1 2006-03-09 Matthias Clasen <mclasen@redhat.com>
3 * Makefile.am: Add ChangeLog.pre-2.8 to EXTRA_DIST.
5 2006-03-08 Tor Lillqvist <tml@novell.com>
7 * glibconfig.h.win32.in: Add G_GUINT64_CONSTANT.
9 2006-03-07 Matthias Clasen <mclasen@redhat.com>
11 * configure.in: Bump version
13 * === Released 2.10.1 ===
17 * glib/gutf8.c (_g_charset_get_aliases): Match the prototype
18 in gconvert.c to fix build problems in NetBSD. (#333651,
19 Julio M. Merino Vidal)
21 2006-03-06 Anders Carlsson <andersca@imendio.com>
24 Make .dylib the shared library suffix on darwin.
25 (#322476, Vladimir Panov)
27 2006-03-06 Anders Carlsson <andersca@imendio.com>
30 Skip checking thread flags on Darwin. Fix suggested by Manish Singh.
31 (#314794, Gregor Riepl)
33 2006-03-01 Tor Lillqvist <tml@novell.com>
35 * glib/gutils.c (g_listenv): Separate implementation on Win32: Use
36 the wide character API on NT-based Windows. Return UTF-8 strings.
38 * glib/glib.symbols: Don't mark g_listenv as PRIVATE, as that
39 meant it wasn't present in the import library. PRIVATE is used
40 only for the backwards-compatibility DLL ABI stability hacks.
42 2006-02-24 Matthias Clasen <mclasen@redhat.com>
44 * configure.in: Bump version
46 * === Released 2.10.0 ===
48 * README.in: Mention the slice allocator and GInitiallyUnowned.
52 * configure.in: Bump version to 2.10.0
54 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
56 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
59 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
61 * tests/completion-test.c: (main): Plug leaks reported by
63 * tests/convert-test.c: (test_iconv_state), (test_one_half):
65 * tests/patterntest.c: (test_compilation): Same
66 * tests/shell-test.c: (do_argv_test): Same.
67 * tests/unicode-caseconv.c: (main): Same.
68 * tests/uri-test.c: (run_to_uri_tests): Same.
71 2006-02-19 Martyn Russell <martyn@imendio.com>
73 * tests/asyncqueue-test.c:
75 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
76 _insert_sorted and _insert_sorted_with_data API.
78 2006-02-18 Matthias Clasen <mclasen@redhat.com>
80 * tests/gobject/Makefile.am: Add paramspec-test
82 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
84 * tests/gobject/gvalue-test.c: Add more tests.
86 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
88 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
90 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
92 * glib/gthreadpool.c: Fix deadlock when signalling the thread
93 which freed a thread pool (#331110, Chris Wilson).
95 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
97 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
98 posix_memalign() in case we actually are going to use it, because
99 we detected a compliant implementation (#328997).
101 * configure.in (enable_included_printf): don't include malloc.h when
102 testing for posix_memalign() funcitonality, since this may break the
103 test on some systems (#328997).
105 2006-02-14 Matthias Clasen <mclasen@redhat.com>
107 * glib/gstrfuncs.c: Improve docs.
109 2006-02-11 Matthias Clasen <mclasen@redhat.com>
111 * configure.in: Bump version
113 2006-02-10 Matthias Clasen <mclasen@redhat.com>
115 * === Released 2.9.6 ===
119 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
121 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
122 the type of varargs arguments. (bug #317679)
124 2006-01-31 Matthias Clasen <mclasen@redhat.com>
126 * autogen.sh: Touch README and INSTALL here to pacify
127 automake. (#329124, Kjartan Maraas, Tim Janik)
129 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
131 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
132 some missing cases where pointed out by Sebastian Wilhelmi.
134 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
136 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
138 2006-01-30 David Schleef <ds@schleef.org>
140 * glib/grand.c: Update URL
142 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
144 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
145 create container classes which are as flexible as a GValue is
147 2006-01-27 Matthias Clasen <mclasen@redhat.com>
149 * configure.in: Bump version
151 * === Released 2.9.5 ===
153 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
158 2006-01-26 Matthias Clasen <mclasen@redhat.com>
160 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
163 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
165 * fixed buglets reported by Jens Granseuer in #328254.
167 * configure.in: free the memory allocated in posix_memalign() tests.
169 * glib/gslice.c: spelling fixes.
