1 2006-02-19 Martyn Russell <martyn@imendio.com>
3 * tests/asyncqueue-test.c:
5 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
6 _insert_sorted and _insert_sorted_with_data API.
8 2006-02-18 Matthias Clasen <mclasen@redhat.com>
10 * tests/gobject/Makefile.am: Add paramspec-test
12 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
14 * tests/gobject/gvalue-test.c: Add more tests.
16 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
18 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
20 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
22 * glib/gthreadpool.c: Fix deadlock when signalling the thread
23 which freed a thread pool (#331110, Chris Wilson).
25 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
27 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
28 posix_memalign() in case we actually are going to use it, because
29 we detected a compliant implementation (#328997).
31 * configure.in (enable_included_printf): don't include malloc.h when
32 testing for posix_memalign() funcitonality, since this may break the
33 test on some systems (#328997).
35 2006-02-14 Matthias Clasen <mclasen@redhat.com>
37 * glib/gstrfuncs.c: Improve docs.
39 2006-02-11 Matthias Clasen <mclasen@redhat.com>
41 * configure.in: Bump version
43 2006-02-10 Matthias Clasen <mclasen@redhat.com>
45 * === Released 2.9.6 ===
49 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
51 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
52 the type of varargs arguments. (bug #317679)
54 2006-01-31 Matthias Clasen <mclasen@redhat.com>
56 * autogen.sh: Touch README and INSTALL here to pacify
57 automake. (#329124, Kjartan Maraas, Tim Janik)
59 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
61 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
62 some missing cases where pointed out by Sebastian Wilhelmi.
64 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
66 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
68 2006-01-30 David Schleef <ds@schleef.org>
70 * glib/grand.c: Update URL
72 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
74 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
75 create container classes which are as flexible as a GValue is
77 2006-01-27 Matthias Clasen <mclasen@redhat.com>
79 * configure.in: Bump version
81 * === Released 2.9.5 ===
83 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
88 2006-01-26 Matthias Clasen <mclasen@redhat.com>
90 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
93 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
95 * fixed buglets reported by Jens Granseuer in #328254.
97 * configure.in: free the memory allocated in posix_memalign() tests.
99 * glib/gslice.c: spelling fixes.
101 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
103 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
104 slices, make sure g_mem_gc_friendly is properly initialized.
106 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
107 the first allocation. applied some branching optimizations.
109 * docs/macros.txt: reflected --enable-gc-friendly change and
110 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
112 * configure.in: changed --enable-gc-friendly=yes to define
113 ENABLE_GC_FRIENDLY_DEFAULT.
115 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
116 if (G_UNLIKELY (g_mem_gc_friendly)).
119 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
120 care of by g_slice_free1().
122 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
124 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
126 * glib/gslice.c: only use posix_memalign() if it's known to work,
127 revert to memalign() otherwise.
129 * configure.in: check for broken posix_memalign() implementations
132 2006-01-24 Matthias Clasen <mclasen@redhat.com>
134 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
135 on all little-endian systems. (#143380, Marc Moorcroft)
137 2006-01-23 Matthias Clasen <mclasen@redhat.com>
139 * configure.in: Change the shared libary extension for hpux-ia64
140 to so. (#328253, The Written Word)
142 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
144 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
145 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
148 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
150 * glib/gslice.c (slice_config_init): initialize GSlice config from
151 G_SLICE environment variable. we support G_SLICE=always-malloc
152 currently, which forces all g_slice_*() allocations to use the system
156 g_parse_debug_string(): added a note about not using g_malloc() here.
157 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
158 g_slice. contains only guesswork in the WIN32 branch.
160 2006-01-18 Matthias Clasen <mclasen@redhat.com>
164 * === Released 2.9.4 ===
168 2006-01-17 Matthias Clasen <mclasen@redhat.com>
170 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
171 call. (#327421, Morten Welinder)
173 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
175 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
176 G_LOCK. Use the unused_thread_queue lock instead. Change
177 g_thread_pool_thread_proxy such that threads only wait on
178 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
179 due to superfluous tasks. Global tasks wait at most for
180 max-idle-time milliseconds. Make sure, that no task is woken up
181 twice for the same event via a wakeup_serial. This fixes #324228.
