1 2006-03-02 Marcus Brinkmann <mb@g10code.de>
3 Implement watches for GIOChannels for write file descriptors on
6 * glib/giowin32.c (GIOWin32Channel): Add a new direction field.
7 (read_thread): Initialize direction.
8 (write_thread): New function.
9 (buffer_write): New function.
10 (g_io_win32_prepare): Handle the G_IO_WIN32_FILE_DESC case for the
12 (g_io_win32_fd_write): Call buffer_write() if there is a writer
14 (g_io_win32_fd_close): Set space_avail_event for writer threads.
15 (g_io_win32_fd_create_watch): Create the writer thread if
16 condition is G_IO_OUT.
17 (g_io_channel_win32_make_pollfd): Likewise here.
19 2006-03-09 Matthias Clasen <mclasen@redhat.com>
21 * Makefile.am: Add ChangeLog.pre-2.8 to EXTRA_DIST.
23 2006-03-08 Tor Lillqvist <tml@novell.com>
25 * glibconfig.h.win32.in: Add G_GUINT64_CONSTANT.
27 2006-03-07 Matthias Clasen <mclasen@redhat.com>
29 * configure.in: Bump version
31 * === Released 2.10.1 ===
35 * glib/gutf8.c (_g_charset_get_aliases): Match the prototype
36 in gconvert.c to fix build problems in NetBSD. (#333651,
37 Julio M. Merino Vidal)
39 2006-03-06 Anders Carlsson <andersca@imendio.com>
42 Make .dylib the shared library suffix on darwin.
43 (#322476, Vladimir Panov)
45 2006-03-06 Anders Carlsson <andersca@imendio.com>
48 Skip checking thread flags on Darwin. Fix suggested by Manish Singh.
49 (#314794, Gregor Riepl)
51 2006-03-01 Tor Lillqvist <tml@novell.com>
53 * glib/gutils.c (g_listenv): Separate implementation on Win32: Use
54 the wide character API on NT-based Windows. Return UTF-8 strings.
56 * glib/glib.symbols: Don't mark g_listenv as PRIVATE, as that
57 meant it wasn't present in the import library. PRIVATE is used
58 only for the backwards-compatibility DLL ABI stability hacks.
60 2006-02-24 Matthias Clasen <mclasen@redhat.com>
62 * configure.in: Bump version
64 * === Released 2.10.0 ===
66 * README.in: Mention the slice allocator and GInitiallyUnowned.
70 * configure.in: Bump version to 2.10.0
72 * tests/utf8-pointer.c: Add tests for g_utf8_strlen().
74 * glib/gutf8.c: Fix boundary cases in g_utf8_strlen().
77 2006-02-23 Kjartan Maraas <kmaraas@gnome.org>
79 * tests/completion-test.c: (main): Plug leaks reported by
81 * tests/convert-test.c: (test_iconv_state), (test_one_half):
83 * tests/patterntest.c: (test_compilation): Same
84 * tests/shell-test.c: (do_argv_test): Same.
85 * tests/unicode-caseconv.c: (main): Same.
86 * tests/uri-test.c: (run_to_uri_tests): Same.
89 2006-02-19 Martyn Russell <martyn@imendio.com>
91 * tests/asyncqueue-test.c:
93 * tests/slist-test.c: Updated to test _sort, _sort_with_data,
94 _insert_sorted and _insert_sorted_with_data API.
96 2006-02-18 Matthias Clasen <mclasen@redhat.com>
98 * tests/gobject/Makefile.am: Add paramspec-test
100 * tests/gobject/paramspec-test.c: Some GParamSpec tests.
102 * tests/gobject/gvalue-test.c: Add more tests.
104 2006-02-17 Kang Jeong-Hee <keizie@gmail.com>
106 * glib/gutf8.c (g_utf8_find_prev_char): Correct documentation typo.
108 2006-02-15 Sebastian Wilhelmi <seppi@seppi.de>
110 * glib/gthreadpool.c: Fix deadlock when signalling the thread
111 which freed a thread pool (#331110, Chris Wilson).
113 Tue Feb 14 17:00:43 2006 Tim Janik <timj@imendio.com>
115 * glib/gslice.c: only define _XOPEN_SOURCE to 600 to get at
116 posix_memalign() in case we actually are going to use it, because
117 we detected a compliant implementation (#328997).
119 * configure.in (enable_included_printf): don't include malloc.h when
120 testing for posix_memalign() funcitonality, since this may break the
121 test on some systems (#328997).
123 2006-02-14 Matthias Clasen <mclasen@redhat.com>
125 * glib/gstrfuncs.c: Improve docs.
