1 2006-03-07 Matthias Clasen <mclasen@redhat.com>
3 * configure.in: Bump version
5 * === Released 2.10.1 ===
9 * glib/gutf8.c (_g_charset_get_aliases): Match the prototype
10 in gconvert.c to fix build problems in NetBSD. (#333651,
11 Julio M. Merino Vidal)
13 2006-03-06 Anders Carlsson <andersca@imendio.com>
16 Make .dylib the shared library suffix on darwin.
17 (#322476, Vladimir Panov)
19 2006-03-06 Anders Carlsson <andersca@imendio.com>
22 Skip checking thread flags on Darwin. Fix suggested by Manish Singh.
23 (#314794, Gregor Riepl)
25 2006-03-01 Tor Lillqvist <tml@novell.com>
27 * glib/gutils.c (g_listenv): Separate implementation on Win32: Use
28 the wide character API on NT-based Windows. Return UTF-8 strings.
30 * glib/glib.symbols: Don't mark g_listenv as PRIVATE, as that
31 meant it wasn't present in the import library. PRIVATE is used
32 only for the backwards-compatibility DLL ABI stability hacks.
34 2006-02-24 Matthias Clasen <mclasen@redhat.com>
36 * configure.in: Bump version
38 * === Released 2.10.0 ===
40 * README.in: Mention the slice allocator and GInitiallyUnowned.
44 * configure.in: Bump version to 2.10.0
46 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
48 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
51 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
53 * tests/completion-test.c: (main): Plug leaks reported by
55 * tests/convert-test.c: (test_iconv_state), (test_one_half):
57 * tests/patterntest.c: (test_compilation): Same
58 * tests/shell-test.c: (do_argv_test): Same.
59 * tests/unicode-caseconv.c: (main): Same.
60 * tests/uri-test.c: (run_to_uri_tests): Same.
63 2006-02-19 Martyn Russell <martyn@imendio.com>
65 * tests/asyncqueue-test.c:
67 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
68 _insert_sorted and _insert_sorted_with_data API.
70 2006-02-18 Matthias Clasen <mclasen@redhat.com>
72 * tests/gobject/Makefile.am: Add paramspec-test
74 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
76 * tests/gobject/gvalue-test.c: Add more tests.
78 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
80 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
82 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
84 * glib/gthreadpool.c: Fix deadlock when signalling the thread
85 which freed a thread pool (#331110, Chris Wilson).
87 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
89 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
90 posix_memalign() in case we actually are going to use it, because
91 we detected a compliant implementation (#328997).
93 * configure.in (enable_included_printf): don't include malloc.h when
94 testing for posix_memalign() funcitonality, since this may break the
95 test on some systems (#328997).
97 2006-02-14 Matthias Clasen <mclasen@redhat.com>
99 * glib/gstrfuncs.c: Improve docs.
101 2006-02-11 Matthias Clasen <mclasen@redhat.com>
103 * configure.in: Bump version
105 2006-02-10 Matthias Clasen <mclasen@redhat.com>
107 * === Released 2.9.6 ===
111 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
113 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
114 the type of varargs arguments. (bug #317679)
116 2006-01-31 Matthias Clasen <mclasen@redhat.com>
118 * autogen.sh: Touch README and INSTALL here to pacify
119 automake. (#329124, Kjartan Maraas, Tim Janik)
121 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
123 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
124 some missing cases where pointed out by Sebastian Wilhelmi.
126 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
128 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
130 2006-01-30 David Schleef <ds@schleef.org>
132 * glib/grand.c: Update URL
134 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
136 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
137 create container classes which are as flexible as a GValue is
139 2006-01-27 Matthias Clasen <mclasen@redhat.com>
141 * configure.in: Bump version
143 * === Released 2.9.5 ===
145 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
150 2006-01-26 Matthias Clasen <mclasen@redhat.com>
152 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
155 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
157 * fixed buglets reported by Jens Granseuer in #328254.
159 * configure.in: free the memory allocated in posix_memalign() tests.
161 * glib/gslice.c: spelling fixes.
163 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
165 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
166 slices, make sure g_mem_gc_friendly is properly initialized.
168 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
169 the first allocation. applied some branching optimizations.
