1 2006-01-16 Matthias Clasen <mclasen@redhat.com>
3 Fix bug 326747, Alberto Ruiz:
5 * glib/gconvert.c (make_valid_utf8): Change this function to
6 replace unknown characters by the Unicode replacement character
7 instead of '?', and don't append "(invalid encoding)".
8 (g_filename_display_name, g_filename_display_basename): Document
9 how to determine if the filename was in an invalid encoding.
11 2006-01-14 Matthias Clasen <mclasen@redhat.com>
13 * glib/gtree.c: Replace the simple recursive implementation
14 by a nonrecursive, threaded implementation by Maurizio
17 2006-01-12 Matthias Clasen <mclasen@redhat.com>
19 * tests/asyncqueue-test.c (main): Fix the
20 assert to not always trigger. (#326558,
23 2006-01-11 Matthias Clasen <mclasen@redhat.com>
25 * tests/convert-test.c: Don't test items_read and
26 items_written in error cases where it is not set.
28 * tests/asyncqueue-test.c: Set error to NULL.
30 2006-01-08 Matthias Clasen <mclasen@redhat.com>
32 * glib/gtree.c: Break some long lines.
34 * tests/tree-test.c: Add more tests.
36 * glib/gslice.c: Include config.h first, pointed out
39 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
41 * docs/reference/gobject/tmpl/gtype.sgml: explain that
42 G_TYPE_INSTANCE_GET_CLASS() does behave different during
45 2006-01-06 Matthias Clasen <mclasen@redhat.com>
47 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
48 that we have posix_memalign(). (#323937, Bogdan Nicula)
50 2006-01-05 Matthias Clasen <mclasen@redhat.com>
52 * configure.in: Bump version
54 * === Released 2.9.2 ===
56 2006-01-05 Hans Breuer <hans@breuer.org>
58 * glib/makefile.msc.in : added gslice.obj
60 2006-01-05 Matthias Clasen <mclasen@redhat.com>
64 * glib/gmain.c: Clarify the documentation of source ids
65 a bit. (#325874, Dan Williams)
67 * configure.in: Fix another strict aliasing problem.
69 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
72 2006-01-04 Manish Singh <yosh@gimp.org>
75 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
76 either, since they have the same issue as g_rmdir. (Related to
79 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
81 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
82 order when creating magazines, so we hand out chunks with ascending
85 2006-01-04 Abel Cheung <maddog@linuxhall.org>
87 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
89 2006-01-03 Martyn Russell <martyn@imendio.com>
91 * docs/reference/glib/glib-sections.txt:
93 * glib/gthreadpool.[ch]:
94 - Added new API g_thread_pool_get_idle_time() and
95 g_thread_pool_set_idle_time(). (#324228).
97 * tests/threadpool-test.c:
98 - Updated test case to do thread pool sorting, thread pool with
99 no sorting and a thread pool with idle thread timeouts.
101 2006-01-03 Matthias Clasen <mclasen@redhat.com>
103 * glib/gmain.h: Add new functions here, too.
105 * glib/glib.symbols: Add new functions.
107 2005-12-20 Michael Meeks <michael.meeks@novell.com>
109 * glib/gmain.c (g_main_context_is_owner): new method
110 to determine if the current thread is the owner of the
113 2006-01-02 Matthias Clasen <mclasen@redhat.com>
117 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
118 expanding to rmdir, since rmdir is not declared in a portable
119 system header we can include in gstdio.h. (#325249, Jani Monoses)
121 2006-01-01 Tor Lillqvist <tml@novell.com>
123 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
124 stdout and stderr to unbuffered. Otherwise the giochannel layer
125 will try to read from them regardless whether the
126 g_io_channel_win32_poll() call here has indicated
127 readability or not. (#325310)
129 2005-12-29 Matthias Clasen <mclasen@redhat.com>
131 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
132 in the docs. (#325273, Declan Naughton)
134 2005-12-27 Manish Singh <yosh@gimp.org>
136 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
137 make sure we really override things.
139 2005-12-27 Matthias Clasen <mclasen@redhat.com>
141 Fix #316221, Michal Benes, Stanislav Brabec;
143 * configure.in: Fix a strict aliasing problem in
144 g_static_mutex_get_mutex().
