1 2006-01-18 Matthias Clasen <mclasen@redhat.com>
5 * === Released 2.9.4 ===
9 2006-01-17 Matthias Clasen <mclasen@redhat.com>
11 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
12 call. (#327421, Morten Welinder)
14 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
16 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
17 G_LOCK. Use the unused_thread_queue lock instead. Change
18 g_thread_pool_thread_proxy such that threads only wait on
19 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
20 due to superfluous tasks. Global tasks wait at most for
21 max-idle-time milliseconds. Make sure, that no task is woken up
22 twice for the same event via a wakeup_serial. This fixes #324228.
24 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
25 invalid NULL into the thread pools. This as well fixes #327290.
27 2006-01-16 Matthias Clasen <mclasen@redhat.com>
29 * configure.in: Bump version
31 * === Released 2.9.3 ===
35 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
36 Rework the strict aliasing fix to not break C++, pointed
37 out by Murray Cumming.
39 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
40 Signal waiting threads, problem noticed by Christian Kellner.
42 2006-01-16 Matthias Clasen <mclasen@redhat.com>
44 Fix bug 326747, Alberto Ruiz:
46 * glib/gconvert.c (make_valid_utf8): Change this function to
47 replace unknown characters by the Unicode replacement character
48 instead of '?', and don't append "(invalid encoding)".
49 (g_filename_display_name, g_filename_display_basename): Document
50 how to determine if the filename was in an invalid encoding.
52 2006-01-14 Matthias Clasen <mclasen@redhat.com>
54 * glib/gtree.c: Replace the simple recursive implementation
55 by a nonrecursive, threaded implementation by Maurizio
58 2006-01-12 Matthias Clasen <mclasen@redhat.com>
60 * tests/asyncqueue-test.c (main): Fix the
61 assert to not always trigger. (#326558,
64 2006-01-11 Matthias Clasen <mclasen@redhat.com>
66 * tests/convert-test.c: Don't test items_read and
67 items_written in error cases where it is not set.
69 * tests/asyncqueue-test.c: Set error to NULL.
71 2006-01-08 Matthias Clasen <mclasen@redhat.com>
73 * glib/gtree.c: Break some long lines.
75 * tests/tree-test.c: Add more tests.
77 * glib/gslice.c: Include config.h first, pointed out
80 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
82 * docs/reference/gobject/tmpl/gtype.sgml: explain that
83 G_TYPE_INSTANCE_GET_CLASS() does behave different during
86 2006-01-06 Matthias Clasen <mclasen@redhat.com>
88 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
89 that we have posix_memalign(). (#323937, Bogdan Nicula)
91 2006-01-05 Matthias Clasen <mclasen@redhat.com>
93 * configure.in: Bump version
95 * === Released 2.9.2 ===
97 2006-01-05 Hans Breuer <hans@breuer.org>
99 * glib/makefile.msc.in : added gslice.obj
101 2006-01-05 Matthias Clasen <mclasen@redhat.com>
105 * glib/gmain.c: Clarify the documentation of source ids
106 a bit. (#325874, Dan Williams)
108 * configure.in: Fix another strict aliasing problem.
110 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
113 2006-01-04 Manish Singh <yosh@gimp.org>
116 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
117 either, since they have the same issue as g_rmdir. (Related to
120 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
122 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
123 order when creating magazines, so we hand out chunks with ascending
126 2006-01-04 Abel Cheung <maddog@linuxhall.org>
128 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
130 2006-01-03 Martyn Russell <martyn@imendio.com>
132 * docs/reference/glib/glib-sections.txt:
134 * glib/gthreadpool.[ch]:
135 - Added new API g_thread_pool_get_idle_time() and
136 g_thread_pool_set_idle_time(). (#324228).
138 * tests/threadpool-test.c:
139 - Updated test case to do thread pool sorting, thread pool with
140 no sorting and a thread pool with idle thread timeouts.
142 2006-01-03 Matthias Clasen <mclasen@redhat.com>
144 * glib/gmain.h: Add new functions here, too.
146 * glib/glib.symbols: Add new functions.