171 * docs/macros.txt: reflected --enable-gc-friendly change and
172 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
174 * configure.in: changed --enable-gc-friendly=yes to define
175 ENABLE_GC_FRIENDLY_DEFAULT.
177 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
178 if (G_UNLIKELY (g_mem_gc_friendly)).
181 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
182 care of by g_slice_free1().
184 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
186 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
188 * glib/gslice.c: only use posix_memalign() if it's known to work,
189 revert to memalign() otherwise.
191 * configure.in: check for broken posix_memalign() implementations
194 2006-01-24 Matthias Clasen <mclasen@redhat.com>
196 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
197 on all little-endian systems. (#143380, Marc Moorcroft)
199 2006-01-23 Matthias Clasen <mclasen@redhat.com>
201 * configure.in: Change the shared libary extension for hpux-ia64
202 to so. (#328253, The Written Word)
204 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
206 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
207 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
210 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
212 * glib/gslice.c (slice_config_init): initialize GSlice config from
213 G_SLICE environment variable. we support G_SLICE=always-malloc
214 currently, which forces all g_slice_*() allocations to use the system
218 g_parse_debug_string(): added a note about not using g_malloc() here.
219 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
220 g_slice. contains only guesswork in the WIN32 branch.
222 2006-01-18 Matthias Clasen <mclasen@redhat.com>
226 * === Released 2.9.4 ===
230 2006-01-17 Matthias Clasen <mclasen@redhat.com>
232 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
233 call. (#327421, Morten Welinder)
235 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
237 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
238 G_LOCK. Use the unused_thread_queue lock instead. Change
239 g_thread_pool_thread_proxy such that threads only wait on
240 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
241 due to superfluous tasks. Global tasks wait at most for
242 max-idle-time milliseconds. Make sure, that no task is woken up
243 twice for the same event via a wakeup_serial. This fixes #324228.
245 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
246 invalid NULL into the thread pools. This as well fixes #327290.
248 2006-01-16 Matthias Clasen <mclasen@redhat.com>
250 * configure.in: Bump version
252 * === Released 2.9.3 ===
256 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
257 Rework the strict aliasing fix to not break C++, pointed
258 out by Murray Cumming.
260 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
261 Signal waiting threads, problem noticed by Christian Kellner.
263 2006-01-16 Matthias Clasen <mclasen@redhat.com>
265 Fix bug 326747, Alberto Ruiz:
267 * glib/gconvert.c (make_valid_utf8): Change this function to
268 replace unknown characters by the Unicode replacement character
269 instead of '?', and don't append "(invalid encoding)".
270 (g_filename_display_name, g_filename_display_basename): Document
271 how to determine if the filename was in an invalid encoding.
273 2006-01-14 Matthias Clasen <mclasen@redhat.com>
275 * glib/gtree.c: Replace the simple recursive implementation
276 by a nonrecursive, threaded implementation by Maurizio
279 2006-01-12 Matthias Clasen <mclasen@redhat.com>
281 * tests/asyncqueue-test.c (main): Fix the
282 assert to not always trigger. (#326558,
285 2006-01-11 Matthias Clasen <mclasen@redhat.com>
287 * tests/convert-test.c: Don't test items_read and
288 items_written in error cases where it is not set.
290 * tests/asyncqueue-test.c: Set error to NULL.
292 2006-01-08 Matthias Clasen <mclasen@redhat.com>
294 * glib/gtree.c: Break some long lines.
296 * tests/tree-test.c: Add more tests.
298 * glib/gslice.c: Include config.h first, pointed out
301 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
303 * docs/reference/gobject/tmpl/gtype.sgml: explain that
304 G_TYPE_INSTANCE_GET_CLASS() does behave different during
307 2006-01-06 Matthias Clasen <mclasen@redhat.com>
309 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
310 that we have posix_memalign(). (#323937, Bogdan Nicula)
312 2006-01-05 Matthias Clasen <mclasen@redhat.com>
314 * configure.in: Bump version
316 * === Released 2.9.2 ===
318 2006-01-05 Hans Breuer <hans@breuer.org>
320 * glib/makefile.msc.in : added gslice.obj
322 2006-01-05 Matthias Clasen <mclasen@redhat.com>
326 * glib/gmain.c: Clarify the documentation of source ids
327 a bit. (#325874, Dan Williams)
329 * configure.in: Fix another strict aliasing problem.
