1 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
3 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
4 posix_memalign() in case we actually are going to use it, because
5 we detected a compliant implementation (#328997).
7 * configure.in (enable_included_printf): don't include malloc.h when
8 testing for posix_memalign() funcitonality, since this may break the
9 test on some systems (#328997).
11 2006-02-14 Matthias Clasen <mclasen@redhat.com>
13 * glib/gstrfuncs.c: Improve docs.
15 2006-02-11 Matthias Clasen <mclasen@redhat.com>
17 * configure.in: Bump version
19 2006-02-10 Matthias Clasen <mclasen@redhat.com>
21 * === Released 2.9.6 ===
25 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
27 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
28 the type of varargs arguments. (bug #317679)
30 2006-01-31 Matthias Clasen <mclasen@redhat.com>
32 * autogen.sh: Touch README and INSTALL here to pacify
33 automake. (#329124, Kjartan Maraas, Tim Janik)
35 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
37 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
38 some missing cases where pointed out by Sebastian Wilhelmi.
40 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
42 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
44 2006-01-30 David Schleef <ds@schleef.org>
46 * glib/grand.c: Update URL
48 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
50 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
51 create container classes which are as flexible as a GValue is
53 2006-01-27 Matthias Clasen <mclasen@redhat.com>
55 * configure.in: Bump version
57 * === Released 2.9.5 ===
59 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
64 2006-01-26 Matthias Clasen <mclasen@redhat.com>
66 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
69 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
71 * fixed buglets reported by Jens Granseuer in #328254.
73 * configure.in: free the memory allocated in posix_memalign() tests.
75 * glib/gslice.c: spelling fixes.
77 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
79 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
80 slices, make sure g_mem_gc_friendly is properly initialized.
82 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
83 the first allocation. applied some branching optimizations.
85 * docs/macros.txt: reflected --enable-gc-friendly change and
86 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
88 * configure.in: changed --enable-gc-friendly=yes to define
89 ENABLE_GC_FRIENDLY_DEFAULT.
91 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
92 if (G_UNLIKELY (g_mem_gc_friendly)).
95 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
96 care of by g_slice_free1().
98 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
100 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
102 * glib/gslice.c: only use posix_memalign() if it's known to work,
103 revert to memalign() otherwise.
105 * configure.in: check for broken posix_memalign() implementations
108 2006-01-24 Matthias Clasen <mclasen@redhat.com>
110 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
111 on all little-endian systems. (#143380, Marc Moorcroft)
113 2006-01-23 Matthias Clasen <mclasen@redhat.com>
115 * configure.in: Change the shared libary extension for hpux-ia64
116 to so. (#328253, The Written Word)
118 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
120 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
121 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
124 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
126 * glib/gslice.c (slice_config_init): initialize GSlice config from
127 G_SLICE environment variable. we support G_SLICE=always-malloc
128 currently, which forces all g_slice_*() allocations to use the system
132 g_parse_debug_string(): added a note about not using g_malloc() here.
133 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
134 g_slice. contains only guesswork in the WIN32 branch.
136 2006-01-18 Matthias Clasen <mclasen@redhat.com>
140 * === Released 2.9.4 ===
144 2006-01-17 Matthias Clasen <mclasen@redhat.com>
146 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
147 call. (#327421, Morten Welinder)
149 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
151 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
152 G_LOCK. Use the unused_thread_queue lock instead. Change
153 g_thread_pool_thread_proxy such that threads only wait on
154 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
155 due to superfluous tasks. Global tasks wait at most for
156 max-idle-time milliseconds. Make sure, that no task is woken up
157 twice for the same event via a wakeup_serial. This fixes #324228.
159 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
160 invalid NULL into the thread pools. This as well fixes #327290.
162 2006-01-16 Matthias Clasen <mclasen@redhat.com>
164 * configure.in: Bump version
166 * === Released 2.9.3 ===
170 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
171 Rework the strict aliasing fix to not break C++, pointed
172 out by Murray Cumming.