127 2006-02-11 Matthias Clasen <mclasen@redhat.com>
129 * configure.in: Bump version
131 2006-02-10 Matthias Clasen <mclasen@redhat.com>
133 * === Released 2.9.6 ===
137 2006-01-31 Behdad Esfahbod <behdad@gnome.org>
139 * docs/reference/glib/tmpl/relations.sgml (g_relation_insert): Specify
140 the type of varargs arguments. (bug #317679)
142 2006-01-31 Matthias Clasen <mclasen@redhat.com>
144 * autogen.sh: Touch README and INSTALL here to pacify
145 automake. (#329124, Kjartan Maraas, Tim Janik)
147 Tue Jan 31 16:45:03 2006 Tim Janik <timj@imendio.com>
149 * glib/gdataset.c: carry out all dtalist pointer accesses atomically,
150 some missing cases where pointed out by Sebastian Wilhelmi.
152 Tue Jan 31 12:33:48 2006 Tim Janik <timj@imendio.com>
154 * configure.in: generate build/win32/vs8/Makefile, to fix build breakage.
156 2006-01-30 David Schleef <ds@schleef.org>
158 * glib/grand.c: Update URL
160 2006-01-29 Sven Herzberg <herzi@gnome-de.org>
162 * docs/reference/gobject/tmpl/param_value_types.sgml: tell how to
163 create container classes which are as flexible as a GValue is
165 2006-01-27 Matthias Clasen <mclasen@redhat.com>
167 * configure.in: Bump version
169 * === Released 2.9.5 ===
171 * glib/glib.symbols: Add g_mem_gc_friendly as an exported
176 2006-01-26 Matthias Clasen <mclasen@redhat.com>
178 * glib/gmem.c (g_mem_init_nomessage): Fix C99isms. (#328705,
181 Wed Jan 25 19:16:57 2006 Tim Janik <timj@imendio.com>
183 * fixed buglets reported by Jens Granseuer in #328254.
185 * configure.in: free the memory allocated in posix_memalign() tests.
187 * glib/gslice.c: spelling fixes.
189 Wed Jan 25 16:39:18 2006 Tim Janik <timj@imendio.com>
191 * glib/gslice.c: honour g_mem_gc_friendly settings when freeing
192 slices, make sure g_mem_gc_friendly is properly initialized.
194 * gmem.[hc]: ensure g_mem_gc_friendly is initialized from G_DEBUG upon
195 the first allocation. applied some branching optimizations.
197 * docs/macros.txt: reflected --enable-gc-friendly change and
198 described ENABLE_GC_FRIENDLY_DEFAULT as well as G_DEBUG=gc-friendly.
200 * configure.in: changed --enable-gc-friendly=yes to define
201 ENABLE_GC_FRIENDLY_DEFAULT.
203 * glib/garray.c: changed ENABLE_GC_FRIENDLY macro #ifdef-s to
204 if (G_UNLIKELY (g_mem_gc_friendly)).
207 * glib/ghash.c: removed ENABLE_GC_FRIENDLY code which is now taken
208 care of by g_slice_free1().
210 * tests/slice-test.c: fixed leaks, reported by Kjartan Maraas.
212 Tue Jan 24 17:49:36 2006 Tim Janik <timj@imendio.com>
214 * glib/gslice.c: only use posix_memalign() if it's known to work,
215 revert to memalign() otherwise.
217 * configure.in: check for broken posix_memalign() implementations
220 2006-01-24 Matthias Clasen <mclasen@redhat.com>
222 * tests/unicode-encoding.c: Use UTF-16LE as target encoding
223 on all little-endian systems. (#143380, Marc Moorcroft)
225 2006-01-23 Matthias Clasen <mclasen@redhat.com>
227 * configure.in: Change the shared libary extension for hpux-ia64
228 to so. (#328253, The Written Word)
230 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
232 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
233 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
236 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
238 * glib/gslice.c (slice_config_init): initialize GSlice config from
239 G_SLICE environment variable. we support G_SLICE=always-malloc
240 currently, which forces all g_slice_*() allocations to use the system
244 g_parse_debug_string(): added a note about not using g_malloc() here.
245 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
246 g_slice. contains only guesswork in the WIN32 branch.
248 2006-01-18 Matthias Clasen <mclasen@redhat.com>
252 * === Released 2.9.4 ===
256 2006-01-17 Matthias Clasen <mclasen@redhat.com>
258 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
259 call. (#327421, Morten Welinder)
261 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
263 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
264 G_LOCK. Use the unused_thread_queue lock instead. Change
265 g_thread_pool_thread_proxy such that threads only wait on
266 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
267 due to superfluous tasks. Global tasks wait at most for
268 max-idle-time milliseconds. Make sure, that no task is woken up
269 twice for the same event via a wakeup_serial. This fixes #324228.
271 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
272 invalid NULL into the thread pools. This as well fixes #327290.
274 2006-01-16 Matthias Clasen <mclasen@redhat.com>
276 * configure.in: Bump version
278 * === Released 2.9.3 ===
282 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
283 Rework the strict aliasing fix to not break C++, pointed
284 out by Murray Cumming.
