen@redhat.com>
160 * glib/gtree.c: Replace the simple recursive implementation
161 by a nonrecursive, threaded implementation by Maurizio
164 2006-01-12 Matthias Clasen <mclasen@redhat.com>
166 * tests/asyncqueue-test.c (main): Fix the
167 assert to not always trigger. (#326558,
170 2006-01-11 Matthias Clasen <mclasen@redhat.com>
172 * tests/convert-test.c: Don't test items_read and
173 items_written in error cases where it is not set.
175 * tests/asyncqueue-test.c: Set error to NULL.
177 2006-01-08 Matthias Clasen <mclasen@redhat.com>
179 * glib/gtree.c: Break some long lines.
181 * tests/tree-test.c: Add more tests.
183 * glib/gslice.c: Include config.h first, pointed out
186 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
188 * docs/reference/gobject/tmpl/gtype.sgml: explain that
189 G_TYPE_INSTANCE_GET_CLASS() does behave different during
192 2006-01-06 Matthias Clasen <mclasen@redhat.com>
194 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
195 that we have posix_memalign(). (#323937, Bogdan Nicula)
197 2006-01-05 Matthias Clasen <mclasen@redhat.com>
199 * configure.in: Bump version
201 * === Released 2.9.2 ===
203 2006-01-05 Hans Breuer <hans@breuer.org>
205 * glib/makefile.msc.in : added gslice.obj
207 2006-01-05 Matthias Clasen <mclasen@redhat.com>
211 * glib/gmain.c: Clarify the documentation of source ids
212 a bit. (#325874, Dan Williams)
214 * configure.in: Fix another strict aliasing problem.
216 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
219 2006-01-04 Manish Singh <yosh@gimp.org>
222 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
223 either, since they have the same issue as g_rmdir. (Related to
226 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
228 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
229 order when creating magazines, so we hand out chunks with ascending
232 2006-01-04 Abel Cheung <maddog@linuxhall.org>
234 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
236 2006-01-03 Martyn Russell <martyn@imendio.com>
238 * docs/reference/glib/glib-sections.txt:
240 * glib/gthreadpool.[ch]:
241 - Added new API g_thread_pool_get_idle_time() and
242 g_thread_pool_set_idle_time(). (#324228).
244 * tests/threadpool-test.c:
245 - Updated test case to do thread pool sorting, thread pool with
246 no sorting and a thread pool with idle thread timeouts.
248 2006-01-03 Matthias Clasen <mclasen@redhat.com>
250 * glib/gmain.h: Add new functions here, too.
252 * glib/glib.symbols: Add new functions.
254 2005-12-20 Michael Meeks <michael.meeks@novell.com>
256 * glib/gmain.c (g_main_context_is_owner): new method
257 to determine if the current thread is the owner of the
260 2006-01-02 Matthias Clasen <mclasen@redhat.com>
264 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
265 expanding to rmdir, since rmdir is not declared in a portable
266 system header we can include in gstdio.h. (#325249, Jani Monoses)
268 2006-01-01 Tor Lillqvist <tml@novell.com>
270 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
271 stdout and stderr to unbuffered. Otherwise the giochannel layer
272 will try to read from them regardless whether the
273 g_io_channel_win32_poll() call here has indicated
274 readability or not. (#325310)
276 2005-12-29 Matthias Clasen <mclasen@redhat.com>
278 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
279 in the docs. (#325273, Declan Naughton)
281 2005-12-27 Manish Singh <yosh@gimp.org>
283 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
284 make sure we really override things.
286 2005-12-27 Matthias Clasen <mclasen@redhat.com>
288 Fix #316221, Michal Benes, Stanislav Brabec;
290 * configure.in: Fix a strict aliasing problem in
291 g_static_mutex_get_mutex().
292 * glib/gthread.h: ...and in
293 g_static_mutex_get_mutex_impl_shortcut().