145 * glib/gthread.h: ...and in
146 g_static_mutex_get_mutex_impl_shortcut().
148 * glib/gdatasetprivate.h: Add a cast to silence compiler
149 warnings. (#321978, Andrew Paprocki)
151 Partial fix for bug #323937, Bogdan Nicula.
153 * configure.in: Check for malloc.h
154 * glib/gslice.c: Don't include malloc.h unconditionally.
156 2005-12-27 Manish Singh <yosh@gimp.org>
158 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
159 (#324950, Dan Yefimov)
161 2005-12-26 Matthias Clasen <mclasen@redhat.com>
163 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
166 2005-12-25 Matthias Clasen <mclasen@redhat.com>
168 * AUTHORS: Update my email
170 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
172 2005-12-21 Matthias Clasen <mclasen@redhat.com>
174 * README.in: Add some notes on when 'make check' may fail.
176 2005-12-20 Matthias Clasen <mclasen@redhat.com>
179 * glib/gthreadpool.h:
180 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
181 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
183 * tests/threadpool-test.c: Test this.
185 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
187 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
188 handle errors without depending on gmessage.c which might not be
189 setup when the error occours.
190 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
191 always freeing can be achieved by adjusting the working set time to
192 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
193 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
194 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
195 minimize wastage if small amounts of differently sized structrues are
196 allocated, this does come at a performance cost of roughly 5% though.
197 fixed up block alignment calculation, so it works for varying
198 block sizes. only use strerror() not g_strerror() since the latter
199 depends on working GQuark and GSlice.
200 mem_error(): implemented in terms of fprintf and vfprintf.
202 * tests/slice-color.c: new program to test cache colorization effects.
204 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
205 G_SLICE_CONFIG_WORKING_SET_MSECS.
207 2005-12-17 Matthias Clasen <mclasen@redhat.com>
209 * glib/goption.c (parse_short_option): Set an error in all
210 failure cases. (#324332, Tim-Philipp Müller)
212 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
214 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
215 and g_atomic_pointer_get. Add g_atomic_int_set and
216 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
217 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
218 as functions (additionally to the macros in the header) for the
219 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
221 2005-12-16 Matthias Clasen <mclasen@redhat.com>
223 * glib/gmem.c (g_allocator_new): Don't return a pointer to
224 a const struct, since apps expect to be able to modify it.
225 (#324179, J. Ali Harlow)
227 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
229 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
230 for systems where the initialization of atomic variables requires a
231 write memory barrier.
233 2005-12-09 Matthias Clasen <mclasen@redhat.com>
235 * configure.in: Bump version
237 * === Released 2.9.1 ===
239 2005-12-09 Alexander Larsson <alexl@redhat.com>
243 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
246 2005-12-08 Matthias Clasen <mclasen@redhat.com>
250 2005-12-07 Martyn Russell <martyn@imendio.com>
252 * glib/gasyncqueue.c:
253 - Call g_queue_insert_sorted() instead of duplicating the code.
254 - Call g_queue_sort() instead of duplicating the code.
255 - Invert sort function results to make sure the same sort function
256 gives the same results across glist, gslist, gqueue and
259 * tests/asyncqueue-test.c:
260 - Updated the sort function to reflect the example in the
261 documentation for gasyncqueue.c.
263 2005-12-07 Martyn Russell <martyn@imendio.com>
265 * docs/reference/glib/glib-sections.txt:
266 * docs/reference/glib/tmpl/linked_lists_double.sgml:
267 * docs/reference/glib/tmpl/linked_lists_single.sgml:
270 - Added g_list_insert_sorted_with_data () and
271 g_slist_insert_sorted_with_data ().
272 - Removed the extra check in g_list_sort() and g_slist_sort() for
273 GCompareDataFunc vs. GCompareFunc.
275 2005-12-07 Tor Lillqvist <tml@novell.com>
277 * glib-gettextize.in: Look up prefix at run-time on Win32,
278 assuming the standard directory structure with glib-gettextize in
281 2005-12-06 Matthias Clasen <mclasen@redhat.com>
283 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
284 we need GTK+ 2.8 to compile against GLib 2.10.
