1 2006-03-07 Matthias Clasen <mclasen@redhat.com>
5 * glib/gutf8.c (_g_charset_get_aliases): Match the prototype
6 in gconvert.c to fix build problems in NetBSD. (#333651,
9 2006-03-06 Anders Carlsson <andersca@imendio.com>
12 Make .dylib the shared library suffix on darwin.
13 (#322476, Vladimir Panov)
15 2006-03-06 Anders Carlsson <andersca@imendio.com>
18 Skip checking thread flags on Darwin. Fix suggested by Manish Singh.
19 (#314794, Gregor Riepl)
21 2006-03-01 Tor Lillqvist <tml@novell.com>
23 * glib/gutils.c (g_listenv): Separate implementation on Win32: Use
24 the wide character API on NT-based Windows. Return UTF-8 strings.
26 * glib/glib.symbols: Don't mark g_listenv as PRIVATE, as that
27 meant it wasn't present in the import library. PRIVATE is used
28 only for the backwards-compatibility DLL ABI stability hacks.
30 2006-02-24 Matthias Clasen <mclasen@redhat.com>
32 * configure.in: Bump version
34 * === Released 2.10.0 ===
36 * README.in: Mention the slice allocator and GInitiallyUnowned.
40 * configure.in: Bump version to 2.10.0
42 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
44 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
47 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
49 * tests/completion-test.c: (main): Plug leaks reported by
51 * tests/convert-test.c: (test_iconv_state), (test_one_half):
53 * tests/patterntest.c: (test_compilation): Same
54 * tests/shell-test.c: (do_argv_test): Same.
55 * tests/unicode-caseconv.c: (main): Same.
56 * tests/uri-test.c: (run_to_uri_tests): Same.
59 2006-02-19 Martyn Russell <martyn@imendio.com>
61 * tests/asyncqueue-test.c:
63 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
64 _insert_sorted and _insert_sorted_with_data API.
66 2006-02-18 Matthias Clasen <mclasen@redhat.com>
68 * tests/gobject/Makefile.am: Add paramspec-test
70 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
72 * tests/gobject/gvalue-test.c: Add more tests.
74 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
76 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
78 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
80 * glib/gthreadpool.c: Fix deadlock when signalling the thread
81 which freed a thread pool (#331110, Chris Wilson).
83 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
85 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
86 posix_memalign() in case we actually are going to use it, because
87 we detected a compliant implementation (#328997).
89 * configure.in (enable_included_printf): don't include malloc.h when
90 testing for posix_memalign() funcitonality, since this may break the
91 test on some systems (#328997).
93 2006-02-14 Matthias Clasen <mclasen@redhat.com>
95 * glib/gstrfuncs.c: Improve docs.
97 2006-02-11 Matthias Clasen <mclasen@redhat.com>
99 * configure.in: Bump version
101 2006-02-10 Matthias Clasen <mclasen@redhat.com>
103 * === Released 2.9.6 ===
107 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
109 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
110 the type of varargs arguments. (bug #317679)
112 2006-01-31 Matthias Clasen <mclasen@redhat.com>
114 * autogen.sh: Touch README and INSTALL here to pacify
115 automake. (#329124, Kjartan Maraas, Tim Janik)
117 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
119 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
120 some missing cases where pointed out by Sebastian Wilhelmi.
122 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
124 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
126 2006-01-30 David Schleef <ds@schleef.org>
128 * glib/grand.c: Update URL
130 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
132 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
133 create container classes which are as flexible as a GValue is
135 2006-01-27 Matthias Clasen <mclasen@redhat.com>
137 * configure.in: Bump version
139 * === Released 2.9.5 ===
141 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
146 2006-01-26 Matthias Clasen <mclasen@redhat.com>
148 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
151 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
153 * fixed buglets reported by Jens Granseuer in #328254.
155 * configure.in: free the memory allocated in posix_memalign() tests.
157 * glib/gslice.c: spelling fixes.
159 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
161 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
162 slices, make sure g_mem_gc_friendly is properly initialized.
164 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
165 the first allocation. applied some branching optimizations.
167 * docs/macros.txt: reflected --enable-gc-friendly change and
168 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
170 * configure.in: changed --enable-gc-friendly=yes to define
171 ENABLE_GC_FRIENDLY_DEFAULT.
173 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
174 if (G_UNLIKELY (g_mem_gc_friendly)).
