1 2006-03-06 Anders Carlsson <andersca@imendio.com>
4 Make .dylib the shared library suffix on darwin.
5 (#322476, Vladimir Panov)
7 2006-03-06 Anders Carlsson <andersca@imendio.com>
10 Skip checking thread flags on Darwin. Fix suggested by Manish Singh.
11 (#314794, Gregor Riepl)
13 2006-03-01 Tor Lillqvist <tml@novell.com>
15 * glib/gutils.c (g_listenv): Separate implementation on Win32: Use
16 the wide character API on NT-based Windows. Return UTF-8 strings.
18 * glib/glib.symbols: Don't mark g_listenv as PRIVATE, as that
19 meant it wasn't present in the import library. PRIVATE is used
20 only for the backwards-compatibility DLL ABI stability hacks.
22 2006-02-24 Matthias Clasen <mclasen@redhat.com>
24 * configure.in: Bump version
26 * === Released 2.10.0 ===
28 * README.in: Mention the slice allocator and GInitiallyUnowned.
32 * configure.in: Bump version to 2.10.0
34 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
36 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
39 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
41 * tests/completion-test.c: (main): Plug leaks reported by
43 * tests/convert-test.c: (test_iconv_state), (test_one_half):
45 * tests/patterntest.c: (test_compilation): Same
46 * tests/shell-test.c: (do_argv_test): Same.
47 * tests/unicode-caseconv.c: (main): Same.
48 * tests/uri-test.c: (run_to_uri_tests): Same.
51 2006-02-19 Martyn Russell <martyn@imendio.com>
53 * tests/asyncqueue-test.c:
55 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
56 _insert_sorted and _insert_sorted_with_data API.
58 2006-02-18 Matthias Clasen <mclasen@redhat.com>
60 * tests/gobject/Makefile.am: Add paramspec-test
62 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
64 * tests/gobject/gvalue-test.c: Add more tests.
66 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
68 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
70 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
72 * glib/gthreadpool.c: Fix deadlock when signalling the thread
73 which freed a thread pool (#331110, Chris Wilson).
75 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
77 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
78 posix_memalign() in case we actually are going to use it, because
79 we detected a compliant implementation (#328997).
81 * configure.in (enable_included_printf): don't include malloc.h when
82 testing for posix_memalign() funcitonality, since this may break the
83 test on some systems (#328997).
85 2006-02-14 Matthias Clasen <mclasen@redhat.com>
87 * glib/gstrfuncs.c: Improve docs.
89 2006-02-11 Matthias Clasen <mclasen@redhat.com>
91 * configure.in: Bump version
93 2006-02-10 Matthias Clasen <mclasen@redhat.com>
95 * === Released 2.9.6 ===
99 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
101 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
102 the type of varargs arguments. (bug #317679)
104 2006-01-31 Matthias Clasen <mclasen@redhat.com>
106 * autogen.sh: Touch README and INSTALL here to pacify
107 automake. (#329124, Kjartan Maraas, Tim Janik)
109 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
111 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
112 some missing cases where pointed out by Sebastian Wilhelmi.
114 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
116 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
118 2006-01-30 David Schleef <ds@schleef.org>
120 * glib/grand.c: Update URL
122 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
124 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
125 create container classes which are as flexible as a GValue is
127 2006-01-27 Matthias Clasen <mclasen@redhat.com>
129 * configure.in: Bump version
131 * === Released 2.9.5 ===
133 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
138 2006-01-26 Matthias Clasen <mclasen@redhat.com>
140 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
143 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
145 * fixed buglets reported by Jens Granseuer in #328254.
147 * configure.in: free the memory allocated in posix_memalign() tests.
149 * glib/gslice.c: spelling fixes.
151 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
153 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
154 slices, make sure g_mem_gc_friendly is properly initialized.
156 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
157 the first allocation. applied some branching optimizations.
159 * docs/macros.txt: reflected --enable-gc-friendly change and
160 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
162 * configure.in: changed --enable-gc-friendly=yes to define
163 ENABLE_GC_FRIENDLY_DEFAULT.
165 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
166 if (G_UNLIKELY (g_mem_gc_friendly)).
169 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
170 care of by g_slice_free1().
172 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
174 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
176 * glib/gslice.c: only use posix_memalign() if it's known to work,
177 revert to memalign() otherwise.