183 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
184 invalid NULL into the thread pools. This as well fixes #327290.
186 2006-01-16 Matthias Clasen <mclasen@redhat.com>
188 * configure.in: Bump version
190 * === Released 2.9.3 ===
194 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
195 Rework the strict aliasing fix to not break C++, pointed
196 out by Murray Cumming.
198 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
199 Signal waiting threads, problem noticed by Christian Kellner.
201 2006-01-16 Matthias Clasen <mclasen@redhat.com>
203 Fix bug 326747, Alberto Ruiz:
205 * glib/gconvert.c (make_valid_utf8): Change this function to
206 replace unknown characters by the Unicode replacement character
207 instead of '?', and don't append "(invalid encoding)".
208 (g_filename_display_name, g_filename_display_basename): Document
209 how to determine if the filename was in an invalid encoding.
211 2006-01-14 Matthias Clasen <mclasen@redhat.com>
213 * glib/gtree.c: Replace the simple recursive implementation
214 by a nonrecursive, threaded implementation by Maurizio
217 2006-01-12 Matthias Clasen <mclasen@redhat.com>
219 * tests/asyncqueue-test.c (main): Fix the
220 assert to not always trigger. (#326558,
223 2006-01-11 Matthias Clasen <mclasen@redhat.com>
225 * tests/convert-test.c: Don't test items_read and
226 items_written in error cases where it is not set.
228 * tests/asyncqueue-test.c: Set error to NULL.
230 2006-01-08 Matthias Clasen <mclasen@redhat.com>
232 * glib/gtree.c: Break some long lines.
234 * tests/tree-test.c: Add more tests.
236 * glib/gslice.c: Include config.h first, pointed out
239 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
241 * docs/reference/gobject/tmpl/gtype.sgml: explain that
242 G_TYPE_INSTANCE_GET_CLASS() does behave different during
245 2006-01-06 Matthias Clasen <mclasen@redhat.com>
247 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
248 that we have posix_memalign(). (#323937, Bogdan Nicula)
250 2006-01-05 Matthias Clasen <mclasen@redhat.com>
252 * configure.in: Bump version
254 * === Released 2.9.2 ===
256 2006-01-05 Hans Breuer <hans@breuer.org>
258 * glib/makefile.msc.in : added gslice.obj
260 2006-01-05 Matthias Clasen <mclasen@redhat.com>
264 * glib/gmain.c: Clarify the documentation of source ids
265 a bit. (#325874, Dan Williams)
267 * configure.in: Fix another strict aliasing problem.
269 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
272 2006-01-04 Manish Singh <yosh@gimp.org>
275 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
276 either, since they have the same issue as g_rmdir. (Related to
279 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
281 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
282 order when creating magazines, so we hand out chunks with ascending
285 2006-01-04 Abel Cheung <maddog@linuxhall.org>
287 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
289 2006-01-03 Martyn Russell <martyn@imendio.com>
291 * docs/reference/glib/glib-sections.txt:
293 * glib/gthreadpool.[ch]:
294 - Added new API g_thread_pool_get_idle_time() and
295 g_thread_pool_set_idle_time(). (#324228).
297 * tests/threadpool-test.c:
298 - Updated test case to do thread pool sorting, thread pool with
299 no sorting and a thread pool with idle thread timeouts.
301 2006-01-03 Matthias Clasen <mclasen@redhat.com>
303 * glib/gmain.h: Add new functions here, too.
305 * glib/glib.symbols: Add new functions.
307 2005-12-20 Michael Meeks <michael.meeks@novell.com>
309 * glib/gmain.c (g_main_context_is_owner): new method
310 to determine if the current thread is the owner of the
313 2006-01-02 Matthias Clasen <mclasen@redhat.com>
317 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
318 expanding to rmdir, since rmdir is not declared in a portable
319 system header we can include in gstdio.h. (#325249, Jani Monoses)
321 2006-01-01 Tor Lillqvist <tml@novell.com>
323 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
324 stdout and stderr to unbuffered. Otherwise the giochannel layer
325 will try to read from them regardless whether the
326 g_io_channel_win32_poll() call here has indicated
327 readability or not. (#325310)
329 2005-12-29 Matthias Clasen <mclasen@redhat.com>
331 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
332 in the docs. (#325273, Declan Naughton)
334 2005-12-27 Manish Singh <yosh@gimp.org>
336 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
337 make sure we really override things.