286 2005-12-05 Manish Singh <yosh@gimp.org>
288 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
289 to override any user settings.
291 2005-12-05 Matthias Clasen <mclasen@redhat.com>
293 * configure.in: Revert an accidential commit
295 2005-12-05 Matthias Clasen <mclasen@redhat.com>
297 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
299 * tests/Makefile.am (TESTS_ENVIRONMENT):
300 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
301 MALLOC_CHECK_ and MALLOC_PERTURB_
303 * tests/run-collate-tests.sh: Run the collation tests explicitly
304 in en_US locale. (#320463)
306 * glib/gmem.h: Really deprecate GMemChunk.
313 * glib/gstrfuncs.c: Add versioned deprecation docs.
315 2005-12-05 Martyn Russell <martyn@imendio.com>
317 * docs/reference/glib/glib-sections.txt:
318 * glib/gasyncqueue.[ch]:
319 - Added support for sorting async queues by with _push_sorted(),
320 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
323 * tests/asyncqueue-test.c:
324 - Added test case for gasyncqueue.c
326 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
328 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
329 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
330 in bug 323178. moved time consuming logic from g_slice_free() out of
331 the inner loop, so g_slice_free_chain_with_offset() provides a real
332 performance benefit over g_slice_free1() now.
334 * glib/gslice.h: renamed g_slice_free_chain() to
335 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
336 a type-safe macro as suggested in bug 323178.
337 simplified the macro implementation of g_slice_free() and implemented
338 it in a type safe manner for all compliers as suggested by Morten
339 Welinder <mortenw@gnome.org>.
344 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
346 2005-12-05 Matthias Clasen <mclasen@redhat.com>
348 * glib/gasyncqueue.c: Add some docs.
350 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
352 * glib/gatomic.c: In the ia64 implementation, use
353 __sync builtin without _si or _di suffix. (#321229,
354 Stanislav Brabec, patch by Andreas Schwab)
356 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
358 * glib/gslice.h: Remove comma at the end of enum.
360 2005-12-04 Matthias Clasen <mclasen@redhat.com>
362 Handle multiple user names with the same UID better.
363 (#319535, Laszlo Peter)
365 * glib/gutils.c (g_get_any_init_do): When determining user
366 data, first look up $LOGNAME. If the UID doesn't match
367 getuid(), fall back to the current behaviour of looking
368 up the user data based on getuid().
370 2005-12-04 Matthias Clasen <mclasen@redhat.com>
372 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
373 to make gcc warn if a function result is ignored. (#145466,
374 Arjan van de Ven, Alex Larsson)
376 * glib/gmem.h: Add the new attribute to g_realloc and
379 2005-12-03 Matthias Clasen <mclasen@redhat.com>
382 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
385 * glib/gmappedfile.c: Make mapping of empty files
388 * glib/gfileutils.c: Don't fork a new process just to
389 fix the permissions of the created temp file. (#321318,
390 Alexis S. L. Carvalho)
392 2005-12-02 Matthias Clasen <mclasen@redhat.com>
394 * README.in: Add a note about Solaris threads.
397 * configure.in: Remove support for Solaris threads.
398 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
400 2005-12-02 Matthias Clasen <mclasen@redhat.com>
402 * configure.in: Define G_GUINT64_CONSTANT in analogy to
403 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
405 2005-12-02 Matthias Clasen <mclasen@redhat.com>
407 * glib/gslice.c: Win32 portability fixes and C99ism removal,
408 pointed out by Kazuki Iwamoto. (#323052)
410 2005-12-02 Matthias Clasen <mclasen@redhat.com>
412 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
413 the IA64 implementation.
415 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
417 * glib/gslice.c: some naming and type size fixups.
419 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
421 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
422 variant into something less verbose, but digestible for gcc-3.4.
424 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
426 * tests/slice-test.c: extended to perform the benchmarking on the old
427 memchunk code if 'O' is selected.
429 * tests/memchunks.c: new file which contains the old GLib mem chunks
430 implementation with prefix old_mem_chunk_.
432 * tests/Makefile.am: added memchunks.c
434 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
436 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
437 acceptable load distribution of the syscall.