179 * configure.in: check for broken posix_memalign() implementations
182 2006-01-24 Matthias Clasen <mclasen@redhat.com>
184 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
185 on all little-endian systems. (#143380, Marc Moorcroft)
187 2006-01-23 Matthias Clasen <mclasen@redhat.com>
189 * configure.in: Change the shared libary extension for hpux-ia64
190 to so. (#328253, The Written Word)
192 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
194 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
195 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
198 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
200 * glib/gslice.c (slice_config_init): initialize GSlice config from
201 G_SLICE environment variable. we support G_SLICE=always-malloc
202 currently, which forces all g_slice_*() allocations to use the system
206 g_parse_debug_string(): added a note about not using g_malloc() here.
207 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
208 g_slice. contains only guesswork in the WIN32 branch.
210 2006-01-18 Matthias Clasen <mclasen@redhat.com>
214 * === Released 2.9.4 ===
218 2006-01-17 Matthias Clasen <mclasen@redhat.com>
220 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
221 call. (#327421, Morten Welinder)
223 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
225 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
226 G_LOCK. Use the unused_thread_queue lock instead. Change
227 g_thread_pool_thread_proxy such that threads only wait on
228 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
229 due to superfluous tasks. Global tasks wait at most for
230 max-idle-time milliseconds. Make sure, that no task is woken up
231 twice for the same event via a wakeup_serial. This fixes #324228.
233 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
234 invalid NULL into the thread pools. This as well fixes #327290.
236 2006-01-16 Matthias Clasen <mclasen@redhat.com>
238 * configure.in: Bump version
240 * === Released 2.9.3 ===
244 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
245 Rework the strict aliasing fix to not break C++, pointed
246 out by Murray Cumming.
248 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
249 Signal waiting threads, problem noticed by Christian Kellner.
251 2006-01-16 Matthias Clasen <mclasen@redhat.com>
253 Fix bug 326747, Alberto Ruiz:
255 * glib/gconvert.c (make_valid_utf8): Change this function to
256 replace unknown characters by the Unicode replacement character
257 instead of '?', and don't append "(invalid encoding)".
258 (g_filename_display_name, g_filename_display_basename): Document
259 how to determine if the filename was in an invalid encoding.
261 2006-01-14 Matthias Clasen <mclasen@redhat.com>
263 * glib/gtree.c: Replace the simple recursive implementation
264 by a nonrecursive, threaded implementation by Maurizio
267 2006-01-12 Matthias Clasen <mclasen@redhat.com>
269 * tests/asyncqueue-test.c (main): Fix the
270 assert to not always trigger. (#326558,
273 2006-01-11 Matthias Clasen <mclasen@redhat.com>
275 * tests/convert-test.c: Don't test items_read and
276 items_written in error cases where it is not set.
278 * tests/asyncqueue-test.c: Set error to NULL.
280 2006-01-08 Matthias Clasen <mclasen@redhat.com>
282 * glib/gtree.c: Break some long lines.
284 * tests/tree-test.c: Add more tests.
286 * glib/gslice.c: Include config.h first, pointed out
289 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
291 * docs/reference/gobject/tmpl/gtype.sgml: explain that
292 G_TYPE_INSTANCE_GET_CLASS() does behave different during
295 2006-01-06 Matthias Clasen <mclasen@redhat.com>
297 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
298 that we have posix_memalign(). (#323937, Bogdan Nicula)
300 2006-01-05 Matthias Clasen <mclasen@redhat.com>
302 * configure.in: Bump version
304 * === Released 2.9.2 ===
306 2006-01-05 Hans Breuer <hans@breuer.org>
308 * glib/makefile.msc.in : added gslice.obj
310 2006-01-05 Matthias Clasen <mclasen@redhat.com>
314 * glib/gmain.c: Clarify the documentation of source ids
315 a bit. (#325874, Dan Williams)
317 * configure.in: Fix another strict aliasing problem.
319 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
322 2006-01-04 Manish Singh <yosh@gimp.org>
325 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
326 either, since they have the same issue as g_rmdir. (Related to
329 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
331 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
332 order when creating magazines, so we hand out chunks with ascending
335 2006-01-04 Abel Cheung <maddog@linuxhall.org>
337 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
339 2006-01-03 Martyn Russell <martyn@imendio.com>
341 * docs/reference/glib/glib-sections.txt:
343 * glib/gthreadpool.[ch]:
344 - Added new API g_thread_pool_get_idle_time() and
345 g_thread_pool_set_idle_time(). (#324228).