286 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
287 Signal waiting threads, problem noticed by Christian Kellner.
289 2006-01-16 Matthias Clasen <mclasen@redhat.com>
291 Fix bug 326747, Alberto Ruiz:
293 * glib/gconvert.c (make_valid_utf8): Change this function to
294 replace unknown characters by the Unicode replacement character
295 instead of '?', and don't append "(invalid encoding)".
296 (g_filename_display_name, g_filename_display_basename): Document
297 how to determine if the filename was in an invalid encoding.
299 2006-01-14 Matthias Clasen <mclasen@redhat.com>
301 * glib/gtree.c: Replace the simple recursive implementation
302 by a nonrecursive, threaded implementation by Maurizio
305 2006-01-12 Matthias Clasen <mclasen@redhat.com>
307 * tests/asyncqueue-test.c (main): Fix the
308 assert to not always trigger. (#326558,
311 2006-01-11 Matthias Clasen <mclasen@redhat.com>
313 * tests/convert-test.c: Don't test items_read and
314 items_written in error cases where it is not set.
316 * tests/asyncqueue-test.c: Set error to NULL.
318 2006-01-08 Matthias Clasen <mclasen@redhat.com>
320 * glib/gtree.c: Break some long lines.
322 * tests/tree-test.c: Add more tests.
324 * glib/gslice.c: Include config.h first, pointed out
327 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
329 * docs/reference/gobject/tmpl/gtype.sgml: explain that
330 G_TYPE_INSTANCE_GET_CLASS() does behave different during
333 2006-01-06 Matthias Clasen <mclasen@redhat.com>
335 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
336 that we have posix_memalign(). (#323937, Bogdan Nicula)
338 2006-01-05 Matthias Clasen <mclasen@redhat.com>
340 * configure.in: Bump version
342 * === Released 2.9.2 ===
344 2006-01-05 Hans Breuer <hans@breuer.org>
346 * glib/makefile.msc.in : added gslice.obj
348 2006-01-05 Matthias Clasen <mclasen@redhat.com>
352 * glib/gmain.c: Clarify the documentation of source ids
353 a bit. (#325874, Dan Williams)
355 * configure.in: Fix another strict aliasing problem.
357 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
360 2006-01-04 Manish Singh <yosh@gimp.org>
363 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
364 either, since they have the same issue as g_rmdir. (Related to
367 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
369 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
370 order when creating magazines, so we hand out chunks with ascending
373 2006-01-04 Abel Cheung <maddog@linuxhall.org>
375 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
377 2006-01-03 Martyn Russell <martyn@imendio.com>
379 * docs/reference/glib/glib-sections.txt:
381 * glib/gthreadpool.[ch]:
382 - Added new API g_thread_pool_get_idle_time() and
383 g_thread_pool_set_idle_time(). (#324228).
385 * tests/threadpool-test.c:
386 - Updated test case to do thread pool sorting, thread pool with
387 no sorting and a thread pool with idle thread timeouts.
389 2006-01-03 Matthias Clasen <mclasen@redhat.com>
391 * glib/gmain.h: Add new functions here, too.
393 * glib/glib.symbols: Add new functions.
395 2005-12-20 Michael Meeks <michael.meeks@novell.com>
397 * glib/gmain.c (g_main_context_is_owner): new method
398 to determine if the current thread is the owner of the
401 2006-01-02 Matthias Clasen <mclasen@redhat.com>
405 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
406 expanding to rmdir, since rmdir is not declared in a portable
407 system header we can include in gstdio.h. (#325249, Jani Monoses)
409 2006-01-01 Tor Lillqvist <tml@novell.com>
411 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
412 stdout and stderr to unbuffered. Otherwise the giochannel layer
413 will try to read from them regardless whether the
414 g_io_channel_win32_poll() call here has indicated
415 readability or not. (#325310)
417 2005-12-29 Matthias Clasen <mclasen@redhat.com>
419 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
420 in the docs. (#325273, Declan Naughton)
422 2005-12-27 Manish Singh <yosh@gimp.org>
424 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
425 make sure we really override things.
427 2005-12-27 Matthias Clasen <mclasen@redhat.com>
429 Fix #316221, Michal Benes, Stanislav Brabec;
431 * configure.in: Fix a strict aliasing problem in
432 g_static_mutex_get_mutex().
433 * glib/gthread.h: ...and in
434 g_static_mutex_get_mutex_impl_shortcut().
436 * glib/gdatasetprivate.h: Add a cast to silence compiler
437 warnings. (#321978, Andrew Paprocki)
439 Partial fix for bug #323937, Bogdan Nicula.
441 * configure.in: Check for malloc.h
442 * glib/gslice.c: Don't include malloc.h unconditionally.