174 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
175 Signal waiting threads, problem noticed by Christian Kellner.
177 2006-01-16 Matthias Clasen <mclasen@redhat.com>
179 Fix bug 326747, Alberto Ruiz:
181 * glib/gconvert.c (make_valid_utf8): Change this function to
182 replace unknown characters by the Unicode replacement character
183 instead of '?', and don't append "(invalid encoding)".
184 (g_filename_display_name, g_filename_display_basename): Document
185 how to determine if the filename was in an invalid encoding.
187 2006-01-14 Matthias Clasen <mclasen@redhat.com>
189 * glib/gtree.c: Replace the simple recursive implementation
190 by a nonrecursive, threaded implementation by Maurizio
193 2006-01-12 Matthias Clasen <mclasen@redhat.com>
195 * tests/asyncqueue-test.c (main): Fix the
196 assert to not always trigger. (#326558,
199 2006-01-11 Matthias Clasen <mclasen@redhat.com>
201 * tests/convert-test.c: Don't test items_read and
202 items_written in error cases where it is not set.
204 * tests/asyncqueue-test.c: Set error to NULL.
206 2006-01-08 Matthias Clasen <mclasen@redhat.com>
208 * glib/gtree.c: Break some long lines.
210 * tests/tree-test.c: Add more tests.
212 * glib/gslice.c: Include config.h first, pointed out
215 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
217 * docs/reference/gobject/tmpl/gtype.sgml: explain that
218 G_TYPE_INSTANCE_GET_CLASS() does behave different during
221 2006-01-06 Matthias Clasen <mclasen@redhat.com>
223 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
224 that we have posix_memalign(). (#323937, Bogdan Nicula)
226 2006-01-05 Matthias Clasen <mclasen@redhat.com>
228 * configure.in: Bump version
230 * === Released 2.9.2 ===
232 2006-01-05 Hans Breuer <hans@breuer.org>
234 * glib/makefile.msc.in : added gslice.obj
236 2006-01-05 Matthias Clasen <mclasen@redhat.com>
240 * glib/gmain.c: Clarify the documentation of source ids
241 a bit. (#325874, Dan Williams)
243 * configure.in: Fix another strict aliasing problem.
245 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
248 2006-01-04 Manish Singh <yosh@gimp.org>
251 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
252 either, since they have the same issue as g_rmdir. (Related to
255 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
257 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
258 order when creating magazines, so we hand out chunks with ascending
261 2006-01-04 Abel Cheung <maddog@linuxhall.org>
263 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
265 2006-01-03 Martyn Russell <martyn@imendio.com>
267 * docs/reference/glib/glib-sections.txt:
269 * glib/gthreadpool.[ch]:
270 - Added new API g_thread_pool_get_idle_time() and
271 g_thread_pool_set_idle_time(). (#324228).
273 * tests/threadpool-test.c:
274 - Updated test case to do thread pool sorting, thread pool with
275 no sorting and a thread pool with idle thread timeouts.
277 2006-01-03 Matthias Clasen <mclasen@redhat.com>
279 * glib/gmain.h: Add new functions here, too.
281 * glib/glib.symbols: Add new functions.
283 2005-12-20 Michael Meeks <michael.meeks@novell.com>
285 * glib/gmain.c (g_main_context_is_owner): new method
286 to determine if the current thread is the owner of the
289 2006-01-02 Matthias Clasen <mclasen@redhat.com>
293 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
294 expanding to rmdir, since rmdir is not declared in a portable
295 system header we can include in gstdio.h. (#325249, Jani Monoses)
297 2006-01-01 Tor Lillqvist <tml@novell.com>
299 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
300 stdout and stderr to unbuffered. Otherwise the giochannel layer
301 will try to read from them regardless whether the
302 g_io_channel_win32_poll() call here has indicated
303 readability or not. (#325310)
305 2005-12-29 Matthias Clasen <mclasen@redhat.com>
307 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
308 in the docs. (#325273, Declan Naughton)
310 2005-12-27 Manish Singh <yosh@gimp.org>
312 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
313 make sure we really override things.