295 * glib/gdatasetprivate.h: Add a cast to silence compiler
296 warnings. (#321978, Andrew Paprocki)
298 Partial fix for bug #323937, Bogdan Nicula.
300 * configure.in: Check for malloc.h
301 * glib/gslice.c: Don't include malloc.h unconditionally.
303 2005-12-27 Manish Singh <yosh@gimp.org>
305 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
306 (#324950, Dan Yefimov)
308 2005-12-26 Matthias Clasen <mclasen@redhat.com>
310 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
313 2005-12-25 Matthias Clasen <mclasen@redhat.com>
315 * AUTHORS: Update my email
317 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
319 2005-12-21 Matthias Clasen <mclasen@redhat.com>
321 * README.in: Add some notes on when 'make check' may fail.
323 2005-12-20 Matthias Clasen <mclasen@redhat.com>
326 * glib/gthreadpool.h:
327 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
328 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
330 * tests/threadpool-test.c: Test this.
332 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
334 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
335 handle errors without depending on gmessage.c which might not be
336 setup when the error occours.
337 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
338 always freeing can be achieved by adjusting the working set time to
339 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
340 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
341 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
342 minimize wastage if small amounts of differently sized structrues are
343 allocated, this does come at a performance cost of roughly 5% though.
344 fixed up block alignment calculation, so it works for varying
345 block sizes. only use strerror() not g_strerror() since the latter
346 depends on working GQuark and GSlice.
347 mem_error(): implemented in terms of fprintf and vfprintf.
349 * tests/slice-color.c: new program to test cache colorization effects.
351 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
352 G_SLICE_CONFIG_WORKING_SET_MSECS.
354 2005-12-17 Matthias Clasen <mclasen@redhat.com>
356 * glib/goption.c (parse_short_option): Set an error in all
357 failure cases. (#324332, Tim-Philipp Müller)
359 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
361 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
362 and g_atomic_pointer_get. Add g_atomic_int_set and
363 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
364 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
365 as functions (additionally to the macros in the header) for the
366 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
368 2005-12-16 Matthias Clasen <mclasen@redhat.com>
370 * glib/gmem.c (g_allocator_new): Don't return a pointer to
371 a const struct, since apps expect to be able to modify it.
372 (#324179, J. Ali Harlow)
374 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
376 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
377 for systems where the initialization of atomic variables requires a
378 write memory barrier.
380 2005-12-09 Matthias Clasen <mclasen@redhat.com>
382 * configure.in: Bump version
384 * === Released 2.9.1 ===
386 2005-12-09 Alexander Larsson <alexl@redhat.com>
390 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
393 2005-12-08 Matthias Clasen <mclasen@redhat.com>
397 2005-12-07 Martyn Russell <martyn@imendio.com>
399 * glib/gasyncqueue.c:
400 - Call g_queue_insert_sorted() instead of duplicating the code.
401 - Call g_queue_sort() instead of duplicating the code.
402 - Invert sort function results to make sure the same sort function
403 gives the same results across glist, gslist, gqueue and
406 * tests/asyncqueue-test.c:
407 - Updated the sort function to reflect the example in the
408 documentation for gasyncqueue.c.
410 2005-12-07 Martyn Russell <martyn@imendio.com>
412 * docs/reference/glib/glib-sections.txt:
413 * docs/reference/glib/tmpl/linked_lists_double.sgml:
414 * docs/reference/glib/tmpl/linked_lists_single.sgml:
417 - Added g_list_insert_sorted_with_data () and
418 g_slist_insert_sorted_with_data ().
419 - Removed the extra check in g_list_sort() and g_slist_sort() for
420 GCompareDataFunc vs. GCompareFunc.
422 2005-12-07 Tor Lillqvist <tml@novell.com>
424 * glib-gettextize.in: Look up prefix at run-time on Win32,
425 assuming the standard directory structure with glib-gettextize in
428 2005-12-06 Matthias Clasen <mclasen@redhat.com>
430 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
431 we need GTK+ 2.8 to compile against GLib 2.10.