148 2005-12-20 Michael Meeks <michael.meeks@novell.com>
150 * glib/gmain.c (g_main_context_is_owner): new method
151 to determine if the current thread is the owner of the
154 2006-01-02 Matthias Clasen <mclasen@redhat.com>
158 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
159 expanding to rmdir, since rmdir is not declared in a portable
160 system header we can include in gstdio.h. (#325249, Jani Monoses)
162 2006-01-01 Tor Lillqvist <tml@novell.com>
164 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
165 stdout and stderr to unbuffered. Otherwise the giochannel layer
166 will try to read from them regardless whether the
167 g_io_channel_win32_poll() call here has indicated
168 readability or not. (#325310)
170 2005-12-29 Matthias Clasen <mclasen@redhat.com>
172 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
173 in the docs. (#325273, Declan Naughton)
175 2005-12-27 Manish Singh <yosh@gimp.org>
177 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
178 make sure we really override things.
180 2005-12-27 Matthias Clasen <mclasen@redhat.com>
182 Fix #316221, Michal Benes, Stanislav Brabec;
184 * configure.in: Fix a strict aliasing problem in
185 g_static_mutex_get_mutex().
186 * glib/gthread.h: ...and in
187 g_static_mutex_get_mutex_impl_shortcut().
189 * glib/gdatasetprivate.h: Add a cast to silence compiler
190 warnings. (#321978, Andrew Paprocki)
192 Partial fix for bug #323937, Bogdan Nicula.
194 * configure.in: Check for malloc.h
195 * glib/gslice.c: Don't include malloc.h unconditionally.
197 2005-12-27 Manish Singh <yosh@gimp.org>
199 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
200 (#324950, Dan Yefimov)
202 2005-12-26 Matthias Clasen <mclasen@redhat.com>
204 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
207 2005-12-25 Matthias Clasen <mclasen@redhat.com>
209 * AUTHORS: Update my email
211 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
213 2005-12-21 Matthias Clasen <mclasen@redhat.com>
215 * README.in: Add some notes on when 'make check' may fail.
217 2005-12-20 Matthias Clasen <mclasen@redhat.com>
220 * glib/gthreadpool.h:
221 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
222 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
224 * tests/threadpool-test.c: Test this.
226 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
228 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
229 handle errors without depending on gmessage.c which might not be
230 setup when the error occours.
231 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
232 always freeing can be achieved by adjusting the working set time to
233 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
234 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
235 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
236 minimize wastage if small amounts of differently sized structrues are
237 allocated, this does come at a performance cost of roughly 5% though.
238 fixed up block alignment calculation, so it works for varying
239 block sizes. only use strerror() not g_strerror() since the latter
240 depends on working GQuark and GSlice.
241 mem_error(): implemented in terms of fprintf and vfprintf.
243 * tests/slice-color.c: new program to test cache colorization effects.
245 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
246 G_SLICE_CONFIG_WORKING_SET_MSECS.
248 2005-12-17 Matthias Clasen <mclasen@redhat.com>
250 * glib/goption.c (parse_short_option): Set an error in all
251 failure cases. (#324332, Tim-Philipp Müller)
253 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
255 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
256 and g_atomic_pointer_get. Add g_atomic_int_set and
257 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
258 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
259 as functions (additionally to the macros in the header) for the
260 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
262 2005-12-16 Matthias Clasen <mclasen@redhat.com>
264 * glib/gmem.c (g_allocator_new): Don't return a pointer to
265 a const struct, since apps expect to be able to modify it.
266 (#324179, J. Ali Harlow)
268 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
270 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
271 for systems where the initialization of atomic variables requires a
272 write memory barrier.
274 2005-12-09 Matthias Clasen <mclasen@redhat.com>
276 * configure.in: Bump version
278 * === Released 2.9.1 ===
280 2005-12-09 Alexander Larsson <alexl@redhat.com>
284 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
287 2005-12-08 Matthias Clasen <mclasen@redhat.com>
291 2005-12-07 Martyn Russell <martyn@imendio.com>
293 * glib/gasyncqueue.c:
294 - Call g_queue_insert_sorted() instead of duplicating the code.
295 - Call g_queue_sort() instead of duplicating the code.
296 - Invert sort function results to make sure the same sort function
297 gives the same results across glist, gslist, gqueue and
300 * tests/asyncqueue-test.c:
301 - Updated the sort function to reflect the example in the
302 documentation for gasyncqueue.c.