339 2005-12-27 Matthias Clasen <mclasen@redhat.com>
341 Fix #316221, Michal Benes, Stanislav Brabec;
343 * configure.in: Fix a strict aliasing problem in
344 g_static_mutex_get_mutex().
345 * glib/gthread.h: ...and in
346 g_static_mutex_get_mutex_impl_shortcut().
348 * glib/gdatasetprivate.h: Add a cast to silence compiler
349 warnings. (#321978, Andrew Paprocki)
351 Partial fix for bug #323937, Bogdan Nicula.
353 * configure.in: Check for malloc.h
354 * glib/gslice.c: Don't include malloc.h unconditionally.
356 2005-12-27 Manish Singh <yosh@gimp.org>
358 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
359 (#324950, Dan Yefimov)
361 2005-12-26 Matthias Clasen <mclasen@redhat.com>
363 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
366 2005-12-25 Matthias Clasen <mclasen@redhat.com>
368 * AUTHORS: Update my email
370 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
372 2005-12-21 Matthias Clasen <mclasen@redhat.com>
374 * README.in: Add some notes on when 'make check' may fail.
376 2005-12-20 Matthias Clasen <mclasen@redhat.com>
379 * glib/gthreadpool.h:
380 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
381 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
383 * tests/threadpool-test.c: Test this.
385 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
387 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
388 handle errors without depending on gmessage.c which might not be
389 setup when the error occours.
390 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
391 always freeing can be achieved by adjusting the working set time to
392 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
393 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
394 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
395 minimize wastage if small amounts of differently sized structrues are
396 allocated, this does come at a performance cost of roughly 5% though.
397 fixed up block alignment calculation, so it works for varying
398 block sizes. only use strerror() not g_strerror() since the latter
399 depends on working GQuark and GSlice.
400 mem_error(): implemented in terms of fprintf and vfprintf.
402 * tests/slice-color.c: new program to test cache colorization effects.
404 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
405 G_SLICE_CONFIG_WORKING_SET_MSECS.
407 2005-12-17 Matthias Clasen <mclasen@redhat.com>
409 * glib/goption.c (parse_short_option): Set an error in all
410 failure cases. (#324332, Tim-Philipp Müller)
412 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
414 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
415 and g_atomic_pointer_get. Add g_atomic_int_set and
416 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
417 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
418 as functions (additionally to the macros in the header) for the
419 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
421 2005-12-16 Matthias Clasen <mclasen@redhat.com>
423 * glib/gmem.c (g_allocator_new): Don't return a pointer to
424 a const struct, since apps expect to be able to modify it.
425 (#324179, J. Ali Harlow)
427 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
429 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
430 for systems where the initialization of atomic variables requires a
431 write memory barrier.
433 2005-12-09 Matthias Clasen <mclasen@redhat.com>
435 * configure.in: Bump version
437 * === Released 2.9.1 ===
439 2005-12-09 Alexander Larsson <alexl@redhat.com>
443 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
446 2005-12-08 Matthias Clasen <mclasen@redhat.com>
450 2005-12-07 Martyn Russell <martyn@imendio.com>
452 * glib/gasyncqueue.c:
453 - Call g_queue_insert_sorted() instead of duplicating the code.
454 - Call g_queue_sort() instead of duplicating the code.
455 - Invert sort function results to make sure the same sort function
456 gives the same results across glist, gslist, gqueue and
459 * tests/asyncqueue-test.c:
460 - Updated the sort function to reflect the example in the
461 documentation for gasyncqueue.c.
