1 2006-02-18 Matthias Clasen <mclasen@redhat.com>
3 * tests/gobject/gvalue-test.c: Add more tests.
5 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
7 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
9 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
11 * glib/gthreadpool.c: Fix deadlock when signalling the thread
12 which freed a thread pool (#331110, Chris Wilson).
14 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
16 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
17 posix_memalign() in case we actually are going to use it, because
18 we detected a compliant implementation (#328997).
20 * configure.in (enable_included_printf): don't include malloc.h when
21 testing for posix_memalign() funcitonality, since this may break the
22 test on some systems (#328997).
24 2006-02-14 Matthias Clasen <mclasen@redhat.com>
26 * glib/gstrfuncs.c: Improve docs.
28 2006-02-11 Matthias Clasen <mclasen@redhat.com>
30 * configure.in: Bump version
32 2006-02-10 Matthias Clasen <mclasen@redhat.com>
34 * === Released 2.9.6 ===
38 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
40 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
41 the type of varargs arguments. (bug #317679)
43 2006-01-31 Matthias Clasen <mclasen@redhat.com>
45 * autogen.sh: Touch README and INSTALL here to pacify
46 automake. (#329124, Kjartan Maraas, Tim Janik)
48 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
50 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
51 some missing cases where pointed out by Sebastian Wilhelmi.
53 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
55 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
57 2006-01-30 David Schleef <ds@schleef.org>
59 * glib/grand.c: Update URL
61 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
63 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
64 create container classes which are as flexible as a GValue is
66 2006-01-27 Matthias Clasen <mclasen@redhat.com>
68 * configure.in: Bump version
70 * === Released 2.9.5 ===
72 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
77 2006-01-26 Matthias Clasen <mclasen@redhat.com>
79 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
82 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
84 * fixed buglets reported by Jens Granseuer in #328254.
86 * configure.in: free the memory allocated in posix_memalign() tests.
88 * glib/gslice.c: spelling fixes.
90 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
92 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
93 slices, make sure g_mem_gc_friendly is properly initialized.
95 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
96 the first allocation. applied some branching optimizations.
98 * docs/macros.txt: reflected --enable-gc-friendly change and
99 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
101 * configure.in: changed --enable-gc-friendly=yes to define
102 ENABLE_GC_FRIENDLY_DEFAULT.
104 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
105 if (G_UNLIKELY (g_mem_gc_friendly)).
108 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
109 care of by g_slice_free1().
111 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
113 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
115 * glib/gslice.c: only use posix_memalign() if it's known to work,
116 revert to memalign() otherwise.
118 * configure.in: check for broken posix_memalign() implementations
121 2006-01-24 Matthias Clasen <mclasen@redhat.com>
123 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
124 on all little-endian systems. (#143380, Marc Moorcroft)
126 2006-01-23 Matthias Clasen <mclasen@redhat.com>
128 * configure.in: Change the shared libary extension for hpux-ia64
129 to so. (#328253, The Written Word)
131 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
133 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
134 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
137 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
139 * glib/gslice.c (slice_config_init): initialize GSlice config from
140 G_SLICE environment variable. we support G_SLICE=always-malloc
141 currently, which forces all g_slice_*() allocations to use the system
145 g_parse_debug_string(): added a note about not using g_malloc() here.
146 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
147 g_slice. contains only guesswork in the WIN32 branch.
149 2006-01-18 Matthias Clasen <mclasen@redhat.com>
153 * === Released 2.9.4 ===
157 2006-01-17 Matthias Clasen <mclasen@redhat.com>
159 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
160 call. (#327421, Morten Welinder)
162 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
164 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
165 G_LOCK. Use the unused_thread_queue lock instead. Change
166 g_thread_pool_thread_proxy such that threads only wait on
167 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
168 due to superfluous tasks. Global tasks wait at most for
169 max-idle-time milliseconds. Make sure, that no task is woken up
170 twice for the same event via a wakeup_serial. This fixes #324228.
172 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
173 invalid NULL into the thread pools. This as well fixes #327290.
