1 2006-03-13 Anders Carlsson <andersca@luminoth.local>
3 * configure.in: Revert fix for #322476, it breaks module loading
4 since libtool on darwin makes shared modules use .so and shared
5 libraries use .dylib. The fix breaks shared module loading everywhere
8 2006-03-02 Marcus Brinkmann <mb@g10code.de>
10 Implement watches for GIOChannels for write file descriptors on
13 * glib/giowin32.c (GIOWin32Channel): Add a new direction field.
14 (read_thread): Initialize direction.
15 (write_thread): New function.
16 (buffer_write): New function.
17 (g_io_win32_prepare): Handle the G_IO_WIN32_FILE_DESC case for the
19 (g_io_win32_fd_write): Call buffer_write() if there is a writer
21 (g_io_win32_fd_close): Set space_avail_event for writer threads.
22 (g_io_win32_fd_create_watch): Create the writer thread if
23 condition is G_IO_OUT.
24 (g_io_channel_win32_make_pollfd): Likewise here.
26 2006-03-09 Matthias Clasen <mclasen@redhat.com>
28 * Makefile.am: Add ChangeLog.pre-2.8 to EXTRA_DIST.
30 2006-03-08 Tor Lillqvist <tml@novell.com>
32 * glibconfig.h.win32.in: Add G_GUINT64_CONSTANT.
34 2006-03-07 Matthias Clasen <mclasen@redhat.com>
36 * configure.in: Bump version
38 * === Released 2.10.1 ===
42 * glib/gutf8.c (_g_charset_get_aliases): Match the prototype
43 in gconvert.c to fix build problems in NetBSD. (#333651,
44 Julio M. Merino Vidal)
46 2006-03-06 Anders Carlsson <andersca@imendio.com>
49 Make .dylib the shared library suffix on darwin.
50 (#322476, Vladimir Panov)
52 2006-03-06 Anders Carlsson <andersca@imendio.com>
55 Skip checking thread flags on Darwin. Fix suggested by Manish Singh.
56 (#314794, Gregor Riepl)
58 2006-03-01 Tor Lillqvist <tml@novell.com>
60 * glib/gutils.c (g_listenv): Separate implementation on Win32: Use
61 the wide character API on NT-based Windows. Return UTF-8 strings.
63 * glib/glib.symbols: Don't mark g_listenv as PRIVATE, as that
64 meant it wasn't present in the import library. PRIVATE is used
65 only for the backwards-compatibility DLL ABI stability hacks.
67 2006-02-24 Matthias Clasen <mclasen@redhat.com>
69 * configure.in: Bump version
71 * === Released 2.10.0 ===
73 * README.in: Mention the slice allocator and GInitiallyUnowned.
77 * configure.in: Bump version to 2.10.0
79 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
81 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
84 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
86 * tests/completion-test.c: (main): Plug leaks reported by
88 * tests/convert-test.c: (test_iconv_state), (test_one_half):
90 * tests/patterntest.c: (test_compilation): Same
91 * tests/shell-test.c: (do_argv_test): Same.
92 * tests/unicode-caseconv.c: (main): Same.
93 * tests/uri-test.c: (run_to_uri_tests): Same.
96 2006-02-19 Martyn Russell <martyn@imendio.com>
98 * tests/asyncqueue-test.c:
100 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
101 _insert_sorted and _insert_sorted_with_data API.
103 2006-02-18 Matthias Clasen <mclasen@redhat.com>
105 * tests/gobject/Makefile.am: Add paramspec-test
107 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
109 * tests/gobject/gvalue-test.c: Add more tests.
111 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
113 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
115 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
117 * glib/gthreadpool.c: Fix deadlock when signalling the thread
118 which freed a thread pool (#331110, Chris Wilson).
120 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
122 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
123 posix_memalign() in case we actually are going to use it, because
124 we detected a compliant implementation (#328997).
126 * configure.in (enable_included_printf): don't include malloc.h when
127 testing for posix_memalign() funcitonality, since this may break the
128 test on some systems (#328997).
130 2006-02-14 Matthias Clasen <mclasen@redhat.com>
132 * glib/gstrfuncs.c: Improve docs.
134 2006-02-11 Matthias Clasen <mclasen@redhat.com>
136 * configure.in: Bump version
138 2006-02-10 Matthias Clasen <mclasen@redhat.com>
140 * === Released 2.9.6 ===
144 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
146 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
147 the type of varargs arguments. (bug #317679)
149 2006-01-31 Matthias Clasen <mclasen@redhat.com>
151 * autogen.sh: Touch README and INSTALL here to pacify
152 automake. (#329124, Kjartan Maraas, Tim Janik)
154 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
156 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
157 some missing cases where pointed out by Sebastian Wilhelmi.