463 2005-12-07 Martyn Russell <martyn@imendio.com>
465 * docs/reference/glib/glib-sections.txt:
466 * docs/reference/glib/tmpl/linked_lists_double.sgml:
467 * docs/reference/glib/tmpl/linked_lists_single.sgml:
470 - Added g_list_insert_sorted_with_data () and
471 g_slist_insert_sorted_with_data ().
472 - Removed the extra check in g_list_sort() and g_slist_sort() for
473 GCompareDataFunc vs. GCompareFunc.
475 2005-12-07 Tor Lillqvist <tml@novell.com>
477 * glib-gettextize.in: Look up prefix at run-time on Win32,
478 assuming the standard directory structure with glib-gettextize in
481 2005-12-06 Matthias Clasen <mclasen@redhat.com>
483 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
484 we need GTK+ 2.8 to compile against GLib 2.10.
486 2005-12-05 Manish Singh <yosh@gimp.org>
488 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
489 to override any user settings.
491 2005-12-05 Matthias Clasen <mclasen@redhat.com>
493 * configure.in: Revert an accidential commit
495 2005-12-05 Matthias Clasen <mclasen@redhat.com>
497 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
499 * tests/Makefile.am (TESTS_ENVIRONMENT):
500 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
501 MALLOC_CHECK_ and MALLOC_PERTURB_
503 * tests/run-collate-tests.sh: Run the collation tests explicitly
504 in en_US locale. (#320463)
506 * glib/gmem.h: Really deprecate GMemChunk.
513 * glib/gstrfuncs.c: Add versioned deprecation docs.
515 2005-12-05 Martyn Russell <martyn@imendio.com>
517 * docs/reference/glib/glib-sections.txt:
518 * glib/gasyncqueue.[ch]:
519 - Added support for sorting async queues by with _push_sorted(),
520 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
523 * tests/asyncqueue-test.c:
524 - Added test case for gasyncqueue.c
526 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
528 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
529 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
530 in bug 323178. moved time consuming logic from g_slice_free() out of
531 the inner loop, so g_slice_free_chain_with_offset() provides a real
532 performance benefit over g_slice_free1() now.
534 * glib/gslice.h: renamed g_slice_free_chain() to
535 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
536 a type-safe macro as suggested in bug 323178.
537 simplified the macro implementation of g_slice_free() and implemented
538 it in a type safe manner for all compliers as suggested by Morten
539 Welinder <mortenw@gnome.org>.
544 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
546 2005-12-05 Matthias Clasen <mclasen@redhat.com>
548 * glib/gasyncqueue.c: Add some docs.
550 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
552 * glib/gatomic.c: In the ia64 implementation, use
553 __sync builtin without _si or _di suffix. (#321229,
554 Stanislav Brabec, patch by Andreas Schwab)
556 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
558 * glib/gslice.h: Remove comma at the end of enum.
560 2005-12-04 Matthias Clasen <mclasen@redhat.com>
562 Handle multiple user names with the same UID better.
563 (#319535, Laszlo Peter)
565 * glib/gutils.c (g_get_any_init_do): When determining user
566 data, first look up $LOGNAME. If the UID doesn't match
567 getuid(), fall back to the current behaviour of looking
568 up the user data based on getuid().
570 2005-12-04 Matthias Clasen <mclasen@redhat.com>
572 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
573 to make gcc warn if a function result is ignored. (#145466,
574 Arjan van de Ven, Alex Larsson)
576 * glib/gmem.h: Add the new attribute to g_realloc and
579 2005-12-03 Matthias Clasen <mclasen@redhat.com>
582 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
585 * glib/gmappedfile.c: Make mapping of empty files
588 * glib/gfileutils.c: Don't fork a new process just to
589 fix the permissions of the created temp file. (#321318,
590 Alexis S. L. Carvalho)
592 2005-12-02 Matthias Clasen <mclasen@redhat.com>
594 * README.in: Add a note about Solaris threads.
597 * configure.in: Remove support for Solaris threads.