315 2005-12-27 Matthias Clasen <mclasen@redhat.com>
317 Fix #316221, Michal Benes, Stanislav Brabec;
319 * configure.in: Fix a strict aliasing problem in
320 g_static_mutex_get_mutex().
321 * glib/gthread.h: ...and in
322 g_static_mutex_get_mutex_impl_shortcut().
324 * glib/gdatasetprivate.h: Add a cast to silence compiler
325 warnings. (#321978, Andrew Paprocki)
327 Partial fix for bug #323937, Bogdan Nicula.
329 * configure.in: Check for malloc.h
330 * glib/gslice.c: Don't include malloc.h unconditionally.
332 2005-12-27 Manish Singh <yosh@gimp.org>
334 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
335 (#324950, Dan Yefimov)
337 2005-12-26 Matthias Clasen <mclasen@redhat.com>
339 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
342 2005-12-25 Matthias Clasen <mclasen@redhat.com>
344 * AUTHORS: Update my email
346 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
348 2005-12-21 Matthias Clasen <mclasen@redhat.com>
350 * README.in: Add some notes on when 'make check' may fail.
352 2005-12-20 Matthias Clasen <mclasen@redhat.com>
355 * glib/gthreadpool.h:
356 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
357 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
359 * tests/threadpool-test.c: Test this.
361 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
363 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
364 handle errors without depending on gmessage.c which might not be
365 setup when the error occours.
366 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
367 always freeing can be achieved by adjusting the working set time to
368 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
369 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
370 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
371 minimize wastage if small amounts of differently sized structrues are
372 allocated, this does come at a performance cost of roughly 5% though.
373 fixed up block alignment calculation, so it works for varying
374 block sizes. only use strerror() not g_strerror() since the latter
375 depends on working GQuark and GSlice.
376 mem_error(): implemented in terms of fprintf and vfprintf.
378 * tests/slice-color.c: new program to test cache colorization effects.
380 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
381 G_SLICE_CONFIG_WORKING_SET_MSECS.
383 2005-12-17 Matthias Clasen <mclasen@redhat.com>
385 * glib/goption.c (parse_short_option): Set an error in all
386 failure cases. (#324332, Tim-Philipp Müller)
388 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
390 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
391 and g_atomic_pointer_get. Add g_atomic_int_set and
392 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
393 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
394 as functions (additionally to the macros in the header) for the
395 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
397 2005-12-16 Matthias Clasen <mclasen@redhat.com>
399 * glib/gmem.c (g_allocator_new): Don't return a pointer to
400 a const struct, since apps expect to be able to modify it.
401 (#324179, J. Ali Harlow)
403 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
405 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
406 for systems where the initialization of atomic variables requires a
407 write memory barrier.
409 2005-12-09 Matthias Clasen <mclasen@redhat.com>
411 * configure.in: Bump version
413 * === Released 2.9.1 ===
415 2005-12-09 Alexander Larsson <alexl@redhat.com>
419 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
422 2005-12-08 Matthias Clasen <mclasen@redhat.com>
426 2005-12-07 Martyn Russell <martyn@imendio.com>
428 * glib/gasyncqueue.c:
429 - Call g_queue_insert_sorted() instead of duplicating the code.
430 - Call g_queue_sort() instead of duplicating the code.
431 - Invert sort function results to make sure the same sort function
432 gives the same results across glist, gslist, gqueue and
435 * tests/asyncqueue-test.c:
436 - Updated the sort function to reflect the example in the
437 documentation for gasyncqueue.c.
439 2005-12-07 Martyn Russell <martyn@imendio.com>
441 * docs/reference/glib/glib-sections.txt:
442 * docs/reference/glib/tmpl/linked_lists_double.sgml:
443 * docs/reference/glib/tmpl/linked_lists_single.sgml:
446 - Added g_list_insert_sorted_with_data () and
447 g_slist_insert_sorted_with_data ().