159 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
161 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
163 2006-01-30 David Schleef <ds@schleef.org>
165 * glib/grand.c: Update URL
167 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
169 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
170 create container classes which are as flexible as a GValue is
172 2006-01-27 Matthias Clasen <mclasen@redhat.com>
174 * configure.in: Bump version
176 * === Released 2.9.5 ===
178 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
183 2006-01-26 Matthias Clasen <mclasen@redhat.com>
185 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
188 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
190 * fixed buglets reported by Jens Granseuer in #328254.
192 * configure.in: free the memory allocated in posix_memalign() tests.
194 * glib/gslice.c: spelling fixes.
196 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
198 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
199 slices, make sure g_mem_gc_friendly is properly initialized.
201 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
202 the first allocation. applied some branching optimizations.
204 * docs/macros.txt: reflected --enable-gc-friendly change and
205 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
207 * configure.in: changed --enable-gc-friendly=yes to define
208 ENABLE_GC_FRIENDLY_DEFAULT.
210 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
211 if (G_UNLIKELY (g_mem_gc_friendly)).
214 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
215 care of by g_slice_free1().
217 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
219 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
221 * glib/gslice.c: only use posix_memalign() if it's known to work,
222 revert to memalign() otherwise.
224 * configure.in: check for broken posix_memalign() implementations
227 2006-01-24 Matthias Clasen <mclasen@redhat.com>
229 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
230 on all little-endian systems. (#143380, Marc Moorcroft)
232 2006-01-23 Matthias Clasen <mclasen@redhat.com>
234 * configure.in: Change the shared libary extension for hpux-ia64
235 to so. (#328253, The Written Word)
237 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
239 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
240 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
243 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
245 * glib/gslice.c (slice_config_init): initialize GSlice config from
246 G_SLICE environment variable. we support G_SLICE=always-malloc
247 currently, which forces all g_slice_*() allocations to use the system
251 g_parse_debug_string(): added a note about not using g_malloc() here.
252 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
253 g_slice. contains only guesswork in the WIN32 branch.
255 2006-01-18 Matthias Clasen <mclasen@redhat.com>
259 * === Released 2.9.4 ===
263 2006-01-17 Matthias Clasen <mclasen@redhat.com>
265 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
266 call. (#327421, Morten Welinder)
268 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
270 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
271 G_LOCK. Use the unused_thread_queue lock instead. Change
272 g_thread_pool_thread_proxy such that threads only wait on
273 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
274 due to superfluous tasks. Global tasks wait at most for
275 max-idle-time milliseconds. Make sure, that no task is woken up
276 twice for the same event via a wakeup_serial. This fixes #324228.
278 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
279 invalid NULL into the thread pools. This as well fixes #327290.
281 2006-01-16 Matthias Clasen <mclasen@redhat.com>
283 * configure.in: Bump version
285 * === Released 2.9.3 ===
289 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
290 Rework the strict aliasing fix to not break C++, pointed
291 out by Murray Cumming.
293 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
294 Signal waiting threads, problem noticed by Christian Kellner.
296 2006-01-16 Matthias Clasen <mclasen@redhat.com>
298 Fix bug 326747, Alberto Ruiz:
300 * glib/gconvert.c (make_valid_utf8): Change this function to
301 replace unknown characters by the Unicode replacement character
302 instead of '?', and don't append "(invalid encoding)".
303 (g_filename_display_name, g_filename_display_basename): Document
304 how to determine if the filename was in an invalid encoding.
306 2006-01-14 Matthias Clasen <mclasen@redhat.com>
308 * glib/gtree.c: Replace the simple recursive implementation
309 by a nonrecursive, threaded implementation by Maurizio
312 2006-01-12 Matthias Clasen <mclasen@redhat.com>
314 * tests/asyncqueue-test.c (main): Fix the
315 assert to not always trigger. (#326558,
318 2006-01-11 Matthias Clasen <mclasen@redhat.com>
320 * tests/convert-test.c: Don't test items_read and
321 items_written in error cases where it is not set.
323 * tests/asyncqueue-test.c: Set error to NULL.
325 2006-01-08 Matthias Clasen <mclasen@redhat.com>
327 * glib/gtree.c: Break some long lines.