598 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
600 2005-12-02 Matthias Clasen <mclasen@redhat.com>
602 * configure.in: Define G_GUINT64_CONSTANT in analogy to
603 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
605 2005-12-02 Matthias Clasen <mclasen@redhat.com>
607 * glib/gslice.c: Win32 portability fixes and C99ism removal,
608 pointed out by Kazuki Iwamoto. (#323052)
610 2005-12-02 Matthias Clasen <mclasen@redhat.com>
612 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
613 the IA64 implementation.
615 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
617 * glib/gslice.c: some naming and type size fixups.
619 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
621 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
622 variant into something less verbose, but digestible for gcc-3.4.
624 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
626 * tests/slice-test.c: extended to perform the benchmarking on the old
627 memchunk code if 'O' is selected.
629 * tests/memchunks.c: new file which contains the old GLib mem chunks
630 implementation with prefix old_mem_chunk_.
632 * tests/Makefile.am: added memchunks.c
634 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
636 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
637 acceptable load distribution of the syscall.
638 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
639 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
641 2005-12-01 Matthias Clasen <mclasen@redhat.com>
643 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
644 double-free base interface members.
646 * tests/Makefile.am: Remove duplicate variable.
648 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
650 * glib/gslice.[hc]: new slice allocator implementation.
652 * tests/slice-test.c: added random slice allocation test.
654 * glib/gthread.[hc]: removed newly added private thread mem API.
656 * glib/gthreadinit.h:
659 * glib/gmem.c: divided glib threading initialisation into three phases,
660 initialisation where private keys and messaging are not available (only
661 needed by gmem.c), initialisation without messaging but private keys
662 available (gslice.c, gmessage.c), and full fledged initialisers that
663 server the rest of glib. initialisation functions got renamed to reflect
664 the limitations of their corresponding phases.
666 * glib/gmem.c: removed memchunk code, defer allocations to
669 * glib/gmem.[hc]: removed g_slice_* skeletons.
671 * glib/glib.symbols: added g_slice_* symbols.
673 * configure.in: check for availability of posix_memalign(3), memalign(3)
676 * glib/Makefile.am: added gslice.[hc].
678 2005-12-01 Tor Lillqvist <tml@novell.com>
680 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
681 trailing slash(es) for non-root folders. stat() fails if non-root
682 folders are specified with trailing slashes. It's too much hassle
683 to demand that callers strip such slashes themselves, especially
684 as it is easy to get it wrong and strip the slash of a root
686 (g_rename): On NT-based Windows, use MoveFileEx() with
687 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
689 2005-11-28 Matthias Clasen <mclasen@redhat.com>
691 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
694 * configure.in: Check whether do { } while (0) works.
696 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
697 G_STMT_END if it works.
699 2005-11-28 Matthias Clasen <mclasen@redhat.com>
701 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
702 * glib/gnode.c (g_node_depth_traverse_level):
703 * glib/gmem.c (g_allocator_new):
704 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
705 (g_hash_table_foreach_remove_or_steal): Silence compiler
708 2005-11-27 Matthias Clasen <mclasen@redhat.com>
710 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
711 Don't read beyond len. (#322520, Christian Persch)
713 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
715 * glib/gdataset.c: access datalist flags via atomic pointer access
716 functions, instead of acquiring the dataset lock. this is faster and
717 also matches the atomic pointer readouts in gdatalistprivate.h and
718 direct pointer modifications required by gobject.c.
720 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
722 * glib/gdataset.c: streamlined initialization code somewhat,
723 removed GData node cache.
725 2005-11-22 Matthias Clasen <mclasen@redhat.com>
727 * glib/ghash.c (g_hash_table_ref):
728 (g_hash_table_unref): Mark these as new API.
730 Avoid double locking in g_intern_string (#322133,
733 * glib/gdataset.c (g_quark_from_string_internal): New
734 internal function which factors out common parts of
735 g_quark_from[_static]_string.
736 (g_quark_from_string, g_quark_from_static_string):
737 Use g_quark_from_string_internal.
738 (g_intern_string, g_intern_static_string): Use
739 g_quark_from_string_internal, and only take the
741 (g_quark_new): Don't store the strings shifted by -1
742 in the g_quarks array.
743 (g_quark_to_string): Adapt to the previous change.
