1 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
3 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
5 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
7 * glib/gthreadpool.c: Fix deadlock when signalling the thread
8 which freed a thread pool (#331110, Chris Wilson).
10 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
12 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
13 posix_memalign() in case we actually are going to use it, because
14 we detected a compliant implementation (#328997).
16 * configure.in (enable_included_printf): don't include malloc.h when
17 testing for posix_memalign() funcitonality, since this may break the
18 test on some systems (#328997).
20 2006-02-14 Matthias Clasen <mclasen@redhat.com>
22 * glib/gstrfuncs.c: Improve docs.
24 2006-02-11 Matthias Clasen <mclasen@redhat.com>
26 * configure.in: Bump version
28 2006-02-10 Matthias Clasen <mclasen@redhat.com>
30 * === Released 2.9.6 ===
34 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
36 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
37 the type of varargs arguments. (bug #317679)
39 2006-01-31 Matthias Clasen <mclasen@redhat.com>
41 * autogen.sh: Touch README and INSTALL here to pacify
42 automake. (#329124, Kjartan Maraas, Tim Janik)
44 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
46 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
47 some missing cases where pointed out by Sebastian Wilhelmi.
49 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
51 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
53 2006-01-30 David Schleef <ds@schleef.org>
55 * glib/grand.c: Update URL
57 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
59 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
60 create container classes which are as flexible as a GValue is
62 2006-01-27 Matthias Clasen <mclasen@redhat.com>
64 * configure.in: Bump version
66 * === Released 2.9.5 ===
68 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
73 2006-01-26 Matthias Clasen <mclasen@redhat.com>
75 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
78 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
80 * fixed buglets reported by Jens Granseuer in #328254.
82 * configure.in: free the memory allocated in posix_memalign() tests.
84 * glib/gslice.c: spelling fixes.
86 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
88 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
89 slices, make sure g_mem_gc_friendly is properly initialized.
91 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
92 the first allocation. applied some branching optimizations.
94 * docs/macros.txt: reflected --enable-gc-friendly change and
95 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
97 * configure.in: changed --enable-gc-friendly=yes to define
98 ENABLE_GC_FRIENDLY_DEFAULT.
100 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
101 if (G_UNLIKELY (g_mem_gc_friendly)).
104 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
105 care of by g_slice_free1().
107 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
109 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
111 * glib/gslice.c: only use posix_memalign() if it's known to work,
112 revert to memalign() otherwise.
114 * configure.in: check for broken posix_memalign() implementations
117 2006-01-24 Matthias Clasen <mclasen@redhat.com>
119 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
120 on all little-endian systems. (#143380, Marc Moorcroft)
122 2006-01-23 Matthias Clasen <mclasen@redhat.com>
124 * configure.in: Change the shared libary extension for hpux-ia64
125 to so. (#328253, The Written Word)
127 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
129 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
130 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
133 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
135 * glib/gslice.c (slice_config_init): initialize GSlice config from
136 G_SLICE environment variable. we support G_SLICE=always-malloc
137 currently, which forces all g_slice_*() allocations to use the system
141 g_parse_debug_string(): added a note about not using g_malloc() here.
142 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
143 g_slice. contains only guesswork in the WIN32 branch.
145 2006-01-18 Matthias Clasen <mclasen@redhat.com>
149 * === Released 2.9.4 ===
153 2006-01-17 Matthias Clasen <mclasen@redhat.com>
155 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
156 call. (#327421, Morten Welinder)
158 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
160 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
161 G_LOCK. Use the unused_thread_queue lock instead. Change
162 g_thread_pool_thread_proxy such that threads only wait on
163 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
164 due to superfluous tasks. Global tasks wait at most for
165 max-idle-time milliseconds. Make sure, that no task is woken up
166 twice for the same event via a wakeup_serial. This fixes #324228.
168 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
169 invalid NULL into the thread pools. This as well fixes #327290.
171 2006-01-16 Matthias Clasen <mclasen@redhat.com>
173 * configure.in: Bump version
175 * === Released 2.9.3 ===
179 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
180 Rework the strict aliasing fix to not break C++, pointed
181 out by Murray Cumming.