433 2005-12-05 Manish Singh <yosh@gimp.org>
435 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
436 to override any user settings.
438 2005-12-05 Matthias Clasen <mclasen@redhat.com>
440 * configure.in: Revert an accidential commit
442 2005-12-05 Matthias Clasen <mclasen@redhat.com>
444 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
446 * tests/Makefile.am (TESTS_ENVIRONMENT):
447 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
448 MALLOC_CHECK_ and MALLOC_PERTURB_
450 * tests/run-collate-tests.sh: Run the collation tests explicitly
451 in en_US locale. (#320463)
453 * glib/gmem.h: Really deprecate GMemChunk.
460 * glib/gstrfuncs.c: Add versioned deprecation docs.
462 2005-12-05 Martyn Russell <martyn@imendio.com>
464 * docs/reference/glib/glib-sections.txt:
465 * glib/gasyncqueue.[ch]:
466 - Added support for sorting async queues by with _push_sorted(),
467 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
470 * tests/asyncqueue-test.c:
471 - Added test case for gasyncqueue.c
473 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
475 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
476 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
477 in bug 323178. moved time consuming logic from g_slice_free() out of
478 the inner loop, so g_slice_free_chain_with_offset() provides a real
479 performance benefit over g_slice_free1() now.
481 * glib/gslice.h: renamed g_slice_free_chain() to
482 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
483 a type-safe macro as suggested in bug 323178.
484 simplified the macro implementation of g_slice_free() and implemented
485 it in a type safe manner for all compliers as suggested by Morten
486 Welinder <mortenw@gnome.org>.
491 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
493 2005-12-05 Matthias Clasen <mclasen@redhat.com>
495 * glib/gasyncqueue.c: Add some docs.
497 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
499 * glib/gatomic.c: In the ia64 implementation, use
500 __sync builtin without _si or _di suffix. (#321229,
501 Stanislav Brabec, patch by Andreas Schwab)
503 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
505 * glib/gslice.h: Remove comma at the end of enum.
507 2005-12-04 Matthias Clasen <mclasen@redhat.com>
509 Handle multiple user names with the same UID better.
510 (#319535, Laszlo Peter)
512 * glib/gutils.c (g_get_any_init_do): When determining user
513 data, first look up $LOGNAME. If the UID doesn't match
514 getuid(), fall back to the current behaviour of looking
515 up the user data based on getuid().
517 2005-12-04 Matthias Clasen <mclasen@redhat.com>
519 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
520 to make gcc warn if a function result is ignored. (#145466,
521 Arjan van de Ven, Alex Larsson)
523 * glib/gmem.h: Add the new attribute to g_realloc and
526 2005-12-03 Matthias Clasen <mclasen@redhat.com>
529 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
532 * glib/gmappedfile.c: Make mapping of empty files
535 * glib/gfileutils.c: Don't fork a new process just to
536 fix the permissions of the created temp file. (#321318,
537 Alexis S. L. Carvalho)
539 2005-12-02 Matthias Clasen <mclasen@redhat.com>
541 * README.in: Add a note about Solaris threads.
544 * configure.in: Remove support for Solaris threads.
545 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
547 2005-12-02 Matthias Clasen <mclasen@redhat.com>
549 * configure.in: Define G_GUINT64_CONSTANT in analogy to
550 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
552 2005-12-02 Matthias Clasen <mclasen@redhat.com>
554 * glib/gslice.c: Win32 portability fixes and C99ism removal,
555 pointed out by Kazuki Iwamoto. (#323052)
557 2005-12-02 Matthias Clasen <mclasen@redhat.com>
559 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
560 the IA64 implementation.
562 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
564 * glib/gslice.c: some naming and type size fixups.
566 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
568 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
569 variant into something less verbose, but digestible for gcc-3.4.
571 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
573 * tests/slice-test.c: extended to perform the benchmarking on the old
574 memchunk code if 'O' is selected.
576 * tests/memchunks.c: new file which contains the old GLib mem chunks
577 implementation with prefix old_mem_chunk_.