329 * tests/tree-test.c: Add more tests.
331 * glib/gslice.c: Include config.h first, pointed out
334 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
336 * docs/reference/gobject/tmpl/gtype.sgml: explain that
337 G_TYPE_INSTANCE_GET_CLASS() does behave different during
340 2006-01-06 Matthias Clasen <mclasen@redhat.com>
342 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
343 that we have posix_memalign(). (#323937, Bogdan Nicula)
345 2006-01-05 Matthias Clasen <mclasen@redhat.com>
347 * configure.in: Bump version
349 * === Released 2.9.2 ===
351 2006-01-05 Hans Breuer <hans@breuer.org>
353 * glib/makefile.msc.in : added gslice.obj
355 2006-01-05 Matthias Clasen <mclasen@redhat.com>
359 * glib/gmain.c: Clarify the documentation of source ids
360 a bit. (#325874, Dan Williams)
362 * configure.in: Fix another strict aliasing problem.
364 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
367 2006-01-04 Manish Singh <yosh@gimp.org>
370 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
371 either, since they have the same issue as g_rmdir. (Related to
374 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
376 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
377 order when creating magazines, so we hand out chunks with ascending
380 2006-01-04 Abel Cheung <maddog@linuxhall.org>
382 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
384 2006-01-03 Martyn Russell <martyn@imendio.com>
386 * docs/reference/glib/glib-sections.txt:
388 * glib/gthreadpool.[ch]:
389 - Added new API g_thread_pool_get_idle_time() and
390 g_thread_pool_set_idle_time(). (#324228).
392 * tests/threadpool-test.c:
393 - Updated test case to do thread pool sorting, thread pool with
394 no sorting and a thread pool with idle thread timeouts.
396 2006-01-03 Matthias Clasen <mclasen@redhat.com>
398 * glib/gmain.h: Add new functions here, too.
400 * glib/glib.symbols: Add new functions.
402 2005-12-20 Michael Meeks <michael.meeks@novell.com>
404 * glib/gmain.c (g_main_context_is_owner): new method
405 to determine if the current thread is the owner of the
408 2006-01-02 Matthias Clasen <mclasen@redhat.com>
412 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
413 expanding to rmdir, since rmdir is not declared in a portable
414 system header we can include in gstdio.h. (#325249, Jani Monoses)
416 2006-01-01 Tor Lillqvist <tml@novell.com>
418 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
419 stdout and stderr to unbuffered. Otherwise the giochannel layer
420 will try to read from them regardless whether the
421 g_io_channel_win32_poll() call here has indicated
422 readability or not. (#325310)
424 2005-12-29 Matthias Clasen <mclasen@redhat.com>
426 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
427 in the docs. (#325273, Declan Naughton)
429 2005-12-27 Manish Singh <yosh@gimp.org>
431 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
432 make sure we really override things.
434 2005-12-27 Matthias Clasen <mclasen@redhat.com>
436 Fix #316221, Michal Benes, Stanislav Brabec;
438 * configure.in: Fix a strict aliasing problem in
439 g_static_mutex_get_mutex().
440 * glib/gthread.h: ...and in
441 g_static_mutex_get_mutex_impl_shortcut().
443 * glib/gdatasetprivate.h: Add a cast to silence compiler
444 warnings. (#321978, Andrew Paprocki)
446 Partial fix for bug #323937, Bogdan Nicula.
448 * configure.in: Check for malloc.h
449 * glib/gslice.c: Don't include malloc.h unconditionally.
451 2005-12-27 Manish Singh <yosh@gimp.org>
453 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
454 (#324950, Dan Yefimov)
456 2005-12-26 Matthias Clasen <mclasen@redhat.com>
458 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
461 2005-12-25 Matthias Clasen <mclasen@redhat.com>
463 * AUTHORS: Update my email
465 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
467 2005-12-21 Matthias Clasen <mclasen@redhat.com>
469 * README.in: Add some notes on when 'make check' may fail.
471 2005-12-20 Matthias Clasen <mclasen@redhat.com>
474 * glib/gthreadpool.h:
475 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
476 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
478 * tests/threadpool-test.c: Test this.
480 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
482 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
483 handle errors without depending on gmessage.c which might not be
484 setup when the error occours.
485 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
486 always freeing can be achieved by adjusting the working set time to
487 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
488 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
489 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
490 minimize wastage if small amounts of differently sized structrues are
491 allocated, this does come at a performance cost of roughly 5% though.