183 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
184 Signal waiting threads, problem noticed by Christian Kellner.
186 2006-01-16 Matthias Clasen <mclasen@redhat.com>
188 Fix bug 326747, Alberto Ruiz:
190 * glib/gconvert.c (make_valid_utf8): Change this function to
191 replace unknown characters by the Unicode replacement character
192 instead of '?', and don't append "(invalid encoding)".
193 (g_filename_display_name, g_filename_display_basename): Document
194 how to determine if the filename was in an invalid encoding.
196 2006-01-14 Matthias Clasen <mclasen@redhat.com>
198 * glib/gtree.c: Replace the simple recursive implementation
199 by a nonrecursive, threaded implementation by Maurizio
202 2006-01-12 Matthias Clasen <mclasen@redhat.com>
204 * tests/asyncqueue-test.c (main): Fix the
205 assert to not always trigger. (#326558,
208 2006-01-11 Matthias Clasen <mclasen@redhat.com>
210 * tests/convert-test.c: Don't test items_read and
211 items_written in error cases where it is not set.
213 * tests/asyncqueue-test.c: Set error to NULL.
215 2006-01-08 Matthias Clasen <mclasen@redhat.com>
217 * glib/gtree.c: Break some long lines.
219 * tests/tree-test.c: Add more tests.
221 * glib/gslice.c: Include config.h first, pointed out
224 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
226 * docs/reference/gobject/tmpl/gtype.sgml: explain that
227 G_TYPE_INSTANCE_GET_CLASS() does behave different during
230 2006-01-06 Matthias Clasen <mclasen@redhat.com>
232 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
233 that we have posix_memalign(). (#323937, Bogdan Nicula)
235 2006-01-05 Matthias Clasen <mclasen@redhat.com>
237 * configure.in: Bump version
239 * === Released 2.9.2 ===
241 2006-01-05 Hans Breuer <hans@breuer.org>
243 * glib/makefile.msc.in : added gslice.obj
245 2006-01-05 Matthias Clasen <mclasen@redhat.com>
249 * glib/gmain.c: Clarify the documentation of source ids
250 a bit. (#325874, Dan Williams)
252 * configure.in: Fix another strict aliasing problem.
254 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
257 2006-01-04 Manish Singh <yosh@gimp.org>
260 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
261 either, since they have the same issue as g_rmdir. (Related to
264 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
266 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
267 order when creating magazines, so we hand out chunks with ascending
270 2006-01-04 Abel Cheung <maddog@linuxhall.org>
272 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
274 2006-01-03 Martyn Russell <martyn@imendio.com>
276 * docs/reference/glib/glib-sections.txt:
278 * glib/gthreadpool.[ch]:
279 - Added new API g_thread_pool_get_idle_time() and
280 g_thread_pool_set_idle_time(). (#324228).
282 * tests/threadpool-test.c:
283 - Updated test case to do thread pool sorting, thread pool with
284 no sorting and a thread pool with idle thread timeouts.
286 2006-01-03 Matthias Clasen <mclasen@redhat.com>
288 * glib/gmain.h: Add new functions here, too.
290 * glib/glib.symbols: Add new functions.
292 2005-12-20 Michael Meeks <michael.meeks@novell.com>
294 * glib/gmain.c (g_main_context_is_owner): new method
295 to determine if the current thread is the owner of the
298 2006-01-02 Matthias Clasen <mclasen@redhat.com>
302 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
303 expanding to rmdir, since rmdir is not declared in a portable
304 system header we can include in gstdio.h. (#325249, Jani Monoses)
306 2006-01-01 Tor Lillqvist <tml@novell.com>
308 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
309 stdout and stderr to unbuffered. Otherwise the giochannel layer
310 will try to read from them regardless whether the
311 g_io_channel_win32_poll() call here has indicated
312 readability or not. (#325310)
314 2005-12-29 Matthias Clasen <mclasen@redhat.com>
316 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
317 in the docs. (#325273, Declan Naughton)
319 2005-12-27 Manish Singh <yosh@gimp.org>
321 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
322 make sure we really override things.