331 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
334 2006-01-04 Manish Singh <yosh@gimp.org>
337 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
338 either, since they have the same issue as g_rmdir. (Related to
341 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
343 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
344 order when creating magazines, so we hand out chunks with ascending
347 2006-01-04 Abel Cheung <maddog@linuxhall.org>
349 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
351 2006-01-03 Martyn Russell <martyn@imendio.com>
353 * docs/reference/glib/glib-sections.txt:
355 * glib/gthreadpool.[ch]:
356 - Added new API g_thread_pool_get_idle_time() and
357 g_thread_pool_set_idle_time(). (#324228).
359 * tests/threadpool-test.c:
360 - Updated test case to do thread pool sorting, thread pool with
361 no sorting and a thread pool with idle thread timeouts.
363 2006-01-03 Matthias Clasen <mclasen@redhat.com>
365 * glib/gmain.h: Add new functions here, too.
367 * glib/glib.symbols: Add new functions.
369 2005-12-20 Michael Meeks <michael.meeks@novell.com>
371 * glib/gmain.c (g_main_context_is_owner): new method
372 to determine if the current thread is the owner of the
375 2006-01-02 Matthias Clasen <mclasen@redhat.com>
379 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
380 expanding to rmdir, since rmdir is not declared in a portable
381 system header we can include in gstdio.h. (#325249, Jani Monoses)
383 2006-01-01 Tor Lillqvist <tml@novell.com>
385 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
386 stdout and stderr to unbuffered. Otherwise the giochannel layer
387 will try to read from them regardless whether the
388 g_io_channel_win32_poll() call here has indicated
389 readability or not. (#325310)
391 2005-12-29 Matthias Clasen <mclasen@redhat.com>
393 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
394 in the docs. (#325273, Declan Naughton)
396 2005-12-27 Manish Singh <yosh@gimp.org>
398 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
399 make sure we really override things.
401 2005-12-27 Matthias Clasen <mclasen@redhat.com>
403 Fix #316221, Michal Benes, Stanislav Brabec;
405 * configure.in: Fix a strict aliasing problem in
406 g_static_mutex_get_mutex().
407 * glib/gthread.h: ...and in
408 g_static_mutex_get_mutex_impl_shortcut().
410 * glib/gdatasetprivate.h: Add a cast to silence compiler
411 warnings. (#321978, Andrew Paprocki)
413 Partial fix for bug #323937, Bogdan Nicula.
415 * configure.in: Check for malloc.h
416 * glib/gslice.c: Don't include malloc.h unconditionally.
418 2005-12-27 Manish Singh <yosh@gimp.org>
420 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
421 (#324950, Dan Yefimov)
423 2005-12-26 Matthias Clasen <mclasen@redhat.com>
425 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
428 2005-12-25 Matthias Clasen <mclasen@redhat.com>
430 * AUTHORS: Update my email
432 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
434 2005-12-21 Matthias Clasen <mclasen@redhat.com>
436 * README.in: Add some notes on when 'make check' may fail.
438 2005-12-20 Matthias Clasen <mclasen@redhat.com>
441 * glib/gthreadpool.h:
442 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
443 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
445 * tests/threadpool-test.c: Test this.
447 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
449 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
450 handle errors without depending on gmessage.c which might not be
451 setup when the error occours.
452 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
453 always freeing can be achieved by adjusting the working set time to
454 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
455 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
456 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
457 minimize wastage if small amounts of differently sized structrues are
458 allocated, this does come at a performance cost of roughly 5% though.
459 fixed up block alignment calculation, so it works for varying
460 block sizes. only use strerror() not g_strerror() since the latter
461 depends on working GQuark and GSlice.
462 mem_error(): implemented in terms of fprintf and vfprintf.
464 * tests/slice-color.c: new program to test cache colorization effects.
466 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
467 G_SLICE_CONFIG_WORKING_SET_MSECS.
469 2005-12-17 Matthias Clasen <mclasen@redhat.com>
471 * glib/goption.c (parse_short_option): Set an error in all
472 failure cases. (#324332, Tim-Philipp Müller)
474 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
476 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
477 and g_atomic_pointer_get. Add g_atomic_int_set and
478 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
479 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
480 as functions (additionally to the macros in the header) for the
481 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
483 2005-12-16 Matthias Clasen <mclasen@redhat.com>
485 * glib/gmem.c (g_allocator_new): Don't return a pointer to
486 a const struct, since apps expect to be able to modify it.