177 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
178 care of by g_slice_free1().
180 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
182 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
184 * glib/gslice.c: only use posix_memalign() if it's known to work,
185 revert to memalign() otherwise.
187 * configure.in: check for broken posix_memalign() implementations
190 2006-01-24 Matthias Clasen <mclasen@redhat.com>
192 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
193 on all little-endian systems. (#143380, Marc Moorcroft)
195 2006-01-23 Matthias Clasen <mclasen@redhat.com>
197 * configure.in: Change the shared libary extension for hpux-ia64
198 to so. (#328253, The Written Word)
200 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
202 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
203 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
206 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
208 * glib/gslice.c (slice_config_init): initialize GSlice config from
209 G_SLICE environment variable. we support G_SLICE=always-malloc
210 currently, which forces all g_slice_*() allocations to use the system
214 g_parse_debug_string(): added a note about not using g_malloc() here.
215 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
216 g_slice. contains only guesswork in the WIN32 branch.
218 2006-01-18 Matthias Clasen <mclasen@redhat.com>
222 * === Released 2.9.4 ===
226 2006-01-17 Matthias Clasen <mclasen@redhat.com>
228 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
229 call. (#327421, Morten Welinder)
231 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
233 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
234 G_LOCK. Use the unused_thread_queue lock instead. Change
235 g_thread_pool_thread_proxy such that threads only wait on
236 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
237 due to superfluous tasks. Global tasks wait at most for
238 max-idle-time milliseconds. Make sure, that no task is woken up
239 twice for the same event via a wakeup_serial. This fixes #324228.
241 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
242 invalid NULL into the thread pools. This as well fixes #327290.
244 2006-01-16 Matthias Clasen <mclasen@redhat.com>
246 * configure.in: Bump version
248 * === Released 2.9.3 ===
252 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
253 Rework the strict aliasing fix to not break C++, pointed
254 out by Murray Cumming.
256 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
257 Signal waiting threads, problem noticed by Christian Kellner.
259 2006-01-16 Matthias Clasen <mclasen@redhat.com>
261 Fix bug 326747, Alberto Ruiz:
263 * glib/gconvert.c (make_valid_utf8): Change this function to
264 replace unknown characters by the Unicode replacement character
265 instead of '?', and don't append "(invalid encoding)".
266 (g_filename_display_name, g_filename_display_basename): Document
267 how to determine if the filename was in an invalid encoding.
269 2006-01-14 Matthias Clasen <mclasen@redhat.com>
271 * glib/gtree.c: Replace the simple recursive implementation
272 by a nonrecursive, threaded implementation by Maurizio
275 2006-01-12 Matthias Clasen <mclasen@redhat.com>
277 * tests/asyncqueue-test.c (main): Fix the
278 assert to not always trigger. (#326558,
281 2006-01-11 Matthias Clasen <mclasen@redhat.com>
283 * tests/convert-test.c: Don't test items_read and
284 items_written in error cases where it is not set.
286 * tests/asyncqueue-test.c: Set error to NULL.
288 2006-01-08 Matthias Clasen <mclasen@redhat.com>
290 * glib/gtree.c: Break some long lines.
292 * tests/tree-test.c: Add more tests.
294 * glib/gslice.c: Include config.h first, pointed out
297 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
299 * docs/reference/gobject/tmpl/gtype.sgml: explain that
300 G_TYPE_INSTANCE_GET_CLASS() does behave different during
303 2006-01-06 Matthias Clasen <mclasen@redhat.com>
305 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
306 that we have posix_memalign(). (#323937, Bogdan Nicula)
308 2006-01-05 Matthias Clasen <mclasen@redhat.com>
310 * configure.in: Bump version
312 * === Released 2.9.2 ===
314 2006-01-05 Hans Breuer <hans@breuer.org>
316 * glib/makefile.msc.in : added gslice.obj
318 2006-01-05 Matthias Clasen <mclasen@redhat.com>
322 * glib/gmain.c: Clarify the documentation of source ids
323 a bit. (#325874, Dan Williams)
325 * configure.in: Fix another strict aliasing problem.