492 fixed up block alignment calculation, so it works for varying
493 block sizes. only use strerror() not g_strerror() since the latter
494 depends on working GQuark and GSlice.
495 mem_error(): implemented in terms of fprintf and vfprintf.
497 * tests/slice-color.c: new program to test cache colorization effects.
499 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
500 G_SLICE_CONFIG_WORKING_SET_MSECS.
502 2005-12-17 Matthias Clasen <mclasen@redhat.com>
504 * glib/goption.c (parse_short_option): Set an error in all
505 failure cases. (#324332, Tim-Philipp Müller)
507 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
509 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
510 and g_atomic_pointer_get. Add g_atomic_int_set and
511 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
512 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
513 as functions (additionally to the macros in the header) for the
514 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
516 2005-12-16 Matthias Clasen <mclasen@redhat.com>
518 * glib/gmem.c (g_allocator_new): Don't return a pointer to
519 a const struct, since apps expect to be able to modify it.
520 (#324179, J. Ali Harlow)
522 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
524 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
525 for systems where the initialization of atomic variables requires a
526 write memory barrier.
528 2005-12-09 Matthias Clasen <mclasen@redhat.com>
530 * configure.in: Bump version
532 * === Released 2.9.1 ===
534 2005-12-09 Alexander Larsson <alexl@redhat.com>
538 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
541 2005-12-08 Matthias Clasen <mclasen@redhat.com>
545 2005-12-07 Martyn Russell <martyn@imendio.com>
547 * glib/gasyncqueue.c:
548 - Call g_queue_insert_sorted() instead of duplicating the code.
549 - Call g_queue_sort() instead of duplicating the code.
550 - Invert sort function results to make sure the same sort function
551 gives the same results across glist, gslist, gqueue and
554 * tests/asyncqueue-test.c:
555 - Updated the sort function to reflect the example in the
556 documentation for gasyncqueue.c.
558 2005-12-07 Martyn Russell <martyn@imendio.com>
560 * docs/reference/glib/glib-sections.txt:
561 * docs/reference/glib/tmpl/linked_lists_double.sgml:
562 * docs/reference/glib/tmpl/linked_lists_single.sgml:
565 - Added g_list_insert_sorted_with_data () and
566 g_slist_insert_sorted_with_data ().
567 - Removed the extra check in g_list_sort() and g_slist_sort() for
568 GCompareDataFunc vs. GCompareFunc.
570 2005-12-07 Tor Lillqvist <tml@novell.com>
572 * glib-gettextize.in: Look up prefix at run-time on Win32,
573 assuming the standard directory structure with glib-gettextize in
576 2005-12-06 Matthias Clasen <mclasen@redhat.com>
578 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
579 we need GTK+ 2.8 to compile against GLib 2.10.
581 2005-12-05 Manish Singh <yosh@gimp.org>
583 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
584 to override any user settings.
586 2005-12-05 Matthias Clasen <mclasen@redhat.com>
588 * configure.in: Revert an accidential commit
590 2005-12-05 Matthias Clasen <mclasen@redhat.com>
592 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
594 * tests/Makefile.am (TESTS_ENVIRONMENT):
595 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
596 MALLOC_CHECK_ and MALLOC_PERTURB_
598 * tests/run-collate-tests.sh: Run the collation tests explicitly
599 in en_US locale. (#320463)
601 * glib/gmem.h: Really deprecate GMemChunk.
608 * glib/gstrfuncs.c: Add versioned deprecation docs.
610 2005-12-05 Martyn Russell <martyn@imendio.com>
612 * docs/reference/glib/glib-sections.txt:
613 * glib/gasyncqueue.[ch]:
614 - Added support for sorting async queues by with _push_sorted(),
615 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
618 * tests/asyncqueue-test.c:
619 - Added test case for gasyncqueue.c
621 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
623 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
624 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
625 in bug 323178. moved time consuming logic from g_slice_free() out of
626 the inner loop, so g_slice_free_chain_with_offset() provides a real
627 performance benefit over g_slice_free1() now.
629 * glib/gslice.h: renamed g_slice_free_chain() to
630 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
631 a type-safe macro as suggested in bug 323178.
632 simplified the macro implementation of g_slice_free() and implemented
633 it in a type safe manner for all compliers as suggested by Morten
634 Welinder <mortenw@gnome.org>.
639 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
641 2005-12-05 Matthias Clasen <mclasen@redhat.com>
643 * glib/gasyncqueue.c: Add some docs.