487 (#324179, J. Ali Harlow)
489 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
491 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
492 for systems where the initialization of atomic variables requires a
493 write memory barrier.
495 2005-12-09 Matthias Clasen <mclasen@redhat.com>
497 * configure.in: Bump version
499 * === Released 2.9.1 ===
501 2005-12-09 Alexander Larsson <alexl@redhat.com>
505 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
508 2005-12-08 Matthias Clasen <mclasen@redhat.com>
512 2005-12-07 Martyn Russell <martyn@imendio.com>
514 * glib/gasyncqueue.c:
515 - Call g_queue_insert_sorted() instead of duplicating the code.
516 - Call g_queue_sort() instead of duplicating the code.
517 - Invert sort function results to make sure the same sort function
518 gives the same results across glist, gslist, gqueue and
521 * tests/asyncqueue-test.c:
522 - Updated the sort function to reflect the example in the
523 documentation for gasyncqueue.c.
525 2005-12-07 Martyn Russell <martyn@imendio.com>
527 * docs/reference/glib/glib-sections.txt:
528 * docs/reference/glib/tmpl/linked_lists_double.sgml:
529 * docs/reference/glib/tmpl/linked_lists_single.sgml:
532 - Added g_list_insert_sorted_with_data () and
533 g_slist_insert_sorted_with_data ().
534 - Removed the extra check in g_list_sort() and g_slist_sort() for
535 GCompareDataFunc vs. GCompareFunc.
537 2005-12-07 Tor Lillqvist <tml@novell.com>
539 * glib-gettextize.in: Look up prefix at run-time on Win32,
540 assuming the standard directory structure with glib-gettextize in
543 2005-12-06 Matthias Clasen <mclasen@redhat.com>
545 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
546 we need GTK+ 2.8 to compile against GLib 2.10.
548 2005-12-05 Manish Singh <yosh@gimp.org>
550 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
551 to override any user settings.
553 2005-12-05 Matthias Clasen <mclasen@redhat.com>
555 * configure.in: Revert an accidential commit
557 2005-12-05 Matthias Clasen <mclasen@redhat.com>
559 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
561 * tests/Makefile.am (TESTS_ENVIRONMENT):
562 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
563 MALLOC_CHECK_ and MALLOC_PERTURB_
565 * tests/run-collate-tests.sh: Run the collation tests explicitly
566 in en_US locale. (#320463)
568 * glib/gmem.h: Really deprecate GMemChunk.
575 * glib/gstrfuncs.c: Add versioned deprecation docs.
577 2005-12-05 Martyn Russell <martyn@imendio.com>
579 * docs/reference/glib/glib-sections.txt:
580 * glib/gasyncqueue.[ch]:
581 - Added support for sorting async queues by with _push_sorted(),
582 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
585 * tests/asyncqueue-test.c:
586 - Added test case for gasyncqueue.c
588 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
590 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
591 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
592 in bug 323178. moved time consuming logic from g_slice_free() out of
593 the inner loop, so g_slice_free_chain_with_offset() provides a real
594 performance benefit over g_slice_free1() now.
596 * glib/gslice.h: renamed g_slice_free_chain() to
597 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
598 a type-safe macro as suggested in bug 323178.
599 simplified the macro implementation of g_slice_free() and implemented
600 it in a type safe manner for all compliers as suggested by Morten
601 Welinder <mortenw@gnome.org>.
606 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
608 2005-12-05 Matthias Clasen <mclasen@redhat.com>
610 * glib/gasyncqueue.c: Add some docs.
612 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
614 * glib/gatomic.c: In the ia64 implementation, use
615 __sync builtin without _si or _di suffix. (#321229,
616 Stanislav Brabec, patch by Andreas Schwab)
618 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
620 * glib/gslice.h: Remove comma at the end of enum.
622 2005-12-04 Matthias Clasen <mclasen@redhat.com>
624 Handle multiple user names with the same UID better.
625 (#319535, Laszlo Peter)
627 * glib/gutils.c (g_get_any_init_do): When determining user
628 data, first look up $LOGNAME. If the UID doesn't match
629 getuid(), fall back to the current behaviour of looking
630 up the user data based on getuid().