347 * tests/threadpool-test.c:
348 - Updated test case to do thread pool sorting, thread pool with
349 no sorting and a thread pool with idle thread timeouts.
351 2006-01-03 Matthias Clasen <mclasen@redhat.com>
353 * glib/gmain.h: Add new functions here, too.
355 * glib/glib.symbols: Add new functions.
357 2005-12-20 Michael Meeks <michael.meeks@novell.com>
359 * glib/gmain.c (g_main_context_is_owner): new method
360 to determine if the current thread is the owner of the
363 2006-01-02 Matthias Clasen <mclasen@redhat.com>
367 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
368 expanding to rmdir, since rmdir is not declared in a portable
369 system header we can include in gstdio.h. (#325249, Jani Monoses)
371 2006-01-01 Tor Lillqvist <tml@novell.com>
373 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
374 stdout and stderr to unbuffered. Otherwise the giochannel layer
375 will try to read from them regardless whether the
376 g_io_channel_win32_poll() call here has indicated
377 readability or not. (#325310)
379 2005-12-29 Matthias Clasen <mclasen@redhat.com>
381 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
382 in the docs. (#325273, Declan Naughton)
384 2005-12-27 Manish Singh <yosh@gimp.org>
386 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
387 make sure we really override things.
389 2005-12-27 Matthias Clasen <mclasen@redhat.com>
391 Fix #316221, Michal Benes, Stanislav Brabec;
393 * configure.in: Fix a strict aliasing problem in
394 g_static_mutex_get_mutex().
395 * glib/gthread.h: ...and in
396 g_static_mutex_get_mutex_impl_shortcut().
398 * glib/gdatasetprivate.h: Add a cast to silence compiler
399 warnings. (#321978, Andrew Paprocki)
401 Partial fix for bug #323937, Bogdan Nicula.
403 * configure.in: Check for malloc.h
404 * glib/gslice.c: Don't include malloc.h unconditionally.
406 2005-12-27 Manish Singh <yosh@gimp.org>
408 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
409 (#324950, Dan Yefimov)
411 2005-12-26 Matthias Clasen <mclasen@redhat.com>
413 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
416 2005-12-25 Matthias Clasen <mclasen@redhat.com>
418 * AUTHORS: Update my email
420 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
422 2005-12-21 Matthias Clasen <mclasen@redhat.com>
424 * README.in: Add some notes on when 'make check' may fail.
426 2005-12-20 Matthias Clasen <mclasen@redhat.com>
429 * glib/gthreadpool.h:
430 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
431 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
433 * tests/threadpool-test.c: Test this.
435 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
437 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
438 handle errors without depending on gmessage.c which might not be
439 setup when the error occours.
440 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
441 always freeing can be achieved by adjusting the working set time to
442 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
443 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
444 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
445 minimize wastage if small amounts of differently sized structrues are
446 allocated, this does come at a performance cost of roughly 5% though.
447 fixed up block alignment calculation, so it works for varying
448 block sizes. only use strerror() not g_strerror() since the latter
449 depends on working GQuark and GSlice.
450 mem_error(): implemented in terms of fprintf and vfprintf.
452 * tests/slice-color.c: new program to test cache colorization effects.
454 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
455 G_SLICE_CONFIG_WORKING_SET_MSECS.
457 2005-12-17 Matthias Clasen <mclasen@redhat.com>
459 * glib/goption.c (parse_short_option): Set an error in all
460 failure cases. (#324332, Tim-Philipp Müller)
462 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
464 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
465 and g_atomic_pointer_get. Add g_atomic_int_set and
466 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
467 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
468 as functions (additionally to the macros in the header) for the
469 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
471 2005-12-16 Matthias Clasen <mclasen@redhat.com>
473 * glib/gmem.c (g_allocator_new): Don't return a pointer to
474 a const struct, since apps expect to be able to modify it.
475 (#324179, J. Ali Harlow)
477 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
479 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
480 for systems where the initialization of atomic variables requires a
481 write memory barrier.