171 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
173 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
174 slices, make sure g_mem_gc_friendly is properly initialized.
176 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
177 the first allocation. applied some branching optimizations.
179 * docs/macros.txt: reflected --enable-gc-friendly change and
180 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
182 * configure.in: changed --enable-gc-friendly=yes to define
183 ENABLE_GC_FRIENDLY_DEFAULT.
185 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
186 if (G_UNLIKELY (g_mem_gc_friendly)).
189 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
190 care of by g_slice_free1().
192 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
194 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
196 * glib/gslice.c: only use posix_memalign() if it's known to work,
197 revert to memalign() otherwise.
199 * configure.in: check for broken posix_memalign() implementations
202 2006-01-24 Matthias Clasen <mclasen@redhat.com>
204 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
205 on all little-endian systems. (#143380, Marc Moorcroft)
207 2006-01-23 Matthias Clasen <mclasen@redhat.com>
209 * configure.in: Change the shared libary extension for hpux-ia64
210 to so. (#328253, The Written Word)
212 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
214 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
215 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
218 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
220 * glib/gslice.c (slice_config_init): initialize GSlice config from
221 G_SLICE environment variable. we support G_SLICE=always-malloc
222 currently, which forces all g_slice_*() allocations to use the system
226 g_parse_debug_string(): added a note about not using g_malloc() here.
227 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
228 g_slice. contains only guesswork in the WIN32 branch.
230 2006-01-18 Matthias Clasen <mclasen@redhat.com>
234 * === Released 2.9.4 ===
238 2006-01-17 Matthias Clasen <mclasen@redhat.com>
240 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
241 call. (#327421, Morten Welinder)
243 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
245 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
246 G_LOCK. Use the unused_thread_queue lock instead. Change
247 g_thread_pool_thread_proxy such that threads only wait on
248 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
249 due to superfluous tasks. Global tasks wait at most for
250 max-idle-time milliseconds. Make sure, that no task is woken up
251 twice for the same event via a wakeup_serial. This fixes #324228.
253 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
254 invalid NULL into the thread pools. This as well fixes #327290.
256 2006-01-16 Matthias Clasen <mclasen@redhat.com>
258 * configure.in: Bump version
260 * === Released 2.9.3 ===
264 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
265 Rework the strict aliasing fix to not break C++, pointed
266 out by Murray Cumming.
268 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
269 Signal waiting threads, problem noticed by Christian Kellner.
271 2006-01-16 Matthias Clasen <mclasen@redhat.com>
273 Fix bug 326747, Alberto Ruiz:
275 * glib/gconvert.c (make_valid_utf8): Change this function to
276 replace unknown characters by the Unicode replacement character
277 instead of '?', and don't append "(invalid encoding)".
278 (g_filename_display_name, g_filename_display_basename): Document
279 how to determine if the filename was in an invalid encoding.
281 2006-01-14 Matthias Clasen <mclasen@redhat.com>
283 * glib/gtree.c: Replace the simple recursive implementation
284 by a nonrecursive, threaded implementation by Maurizio
287 2006-01-12 Matthias Clasen <mclasen@redhat.com>
289 * tests/asyncqueue-test.c (main): Fix the
290 assert to not always trigger. (#326558,
293 2006-01-11 Matthias Clasen <mclasen@redhat.com>
295 * tests/convert-test.c: Don't test items_read and
296 items_written in error cases where it is not set.
298 * tests/asyncqueue-test.c: Set error to NULL.
300 2006-01-08 Matthias Clasen <mclasen@redhat.com>
302 * glib/gtree.c: Break some long lines.
304 * tests/tree-test.c: Add more tests.
306 * glib/gslice.c: Include config.h first, pointed out
309 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
311 * docs/reference/gobject/tmpl/gtype.sgml: explain that
312 G_TYPE_INSTANCE_GET_CLASS() does behave different during
315 2006-01-06 Matthias Clasen <mclasen@redhat.com>
317 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
318 that we have posix_memalign(). (#323937, Bogdan Nicula)
320 2006-01-05 Matthias Clasen <mclasen@redhat.com>
322 * configure.in: Bump version
324 * === Released 2.9.2 ===
326 2006-01-05 Hans Breuer <hans@breuer.org>
328 * glib/makefile.msc.in : added gslice.obj
330 2006-01-05 Matthias Clasen <mclasen@redhat.com>
334 * glib/gmain.c: Clarify the documentation of source ids
335 a bit. (#325874, Dan Williams)
337 * configure.in: Fix another strict aliasing problem.