579 * tests/Makefile.am: added memchunks.c
581 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
583 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
584 acceptable load distribution of the syscall.
585 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
586 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
588 2005-12-01 Matthias Clasen <mclasen@redhat.com>
590 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
591 double-free base interface members.
593 * tests/Makefile.am: Remove duplicate variable.
595 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
597 * glib/gslice.[hc]: new slice allocator implementation.
599 * tests/slice-test.c: added random slice allocation test.
601 * glib/gthread.[hc]: removed newly added private thread mem API.
603 * glib/gthreadinit.h:
606 * glib/gmem.c: divided glib threading initialisation into three phases,
607 initialisation where private keys and messaging are not available (only
608 needed by gmem.c), initialisation without messaging but private keys
609 available (gslice.c, gmessage.c), and full fledged initialisers that
610 server the rest of glib. initialisation functions got renamed to reflect
611 the limitations of their corresponding phases.
613 * glib/gmem.c: removed memchunk code, defer allocations to
616 * glib/gmem.[hc]: removed g_slice_* skeletons.
618 * glib/glib.symbols: added g_slice_* symbols.
620 * configure.in: check for availability of posix_memalign(3), memalign(3)
623 * glib/Makefile.am: added gslice.[hc].
625 2005-12-01 Tor Lillqvist <tml@novell.com>
627 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
628 trailing slash(es) for non-root folders. stat() fails if non-root
629 folders are specified with trailing slashes. It's too much hassle
630 to demand that callers strip such slashes themselves, especially
631 as it is easy to get it wrong and strip the slash of a root
633 (g_rename): On NT-based Windows, use MoveFileEx() with
634 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
636 2005-11-28 Matthias Clasen <mclasen@redhat.com>
638 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
641 * configure.in: Check whether do { } while (0) works.
643 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
644 G_STMT_END if it works.
646 2005-11-28 Matthias Clasen <mclasen@redhat.com>
648 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
649 * glib/gnode.c (g_node_depth_traverse_level):
650 * glib/gmem.c (g_allocator_new):
651 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
652 (g_hash_table_foreach_remove_or_steal): Silence compiler
655 2005-11-27 Matthias Clasen <mclasen@redhat.com>
657 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
658 Don't read beyond len. (#322520, Christian Persch)
660 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
662 * glib/gdataset.c: access datalist flags via atomic pointer access
663 functions, instead of acquiring the dataset lock. this is faster and
664 also matches the atomic pointer readouts in gdatalistprivate.h and
665 direct pointer modifications required by gobject.c.
667 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
669 * glib/gdataset.c: streamlined initialization code somewhat,
670 removed GData node cache.
672 2005-11-22 Matthias Clasen <mclasen@redhat.com>
674 * glib/ghash.c (g_hash_table_ref):
675 (g_hash_table_unref): Mark these as new API.
677 Avoid double locking in g_intern_string (#322133,
680 * glib/gdataset.c (g_quark_from_string_internal): New
681 internal function which factors out common parts of
682 g_quark_from[_static]_string.
683 (g_quark_from_string, g_quark_from_static_string):
684 Use g_quark_from_string_internal.
685 (g_intern_string, g_intern_static_string): Use
686 g_quark_from_string_internal, and only take the
688 (g_quark_new): Don't store the strings shifted by -1
689 in the g_quarks array.
690 (g_quark_to_string): Adapt to the previous change.
692 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
696 g_hash_table_new_full(): create hash tables with a ref count of 1.
697 g_hash_table_ref(): atomically ref_count+=1
698 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
699 when refcount reaches 0.
700 g_hash_table_destroy(): just destroy keys and values, unref by 1.
701 g_hash_table_insert():
702 g_hash_table_replace(): assert ref_count>0.