745 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
749 g_hash_table_new_full(): create hash tables with a ref count of 1.
750 g_hash_table_ref(): atomically ref_count+=1
751 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
752 when refcount reaches 0.
753 g_hash_table_destroy(): just destroy keys and values, unref by 1.
754 g_hash_table_insert():
755 g_hash_table_replace(): assert ref_count>0.
758 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
761 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
763 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
764 variables. (#321984, Andrew Paprocki)
766 2005-11-18 Matthias Clasen <mclasen@redhat.com>
768 * configure.in: Bump version
770 2005-11-17 Matthias Clasen <mclasen@redhat.com>
772 * === Released 2.9.0 ===
774 * glib/glib.symbols: Updates
776 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
777 hints to the docs. (Paolo Borelli)
781 * glib/gthread.c (g_thread_foreach): Mark as new api.
783 * README.in: Updates.
787 * glib/gdate.c (g_date_set_time_t):
788 (g_date_set_time_val): Add functions to set a date from
789 a time_t and from a GTimeVal.
790 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
793 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
794 and g_utf8_offset_to_pointer.
796 * glib/gutf8.c (g_utf8_pointer_to_offset)
797 (g_utf8_offset_to_pointer): Handle negative offsets, and use
798 "stutter stepping" for going backwards. (#320638, Larry
803 * glib/gthread.c: const correctness fixes, found
804 by Arjan van de Ven and gcc.
806 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
808 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
809 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
811 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
813 * glib/guniprop.c: Use bit hacks instead when checking a general
814 category value against multiple values.
816 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
818 2005-11-10 Simos Xenitellis <simos@gnome.org>
820 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
822 2005-11-05 Tor Lillqvist <tml@novell.com>
824 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
825 (#320688) Instead set error and return failure.
827 2005-11-07 Matthias Clasen <mclasen@redhat.com>
829 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
830 while holding the context lock. (#320886, Andy Wingo)
832 2005-11-07 Billy Biggs <vektor@dumbterm.net>
834 * tests/refcount/Makefile.am: Add a missing reference to libglib
835 in the LDADD for this test.
837 2005-11-07 Matthias Clasen <mclasen@redhat.com>
839 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
840 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
841 GLib 2.9. (#320755, Luca Ferretti)
843 2005-11-04 Matthias Clasen <mclasen@redhat.com>
845 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
846 use a faster lookup table for g_unichar_get_mirror_char().
847 * glib/gmirroringtable.h: The new table.
848 * glib/gunichartables.h: Remove bidi_mirroring_table.
849 * glib/gen-unicode-tables.pl: Don't generate the mirroring
851 * glib/glib-mirroring-tab/*: Sources for the program
852 which generated gmirroringtable.h.
854 * glib/glist.c: Avoid some code duplication.
856 * glib/gscanner.h: Include gdataset.h. (#320322)
858 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
859 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
860 to help find critical warnings in applications. (#320017,
863 2005-11-02 Tor Lillqvist <tml@novell.com>
865 * glib/glib.symbols: Remove large amount of trailing whitespace
866 from one line. Remove #ifdef G_UNDEFINED from around
867 g_hash_table_new and g_hash_table_new_full symbols, what was the
868 purpose of that? Add the g_slice functions.
870 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
872 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
873 added g_slice_*() API to allocate and cache small bits of memory.
874 an actuall allocator implementation for g_slice_*() is still pending.
876 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
877 changed GRealThread list to use in-structure *next; fields instead
878 of GSList, in order for thread iteration to not depenend on g_slice_*()
880 _g_thread_mem_private_get():
881 _g_thread_mem_private_set(): added accessors for private memory,
882 needed because the ordinary GPrivate implementation relies on GArray
883 and GSList and therefore indirectly on working g_slice_*() allocations.
886 g_thread_foreach(): new public API function to loop over all existing threads.
900 removed GAllocator and free list usages and accompanying locks.
901 use g_slice_*() API to allocate and cache small bits of memory.
903 * glib/ghook.h: removed GMemChunk field from public API.
906 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
908 * glib/gnode.h: deprecate allocator API.