645 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
647 * glib/gatomic.c: In the ia64 implementation, use
648 __sync builtin without _si or _di suffix. (#321229,
649 Stanislav Brabec, patch by Andreas Schwab)
651 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
653 * glib/gslice.h: Remove comma at the end of enum.
655 2005-12-04 Matthias Clasen <mclasen@redhat.com>
657 Handle multiple user names with the same UID better.
658 (#319535, Laszlo Peter)
660 * glib/gutils.c (g_get_any_init_do): When determining user
661 data, first look up $LOGNAME. If the UID doesn't match
662 getuid(), fall back to the current behaviour of looking
663 up the user data based on getuid().
665 2005-12-04 Matthias Clasen <mclasen@redhat.com>
667 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
668 to make gcc warn if a function result is ignored. (#145466,
669 Arjan van de Ven, Alex Larsson)
671 * glib/gmem.h: Add the new attribute to g_realloc and
674 2005-12-03 Matthias Clasen <mclasen@redhat.com>
677 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
680 * glib/gmappedfile.c: Make mapping of empty files
683 * glib/gfileutils.c: Don't fork a new process just to
684 fix the permissions of the created temp file. (#321318,
685 Alexis S. L. Carvalho)
687 2005-12-02 Matthias Clasen <mclasen@redhat.com>
689 * README.in: Add a note about Solaris threads.
692 * configure.in: Remove support for Solaris threads.
693 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
695 2005-12-02 Matthias Clasen <mclasen@redhat.com>
697 * configure.in: Define G_GUINT64_CONSTANT in analogy to
698 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
700 2005-12-02 Matthias Clasen <mclasen@redhat.com>
702 * glib/gslice.c: Win32 portability fixes and C99ism removal,
703 pointed out by Kazuki Iwamoto. (#323052)
705 2005-12-02 Matthias Clasen <mclasen@redhat.com>
707 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
708 the IA64 implementation.
710 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
712 * glib/gslice.c: some naming and type size fixups.
714 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
716 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
717 variant into something less verbose, but digestible for gcc-3.4.
719 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
721 * tests/slice-test.c: extended to perform the benchmarking on the old
722 memchunk code if 'O' is selected.
724 * tests/memchunks.c: new file which contains the old GLib mem chunks
725 implementation with prefix old_mem_chunk_.
727 * tests/Makefile.am: added memchunks.c
729 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
731 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
732 acceptable load distribution of the syscall.
733 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
734 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
736 2005-12-01 Matthias Clasen <mclasen@redhat.com>
738 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
739 double-free base interface members.
741 * tests/Makefile.am: Remove duplicate variable.
743 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
745 * glib/gslice.[hc]: new slice allocator implementation.
747 * tests/slice-test.c: added random slice allocation test.
749 * glib/gthread.[hc]: removed newly added private thread mem API.
751 * glib/gthreadinit.h:
754 * glib/gmem.c: divided glib threading initialisation into three phases,
755 initialisation where private keys and messaging are not available (only
756 needed by gmem.c), initialisation without messaging but private keys
757 available (gslice.c, gmessage.c), and full fledged initialisers that
758 server the rest of glib. initialisation functions got renamed to reflect
759 the limitations of their corresponding phases.
761 * glib/gmem.c: removed memchunk code, defer allocations to
764 * glib/gmem.[hc]: removed g_slice_* skeletons.
766 * glib/glib.symbols: added g_slice_* symbols.
768 * configure.in: check for availability of posix_memalign(3), memalign(3)
771 * glib/Makefile.am: added gslice.[hc].
773 2005-12-01 Tor Lillqvist <tml@novell.com>
775 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
776 trailing slash(es) for non-root folders. stat() fails if non-root
777 folders are specified with trailing slashes. It's too much hassle
778 to demand that callers strip such slashes themselves, especially
779 as it is easy to get it wrong and strip the slash of a root
781 (g_rename): On NT-based Windows, use MoveFileEx() with
782 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
784 2005-11-28 Matthias Clasen <mclasen@redhat.com>
786 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
789 * configure.in: Check whether do { } while (0) works.
791 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
792 G_STMT_END if it works.
794 2005-11-28 Matthias Clasen <mclasen@redhat.com>
796 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
797 * glib/gnode.c (g_node_depth_traverse_level):
798 * glib/gmem.c (g_allocator_new):
799 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
800 (g_hash_table_foreach_remove_or_steal): Silence compiler
803 2005-11-27 Matthias Clasen <mclasen@redhat.com>
805 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
806 Don't read beyond len. (#322520, Christian Persch)
808 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
810 * glib/gdataset.c: access datalist flags via atomic pointer access
811 functions, instead of acquiring the dataset lock. this is faster and
812 also matches the atomic pointer readouts in gdatalistprivate.h and
813 direct pointer modifications required by gobject.c.