339 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
342 2006-01-04 Manish Singh <yosh@gimp.org>
345 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
346 either, since they have the same issue as g_rmdir. (Related to
349 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
351 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
352 order when creating magazines, so we hand out chunks with ascending
355 2006-01-04 Abel Cheung <maddog@linuxhall.org>
357 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
359 2006-01-03 Martyn Russell <martyn@imendio.com>
361 * docs/reference/glib/glib-sections.txt:
363 * glib/gthreadpool.[ch]:
364 - Added new API g_thread_pool_get_idle_time() and
365 g_thread_pool_set_idle_time(). (#324228).
367 * tests/threadpool-test.c:
368 - Updated test case to do thread pool sorting, thread pool with
369 no sorting and a thread pool with idle thread timeouts.
371 2006-01-03 Matthias Clasen <mclasen@redhat.com>
373 * glib/gmain.h: Add new functions here, too.
375 * glib/glib.symbols: Add new functions.
377 2005-12-20 Michael Meeks <michael.meeks@novell.com>
379 * glib/gmain.c (g_main_context_is_owner): new method
380 to determine if the current thread is the owner of the
383 2006-01-02 Matthias Clasen <mclasen@redhat.com>
387 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
388 expanding to rmdir, since rmdir is not declared in a portable
389 system header we can include in gstdio.h. (#325249, Jani Monoses)
391 2006-01-01 Tor Lillqvist <tml@novell.com>
393 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
394 stdout and stderr to unbuffered. Otherwise the giochannel layer
395 will try to read from them regardless whether the
396 g_io_channel_win32_poll() call here has indicated
397 readability or not. (#325310)
399 2005-12-29 Matthias Clasen <mclasen@redhat.com>
401 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
402 in the docs. (#325273, Declan Naughton)
404 2005-12-27 Manish Singh <yosh@gimp.org>
406 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
407 make sure we really override things.
409 2005-12-27 Matthias Clasen <mclasen@redhat.com>
411 Fix #316221, Michal Benes, Stanislav Brabec;
413 * configure.in: Fix a strict aliasing problem in
414 g_static_mutex_get_mutex().
415 * glib/gthread.h: ...and in
416 g_static_mutex_get_mutex_impl_shortcut().
418 * glib/gdatasetprivate.h: Add a cast to silence compiler
419 warnings. (#321978, Andrew Paprocki)
421 Partial fix for bug #323937, Bogdan Nicula.
423 * configure.in: Check for malloc.h
424 * glib/gslice.c: Don't include malloc.h unconditionally.
426 2005-12-27 Manish Singh <yosh@gimp.org>
428 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
429 (#324950, Dan Yefimov)
431 2005-12-26 Matthias Clasen <mclasen@redhat.com>
433 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
436 2005-12-25 Matthias Clasen <mclasen@redhat.com>
438 * AUTHORS: Update my email
440 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
442 2005-12-21 Matthias Clasen <mclasen@redhat.com>
444 * README.in: Add some notes on when 'make check' may fail.
446 2005-12-20 Matthias Clasen <mclasen@redhat.com>
449 * glib/gthreadpool.h:
450 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
451 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
453 * tests/threadpool-test.c: Test this.
455 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
457 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
458 handle errors without depending on gmessage.c which might not be
459 setup when the error occours.
460 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
461 always freeing can be achieved by adjusting the working set time to
462 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
463 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
464 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
465 minimize wastage if small amounts of differently sized structrues are
466 allocated, this does come at a performance cost of roughly 5% though.
467 fixed up block alignment calculation, so it works for varying
468 block sizes. only use strerror() not g_strerror() since the latter
469 depends on working GQuark and GSlice.
470 mem_error(): implemented in terms of fprintf and vfprintf.
472 * tests/slice-color.c: new program to test cache colorization effects.
474 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
475 G_SLICE_CONFIG_WORKING_SET_MSECS.
477 2005-12-17 Matthias Clasen <mclasen@redhat.com>
479 * glib/goption.c (parse_short_option): Set an error in all
480 failure cases. (#324332, Tim-Philipp Müller)
482 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
484 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
485 and g_atomic_pointer_get. Add g_atomic_int_set and
486 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
487 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
488 as functions (additionally to the macros in the header) for the
489 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
491 2005-12-16 Matthias Clasen <mclasen@redhat.com>
493 * glib/gmem.c (g_allocator_new): Don't return a pointer to
494 a const struct, since apps expect to be able to modify it.