448 - Removed the extra check in g_list_sort() and g_slist_sort() for
449 GCompareDataFunc vs. GCompareFunc.
451 2005-12-07 Tor Lillqvist <tml@novell.com>
453 * glib-gettextize.in: Look up prefix at run-time on Win32,
454 assuming the standard directory structure with glib-gettextize in
457 2005-12-06 Matthias Clasen <mclasen@redhat.com>
459 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
460 we need GTK+ 2.8 to compile against GLib 2.10.
462 2005-12-05 Manish Singh <yosh@gimp.org>
464 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
465 to override any user settings.
467 2005-12-05 Matthias Clasen <mclasen@redhat.com>
469 * configure.in: Revert an accidential commit
471 2005-12-05 Matthias Clasen <mclasen@redhat.com>
473 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
475 * tests/Makefile.am (TESTS_ENVIRONMENT):
476 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
477 MALLOC_CHECK_ and MALLOC_PERTURB_
479 * tests/run-collate-tests.sh: Run the collation tests explicitly
480 in en_US locale. (#320463)
482 * glib/gmem.h: Really deprecate GMemChunk.
489 * glib/gstrfuncs.c: Add versioned deprecation docs.
491 2005-12-05 Martyn Russell <martyn@imendio.com>
493 * docs/reference/glib/glib-sections.txt:
494 * glib/gasyncqueue.[ch]:
495 - Added support for sorting async queues by with _push_sorted(),
496 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
499 * tests/asyncqueue-test.c:
500 - Added test case for gasyncqueue.c
502 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
504 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
505 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
506 in bug 323178. moved time consuming logic from g_slice_free() out of
507 the inner loop, so g_slice_free_chain_with_offset() provides a real
508 performance benefit over g_slice_free1() now.
510 * glib/gslice.h: renamed g_slice_free_chain() to
511 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
512 a type-safe macro as suggested in bug 323178.
513 simplified the macro implementation of g_slice_free() and implemented
514 it in a type safe manner for all compliers as suggested by Morten
515 Welinder <mortenw@gnome.org>.
520 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
522 2005-12-05 Matthias Clasen <mclasen@redhat.com>
524 * glib/gasyncqueue.c: Add some docs.
526 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
528 * glib/gatomic.c: In the ia64 implementation, use
529 __sync builtin without _si or _di suffix. (#321229,
530 Stanislav Brabec, patch by Andreas Schwab)
532 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
534 * glib/gslice.h: Remove comma at the end of enum.
536 2005-12-04 Matthias Clasen <mclasen@redhat.com>
538 Handle multiple user names with the same UID better.
539 (#319535, Laszlo Peter)
541 * glib/gutils.c (g_get_any_init_do): When determining user
542 data, first look up $LOGNAME. If the UID doesn't match
543 getuid(), fall back to the current behaviour of looking
544 up the user data based on getuid().
546 2005-12-04 Matthias Clasen <mclasen@redhat.com>
548 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
549 to make gcc warn if a function result is ignored. (#145466,
550 Arjan van de Ven, Alex Larsson)
552 * glib/gmem.h: Add the new attribute to g_realloc and
555 2005-12-03 Matthias Clasen <mclasen@redhat.com>
558 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
561 * glib/gmappedfile.c: Make mapping of empty files
564 * glib/gfileutils.c: Don't fork a new process just to
565 fix the permissions of the created temp file. (#321318,
566 Alexis S. L. Carvalho)
568 2005-12-02 Matthias Clasen <mclasen@redhat.com>
570 * README.in: Add a note about Solaris threads.
573 * configure.in: Remove support for Solaris threads.
574 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
576 2005-12-02 Matthias Clasen <mclasen@redhat.com>
578 * configure.in: Define G_GUINT64_CONSTANT in analogy to
579 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
581 2005-12-02 Matthias Clasen <mclasen@redhat.com>
583 * glib/gslice.c: Win32 portability fixes and C99ism removal,
584 pointed out by Kazuki Iwamoto. (#323052)
586 2005-12-02 Matthias Clasen <mclasen@redhat.com>
588 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
589 the IA64 implementation.