175 2006-01-16 Matthias Clasen <mclasen@redhat.com>
177 * configure.in: Bump version
179 * === Released 2.9.3 ===
183 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
184 Rework the strict aliasing fix to not break C++, pointed
185 out by Murray Cumming.
187 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
188 Signal waiting threads, problem noticed by Christian Kellner.
190 2006-01-16 Matthias Clasen <mclasen@redhat.com>
192 Fix bug 326747, Alberto Ruiz:
194 * glib/gconvert.c (make_valid_utf8): Change this function to
195 replace unknown characters by the Unicode replacement character
196 instead of '?', and don't append "(invalid encoding)".
197 (g_filename_display_name, g_filename_display_basename): Document
198 how to determine if the filename was in an invalid encoding.
200 2006-01-14 Matthias Clasen <mclasen@redhat.com>
202 * glib/gtree.c: Replace the simple recursive implementation
203 by a nonrecursive, threaded implementation by Maurizio
206 2006-01-12 Matthias Clasen <mclasen@redhat.com>
208 * tests/asyncqueue-test.c (main): Fix the
209 assert to not always trigger. (#326558,
212 2006-01-11 Matthias Clasen <mclasen@redhat.com>
214 * tests/convert-test.c: Don't test items_read and
215 items_written in error cases where it is not set.
217 * tests/asyncqueue-test.c: Set error to NULL.
219 2006-01-08 Matthias Clasen <mclasen@redhat.com>
221 * glib/gtree.c: Break some long lines.
223 * tests/tree-test.c: Add more tests.
225 * glib/gslice.c: Include config.h first, pointed out
228 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
230 * docs/reference/gobject/tmpl/gtype.sgml: explain that
231 G_TYPE_INSTANCE_GET_CLASS() does behave different during
234 2006-01-06 Matthias Clasen <mclasen@redhat.com>
236 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
237 that we have posix_memalign(). (#323937, Bogdan Nicula)
239 2006-01-05 Matthias Clasen <mclasen@redhat.com>
241 * configure.in: Bump version
243 * === Released 2.9.2 ===
245 2006-01-05 Hans Breuer <hans@breuer.org>
247 * glib/makefile.msc.in : added gslice.obj
249 2006-01-05 Matthias Clasen <mclasen@redhat.com>
253 * glib/gmain.c: Clarify the documentation of source ids
254 a bit. (#325874, Dan Williams)
256 * configure.in: Fix another strict aliasing problem.
258 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
261 2006-01-04 Manish Singh <yosh@gimp.org>
264 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
265 either, since they have the same issue as g_rmdir. (Related to
268 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
270 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
271 order when creating magazines, so we hand out chunks with ascending
274 2006-01-04 Abel Cheung <maddog@linuxhall.org>
276 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
278 2006-01-03 Martyn Russell <martyn@imendio.com>
280 * docs/reference/glib/glib-sections.txt:
282 * glib/gthreadpool.[ch]:
283 - Added new API g_thread_pool_get_idle_time() and
284 g_thread_pool_set_idle_time(). (#324228).
286 * tests/threadpool-test.c:
287 - Updated test case to do thread pool sorting, thread pool with
288 no sorting and a thread pool with idle thread timeouts.
290 2006-01-03 Matthias Clasen <mclasen@redhat.com>
292 * glib/gmain.h: Add new functions here, too.
294 * glib/glib.symbols: Add new functions.
296 2005-12-20 Michael Meeks <michael.meeks@novell.com>
298 * glib/gmain.c (g_main_context_is_owner): new method
299 to determine if the current thread is the owner of the
302 2006-01-02 Matthias Clasen <mclasen@redhat.com>
306 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
307 expanding to rmdir, since rmdir is not declared in a portable
308 system header we can include in gstdio.h. (#325249, Jani Monoses)
310 2006-01-01 Tor Lillqvist <tml@novell.com>
312 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
313 stdout and stderr to unbuffered. Otherwise the giochannel layer
314 will try to read from them regardless whether the
315 g_io_channel_win32_poll() call here has indicated
316 readability or not. (#325310)
318 2005-12-29 Matthias Clasen <mclasen@redhat.com>
320 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
321 in the docs. (#325273, Declan Naughton)
323 2005-12-27 Manish Singh <yosh@gimp.org>
325 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
326 make sure we really override things.