495 (#324179, J. Ali Harlow)
497 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
499 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
500 for systems where the initialization of atomic variables requires a
501 write memory barrier.
503 2005-12-09 Matthias Clasen <mclasen@redhat.com>
505 * configure.in: Bump version
507 * === Released 2.9.1 ===
509 2005-12-09 Alexander Larsson <alexl@redhat.com>
513 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
516 2005-12-08 Matthias Clasen <mclasen@redhat.com>
520 2005-12-07 Martyn Russell <martyn@imendio.com>
522 * glib/gasyncqueue.c:
523 - Call g_queue_insert_sorted() instead of duplicating the code.
524 - Call g_queue_sort() instead of duplicating the code.
525 - Invert sort function results to make sure the same sort function
526 gives the same results across glist, gslist, gqueue and
529 * tests/asyncqueue-test.c:
530 - Updated the sort function to reflect the example in the
531 documentation for gasyncqueue.c.
533 2005-12-07 Martyn Russell <martyn@imendio.com>
535 * docs/reference/glib/glib-sections.txt:
536 * docs/reference/glib/tmpl/linked_lists_double.sgml:
537 * docs/reference/glib/tmpl/linked_lists_single.sgml:
540 - Added g_list_insert_sorted_with_data () and
541 g_slist_insert_sorted_with_data ().
542 - Removed the extra check in g_list_sort() and g_slist_sort() for
543 GCompareDataFunc vs. GCompareFunc.
545 2005-12-07 Tor Lillqvist <tml@novell.com>
547 * glib-gettextize.in: Look up prefix at run-time on Win32,
548 assuming the standard directory structure with glib-gettextize in
551 2005-12-06 Matthias Clasen <mclasen@redhat.com>
553 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
554 we need GTK+ 2.8 to compile against GLib 2.10.
556 2005-12-05 Manish Singh <yosh@gimp.org>
558 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
559 to override any user settings.
561 2005-12-05 Matthias Clasen <mclasen@redhat.com>
563 * configure.in: Revert an accidential commit
565 2005-12-05 Matthias Clasen <mclasen@redhat.com>
567 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
569 * tests/Makefile.am (TESTS_ENVIRONMENT):
570 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
571 MALLOC_CHECK_ and MALLOC_PERTURB_
573 * tests/run-collate-tests.sh: Run the collation tests explicitly
574 in en_US locale. (#320463)
576 * glib/gmem.h: Really deprecate GMemChunk.
583 * glib/gstrfuncs.c: Add versioned deprecation docs.
585 2005-12-05 Martyn Russell <martyn@imendio.com>
587 * docs/reference/glib/glib-sections.txt:
588 * glib/gasyncqueue.[ch]:
589 - Added support for sorting async queues by with _push_sorted(),
590 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
593 * tests/asyncqueue-test.c:
594 - Added test case for gasyncqueue.c
596 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
598 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
599 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
600 in bug 323178. moved time consuming logic from g_slice_free() out of
601 the inner loop, so g_slice_free_chain_with_offset() provides a real
602 performance benefit over g_slice_free1() now.
604 * glib/gslice.h: renamed g_slice_free_chain() to
605 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
606 a type-safe macro as suggested in bug 323178.
607 simplified the macro implementation of g_slice_free() and implemented
608 it in a type safe manner for all compliers as suggested by Morten
609 Welinder <mortenw@gnome.org>.
614 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
616 2005-12-05 Matthias Clasen <mclasen@redhat.com>
618 * glib/gasyncqueue.c: Add some docs.
620 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
622 * glib/gatomic.c: In the ia64 implementation, use
623 __sync builtin without _si or _di suffix. (#321229,
624 Stanislav Brabec, patch by Andreas Schwab)
626 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
628 * glib/gslice.h: Remove comma at the end of enum.
630 2005-12-04 Matthias Clasen <mclasen@redhat.com>
632 Handle multiple user names with the same UID better.
633 (#319535, Laszlo Peter)
635 * glib/gutils.c (g_get_any_init_do): When determining user
636 data, first look up $LOGNAME. If the UID doesn't match
637 getuid(), fall back to the current behaviour of looking
638 up the user data based on getuid().