304 2005-12-07 Martyn Russell <martyn@imendio.com>
306 * docs/reference/glib/glib-sections.txt:
307 * docs/reference/glib/tmpl/linked_lists_double.sgml:
308 * docs/reference/glib/tmpl/linked_lists_single.sgml:
311 - Added g_list_insert_sorted_with_data () and
312 g_slist_insert_sorted_with_data ().
313 - Removed the extra check in g_list_sort() and g_slist_sort() for
314 GCompareDataFunc vs. GCompareFunc.
316 2005-12-07 Tor Lillqvist <tml@novell.com>
318 * glib-gettextize.in: Look up prefix at run-time on Win32,
319 assuming the standard directory structure with glib-gettextize in
322 2005-12-06 Matthias Clasen <mclasen@redhat.com>
324 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
325 we need GTK+ 2.8 to compile against GLib 2.10.
327 2005-12-05 Manish Singh <yosh@gimp.org>
329 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
330 to override any user settings.
332 2005-12-05 Matthias Clasen <mclasen@redhat.com>
334 * configure.in: Revert an accidential commit
336 2005-12-05 Matthias Clasen <mclasen@redhat.com>
338 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
340 * tests/Makefile.am (TESTS_ENVIRONMENT):
341 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
342 MALLOC_CHECK_ and MALLOC_PERTURB_
344 * tests/run-collate-tests.sh: Run the collation tests explicitly
345 in en_US locale. (#320463)
347 * glib/gmem.h: Really deprecate GMemChunk.
354 * glib/gstrfuncs.c: Add versioned deprecation docs.
356 2005-12-05 Martyn Russell <martyn@imendio.com>
358 * docs/reference/glib/glib-sections.txt:
359 * glib/gasyncqueue.[ch]:
360 - Added support for sorting async queues by with _push_sorted(),
361 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
364 * tests/asyncqueue-test.c:
365 - Added test case for gasyncqueue.c
367 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
369 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
370 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
371 in bug 323178. moved time consuming logic from g_slice_free() out of
372 the inner loop, so g_slice_free_chain_with_offset() provides a real
373 performance benefit over g_slice_free1() now.
375 * glib/gslice.h: renamed g_slice_free_chain() to
376 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
377 a type-safe macro as suggested in bug 323178.
378 simplified the macro implementation of g_slice_free() and implemented
379 it in a type safe manner for all compliers as suggested by Morten
380 Welinder <mortenw@gnome.org>.
385 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
387 2005-12-05 Matthias Clasen <mclasen@redhat.com>
389 * glib/gasyncqueue.c: Add some docs.
391 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
393 * glib/gatomic.c: In the ia64 implementation, use
394 __sync builtin without _si or _di suffix. (#321229,
395 Stanislav Brabec, patch by Andreas Schwab)
397 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
399 * glib/gslice.h: Remove comma at the end of enum.
401 2005-12-04 Matthias Clasen <mclasen@redhat.com>
403 Handle multiple user names with the same UID better.
404 (#319535, Laszlo Peter)
406 * glib/gutils.c (g_get_any_init_do): When determining user
407 data, first look up $LOGNAME. If the UID doesn't match
408 getuid(), fall back to the current behaviour of looking
409 up the user data based on getuid().
411 2005-12-04 Matthias Clasen <mclasen@redhat.com>
413 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
414 to make gcc warn if a function result is ignored. (#145466,
415 Arjan van de Ven, Alex Larsson)
417 * glib/gmem.h: Add the new attribute to g_realloc and
420 2005-12-03 Matthias Clasen <mclasen@redhat.com>
423 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
426 * glib/gmappedfile.c: Make mapping of empty files
429 * glib/gfileutils.c: Don't fork a new process just to
430 fix the permissions of the created temp file. (#321318,
431 Alexis S. L. Carvalho)
433 2005-12-02 Matthias Clasen <mclasen@redhat.com>
435 * README.in: Add a note about Solaris threads.
438 * configure.in: Remove support for Solaris threads.
439 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
441 2005-12-02 Matthias Clasen <mclasen@redhat.com>
443 * configure.in: Define G_GUINT64_CONSTANT in analogy to
444 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
446 2005-12-02 Matthias Clasen <mclasen@redhat.com>
448 * glib/gslice.c: Win32 portability fixes and C99ism removal,
449 pointed out by Kazuki Iwamoto. (#323052)
451 2005-12-02 Matthias Clasen <mclasen@redhat.com>
453 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
454 the IA64 implementation.