444 2005-12-27 Manish Singh <yosh@gimp.org>
446 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
447 (#324950, Dan Yefimov)
449 2005-12-26 Matthias Clasen <mclasen@redhat.com>
451 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
454 2005-12-25 Matthias Clasen <mclasen@redhat.com>
456 * AUTHORS: Update my email
458 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
460 2005-12-21 Matthias Clasen <mclasen@redhat.com>
462 * README.in: Add some notes on when 'make check' may fail.
464 2005-12-20 Matthias Clasen <mclasen@redhat.com>
467 * glib/gthreadpool.h:
468 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
469 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
471 * tests/threadpool-test.c: Test this.
473 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
475 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
476 handle errors without depending on gmessage.c which might not be
477 setup when the error occours.
478 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
479 always freeing can be achieved by adjusting the working set time to
480 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
481 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
482 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
483 minimize wastage if small amounts of differently sized structrues are
484 allocated, this does come at a performance cost of roughly 5% though.
485 fixed up block alignment calculation, so it works for varying
486 block sizes. only use strerror() not g_strerror() since the latter
487 depends on working GQuark and GSlice.
488 mem_error(): implemented in terms of fprintf and vfprintf.
490 * tests/slice-color.c: new program to test cache colorization effects.
492 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
493 G_SLICE_CONFIG_WORKING_SET_MSECS.
495 2005-12-17 Matthias Clasen <mclasen@redhat.com>
497 * glib/goption.c (parse_short_option): Set an error in all
498 failure cases. (#324332, Tim-Philipp Müller)
500 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
502 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
503 and g_atomic_pointer_get. Add g_atomic_int_set and
504 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
505 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
506 as functions (additionally to the macros in the header) for the
507 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
509 2005-12-16 Matthias Clasen <mclasen@redhat.com>
511 * glib/gmem.c (g_allocator_new): Don't return a pointer to
512 a const struct, since apps expect to be able to modify it.
513 (#324179, J. Ali Harlow)
515 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
517 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
518 for systems where the initialization of atomic variables requires a
519 write memory barrier.
521 2005-12-09 Matthias Clasen <mclasen@redhat.com>
523 * configure.in: Bump version
525 * === Released 2.9.1 ===
527 2005-12-09 Alexander Larsson <alexl@redhat.com>
531 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
534 2005-12-08 Matthias Clasen <mclasen@redhat.com>
538 2005-12-07 Martyn Russell <martyn@imendio.com>
540 * glib/gasyncqueue.c:
541 - Call g_queue_insert_sorted() instead of duplicating the code.
542 - Call g_queue_sort() instead of duplicating the code.
543 - Invert sort function results to make sure the same sort function
544 gives the same results across glist, gslist, gqueue and
547 * tests/asyncqueue-test.c:
548 - Updated the sort function to reflect the example in the
549 documentation for gasyncqueue.c.
551 2005-12-07 Martyn Russell <martyn@imendio.com>
553 * docs/reference/glib/glib-sections.txt:
554 * docs/reference/glib/tmpl/linked_lists_double.sgml:
555 * docs/reference/glib/tmpl/linked_lists_single.sgml:
558 - Added g_list_insert_sorted_with_data () and
559 g_slist_insert_sorted_with_data ().
560 - Removed the extra check in g_list_sort() and g_slist_sort() for
561 GCompareDataFunc vs. GCompareFunc.
563 2005-12-07 Tor Lillqvist <tml@novell.com>
565 * glib-gettextize.in: Look up prefix at run-time on Win32,
566 assuming the standard directory structure with glib-gettextize in
569 2005-12-06 Matthias Clasen <mclasen@redhat.com>
571 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
572 we need GTK+ 2.8 to compile against GLib 2.10.
574 2005-12-05 Manish Singh <yosh@gimp.org>
576 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
577 to override any user settings.
579 2005-12-05 Matthias Clasen <mclasen@redhat.com>
581 * configure.in: Revert an accidential commit
583 2005-12-05 Matthias Clasen <mclasen@redhat.com>
585 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
587 * tests/Makefile.am (TESTS_ENVIRONMENT):
588 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
589 MALLOC_CHECK_ and MALLOC_PERTURB_
591 * tests/run-collate-tests.sh: Run the collation tests explicitly
592 in en_US locale. (#320463)
594 * glib/gmem.h: Really deprecate GMemChunk.
601 * glib/gstrfuncs.c: Add versioned deprecation docs.
603 2005-12-05 Martyn Russell <martyn@imendio.com>
605 * docs/reference/glib/glib-sections.txt:
606 * glib/gasyncqueue.[ch]:
607 - Added support for sorting async queues by with _push_sorted(),
608 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
611 * tests/asyncqueue-test.c:
612 - Added test case for gasyncqueue.c
614 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
616 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
617 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
618 in bug 323178. moved time consuming logic from g_slice_free() out of
619 the inner loop, so g_slice_free_chain_with_offset() provides a real
620 performance benefit over g_slice_free1() now.