483 2005-12-09 Matthias Clasen <mclasen@redhat.com>
485 * configure.in: Bump version
487 * === Released 2.9.1 ===
489 2005-12-09 Alexander Larsson <alexl@redhat.com>
493 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
496 2005-12-08 Matthias Clasen <mclasen@redhat.com>
500 2005-12-07 Martyn Russell <martyn@imendio.com>
502 * glib/gasyncqueue.c:
503 - Call g_queue_insert_sorted() instead of duplicating the code.
504 - Call g_queue_sort() instead of duplicating the code.
505 - Invert sort function results to make sure the same sort function
506 gives the same results across glist, gslist, gqueue and
509 * tests/asyncqueue-test.c:
510 - Updated the sort function to reflect the example in the
511 documentation for gasyncqueue.c.
513 2005-12-07 Martyn Russell <martyn@imendio.com>
515 * docs/reference/glib/glib-sections.txt:
516 * docs/reference/glib/tmpl/linked_lists_double.sgml:
517 * docs/reference/glib/tmpl/linked_lists_single.sgml:
520 - Added g_list_insert_sorted_with_data () and
521 g_slist_insert_sorted_with_data ().
522 - Removed the extra check in g_list_sort() and g_slist_sort() for
523 GCompareDataFunc vs. GCompareFunc.
525 2005-12-07 Tor Lillqvist <tml@novell.com>
527 * glib-gettextize.in: Look up prefix at run-time on Win32,
528 assuming the standard directory structure with glib-gettextize in
531 2005-12-06 Matthias Clasen <mclasen@redhat.com>
533 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
534 we need GTK+ 2.8 to compile against GLib 2.10.
536 2005-12-05 Manish Singh <yosh@gimp.org>
538 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
539 to override any user settings.
541 2005-12-05 Matthias Clasen <mclasen@redhat.com>
543 * configure.in: Revert an accidential commit
545 2005-12-05 Matthias Clasen <mclasen@redhat.com>
547 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
549 * tests/Makefile.am (TESTS_ENVIRONMENT):
550 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
551 MALLOC_CHECK_ and MALLOC_PERTURB_
553 * tests/run-collate-tests.sh: Run the collation tests explicitly
554 in en_US locale. (#320463)
556 * glib/gmem.h: Really deprecate GMemChunk.
563 * glib/gstrfuncs.c: Add versioned deprecation docs.
565 2005-12-05 Martyn Russell <martyn@imendio.com>
567 * docs/reference/glib/glib-sections.txt:
568 * glib/gasyncqueue.[ch]:
569 - Added support for sorting async queues by with _push_sorted(),
570 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
573 * tests/asyncqueue-test.c:
574 - Added test case for gasyncqueue.c
576 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
578 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
579 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
580 in bug 323178. moved time consuming logic from g_slice_free() out of
581 the inner loop, so g_slice_free_chain_with_offset() provides a real
582 performance benefit over g_slice_free1() now.
584 * glib/gslice.h: renamed g_slice_free_chain() to
585 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
586 a type-safe macro as suggested in bug 323178.
587 simplified the macro implementation of g_slice_free() and implemented
588 it in a type safe manner for all compliers as suggested by Morten
589 Welinder <mortenw@gnome.org>.
594 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
596 2005-12-05 Matthias Clasen <mclasen@redhat.com>
598 * glib/gasyncqueue.c: Add some docs.
600 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
602 * glib/gatomic.c: In the ia64 implementation, use
603 __sync builtin without _si or _di suffix. (#321229,
604 Stanislav Brabec, patch by Andreas Schwab)
606 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
608 * glib/gslice.h: Remove comma at the end of enum.
610 2005-12-04 Matthias Clasen <mclasen@redhat.com>
612 Handle multiple user names with the same UID better.
613 (#319535, Laszlo Peter)
615 * glib/gutils.c (g_get_any_init_do): When determining user
616 data, first look up $LOGNAME. If the UID doesn't match
617 getuid(), fall back to the current behaviour of looking
618 up the user data based on getuid().