327 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
330 2006-01-04 Manish Singh <yosh@gimp.org>
333 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
334 either, since they have the same issue as g_rmdir. (Related to
337 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
339 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
340 order when creating magazines, so we hand out chunks with ascending
343 2006-01-04 Abel Cheung <maddog@linuxhall.org>
345 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
347 2006-01-03 Martyn Russell <martyn@imendio.com>
349 * docs/reference/glib/glib-sections.txt:
351 * glib/gthreadpool.[ch]:
352 - Added new API g_thread_pool_get_idle_time() and
353 g_thread_pool_set_idle_time(). (#324228).
355 * tests/threadpool-test.c:
356 - Updated test case to do thread pool sorting, thread pool with
357 no sorting and a thread pool with idle thread timeouts.
359 2006-01-03 Matthias Clasen <mclasen@redhat.com>
361 * glib/gmain.h: Add new functions here, too.
363 * glib/glib.symbols: Add new functions.
365 2005-12-20 Michael Meeks <michael.meeks@novell.com>
367 * glib/gmain.c (g_main_context_is_owner): new method
368 to determine if the current thread is the owner of the
371 2006-01-02 Matthias Clasen <mclasen@redhat.com>
375 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
376 expanding to rmdir, since rmdir is not declared in a portable
377 system header we can include in gstdio.h. (#325249, Jani Monoses)
379 2006-01-01 Tor Lillqvist <tml@novell.com>
381 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
382 stdout and stderr to unbuffered. Otherwise the giochannel layer
383 will try to read from them regardless whether the
384 g_io_channel_win32_poll() call here has indicated
385 readability or not. (#325310)
387 2005-12-29 Matthias Clasen <mclasen@redhat.com>
389 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
390 in the docs. (#325273, Declan Naughton)
392 2005-12-27 Manish Singh <yosh@gimp.org>
394 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
395 make sure we really override things.
397 2005-12-27 Matthias Clasen <mclasen@redhat.com>
399 Fix #316221, Michal Benes, Stanislav Brabec;
401 * configure.in: Fix a strict aliasing problem in
402 g_static_mutex_get_mutex().
403 * glib/gthread.h: ...and in
404 g_static_mutex_get_mutex_impl_shortcut().
406 * glib/gdatasetprivate.h: Add a cast to silence compiler
407 warnings. (#321978, Andrew Paprocki)
409 Partial fix for bug #323937, Bogdan Nicula.
411 * configure.in: Check for malloc.h
412 * glib/gslice.c: Don't include malloc.h unconditionally.
414 2005-12-27 Manish Singh <yosh@gimp.org>
416 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
417 (#324950, Dan Yefimov)
419 2005-12-26 Matthias Clasen <mclasen@redhat.com>
421 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
424 2005-12-25 Matthias Clasen <mclasen@redhat.com>
426 * AUTHORS: Update my email
428 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
430 2005-12-21 Matthias Clasen <mclasen@redhat.com>
432 * README.in: Add some notes on when 'make check' may fail.
434 2005-12-20 Matthias Clasen <mclasen@redhat.com>
437 * glib/gthreadpool.h:
438 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
439 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
441 * tests/threadpool-test.c: Test this.
443 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
445 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
446 handle errors without depending on gmessage.c which might not be
447 setup when the error occours.
448 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
449 always freeing can be achieved by adjusting the working set time to
450 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
451 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
452 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
453 minimize wastage if small amounts of differently sized structrues are
454 allocated, this does come at a performance cost of roughly 5% though.
455 fixed up block alignment calculation, so it works for varying
456 block sizes. only use strerror() not g_strerror() since the latter
457 depends on working GQuark and GSlice.
458 mem_error(): implemented in terms of fprintf and vfprintf.
460 * tests/slice-color.c: new program to test cache colorization effects.
462 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
463 G_SLICE_CONFIG_WORKING_SET_MSECS.
465 2005-12-17 Matthias Clasen <mclasen@redhat.com>
467 * glib/goption.c (parse_short_option): Set an error in all
468 failure cases. (#324332, Tim-Philipp Müller)
470 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
472 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
473 and g_atomic_pointer_get. Add g_atomic_int_set and
474 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
475 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
476 as functions (additionally to the macros in the header) for the
477 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
479 2005-12-16 Matthias Clasen <mclasen@redhat.com>
481 * glib/gmem.c (g_allocator_new): Don't return a pointer to
482 a const struct, since apps expect to be able to modify it.