438 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
439 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
441 2005-12-01 Matthias Clasen <mclasen@redhat.com>
443 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
444 double-free base interface members.
446 * tests/Makefile.am: Remove duplicate variable.
448 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
450 * glib/gslice.[hc]: new slice allocator implementation.
452 * tests/slice-test.c: added random slice allocation test.
454 * glib/gthread.[hc]: removed newly added private thread mem API.
456 * glib/gthreadinit.h:
459 * glib/gmem.c: divided glib threading initialisation into three phases,
460 initialisation where private keys and messaging are not available (only
461 needed by gmem.c), initialisation without messaging but private keys
462 available (gslice.c, gmessage.c), and full fledged initialisers that
463 server the rest of glib. initialisation functions got renamed to reflect
464 the limitations of their corresponding phases.
466 * glib/gmem.c: removed memchunk code, defer allocations to
469 * glib/gmem.[hc]: removed g_slice_* skeletons.
471 * glib/glib.symbols: added g_slice_* symbols.
473 * configure.in: check for availability of posix_memalign(3), memalign(3)
476 * glib/Makefile.am: added gslice.[hc].
478 2005-12-01 Tor Lillqvist <tml@novell.com>
480 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
481 trailing slash(es) for non-root folders. stat() fails if non-root
482 folders are specified with trailing slashes. It's too much hassle
483 to demand that callers strip such slashes themselves, especially
484 as it is easy to get it wrong and strip the slash of a root
486 (g_rename): On NT-based Windows, use MoveFileEx() with
487 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
489 2005-11-28 Matthias Clasen <mclasen@redhat.com>
491 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
494 * configure.in: Check whether do { } while (0) works.
496 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
497 G_STMT_END if it works.
499 2005-11-28 Matthias Clasen <mclasen@redhat.com>
501 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
502 * glib/gnode.c (g_node_depth_traverse_level):
503 * glib/gmem.c (g_allocator_new):
504 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
505 (g_hash_table_foreach_remove_or_steal): Silence compiler
508 2005-11-27 Matthias Clasen <mclasen@redhat.com>
510 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
511 Don't read beyond len. (#322520, Christian Persch)
513 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
515 * glib/gdataset.c: access datalist flags via atomic pointer access
516 functions, instead of acquiring the dataset lock. this is faster and
517 also matches the atomic pointer readouts in gdatalistprivate.h and
518 direct pointer modifications required by gobject.c.
520 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
522 * glib/gdataset.c: streamlined initialization code somewhat,
523 removed GData node cache.
525 2005-11-22 Matthias Clasen <mclasen@redhat.com>
527 * glib/ghash.c (g_hash_table_ref):
528 (g_hash_table_unref): Mark these as new API.
530 Avoid double locking in g_intern_string (#322133,
533 * glib/gdataset.c (g_quark_from_string_internal): New
534 internal function which factors out common parts of
535 g_quark_from[_static]_string.
536 (g_quark_from_string, g_quark_from_static_string):
537 Use g_quark_from_string_internal.
538 (g_intern_string, g_intern_static_string): Use
539 g_quark_from_string_internal, and only take the
541 (g_quark_new): Don't store the strings shifted by -1
542 in the g_quarks array.
543 (g_quark_to_string): Adapt to the previous change.
545 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
549 g_hash_table_new_full(): create hash tables with a ref count of 1.
550 g_hash_table_ref(): atomically ref_count+=1
551 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
552 when refcount reaches 0.
553 g_hash_table_destroy(): just destroy keys and values, unref by 1.
554 g_hash_table_insert():
555 g_hash_table_replace(): assert ref_count>0.
558 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
561 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
563 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
564 variables. (#321984, Andrew Paprocki)
566 2005-11-18 Matthias Clasen <mclasen@redhat.com>
568 * configure.in: Bump version
570 2005-11-17 Matthias Clasen <mclasen@redhat.com>
572 * === Released 2.9.0 ===
574 * glib/glib.symbols: Updates
576 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
577 hints to the docs. (Paolo Borelli)
581 * glib/gthread.c (g_thread_foreach): Mark as new api.