591 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
593 * glib/gslice.c: some naming and type size fixups.
595 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
597 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
598 variant into something less verbose, but digestible for gcc-3.4.
600 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
602 * tests/slice-test.c: extended to perform the benchmarking on the old
603 memchunk code if 'O' is selected.
605 * tests/memchunks.c: new file which contains the old GLib mem chunks
606 implementation with prefix old_mem_chunk_.
608 * tests/Makefile.am: added memchunks.c
610 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
612 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
613 acceptable load distribution of the syscall.
614 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
615 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
617 2005-12-01 Matthias Clasen <mclasen@redhat.com>
619 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
620 double-free base interface members.
622 * tests/Makefile.am: Remove duplicate variable.
624 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
626 * glib/gslice.[hc]: new slice allocator implementation.
628 * tests/slice-test.c: added random slice allocation test.
630 * glib/gthread.[hc]: removed newly added private thread mem API.
632 * glib/gthreadinit.h:
635 * glib/gmem.c: divided glib threading initialisation into three phases,
636 initialisation where private keys and messaging are not available (only
637 needed by gmem.c), initialisation without messaging but private keys
638 available (gslice.c, gmessage.c), and full fledged initialisers that
639 server the rest of glib. initialisation functions got renamed to reflect
640 the limitations of their corresponding phases.
642 * glib/gmem.c: removed memchunk code, defer allocations to
645 * glib/gmem.[hc]: removed g_slice_* skeletons.
647 * glib/glib.symbols: added g_slice_* symbols.
649 * configure.in: check for availability of posix_memalign(3), memalign(3)
652 * glib/Makefile.am: added gslice.[hc].
654 2005-12-01 Tor Lillqvist <tml@novell.com>
656 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
657 trailing slash(es) for non-root folders. stat() fails if non-root
658 folders are specified with trailing slashes. It's too much hassle
659 to demand that callers strip such slashes themselves, especially
660 as it is easy to get it wrong and strip the slash of a root
662 (g_rename): On NT-based Windows, use MoveFileEx() with
663 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
665 2005-11-28 Matthias Clasen <mclasen@redhat.com>
667 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
670 * configure.in: Check whether do { } while (0) works.
672 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
673 G_STMT_END if it works.
675 2005-11-28 Matthias Clasen <mclasen@redhat.com>
677 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
678 * glib/gnode.c (g_node_depth_traverse_level):
679 * glib/gmem.c (g_allocator_new):
680 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
681 (g_hash_table_foreach_remove_or_steal): Silence compiler
684 2005-11-27 Matthias Clasen <mclasen@redhat.com>
686 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
687 Don't read beyond len. (#322520, Christian Persch)
689 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
691 * glib/gdataset.c: access datalist flags via atomic pointer access
692 functions, instead of acquiring the dataset lock. this is faster and
693 also matches the atomic pointer readouts in gdatalistprivate.h and
694 direct pointer modifications required by gobject.c.
696 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
698 * glib/gdataset.c: streamlined initialization code somewhat,
699 removed GData node cache.
701 2005-11-22 Matthias Clasen <mclasen@redhat.com>
703 * glib/ghash.c (g_hash_table_ref):
704 (g_hash_table_unref): Mark these as new API.
706 Avoid double locking in g_intern_string (#322133,
709 * glib/gdataset.c (g_quark_from_string_internal): New
710 internal function which factors out common parts of
711 g_quark_from[_static]_string.
712 (g_quark_from_string, g_quark_from_static_string):
713 Use g_quark_from_string_internal.
714 (g_intern_string, g_intern_static_string): Use
715 g_quark_from_string_internal, and only take the
717 (g_quark_new): Don't store the strings shifted by -1
718 in the g_quarks array.
719 (g_quark_to_string): Adapt to the previous change.
721 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
725 g_hash_table_new_full(): create hash tables with a ref count of 1.
726 g_hash_table_ref(): atomically ref_count+=1
727 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
728 when refcount reaches 0.
729 g_hash_table_destroy(): just destroy keys and values, unref by 1.