640 2005-12-04 Matthias Clasen <mclasen@redhat.com>
642 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
643 to make gcc warn if a function result is ignored. (#145466,
644 Arjan van de Ven, Alex Larsson)
646 * glib/gmem.h: Add the new attribute to g_realloc and
649 2005-12-03 Matthias Clasen <mclasen@redhat.com>
652 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
655 * glib/gmappedfile.c: Make mapping of empty files
658 * glib/gfileutils.c: Don't fork a new process just to
659 fix the permissions of the created temp file. (#321318,
660 Alexis S. L. Carvalho)
662 2005-12-02 Matthias Clasen <mclasen@redhat.com>
664 * README.in: Add a note about Solaris threads.
667 * configure.in: Remove support for Solaris threads.
668 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
670 2005-12-02 Matthias Clasen <mclasen@redhat.com>
672 * configure.in: Define G_GUINT64_CONSTANT in analogy to
673 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
675 2005-12-02 Matthias Clasen <mclasen@redhat.com>
677 * glib/gslice.c: Win32 portability fixes and C99ism removal,
678 pointed out by Kazuki Iwamoto. (#323052)
680 2005-12-02 Matthias Clasen <mclasen@redhat.com>
682 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
683 the IA64 implementation.
685 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
687 * glib/gslice.c: some naming and type size fixups.
689 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
691 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
692 variant into something less verbose, but digestible for gcc-3.4.
694 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
696 * tests/slice-test.c: extended to perform the benchmarking on the old
697 memchunk code if 'O' is selected.
699 * tests/memchunks.c: new file which contains the old GLib mem chunks
700 implementation with prefix old_mem_chunk_.
702 * tests/Makefile.am: added memchunks.c
704 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
706 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
707 acceptable load distribution of the syscall.
708 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
709 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
711 2005-12-01 Matthias Clasen <mclasen@redhat.com>
713 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
714 double-free base interface members.
716 * tests/Makefile.am: Remove duplicate variable.
718 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
720 * glib/gslice.[hc]: new slice allocator implementation.
722 * tests/slice-test.c: added random slice allocation test.
724 * glib/gthread.[hc]: removed newly added private thread mem API.
726 * glib/gthreadinit.h:
729 * glib/gmem.c: divided glib threading initialisation into three phases,
730 initialisation where private keys and messaging are not available (only
731 needed by gmem.c), initialisation without messaging but private keys
732 available (gslice.c, gmessage.c), and full fledged initialisers that
733 server the rest of glib. initialisation functions got renamed to reflect
734 the limitations of their corresponding phases.
736 * glib/gmem.c: removed memchunk code, defer allocations to
739 * glib/gmem.[hc]: removed g_slice_* skeletons.
741 * glib/glib.symbols: added g_slice_* symbols.
743 * configure.in: check for availability of posix_memalign(3), memalign(3)
746 * glib/Makefile.am: added gslice.[hc].
748 2005-12-01 Tor Lillqvist <tml@novell.com>
750 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
751 trailing slash(es) for non-root folders. stat() fails if non-root
752 folders are specified with trailing slashes. It's too much hassle
753 to demand that callers strip such slashes themselves, especially
754 as it is easy to get it wrong and strip the slash of a root
756 (g_rename): On NT-based Windows, use MoveFileEx() with
757 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
759 2005-11-28 Matthias Clasen <mclasen@redhat.com>
761 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
764 * configure.in: Check whether do { } while (0) works.
766 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
767 G_STMT_END if it works.
769 2005-11-28 Matthias Clasen <mclasen@redhat.com>
771 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
772 * glib/gnode.c (g_node_depth_traverse_level):
773 * glib/gmem.c (g_allocator_new):
774 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
775 (g_hash_table_foreach_remove_or_steal): Silence compiler
778 2005-11-27 Matthias Clasen <mclasen@redhat.com>
780 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
781 Don't read beyond len. (#322520, Christian Persch)
783 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
785 * glib/gdataset.c: access datalist flags via atomic pointer access
786 functions, instead of acquiring the dataset lock. this is faster and
787 also matches the atomic pointer readouts in gdatalistprivate.h and
788 direct pointer modifications required by gobject.c.
790 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
792 * glib/gdataset.c: streamlined initialization code somewhat,
793 removed GData node cache.
