1 2006-01-18 Matthias Clasen <mclasen@redhat.com>
3 * === Released 2.9.4 ===
7 2006-01-17 Matthias Clasen <mclasen@redhat.com>
9 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
10 call. (#327421, Morten Welinder)
12 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
14 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
15 G_LOCK. Use the unused_thread_queue lock instead. Change
16 g_thread_pool_thread_proxy such that threads only wait on
17 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
18 due to superfluous tasks. Global tasks wait at most for
19 max-idle-time milliseconds. Make sure, that no task is woken up
20 twice for the same event via a wakeup_serial. This fixes #324228.
22 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
23 invalid NULL into the thread pools. This as well fixes #327290.
25 2006-01-16 Matthias Clasen <mclasen@redhat.com>
27 * configure.in: Bump version
29 * === Released 2.9.3 ===
33 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
34 Rework the strict aliasing fix to not break C++, pointed
35 out by Murray Cumming.
37 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
38 Signal waiting threads, problem noticed by Christian Kellner.
40 2006-01-16 Matthias Clasen <mclasen@redhat.com>
42 Fix bug 326747, Alberto Ruiz:
44 * glib/gconvert.c (make_valid_utf8): Change this function to
45 replace unknown characters by the Unicode replacement character
46 instead of '?', and don't append "(invalid encoding)".
47 (g_filename_display_name, g_filename_display_basename): Document
48 how to determine if the filename was in an invalid encoding.
50 2006-01-14 Matthias Clasen <mclasen@redhat.com>
52 * glib/gtree.c: Replace the simple recursive implementation
53 by a nonrecursive, threaded implementation by Maurizio
56 2006-01-12 Matthias Clasen <mclasen@redhat.com>
58 * tests/asyncqueue-test.c (main): Fix the
59 assert to not always trigger. (#326558,
62 2006-01-11 Matthias Clasen <mclasen@redhat.com>
64 * tests/convert-test.c: Don't test items_read and
65 items_written in error cases where it is not set.
67 * tests/asyncqueue-test.c: Set error to NULL.
69 2006-01-08 Matthias Clasen <mclasen@redhat.com>
71 * glib/gtree.c: Break some long lines.
73 * tests/tree-test.c: Add more tests.
75 * glib/gslice.c: Include config.h first, pointed out
78 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
80 * docs/reference/gobject/tmpl/gtype.sgml: explain that
81 G_TYPE_INSTANCE_GET_CLASS() does behave different during
84 2006-01-06 Matthias Clasen <mclasen@redhat.com>
86 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
87 that we have posix_memalign(). (#323937, Bogdan Nicula)
89 2006-01-05 Matthias Clasen <mclasen@redhat.com>
91 * configure.in: Bump version
93 * === Released 2.9.2 ===
95 2006-01-05 Hans Breuer <hans@breuer.org>
97 * glib/makefile.msc.in : added gslice.obj
99 2006-01-05 Matthias Clasen <mclasen@redhat.com>
103 * glib/gmain.c: Clarify the documentation of source ids
104 a bit. (#325874, Dan Williams)
106 * configure.in: Fix another strict aliasing problem.
108 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
111 2006-01-04 Manish Singh <yosh@gimp.org>
114 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
115 either, since they have the same issue as g_rmdir. (Related to
118 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
120 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
121 order when creating magazines, so we hand out chunks with ascending
124 2006-01-04 Abel Cheung <maddog@linuxhall.org>
126 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
128 2006-01-03 Martyn Russell <martyn@imendio.com>
130 * docs/reference/glib/glib-sections.txt:
132 * glib/gthreadpool.[ch]:
133 - Added new API g_thread_pool_get_idle_time() and
134 g_thread_pool_set_idle_time(). (#324228).
136 * tests/threadpool-test.c:
137 - Updated test case to do thread pool sorting, thread pool with
138 no sorting and a thread pool with idle thread timeouts.
140 2006-01-03 Matthias Clasen <mclasen@redhat.com>
142 * glib/gmain.h: Add new functions here, too.
144 * glib/glib.symbols: Add new functions.