324 2005-12-27 Matthias Clasen <mclasen@redhat.com>
326 Fix #316221, Michal Benes, Stanislav Brabec;
328 * configure.in: Fix a strict aliasing problem in
329 g_static_mutex_get_mutex().
330 * glib/gthread.h: ...and in
331 g_static_mutex_get_mutex_impl_shortcut().
333 * glib/gdatasetprivate.h: Add a cast to silence compiler
334 warnings. (#321978, Andrew Paprocki)
336 Partial fix for bug #323937, Bogdan Nicula.
338 * configure.in: Check for malloc.h
339 * glib/gslice.c: Don't include malloc.h unconditionally.
341 2005-12-27 Manish Singh <yosh@gimp.org>
343 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
344 (#324950, Dan Yefimov)
346 2005-12-26 Matthias Clasen <mclasen@redhat.com>
348 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
351 2005-12-25 Matthias Clasen <mclasen@redhat.com>
353 * AUTHORS: Update my email
355 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
357 2005-12-21 Matthias Clasen <mclasen@redhat.com>
359 * README.in: Add some notes on when 'make check' may fail.
361 2005-12-20 Matthias Clasen <mclasen@redhat.com>
364 * glib/gthreadpool.h:
365 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
366 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
368 * tests/threadpool-test.c: Test this.
370 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
372 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
373 handle errors without depending on gmessage.c which might not be
374 setup when the error occours.
375 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
376 always freeing can be achieved by adjusting the working set time to
377 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
378 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
379 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
380 minimize wastage if small amounts of differently sized structrues are
381 allocated, this does come at a performance cost of roughly 5% though.
382 fixed up block alignment calculation, so it works for varying
383 block sizes. only use strerror() not g_strerror() since the latter
384 depends on working GQuark and GSlice.
385 mem_error(): implemented in terms of fprintf and vfprintf.
387 * tests/slice-color.c: new program to test cache colorization effects.
389 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
390 G_SLICE_CONFIG_WORKING_SET_MSECS.
392 2005-12-17 Matthias Clasen <mclasen@redhat.com>
394 * glib/goption.c (parse_short_option): Set an error in all
395 failure cases. (#324332, Tim-Philipp Müller)
397 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
399 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
400 and g_atomic_pointer_get. Add g_atomic_int_set and
401 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
402 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
403 as functions (additionally to the macros in the header) for the
404 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
406 2005-12-16 Matthias Clasen <mclasen@redhat.com>
408 * glib/gmem.c (g_allocator_new): Don't return a pointer to
409 a const struct, since apps expect to be able to modify it.
410 (#324179, J. Ali Harlow)
412 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
414 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
415 for systems where the initialization of atomic variables requires a
416 write memory barrier.
418 2005-12-09 Matthias Clasen <mclasen@redhat.com>
420 * configure.in: Bump version
422 * === Released 2.9.1 ===
424 2005-12-09 Alexander Larsson <alexl@redhat.com>
428 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
431 2005-12-08 Matthias Clasen <mclasen@redhat.com>
435 2005-12-07 Martyn Russell <martyn@imendio.com>
437 * glib/gasyncqueue.c:
438 - Call g_queue_insert_sorted() instead of duplicating the code.
439 - Call g_queue_sort() instead of duplicating the code.
440 - Invert sort function results to make sure the same sort function
441 gives the same results across glist, gslist, gqueue and
444 * tests/asyncqueue-test.c:
445 - Updated the sort function to reflect the example in the
446 documentation for gasyncqueue.c.
448 2005-12-07 Martyn Russell <martyn@imendio.com>
450 * docs/reference/glib/glib-sections.txt:
451 * docs/reference/glib/tmpl/linked_lists_double.sgml:
452 * docs/reference/glib/tmpl/linked_lists_single.sgml:
455 - Added g_list_insert_sorted_with_data () and
456 g_slist_insert_sorted_with_data ().