622 * glib/gslice.h: renamed g_slice_free_chain() to
623 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
624 a type-safe macro as suggested in bug 323178.
625 simplified the macro implementation of g_slice_free() and implemented
626 it in a type safe manner for all compliers as suggested by Morten
627 Welinder <mortenw@gnome.org>.
632 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
634 2005-12-05 Matthias Clasen <mclasen@redhat.com>
636 * glib/gasyncqueue.c: Add some docs.
638 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
640 * glib/gatomic.c: In the ia64 implementation, use
641 __sync builtin without _si or _di suffix. (#321229,
642 Stanislav Brabec, patch by Andreas Schwab)
644 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
646 * glib/gslice.h: Remove comma at the end of enum.
648 2005-12-04 Matthias Clasen <mclasen@redhat.com>
650 Handle multiple user names with the same UID better.
651 (#319535, Laszlo Peter)
653 * glib/gutils.c (g_get_any_init_do): When determining user
654 data, first look up $LOGNAME. If the UID doesn't match
655 getuid(), fall back to the current behaviour of looking
656 up the user data based on getuid().
658 2005-12-04 Matthias Clasen <mclasen@redhat.com>
660 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
661 to make gcc warn if a function result is ignored. (#145466,
662 Arjan van de Ven, Alex Larsson)
664 * glib/gmem.h: Add the new attribute to g_realloc and
667 2005-12-03 Matthias Clasen <mclasen@redhat.com>
670 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
673 * glib/gmappedfile.c: Make mapping of empty files
676 * glib/gfileutils.c: Don't fork a new process just to
677 fix the permissions of the created temp file. (#321318,
678 Alexis S. L. Carvalho)
680 2005-12-02 Matthias Clasen <mclasen@redhat.com>
682 * README.in: Add a note about Solaris threads.
685 * configure.in: Remove support for Solaris threads.
686 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
688 2005-12-02 Matthias Clasen <mclasen@redhat.com>
690 * configure.in: Define G_GUINT64_CONSTANT in analogy to
691 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
693 2005-12-02 Matthias Clasen <mclasen@redhat.com>
695 * glib/gslice.c: Win32 portability fixes and C99ism removal,
696 pointed out by Kazuki Iwamoto. (#323052)
698 2005-12-02 Matthias Clasen <mclasen@redhat.com>
700 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
701 the IA64 implementation.
703 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
705 * glib/gslice.c: some naming and type size fixups.
707 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
709 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
710 variant into something less verbose, but digestible for gcc-3.4.
712 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
714 * tests/slice-test.c: extended to perform the benchmarking on the old
715 memchunk code if 'O' is selected.
717 * tests/memchunks.c: new file which contains the old GLib mem chunks
718 implementation with prefix old_mem_chunk_.
720 * tests/Makefile.am: added memchunks.c
722 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
724 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
725 acceptable load distribution of the syscall.
726 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
727 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
729 2005-12-01 Matthias Clasen <mclasen@redhat.com>
731 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
732 double-free base interface members.
734 * tests/Makefile.am: Remove duplicate variable.
736 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
738 * glib/gslice.[hc]: new slice allocator implementation.
740 * tests/slice-test.c: added random slice allocation test.
742 * glib/gthread.[hc]: removed newly added private thread mem API.
744 * glib/gthreadinit.h:
747 * glib/gmem.c: divided glib threading initialisation into three phases,
748 initialisation where private keys and messaging are not available (only
749 needed by gmem.c), initialisation without messaging but private keys
750 available (gslice.c, gmessage.c), and full fledged initialisers that
751 server the rest of glib. initialisation functions got renamed to reflect
752 the limitations of their corresponding phases.
754 * glib/gmem.c: removed memchunk code, defer allocations to
757 * glib/gmem.[hc]: removed g_slice_* skeletons.
759 * glib/glib.symbols: added g_slice_* symbols.
761 * configure.in: check for availability of posix_memalign(3), memalign(3)
764 * glib/Makefile.am: added gslice.[hc].
766 2005-12-01 Tor Lillqvist <tml@novell.com>
768 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
769 trailing slash(es) for non-root folders. stat() fails if non-root
770 folders are specified with trailing slashes. It's too much hassle
771 to demand that callers strip such slashes themselves, especially
772 as it is easy to get it wrong and strip the slash of a root
774 (g_rename): On NT-based Windows, use MoveFileEx() with
775 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
777 2005-11-28 Matthias Clasen <mclasen@redhat.com>
779 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
782 * configure.in: Check whether do { } while (0) works.