146 2005-12-20 Michael Meeks <michael.meeks@novell.com>
148 * glib/gmain.c (g_main_context_is_owner): new method
149 to determine if the current thread is the owner of the
152 2006-01-02 Matthias Clasen <mclasen@redhat.com>
156 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
157 expanding to rmdir, since rmdir is not declared in a portable
158 system header we can include in gstdio.h. (#325249, Jani Monoses)
160 2006-01-01 Tor Lillqvist <tml@novell.com>
162 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
163 stdout and stderr to unbuffered. Otherwise the giochannel layer
164 will try to read from them regardless whether the
165 g_io_channel_win32_poll() call here has indicated
166 readability or not. (#325310)
168 2005-12-29 Matthias Clasen <mclasen@redhat.com>
170 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
171 in the docs. (#325273, Declan Naughton)
173 2005-12-27 Manish Singh <yosh@gimp.org>
175 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
176 make sure we really override things.
178 2005-12-27 Matthias Clasen <mclasen@redhat.com>
180 Fix #316221, Michal Benes, Stanislav Brabec;
182 * configure.in: Fix a strict aliasing problem in
183 g_static_mutex_get_mutex().
184 * glib/gthread.h: ...and in
185 g_static_mutex_get_mutex_impl_shortcut().
187 * glib/gdatasetprivate.h: Add a cast to silence compiler
188 warnings. (#321978, Andrew Paprocki)
190 Partial fix for bug #323937, Bogdan Nicula.
192 * configure.in: Check for malloc.h
193 * glib/gslice.c: Don't include malloc.h unconditionally.
195 2005-12-27 Manish Singh <yosh@gimp.org>
197 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
198 (#324950, Dan Yefimov)
200 2005-12-26 Matthias Clasen <mclasen@redhat.com>
202 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
205 2005-12-25 Matthias Clasen <mclasen@redhat.com>
207 * AUTHORS: Update my email
209 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
211 2005-12-21 Matthias Clasen <mclasen@redhat.com>
213 * README.in: Add some notes on when 'make check' may fail.
215 2005-12-20 Matthias Clasen <mclasen@redhat.com>
218 * glib/gthreadpool.h:
219 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
220 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
222 * tests/threadpool-test.c: Test this.
224 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
226 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
227 handle errors without depending on gmessage.c which might not be
228 setup when the error occours.
229 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
230 always freeing can be achieved by adjusting the working set time to
231 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
232 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
233 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
234 minimize wastage if small amounts of differently sized structrues are
235 allocated, this does come at a performance cost of roughly 5% though.
236 fixed up block alignment calculation, so it works for varying
237 block sizes. only use strerror() not g_strerror() since the latter
238 depends on working GQuark and GSlice.
239 mem_error(): implemented in terms of fprintf and vfprintf.
241 * tests/slice-color.c: new program to test cache colorization effects.
243 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
244 G_SLICE_CONFIG_WORKING_SET_MSECS.
246 2005-12-17 Matthias Clasen <mclasen@redhat.com>
248 * glib/goption.c (parse_short_option): Set an error in all
249 failure cases. (#324332, Tim-Philipp Müller)
251 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
253 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
254 and g_atomic_pointer_get. Add g_atomic_int_set and
255 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
256 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
257 as functions (additionally to the macros in the header) for the
258 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
260 2005-12-16 Matthias Clasen <mclasen@redhat.com>
262 * glib/gmem.c (g_allocator_new): Don't return a pointer to
263 a const struct, since apps expect to be able to modify it.
264 (#324179, J. Ali Harlow)
266 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
268 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
269 for systems where the initialization of atomic variables requires a
270 write memory barrier.
272 2005-12-09 Matthias Clasen <mclasen@redhat.com>
274 * configure.in: Bump version
276 * === Released 2.9.1 ===
278 2005-12-09 Alexander Larsson <alexl@redhat.com>
282 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
285 2005-12-08 Matthias Clasen <mclasen@redhat.com>
289 2005-12-07 Martyn Russell <martyn@imendio.com>
291 * glib/gasyncqueue.c:
292 - Call g_queue_insert_sorted() instead of duplicating the code.
293 - Call g_queue_sort() instead of duplicating the code.
294 - Invert sort function results to make sure the same sort function
295 gives the same results across glist, gslist, gqueue and
298 * tests/asyncqueue-test.c:
299 - Updated the sort function to reflect the example in the
300 documentation for gasyncqueue.c.