815 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
817 * glib/gdataset.c: streamlined initialization code somewhat,
818 removed GData node cache.
820 2005-11-22 Matthias Clasen <mclasen@redhat.com>
822 * glib/ghash.c (g_hash_table_ref):
823 (g_hash_table_unref): Mark these as new API.
825 Avoid double locking in g_intern_string (#322133,
828 * glib/gdataset.c (g_quark_from_string_internal): New
829 internal function which factors out common parts of
830 g_quark_from[_static]_string.
831 (g_quark_from_string, g_quark_from_static_string):
832 Use g_quark_from_string_internal.
833 (g_intern_string, g_intern_static_string): Use
834 g_quark_from_string_internal, and only take the
836 (g_quark_new): Don't store the strings shifted by -1
837 in the g_quarks array.
838 (g_quark_to_string): Adapt to the previous change.
840 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
844 g_hash_table_new_full(): create hash tables with a ref count of 1.
845 g_hash_table_ref(): atomically ref_count+=1
846 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
847 when refcount reaches 0.
848 g_hash_table_destroy(): just destroy keys and values, unref by 1.
849 g_hash_table_insert():
850 g_hash_table_replace(): assert ref_count>0.
853 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
856 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
858 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
859 variables. (#321984, Andrew Paprocki)
861 2005-11-18 Matthias Clasen <mclasen@redhat.com>
863 * configure.in: Bump version
865 2005-11-17 Matthias Clasen <mclasen@redhat.com>
867 * === Released 2.9.0 ===
869 * glib/glib.symbols: Updates
871 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
872 hints to the docs. (Paolo Borelli)
876 * glib/gthread.c (g_thread_foreach): Mark as new api.
878 * README.in: Updates.
882 * glib/gdate.c (g_date_set_time_t):
883 (g_date_set_time_val): Add functions to set a date from
884 a time_t and from a GTimeVal.
885 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
888 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
889 and g_utf8_offset_to_pointer.
891 * glib/gutf8.c (g_utf8_pointer_to_offset)
892 (g_utf8_offset_to_pointer): Handle negative offsets, and use
893 "stutter stepping" for going backwards. (#320638, Larry
898 * glib/gthread.c: const correctness fixes, found
899 by Arjan van de Ven and gcc.
901 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
903 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
904 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
906 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
908 * glib/guniprop.c: Use bit hacks instead when checking a general
909 category value against multiple values.
911 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
913 2005-11-10 Simos Xenitellis <simos@gnome.org>
915 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
917 2005-11-05 Tor Lillqvist <tml@novell.com>
919 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
920 (#320688) Instead set error and return failure.
922 2005-11-07 Matthias Clasen <mclasen@redhat.com>
924 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
925 while holding the context lock. (#320886, Andy Wingo)
927 2005-11-07 Billy Biggs <vektor@dumbterm.net>
929 * tests/refcount/Makefile.am: Add a missing reference to libglib
930 in the LDADD for this test.
932 2005-11-07 Matthias Clasen <mclasen@redhat.com>
934 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
935 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
936 GLib 2.9. (#320755, Luca Ferretti)
938 2005-11-04 Matthias Clasen <mclasen@redhat.com>
940 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
941 use a faster lookup table for g_unichar_get_mirror_char().
942 * glib/gmirroringtable.h: The new table.
943 * glib/gunichartables.h: Remove bidi_mirroring_table.
944 * glib/gen-unicode-tables.pl: Don't generate the mirroring
946 * glib/glib-mirroring-tab/*: Sources for the program
947 which generated gmirroringtable.h.
949 * glib/glist.c: Avoid some code duplication.
951 * glib/gscanner.h: Include gdataset.h. (#320322)
953 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
954 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
955 to help find critical warnings in applications. (#320017,
958 2005-11-02 Tor Lillqvist <tml@novell.com>
960 * glib/glib.symbols: Remove large amount of trailing whitespace
961 from one line. Remove #ifdef G_UNDEFINED from around
962 g_hash_table_new and g_hash_table_new_full symbols, what was the
963 purpose of that? Add the g_slice functions.
965 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
967 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
968 added g_slice_*() API to allocate and cache small bits of memory.