632 2005-12-04 Matthias Clasen <mclasen@redhat.com>
634 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
635 to make gcc warn if a function result is ignored. (#145466,
636 Arjan van de Ven, Alex Larsson)
638 * glib/gmem.h: Add the new attribute to g_realloc and
641 2005-12-03 Matthias Clasen <mclasen@redhat.com>
644 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
647 * glib/gmappedfile.c: Make mapping of empty files
650 * glib/gfileutils.c: Don't fork a new process just to
651 fix the permissions of the created temp file. (#321318,
652 Alexis S. L. Carvalho)
654 2005-12-02 Matthias Clasen <mclasen@redhat.com>
656 * README.in: Add a note about Solaris threads.
659 * configure.in: Remove support for Solaris threads.
660 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
662 2005-12-02 Matthias Clasen <mclasen@redhat.com>
664 * configure.in: Define G_GUINT64_CONSTANT in analogy to
665 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
667 2005-12-02 Matthias Clasen <mclasen@redhat.com>
669 * glib/gslice.c: Win32 portability fixes and C99ism removal,
670 pointed out by Kazuki Iwamoto. (#323052)
672 2005-12-02 Matthias Clasen <mclasen@redhat.com>
674 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
675 the IA64 implementation.
677 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
679 * glib/gslice.c: some naming and type size fixups.
681 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
683 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
684 variant into something less verbose, but digestible for gcc-3.4.
686 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
688 * tests/slice-test.c: extended to perform the benchmarking on the old
689 memchunk code if 'O' is selected.
691 * tests/memchunks.c: new file which contains the old GLib mem chunks
692 implementation with prefix old_mem_chunk_.
694 * tests/Makefile.am: added memchunks.c
696 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
698 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
699 acceptable load distribution of the syscall.
700 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
701 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
703 2005-12-01 Matthias Clasen <mclasen@redhat.com>
705 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
706 double-free base interface members.
708 * tests/Makefile.am: Remove duplicate variable.
710 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
712 * glib/gslice.[hc]: new slice allocator implementation.
714 * tests/slice-test.c: added random slice allocation test.
716 * glib/gthread.[hc]: removed newly added private thread mem API.
718 * glib/gthreadinit.h:
721 * glib/gmem.c: divided glib threading initialisation into three phases,
722 initialisation where private keys and messaging are not available (only
723 needed by gmem.c), initialisation without messaging but private keys
724 available (gslice.c, gmessage.c), and full fledged initialisers that
725 server the rest of glib. initialisation functions got renamed to reflect
726 the limitations of their corresponding phases.
728 * glib/gmem.c: removed memchunk code, defer allocations to
731 * glib/gmem.[hc]: removed g_slice_* skeletons.
733 * glib/glib.symbols: added g_slice_* symbols.
735 * configure.in: check for availability of posix_memalign(3), memalign(3)
738 * glib/Makefile.am: added gslice.[hc].
740 2005-12-01 Tor Lillqvist <tml@novell.com>
742 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
743 trailing slash(es) for non-root folders. stat() fails if non-root
744 folders are specified with trailing slashes. It's too much hassle
745 to demand that callers strip such slashes themselves, especially
746 as it is easy to get it wrong and strip the slash of a root
748 (g_rename): On NT-based Windows, use MoveFileEx() with
749 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
751 2005-11-28 Matthias Clasen <mclasen@redhat.com>
753 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
756 * configure.in: Check whether do { } while (0) works.
758 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
759 G_STMT_END if it works.
761 2005-11-28 Matthias Clasen <mclasen@redhat.com>
763 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
764 * glib/gnode.c (g_node_depth_traverse_level):
765 * glib/gmem.c (g_allocator_new):
766 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
767 (g_hash_table_foreach_remove_or_steal): Silence compiler
770 2005-11-27 Matthias Clasen <mclasen@redhat.com>
772 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
773 Don't read beyond len. (#322520, Christian Persch)
775 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
777 * glib/gdataset.c: access datalist flags via atomic pointer access
778 functions, instead of acquiring the dataset lock. this is faster and
779 also matches the atomic pointer readouts in gdatalistprivate.h and
780 direct pointer modifications required by gobject.c.
782 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
784 * glib/gdataset.c: streamlined initialization code somewhat,
785 removed GData node cache.