302 2005-12-07 Martyn Russell <martyn@imendio.com>
304 * docs/reference/glib/glib-sections.txt:
305 * docs/reference/glib/tmpl/linked_lists_double.sgml:
306 * docs/reference/glib/tmpl/linked_lists_single.sgml:
309 - Added g_list_insert_sorted_with_data () and
310 g_slist_insert_sorted_with_data ().
311 - Removed the extra check in g_list_sort() and g_slist_sort() for
312 GCompareDataFunc vs. GCompareFunc.
314 2005-12-07 Tor Lillqvist <tml@novell.com>
316 * glib-gettextize.in: Look up prefix at run-time on Win32,
317 assuming the standard directory structure with glib-gettextize in
320 2005-12-06 Matthias Clasen <mclasen@redhat.com>
322 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
323 we need GTK+ 2.8 to compile against GLib 2.10.
325 2005-12-05 Manish Singh <yosh@gimp.org>
327 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
328 to override any user settings.
330 2005-12-05 Matthias Clasen <mclasen@redhat.com>
332 * configure.in: Revert an accidential commit
334 2005-12-05 Matthias Clasen <mclasen@redhat.com>
336 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
338 * tests/Makefile.am (TESTS_ENVIRONMENT):
339 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
340 MALLOC_CHECK_ and MALLOC_PERTURB_
342 * tests/run-collate-tests.sh: Run the collation tests explicitly
343 in en_US locale. (#320463)
345 * glib/gmem.h: Really deprecate GMemChunk.
352 * glib/gstrfuncs.c: Add versioned deprecation docs.
354 2005-12-05 Martyn Russell <martyn@imendio.com>
356 * docs/reference/glib/glib-sections.txt:
357 * glib/gasyncqueue.[ch]:
358 - Added support for sorting async queues by with _push_sorted(),
359 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
362 * tests/asyncqueue-test.c:
363 - Added test case for gasyncqueue.c
365 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
367 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
368 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
369 in bug 323178. moved time consuming logic from g_slice_free() out of
370 the inner loop, so g_slice_free_chain_with_offset() provides a real
371 performance benefit over g_slice_free1() now.
373 * glib/gslice.h: renamed g_slice_free_chain() to
374 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
375 a type-safe macro as suggested in bug 323178.
376 simplified the macro implementation of g_slice_free() and implemented
377 it in a type safe manner for all compliers as suggested by Morten
378 Welinder <mortenw@gnome.org>.
383 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
385 2005-12-05 Matthias Clasen <mclasen@redhat.com>
387 * glib/gasyncqueue.c: Add some docs.
389 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
391 * glib/gatomic.c: In the ia64 implementation, use
392 __sync builtin without _si or _di suffix. (#321229,
393 Stanislav Brabec, patch by Andreas Schwab)
395 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
397 * glib/gslice.h: Remove comma at the end of enum.
399 2005-12-04 Matthias Clasen <mclasen@redhat.com>
401 Handle multiple user names with the same UID better.
402 (#319535, Laszlo Peter)
404 * glib/gutils.c (g_get_any_init_do): When determining user
405 data, first look up $LOGNAME. If the UID doesn't match
406 getuid(), fall back to the current behaviour of looking
407 up the user data based on getuid().
409 2005-12-04 Matthias Clasen <mclasen@redhat.com>
411 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
412 to make gcc warn if a function result is ignored. (#145466,
413 Arjan van de Ven, Alex Larsson)
415 * glib/gmem.h: Add the new attribute to g_realloc and
418 2005-12-03 Matthias Clasen <mclasen@redhat.com>
421 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
424 * glib/gmappedfile.c: Make mapping of empty files
427 * glib/gfileutils.c: Don't fork a new process just to
428 fix the permissions of the created temp file. (#321318,
429 Alexis S. L. Carvalho)
431 2005-12-02 Matthias Clasen <mclasen@redhat.com>
433 * README.in: Add a note about Solaris threads.
436 * configure.in: Remove support for Solaris threads.
437 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
439 2005-12-02 Matthias Clasen <mclasen@redhat.com>
441 * configure.in: Define G_GUINT64_CONSTANT in analogy to
442 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
444 2005-12-02 Matthias Clasen <mclasen@redhat.com>
446 * glib/gslice.c: Win32 portability fixes and C99ism removal,
447 pointed out by Kazuki Iwamoto. (#323052)
449 2005-12-02 Matthias Clasen <mclasen@redhat.com>
451 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
452 the IA64 implementation.