620 2005-12-04 Matthias Clasen <mclasen@redhat.com>
622 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
623 to make gcc warn if a function result is ignored. (#145466,
624 Arjan van de Ven, Alex Larsson)
626 * glib/gmem.h: Add the new attribute to g_realloc and
629 2005-12-03 Matthias Clasen <mclasen@redhat.com>
632 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
635 * glib/gmappedfile.c: Make mapping of empty files
638 * glib/gfileutils.c: Don't fork a new process just to
639 fix the permissions of the created temp file. (#321318,
640 Alexis S. L. Carvalho)
642 2005-12-02 Matthias Clasen <mclasen@redhat.com>
644 * README.in: Add a note about Solaris threads.
647 * configure.in: Remove support for Solaris threads.
648 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
650 2005-12-02 Matthias Clasen <mclasen@redhat.com>
652 * configure.in: Define G_GUINT64_CONSTANT in analogy to
653 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
655 2005-12-02 Matthias Clasen <mclasen@redhat.com>
657 * glib/gslice.c: Win32 portability fixes and C99ism removal,
658 pointed out by Kazuki Iwamoto. (#323052)
660 2005-12-02 Matthias Clasen <mclasen@redhat.com>
662 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
663 the IA64 implementation.
665 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
667 * glib/gslice.c: some naming and type size fixups.
669 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
671 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
672 variant into something less verbose, but digestible for gcc-3.4.
674 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
676 * tests/slice-test.c: extended to perform the benchmarking on the old
677 memchunk code if 'O' is selected.
679 * tests/memchunks.c: new file which contains the old GLib mem chunks
680 implementation with prefix old_mem_chunk_.
682 * tests/Makefile.am: added memchunks.c
684 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
686 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
687 acceptable load distribution of the syscall.
688 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
689 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
691 2005-12-01 Matthias Clasen <mclasen@redhat.com>
693 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
694 double-free base interface members.
696 * tests/Makefile.am: Remove duplicate variable.
698 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
700 * glib/gslice.[hc]: new slice allocator implementation.
702 * tests/slice-test.c: added random slice allocation test.
704 * glib/gthread.[hc]: removed newly added private thread mem API.
706 * glib/gthreadinit.h:
709 * glib/gmem.c: divided glib threading initialisation into three phases,
710 initialisation where private keys and messaging are not available (only
711 needed by gmem.c), initialisation without messaging but private keys
712 available (gslice.c, gmessage.c), and full fledged initialisers that
713 server the rest of glib. initialisation functions got renamed to reflect
714 the limitations of their corresponding phases.
716 * glib/gmem.c: removed memchunk code, defer allocations to
719 * glib/gmem.[hc]: removed g_slice_* skeletons.
721 * glib/glib.symbols: added g_slice_* symbols.
723 * configure.in: check for availability of posix_memalign(3), memalign(3)
726 * glib/Makefile.am: added gslice.[hc].
728 2005-12-01 Tor Lillqvist <tml@novell.com>
730 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
731 trailing slash(es) for non-root folders. stat() fails if non-root
732 folders are specified with trailing slashes. It's too much hassle
733 to demand that callers strip such slashes themselves, especially
734 as it is easy to get it wrong and strip the slash of a root
736 (g_rename): On NT-based Windows, use MoveFileEx() with
737 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
739 2005-11-28 Matthias Clasen <mclasen@redhat.com>
741 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
744 * configure.in: Check whether do { } while (0) works.
746 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
747 G_STMT_END if it works.
749 2005-11-28 Matthias Clasen <mclasen@redhat.com>
751 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
752 * glib/gnode.c (g_node_depth_traverse_level):
753 * glib/gmem.c (g_allocator_new):
754 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
755 (g_hash_table_foreach_remove_or_steal): Silence compiler
758 2005-11-27 Matthias Clasen <mclasen@redhat.com>
760 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
761 Don't read beyond len. (#322520, Christian Persch)
763 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
765 * glib/gdataset.c: access datalist flags via atomic pointer access
766 functions, instead of acquiring the dataset lock. this is faster and
767 also matches the atomic pointer readouts in gdatalistprivate.h and
768 direct pointer modifications required by gobject.c.
770 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
772 * glib/gdataset.c: streamlined initialization code somewhat,
773 removed GData node cache.
775 2005-11-22 Matthias Clasen <mclasen@redhat.com>
777 * glib/ghash.c (g_hash_table_ref):
778 (g_hash_table_unref): Mark these as new API.