457 - Removed the extra check in g_list_sort() and g_slist_sort() for
458 GCompareDataFunc vs. GCompareFunc.
460 2005-12-07 Tor Lillqvist <tml@novell.com>
462 * glib-gettextize.in: Look up prefix at run-time on Win32,
463 assuming the standard directory structure with glib-gettextize in
466 2005-12-06 Matthias Clasen <mclasen@redhat.com>
468 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
469 we need GTK+ 2.8 to compile against GLib 2.10.
471 2005-12-05 Manish Singh <yosh@gimp.org>
473 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
474 to override any user settings.
476 2005-12-05 Matthias Clasen <mclasen@redhat.com>
478 * configure.in: Revert an accidential commit
480 2005-12-05 Matthias Clasen <mclasen@redhat.com>
482 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
484 * tests/Makefile.am (TESTS_ENVIRONMENT):
485 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
486 MALLOC_CHECK_ and MALLOC_PERTURB_
488 * tests/run-collate-tests.sh: Run the collation tests explicitly
489 in en_US locale. (#320463)
491 * glib/gmem.h: Really deprecate GMemChunk.
498 * glib/gstrfuncs.c: Add versioned deprecation docs.
500 2005-12-05 Martyn Russell <martyn@imendio.com>
502 * docs/reference/glib/glib-sections.txt:
503 * glib/gasyncqueue.[ch]:
504 - Added support for sorting async queues by with _push_sorted(),
505 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
508 * tests/asyncqueue-test.c:
509 - Added test case for gasyncqueue.c
511 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
513 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
514 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
515 in bug 323178. moved time consuming logic from g_slice_free() out of
516 the inner loop, so g_slice_free_chain_with_offset() provides a real
517 performance benefit over g_slice_free1() now.
519 * glib/gslice.h: renamed g_slice_free_chain() to
520 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
521 a type-safe macro as suggested in bug 323178.
522 simplified the macro implementation of g_slice_free() and implemented
523 it in a type safe manner for all compliers as suggested by Morten
524 Welinder <mortenw@gnome.org>.
529 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
531 2005-12-05 Matthias Clasen <mclasen@redhat.com>
533 * glib/gasyncqueue.c: Add some docs.
535 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
537 * glib/gatomic.c: In the ia64 implementation, use
538 __sync builtin without _si or _di suffix. (#321229,
539 Stanislav Brabec, patch by Andreas Schwab)
541 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
543 * glib/gslice.h: Remove comma at the end of enum.
545 2005-12-04 Matthias Clasen <mclasen@redhat.com>
547 Handle multiple user names with the same UID better.
548 (#319535, Laszlo Peter)
550 * glib/gutils.c (g_get_any_init_do): When determining user
551 data, first look up $LOGNAME. If the UID doesn't match
552 getuid(), fall back to the current behaviour of looking
553 up the user data based on getuid().
555 2005-12-04 Matthias Clasen <mclasen@redhat.com>
557 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
558 to make gcc warn if a function result is ignored. (#145466,
559 Arjan van de Ven, Alex Larsson)
561 * glib/gmem.h: Add the new attribute to g_realloc and
564 2005-12-03 Matthias Clasen <mclasen@redhat.com>
567 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
570 * glib/gmappedfile.c: Make mapping of empty files
573 * glib/gfileutils.c: Don't fork a new process just to
574 fix the permissions of the created temp file. (#321318,
575 Alexis S. L. Carvalho)
577 2005-12-02 Matthias Clasen <mclasen@redhat.com>
579 * README.in: Add a note about Solaris threads.
582 * configure.in: Remove support for Solaris threads.
583 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
585 2005-12-02 Matthias Clasen <mclasen@redhat.com>
587 * configure.in: Define G_GUINT64_CONSTANT in analogy to
588 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
590 2005-12-02 Matthias Clasen <mclasen@redhat.com>
592 * glib/gslice.c: Win32 portability fixes and C99ism removal,
593 pointed out by Kazuki Iwamoto. (#323052)
595 2005-12-02 Matthias Clasen <mclasen@redhat.com>
597 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
598 the IA64 implementation.