454 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
456 * glib/gslice.c: some naming and type size fixups.
458 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
460 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
461 variant into something less verbose, but digestible for gcc-3.4.
463 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
465 * tests/slice-test.c: extended to perform the benchmarking on the old
466 memchunk code if 'O' is selected.
468 * tests/memchunks.c: new file which contains the old GLib mem chunks
469 implementation with prefix old_mem_chunk_.
471 * tests/Makefile.am: added memchunks.c
473 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
475 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
476 acceptable load distribution of the syscall.
477 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
478 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
480 2005-12-01 Matthias Clasen <mclasen@redhat.com>
482 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
483 double-free base interface members.
485 * tests/Makefile.am: Remove duplicate variable.
487 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
489 * glib/gslice.[hc]: new slice allocator implementation.
491 * tests/slice-test.c: added random slice allocation test.
493 * glib/gthread.[hc]: removed newly added private thread mem API.
495 * glib/gthreadinit.h:
498 * glib/gmem.c: divided glib threading initialisation into three phases,
499 initialisation where private keys and messaging are not available (only
500 needed by gmem.c), initialisation without messaging but private keys
501 available (gslice.c, gmessage.c), and full fledged initialisers that
502 server the rest of glib. initialisation functions got renamed to reflect
503 the limitations of their corresponding phases.
505 * glib/gmem.c: removed memchunk code, defer allocations to
508 * glib/gmem.[hc]: removed g_slice_* skeletons.
510 * glib/glib.symbols: added g_slice_* symbols.
512 * configure.in: check for availability of posix_memalign(3), memalign(3)
515 * glib/Makefile.am: added gslice.[hc].
517 2005-12-01 Tor Lillqvist <tml@novell.com>
519 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
520 trailing slash(es) for non-root folders. stat() fails if non-root
521 folders are specified with trailing slashes. It's too much hassle
522 to demand that callers strip such slashes themselves, especially
523 as it is easy to get it wrong and strip the slash of a root
525 (g_rename): On NT-based Windows, use MoveFileEx() with
526 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
528 2005-11-28 Matthias Clasen <mclasen@redhat.com>
530 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
533 * configure.in: Check whether do { } while (0) works.
535 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
536 G_STMT_END if it works.
538 2005-11-28 Matthias Clasen <mclasen@redhat.com>
540 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
541 * glib/gnode.c (g_node_depth_traverse_level):
542 * glib/gmem.c (g_allocator_new):
543 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
544 (g_hash_table_foreach_remove_or_steal): Silence compiler
547 2005-11-27 Matthias Clasen <mclasen@redhat.com>
549 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
550 Don't read beyond len. (#322520, Christian Persch)
552 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
554 * glib/gdataset.c: access datalist flags via atomic pointer access
555 functions, instead of acquiring the dataset lock. this is faster and
556 also matches the atomic pointer readouts in gdatalistprivate.h and
557 direct pointer modifications required by gobject.c.
559 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
561 * glib/gdataset.c: streamlined initialization code somewhat,
562 removed GData node cache.
564 2005-11-22 Matthias Clasen <mclasen@redhat.com>
566 * glib/ghash.c (g_hash_table_ref):
567 (g_hash_table_unref): Mark these as new API.
569 Avoid double locking in g_intern_string (#322133,
572 * glib/gdataset.c (g_quark_from_string_internal): New
573 internal function which factors out common parts of
574 g_quark_from[_static]_string.
575 (g_quark_from_string, g_quark_from_static_string):
576 Use g_quark_from_string_internal.
577 (g_intern_string, g_intern_static_string): Use
578 g_quark_from_string_internal, and only take the
580 (g_quark_new): Don't store the strings shifted by -1
581 in the g_quarks array.
582 (g_quark_to_string): Adapt to the previous change.
584 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
588 g_hash_table_new_full(): create hash tables with a ref count of 1.
589 g_hash_table_ref(): atomically ref_count+=1
590 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
591 when refcount reaches 0.
592 g_hash_table_destroy(): just destroy keys and values, unref by 1.