583 * README.in: Updates.
587 * glib/gdate.c (g_date_set_time_t):
588 (g_date_set_time_val): Add functions to set a date from
589 a time_t and from a GTimeVal.
590 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
593 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
594 and g_utf8_offset_to_pointer.
596 * glib/gutf8.c (g_utf8_pointer_to_offset)
597 (g_utf8_offset_to_pointer): Handle negative offsets, and use
598 "stutter stepping" for going backwards. (#320638, Larry
603 * glib/gthread.c: const correctness fixes, found
604 by Arjan van de Ven and gcc.
606 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
608 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
609 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
611 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
613 * glib/guniprop.c: Use bit hacks instead when checking a general
614 category value against multiple values.
616 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
618 2005-11-10 Simos Xenitellis <simos@gnome.org>
620 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
622 2005-11-05 Tor Lillqvist <tml@novell.com>
624 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
625 (#320688) Instead set error and return failure.
627 2005-11-07 Matthias Clasen <mclasen@redhat.com>
629 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
630 while holding the context lock. (#320886, Andy Wingo)
632 2005-11-07 Billy Biggs <vektor@dumbterm.net>
634 * tests/refcount/Makefile.am: Add a missing reference to libglib
635 in the LDADD for this test.
637 2005-11-07 Matthias Clasen <mclasen@redhat.com>
639 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
640 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
641 GLib 2.9. (#320755, Luca Ferretti)
643 2005-11-04 Matthias Clasen <mclasen@redhat.com>
645 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
646 use a faster lookup table for g_unichar_get_mirror_char().
647 * glib/gmirroringtable.h: The new table.
648 * glib/gunichartables.h: Remove bidi_mirroring_table.
649 * glib/gen-unicode-tables.pl: Don't generate the mirroring
651 * glib/glib-mirroring-tab/*: Sources for the program
652 which generated gmirroringtable.h.
654 * glib/glist.c: Avoid some code duplication.
656 * glib/gscanner.h: Include gdataset.h. (#320322)
658 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
659 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
660 to help find critical warnings in applications. (#320017,
663 2005-11-02 Tor Lillqvist <tml@novell.com>
665 * glib/glib.symbols: Remove large amount of trailing whitespace
666 from one line. Remove #ifdef G_UNDEFINED from around
667 g_hash_table_new and g_hash_table_new_full symbols, what was the
668 purpose of that? Add the g_slice functions.
670 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
672 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
673 added g_slice_*() API to allocate and cache small bits of memory.
674 an actuall allocator implementation for g_slice_*() is still pending.
676 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
677 changed GRealThread list to use in-structure *next; fields instead
678 of GSList, in order for thread iteration to not depenend on g_slice_*()
680 _g_thread_mem_private_get():
681 _g_thread_mem_private_set(): added accessors for private memory,
682 needed because the ordinary GPrivate implementation relies on GArray
683 and GSList and therefore indirectly on working g_slice_*() allocations.
686 g_thread_foreach(): new public API function to loop over all existing threads.
700 removed GAllocator and free list usages and accompanying locks.
701 use g_slice_*() API to allocate and cache small bits of memory.
703 * glib/ghook.h: removed GMemChunk field from public API.
706 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
708 * glib/gnode.h: deprecate allocator API.
710 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
711 be used for poll rec lists.
713 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
714 g_relation_destroy(): free all tuples from the all_tuples hash table,
715 this effectively maintains the life time track keeping of tuples.
716 g_relation_delete_tuple(): free tuples which are removed from the
717 all_tuples hash table. this fixes a temporary leak that was present
718 in the memchunk code until the destruction of the relation.
720 2005-10-29 Matthias Clasen <mclasen@redhat.com>
722 * tests/convert-test.c: Add some tests for conversions between
723 UTF-8, UCS-4 and UTF-16.
725 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
726 of len == -1, noticed by Morten Welinder.
728 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
730 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
732 2005-10-26 Matthias Clasen <mclasen@redhat.com>
734 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
735 in the error case, and add some documentation. (#319806, Morten
738 2005-10-19 Manish Singh <yosh@gimp.org>
740 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
741 config.h symbol automatically. Fixes bug #313731.