328 2005-12-27 Matthias Clasen <mclasen@redhat.com>
330 Fix #316221, Michal Benes, Stanislav Brabec;
332 * configure.in: Fix a strict aliasing problem in
333 g_static_mutex_get_mutex().
334 * glib/gthread.h: ...and in
335 g_static_mutex_get_mutex_impl_shortcut().
337 * glib/gdatasetprivate.h: Add a cast to silence compiler
338 warnings. (#321978, Andrew Paprocki)
340 Partial fix for bug #323937, Bogdan Nicula.
342 * configure.in: Check for malloc.h
343 * glib/gslice.c: Don't include malloc.h unconditionally.
345 2005-12-27 Manish Singh <yosh@gimp.org>
347 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
348 (#324950, Dan Yefimov)
350 2005-12-26 Matthias Clasen <mclasen@redhat.com>
352 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
355 2005-12-25 Matthias Clasen <mclasen@redhat.com>
357 * AUTHORS: Update my email
359 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
361 2005-12-21 Matthias Clasen <mclasen@redhat.com>
363 * README.in: Add some notes on when 'make check' may fail.
365 2005-12-20 Matthias Clasen <mclasen@redhat.com>
368 * glib/gthreadpool.h:
369 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
370 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
372 * tests/threadpool-test.c: Test this.
374 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
376 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
377 handle errors without depending on gmessage.c which might not be
378 setup when the error occours.
379 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
380 always freeing can be achieved by adjusting the working set time to
381 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
382 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
383 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
384 minimize wastage if small amounts of differently sized structrues are
385 allocated, this does come at a performance cost of roughly 5% though.
386 fixed up block alignment calculation, so it works for varying
387 block sizes. only use strerror() not g_strerror() since the latter
388 depends on working GQuark and GSlice.
389 mem_error(): implemented in terms of fprintf and vfprintf.
391 * tests/slice-color.c: new program to test cache colorization effects.
393 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
394 G_SLICE_CONFIG_WORKING_SET_MSECS.
396 2005-12-17 Matthias Clasen <mclasen@redhat.com>
398 * glib/goption.c (parse_short_option): Set an error in all
399 failure cases. (#324332, Tim-Philipp Müller)
401 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
403 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
404 and g_atomic_pointer_get. Add g_atomic_int_set and
405 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
406 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
407 as functions (additionally to the macros in the header) for the
408 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
410 2005-12-16 Matthias Clasen <mclasen@redhat.com>
412 * glib/gmem.c (g_allocator_new): Don't return a pointer to
413 a const struct, since apps expect to be able to modify it.
414 (#324179, J. Ali Harlow)
416 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
418 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
419 for systems where the initialization of atomic variables requires a
420 write memory barrier.
422 2005-12-09 Matthias Clasen <mclasen@redhat.com>
424 * configure.in: Bump version
426 * === Released 2.9.1 ===
428 2005-12-09 Alexander Larsson <alexl@redhat.com>
432 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
435 2005-12-08 Matthias Clasen <mclasen@redhat.com>
439 2005-12-07 Martyn Russell <martyn@imendio.com>
441 * glib/gasyncqueue.c:
442 - Call g_queue_insert_sorted() instead of duplicating the code.
443 - Call g_queue_sort() instead of duplicating the code.
444 - Invert sort function results to make sure the same sort function
445 gives the same results across glist, gslist, gqueue and
448 * tests/asyncqueue-test.c:
449 - Updated the sort function to reflect the example in the
450 documentation for gasyncqueue.c.
452 2005-12-07 Martyn Russell <martyn@imendio.com>
454 * docs/reference/glib/glib-sections.txt:
455 * docs/reference/glib/tmpl/linked_lists_double.sgml:
456 * docs/reference/glib/tmpl/linked_lists_single.sgml:
459 - Added g_list_insert_sorted_with_data () and
460 g_slist_insert_sorted_with_data ().