780 Avoid double locking in g_intern_string (#322133,
783 * glib/gdataset.c (g_quark_from_string_internal): New
784 internal function which factors out common parts of
785 g_quark_from[_static]_string.
786 (g_quark_from_string, g_quark_from_static_string):
787 Use g_quark_from_string_internal.
788 (g_intern_string, g_intern_static_string): Use
789 g_quark_from_string_internal, and only take the
791 (g_quark_new): Don't store the strings shifted by -1
792 in the g_quarks array.
793 (g_quark_to_string): Adapt to the previous change.
795 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
799 g_hash_table_new_full(): create hash tables with a ref count of 1.
800 g_hash_table_ref(): atomically ref_count+=1
801 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
802 when refcount reaches 0.
803 g_hash_table_destroy(): just destroy keys and values, unref by 1.
804 g_hash_table_insert():
805 g_hash_table_replace(): assert ref_count>0.
808 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
811 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
813 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
814 variables. (#321984, Andrew Paprocki)
816 2005-11-18 Matthias Clasen <mclasen@redhat.com>
818 * configure.in: Bump version
820 2005-11-17 Matthias Clasen <mclasen@redhat.com>
822 * === Released 2.9.0 ===
824 * glib/glib.symbols: Updates
826 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
827 hints to the docs. (Paolo Borelli)
831 * glib/gthread.c (g_thread_foreach): Mark as new api.
833 * README.in: Updates.
837 * glib/gdate.c (g_date_set_time_t):
838 (g_date_set_time_val): Add functions to set a date from
839 a time_t and from a GTimeVal.
840 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
843 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
844 and g_utf8_offset_to_pointer.
846 * glib/gutf8.c (g_utf8_pointer_to_offset)
847 (g_utf8_offset_to_pointer): Handle negative offsets, and use
848 "stutter stepping" for going backwards. (#320638, Larry
853 * glib/gthread.c: const correctness fixes, found
854 by Arjan van de Ven and gcc.
856 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
858 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
859 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
861 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
863 * glib/guniprop.c: Use bit hacks instead when checking a general
864 category value against multiple values.
866 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
868 2005-11-10 Simos Xenitellis <simos@gnome.org>
870 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
872 2005-11-05 Tor Lillqvist <tml@novell.com>
874 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
875 (#320688) Instead set error and return failure.
877 2005-11-07 Matthias Clasen <mclasen@redhat.com>
879 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
880 while holding the context lock. (#320886, Andy Wingo)
882 2005-11-07 Billy Biggs <vektor@dumbterm.net>
884 * tests/refcount/Makefile.am: Add a missing reference to libglib
885 in the LDADD for this test.
887 2005-11-07 Matthias Clasen <mclasen@redhat.com>
889 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
890 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
891 GLib 2.9. (#320755, Luca Ferretti)
893 2005-11-04 Matthias Clasen <mclasen@redhat.com>
895 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
896 use a faster lookup table for g_unichar_get_mirror_char().
897 * glib/gmirroringtable.h: The new table.
898 * glib/gunichartables.h: Remove bidi_mirroring_table.
899 * glib/gen-unicode-tables.pl: Don't generate the mirroring
901 * glib/glib-mirroring-tab/*: Sources for the program
902 which generated gmirroringtable.h.
904 * glib/glist.c: Avoid some code duplication.
906 * glib/gscanner.h: Include gdataset.h. (#320322)
908 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
909 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
910 to help find critical warnings in applications. (#320017,
913 2005-11-02 Tor Lillqvist <tml@novell.com>
915 * glib/glib.symbols: Remove large amount of trailing whitespace
916 from one line. Remove #ifdef G_UNDEFINED from around
917 g_hash_table_new and g_hash_table_new_full symbols, what was the
918 purpose of that? Add the g_slice functions.
920 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
922 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
923 added g_slice_*() API to allocate and cache small bits of memory.
924 an actuall allocator implementation for g_slice_*() is still pending.
926 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
927 changed GRealThread list to use in-structure *next; fields instead
928 of GSList, in order for thread iteration to not depenend on g_slice_*()
930 _g_thread_mem_private_get():
931 _g_thread_mem_private_set(): added accessors for private memory,
932 needed because the ordinary GPrivate implementation relies on GArray
933 and GSList and therefore indirectly on working g_slice_*() allocations.