705 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
708 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
710 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
711 variables. (#321984, Andrew Paprocki)
713 2005-11-18 Matthias Clasen <mclasen@redhat.com>
715 * configure.in: Bump version
717 2005-11-17 Matthias Clasen <mclasen@redhat.com>
719 * === Released 2.9.0 ===
721 * glib/glib.symbols: Updates
723 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
724 hints to the docs. (Paolo Borelli)
728 * glib/gthread.c (g_thread_foreach): Mark as new api.
730 * README.in: Updates.
734 * glib/gdate.c (g_date_set_time_t):
735 (g_date_set_time_val): Add functions to set a date from
736 a time_t and from a GTimeVal.
737 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
740 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
741 and g_utf8_offset_to_pointer.
743 * glib/gutf8.c (g_utf8_pointer_to_offset)
744 (g_utf8_offset_to_pointer): Handle negative offsets, and use
745 "stutter stepping" for going backwards. (#320638, Larry
750 * glib/gthread.c: const correctness fixes, found
751 by Arjan van de Ven and gcc.
753 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
755 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
756 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
758 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
760 * glib/guniprop.c: Use bit hacks instead when checking a general
761 category value against multiple values.
763 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
765 2005-11-10 Simos Xenitellis <simos@gnome.org>
767 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
769 2005-11-05 Tor Lillqvist <tml@novell.com>
771 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
772 (#320688) Instead set error and return failure.
774 2005-11-07 Matthias Clasen <mclasen@redhat.com>
776 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
777 while holding the context lock. (#320886, Andy Wingo)
779 2005-11-07 Billy Biggs <vektor@dumbterm.net>
781 * tests/refcount/Makefile.am: Add a missing reference to libglib
782 in the LDADD for this test.
784 2005-11-07 Matthias Clasen <mclasen@redhat.com>
786 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
787 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
788 GLib 2.9. (#320755, Luca Ferretti)
790 2005-11-04 Matthias Clasen <mclasen@redhat.com>
792 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
793 use a faster lookup table for g_unichar_get_mirror_char().
794 * glib/gmirroringtable.h: The new table.
795 * glib/gunichartables.h: Remove bidi_mirroring_table.
796 * glib/gen-unicode-tables.pl: Don't generate the mirroring
798 * glib/glib-mirroring-tab/*: Sources for the program
799 which generated gmirroringtable.h.
801 * glib/glist.c: Avoid some code duplication.
803 * glib/gscanner.h: Include gdataset.h. (#320322)
805 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
806 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
807 to help find critical warnings in applications. (#320017,
810 2005-11-02 Tor Lillqvist <tml@novell.com>
812 * glib/glib.symbols: Remove large amount of trailing whitespace
813 from one line. Remove #ifdef G_UNDEFINED from around
814 g_hash_table_new and g_hash_table_new_full symbols, what was the
815 purpose of that? Add the g_slice functions.
817 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
819 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
820 added g_slice_*() API to allocate and cache small bits of memory.
821 an actuall allocator implementation for g_slice_*() is still pending.
823 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
824 changed GRealThread list to use in-structure *next; fields instead
825 of GSList, in order for thread iteration to not depenend on g_slice_*()
827 _g_thread_mem_private_get():
828 _g_thread_mem_private_set(): added accessors for private memory,
829 needed because the ordinary GPrivate implementation relies on GArray
830 and GSList and therefore indirectly on working g_slice_*() allocations.
833 g_thread_foreach(): new public API function to loop over all existing threads.
847 removed GAllocator and free list usages and accompanying locks.
848 use g_slice_*() API to allocate and cache small bits of memory.
850 * glib/ghook.h: removed GMemChunk field from public API.
853 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
855 * glib/gnode.h: deprecate allocator API.
857 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
858 be used for poll rec lists.
860 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
861 g_relation_destroy(): free all tuples from the all_tuples hash table,
862 this effectively maintains the life time track keeping of tuples.
863 g_relation_delete_tuple(): free tuples which are removed from the
864 all_tuples hash table. this fixes a temporary leak that was present
865 in the memchunk code until the destruction of the relation.