969 an actuall allocator implementation for g_slice_*() is still pending.
971 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
972 changed GRealThread list to use in-structure *next; fields instead
973 of GSList, in order for thread iteration to not depenend on g_slice_*()
975 _g_thread_mem_private_get():
976 _g_thread_mem_private_set(): added accessors for private memory,
977 needed because the ordinary GPrivate implementation relies on GArray
978 and GSList and therefore indirectly on working g_slice_*() allocations.
981 g_thread_foreach(): new public API function to loop over all existing threads.
995 removed GAllocator and free list usages and accompanying locks.
996 use g_slice_*() API to allocate and cache small bits of memory.
998 * glib/ghook.h: removed GMemChunk field from public API.
1001 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
1003 * glib/gnode.h: deprecate allocator API.
1005 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
1006 be used for poll rec lists.
1008 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
1009 g_relation_destroy(): free all tuples from the all_tuples hash table,
1010 this effectively maintains the life time track keeping of tuples.
1011 g_relation_delete_tuple(): free tuples which are removed from the
1012 all_tuples hash table. this fixes a temporary leak that was present
1013 in the memchunk code until the destruction of the relation.
1015 2005-10-29 Matthias Clasen <mclasen@redhat.com>
1017 * tests/convert-test.c: Add some tests for conversions between
1018 UTF-8, UCS-4 and UTF-16.
1020 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
1021 of len == -1, noticed by Morten Welinder.
1023 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
1025 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
1027 2005-10-26 Matthias Clasen <mclasen@redhat.com>
1029 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
1030 in the error case, and add some documentation. (#319806, Morten
1033 2005-10-19 Manish Singh <yosh@gimp.org>
1035 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
1036 config.h symbol automatically. Fixes bug #313731.
1038 2005-10-19 Tor Lillqvist <tml@novell.com>
1040 * glib/gwin32.c (g_win32_get_package_installation_directory):
1041 Return a g_strdup()ed copy of the value stored in the hash table,
1042 so that it can be g_free()d without leaving a dangling pointer in
1043 the hash table. (#319232)
1045 2005-10-06 Matthias Clasen <mclasen@redhat.com>
1047 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
1048 all-zero sequences correctly. (#317930, Sebastien Bacher)
1050 2005-10-05 Matthias Clasen <mclasen@redhat.com>
1052 * glib/gmarkup.c (g_markup_escape_text): Doc addition
1054 * Makefile.am: Add Collation tests.
1056 * tests/collate/*: Inputs and expected outputs for collation tests.
1058 * tests/run-collate-tests.sh: Script to run collation tests.
1060 * tests/unicode-collate.c (main): Rework slightly to make
1061 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
1063 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
1065 * docs/reference/glib/tmpl/unicode.sgml:
1066 * glib/gen-unicode-tables.pl:
1068 * glib/gunichartables.h:
1070 * tests/casefold.txt:
1071 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1072 GUnicodeBreakType types. That may break some applications, like
1075 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1077 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1078 docs. (#314393, Matthew F. Barnes)
1080 * glib/glib.symbols:
1081 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1082 they are already declared in glib.h. This doesn't break documented
1083 use of gprintf.h, but should probably be pointed out in the
1084 release notes for 2.10. (#314232, Behdad Esfahbod)
1086 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1088 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1089 complexity optimization, based on a patch from Matthias Clasen.
1091 * tests/patterntest.c: more tests, mostly from matthias.
1093 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1095 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1096 (#316703, Mark Drago)
1098 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1100 * glib/gpattern.c: applied a patch from matthias which checks on the
1101 upper bound of GPatternSpec length to optimize matches.
1104 * tests/patterntest.c: added more match cases.
1106 2005-09-16 Tor Lillqvist <tml@novell.com>
1108 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1109 warning to doc comment that these functions should not be used on
1110 encodings like CP932.
1112 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1114 * tests/keyfile-test.c: Add a test for grup names of length 1.
1116 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1117 of length 1. (#316309)
1119 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1121 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1123 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1125 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1126 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1129 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1131 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1132 (child_watch_helper_thread): Remove some dead code. Closes
1135 2005-09-07 Tor Lillqvist <tml@novell.com>
1137 * glib/Makefile.am: Create also a console version of the
1138 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1139 It's otherwise identical to gspawn-win32-helper.exe, except marked
1140 as a console application (linked without the -mwindows option).