600 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
602 * glib/gslice.c: some naming and type size fixups.
604 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
606 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
607 variant into something less verbose, but digestible for gcc-3.4.
609 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
611 * tests/slice-test.c: extended to perform the benchmarking on the old
612 memchunk code if 'O' is selected.
614 * tests/memchunks.c: new file which contains the old GLib mem chunks
615 implementation with prefix old_mem_chunk_.
617 * tests/Makefile.am: added memchunks.c
619 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
621 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
622 acceptable load distribution of the syscall.
623 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
624 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
626 2005-12-01 Matthias Clasen <mclasen@redhat.com>
628 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
629 double-free base interface members.
631 * tests/Makefile.am: Remove duplicate variable.
633 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
635 * glib/gslice.[hc]: new slice allocator implementation.
637 * tests/slice-test.c: added random slice allocation test.
639 * glib/gthread.[hc]: removed newly added private thread mem API.
641 * glib/gthreadinit.h:
644 * glib/gmem.c: divided glib threading initialisation into three phases,
645 initialisation where private keys and messaging are not available (only
646 needed by gmem.c), initialisation without messaging but private keys
647 available (gslice.c, gmessage.c), and full fledged initialisers that
648 server the rest of glib. initialisation functions got renamed to reflect
649 the limitations of their corresponding phases.
651 * glib/gmem.c: removed memchunk code, defer allocations to
654 * glib/gmem.[hc]: removed g_slice_* skeletons.
656 * glib/glib.symbols: added g_slice_* symbols.
658 * configure.in: check for availability of posix_memalign(3), memalign(3)
661 * glib/Makefile.am: added gslice.[hc].
663 2005-12-01 Tor Lillqvist <tml@novell.com>
665 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
666 trailing slash(es) for non-root folders. stat() fails if non-root
667 folders are specified with trailing slashes. It's too much hassle
668 to demand that callers strip such slashes themselves, especially
669 as it is easy to get it wrong and strip the slash of a root
671 (g_rename): On NT-based Windows, use MoveFileEx() with
672 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
674 2005-11-28 Matthias Clasen <mclasen@redhat.com>
676 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
679 * configure.in: Check whether do { } while (0) works.
681 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
682 G_STMT_END if it works.
684 2005-11-28 Matthias Clasen <mclasen@redhat.com>
686 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
687 * glib/gnode.c (g_node_depth_traverse_level):
688 * glib/gmem.c (g_allocator_new):
689 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
690 (g_hash_table_foreach_remove_or_steal): Silence compiler
693 2005-11-27 Matthias Clasen <mclasen@redhat.com>
695 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
696 Don't read beyond len. (#322520, Christian Persch)
698 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
700 * glib/gdataset.c: access datalist flags via atomic pointer access
701 functions, instead of acquiring the dataset lock. this is faster and
702 also matches the atomic pointer readouts in gdatalistprivate.h and
703 direct pointer modifications required by gobject.c.
705 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
707 * glib/gdataset.c: streamlined initialization code somewhat,
708 removed GData node cache.
710 2005-11-22 Matthias Clasen <mclasen@redhat.com>
712 * glib/ghash.c (g_hash_table_ref):
713 (g_hash_table_unref): Mark these as new API.
715 Avoid double locking in g_intern_string (#322133,
718 * glib/gdataset.c (g_quark_from_string_internal): New
719 internal function which factors out common parts of
720 g_quark_from[_static]_string.
721 (g_quark_from_string, g_quark_from_static_string):
722 Use g_quark_from_string_internal.
723 (g_intern_string, g_intern_static_string): Use
724 g_quark_from_string_internal, and only take the
726 (g_quark_new): Don't store the strings shifted by -1
727 in the g_quarks array.
728 (g_quark_to_string): Adapt to the previous change.
730 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
734 g_hash_table_new_full(): create hash tables with a ref count of 1.
735 g_hash_table_ref(): atomically ref_count+=1
736 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
737 when refcount reaches 0.