867 2005-10-29 Matthias Clasen <mclasen@redhat.com>
869 * tests/convert-test.c: Add some tests for conversions between
870 UTF-8, UCS-4 and UTF-16.
872 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
873 of len == -1, noticed by Morten Welinder.
875 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
877 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
879 2005-10-26 Matthias Clasen <mclasen@redhat.com>
881 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
882 in the error case, and add some documentation. (#319806, Morten
885 2005-10-19 Manish Singh <yosh@gimp.org>
887 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
888 config.h symbol automatically. Fixes bug #313731.
890 2005-10-19 Tor Lillqvist <tml@novell.com>
892 * glib/gwin32.c (g_win32_get_package_installation_directory):
893 Return a g_strdup()ed copy of the value stored in the hash table,
894 so that it can be g_free()d without leaving a dangling pointer in
895 the hash table. (#319232)
897 2005-10-06 Matthias Clasen <mclasen@redhat.com>
899 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
900 all-zero sequences correctly. (#317930, Sebastien Bacher)
902 2005-10-05 Matthias Clasen <mclasen@redhat.com>
904 * glib/gmarkup.c (g_markup_escape_text): Doc addition
906 * Makefile.am: Add Collation tests.
908 * tests/collate/*: Inputs and expected outputs for collation tests.
910 * tests/run-collate-tests.sh: Script to run collation tests.
912 * tests/unicode-collate.c (main): Rework slightly to make
913 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
915 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
917 * docs/reference/glib/tmpl/unicode.sgml:
918 * glib/gen-unicode-tables.pl:
920 * glib/gunichartables.h:
922 * tests/casefold.txt:
923 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
924 GUnicodeBreakType types. That may break some applications, like
927 2005-09-26 Matthias Clasen <mclasen@redhat.com>
929 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
930 docs. (#314393, Matthew F. Barnes)
933 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
934 they are already declared in glib.h. This doesn't break documented
935 use of gprintf.h, but should probably be pointed out in the
936 release notes for 2.10. (#314232, Behdad Esfahbod)
938 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
940 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
941 complexity optimization, based on a patch from Matthias Clasen.
943 * tests/patterntest.c: more tests, mostly from matthias.
945 2005-09-20 Matthias Clasen <mclasen@redhat.com>
947 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
948 (#316703, Mark Drago)
950 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
952 * glib/gpattern.c: applied a patch from matthias which checks on the
953 upper bound of GPatternSpec length to optimize matches.
956 * tests/patterntest.c: added more match cases.
958 2005-09-16 Tor Lillqvist <tml@novell.com>
960 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
961 warning to doc comment that these functions should not be used on
962 encodings like CP932.
964 2005-09-14 Matthias Clasen <mclasen@redhat.com>
966 * tests/keyfile-test.c: Add a test for grup names of length 1.
968 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
969 of length 1. (#316309)
971 2005-09-12 Matthias Clasen <mclasen@redhat.com>
973 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
975 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
977 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
978 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
981 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
983 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
984 (child_watch_helper_thread): Remove some dead code. Closes
987 2005-09-07 Tor Lillqvist <tml@novell.com>
989 * glib/Makefile.am: Create also a console version of the
990 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
991 It's otherwise identical to gspawn-win32-helper.exe, except marked
992 as a console application (linked without the -mwindows option).
994 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
995 Drop the dont_wait parameter. Its truth value correlated 100% with
996 the NULLness of the exit_status parameter anyway, so it's enough
997 to check whether exit_status is NULL. Invert the sense of the
998 dont_return_handle parameter and rename it to do_return_handle, to
999 make the code easier to read by avoiding double negations.
1001 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1002 to do_spawn_with_pipes() accordingly.
1004 (do_spawn_with_pipes): If we have a console, use the console
1005 version of the helper program, otherwise use the GUI one. This
1006 avoids extra console windows opening up in some situations. (In
1007 case a console application uses the GUI gspawn-win32-helper.exe to
1008 spawn another console application we would get a separate console
1009 for the spawned console application).