483 (#324179, J. Ali Harlow)
485 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
487 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
488 for systems where the initialization of atomic variables requires a
489 write memory barrier.
491 2005-12-09 Matthias Clasen <mclasen@redhat.com>
493 * configure.in: Bump version
495 * === Released 2.9.1 ===
497 2005-12-09 Alexander Larsson <alexl@redhat.com>
501 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
504 2005-12-08 Matthias Clasen <mclasen@redhat.com>
508 2005-12-07 Martyn Russell <martyn@imendio.com>
510 * glib/gasyncqueue.c:
511 - Call g_queue_insert_sorted() instead of duplicating the code.
512 - Call g_queue_sort() instead of duplicating the code.
513 - Invert sort function results to make sure the same sort function
514 gives the same results across glist, gslist, gqueue and
517 * tests/asyncqueue-test.c:
518 - Updated the sort function to reflect the example in the
519 documentation for gasyncqueue.c.
521 2005-12-07 Martyn Russell <martyn@imendio.com>
523 * docs/reference/glib/glib-sections.txt:
524 * docs/reference/glib/tmpl/linked_lists_double.sgml:
525 * docs/reference/glib/tmpl/linked_lists_single.sgml:
528 - Added g_list_insert_sorted_with_data () and
529 g_slist_insert_sorted_with_data ().
530 - Removed the extra check in g_list_sort() and g_slist_sort() for
531 GCompareDataFunc vs. GCompareFunc.
533 2005-12-07 Tor Lillqvist <tml@novell.com>
535 * glib-gettextize.in: Look up prefix at run-time on Win32,
536 assuming the standard directory structure with glib-gettextize in
539 2005-12-06 Matthias Clasen <mclasen@redhat.com>
541 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
542 we need GTK+ 2.8 to compile against GLib 2.10.
544 2005-12-05 Manish Singh <yosh@gimp.org>
546 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
547 to override any user settings.
549 2005-12-05 Matthias Clasen <mclasen@redhat.com>
551 * configure.in: Revert an accidential commit
553 2005-12-05 Matthias Clasen <mclasen@redhat.com>
555 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
557 * tests/Makefile.am (TESTS_ENVIRONMENT):
558 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
559 MALLOC_CHECK_ and MALLOC_PERTURB_
561 * tests/run-collate-tests.sh: Run the collation tests explicitly
562 in en_US locale. (#320463)
564 * glib/gmem.h: Really deprecate GMemChunk.
571 * glib/gstrfuncs.c: Add versioned deprecation docs.
573 2005-12-05 Martyn Russell <martyn@imendio.com>
575 * docs/reference/glib/glib-sections.txt:
576 * glib/gasyncqueue.[ch]:
577 - Added support for sorting async queues by with _push_sorted(),
578 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
581 * tests/asyncqueue-test.c:
582 - Added test case for gasyncqueue.c
584 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
586 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
587 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
588 in bug 323178. moved time consuming logic from g_slice_free() out of
589 the inner loop, so g_slice_free_chain_with_offset() provides a real
590 performance benefit over g_slice_free1() now.
592 * glib/gslice.h: renamed g_slice_free_chain() to
593 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
594 a type-safe macro as suggested in bug 323178.
595 simplified the macro implementation of g_slice_free() and implemented
596 it in a type safe manner for all compliers as suggested by Morten
597 Welinder <mortenw@gnome.org>.
602 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
604 2005-12-05 Matthias Clasen <mclasen@redhat.com>
606 * glib/gasyncqueue.c: Add some docs.
608 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
610 * glib/gatomic.c: In the ia64 implementation, use
611 __sync builtin without _si or _di suffix. (#321229,
612 Stanislav Brabec, patch by Andreas Schwab)
614 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
616 * glib/gslice.h: Remove comma at the end of enum.
618 2005-12-04 Matthias Clasen <mclasen@redhat.com>
620 Handle multiple user names with the same UID better.
621 (#319535, Laszlo Peter)
623 * glib/gutils.c (g_get_any_init_do): When determining user
624 data, first look up $LOGNAME. If the UID doesn't match
625 getuid(), fall back to the current behaviour of looking
626 up the user data based on getuid().