738 g_hash_table_destroy(): just destroy keys and values, unref by 1.
739 g_hash_table_insert():
740 g_hash_table_replace(): assert ref_count>0.
743 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
746 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
748 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
749 variables. (#321984, Andrew Paprocki)
751 2005-11-18 Matthias Clasen <mclasen@redhat.com>
753 * configure.in: Bump version
755 2005-11-17 Matthias Clasen <mclasen@redhat.com>
757 * === Released 2.9.0 ===
759 * glib/glib.symbols: Updates
761 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
762 hints to the docs. (Paolo Borelli)
766 * glib/gthread.c (g_thread_foreach): Mark as new api.
768 * README.in: Updates.
772 * glib/gdate.c (g_date_set_time_t):
773 (g_date_set_time_val): Add functions to set a date from
774 a time_t and from a GTimeVal.
775 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
778 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
779 and g_utf8_offset_to_pointer.
781 * glib/gutf8.c (g_utf8_pointer_to_offset)
782 (g_utf8_offset_to_pointer): Handle negative offsets, and use
783 "stutter stepping" for going backwards. (#320638, Larry
788 * glib/gthread.c: const correctness fixes, found
789 by Arjan van de Ven and gcc.
791 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
793 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
794 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
796 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
798 * glib/guniprop.c: Use bit hacks instead when checking a general
799 category value against multiple values.
801 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
803 2005-11-10 Simos Xenitellis <simos@gnome.org>
805 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
807 2005-11-05 Tor Lillqvist <tml@novell.com>
809 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
810 (#320688) Instead set error and return failure.
812 2005-11-07 Matthias Clasen <mclasen@redhat.com>
814 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
815 while holding the context lock. (#320886, Andy Wingo)
817 2005-11-07 Billy Biggs <vektor@dumbterm.net>
819 * tests/refcount/Makefile.am: Add a missing reference to libglib
820 in the LDADD for this test.
822 2005-11-07 Matthias Clasen <mclasen@redhat.com>
824 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
825 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
826 GLib 2.9. (#320755, Luca Ferretti)
828 2005-11-04 Matthias Clasen <mclasen@redhat.com>
830 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
831 use a faster lookup table for g_unichar_get_mirror_char().
832 * glib/gmirroringtable.h: The new table.
833 * glib/gunichartables.h: Remove bidi_mirroring_table.
834 * glib/gen-unicode-tables.pl: Don't generate the mirroring
836 * glib/glib-mirroring-tab/*: Sources for the program
837 which generated gmirroringtable.h.
839 * glib/glist.c: Avoid some code duplication.
841 * glib/gscanner.h: Include gdataset.h. (#320322)
843 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
844 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
845 to help find critical warnings in applications. (#320017,
848 2005-11-02 Tor Lillqvist <tml@novell.com>
850 * glib/glib.symbols: Remove large amount of trailing whitespace
851 from one line. Remove #ifdef G_UNDEFINED from around
852 g_hash_table_new and g_hash_table_new_full symbols, what was the
853 purpose of that? Add the g_slice functions.
855 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
857 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
858 added g_slice_*() API to allocate and cache small bits of memory.
859 an actuall allocator implementation for g_slice_*() is still pending.
861 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
862 changed GRealThread list to use in-structure *next; fields instead
863 of GSList, in order for thread iteration to not depenend on g_slice_*()
865 _g_thread_mem_private_get():
866 _g_thread_mem_private_set(): added accessors for private memory,
867 needed because the ordinary GPrivate implementation relies on GArray
868 and GSList and therefore indirectly on working g_slice_*() allocations.
871 g_thread_foreach(): new public API function to loop over all existing threads.
885 removed GAllocator and free list usages and accompanying locks.
886 use g_slice_*() API to allocate and cache small bits of memory.
888 * glib/ghook.h: removed GMemChunk field from public API.
891 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
893 * glib/gnode.h: deprecate allocator API.
895 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
896 be used for poll rec lists.
898 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
899 g_relation_destroy(): free all tuples from the all_tuples hash table,
900 this effectively maintains the life time track keeping of tuples.