593 g_hash_table_insert():
594 g_hash_table_replace(): assert ref_count>0.
597 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
600 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
602 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
603 variables. (#321984, Andrew Paprocki)
605 2005-11-18 Matthias Clasen <mclasen@redhat.com>
607 * configure.in: Bump version
609 2005-11-17 Matthias Clasen <mclasen@redhat.com>
611 * === Released 2.9.0 ===
613 * glib/glib.symbols: Updates
615 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
616 hints to the docs. (Paolo Borelli)
620 * glib/gthread.c (g_thread_foreach): Mark as new api.
622 * README.in: Updates.
626 * glib/gdate.c (g_date_set_time_t):
627 (g_date_set_time_val): Add functions to set a date from
628 a time_t and from a GTimeVal.
629 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
632 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
633 and g_utf8_offset_to_pointer.
635 * glib/gutf8.c (g_utf8_pointer_to_offset)
636 (g_utf8_offset_to_pointer): Handle negative offsets, and use
637 "stutter stepping" for going backwards. (#320638, Larry
642 * glib/gthread.c: const correctness fixes, found
643 by Arjan van de Ven and gcc.
645 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
647 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
648 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
650 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
652 * glib/guniprop.c: Use bit hacks instead when checking a general
653 category value against multiple values.
655 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
657 2005-11-10 Simos Xenitellis <simos@gnome.org>
659 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
661 2005-11-05 Tor Lillqvist <tml@novell.com>
663 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
664 (#320688) Instead set error and return failure.
666 2005-11-07 Matthias Clasen <mclasen@redhat.com>
668 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
669 while holding the context lock. (#320886, Andy Wingo)
671 2005-11-07 Billy Biggs <vektor@dumbterm.net>
673 * tests/refcount/Makefile.am: Add a missing reference to libglib
674 in the LDADD for this test.
676 2005-11-07 Matthias Clasen <mclasen@redhat.com>
678 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
679 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
680 GLib 2.9. (#320755, Luca Ferretti)
682 2005-11-04 Matthias Clasen <mclasen@redhat.com>
684 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
685 use a faster lookup table for g_unichar_get_mirror_char().
686 * glib/gmirroringtable.h: The new table.
687 * glib/gunichartables.h: Remove bidi_mirroring_table.
688 * glib/gen-unicode-tables.pl: Don't generate the mirroring
690 * glib/glib-mirroring-tab/*: Sources for the program
691 which generated gmirroringtable.h.
693 * glib/glist.c: Avoid some code duplication.
695 * glib/gscanner.h: Include gdataset.h. (#320322)
697 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
698 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
699 to help find critical warnings in applications. (#320017,
702 2005-11-02 Tor Lillqvist <tml@novell.com>
704 * glib/glib.symbols: Remove large amount of trailing whitespace
705 from one line. Remove #ifdef G_UNDEFINED from around
706 g_hash_table_new and g_hash_table_new_full symbols, what was the
707 purpose of that? Add the g_slice functions.
709 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
711 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
712 added g_slice_*() API to allocate and cache small bits of memory.
713 an actuall allocator implementation for g_slice_*() is still pending.
715 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
716 changed GRealThread list to use in-structure *next; fields instead
717 of GSList, in order for thread iteration to not depenend on g_slice_*()
719 _g_thread_mem_private_get():
720 _g_thread_mem_private_set(): added accessors for private memory,
721 needed because the ordinary GPrivate implementation relies on GArray
722 and GSList and therefore indirectly on working g_slice_*() allocations.
725 g_thread_foreach(): new public API function to loop over all existing threads.
739 removed GAllocator and free list usages and accompanying locks.
740 use g_slice_*() API to allocate and cache small bits of memory.
742 * glib/ghook.h: removed GMemChunk field from public API.
745 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
747 * glib/gnode.h: deprecate allocator API.
749 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
750 be used for poll rec lists.
752 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
753 g_relation_destroy(): free all tuples from the all_tuples hash table,
754 this effectively maintains the life time track keeping of tuples.
755 g_relation_delete_tuple(): free tuples which are removed from the
756 all_tuples hash table. this fixes a temporary leak that was present
757 in the memchunk code until the destruction of the relation.