628 2005-12-04 Matthias Clasen <mclasen@redhat.com>
630 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
631 to make gcc warn if a function result is ignored. (#145466,
632 Arjan van de Ven, Alex Larsson)
634 * glib/gmem.h: Add the new attribute to g_realloc and
637 2005-12-03 Matthias Clasen <mclasen@redhat.com>
640 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
643 * glib/gmappedfile.c: Make mapping of empty files
646 * glib/gfileutils.c: Don't fork a new process just to
647 fix the permissions of the created temp file. (#321318,
648 Alexis S. L. Carvalho)
650 2005-12-02 Matthias Clasen <mclasen@redhat.com>
652 * README.in: Add a note about Solaris threads.
655 * configure.in: Remove support for Solaris threads.
656 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
658 2005-12-02 Matthias Clasen <mclasen@redhat.com>
660 * configure.in: Define G_GUINT64_CONSTANT in analogy to
661 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
663 2005-12-02 Matthias Clasen <mclasen@redhat.com>
665 * glib/gslice.c: Win32 portability fixes and C99ism removal,
666 pointed out by Kazuki Iwamoto. (#323052)
668 2005-12-02 Matthias Clasen <mclasen@redhat.com>
670 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
671 the IA64 implementation.
673 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
675 * glib/gslice.c: some naming and type size fixups.
677 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
679 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
680 variant into something less verbose, but digestible for gcc-3.4.
682 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
684 * tests/slice-test.c: extended to perform the benchmarking on the old
685 memchunk code if 'O' is selected.
687 * tests/memchunks.c: new file which contains the old GLib mem chunks
688 implementation with prefix old_mem_chunk_.
690 * tests/Makefile.am: added memchunks.c
692 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
694 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
695 acceptable load distribution of the syscall.
696 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
697 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
699 2005-12-01 Matthias Clasen <mclasen@redhat.com>
701 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
702 double-free base interface members.
704 * tests/Makefile.am: Remove duplicate variable.
706 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
708 * glib/gslice.[hc]: new slice allocator implementation.
710 * tests/slice-test.c: added random slice allocation test.
712 * glib/gthread.[hc]: removed newly added private thread mem API.
714 * glib/gthreadinit.h:
717 * glib/gmem.c: divided glib threading initialisation into three phases,
718 initialisation where private keys and messaging are not available (only
719 needed by gmem.c), initialisation without messaging but private keys
720 available (gslice.c, gmessage.c), and full fledged initialisers that
721 server the rest of glib. initialisation functions got renamed to reflect
722 the limitations of their corresponding phases.
724 * glib/gmem.c: removed memchunk code, defer allocations to
727 * glib/gmem.[hc]: removed g_slice_* skeletons.
729 * glib/glib.symbols: added g_slice_* symbols.
731 * configure.in: check for availability of posix_memalign(3), memalign(3)
734 * glib/Makefile.am: added gslice.[hc].
736 2005-12-01 Tor Lillqvist <tml@novell.com>
738 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
739 trailing slash(es) for non-root folders. stat() fails if non-root
740 folders are specified with trailing slashes. It's too much hassle
741 to demand that callers strip such slashes themselves, especially
742 as it is easy to get it wrong and strip the slash of a root
744 (g_rename): On NT-based Windows, use MoveFileEx() with
745 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
747 2005-11-28 Matthias Clasen <mclasen@redhat.com>
749 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
752 * configure.in: Check whether do { } while (0) works.
754 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
755 G_STMT_END if it works.
757 2005-11-28 Matthias Clasen <mclasen@redhat.com>
759 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
760 * glib/gnode.c (g_node_depth_traverse_level):
761 * glib/gmem.c (g_allocator_new):
762 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
763 (g_hash_table_foreach_remove_or_steal): Silence compiler
766 2005-11-27 Matthias Clasen <mclasen@redhat.com>
768 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
769 Don't read beyond len. (#322520, Christian Persch)
771 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
773 * glib/gdataset.c: access datalist flags via atomic pointer access
774 functions, instead of acquiring the dataset lock. this is faster and
775 also matches the atomic pointer readouts in gdatalistprivate.h and
776 direct pointer modifications required by gobject.c.
778 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
780 * glib/gdataset.c: streamlined initialization code somewhat,
781 removed GData node cache.
783 2005-11-22 Matthias Clasen <mclasen@redhat.com>
785 * glib/ghash.c (g_hash_table_ref):
786 (g_hash_table_unref): Mark these as new API.