456 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
458 * glib/gslice.c: some naming and type size fixups.
460 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
462 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
463 variant into something less verbose, but digestible for gcc-3.4.
465 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
467 * tests/slice-test.c: extended to perform the benchmarking on the old
468 memchunk code if 'O' is selected.
470 * tests/memchunks.c: new file which contains the old GLib mem chunks
471 implementation with prefix old_mem_chunk_.
473 * tests/Makefile.am: added memchunks.c
475 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
477 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
478 acceptable load distribution of the syscall.
479 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
480 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
482 2005-12-01 Matthias Clasen <mclasen@redhat.com>
484 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
485 double-free base interface members.
487 * tests/Makefile.am: Remove duplicate variable.
489 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
491 * glib/gslice.[hc]: new slice allocator implementation.
493 * tests/slice-test.c: added random slice allocation test.
495 * glib/gthread.[hc]: removed newly added private thread mem API.
497 * glib/gthreadinit.h:
500 * glib/gmem.c: divided glib threading initialisation into three phases,
501 initialisation where private keys and messaging are not available (only
502 needed by gmem.c), initialisation without messaging but private keys
503 available (gslice.c, gmessage.c), and full fledged initialisers that
504 server the rest of glib. initialisation functions got renamed to reflect
505 the limitations of their corresponding phases.
507 * glib/gmem.c: removed memchunk code, defer allocations to
510 * glib/gmem.[hc]: removed g_slice_* skeletons.
512 * glib/glib.symbols: added g_slice_* symbols.
514 * configure.in: check for availability of posix_memalign(3), memalign(3)
517 * glib/Makefile.am: added gslice.[hc].
519 2005-12-01 Tor Lillqvist <tml@novell.com>
521 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
522 trailing slash(es) for non-root folders. stat() fails if non-root
523 folders are specified with trailing slashes. It's too much hassle
524 to demand that callers strip such slashes themselves, especially
525 as it is easy to get it wrong and strip the slash of a root
527 (g_rename): On NT-based Windows, use MoveFileEx() with
528 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
530 2005-11-28 Matthias Clasen <mclasen@redhat.com>
532 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
535 * configure.in: Check whether do { } while (0) works.
537 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
538 G_STMT_END if it works.
540 2005-11-28 Matthias Clasen <mclasen@redhat.com>
542 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
543 * glib/gnode.c (g_node_depth_traverse_level):
544 * glib/gmem.c (g_allocator_new):
545 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
546 (g_hash_table_foreach_remove_or_steal): Silence compiler
549 2005-11-27 Matthias Clasen <mclasen@redhat.com>
551 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
552 Don't read beyond len. (#322520, Christian Persch)
554 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
556 * glib/gdataset.c: access datalist flags via atomic pointer access
557 functions, instead of acquiring the dataset lock. this is faster and
558 also matches the atomic pointer readouts in gdatalistprivate.h and
559 direct pointer modifications required by gobject.c.
561 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
563 * glib/gdataset.c: streamlined initialization code somewhat,
564 removed GData node cache.
566 2005-11-22 Matthias Clasen <mclasen@redhat.com>
568 * glib/ghash.c (g_hash_table_ref):
569 (g_hash_table_unref): Mark these as new API.
571 Avoid double locking in g_intern_string (#322133,
574 * glib/gdataset.c (g_quark_from_string_internal): New
575 internal function which factors out common parts of
576 g_quark_from[_static]_string.
577 (g_quark_from_string, g_quark_from_static_string):
578 Use g_quark_from_string_internal.
579 (g_intern_string, g_intern_static_string): Use
580 g_quark_from_string_internal, and only take the
582 (g_quark_new): Don't store the strings shifted by -1
583 in the g_quarks array.
584 (g_quark_to_string): Adapt to the previous change.
586 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
590 g_hash_table_new_full(): create hash tables with a ref count of 1.
591 g_hash_table_ref(): atomically ref_count+=1
592 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
593 when refcount reaches 0.
594 g_hash_table_destroy(): just destroy keys and values, unref by 1.