759 2005-10-29 Matthias Clasen <mclasen@redhat.com>
761 * tests/convert-test.c: Add some tests for conversions between
762 UTF-8, UCS-4 and UTF-16.
764 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
765 of len == -1, noticed by Morten Welinder.
767 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
769 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
771 2005-10-26 Matthias Clasen <mclasen@redhat.com>
773 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
774 in the error case, and add some documentation. (#319806, Morten
777 2005-10-19 Manish Singh <yosh@gimp.org>
779 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
780 config.h symbol automatically. Fixes bug #313731.
782 2005-10-19 Tor Lillqvist <tml@novell.com>
784 * glib/gwin32.c (g_win32_get_package_installation_directory):
785 Return a g_strdup()ed copy of the value stored in the hash table,
786 so that it can be g_free()d without leaving a dangling pointer in
787 the hash table. (#319232)
789 2005-10-06 Matthias Clasen <mclasen@redhat.com>
791 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
792 all-zero sequences correctly. (#317930, Sebastien Bacher)
794 2005-10-05 Matthias Clasen <mclasen@redhat.com>
796 * glib/gmarkup.c (g_markup_escape_text): Doc addition
798 * Makefile.am: Add Collation tests.
800 * tests/collate/*: Inputs and expected outputs for collation tests.
802 * tests/run-collate-tests.sh: Script to run collation tests.
804 * tests/unicode-collate.c (main): Rework slightly to make
805 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
807 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
809 * docs/reference/glib/tmpl/unicode.sgml:
810 * glib/gen-unicode-tables.pl:
812 * glib/gunichartables.h:
814 * tests/casefold.txt:
815 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
816 GUnicodeBreakType types. That may break some applications, like
819 2005-09-26 Matthias Clasen <mclasen@redhat.com>
821 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
822 docs. (#314393, Matthew F. Barnes)
825 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
826 they are already declared in glib.h. This doesn't break documented
827 use of gprintf.h, but should probably be pointed out in the
828 release notes for 2.10. (#314232, Behdad Esfahbod)
830 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
832 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
833 complexity optimization, based on a patch from Matthias Clasen.
835 * tests/patterntest.c: more tests, mostly from matthias.
837 2005-09-20 Matthias Clasen <mclasen@redhat.com>
839 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
840 (#316703, Mark Drago)
842 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
844 * glib/gpattern.c: applied a patch from matthias which checks on the
845 upper bound of GPatternSpec length to optimize matches.
848 * tests/patterntest.c: added more match cases.
850 2005-09-16 Tor Lillqvist <tml@novell.com>
852 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
853 warning to doc comment that these functions should not be used on
854 encodings like CP932.
856 2005-09-14 Matthias Clasen <mclasen@redhat.com>
858 * tests/keyfile-test.c: Add a test for grup names of length 1.
860 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
861 of length 1. (#316309)
863 2005-09-12 Matthias Clasen <mclasen@redhat.com>
865 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
867 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
869 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
870 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
873 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
875 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
876 (child_watch_helper_thread): Remove some dead code. Closes
879 2005-09-07 Tor Lillqvist <tml@novell.com>
881 * glib/Makefile.am: Create also a console version of the
882 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
883 It's otherwise identical to gspawn-win32-helper.exe, except marked
884 as a console application (linked without the -mwindows option).
886 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
887 Drop the dont_wait parameter. Its truth value correlated 100% with
888 the NULLness of the exit_status parameter anyway, so it's enough
889 to check whether exit_status is NULL. Invert the sense of the
890 dont_return_handle parameter and rename it to do_return_handle, to
891 make the code easier to read by avoiding double negations.
893 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
894 to do_spawn_with_pipes() accordingly.
896 (do_spawn_with_pipes): If we have a console, use the console
897 version of the helper program, otherwise use the GUI one. This
898 avoids extra console windows opening up in some situations. (In
899 case a console application uses the GUI gspawn-win32-helper.exe to
900 spawn another console application we would get a separate console
901 for the spawned console application).
903 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
905 2005-09-05 Matthias Clasen <mclasen@redhat.com>
907 * glib/gmappedfile.c (g_mapped_file_new): Report an error
908 if the file is too large. (#315275, Kjartan Maraas)
910 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
911 of read() is signed. (#315273, Kjartan Maraas)
913 2005-08-31 Tor Lillqvist <tml@novell.com>
915 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
916 the EXE and each DLL have their own atexit function chains.