730 g_hash_table_insert():
731 g_hash_table_replace(): assert ref_count>0.
734 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
737 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
739 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
740 variables. (#321984, Andrew Paprocki)
742 2005-11-18 Matthias Clasen <mclasen@redhat.com>
744 * configure.in: Bump version
746 2005-11-17 Matthias Clasen <mclasen@redhat.com>
748 * === Released 2.9.0 ===
750 * glib/glib.symbols: Updates
752 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
753 hints to the docs. (Paolo Borelli)
757 * glib/gthread.c (g_thread_foreach): Mark as new api.
759 * README.in: Updates.
763 * glib/gdate.c (g_date_set_time_t):
764 (g_date_set_time_val): Add functions to set a date from
765 a time_t and from a GTimeVal.
766 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
769 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
770 and g_utf8_offset_to_pointer.
772 * glib/gutf8.c (g_utf8_pointer_to_offset)
773 (g_utf8_offset_to_pointer): Handle negative offsets, and use
774 "stutter stepping" for going backwards. (#320638, Larry
779 * glib/gthread.c: const correctness fixes, found
780 by Arjan van de Ven and gcc.
782 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
784 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
785 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
787 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
789 * glib/guniprop.c: Use bit hacks instead when checking a general
790 category value against multiple values.
792 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
794 2005-11-10 Simos Xenitellis <simos@gnome.org>
796 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
798 2005-11-05 Tor Lillqvist <tml@novell.com>
800 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
801 (#320688) Instead set error and return failure.
803 2005-11-07 Matthias Clasen <mclasen@redhat.com>
805 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
806 while holding the context lock. (#320886, Andy Wingo)
808 2005-11-07 Billy Biggs <vektor@dumbterm.net>
810 * tests/refcount/Makefile.am: Add a missing reference to libglib
811 in the LDADD for this test.
813 2005-11-07 Matthias Clasen <mclasen@redhat.com>
815 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
816 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
817 GLib 2.9. (#320755, Luca Ferretti)
819 2005-11-04 Matthias Clasen <mclasen@redhat.com>
821 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
822 use a faster lookup table for g_unichar_get_mirror_char().
823 * glib/gmirroringtable.h: The new table.
824 * glib/gunichartables.h: Remove bidi_mirroring_table.
825 * glib/gen-unicode-tables.pl: Don't generate the mirroring
827 * glib/glib-mirroring-tab/*: Sources for the program
828 which generated gmirroringtable.h.
830 * glib/glist.c: Avoid some code duplication.
832 * glib/gscanner.h: Include gdataset.h. (#320322)
834 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
835 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
836 to help find critical warnings in applications. (#320017,
839 2005-11-02 Tor Lillqvist <tml@novell.com>
841 * glib/glib.symbols: Remove large amount of trailing whitespace
842 from one line. Remove #ifdef G_UNDEFINED from around
843 g_hash_table_new and g_hash_table_new_full symbols, what was the
844 purpose of that? Add the g_slice functions.
846 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
848 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
849 added g_slice_*() API to allocate and cache small bits of memory.
850 an actuall allocator implementation for g_slice_*() is still pending.
852 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
853 changed GRealThread list to use in-structure *next; fields instead
854 of GSList, in order for thread iteration to not depenend on g_slice_*()
856 _g_thread_mem_private_get():
857 _g_thread_mem_private_set(): added accessors for private memory,
858 needed because the ordinary GPrivate implementation relies on GArray
859 and GSList and therefore indirectly on working g_slice_*() allocations.
862 g_thread_foreach(): new public API function to loop over all existing threads.
876 removed GAllocator and free list usages and accompanying locks.
877 use g_slice_*() API to allocate and cache small bits of memory.
879 * glib/ghook.h: removed GMemChunk field from public API.
882 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
884 * glib/gnode.h: deprecate allocator API.
886 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
887 be used for poll rec lists.
889 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
890 g_relation_destroy(): free all tuples from the all_tuples hash table,
891 this effectively maintains the life time track keeping of tuples.