743 2005-10-19 Tor Lillqvist <tml@novell.com>
745 * glib/gwin32.c (g_win32_get_package_installation_directory):
746 Return a g_strdup()ed copy of the value stored in the hash table,
747 so that it can be g_free()d without leaving a dangling pointer in
748 the hash table. (#319232)
750 2005-10-06 Matthias Clasen <mclasen@redhat.com>
752 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
753 all-zero sequences correctly. (#317930, Sebastien Bacher)
755 2005-10-05 Matthias Clasen <mclasen@redhat.com>
757 * glib/gmarkup.c (g_markup_escape_text): Doc addition
759 * Makefile.am: Add Collation tests.
761 * tests/collate/*: Inputs and expected outputs for collation tests.
763 * tests/run-collate-tests.sh: Script to run collation tests.
765 * tests/unicode-collate.c (main): Rework slightly to make
766 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
768 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
770 * docs/reference/glib/tmpl/unicode.sgml:
771 * glib/gen-unicode-tables.pl:
773 * glib/gunichartables.h:
775 * tests/casefold.txt:
776 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
777 GUnicodeBreakType types. That may break some applications, like
780 2005-09-26 Matthias Clasen <mclasen@redhat.com>
782 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
783 docs. (#314393, Matthew F. Barnes)
786 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
787 they are already declared in glib.h. This doesn't break documented
788 use of gprintf.h, but should probably be pointed out in the
789 release notes for 2.10. (#314232, Behdad Esfahbod)
791 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
793 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
794 complexity optimization, based on a patch from Matthias Clasen.
796 * tests/patterntest.c: more tests, mostly from matthias.
798 2005-09-20 Matthias Clasen <mclasen@redhat.com>
800 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
801 (#316703, Mark Drago)
803 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
805 * glib/gpattern.c: applied a patch from matthias which checks on the
806 upper bound of GPatternSpec length to optimize matches.
809 * tests/patterntest.c: added more match cases.
811 2005-09-16 Tor Lillqvist <tml@novell.com>
813 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
814 warning to doc comment that these functions should not be used on
815 encodings like CP932.
817 2005-09-14 Matthias Clasen <mclasen@redhat.com>
819 * tests/keyfile-test.c: Add a test for grup names of length 1.
821 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
822 of length 1. (#316309)
824 2005-09-12 Matthias Clasen <mclasen@redhat.com>
826 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
828 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
830 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
831 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
834 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
836 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
837 (child_watch_helper_thread): Remove some dead code. Closes
840 2005-09-07 Tor Lillqvist <tml@novell.com>
842 * glib/Makefile.am: Create also a console version of the
843 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
844 It's otherwise identical to gspawn-win32-helper.exe, except marked
845 as a console application (linked without the -mwindows option).
847 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
848 Drop the dont_wait parameter. Its truth value correlated 100% with
849 the NULLness of the exit_status parameter anyway, so it's enough
850 to check whether exit_status is NULL. Invert the sense of the
851 dont_return_handle parameter and rename it to do_return_handle, to
852 make the code easier to read by avoiding double negations.
854 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
855 to do_spawn_with_pipes() accordingly.
857 (do_spawn_with_pipes): If we have a console, use the console
858 version of the helper program, otherwise use the GUI one. This
859 avoids extra console windows opening up in some situations. (In
860 case a console application uses the GUI gspawn-win32-helper.exe to
861 spawn another console application we would get a separate console
862 for the spawned console application).
864 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
866 2005-09-05 Matthias Clasen <mclasen@redhat.com>
868 * glib/gmappedfile.c (g_mapped_file_new): Report an error
869 if the file is too large. (#315275, Kjartan Maraas)
871 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
872 of read() is signed. (#315273, Kjartan Maraas)
874 2005-08-31 Tor Lillqvist <tml@novell.com>
876 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
877 the EXE and each DLL have their own atexit function chains.
879 #define g_atexit as atexit instead. This means it has a
880 better chance of doing what the caller wants. For instance,
881 gtkhtml calls g_atexit() registering a function in gtkhtml
882 itself. This caused a crash when g_atexit() was implemented as a
883 function in the GLib DLL. The gtkhtml DLL was already unloaded by
884 the time the GLib DLL got unloaded.