936 g_thread_foreach(): new public API function to loop over all existing threads.
950 removed GAllocator and free list usages and accompanying locks.
951 use g_slice_*() API to allocate and cache small bits of memory.
953 * glib/ghook.h: removed GMemChunk field from public API.
956 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
958 * glib/gnode.h: deprecate allocator API.
960 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
961 be used for poll rec lists.
963 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
964 g_relation_destroy(): free all tuples from the all_tuples hash table,
965 this effectively maintains the life time track keeping of tuples.
966 g_relation_delete_tuple(): free tuples which are removed from the
967 all_tuples hash table. this fixes a temporary leak that was present
968 in the memchunk code until the destruction of the relation.
970 2005-10-29 Matthias Clasen <mclasen@redhat.com>
972 * tests/convert-test.c: Add some tests for conversions between
973 UTF-8, UCS-4 and UTF-16.
975 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
976 of len == -1, noticed by Morten Welinder.
978 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
980 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
982 2005-10-26 Matthias Clasen <mclasen@redhat.com>
984 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
985 in the error case, and add some documentation. (#319806, Morten
988 2005-10-19 Manish Singh <yosh@gimp.org>
990 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
991 config.h symbol automatically. Fixes bug #313731.
993 2005-10-19 Tor Lillqvist <tml@novell.com>
995 * glib/gwin32.c (g_win32_get_package_installation_directory):
996 Return a g_strdup()ed copy of the value stored in the hash table,
997 so that it can be g_free()d without leaving a dangling pointer in
998 the hash table. (#319232)
1000 2005-10-06 Matthias Clasen <mclasen@redhat.com>
1002 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
1003 all-zero sequences correctly. (#317930, Sebastien Bacher)
1005 2005-10-05 Matthias Clasen <mclasen@redhat.com>
1007 * glib/gmarkup.c (g_markup_escape_text): Doc addition
1009 * Makefile.am: Add Collation tests.
1011 * tests/collate/*: Inputs and expected outputs for collation tests.
1013 * tests/run-collate-tests.sh: Script to run collation tests.
1015 * tests/unicode-collate.c (main): Rework slightly to make
1016 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
1018 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
1020 * docs/reference/glib/tmpl/unicode.sgml:
1021 * glib/gen-unicode-tables.pl:
1023 * glib/gunichartables.h:
1025 * tests/casefold.txt:
1026 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1027 GUnicodeBreakType types. That may break some applications, like
1030 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1032 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1033 docs. (#314393, Matthew F. Barnes)
1035 * glib/glib.symbols:
1036 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1037 they are already declared in glib.h. This doesn't break documented
1038 use of gprintf.h, but should probably be pointed out in the
1039 release notes for 2.10. (#314232, Behdad Esfahbod)
1041 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1043 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1044 complexity optimization, based on a patch from Matthias Clasen.
1046 * tests/patterntest.c: more tests, mostly from matthias.
1048 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1050 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1051 (#316703, Mark Drago)
1053 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1055 * glib/gpattern.c: applied a patch from matthias which checks on the
1056 upper bound of GPatternSpec length to optimize matches.
1059 * tests/patterntest.c: added more match cases.
1061 2005-09-16 Tor Lillqvist <tml@novell.com>
1063 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1064 warning to doc comment that these functions should not be used on
1065 encodings like CP932.
1067 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1069 * tests/keyfile-test.c: Add a test for grup names of length 1.
1071 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1072 of length 1. (#316309)
1074 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1076 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1078 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1080 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1081 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1084 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1086 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1087 (child_watch_helper_thread): Remove some dead code. Closes
1090 2005-09-07 Tor Lillqvist <tml@novell.com>
1092 * glib/Makefile.am: Create also a console version of the
1093 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1094 It's otherwise identical to gspawn-win32-helper.exe, except marked
1095 as a console application (linked without the -mwindows option).
1097 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1098 Drop the dont_wait parameter. Its truth value correlated 100% with
1099 the NULLness of the exit_status parameter anyway, so it's enough
1100 to check whether exit_status is NULL. Invert the sense of the
1101 dont_return_handle parameter and rename it to do_return_handle, to
1102 make the code easier to read by avoiding double negations.