787 2005-11-22 Matthias Clasen <mclasen@redhat.com>
789 * glib/ghash.c (g_hash_table_ref):
790 (g_hash_table_unref): Mark these as new API.
792 Avoid double locking in g_intern_string (#322133,
795 * glib/gdataset.c (g_quark_from_string_internal): New
796 internal function which factors out common parts of
797 g_quark_from[_static]_string.
798 (g_quark_from_string, g_quark_from_static_string):
799 Use g_quark_from_string_internal.
800 (g_intern_string, g_intern_static_string): Use
801 g_quark_from_string_internal, and only take the
803 (g_quark_new): Don't store the strings shifted by -1
804 in the g_quarks array.
805 (g_quark_to_string): Adapt to the previous change.
807 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
811 g_hash_table_new_full(): create hash tables with a ref count of 1.
812 g_hash_table_ref(): atomically ref_count+=1
813 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
814 when refcount reaches 0.
815 g_hash_table_destroy(): just destroy keys and values, unref by 1.
816 g_hash_table_insert():
817 g_hash_table_replace(): assert ref_count>0.
820 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
823 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
825 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
826 variables. (#321984, Andrew Paprocki)
828 2005-11-18 Matthias Clasen <mclasen@redhat.com>
830 * configure.in: Bump version
832 2005-11-17 Matthias Clasen <mclasen@redhat.com>
834 * === Released 2.9.0 ===
836 * glib/glib.symbols: Updates
838 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
839 hints to the docs. (Paolo Borelli)
843 * glib/gthread.c (g_thread_foreach): Mark as new api.
845 * README.in: Updates.
849 * glib/gdate.c (g_date_set_time_t):
850 (g_date_set_time_val): Add functions to set a date from
851 a time_t and from a GTimeVal.
852 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
855 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
856 and g_utf8_offset_to_pointer.
858 * glib/gutf8.c (g_utf8_pointer_to_offset)
859 (g_utf8_offset_to_pointer): Handle negative offsets, and use
860 "stutter stepping" for going backwards. (#320638, Larry
865 * glib/gthread.c: const correctness fixes, found
866 by Arjan van de Ven and gcc.
868 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
870 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
871 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
873 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
875 * glib/guniprop.c: Use bit hacks instead when checking a general
876 category value against multiple values.
878 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
880 2005-11-10 Simos Xenitellis <simos@gnome.org>
882 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
884 2005-11-05 Tor Lillqvist <tml@novell.com>
886 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
887 (#320688) Instead set error and return failure.
889 2005-11-07 Matthias Clasen <mclasen@redhat.com>
891 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
892 while holding the context lock. (#320886, Andy Wingo)
894 2005-11-07 Billy Biggs <vektor@dumbterm.net>
896 * tests/refcount/Makefile.am: Add a missing reference to libglib
897 in the LDADD for this test.
899 2005-11-07 Matthias Clasen <mclasen@redhat.com>
901 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
902 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
903 GLib 2.9. (#320755, Luca Ferretti)
905 2005-11-04 Matthias Clasen <mclasen@redhat.com>
907 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
908 use a faster lookup table for g_unichar_get_mirror_char().
909 * glib/gmirroringtable.h: The new table.
910 * glib/gunichartables.h: Remove bidi_mirroring_table.
911 * glib/gen-unicode-tables.pl: Don't generate the mirroring
913 * glib/glib-mirroring-tab/*: Sources for the program
914 which generated gmirroringtable.h.
916 * glib/glist.c: Avoid some code duplication.
918 * glib/gscanner.h: Include gdataset.h. (#320322)
920 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
921 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
922 to help find critical warnings in applications. (#320017,
925 2005-11-02 Tor Lillqvist <tml@novell.com>
927 * glib/glib.symbols: Remove large amount of trailing whitespace
928 from one line. Remove #ifdef G_UNDEFINED from around
929 g_hash_table_new and g_hash_table_new_full symbols, what was the
930 purpose of that? Add the g_slice functions.
932 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
934 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
935 added g_slice_*() API to allocate and cache small bits of memory.
936 an actuall allocator implementation for g_slice_*() is still pending.
938 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
939 changed GRealThread list to use in-structure *next; fields instead
940 of GSList, in order for thread iteration to not depenend on g_slice_*()
942 _g_thread_mem_private_get():
943 _g_thread_mem_private_set(): added accessors for private memory,
944 needed because the ordinary GPrivate implementation relies on GArray
945 and GSList and therefore indirectly on working g_slice_*() allocations.