1142 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1143 Drop the dont_wait parameter. Its truth value correlated 100% with
1144 the NULLness of the exit_status parameter anyway, so it's enough
1145 to check whether exit_status is NULL. Invert the sense of the
1146 dont_return_handle parameter and rename it to do_return_handle, to
1147 make the code easier to read by avoiding double negations.
1149 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1150 to do_spawn_with_pipes() accordingly.
1152 (do_spawn_with_pipes): If we have a console, use the console
1153 version of the helper program, otherwise use the GUI one. This
1154 avoids extra console windows opening up in some situations. (In
1155 case a console application uses the GUI gspawn-win32-helper.exe to
1156 spawn another console application we would get a separate console
1157 for the spawned console application).
1159 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1161 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1163 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1164 if the file is too large. (#315275, Kjartan Maraas)
1166 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1167 of read() is signed. (#315273, Kjartan Maraas)
1169 2005-08-31 Tor Lillqvist <tml@novell.com>
1171 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1172 the EXE and each DLL have their own atexit function chains.
1174 #define g_atexit as atexit instead. This means it has a
1175 better chance of doing what the caller wants. For instance,
1176 gtkhtml calls g_atexit() registering a function in gtkhtml
1177 itself. This caused a crash when g_atexit() was implemented as a
1178 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1179 the time the GLib DLL got unloaded.
1181 * glib/gutils.c: #undef the #define mentioned above, to also get a
1182 real g_atexit() into the DLL for backward compatibility. Document
1183 the Windows behaviour of g_atexit(), and document the varying ways
1184 atexit() can behave in the context of dynamically loaded modules
1187 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1189 * glib/glib.symbols:
1191 * glib/gdataset.c: Add string interning functions.
1193 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1195 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1198 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1200 * configure.in: Bump version to 2.9.0
1202 2005-08-25 Tor Lillqvist <tml@novell.com>
1204 Make also the g_spawn*() functions take parameters in the GLib
1205 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1206 API or ABI. Like the other GLib API that was earlier changed to
1207 use UTF-8 on Windows, the names of the functions that take UTF-8
1208 have _utf8 suffixes added by using preprocessor macros in the
1209 header file. The old names are kept for functions with the old
1210 behaviour, taking parameters in the system codepage, for DLL ABI
1213 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1214 g_spawn*() functions.
1216 * glib/gspawn-win32.c: Use wide-char API on NT-based
1217 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1218 system codepage (Win9x) and call the C library _wspawn*() or
1219 spawn*() functions respectvely. Add DLL ABI stability versions
1220 that take parameters in the system codepage.
1222 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1223 wide-char versions of argv and envp, and use wide-char API to
1224 change directory and spawn the program to run. Remove the verbose
1225 debugging output, it was too complex to modify for the wide-char
1226 features. (Just add temporary debugging printouts if needed, no
1227 need to have them permanently in the source.)
1229 * glib/gspawn.c: Corresponding documentation updates.
1231 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1232 symbols as PRIVATE, add the new _utf8-suffixed ones.
1234 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1236 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1237 guint64 otherwise and that can produce warnings about comparison
1238 between signed and unsigned.
1240 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1242 * glib/gutils.c: Fix the crt_externs.h include.
1244 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1246 * NEWS: Fix spelling of my first name.
1248 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1252 * === Released 2.8.1 ===
1256 2005-08-20 Hans Breuer <hans@breuer.org>
1258 * glib/makefile.msc.in : link with ws2_32.lib
1260 2005-08-18 Tor Lillqvist <tml@novell.com>
1262 * configure.in: Check for <sys/wait.h>
1264 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1266 2005-08-18 Ross Burton <ross@burtonini.com>
1269 Optimise single-character insertions.
1274 * tests/string-test.c:
1275 Add tests for new optimisation, and fix a leak.
1277 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1279 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1281 * glib/gutils.c: On Darwin, include crt-externs.h and
1282 define environ using _NSGetEnviron(). (#313731)
1284 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1286 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1287 (g_get_any_init_do): ... to this new function.
1288 (g_get_any_init): Declare as inline.
1289 (g_get_any_init_locked): New inline function, does the locking.
1290 Make use of these two throughout the code.
1292 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1294 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1295 the child process and then simply return. This makes
1296 The "S" option work as documented in g_on_error_query().
1297 (#313125, Matthew F. Barnes)
1299 * glib/gunicode.h: Update the link to Unicode category
1300 values. (#313369, Behnam Esfahbod)
1302 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1303 a little. (#311727, Tristan van Berkom)
1305 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1306 check work on ia64 too, where some symbols ended up
1307 in yet another section.