788 Avoid double locking in g_intern_string (#322133,
791 * glib/gdataset.c (g_quark_from_string_internal): New
792 internal function which factors out common parts of
793 g_quark_from[_static]_string.
794 (g_quark_from_string, g_quark_from_static_string):
795 Use g_quark_from_string_internal.
796 (g_intern_string, g_intern_static_string): Use
797 g_quark_from_string_internal, and only take the
799 (g_quark_new): Don't store the strings shifted by -1
800 in the g_quarks array.
801 (g_quark_to_string): Adapt to the previous change.
803 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
807 g_hash_table_new_full(): create hash tables with a ref count of 1.
808 g_hash_table_ref(): atomically ref_count+=1
809 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
810 when refcount reaches 0.
811 g_hash_table_destroy(): just destroy keys and values, unref by 1.
812 g_hash_table_insert():
813 g_hash_table_replace(): assert ref_count>0.
816 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
819 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
821 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
822 variables. (#321984, Andrew Paprocki)
824 2005-11-18 Matthias Clasen <mclasen@redhat.com>
826 * configure.in: Bump version
828 2005-11-17 Matthias Clasen <mclasen@redhat.com>
830 * === Released 2.9.0 ===
832 * glib/glib.symbols: Updates
834 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
835 hints to the docs. (Paolo Borelli)
839 * glib/gthread.c (g_thread_foreach): Mark as new api.
841 * README.in: Updates.
845 * glib/gdate.c (g_date_set_time_t):
846 (g_date_set_time_val): Add functions to set a date from
847 a time_t and from a GTimeVal.
848 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
851 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
852 and g_utf8_offset_to_pointer.
854 * glib/gutf8.c (g_utf8_pointer_to_offset)
855 (g_utf8_offset_to_pointer): Handle negative offsets, and use
856 "stutter stepping" for going backwards. (#320638, Larry
861 * glib/gthread.c: const correctness fixes, found
862 by Arjan van de Ven and gcc.
864 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
866 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
867 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
869 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
871 * glib/guniprop.c: Use bit hacks instead when checking a general
872 category value against multiple values.
874 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
876 2005-11-10 Simos Xenitellis <simos@gnome.org>
878 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
880 2005-11-05 Tor Lillqvist <tml@novell.com>
882 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
883 (#320688) Instead set error and return failure.
885 2005-11-07 Matthias Clasen <mclasen@redhat.com>
887 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
888 while holding the context lock. (#320886, Andy Wingo)
890 2005-11-07 Billy Biggs <vektor@dumbterm.net>
892 * tests/refcount/Makefile.am: Add a missing reference to libglib
893 in the LDADD for this test.
895 2005-11-07 Matthias Clasen <mclasen@redhat.com>
897 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
898 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
899 GLib 2.9. (#320755, Luca Ferretti)
901 2005-11-04 Matthias Clasen <mclasen@redhat.com>
903 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
904 use a faster lookup table for g_unichar_get_mirror_char().
905 * glib/gmirroringtable.h: The new table.
906 * glib/gunichartables.h: Remove bidi_mirroring_table.
907 * glib/gen-unicode-tables.pl: Don't generate the mirroring
909 * glib/glib-mirroring-tab/*: Sources for the program
910 which generated gmirroringtable.h.
912 * glib/glist.c: Avoid some code duplication.
914 * glib/gscanner.h: Include gdataset.h. (#320322)
916 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
917 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
918 to help find critical warnings in applications. (#320017,
921 2005-11-02 Tor Lillqvist <tml@novell.com>
923 * glib/glib.symbols: Remove large amount of trailing whitespace
924 from one line. Remove #ifdef G_UNDEFINED from around
925 g_hash_table_new and g_hash_table_new_full symbols, what was the
926 purpose of that? Add the g_slice functions.
928 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
930 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
931 added g_slice_*() API to allocate and cache small bits of memory.
932 an actuall allocator implementation for g_slice_*() is still pending.
934 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
935 changed GRealThread list to use in-structure *next; fields instead
936 of GSList, in order for thread iteration to not depenend on g_slice_*()
938 _g_thread_mem_private_get():
939 _g_thread_mem_private_set(): added accessors for private memory,
940 needed because the ordinary GPrivate implementation relies on GArray
941 and GSList and therefore indirectly on working g_slice_*() allocations.