948 g_thread_foreach(): new public API function to loop over all existing threads.
962 removed GAllocator and free list usages and accompanying locks.
963 use g_slice_*() API to allocate and cache small bits of memory.
965 * glib/ghook.h: removed GMemChunk field from public API.
968 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
970 * glib/gnode.h: deprecate allocator API.
972 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
973 be used for poll rec lists.
975 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
976 g_relation_destroy(): free all tuples from the all_tuples hash table,
977 this effectively maintains the life time track keeping of tuples.
978 g_relation_delete_tuple(): free tuples which are removed from the
979 all_tuples hash table. this fixes a temporary leak that was present
980 in the memchunk code until the destruction of the relation.
982 2005-10-29 Matthias Clasen <mclasen@redhat.com>
984 * tests/convert-test.c: Add some tests for conversions between
985 UTF-8, UCS-4 and UTF-16.
987 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
988 of len == -1, noticed by Morten Welinder.
990 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
992 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
994 2005-10-26 Matthias Clasen <mclasen@redhat.com>
996 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
997 in the error case, and add some documentation. (#319806, Morten
1000 2005-10-19 Manish Singh <yosh@gimp.org>
1002 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
1003 config.h symbol automatically. Fixes bug #313731.
1005 2005-10-19 Tor Lillqvist <tml@novell.com>
1007 * glib/gwin32.c (g_win32_get_package_installation_directory):
1008 Return a g_strdup()ed copy of the value stored in the hash table,
1009 so that it can be g_free()d without leaving a dangling pointer in
1010 the hash table. (#319232)
1012 2005-10-06 Matthias Clasen <mclasen@redhat.com>
1014 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
1015 all-zero sequences correctly. (#317930, Sebastien Bacher)
1017 2005-10-05 Matthias Clasen <mclasen@redhat.com>
1019 * glib/gmarkup.c (g_markup_escape_text): Doc addition
1021 * Makefile.am: Add Collation tests.
1023 * tests/collate/*: Inputs and expected outputs for collation tests.
1025 * tests/run-collate-tests.sh: Script to run collation tests.
1027 * tests/unicode-collate.c (main): Rework slightly to make
1028 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
1030 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
1032 * docs/reference/glib/tmpl/unicode.sgml:
1033 * glib/gen-unicode-tables.pl:
1035 * glib/gunichartables.h:
1037 * tests/casefold.txt:
1038 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1039 GUnicodeBreakType types. That may break some applications, like
1042 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1044 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1045 docs. (#314393, Matthew F. Barnes)
1047 * glib/glib.symbols:
1048 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1049 they are already declared in glib.h. This doesn't break documented
1050 use of gprintf.h, but should probably be pointed out in the
1051 release notes for 2.10. (#314232, Behdad Esfahbod)
1053 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1055 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1056 complexity optimization, based on a patch from Matthias Clasen.
1058 * tests/patterntest.c: more tests, mostly from matthias.
1060 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1062 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1063 (#316703, Mark Drago)
1065 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1067 * glib/gpattern.c: applied a patch from matthias which checks on the
1068 upper bound of GPatternSpec length to optimize matches.
1071 * tests/patterntest.c: added more match cases.
1073 2005-09-16 Tor Lillqvist <tml@novell.com>
1075 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1076 warning to doc comment that these functions should not be used on
1077 encodings like CP932.
1079 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1081 * tests/keyfile-test.c: Add a test for grup names of length 1.
1083 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1084 of length 1. (#316309)
1086 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1088 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1090 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1092 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1093 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1096 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1098 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1099 (child_watch_helper_thread): Remove some dead code. Closes
1102 2005-09-07 Tor Lillqvist <tml@novell.com>
1104 * glib/Makefile.am: Create also a console version of the
1105 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1106 It's otherwise identical to gspawn-win32-helper.exe, except marked
1107 as a console application (linked without the -mwindows option).
1109 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1110 Drop the dont_wait parameter. Its truth value correlated 100% with
1111 the NULLness of the exit_status parameter anyway, so it's enough
1112 to check whether exit_status is NULL. Invert the sense of the
1113 dont_return_handle parameter and rename it to do_return_handle, to
1114 make the code easier to read by avoiding double negations.