1011 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1013 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1015 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1016 if the file is too large. (#315275, Kjartan Maraas)
1018 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1019 of read() is signed. (#315273, Kjartan Maraas)
1021 2005-08-31 Tor Lillqvist <tml@novell.com>
1023 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1024 the EXE and each DLL have their own atexit function chains.
1026 #define g_atexit as atexit instead. This means it has a
1027 better chance of doing what the caller wants. For instance,
1028 gtkhtml calls g_atexit() registering a function in gtkhtml
1029 itself. This caused a crash when g_atexit() was implemented as a
1030 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1031 the time the GLib DLL got unloaded.
1033 * glib/gutils.c: #undef the #define mentioned above, to also get a
1034 real g_atexit() into the DLL for backward compatibility. Document
1035 the Windows behaviour of g_atexit(), and document the varying ways
1036 atexit() can behave in the context of dynamically loaded modules
1039 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1041 * glib/glib.symbols:
1043 * glib/gdataset.c: Add string interning functions.
1045 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1047 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1050 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1052 * configure.in: Bump version to 2.9.0
1054 2005-08-25 Tor Lillqvist <tml@novell.com>
1056 Make also the g_spawn*() functions take parameters in the GLib
1057 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1058 API or ABI. Like the other GLib API that was earlier changed to
1059 use UTF-8 on Windows, the names of the functions that take UTF-8
1060 have _utf8 suffixes added by using preprocessor macros in the
1061 header file. The old names are kept for functions with the old
1062 behaviour, taking parameters in the system codepage, for DLL ABI
1065 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1066 g_spawn*() functions.
1068 * glib/gspawn-win32.c: Use wide-char API on NT-based
1069 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1070 system codepage (Win9x) and call the C library _wspawn*() or
1071 spawn*() functions respectvely. Add DLL ABI stability versions
1072 that take parameters in the system codepage.
1074 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1075 wide-char versions of argv and envp, and use wide-char API to
1076 change directory and spawn the program to run. Remove the verbose
1077 debugging output, it was too complex to modify for the wide-char
1078 features. (Just add temporary debugging printouts if needed, no
1079 need to have them permanently in the source.)
1081 * glib/gspawn.c: Corresponding documentation updates.
1083 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1084 symbols as PRIVATE, add the new _utf8-suffixed ones.
1086 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1088 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1089 guint64 otherwise and that can produce warnings about comparison
1090 between signed and unsigned.
1092 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1094 * glib/gutils.c: Fix the crt_externs.h include.
1096 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1098 * NEWS: Fix spelling of my first name.
1100 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1104 * === Released 2.8.1 ===
1108 2005-08-20 Hans Breuer <hans@breuer.org>
1110 * glib/makefile.msc.in : link with ws2_32.lib
1112 2005-08-18 Tor Lillqvist <tml@novell.com>
1114 * configure.in: Check for <sys/wait.h>
1116 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1118 2005-08-18 Ross Burton <ross@burtonini.com>
1121 Optimise single-character insertions.
1126 * tests/string-test.c:
1127 Add tests for new optimisation, and fix a leak.
1129 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1131 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1133 * glib/gutils.c: On Darwin, include crt-externs.h and
1134 define environ using _NSGetEnviron(). (#313731)
1136 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1138 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1139 (g_get_any_init_do): ... to this new function.
1140 (g_get_any_init): Declare as inline.
1141 (g_get_any_init_locked): New inline function, does the locking.
1142 Make use of these two throughout the code.
1144 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1146 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1147 the child process and then simply return. This makes
1148 The "S" option work as documented in g_on_error_query().
1149 (#313125, Matthew F. Barnes)
1151 * glib/gunicode.h: Update the link to Unicode category
1152 values. (#313369, Behnam Esfahbod)
1154 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1155 a little. (#311727, Tristan van Berkom)
1157 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1158 check work on ia64 too, where some symbols ended up
1159 in yet another section.