461 - Removed the extra check in g_list_sort() and g_slist_sort() for
462 GCompareDataFunc vs. GCompareFunc.
464 2005-12-07 Tor Lillqvist <tml@novell.com>
466 * glib-gettextize.in: Look up prefix at run-time on Win32,
467 assuming the standard directory structure with glib-gettextize in
470 2005-12-06 Matthias Clasen <mclasen@redhat.com>
472 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
473 we need GTK+ 2.8 to compile against GLib 2.10.
475 2005-12-05 Manish Singh <yosh@gimp.org>
477 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
478 to override any user settings.
480 2005-12-05 Matthias Clasen <mclasen@redhat.com>
482 * configure.in: Revert an accidential commit
484 2005-12-05 Matthias Clasen <mclasen@redhat.com>
486 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
488 * tests/Makefile.am (TESTS_ENVIRONMENT):
489 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
490 MALLOC_CHECK_ and MALLOC_PERTURB_
492 * tests/run-collate-tests.sh: Run the collation tests explicitly
493 in en_US locale. (#320463)
495 * glib/gmem.h: Really deprecate GMemChunk.
502 * glib/gstrfuncs.c: Add versioned deprecation docs.
504 2005-12-05 Martyn Russell <martyn@imendio.com>
506 * docs/reference/glib/glib-sections.txt:
507 * glib/gasyncqueue.[ch]:
508 - Added support for sorting async queues by with _push_sorted(),
509 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
512 * tests/asyncqueue-test.c:
513 - Added test case for gasyncqueue.c
515 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
517 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
518 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
519 in bug 323178. moved time consuming logic from g_slice_free() out of
520 the inner loop, so g_slice_free_chain_with_offset() provides a real
521 performance benefit over g_slice_free1() now.
523 * glib/gslice.h: renamed g_slice_free_chain() to
524 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
525 a type-safe macro as suggested in bug 323178.
526 simplified the macro implementation of g_slice_free() and implemented
527 it in a type safe manner for all compliers as suggested by Morten
528 Welinder <mortenw@gnome.org>.
533 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
535 2005-12-05 Matthias Clasen <mclasen@redhat.com>
537 * glib/gasyncqueue.c: Add some docs.
539 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
541 * glib/gatomic.c: In the ia64 implementation, use
542 __sync builtin without _si or _di suffix. (#321229,
543 Stanislav Brabec, patch by Andreas Schwab)
545 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
547 * glib/gslice.h: Remove comma at the end of enum.
549 2005-12-04 Matthias Clasen <mclasen@redhat.com>
551 Handle multiple user names with the same UID better.
552 (#319535, Laszlo Peter)
554 * glib/gutils.c (g_get_any_init_do): When determining user
555 data, first look up $LOGNAME. If the UID doesn't match
556 getuid(), fall back to the current behaviour of looking
557 up the user data based on getuid().
559 2005-12-04 Matthias Clasen <mclasen@redhat.com>
561 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
562 to make gcc warn if a function result is ignored. (#145466,
563 Arjan van de Ven, Alex Larsson)
565 * glib/gmem.h: Add the new attribute to g_realloc and
568 2005-12-03 Matthias Clasen <mclasen@redhat.com>
571 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
574 * glib/gmappedfile.c: Make mapping of empty files
577 * glib/gfileutils.c: Don't fork a new process just to
578 fix the permissions of the created temp file. (#321318,
579 Alexis S. L. Carvalho)
581 2005-12-02 Matthias Clasen <mclasen@redhat.com>
583 * README.in: Add a note about Solaris threads.
586 * configure.in: Remove support for Solaris threads.
587 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
589 2005-12-02 Matthias Clasen <mclasen@redhat.com>
591 * configure.in: Define G_GUINT64_CONSTANT in analogy to
592 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
594 2005-12-02 Matthias Clasen <mclasen@redhat.com>
596 * glib/gslice.c: Win32 portability fixes and C99ism removal,
597 pointed out by Kazuki Iwamoto. (#323052)
599 2005-12-02 Matthias Clasen <mclasen@redhat.com>
601 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
602 the IA64 implementation.