910 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
911 be used for poll rec lists.
913 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
914 g_relation_destroy(): free all tuples from the all_tuples hash table,
915 this effectively maintains the life time track keeping of tuples.
916 g_relation_delete_tuple(): free tuples which are removed from the
917 all_tuples hash table. this fixes a temporary leak that was present
918 in the memchunk code until the destruction of the relation.
920 2005-10-29 Matthias Clasen <mclasen@redhat.com>
922 * tests/convert-test.c: Add some tests for conversions between
923 UTF-8, UCS-4 and UTF-16.
925 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
926 of len == -1, noticed by Morten Welinder.
928 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
930 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
932 2005-10-26 Matthias Clasen <mclasen@redhat.com>
934 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
935 in the error case, and add some documentation. (#319806, Morten
938 2005-10-19 Manish Singh <yosh@gimp.org>
940 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
941 config.h symbol automatically. Fixes bug #313731.
943 2005-10-19 Tor Lillqvist <tml@novell.com>
945 * glib/gwin32.c (g_win32_get_package_installation_directory):
946 Return a g_strdup()ed copy of the value stored in the hash table,
947 so that it can be g_free()d without leaving a dangling pointer in
948 the hash table. (#319232)
950 2005-10-06 Matthias Clasen <mclasen@redhat.com>
952 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
953 all-zero sequences correctly. (#317930, Sebastien Bacher)
955 2005-10-05 Matthias Clasen <mclasen@redhat.com>
957 * glib/gmarkup.c (g_markup_escape_text): Doc addition
959 * Makefile.am: Add Collation tests.
961 * tests/collate/*: Inputs and expected outputs for collation tests.
963 * tests/run-collate-tests.sh: Script to run collation tests.
965 * tests/unicode-collate.c (main): Rework slightly to make
966 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
968 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
970 * docs/reference/glib/tmpl/unicode.sgml:
971 * glib/gen-unicode-tables.pl:
973 * glib/gunichartables.h:
975 * tests/casefold.txt:
976 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
977 GUnicodeBreakType types. That may break some applications, like
980 2005-09-26 Matthias Clasen <mclasen@redhat.com>
982 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
983 docs. (#314393, Matthew F. Barnes)
986 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
987 they are already declared in glib.h. This doesn't break documented
988 use of gprintf.h, but should probably be pointed out in the
989 release notes for 2.10. (#314232, Behdad Esfahbod)
991 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
993 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
994 complexity optimization, based on a patch from Matthias Clasen.
996 * tests/patterntest.c: more tests, mostly from matthias.
998 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1000 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1001 (#316703, Mark Drago)
1003 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1005 * glib/gpattern.c: applied a patch from matthias which checks on the
1006 upper bound of GPatternSpec length to optimize matches.
1009 * tests/patterntest.c: added more match cases.
1011 2005-09-16 Tor Lillqvist <tml@novell.com>
1013 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1014 warning to doc comment that these functions should not be used on
1015 encodings like CP932.
1017 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1019 * tests/keyfile-test.c: Add a test for grup names of length 1.
1021 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1022 of length 1. (#316309)
1024 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1026 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1028 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1030 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1031 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1034 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1036 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1037 (child_watch_helper_thread): Remove some dead code. Closes
1040 2005-09-07 Tor Lillqvist <tml@novell.com>
1042 * glib/Makefile.am: Create also a console version of the
1043 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1044 It's otherwise identical to gspawn-win32-helper.exe, except marked
1045 as a console application (linked without the -mwindows option).
1047 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1048 Drop the dont_wait parameter. Its truth value correlated 100% with
1049 the NULLness of the exit_status parameter anyway, so it's enough
1050 to check whether exit_status is NULL. Invert the sense of the
1051 dont_return_handle parameter and rename it to do_return_handle, to
1052 make the code easier to read by avoiding double negations.
1054 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1055 to do_spawn_with_pipes() accordingly.