784 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
785 G_STMT_END if it works.
787 2005-11-28 Matthias Clasen <mclasen@redhat.com>
789 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
790 * glib/gnode.c (g_node_depth_traverse_level):
791 * glib/gmem.c (g_allocator_new):
792 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
793 (g_hash_table_foreach_remove_or_steal): Silence compiler
796 2005-11-27 Matthias Clasen <mclasen@redhat.com>
798 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
799 Don't read beyond len. (#322520, Christian Persch)
801 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
803 * glib/gdataset.c: access datalist flags via atomic pointer access
804 functions, instead of acquiring the dataset lock. this is faster and
805 also matches the atomic pointer readouts in gdatalistprivate.h and
806 direct pointer modifications required by gobject.c.
808 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
810 * glib/gdataset.c: streamlined initialization code somewhat,
811 removed GData node cache.
813 2005-11-22 Matthias Clasen <mclasen@redhat.com>
815 * glib/ghash.c (g_hash_table_ref):
816 (g_hash_table_unref): Mark these as new API.
818 Avoid double locking in g_intern_string (#322133,
821 * glib/gdataset.c (g_quark_from_string_internal): New
822 internal function which factors out common parts of
823 g_quark_from[_static]_string.
824 (g_quark_from_string, g_quark_from_static_string):
825 Use g_quark_from_string_internal.
826 (g_intern_string, g_intern_static_string): Use
827 g_quark_from_string_internal, and only take the
829 (g_quark_new): Don't store the strings shifted by -1
830 in the g_quarks array.
831 (g_quark_to_string): Adapt to the previous change.
833 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
837 g_hash_table_new_full(): create hash tables with a ref count of 1.
838 g_hash_table_ref(): atomically ref_count+=1
839 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
840 when refcount reaches 0.
841 g_hash_table_destroy(): just destroy keys and values, unref by 1.
842 g_hash_table_insert():
843 g_hash_table_replace(): assert ref_count>0.
846 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
849 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
851 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
852 variables. (#321984, Andrew Paprocki)
854 2005-11-18 Matthias Clasen <mclasen@redhat.com>
856 * configure.in: Bump version
858 2005-11-17 Matthias Clasen <mclasen@redhat.com>
860 * === Released 2.9.0 ===
862 * glib/glib.symbols: Updates
864 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
865 hints to the docs. (Paolo Borelli)
869 * glib/gthread.c (g_thread_foreach): Mark as new api.
871 * README.in: Updates.
875 * glib/gdate.c (g_date_set_time_t):
876 (g_date_set_time_val): Add functions to set a date from
877 a time_t and from a GTimeVal.
878 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
881 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
882 and g_utf8_offset_to_pointer.
884 * glib/gutf8.c (g_utf8_pointer_to_offset)
885 (g_utf8_offset_to_pointer): Handle negative offsets, and use
886 "stutter stepping" for going backwards. (#320638, Larry
891 * glib/gthread.c: const correctness fixes, found
892 by Arjan van de Ven and gcc.
894 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
896 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
897 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
899 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
901 * glib/guniprop.c: Use bit hacks instead when checking a general
902 category value against multiple values.
904 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
906 2005-11-10 Simos Xenitellis <simos@gnome.org>
908 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
910 2005-11-05 Tor Lillqvist <tml@novell.com>
912 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
913 (#320688) Instead set error and return failure.
915 2005-11-07 Matthias Clasen <mclasen@redhat.com>
917 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
918 while holding the context lock. (#320886, Andy Wingo)
920 2005-11-07 Billy Biggs <vektor@dumbterm.net>
922 * tests/refcount/Makefile.am: Add a missing reference to libglib
923 in the LDADD for this test.
925 2005-11-07 Matthias Clasen <mclasen@redhat.com>
927 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
928 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
929 GLib 2.9. (#320755, Luca Ferretti)
931 2005-11-04 Matthias Clasen <mclasen@redhat.com>
933 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
934 use a faster lookup table for g_unichar_get_mirror_char().
935 * glib/gmirroringtable.h: The new table.
936 * glib/gunichartables.h: Remove bidi_mirroring_table.
937 * glib/gen-unicode-tables.pl: Don't generate the mirroring
939 * glib/glib-mirroring-tab/*: Sources for the program
940 which generated gmirroringtable.h.
942 * glib/glist.c: Avoid some code duplication.
944 * glib/gscanner.h: Include gdataset.h. (#320322)
946 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
947 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
948 to help find critical warnings in applications. (#320017,
951 2005-11-02 Tor Lillqvist <tml@novell.com>
953 * glib/glib.symbols: Remove large amount of trailing whitespace
954 from one line. Remove #ifdef G_UNDEFINED from around
955 g_hash_table_new and g_hash_table_new_full symbols, what was the
956 purpose of that? Add the g_slice functions.
958 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
960 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
961 added g_slice_*() API to allocate and cache small bits of memory.
962 an actuall allocator implementation for g_slice_*() is still pending.