604 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
606 * glib/gslice.c: some naming and type size fixups.
608 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
610 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
611 variant into something less verbose, but digestible for gcc-3.4.
613 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
615 * tests/slice-test.c: extended to perform the benchmarking on the old
616 memchunk code if 'O' is selected.
618 * tests/memchunks.c: new file which contains the old GLib mem chunks
619 implementation with prefix old_mem_chunk_.
621 * tests/Makefile.am: added memchunks.c
623 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
625 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
626 acceptable load distribution of the syscall.
627 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
628 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
630 2005-12-01 Matthias Clasen <mclasen@redhat.com>
632 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
633 double-free base interface members.
635 * tests/Makefile.am: Remove duplicate variable.
637 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
639 * glib/gslice.[hc]: new slice allocator implementation.
641 * tests/slice-test.c: added random slice allocation test.
643 * glib/gthread.[hc]: removed newly added private thread mem API.
645 * glib/gthreadinit.h:
648 * glib/gmem.c: divided glib threading initialisation into three phases,
649 initialisation where private keys and messaging are not available (only
650 needed by gmem.c), initialisation without messaging but private keys
651 available (gslice.c, gmessage.c), and full fledged initialisers that
652 server the rest of glib. initialisation functions got renamed to reflect
653 the limitations of their corresponding phases.
655 * glib/gmem.c: removed memchunk code, defer allocations to
658 * glib/gmem.[hc]: removed g_slice_* skeletons.
660 * glib/glib.symbols: added g_slice_* symbols.
662 * configure.in: check for availability of posix_memalign(3), memalign(3)
665 * glib/Makefile.am: added gslice.[hc].
667 2005-12-01 Tor Lillqvist <tml@novell.com>
669 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
670 trailing slash(es) for non-root folders. stat() fails if non-root
671 folders are specified with trailing slashes. It's too much hassle
672 to demand that callers strip such slashes themselves, especially
673 as it is easy to get it wrong and strip the slash of a root
675 (g_rename): On NT-based Windows, use MoveFileEx() with
676 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
678 2005-11-28 Matthias Clasen <mclasen@redhat.com>
680 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
683 * configure.in: Check whether do { } while (0) works.
685 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
686 G_STMT_END if it works.
688 2005-11-28 Matthias Clasen <mclasen@redhat.com>
690 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
691 * glib/gnode.c (g_node_depth_traverse_level):
692 * glib/gmem.c (g_allocator_new):
693 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
694 (g_hash_table_foreach_remove_or_steal): Silence compiler
697 2005-11-27 Matthias Clasen <mclasen@redhat.com>
699 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
700 Don't read beyond len. (#322520, Christian Persch)
702 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
704 * glib/gdataset.c: access datalist flags via atomic pointer access
705 functions, instead of acquiring the dataset lock. this is faster and
706 also matches the atomic pointer readouts in gdatalistprivate.h and
707 direct pointer modifications required by gobject.c.
709 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
711 * glib/gdataset.c: streamlined initialization code somewhat,
712 removed GData node cache.
714 2005-11-22 Matthias Clasen <mclasen@redhat.com>
716 * glib/ghash.c (g_hash_table_ref):
717 (g_hash_table_unref): Mark these as new API.
719 Avoid double locking in g_intern_string (#322133,
722 * glib/gdataset.c (g_quark_from_string_internal): New
723 internal function which factors out common parts of
724 g_quark_from[_static]_string.
725 (g_quark_from_string, g_quark_from_static_string):
726 Use g_quark_from_string_internal.
727 (g_intern_string, g_intern_static_string): Use
728 g_quark_from_string_internal, and only take the
730 (g_quark_new): Don't store the strings shifted by -1
731 in the g_quarks array.
732 (g_quark_to_string): Adapt to the previous change.
734 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
738 g_hash_table_new_full(): create hash tables with a ref count of 1.