1057 (do_spawn_with_pipes): If we have a console, use the console
1058 version of the helper program, otherwise use the GUI one. This
1059 avoids extra console windows opening up in some situations. (In
1060 case a console application uses the GUI gspawn-win32-helper.exe to
1061 spawn another console application we would get a separate console
1062 for the spawned console application).
1064 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1066 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1068 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1069 if the file is too large. (#315275, Kjartan Maraas)
1071 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1072 of read() is signed. (#315273, Kjartan Maraas)
1074 2005-08-31 Tor Lillqvist <tml@novell.com>
1076 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1077 the EXE and each DLL have their own atexit function chains.
1079 #define g_atexit as atexit instead. This means it has a
1080 better chance of doing what the caller wants. For instance,
1081 gtkhtml calls g_atexit() registering a function in gtkhtml
1082 itself. This caused a crash when g_atexit() was implemented as a
1083 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1084 the time the GLib DLL got unloaded.
1086 * glib/gutils.c: #undef the #define mentioned above, to also get a
1087 real g_atexit() into the DLL for backward compatibility. Document
1088 the Windows behaviour of g_atexit(), and document the varying ways
1089 atexit() can behave in the context of dynamically loaded modules
1092 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1094 * glib/glib.symbols:
1096 * glib/gdataset.c: Add string interning functions.
1098 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1100 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1103 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1105 * configure.in: Bump version to 2.9.0
1107 2005-08-25 Tor Lillqvist <tml@novell.com>
1109 Make also the g_spawn*() functions take parameters in the GLib
1110 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1111 API or ABI. Like the other GLib API that was earlier changed to
1112 use UTF-8 on Windows, the names of the functions that take UTF-8
1113 have _utf8 suffixes added by using preprocessor macros in the
1114 header file. The old names are kept for functions with the old
1115 behaviour, taking parameters in the system codepage, for DLL ABI
1118 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1119 g_spawn*() functions.
1121 * glib/gspawn-win32.c: Use wide-char API on NT-based
1122 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1123 system codepage (Win9x) and call the C library _wspawn*() or
1124 spawn*() functions respectvely. Add DLL ABI stability versions
1125 that take parameters in the system codepage.
1127 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1128 wide-char versions of argv and envp, and use wide-char API to
1129 change directory and spawn the program to run. Remove the verbose
1130 debugging output, it was too complex to modify for the wide-char
1131 features. (Just add temporary debugging printouts if needed, no
1132 need to have them permanently in the source.)
1134 * glib/gspawn.c: Corresponding documentation updates.
1136 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1137 symbols as PRIVATE, add the new _utf8-suffixed ones.
1139 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1141 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1142 guint64 otherwise and that can produce warnings about comparison
1143 between signed and unsigned.
1145 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1147 * glib/gutils.c: Fix the crt_externs.h include.
1149 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1151 * NEWS: Fix spelling of my first name.
1153 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1157 * === Released 2.8.1 ===
1161 2005-08-20 Hans Breuer <hans@breuer.org>
1163 * glib/makefile.msc.in : link with ws2_32.lib
1165 2005-08-18 Tor Lillqvist <tml@novell.com>
1167 * configure.in: Check for <sys/wait.h>
1169 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1171 2005-08-18 Ross Burton <ross@burtonini.com>
1174 Optimise single-character insertions.
1179 * tests/string-test.c:
1180 Add tests for new optimisation, and fix a leak.
1182 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1184 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1186 * glib/gutils.c: On Darwin, include crt-externs.h and
1187 define environ using _NSGetEnviron(). (#313731)
1189 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1191 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1192 (g_get_any_init_do): ... to this new function.
1193 (g_get_any_init): Declare as inline.
1194 (g_get_any_init_locked): New inline function, does the locking.
1195 Make use of these two throughout the code.
1197 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1199 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1200 the child process and then simply return. This makes
1201 The "S" option work as documented in g_on_error_query().
1202 (#313125, Matthew F. Barnes)
1204 * glib/gunicode.h: Update the link to Unicode category
1205 values. (#313369, Behnam Esfahbod)
1207 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1208 a little. (#311727, Tristan van Berkom)
1210 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1211 check work on ia64 too, where some symbols ended up
1212 in yet another section.