892 g_relation_delete_tuple(): free tuples which are removed from the
893 all_tuples hash table. this fixes a temporary leak that was present
894 in the memchunk code until the destruction of the relation.
896 2005-10-29 Matthias Clasen <mclasen@redhat.com>
898 * tests/convert-test.c: Add some tests for conversions between
899 UTF-8, UCS-4 and UTF-16.
901 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
902 of len == -1, noticed by Morten Welinder.
904 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
906 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
908 2005-10-26 Matthias Clasen <mclasen@redhat.com>
910 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
911 in the error case, and add some documentation. (#319806, Morten
914 2005-10-19 Manish Singh <yosh@gimp.org>
916 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
917 config.h symbol automatically. Fixes bug #313731.
919 2005-10-19 Tor Lillqvist <tml@novell.com>
921 * glib/gwin32.c (g_win32_get_package_installation_directory):
922 Return a g_strdup()ed copy of the value stored in the hash table,
923 so that it can be g_free()d without leaving a dangling pointer in
924 the hash table. (#319232)
926 2005-10-06 Matthias Clasen <mclasen@redhat.com>
928 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
929 all-zero sequences correctly. (#317930, Sebastien Bacher)
931 2005-10-05 Matthias Clasen <mclasen@redhat.com>
933 * glib/gmarkup.c (g_markup_escape_text): Doc addition
935 * Makefile.am: Add Collation tests.
937 * tests/collate/*: Inputs and expected outputs for collation tests.
939 * tests/run-collate-tests.sh: Script to run collation tests.
941 * tests/unicode-collate.c (main): Rework slightly to make
942 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
944 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
946 * docs/reference/glib/tmpl/unicode.sgml:
947 * glib/gen-unicode-tables.pl:
949 * glib/gunichartables.h:
951 * tests/casefold.txt:
952 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
953 GUnicodeBreakType types. That may break some applications, like
956 2005-09-26 Matthias Clasen <mclasen@redhat.com>
958 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
959 docs. (#314393, Matthew F. Barnes)
962 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
963 they are already declared in glib.h. This doesn't break documented
964 use of gprintf.h, but should probably be pointed out in the
965 release notes for 2.10. (#314232, Behdad Esfahbod)
967 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
969 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
970 complexity optimization, based on a patch from Matthias Clasen.
972 * tests/patterntest.c: more tests, mostly from matthias.
974 2005-09-20 Matthias Clasen <mclasen@redhat.com>
976 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
977 (#316703, Mark Drago)
979 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
981 * glib/gpattern.c: applied a patch from matthias which checks on the
982 upper bound of GPatternSpec length to optimize matches.
985 * tests/patterntest.c: added more match cases.
987 2005-09-16 Tor Lillqvist <tml@novell.com>
989 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
990 warning to doc comment that these functions should not be used on
991 encodings like CP932.
993 2005-09-14 Matthias Clasen <mclasen@redhat.com>
995 * tests/keyfile-test.c: Add a test for grup names of length 1.
997 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
998 of length 1. (#316309)
1000 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1002 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1004 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1006 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1007 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1010 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1012 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1013 (child_watch_helper_thread): Remove some dead code. Closes
1016 2005-09-07 Tor Lillqvist <tml@novell.com>
1018 * glib/Makefile.am: Create also a console version of the
1019 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1020 It's otherwise identical to gspawn-win32-helper.exe, except marked
1021 as a console application (linked without the -mwindows option).
1023 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1024 Drop the dont_wait parameter. Its truth value correlated 100% with
1025 the NULLness of the exit_status parameter anyway, so it's enough
1026 to check whether exit_status is NULL. Invert the sense of the
1027 dont_return_handle parameter and rename it to do_return_handle, to
1028 make the code easier to read by avoiding double negations.
1030 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1031 to do_spawn_with_pipes() accordingly.
1033 (do_spawn_with_pipes): If we have a console, use the console
1034 version of the helper program, otherwise use the GUI one. This
1035 avoids extra console windows opening up in some situations. (In
1036 case a console application uses the GUI gspawn-win32-helper.exe to
1037 spawn another console application we would get a separate console
1038 for the spawned console application).