595 g_hash_table_insert():
596 g_hash_table_replace(): assert ref_count>0.
599 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
602 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
604 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
605 variables. (#321984, Andrew Paprocki)
607 2005-11-18 Matthias Clasen <mclasen@redhat.com>
609 * configure.in: Bump version
611 2005-11-17 Matthias Clasen <mclasen@redhat.com>
613 * === Released 2.9.0 ===
615 * glib/glib.symbols: Updates
617 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
618 hints to the docs. (Paolo Borelli)
622 * glib/gthread.c (g_thread_foreach): Mark as new api.
624 * README.in: Updates.
628 * glib/gdate.c (g_date_set_time_t):
629 (g_date_set_time_val): Add functions to set a date from
630 a time_t and from a GTimeVal.
631 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
634 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
635 and g_utf8_offset_to_pointer.
637 * glib/gutf8.c (g_utf8_pointer_to_offset)
638 (g_utf8_offset_to_pointer): Handle negative offsets, and use
639 "stutter stepping" for going backwards. (#320638, Larry
644 * glib/gthread.c: const correctness fixes, found
645 by Arjan van de Ven and gcc.
647 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
649 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
650 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
652 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
654 * glib/guniprop.c: Use bit hacks instead when checking a general
655 category value against multiple values.
657 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
659 2005-11-10 Simos Xenitellis <simos@gnome.org>
661 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
663 2005-11-05 Tor Lillqvist <tml@novell.com>
665 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
666 (#320688) Instead set error and return failure.
668 2005-11-07 Matthias Clasen <mclasen@redhat.com>
670 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
671 while holding the context lock. (#320886, Andy Wingo)
673 2005-11-07 Billy Biggs <vektor@dumbterm.net>
675 * tests/refcount/Makefile.am: Add a missing reference to libglib
676 in the LDADD for this test.
678 2005-11-07 Matthias Clasen <mclasen@redhat.com>
680 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
681 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
682 GLib 2.9. (#320755, Luca Ferretti)
684 2005-11-04 Matthias Clasen <mclasen@redhat.com>
686 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
687 use a faster lookup table for g_unichar_get_mirror_char().
688 * glib/gmirroringtable.h: The new table.
689 * glib/gunichartables.h: Remove bidi_mirroring_table.
690 * glib/gen-unicode-tables.pl: Don't generate the mirroring
692 * glib/glib-mirroring-tab/*: Sources for the program
693 which generated gmirroringtable.h.
695 * glib/glist.c: Avoid some code duplication.
697 * glib/gscanner.h: Include gdataset.h. (#320322)
699 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
700 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
701 to help find critical warnings in applications. (#320017,
704 2005-11-02 Tor Lillqvist <tml@novell.com>
706 * glib/glib.symbols: Remove large amount of trailing whitespace
707 from one line. Remove #ifdef G_UNDEFINED from around
708 g_hash_table_new and g_hash_table_new_full symbols, what was the
709 purpose of that? Add the g_slice functions.
711 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
713 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
714 added g_slice_*() API to allocate and cache small bits of memory.
715 an actuall allocator implementation for g_slice_*() is still pending.
717 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
718 changed GRealThread list to use in-structure *next; fields instead
719 of GSList, in order for thread iteration to not depenend on g_slice_*()
721 _g_thread_mem_private_get():
722 _g_thread_mem_private_set(): added accessors for private memory,
723 needed because the ordinary GPrivate implementation relies on GArray
724 and GSList and therefore indirectly on working g_slice_*() allocations.
727 g_thread_foreach(): new public API function to loop over all existing threads.
741 removed GAllocator and free list usages and accompanying locks.
742 use g_slice_*() API to allocate and cache small bits of memory.
744 * glib/ghook.h: removed GMemChunk field from public API.
747 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
749 * glib/gnode.h: deprecate allocator API.
751 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
752 be used for poll rec lists.
754 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
755 g_relation_destroy(): free all tuples from the all_tuples hash table,
756 this effectively maintains the life time track keeping of tuples.
757 g_relation_delete_tuple(): free tuples which are removed from the
758 all_tuples hash table. this fixes a temporary leak that was present
759 in the memchunk code until the destruction of the relation.