1104 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1105 to do_spawn_with_pipes() accordingly.
1107 (do_spawn_with_pipes): If we have a console, use the console
1108 version of the helper program, otherwise use the GUI one. This
1109 avoids extra console windows opening up in some situations. (In
1110 case a console application uses the GUI gspawn-win32-helper.exe to
1111 spawn another console application we would get a separate console
1112 for the spawned console application).
1114 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1116 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1118 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1119 if the file is too large. (#315275, Kjartan Maraas)
1121 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1122 of read() is signed. (#315273, Kjartan Maraas)
1124 2005-08-31 Tor Lillqvist <tml@novell.com>
1126 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1127 the EXE and each DLL have their own atexit function chains.
1129 #define g_atexit as atexit instead. This means it has a
1130 better chance of doing what the caller wants. For instance,
1131 gtkhtml calls g_atexit() registering a function in gtkhtml
1132 itself. This caused a crash when g_atexit() was implemented as a
1133 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1134 the time the GLib DLL got unloaded.
1136 * glib/gutils.c: #undef the #define mentioned above, to also get a
1137 real g_atexit() into the DLL for backward compatibility. Document
1138 the Windows behaviour of g_atexit(), and document the varying ways
1139 atexit() can behave in the context of dynamically loaded modules
1142 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1144 * glib/glib.symbols:
1146 * glib/gdataset.c: Add string interning functions.
1148 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1150 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1153 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1155 * configure.in: Bump version to 2.9.0
1157 2005-08-25 Tor Lillqvist <tml@novell.com>
1159 Make also the g_spawn*() functions take parameters in the GLib
1160 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1161 API or ABI. Like the other GLib API that was earlier changed to
1162 use UTF-8 on Windows, the names of the functions that take UTF-8
1163 have _utf8 suffixes added by using preprocessor macros in the
1164 header file. The old names are kept for functions with the old
1165 behaviour, taking parameters in the system codepage, for DLL ABI
1168 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1169 g_spawn*() functions.
1171 * glib/gspawn-win32.c: Use wide-char API on NT-based
1172 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1173 system codepage (Win9x) and call the C library _wspawn*() or
1174 spawn*() functions respectvely. Add DLL ABI stability versions
1175 that take parameters in the system codepage.
1177 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1178 wide-char versions of argv and envp, and use wide-char API to
1179 change directory and spawn the program to run. Remove the verbose
1180 debugging output, it was too complex to modify for the wide-char
1181 features. (Just add temporary debugging printouts if needed, no
1182 need to have them permanently in the source.)
1184 * glib/gspawn.c: Corresponding documentation updates.
1186 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1187 symbols as PRIVATE, add the new _utf8-suffixed ones.
1189 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1191 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1192 guint64 otherwise and that can produce warnings about comparison
1193 between signed and unsigned.
1195 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1197 * glib/gutils.c: Fix the crt_externs.h include.
1199 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1201 * NEWS: Fix spelling of my first name.
1203 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1207 * === Released 2.8.1 ===
1211 2005-08-20 Hans Breuer <hans@breuer.org>
1213 * glib/makefile.msc.in : link with ws2_32.lib
1215 2005-08-18 Tor Lillqvist <tml@novell.com>
1217 * configure.in: Check for <sys/wait.h>
1219 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1221 2005-08-18 Ross Burton <ross@burtonini.com>
1224 Optimise single-character insertions.
1229 * tests/string-test.c:
1230 Add tests for new optimisation, and fix a leak.
1232 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1234 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1236 * glib/gutils.c: On Darwin, include crt-externs.h and
1237 define environ using _NSGetEnviron(). (#313731)
1239 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1241 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1242 (g_get_any_init_do): ... to this new function.
1243 (g_get_any_init): Declare as inline.
1244 (g_get_any_init_locked): New inline function, does the locking.
1245 Make use of these two throughout the code.
1247 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1249 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1250 the child process and then simply return. This makes
1251 The "S" option work as documented in g_on_error_query().
1252 (#313125, Matthew F. Barnes)
1254 * glib/gunicode.h: Update the link to Unicode category
1255 values. (#313369, Behnam Esfahbod)
1257 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1258 a little. (#311727, Tristan van Berkom)
1260 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1261 check work on ia64 too, where some symbols ended up
1262 in yet another section.