739 g_hash_table_ref(): atomically ref_count+=1
740 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
741 when refcount reaches 0.
742 g_hash_table_destroy(): just destroy keys and values, unref by 1.
743 g_hash_table_insert():
744 g_hash_table_replace(): assert ref_count>0.
747 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
750 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
752 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
753 variables. (#321984, Andrew Paprocki)
755 2005-11-18 Matthias Clasen <mclasen@redhat.com>
757 * configure.in: Bump version
759 2005-11-17 Matthias Clasen <mclasen@redhat.com>
761 * === Released 2.9.0 ===
763 * glib/glib.symbols: Updates
765 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
766 hints to the docs. (Paolo Borelli)
770 * glib/gthread.c (g_thread_foreach): Mark as new api.
772 * README.in: Updates.
776 * glib/gdate.c (g_date_set_time_t):
777 (g_date_set_time_val): Add functions to set a date from
778 a time_t and from a GTimeVal.
779 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
782 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
783 and g_utf8_offset_to_pointer.
785 * glib/gutf8.c (g_utf8_pointer_to_offset)
786 (g_utf8_offset_to_pointer): Handle negative offsets, and use
787 "stutter stepping" for going backwards. (#320638, Larry
792 * glib/gthread.c: const correctness fixes, found
793 by Arjan van de Ven and gcc.
795 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
797 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
798 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
800 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
802 * glib/guniprop.c: Use bit hacks instead when checking a general
803 category value against multiple values.
805 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
807 2005-11-10 Simos Xenitellis <simos@gnome.org>
809 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
811 2005-11-05 Tor Lillqvist <tml@novell.com>
813 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
814 (#320688) Instead set error and return failure.
816 2005-11-07 Matthias Clasen <mclasen@redhat.com>
818 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
819 while holding the context lock. (#320886, Andy Wingo)
821 2005-11-07 Billy Biggs <vektor@dumbterm.net>
823 * tests/refcount/Makefile.am: Add a missing reference to libglib
824 in the LDADD for this test.
826 2005-11-07 Matthias Clasen <mclasen@redhat.com>
828 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
829 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
830 GLib 2.9. (#320755, Luca Ferretti)
832 2005-11-04 Matthias Clasen <mclasen@redhat.com>
834 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
835 use a faster lookup table for g_unichar_get_mirror_char().
836 * glib/gmirroringtable.h: The new table.
837 * glib/gunichartables.h: Remove bidi_mirroring_table.
838 * glib/gen-unicode-tables.pl: Don't generate the mirroring
840 * glib/glib-mirroring-tab/*: Sources for the program
841 which generated gmirroringtable.h.
843 * glib/glist.c: Avoid some code duplication.
845 * glib/gscanner.h: Include gdataset.h. (#320322)
847 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
848 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
849 to help find critical warnings in applications. (#320017,
852 2005-11-02 Tor Lillqvist <tml@novell.com>
854 * glib/glib.symbols: Remove large amount of trailing whitespace
855 from one line. Remove #ifdef G_UNDEFINED from around
856 g_hash_table_new and g_hash_table_new_full symbols, what was the
857 purpose of that? Add the g_slice functions.
859 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
861 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
862 added g_slice_*() API to allocate and cache small bits of memory.
863 an actuall allocator implementation for g_slice_*() is still pending.
865 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
866 changed GRealThread list to use in-structure *next; fields instead
867 of GSList, in order for thread iteration to not depenend on g_slice_*()
869 _g_thread_mem_private_get():
870 _g_thread_mem_private_set(): added accessors for private memory,
871 needed because the ordinary GPrivate implementation relies on GArray
872 and GSList and therefore indirectly on working g_slice_*() allocations.
875 g_thread_foreach(): new public API function to loop over all existing threads.
889 removed GAllocator and free list usages and accompanying locks.
890 use g_slice_*() API to allocate and cache small bits of memory.
892 * glib/ghook.h: removed GMemChunk field from public API.
895 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
897 * glib/gnode.h: deprecate allocator API.
899 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
900 be used for poll rec lists.