795 2005-11-22 Matthias Clasen <mclasen@redhat.com>
797 * glib/ghash.c (g_hash_table_ref):
798 (g_hash_table_unref): Mark these as new API.
800 Avoid double locking in g_intern_string (#322133,
803 * glib/gdataset.c (g_quark_from_string_internal): New
804 internal function which factors out common parts of
805 g_quark_from[_static]_string.
806 (g_quark_from_string, g_quark_from_static_string):
807 Use g_quark_from_string_internal.
808 (g_intern_string, g_intern_static_string): Use
809 g_quark_from_string_internal, and only take the
811 (g_quark_new): Don't store the strings shifted by -1
812 in the g_quarks array.
813 (g_quark_to_string): Adapt to the previous change.
815 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
819 g_hash_table_new_full(): create hash tables with a ref count of 1.
820 g_hash_table_ref(): atomically ref_count+=1
821 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
822 when refcount reaches 0.
823 g_hash_table_destroy(): just destroy keys and values, unref by 1.
824 g_hash_table_insert():
825 g_hash_table_replace(): assert ref_count>0.
828 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
831 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
833 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
834 variables. (#321984, Andrew Paprocki)
836 2005-11-18 Matthias Clasen <mclasen@redhat.com>
838 * configure.in: Bump version
840 2005-11-17 Matthias Clasen <mclasen@redhat.com>
842 * === Released 2.9.0 ===
844 * glib/glib.symbols: Updates
846 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
847 hints to the docs. (Paolo Borelli)
851 * glib/gthread.c (g_thread_foreach): Mark as new api.
853 * README.in: Updates.
857 * glib/gdate.c (g_date_set_time_t):
858 (g_date_set_time_val): Add functions to set a date from
859 a time_t and from a GTimeVal.
860 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
863 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
864 and g_utf8_offset_to_pointer.
866 * glib/gutf8.c (g_utf8_pointer_to_offset)
867 (g_utf8_offset_to_pointer): Handle negative offsets, and use
868 "stutter stepping" for going backwards. (#320638, Larry
873 * glib/gthread.c: const correctness fixes, found
874 by Arjan van de Ven and gcc.
876 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
878 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
879 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
881 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
883 * glib/guniprop.c: Use bit hacks instead when checking a general
884 category value against multiple values.
886 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
888 2005-11-10 Simos Xenitellis <simos@gnome.org>
890 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
892 2005-11-05 Tor Lillqvist <tml@novell.com>
894 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
895 (#320688) Instead set error and return failure.
897 2005-11-07 Matthias Clasen <mclasen@redhat.com>
899 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
900 while holding the context lock. (#320886, Andy Wingo)
902 2005-11-07 Billy Biggs <vektor@dumbterm.net>
904 * tests/refcount/Makefile.am: Add a missing reference to libglib
905 in the LDADD for this test.
907 2005-11-07 Matthias Clasen <mclasen@redhat.com>
909 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
910 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
911 GLib 2.9. (#320755, Luca Ferretti)
913 2005-11-04 Matthias Clasen <mclasen@redhat.com>
915 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
916 use a faster lookup table for g_unichar_get_mirror_char().
917 * glib/gmirroringtable.h: The new table.
918 * glib/gunichartables.h: Remove bidi_mirroring_table.
919 * glib/gen-unicode-tables.pl: Don't generate the mirroring
921 * glib/glib-mirroring-tab/*: Sources for the program
922 which generated gmirroringtable.h.
924 * glib/glist.c: Avoid some code duplication.
926 * glib/gscanner.h: Include gdataset.h. (#320322)
928 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
929 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
930 to help find critical warnings in applications. (#320017,
933 2005-11-02 Tor Lillqvist <tml@novell.com>
935 * glib/glib.symbols: Remove large amount of trailing whitespace
936 from one line. Remove #ifdef G_UNDEFINED from around
937 g_hash_table_new and g_hash_table_new_full symbols, what was the
938 purpose of that? Add the g_slice functions.
940 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
942 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
943 added g_slice_*() API to allocate and cache small bits of memory.
944 an actuall allocator implementation for g_slice_*() is still pending.