761 2005-10-29 Matthias Clasen <mclasen@redhat.com>
763 * tests/convert-test.c: Add some tests for conversions between
764 UTF-8, UCS-4 and UTF-16.
766 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
767 of len == -1, noticed by Morten Welinder.
769 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
771 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
773 2005-10-26 Matthias Clasen <mclasen@redhat.com>
775 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
776 in the error case, and add some documentation. (#319806, Morten
779 2005-10-19 Manish Singh <yosh@gimp.org>
781 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
782 config.h symbol automatically. Fixes bug #313731.
784 2005-10-19 Tor Lillqvist <tml@novell.com>
786 * glib/gwin32.c (g_win32_get_package_installation_directory):
787 Return a g_strdup()ed copy of the value stored in the hash table,
788 so that it can be g_free()d without leaving a dangling pointer in
789 the hash table. (#319232)
791 2005-10-06 Matthias Clasen <mclasen@redhat.com>
793 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
794 all-zero sequences correctly. (#317930, Sebastien Bacher)
796 2005-10-05 Matthias Clasen <mclasen@redhat.com>
798 * glib/gmarkup.c (g_markup_escape_text): Doc addition
800 * Makefile.am: Add Collation tests.
802 * tests/collate/*: Inputs and expected outputs for collation tests.
804 * tests/run-collate-tests.sh: Script to run collation tests.
806 * tests/unicode-collate.c (main): Rework slightly to make
807 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
809 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
811 * docs/reference/glib/tmpl/unicode.sgml:
812 * glib/gen-unicode-tables.pl:
814 * glib/gunichartables.h:
816 * tests/casefold.txt:
817 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
818 GUnicodeBreakType types. That may break some applications, like
821 2005-09-26 Matthias Clasen <mclasen@redhat.com>
823 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
824 docs. (#314393, Matthew F. Barnes)
827 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
828 they are already declared in glib.h. This doesn't break documented
829 use of gprintf.h, but should probably be pointed out in the
830 release notes for 2.10. (#314232, Behdad Esfahbod)
832 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
834 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
835 complexity optimization, based on a patch from Matthias Clasen.
837 * tests/patterntest.c: more tests, mostly from matthias.
839 2005-09-20 Matthias Clasen <mclasen@redhat.com>
841 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
842 (#316703, Mark Drago)
844 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
846 * glib/gpattern.c: applied a patch from matthias which checks on the
847 upper bound of GPatternSpec length to optimize matches.
850 * tests/patterntest.c: added more match cases.
852 2005-09-16 Tor Lillqvist <tml@novell.com>
854 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
855 warning to doc comment that these functions should not be used on
856 encodings like CP932.
858 2005-09-14 Matthias Clasen <mclasen@redhat.com>
860 * tests/keyfile-test.c: Add a test for grup names of length 1.
862 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
863 of length 1. (#316309)
865 2005-09-12 Matthias Clasen <mclasen@redhat.com>
867 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
869 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
871 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
872 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
875 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
877 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
878 (child_watch_helper_thread): Remove some dead code. Closes
881 2005-09-07 Tor Lillqvist <tml@novell.com>
883 * glib/Makefile.am: Create also a console version of the
884 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
885 It's otherwise identical to gspawn-win32-helper.exe, except marked
886 as a console application (linked without the -mwindows option).
888 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
889 Drop the dont_wait parameter. Its truth value correlated 100% with
890 the NULLness of the exit_status parameter anyway, so it's enough
891 to check whether exit_status is NULL. Invert the sense of the
892 dont_return_handle parameter and rename it to do_return_handle, to
893 make the code easier to read by avoiding double negations.
895 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
896 to do_spawn_with_pipes() accordingly.
898 (do_spawn_with_pipes): If we have a console, use the console
899 version of the helper program, otherwise use the GUI one. This
900 avoids extra console windows opening up in some situations. (In
901 case a console application uses the GUI gspawn-win32-helper.exe to
902 spawn another console application we would get a separate console
903 for the spawned console application).
905 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
907 2005-09-05 Matthias Clasen <mclasen@redhat.com>
909 * glib/gmappedfile.c (g_mapped_file_new): Report an error
910 if the file is too large. (#315275, Kjartan Maraas)
912 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
913 of read() is signed. (#315273, Kjartan Maraas)
915 2005-08-31 Tor Lillqvist <tml@novell.com>
917 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
918 the EXE and each DLL have their own atexit function chains.