902 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
903 g_relation_destroy(): free all tuples from the all_tuples hash table,
904 this effectively maintains the life time track keeping of tuples.
905 g_relation_delete_tuple(): free tuples which are removed from the
906 all_tuples hash table. this fixes a temporary leak that was present
907 in the memchunk code until the destruction of the relation.
909 2005-10-29 Matthias Clasen <mclasen@redhat.com>
911 * tests/convert-test.c: Add some tests for conversions between
912 UTF-8, UCS-4 and UTF-16.
914 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
915 of len == -1, noticed by Morten Welinder.
917 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
919 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
921 2005-10-26 Matthias Clasen <mclasen@redhat.com>
923 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
924 in the error case, and add some documentation. (#319806, Morten
927 2005-10-19 Manish Singh <yosh@gimp.org>
929 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
930 config.h symbol automatically. Fixes bug #313731.
932 2005-10-19 Tor Lillqvist <tml@novell.com>
934 * glib/gwin32.c (g_win32_get_package_installation_directory):
935 Return a g_strdup()ed copy of the value stored in the hash table,
936 so that it can be g_free()d without leaving a dangling pointer in
937 the hash table. (#319232)
939 2005-10-06 Matthias Clasen <mclasen@redhat.com>
941 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
942 all-zero sequences correctly. (#317930, Sebastien Bacher)
944 2005-10-05 Matthias Clasen <mclasen@redhat.com>
946 * glib/gmarkup.c (g_markup_escape_text): Doc addition
948 * Makefile.am: Add Collation tests.
950 * tests/collate/*: Inputs and expected outputs for collation tests.
952 * tests/run-collate-tests.sh: Script to run collation tests.
954 * tests/unicode-collate.c (main): Rework slightly to make
955 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
957 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
959 * docs/reference/glib/tmpl/unicode.sgml:
960 * glib/gen-unicode-tables.pl:
962 * glib/gunichartables.h:
964 * tests/casefold.txt:
965 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
966 GUnicodeBreakType types. That may break some applications, like
969 2005-09-26 Matthias Clasen <mclasen@redhat.com>
971 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
972 docs. (#314393, Matthew F. Barnes)
975 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
976 they are already declared in glib.h. This doesn't break documented
977 use of gprintf.h, but should probably be pointed out in the
978 release notes for 2.10. (#314232, Behdad Esfahbod)
980 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
982 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
983 complexity optimization, based on a patch from Matthias Clasen.
985 * tests/patterntest.c: more tests, mostly from matthias.
987 2005-09-20 Matthias Clasen <mclasen@redhat.com>
989 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
990 (#316703, Mark Drago)
992 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
994 * glib/gpattern.c: applied a patch from matthias which checks on the
995 upper bound of GPatternSpec length to optimize matches.
998 * tests/patterntest.c: added more match cases.
1000 2005-09-16 Tor Lillqvist <tml@novell.com>
1002 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1003 warning to doc comment that these functions should not be used on
1004 encodings like CP932.
1006 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1008 * tests/keyfile-test.c: Add a test for grup names of length 1.
1010 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1011 of length 1. (#316309)
1013 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1015 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1017 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1019 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1020 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1023 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1025 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1026 (child_watch_helper_thread): Remove some dead code. Closes
1029 2005-09-07 Tor Lillqvist <tml@novell.com>
1031 * glib/Makefile.am: Create also a console version of the
1032 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1033 It's otherwise identical to gspawn-win32-helper.exe, except marked
1034 as a console application (linked without the -mwindows option).
1036 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1037 Drop the dont_wait parameter. Its truth value correlated 100% with
1038 the NULLness of the exit_status parameter anyway, so it's enough
1039 to check whether exit_status is NULL. Invert the sense of the
1040 dont_return_handle parameter and rename it to do_return_handle, to
1041 make the code easier to read by avoiding double negations.
1043 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1044 to do_spawn_with_pipes() accordingly.
1046 (do_spawn_with_pipes): If we have a console, use the console
1047 version of the helper program, otherwise use the GUI one. This
1048 avoids extra console windows opening up in some situations. (In
1049 case a console application uses the GUI gspawn-win32-helper.exe to
1050 spawn another console application we would get a separate console
1051 for the spawned console application).