918 #define g_atexit as atexit instead. This means it has a
919 better chance of doing what the caller wants. For instance,
920 gtkhtml calls g_atexit() registering a function in gtkhtml
921 itself. This caused a crash when g_atexit() was implemented as a
922 function in the GLib DLL. The gtkhtml DLL was already unloaded by
923 the time the GLib DLL got unloaded.
925 * glib/gutils.c: #undef the #define mentioned above, to also get a
926 real g_atexit() into the DLL for backward compatibility. Document
927 the Windows behaviour of g_atexit(), and document the varying ways
928 atexit() can behave in the context of dynamically loaded modules
931 2005-08-31 Matthias Clasen <mclasen@redhat.com>
935 * glib/gdataset.c: Add string interning functions.
937 2005-08-28 Matthias Clasen <mclasen@redhat.com>
939 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
942 2005-08-26 Matthias Clasen <mclasen@redhat.com>
944 * configure.in: Bump version to 2.9.0
946 2005-08-25 Tor Lillqvist <tml@novell.com>
948 Make also the g_spawn*() functions take parameters in the GLib
949 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
950 API or ABI. Like the other GLib API that was earlier changed to
951 use UTF-8 on Windows, the names of the functions that take UTF-8
952 have _utf8 suffixes added by using preprocessor macros in the
953 header file. The old names are kept for functions with the old
954 behaviour, taking parameters in the system codepage, for DLL ABI
957 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
958 g_spawn*() functions.
960 * glib/gspawn-win32.c: Use wide-char API on NT-based
961 Windows. Convert parameters from UTF-8 to wide chars (NT) or
962 system codepage (Win9x) and call the C library _wspawn*() or
963 spawn*() functions respectvely. Add DLL ABI stability versions
964 that take parameters in the system codepage.
966 * glib/gspawn-win32-helper.c: On NT-based Windows use the
967 wide-char versions of argv and envp, and use wide-char API to
968 change directory and spawn the program to run. Remove the verbose
969 debugging output, it was too complex to modify for the wide-char
970 features. (Just add temporary debugging printouts if needed, no
971 need to have them permanently in the source.)
973 * glib/gspawn.c: Corresponding documentation updates.
975 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
976 symbols as PRIVATE, add the new _utf8-suffixed ones.
978 2005-08-24 Stepan Kasal <kasal@ucw.cz>
980 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
981 guint64 otherwise and that can produce warnings about comparison
982 between signed and unsigned.
984 2005-08-23 Matthias Clasen <mclasen@redhat.com>
986 * glib/gutils.c: Fix the crt_externs.h include.
988 2005-08-23 Stepan Kasal <kasal@ucw.cz>
990 * NEWS: Fix spelling of my first name.
992 2005-08-23 Matthias Clasen <mclasen@redhat.com>
996 * === Released 2.8.1 ===
1000 2005-08-20 Hans Breuer <hans@breuer.org>
1002 * glib/makefile.msc.in : link with ws2_32.lib
1004 2005-08-18 Tor Lillqvist <tml@novell.com>
1006 * configure.in: Check for <sys/wait.h>
1008 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1010 2005-08-18 Ross Burton <ross@burtonini.com>
1013 Optimise single-character insertions.
1018 * tests/string-test.c:
1019 Add tests for new optimisation, and fix a leak.
1021 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1023 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1025 * glib/gutils.c: On Darwin, include crt-externs.h and
1026 define environ using _NSGetEnviron(). (#313731)
1028 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1030 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1031 (g_get_any_init_do): ... to this new function.
1032 (g_get_any_init): Declare as inline.
1033 (g_get_any_init_locked): New inline function, does the locking.
1034 Make use of these two throughout the code.
1036 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1038 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1039 the child process and then simply return. This makes
1040 The "S" option work as documented in g_on_error_query().
1041 (#313125, Matthew F. Barnes)
1043 * glib/gunicode.h: Update the link to Unicode category
1044 values. (#313369, Behnam Esfahbod)
1046 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1047 a little. (#311727, Tristan van Berkom)
1049 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1050 check work on ia64 too, where some symbols ended up
1051 in yet another section.