886 * glib/gutils.c: #undef the #define mentioned above, to also get a
887 real g_atexit() into the DLL for backward compatibility. Document
888 the Windows behaviour of g_atexit(), and document the varying ways
889 atexit() can behave in the context of dynamically loaded modules
892 2005-08-31 Matthias Clasen <mclasen@redhat.com>
896 * glib/gdataset.c: Add string interning functions.
898 2005-08-28 Matthias Clasen <mclasen@redhat.com>
900 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
903 2005-08-26 Matthias Clasen <mclasen@redhat.com>
905 * configure.in: Bump version to 2.9.0
907 2005-08-25 Tor Lillqvist <tml@novell.com>
909 Make also the g_spawn*() functions take parameters in the GLib
910 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
911 API or ABI. Like the other GLib API that was earlier changed to
912 use UTF-8 on Windows, the names of the functions that take UTF-8
913 have _utf8 suffixes added by using preprocessor macros in the
914 header file. The old names are kept for functions with the old
915 behaviour, taking parameters in the system codepage, for DLL ABI
918 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
919 g_spawn*() functions.
921 * glib/gspawn-win32.c: Use wide-char API on NT-based
922 Windows. Convert parameters from UTF-8 to wide chars (NT) or
923 system codepage (Win9x) and call the C library _wspawn*() or
924 spawn*() functions respectvely. Add DLL ABI stability versions
925 that take parameters in the system codepage.
927 * glib/gspawn-win32-helper.c: On NT-based Windows use the
928 wide-char versions of argv and envp, and use wide-char API to
929 change directory and spawn the program to run. Remove the verbose
930 debugging output, it was too complex to modify for the wide-char
931 features. (Just add temporary debugging printouts if needed, no
932 need to have them permanently in the source.)
934 * glib/gspawn.c: Corresponding documentation updates.
936 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
937 symbols as PRIVATE, add the new _utf8-suffixed ones.
939 2005-08-24 Stepan Kasal <kasal@ucw.cz>
941 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
942 guint64 otherwise and that can produce warnings about comparison
943 between signed and unsigned.
945 2005-08-23 Matthias Clasen <mclasen@redhat.com>
947 * glib/gutils.c: Fix the crt_externs.h include.
949 2005-08-23 Stepan Kasal <kasal@ucw.cz>
951 * NEWS: Fix spelling of my first name.
953 2005-08-23 Matthias Clasen <mclasen@redhat.com>
957 * === Released 2.8.1 ===
961 2005-08-20 Hans Breuer <hans@breuer.org>
963 * glib/makefile.msc.in : link with ws2_32.lib
965 2005-08-18 Tor Lillqvist <tml@novell.com>
967 * configure.in: Check for <sys/wait.h>
969 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
971 2005-08-18 Ross Burton <ross@burtonini.com>
974 Optimise single-character insertions.
979 * tests/string-test.c:
980 Add tests for new optimisation, and fix a leak.
982 2005-08-17 Matthias Clasen <mclasen@redhat.com>
984 * configure.in: Check for crt_externs.h and _NSGetEnviron.
986 * glib/gutils.c: On Darwin, include crt-externs.h and
987 define environ using _NSGetEnviron(). (#313731)
989 2005-08-16 Stepan Kasal <kasal@ucw.cz>
991 * glib/gutils.c (g_get_any_init): Move the body of the big if...
992 (g_get_any_init_do): ... to this new function.
993 (g_get_any_init): Declare as inline.
994 (g_get_any_init_locked): New inline function, does the locking.
995 Make use of these two throughout the code.
997 2005-08-15 Matthias Clasen <mclasen@redhat.com>
999 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1000 the child process and then simply return. This makes
1001 The "S" option work as documented in g_on_error_query().
1002 (#313125, Matthew F. Barnes)
1004 * glib/gunicode.h: Update the link to Unicode category
1005 values. (#313369, Behnam Esfahbod)
1007 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1008 a little. (#311727, Tristan van Berkom)
1010 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1011 check work on ia64 too, where some symbols ended up
1012 in yet another section.