1116 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1117 to do_spawn_with_pipes() accordingly.
1119 (do_spawn_with_pipes): If we have a console, use the console
1120 version of the helper program, otherwise use the GUI one. This
1121 avoids extra console windows opening up in some situations. (In
1122 case a console application uses the GUI gspawn-win32-helper.exe to
1123 spawn another console application we would get a separate console
1124 for the spawned console application).
1126 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1128 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1130 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1131 if the file is too large. (#315275, Kjartan Maraas)
1133 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1134 of read() is signed. (#315273, Kjartan Maraas)
1136 2005-08-31 Tor Lillqvist <tml@novell.com>
1138 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1139 the EXE and each DLL have their own atexit function chains.
1141 #define g_atexit as atexit instead. This means it has a
1142 better chance of doing what the caller wants. For instance,
1143 gtkhtml calls g_atexit() registering a function in gtkhtml
1144 itself. This caused a crash when g_atexit() was implemented as a
1145 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1146 the time the GLib DLL got unloaded.
1148 * glib/gutils.c: #undef the #define mentioned above, to also get a
1149 real g_atexit() into the DLL for backward compatibility. Document
1150 the Windows behaviour of g_atexit(), and document the varying ways
1151 atexit() can behave in the context of dynamically loaded modules
1154 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1156 * glib/glib.symbols:
1158 * glib/gdataset.c: Add string interning functions.
1160 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1162 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1165 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1167 * configure.in: Bump version to 2.9.0
1169 2005-08-25 Tor Lillqvist <tml@novell.com>
1171 Make also the g_spawn*() functions take parameters in the GLib
1172 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1173 API or ABI. Like the other GLib API that was earlier changed to
1174 use UTF-8 on Windows, the names of the functions that take UTF-8
1175 have _utf8 suffixes added by using preprocessor macros in the
1176 header file. The old names are kept for functions with the old
1177 behaviour, taking parameters in the system codepage, for DLL ABI
1180 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1181 g_spawn*() functions.
1183 * glib/gspawn-win32.c: Use wide-char API on NT-based
1184 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1185 system codepage (Win9x) and call the C library _wspawn*() or
1186 spawn*() functions respectvely. Add DLL ABI stability versions
1187 that take parameters in the system codepage.
1189 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1190 wide-char versions of argv and envp, and use wide-char API to
1191 change directory and spawn the program to run. Remove the verbose
1192 debugging output, it was too complex to modify for the wide-char
1193 features. (Just add temporary debugging printouts if needed, no
1194 need to have them permanently in the source.)
1196 * glib/gspawn.c: Corresponding documentation updates.
1198 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1199 symbols as PRIVATE, add the new _utf8-suffixed ones.
1201 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1203 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1204 guint64 otherwise and that can produce warnings about comparison
1205 between signed and unsigned.
1207 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1209 * glib/gutils.c: Fix the crt_externs.h include.
1211 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1213 * NEWS: Fix spelling of my first name.
1215 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1219 * === Released 2.8.1 ===
1223 2005-08-20 Hans Breuer <hans@breuer.org>
1225 * glib/makefile.msc.in : link with ws2_32.lib
1227 2005-08-18 Tor Lillqvist <tml@novell.com>
1229 * configure.in: Check for <sys/wait.h>
1231 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1233 2005-08-18 Ross Burton <ross@burtonini.com>
1236 Optimise single-character insertions.
1241 * tests/string-test.c:
1242 Add tests for new optimisation, and fix a leak.
1244 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1246 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1248 * glib/gutils.c: On Darwin, include crt-externs.h and
1249 define environ using _NSGetEnviron(). (#313731)
1251 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1253 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1254 (g_get_any_init_do): ... to this new function.
1255 (g_get_any_init): Declare as inline.
1256 (g_get_any_init_locked): New inline function, does the locking.
1257 Make use of these two throughout the code.
1259 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1261 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1262 the child process and then simply return. This makes
1263 The "S" option work as documented in g_on_error_query().
1264 (#313125, Matthew F. Barnes)
1266 * glib/gunicode.h: Update the link to Unicode category
1267 values. (#313369, Behnam Esfahbod)
1269 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1270 a little. (#311727, Tristan van Berkom)
1272 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1273 check work on ia64 too, where some symbols ended up
1274 in yet another section.