944 g_thread_foreach(): new public API function to loop over all existing threads.
958 removed GAllocator and free list usages and accompanying locks.
959 use g_slice_*() API to allocate and cache small bits of memory.
961 * glib/ghook.h: removed GMemChunk field from public API.
964 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
966 * glib/gnode.h: deprecate allocator API.
968 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
969 be used for poll rec lists.
971 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
972 g_relation_destroy(): free all tuples from the all_tuples hash table,
973 this effectively maintains the life time track keeping of tuples.
974 g_relation_delete_tuple(): free tuples which are removed from the
975 all_tuples hash table. this fixes a temporary leak that was present
976 in the memchunk code until the destruction of the relation.
978 2005-10-29 Matthias Clasen <mclasen@redhat.com>
980 * tests/convert-test.c: Add some tests for conversions between
981 UTF-8, UCS-4 and UTF-16.
983 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
984 of len == -1, noticed by Morten Welinder.
986 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
988 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
990 2005-10-26 Matthias Clasen <mclasen@redhat.com>
992 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
993 in the error case, and add some documentation. (#319806, Morten
996 2005-10-19 Manish Singh <yosh@gimp.org>
998 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
999 config.h symbol automatically. Fixes bug #313731.
1001 2005-10-19 Tor Lillqvist <tml@novell.com>
1003 * glib/gwin32.c (g_win32_get_package_installation_directory):
1004 Return a g_strdup()ed copy of the value stored in the hash table,
1005 so that it can be g_free()d without leaving a dangling pointer in
1006 the hash table. (#319232)
1008 2005-10-06 Matthias Clasen <mclasen@redhat.com>
1010 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
1011 all-zero sequences correctly. (#317930, Sebastien Bacher)
1013 2005-10-05 Matthias Clasen <mclasen@redhat.com>
1015 * glib/gmarkup.c (g_markup_escape_text): Doc addition
1017 * Makefile.am: Add Collation tests.
1019 * tests/collate/*: Inputs and expected outputs for collation tests.
1021 * tests/run-collate-tests.sh: Script to run collation tests.
1023 * tests/unicode-collate.c (main): Rework slightly to make
1024 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
1026 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
1028 * docs/reference/glib/tmpl/unicode.sgml:
1029 * glib/gen-unicode-tables.pl:
1031 * glib/gunichartables.h:
1033 * tests/casefold.txt:
1034 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1035 GUnicodeBreakType types. That may break some applications, like
1038 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1040 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1041 docs. (#314393, Matthew F. Barnes)
1043 * glib/glib.symbols:
1044 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1045 they are already declared in glib.h. This doesn't break documented
1046 use of gprintf.h, but should probably be pointed out in the
1047 release notes for 2.10. (#314232, Behdad Esfahbod)
1049 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1051 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1052 complexity optimization, based on a patch from Matthias Clasen.
1054 * tests/patterntest.c: more tests, mostly from matthias.
1056 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1058 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1059 (#316703, Mark Drago)
1061 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1063 * glib/gpattern.c: applied a patch from matthias which checks on the
1064 upper bound of GPatternSpec length to optimize matches.
1067 * tests/patterntest.c: added more match cases.
1069 2005-09-16 Tor Lillqvist <tml@novell.com>
1071 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1072 warning to doc comment that these functions should not be used on
1073 encodings like CP932.
1075 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1077 * tests/keyfile-test.c: Add a test for grup names of length 1.
1079 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1080 of length 1. (#316309)
1082 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1084 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1086 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1088 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1089 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1092 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1094 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1095 (child_watch_helper_thread): Remove some dead code. Closes
1098 2005-09-07 Tor Lillqvist <tml@novell.com>
1100 * glib/Makefile.am: Create also a console version of the
1101 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1102 It's otherwise identical to gspawn-win32-helper.exe, except marked
1103 as a console application (linked without the -mwindows option).
1105 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1106 Drop the dont_wait parameter. Its truth value correlated 100% with
1107 the NULLness of the exit_status parameter anyway, so it's enough
1108 to check whether exit_status is NULL. Invert the sense of the
1109 dont_return_handle parameter and rename it to do_return_handle, to
1110 make the code easier to read by avoiding double negations.