946 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
947 changed GRealThread list to use in-structure *next; fields instead
948 of GSList, in order for thread iteration to not depenend on g_slice_*()
950 _g_thread_mem_private_get():
951 _g_thread_mem_private_set(): added accessors for private memory,
952 needed because the ordinary GPrivate implementation relies on GArray
953 and GSList and therefore indirectly on working g_slice_*() allocations.
956 g_thread_foreach(): new public API function to loop over all existing threads.
970 removed GAllocator and free list usages and accompanying locks.
971 use g_slice_*() API to allocate and cache small bits of memory.
973 * glib/ghook.h: removed GMemChunk field from public API.
976 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
978 * glib/gnode.h: deprecate allocator API.
980 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
981 be used for poll rec lists.
983 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
984 g_relation_destroy(): free all tuples from the all_tuples hash table,
985 this effectively maintains the life time track keeping of tuples.
986 g_relation_delete_tuple(): free tuples which are removed from the
987 all_tuples hash table. this fixes a temporary leak that was present
988 in the memchunk code until the destruction of the relation.
990 2005-10-29 Matthias Clasen <mclasen@redhat.com>
992 * tests/convert-test.c: Add some tests for conversions between
993 UTF-8, UCS-4 and UTF-16.
995 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
996 of len == -1, noticed by Morten Welinder.
998 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
1000 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
1002 2005-10-26 Matthias Clasen <mclasen@redhat.com>
1004 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
1005 in the error case, and add some documentation. (#319806, Morten
1008 2005-10-19 Manish Singh <yosh@gimp.org>
1010 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
1011 config.h symbol automatically. Fixes bug #313731.
1013 2005-10-19 Tor Lillqvist <tml@novell.com>
1015 * glib/gwin32.c (g_win32_get_package_installation_directory):
1016 Return a g_strdup()ed copy of the value stored in the hash table,
1017 so that it can be g_free()d without leaving a dangling pointer in
1018 the hash table. (#319232)
1020 2005-10-06 Matthias Clasen <mclasen@redhat.com>
1022 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
1023 all-zero sequences correctly. (#317930, Sebastien Bacher)
1025 2005-10-05 Matthias Clasen <mclasen@redhat.com>
1027 * glib/gmarkup.c (g_markup_escape_text): Doc addition
1029 * Makefile.am: Add Collation tests.
1031 * tests/collate/*: Inputs and expected outputs for collation tests.
1033 * tests/run-collate-tests.sh: Script to run collation tests.
1035 * tests/unicode-collate.c (main): Rework slightly to make
1036 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
1038 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
1040 * docs/reference/glib/tmpl/unicode.sgml:
1041 * glib/gen-unicode-tables.pl:
1043 * glib/gunichartables.h:
1045 * tests/casefold.txt:
1046 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1047 GUnicodeBreakType types. That may break some applications, like
1050 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1052 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1053 docs. (#314393, Matthew F. Barnes)
1055 * glib/glib.symbols:
1056 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1057 they are already declared in glib.h. This doesn't break documented
1058 use of gprintf.h, but should probably be pointed out in the
1059 release notes for 2.10. (#314232, Behdad Esfahbod)
1061 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1063 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1064 complexity optimization, based on a patch from Matthias Clasen.
1066 * tests/patterntest.c: more tests, mostly from matthias.
1068 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1070 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1071 (#316703, Mark Drago)
1073 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1075 * glib/gpattern.c: applied a patch from matthias which checks on the
1076 upper bound of GPatternSpec length to optimize matches.
1079 * tests/patterntest.c: added more match cases.
1081 2005-09-16 Tor Lillqvist <tml@novell.com>
1083 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1084 warning to doc comment that these functions should not be used on
1085 encodings like CP932.
1087 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1089 * tests/keyfile-test.c: Add a test for grup names of length 1.
1091 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1092 of length 1. (#316309)
1094 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1096 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1098 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1100 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1101 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1104 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1106 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1107 (child_watch_helper_thread): Remove some dead code. Closes
1110 2005-09-07 Tor Lillqvist <tml@novell.com>
1112 * glib/Makefile.am: Create also a console version of the
1113 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1114 It's otherwise identical to gspawn-win32-helper.exe, except marked
1115 as a console application (linked without the -mwindows option).
1117 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1118 Drop the dont_wait parameter. Its truth value correlated 100% with
1119 the NULLness of the exit_status parameter anyway, so it's enough
1120 to check whether exit_status is NULL. Invert the sense of the
1121 dont_return_handle parameter and rename it to do_return_handle, to
1122 make the code easier to read by avoiding double negations.