964 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
965 changed GRealThread list to use in-structure *next; fields instead
966 of GSList, in order for thread iteration to not depenend on g_slice_*()
968 _g_thread_mem_private_get():
969 _g_thread_mem_private_set(): added accessors for private memory,
970 needed because the ordinary GPrivate implementation relies on GArray
971 and GSList and therefore indirectly on working g_slice_*() allocations.
974 g_thread_foreach(): new public API function to loop over all existing threads.
988 removed GAllocator and free list usages and accompanying locks.
989 use g_slice_*() API to allocate and cache small bits of memory.
991 * glib/ghook.h: removed GMemChunk field from public API.
994 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
996 * glib/gnode.h: deprecate allocator API.
998 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
999 be used for poll rec lists.
1001 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
1002 g_relation_destroy(): free all tuples from the all_tuples hash table,
1003 this effectively maintains the life time track keeping of tuples.
1004 g_relation_delete_tuple(): free tuples which are removed from the
1005 all_tuples hash table. this fixes a temporary leak that was present
1006 in the memchunk code until the destruction of the relation.
1008 2005-10-29 Matthias Clasen <mclasen@redhat.com>
1010 * tests/convert-test.c: Add some tests for conversions between
1011 UTF-8, UCS-4 and UTF-16.
1013 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
1014 of len == -1, noticed by Morten Welinder.
1016 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
1018 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
1020 2005-10-26 Matthias Clasen <mclasen@redhat.com>
1022 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
1023 in the error case, and add some documentation. (#319806, Morten
1026 2005-10-19 Manish Singh <yosh@gimp.org>
1028 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
1029 config.h symbol automatically. Fixes bug #313731.
1031 2005-10-19 Tor Lillqvist <tml@novell.com>
1033 * glib/gwin32.c (g_win32_get_package_installation_directory):
1034 Return a g_strdup()ed copy of the value stored in the hash table,
1035 so that it can be g_free()d without leaving a dangling pointer in
1036 the hash table. (#319232)
1038 2005-10-06 Matthias Clasen <mclasen@redhat.com>
1040 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
1041 all-zero sequences correctly. (#317930, Sebastien Bacher)
1043 2005-10-05 Matthias Clasen <mclasen@redhat.com>
1045 * glib/gmarkup.c (g_markup_escape_text): Doc addition
1047 * Makefile.am: Add Collation tests.
1049 * tests/collate/*: Inputs and expected outputs for collation tests.
1051 * tests/run-collate-tests.sh: Script to run collation tests.
1053 * tests/unicode-collate.c (main): Rework slightly to make
1054 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
1056 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
1058 * docs/reference/glib/tmpl/unicode.sgml:
1059 * glib/gen-unicode-tables.pl:
1061 * glib/gunichartables.h:
1063 * tests/casefold.txt:
1064 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
1065 GUnicodeBreakType types. That may break some applications, like
1068 2005-09-26 Matthias Clasen <mclasen@redhat.com>
1070 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
1071 docs. (#314393, Matthew F. Barnes)
1073 * glib/glib.symbols:
1074 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
1075 they are already declared in glib.h. This doesn't break documented
1076 use of gprintf.h, but should probably be pointed out in the
1077 release notes for 2.10. (#314232, Behdad Esfahbod)
1079 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
1081 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
1082 complexity optimization, based on a patch from Matthias Clasen.
1084 * tests/patterntest.c: more tests, mostly from matthias.
1086 2005-09-20 Matthias Clasen <mclasen@redhat.com>
1088 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
1089 (#316703, Mark Drago)
1091 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
1093 * glib/gpattern.c: applied a patch from matthias which checks on the
1094 upper bound of GPatternSpec length to optimize matches.
1097 * tests/patterntest.c: added more match cases.
1099 2005-09-16 Tor Lillqvist <tml@novell.com>
1101 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
1102 warning to doc comment that these functions should not be used on
1103 encodings like CP932.
1105 2005-09-14 Matthias Clasen <mclasen@redhat.com>
1107 * tests/keyfile-test.c: Add a test for grup names of length 1.
1109 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
1110 of length 1. (#316309)
1112 2005-09-12 Matthias Clasen <mclasen@redhat.com>
1114 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
1116 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
1118 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
1119 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
1122 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
1124 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
1125 (child_watch_helper_thread): Remove some dead code. Closes
1128 2005-09-07 Tor Lillqvist <tml@novell.com>
1130 * glib/Makefile.am: Create also a console version of the
1131 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
1132 It's otherwise identical to gspawn-win32-helper.exe, except marked
1133 as a console application (linked without the -mwindows option).
1135 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
1136 Drop the dont_wait parameter. Its truth value correlated 100% with
1137 the NULLness of the exit_status parameter anyway, so it's enough
1138 to check whether exit_status is NULL. Invert the sense of the
1139 dont_return_handle parameter and rename it to do_return_handle, to
1140 make the code easier to read by avoiding double negations.