901 g_relation_delete_tuple(): free tuples which are removed from the
902 all_tuples hash table. this fixes a temporary leak that was present
903 in the memchunk code until the destruction of the relation.
905 2005-10-29 Matthias Clasen <mclasen@redhat.com>
907 * tests/convert-test.c: Add some tests for conversions between
908 UTF-8, UCS-4 and UTF-16.
910 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
911 of len == -1, noticed by Morten Welinder.
913 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
915 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
917 2005-10-26 Matthias Clasen <mclasen@redhat.com>
919 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
920 in the error case, and add some documentation. (#319806, Morten
923 2005-10-19 Manish Singh <yosh@gimp.org>
925 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
926 config.h symbol automatically. Fixes bug #313731.
928 2005-10-19 Tor Lillqvist <tml@novell.com>
930 * glib/gwin32.c (g_win32_get_package_installation_directory):
931 Return a g_strdup()ed copy of the value stored in the hash table,
932 so that it can be g_free()d without leaving a dangling pointer in
933 the hash table. (#319232)
935 2005-10-06 Matthias Clasen <mclasen@redhat.com>
937 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
938 all-zero sequences correctly. (#317930, Sebastien Bacher)
940 2005-10-05 Matthias Clasen <mclasen@redhat.com>
942 * glib/gmarkup.c (g_markup_escape_text): Doc addition
944 * Makefile.am: Add Collation tests.
946 * tests/collate/*: Inputs and expected outputs for collation tests.
948 * tests/run-collate-tests.sh: Script to run collation tests.
950 * tests/unicode-collate.c (main): Rework slightly to make
951 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
953 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
955 * docs/reference/glib/tmpl/unicode.sgml:
956 * glib/gen-unicode-tables.pl:
958 * glib/gunichartables.h:
960 * tests/casefold.txt:
961 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
962 GUnicodeBreakType types. That may break some applications, like
965 2005-09-26 Matthias Clasen <mclasen@redhat.com>
967 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
968 docs. (#314393, Matthew F. Barnes)
971 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
972 they are already declared in glib.h. This doesn't break documented
973 use of gprintf.h, but should probably be pointed out in the
974 release notes for 2.10. (#314232, Behdad Esfahbod)
976 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
978 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
979 complexity optimization, based on a patch from Matthias Clasen.
981 * tests/patterntest.c: more tests, mostly from matthias.
983 2005-09-20 Matthias Clasen <mclasen@redhat.com>
985 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
986 (#316703, Mark Drago)
988 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
990 * glib/gpattern.c: applied a patch from matthias which checks on the
991 upper bound of GPatternSpec length to optimize matches.
994 * tests/patterntest.c: added more match cases.
996 2005-09-16 Tor Lillqvist <tml@novell.com>
998 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
999 warning to doc comment that these functions should not be used on
1000 encodings like CP932.
1002 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1004 * tests/keyfile-test.c: Add a test for grup names of length 1.
1006 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1007 of length 1. (#316309)
1009 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1011 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1013 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1015 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1016 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1019 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1021 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1022 (child_watch_helper_thread): Remove some dead code. Closes
1025 2005-09-07 Tor Lillqvist <tml@novell.com>
1027 * glib/Makefile.am: Create also a console version of the
1028 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1029 It's otherwise identical to gspawn-win32-helper.exe, except marked
1030 as a console application (linked without the -mwindows option).
1032 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1033 Drop the dont_wait parameter. Its truth value correlated 100% with
1034 the NULLness of the exit_status parameter anyway, so it's enough
1035 to check whether exit_status is NULL. Invert the sense of the
1036 dont_return_handle parameter and rename it to do_return_handle, to
1037 make the code easier to read by avoiding double negations.
1039 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1040 to do_spawn_with_pipes() accordingly.
1042 (do_spawn_with_pipes): If we have a console, use the console
1043 version of the helper program, otherwise use the GUI one. This
1044 avoids extra console windows opening up in some situations. (In
1045 case a console application uses the GUI gspawn-win32-helper.exe to
1046 spawn another console application we would get a separate console
1047 for the spawned console application).