1040 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1042 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1044 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1045 if the file is too large. (#315275, Kjartan Maraas)
1047 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1048 of read() is signed. (#315273, Kjartan Maraas)
1050 2005-08-31 Tor Lillqvist <tml@novell.com>
1052 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1053 the EXE and each DLL have their own atexit function chains.
1055 #define g_atexit as atexit instead. This means it has a
1056 better chance of doing what the caller wants. For instance,
1057 gtkhtml calls g_atexit() registering a function in gtkhtml
1058 itself. This caused a crash when g_atexit() was implemented as a
1059 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1060 the time the GLib DLL got unloaded.
1062 * glib/gutils.c: #undef the #define mentioned above, to also get a
1063 real g_atexit() into the DLL for backward compatibility. Document
1064 the Windows behaviour of g_atexit(), and document the varying ways
1065 atexit() can behave in the context of dynamically loaded modules
1068 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1070 * glib/glib.symbols:
1072 * glib/gdataset.c: Add string interning functions.
1074 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1076 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1079 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1081 * configure.in: Bump version to 2.9.0
1083 2005-08-25 Tor Lillqvist <tml@novell.com>
1085 Make also the g_spawn*() functions take parameters in the GLib
1086 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1087 API or ABI. Like the other GLib API that was earlier changed to
1088 use UTF-8 on Windows, the names of the functions that take UTF-8
1089 have _utf8 suffixes added by using preprocessor macros in the
1090 header file. The old names are kept for functions with the old
1091 behaviour, taking parameters in the system codepage, for DLL ABI
1094 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1095 g_spawn*() functions.
1097 * glib/gspawn-win32.c: Use wide-char API on NT-based
1098 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1099 system codepage (Win9x) and call the C library _wspawn*() or
1100 spawn*() functions respectvely. Add DLL ABI stability versions
1101 that take parameters in the system codepage.
1103 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1104 wide-char versions of argv and envp, and use wide-char API to
1105 change directory and spawn the program to run. Remove the verbose
1106 debugging output, it was too complex to modify for the wide-char
1107 features. (Just add temporary debugging printouts if needed, no
1108 need to have them permanently in the source.)
1110 * glib/gspawn.c: Corresponding documentation updates.
1112 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1113 symbols as PRIVATE, add the new _utf8-suffixed ones.
1115 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1117 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1118 guint64 otherwise and that can produce warnings about comparison
1119 between signed and unsigned.
1121 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1123 * glib/gutils.c: Fix the crt_externs.h include.
1125 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1127 * NEWS: Fix spelling of my first name.
1129 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1133 * === Released 2.8.1 ===
1137 2005-08-20 Hans Breuer <hans@breuer.org>
1139 * glib/makefile.msc.in : link with ws2_32.lib
1141 2005-08-18 Tor Lillqvist <tml@novell.com>
1143 * configure.in: Check for <sys/wait.h>
1145 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1147 2005-08-18 Ross Burton <ross@burtonini.com>
1150 Optimise single-character insertions.
1155 * tests/string-test.c:
1156 Add tests for new optimisation, and fix a leak.
1158 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1160 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1162 * glib/gutils.c: On Darwin, include crt-externs.h and
1163 define environ using _NSGetEnviron(). (#313731)
1165 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1167 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1168 (g_get_any_init_do): ... to this new function.
1169 (g_get_any_init): Declare as inline.
1170 (g_get_any_init_locked): New inline function, does the locking.
1171 Make use of these two throughout the code.
1173 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1175 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1176 the child process and then simply return. This makes
1177 The "S" option work as documented in g_on_error_query().
1178 (#313125, Matthew F. Barnes)
1180 * glib/gunicode.h: Update the link to Unicode category
1181 values. (#313369, Behnam Esfahbod)
1183 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1184 a little. (#311727, Tristan van Berkom)
1186 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1187 check work on ia64 too, where some symbols ended up
1188 in yet another section.