1053 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1055 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1057 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1058 if the file is too large. (#315275, Kjartan Maraas)
1060 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1061 of read() is signed. (#315273, Kjartan Maraas)
1063 2005-08-31 Tor Lillqvist <tml@novell.com>
1065 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1066 the EXE and each DLL have their own atexit function chains.
1068 #define g_atexit as atexit instead. This means it has a
1069 better chance of doing what the caller wants. For instance,
1070 gtkhtml calls g_atexit() registering a function in gtkhtml
1071 itself. This caused a crash when g_atexit() was implemented as a
1072 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1073 the time the GLib DLL got unloaded.
1075 * glib/gutils.c: #undef the #define mentioned above, to also get a
1076 real g_atexit() into the DLL for backward compatibility. Document
1077 the Windows behaviour of g_atexit(), and document the varying ways
1078 atexit() can behave in the context of dynamically loaded modules
1081 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1083 * glib/glib.symbols:
1085 * glib/gdataset.c: Add string interning functions.
1087 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1089 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1092 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1094 * configure.in: Bump version to 2.9.0
1096 2005-08-25 Tor Lillqvist <tml@novell.com>
1098 Make also the g_spawn*() functions take parameters in the GLib
1099 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1100 API or ABI. Like the other GLib API that was earlier changed to
1101 use UTF-8 on Windows, the names of the functions that take UTF-8
1102 have _utf8 suffixes added by using preprocessor macros in the
1103 header file. The old names are kept for functions with the old
1104 behaviour, taking parameters in the system codepage, for DLL ABI
1107 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1108 g_spawn*() functions.
1110 * glib/gspawn-win32.c: Use wide-char API on NT-based
1111 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1112 system codepage (Win9x) and call the C library _wspawn*() or
1113 spawn*() functions respectvely. Add DLL ABI stability versions
1114 that take parameters in the system codepage.
1116 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1117 wide-char versions of argv and envp, and use wide-char API to
1118 change directory and spawn the program to run. Remove the verbose
1119 debugging output, it was too complex to modify for the wide-char
1120 features. (Just add temporary debugging printouts if needed, no
1121 need to have them permanently in the source.)
1123 * glib/gspawn.c: Corresponding documentation updates.
1125 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1126 symbols as PRIVATE, add the new _utf8-suffixed ones.
1128 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1130 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1131 guint64 otherwise and that can produce warnings about comparison
1132 between signed and unsigned.
1134 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1136 * glib/gutils.c: Fix the crt_externs.h include.
1138 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1140 * NEWS: Fix spelling of my first name.
1142 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1146 * === Released 2.8.1 ===
1150 2005-08-20 Hans Breuer <hans@breuer.org>
1152 * glib/makefile.msc.in : link with ws2_32.lib
1154 2005-08-18 Tor Lillqvist <tml@novell.com>
1156 * configure.in: Check for <sys/wait.h>
1158 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1160 2005-08-18 Ross Burton <ross@burtonini.com>
1163 Optimise single-character insertions.
1168 * tests/string-test.c:
1169 Add tests for new optimisation, and fix a leak.
1171 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1173 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1175 * glib/gutils.c: On Darwin, include crt-externs.h and
1176 define environ using _NSGetEnviron(). (#313731)
1178 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1180 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1181 (g_get_any_init_do): ... to this new function.
1182 (g_get_any_init): Declare as inline.
1183 (g_get_any_init_locked): New inline function, does the locking.
1184 Make use of these two throughout the code.
1186 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1188 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1189 the child process and then simply return. This makes
1190 The "S" option work as documented in g_on_error_query().
1191 (#313125, Matthew F. Barnes)
1193 * glib/gunicode.h: Update the link to Unicode category
1194 values. (#313369, Behnam Esfahbod)
1196 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1197 a little. (#311727, Tristan van Berkom)
1199 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1200 check work on ia64 too, where some symbols ended up
1201 in yet another section.