920 #define g_atexit as atexit instead. This means it has a
921 better chance of doing what the caller wants. For instance,
922 gtkhtml calls g_atexit() registering a function in gtkhtml
923 itself. This caused a crash when g_atexit() was implemented as a
924 function in the GLib DLL. The gtkhtml DLL was already unloaded by
925 the time the GLib DLL got unloaded.
927 * glib/gutils.c: #undef the #define mentioned above, to also get a
928 real g_atexit() into the DLL for backward compatibility. Document
929 the Windows behaviour of g_atexit(), and document the varying ways
930 atexit() can behave in the context of dynamically loaded modules
933 2005-08-31 Matthias Clasen <mclasen@redhat.com>
937 * glib/gdataset.c: Add string interning functions.
939 2005-08-28 Matthias Clasen <mclasen@redhat.com>
941 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
944 2005-08-26 Matthias Clasen <mclasen@redhat.com>
946 * configure.in: Bump version to 2.9.0
948 2005-08-25 Tor Lillqvist <tml@novell.com>
950 Make also the g_spawn*() functions take parameters in the GLib
951 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
952 API or ABI. Like the other GLib API that was earlier changed to
953 use UTF-8 on Windows, the names of the functions that take UTF-8
954 have _utf8 suffixes added by using preprocessor macros in the
955 header file. The old names are kept for functions with the old
956 behaviour, taking parameters in the system codepage, for DLL ABI
959 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
960 g_spawn*() functions.
962 * glib/gspawn-win32.c: Use wide-char API on NT-based
963 Windows. Convert parameters from UTF-8 to wide chars (NT) or
964 system codepage (Win9x) and call the C library _wspawn*() or
965 spawn*() functions respectvely. Add DLL ABI stability versions
966 that take parameters in the system codepage.
968 * glib/gspawn-win32-helper.c: On NT-based Windows use the
969 wide-char versions of argv and envp, and use wide-char API to
970 change directory and spawn the program to run. Remove the verbose
971 debugging output, it was too complex to modify for the wide-char
972 features. (Just add temporary debugging printouts if needed, no
973 need to have them permanently in the source.)
975 * glib/gspawn.c: Corresponding documentation updates.
977 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
978 symbols as PRIVATE, add the new _utf8-suffixed ones.
980 2005-08-24 Stepan Kasal <kasal@ucw.cz>
982 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
983 guint64 otherwise and that can produce warnings about comparison
984 between signed and unsigned.
986 2005-08-23 Matthias Clasen <mclasen@redhat.com>
988 * glib/gutils.c: Fix the crt_externs.h include.
990 2005-08-23 Stepan Kasal <kasal@ucw.cz>
992 * NEWS: Fix spelling of my first name.
994 2005-08-23 Matthias Clasen <mclasen@redhat.com>
998 * === Released 2.8.1 ===
1002 2005-08-20 Hans Breuer <hans@breuer.org>
1004 * glib/makefile.msc.in : link with ws2_32.lib
1006 2005-08-18 Tor Lillqvist <tml@novell.com>
1008 * configure.in: Check for <sys/wait.h>
1010 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1012 2005-08-18 Ross Burton <ross@burtonini.com>
1015 Optimise single-character insertions.
1020 * tests/string-test.c:
1021 Add tests for new optimisation, and fix a leak.
1023 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1025 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1027 * glib/gutils.c: On Darwin, include crt-externs.h and
1028 define environ using _NSGetEnviron(). (#313731)
1030 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1032 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1033 (g_get_any_init_do): ... to this new function.
1034 (g_get_any_init): Declare as inline.
1035 (g_get_any_init_locked): New inline function, does the locking.
1036 Make use of these two throughout the code.
1038 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1040 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1041 the child process and then simply return. This makes
1042 The "S" option work as documented in g_on_error_query().
1043 (#313125, Matthew F. Barnes)
1045 * glib/gunicode.h: Update the link to Unicode category
1046 values. (#313369, Behnam Esfahbod)
1048 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1049 a little. (#311727, Tristan van Berkom)
1051 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1052 check work on ia64 too, where some symbols ended up
1053 in yet another section.