1142 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
1143 to do_spawn_with_pipes() accordingly.
1145 (do_spawn_with_pipes): If we have a console, use the console
1146 version of the helper program, otherwise use the GUI one. This
1147 avoids extra console windows opening up in some situations. (In
1148 case a console application uses the GUI gspawn-win32-helper.exe to
1149 spawn another console application we would get a separate console
1150 for the spawned console application).
1152 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
1154 2005-09-05 Matthias Clasen <mclasen@redhat.com>
1156 * glib/gmappedfile.c (g_mapped_file_new): Report an error
1157 if the file is too large. (#315275, Kjartan Maraas)
1159 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
1160 of read() is signed. (#315273, Kjartan Maraas)
1162 2005-08-31 Tor Lillqvist <tml@novell.com>
1164 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
1165 the EXE and each DLL have their own atexit function chains.
1167 #define g_atexit as atexit instead. This means it has a
1168 better chance of doing what the caller wants. For instance,
1169 gtkhtml calls g_atexit() registering a function in gtkhtml
1170 itself. This caused a crash when g_atexit() was implemented as a
1171 function in the GLib DLL. The gtkhtml DLL was already unloaded by
1172 the time the GLib DLL got unloaded.
1174 * glib/gutils.c: #undef the #define mentioned above, to also get a
1175 real g_atexit() into the DLL for backward compatibility. Document
1176 the Windows behaviour of g_atexit(), and document the varying ways
1177 atexit() can behave in the context of dynamically loaded modules
1180 2005-08-31 Matthias Clasen <mclasen@redhat.com>
1182 * glib/glib.symbols:
1184 * glib/gdataset.c: Add string interning functions.
1186 2005-08-28 Matthias Clasen <mclasen@redhat.com>
1188 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
1191 2005-08-26 Matthias Clasen <mclasen@redhat.com>
1193 * configure.in: Bump version to 2.9.0
1195 2005-08-25 Tor Lillqvist <tml@novell.com>
1197 Make also the g_spawn*() functions take parameters in the GLib
1198 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
1199 API or ABI. Like the other GLib API that was earlier changed to
1200 use UTF-8 on Windows, the names of the functions that take UTF-8
1201 have _utf8 suffixes added by using preprocessor macros in the
1202 header file. The old names are kept for functions with the old
1203 behaviour, taking parameters in the system codepage, for DLL ABI
1206 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
1207 g_spawn*() functions.
1209 * glib/gspawn-win32.c: Use wide-char API on NT-based
1210 Windows. Convert parameters from UTF-8 to wide chars (NT) or
1211 system codepage (Win9x) and call the C library _wspawn*() or
1212 spawn*() functions respectvely. Add DLL ABI stability versions
1213 that take parameters in the system codepage.
1215 * glib/gspawn-win32-helper.c: On NT-based Windows use the
1216 wide-char versions of argv and envp, and use wide-char API to
1217 change directory and spawn the program to run. Remove the verbose
1218 debugging output, it was too complex to modify for the wide-char
1219 features. (Just add temporary debugging printouts if needed, no
1220 need to have them permanently in the source.)
1222 * glib/gspawn.c: Corresponding documentation updates.
1224 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
1225 symbols as PRIVATE, add the new _utf8-suffixed ones.
1227 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1229 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1230 guint64 otherwise and that can produce warnings about comparison
1231 between signed and unsigned.
1233 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1235 * glib/gutils.c: Fix the crt_externs.h include.
1237 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1239 * NEWS: Fix spelling of my first name.
1241 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1245 * === Released 2.8.1 ===
1249 2005-08-20 Hans Breuer <hans@breuer.org>
1251 * glib/makefile.msc.in : link with ws2_32.lib
1253 2005-08-18 Tor Lillqvist <tml@novell.com>
1255 * configure.in: Check for <sys/wait.h>
1257 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1259 2005-08-18 Ross Burton <ross@burtonini.com>
1262 Optimise single-character insertions.
1267 * tests/string-test.c:
1268 Add tests for new optimisation, and fix a leak.
1270 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1272 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1274 * glib/gutils.c: On Darwin, include crt-externs.h and
1275 define environ using _NSGetEnviron(). (#313731)
1277 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1279 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1280 (g_get_any_init_do): ... to this new function.
1281 (g_get_any_init): Declare as inline.
1282 (g_get_any_init_locked): New inline function, does the locking.
1283 Make use of these two throughout the code.
1285 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1287 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1288 the child process and then simply return. This makes
1289 The "S" option work as documented in g_on_error_query().
1290 (#313125, Matthew F. Barnes)
1292 * glib/gunicode.h: Update the link to Unicode category
1293 values. (#313369, Behnam Esfahbod)
1295 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1296 a little. (#311727, Tristan van Berkom)
1298 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1299 check work on ia64 too, where some symbols ended up
1300 in yet another section.