1049 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1051 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1053 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1054 if the file is too large. (#315275, Kjartan Maraas)
1056 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1057 of read() is signed. (#315273, Kjartan Maraas)
1059 2005-08-31 Tor Lillqvist <tml@novell.com>
1061 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1062 the EXE and each DLL have their own atexit function chains.
1064 #define g_atexit as atexit instead. This means it has a
1065 better chance of doing what the caller wants. For instance,
1066 gtkhtml calls g_atexit() registering a function in gtkhtml
1067 itself. This caused a crash when g_atexit() was implemented as a
1068 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1069 the time the GLib DLL got unloaded.
1071 * glib/gutils.c: #undef the #define mentioned above, to also get a
1072 real g_atexit() into the DLL for backward compatibility. Document
1073 the Windows behaviour of g_atexit(), and document the varying ways
1074 atexit() can behave in the context of dynamically loaded modules
1077 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1079 * glib/glib.symbols:
1081 * glib/gdataset.c: Add string interning functions.
1083 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1085 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1088 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1090 * configure.in: Bump version to 2.9.0
1092 2005-08-25 Tor Lillqvist <tml@novell.com>
1094 Make also the g_spawn*() functions take parameters in the GLib
1095 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1096 API or ABI. Like the other GLib API that was earlier changed to
1097 use UTF-8 on Windows, the names of the functions that take UTF-8
1098 have _utf8 suffixes added by using preprocessor macros in the
1099 header file. The old names are kept for functions with the old
1100 behaviour, taking parameters in the system codepage, for DLL ABI
1103 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1104 g_spawn*() functions.
1106 * glib/gspawn-win32.c: Use wide-char API on NT-based
1107 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1108 system codepage (Win9x) and call the C library _wspawn*() or
1109 spawn*() functions respectvely. Add DLL ABI stability versions
1110 that take parameters in the system codepage.
1112 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1113 wide-char versions of argv and envp, and use wide-char API to
1114 change directory and spawn the program to run. Remove the verbose
1115 debugging output, it was too complex to modify for the wide-char
1116 features. (Just add temporary debugging printouts if needed, no
1117 need to have them permanently in the source.)
1119 * glib/gspawn.c: Corresponding documentation updates.
1121 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1122 symbols as PRIVATE, add the new _utf8-suffixed ones.
1124 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1126 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1127 guint64 otherwise and that can produce warnings about comparison
1128 between signed and unsigned.
1130 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1132 * glib/gutils.c: Fix the crt_externs.h include.
1134 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1136 * NEWS: Fix spelling of my first name.
1138 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1142 * === Released 2.8.1 ===
1146 2005-08-20 Hans Breuer <hans@breuer.org>
1148 * glib/makefile.msc.in : link with ws2_32.lib
1150 2005-08-18 Tor Lillqvist <tml@novell.com>
1152 * configure.in: Check for <sys/wait.h>
1154 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1156 2005-08-18 Ross Burton <ross@burtonini.com>
1159 Optimise single-character insertions.
1164 * tests/string-test.c:
1165 Add tests for new optimisation, and fix a leak.
1167 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1169 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1171 * glib/gutils.c: On Darwin, include crt-externs.h and
1172 define environ using _NSGetEnviron(). (#313731)
1174 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1176 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1177 (g_get_any_init_do): ... to this new function.
1178 (g_get_any_init): Declare as inline.
1179 (g_get_any_init_locked): New inline function, does the locking.
1180 Make use of these two throughout the code.
1182 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1184 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1185 the child process and then simply return. This makes
1186 The "S" option work as documented in g_on_error_query().
1187 (#313125, Matthew F. Barnes)
1189 * glib/gunicode.h: Update the link to Unicode category
1190 values. (#313369, Behnam Esfahbod)
1192 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1193 a little. (#311727, Tristan van Berkom)
1195 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1196 check work on ia64 too, where some symbols ended up
1197 in yet another section.