1124 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1125 to do_spawn_with_pipes() accordingly.
1127 (do_spawn_with_pipes): If we have a console, use the console
1128 version of the helper program, otherwise use the GUI one. This
1129 avoids extra console windows opening up in some situations. (In
1130 case a console application uses the GUI gspawn-win32-helper.exe to
1131 spawn another console application we would get a separate console
1132 for the spawned console application).
1134 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1136 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1138 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1139 if the file is too large. (#315275, Kjartan Maraas)
1141 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1142 of read() is signed. (#315273, Kjartan Maraas)
1144 2005-08-31 Tor Lillqvist <tml@novell.com>
1146 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1147 the EXE and each DLL have their own atexit function chains.
1149 #define g_atexit as atexit instead. This means it has a
1150 better chance of doing what the caller wants. For instance,
1151 gtkhtml calls g_atexit() registering a function in gtkhtml
1152 itself. This caused a crash when g_atexit() was implemented as a
1153 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1154 the time the GLib DLL got unloaded.
1156 * glib/gutils.c: #undef the #define mentioned above, to also get a
1157 real g_atexit() into the DLL for backward compatibility. Document
1158 the Windows behaviour of g_atexit(), and document the varying ways
1159 atexit() can behave in the context of dynamically loaded modules
1162 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1164 * glib/glib.symbols:
1166 * glib/gdataset.c: Add string interning functions.
1168 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1170 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1173 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1175 * configure.in: Bump version to 2.9.0
1177 2005-08-25 Tor Lillqvist <tml@novell.com>
1179 Make also the g_spawn*() functions take parameters in the GLib
1180 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1181 API or ABI. Like the other GLib API that was earlier changed to
1182 use UTF-8 on Windows, the names of the functions that take UTF-8
1183 have _utf8 suffixes added by using preprocessor macros in the
1184 header file. The old names are kept for functions with the old
1185 behaviour, taking parameters in the system codepage, for DLL ABI
1188 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1189 g_spawn*() functions.
1191 * glib/gspawn-win32.c: Use wide-char API on NT-based
1192 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1193 system codepage (Win9x) and call the C library _wspawn*() or
1194 spawn*() functions respectvely. Add DLL ABI stability versions
1195 that take parameters in the system codepage.
1197 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1198 wide-char versions of argv and envp, and use wide-char API to
1199 change directory and spawn the program to run. Remove the verbose
1200 debugging output, it was too complex to modify for the wide-char
1201 features. (Just add temporary debugging printouts if needed, no
1202 need to have them permanently in the source.)
1204 * glib/gspawn.c: Corresponding documentation updates.
1206 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1207 symbols as PRIVATE, add the new _utf8-suffixed ones.
1209 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1211 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1212 guint64 otherwise and that can produce warnings about comparison
1213 between signed and unsigned.
1215 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1217 * glib/gutils.c: Fix the crt_externs.h include.
1219 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1221 * NEWS: Fix spelling of my first name.
1223 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1227 * === Released 2.8.1 ===
1231 2005-08-20 Hans Breuer <hans@breuer.org>
1233 * glib/makefile.msc.in : link with ws2_32.lib
1235 2005-08-18 Tor Lillqvist <tml@novell.com>
1237 * configure.in: Check for <sys/wait.h>
1239 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1241 2005-08-18 Ross Burton <ross@burtonini.com>
1244 Optimise single-character insertions.
1249 * tests/string-test.c:
1250 Add tests for new optimisation, and fix a leak.
1252 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1254 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1256 * glib/gutils.c: On Darwin, include crt-externs.h and
1257 define environ using _NSGetEnviron(). (#313731)
1259 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1261 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1262 (g_get_any_init_do): ... to this new function.
1263 (g_get_any_init): Declare as inline.
1264 (g_get_any_init_locked): New inline function, does the locking.
1265 Make use of these two throughout the code.
1267 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1269 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1270 the child process and then simply return. This makes
1271 The "S" option work as documented in g_on_error_query().
1272 (#313125, Matthew F. Barnes)
1274 * glib/gunicode.h: Update the link to Unicode category
1275 values. (#313369, Behnam Esfahbod)
1277 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1278 a little. (#311727, Tristan van Berkom)
1280 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1281 check work on ia64 too, where some symbols ended up
1282 in yet another section.