1112 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1113 to do_spawn_with_pipes() accordingly.
1115 (do_spawn_with_pipes): If we have a console, use the console
1116 version of the helper program, otherwise use the GUI one. This
1117 avoids extra console windows opening up in some situations. (In
1118 case a console application uses the GUI gspawn-win32-helper.exe to
1119 spawn another console application we would get a separate console
1120 for the spawned console application).
1122 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1124 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1126 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1127 if the file is too large. (#315275, Kjartan Maraas)
1129 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1130 of read() is signed. (#315273, Kjartan Maraas)
1132 2005-08-31 Tor Lillqvist <tml@novell.com>
1134 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1135 the EXE and each DLL have their own atexit function chains.
1137 #define g_atexit as atexit instead. This means it has a
1138 better chance of doing what the caller wants. For instance,
1139 gtkhtml calls g_atexit() registering a function in gtkhtml
1140 itself. This caused a crash when g_atexit() was implemented as a
1141 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1142 the time the GLib DLL got unloaded.
1144 * glib/gutils.c: #undef the #define mentioned above, to also get a
1145 real g_atexit() into the DLL for backward compatibility. Document
1146 the Windows behaviour of g_atexit(), and document the varying ways
1147 atexit() can behave in the context of dynamically loaded modules
1150 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1152 * glib/glib.symbols:
1154 * glib/gdataset.c: Add string interning functions.
1156 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1158 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1161 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1163 * configure.in: Bump version to 2.9.0
1165 2005-08-25 Tor Lillqvist <tml@novell.com>
1167 Make also the g_spawn*() functions take parameters in the GLib
1168 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1169 API or ABI. Like the other GLib API that was earlier changed to
1170 use UTF-8 on Windows, the names of the functions that take UTF-8
1171 have _utf8 suffixes added by using preprocessor macros in the
1172 header file. The old names are kept for functions with the old
1173 behaviour, taking parameters in the system codepage, for DLL ABI
1176 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1177 g_spawn*() functions.
1179 * glib/gspawn-win32.c: Use wide-char API on NT-based
1180 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1181 system codepage (Win9x) and call the C library _wspawn*() or
1182 spawn*() functions respectvely. Add DLL ABI stability versions
1183 that take parameters in the system codepage.
1185 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1186 wide-char versions of argv and envp, and use wide-char API to
1187 change directory and spawn the program to run. Remove the verbose
1188 debugging output, it was too complex to modify for the wide-char
1189 features. (Just add temporary debugging printouts if needed, no
1190 need to have them permanently in the source.)
1192 * glib/gspawn.c: Corresponding documentation updates.
1194 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1195 symbols as PRIVATE, add the new _utf8-suffixed ones.
1197 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1199 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1200 guint64 otherwise and that can produce warnings about comparison
1201 between signed and unsigned.
1203 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1205 * glib/gutils.c: Fix the crt_externs.h include.
1207 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1209 * NEWS: Fix spelling of my first name.
1211 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1215 * === Released 2.8.1 ===
1219 2005-08-20 Hans Breuer <hans@breuer.org>
1221 * glib/makefile.msc.in : link with ws2_32.lib
1223 2005-08-18 Tor Lillqvist <tml@novell.com>
1225 * configure.in: Check for <sys/wait.h>
1227 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1229 2005-08-18 Ross Burton <ross@burtonini.com>
1232 Optimise single-character insertions.
1237 * tests/string-test.c:
1238 Add tests for new optimisation, and fix a leak.
1240 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1242 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1244 * glib/gutils.c: On Darwin, include crt-externs.h and
1245 define environ using _NSGetEnviron(). (#313731)
1247 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1249 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1250 (g_get_any_init_do): ... to this new function.
1251 (g_get_any_init): Declare as inline.
1252 (g_get_any_init_locked): New inline function, does the locking.
1253 Make use of these two throughout the code.
1255 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1257 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1258 the child process and then simply return. This makes
1259 The "S" option work as documented in g_on_error_query().
1260 (#313125, Matthew F. Barnes)
1262 * glib/gunicode.h: Update the link to Unicode category
1263 values. (#313369, Behnam Esfahbod)
1265 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1266 a little. (#311727, Tristan van Berkom)
1268 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1269 check work on ia64 too, where some symbols ended up
1270 in yet another section.