1 Mon Jan 23 17:30:33 2006 Tim Janik <timj@imendio.com>
3 * glib/gutils.c (_g_getenv_nomalloc): wiped out all the wonderfull
4 G_OS_WIN32 code i wrote ;-[) after tml told me windows has getenv()
7 Mon Jan 23 16:46:20 2006 Tim Janik <timj@imendio.com>
9 * glib/gslice.c (slice_config_init): initialize GSlice config from
10 G_SLICE environemtn variable. we support G_SLICE=always-malloc
11 currently, which forces all g_slice_*() allocations to use the system
15 g_parse_debug_string(): added a note about not using g_malloc() here.
16 _g_getenv_nomalloc(): getenv() variant that doesn't use g_malloc or
17 g_slice. contains only guesswork in the WIN32 branch.
19 2006-01-18 Matthias Clasen <mclasen@redhat.com>
23 * === Released 2.9.4 ===
27 2006-01-17 Matthias Clasen <mclasen@redhat.com>
29 * glib/gkeyfile.c (g_key_file_get_groups): Correct a g_new0()
30 call. (#327421, Morten Welinder)
32 2006-01-17 Sebastian Wilhelmi <seppi@seppi.de>
34 * glib/gthreadpool.c: To avoid deadlocks get rid of the settings
35 G_LOCK. Use the unused_thread_queue lock instead. Change
36 g_thread_pool_thread_proxy such that threads only wait on
37 non-exlusive pools for at most a 1/2 second. Do not reorder tasks
38 due to superfluous tasks. Global tasks wait at most for
39 max-idle-time milliseconds. Make sure, that no task is woken up
40 twice for the same event via a wakeup_serial. This fixes #324228.
42 * tests/threadpool-test.c: Adapt test accordingly. Do not pass
43 invalid NULL into the thread pools. This as well fixes #327290.
45 2006-01-16 Matthias Clasen <mclasen@redhat.com>
47 * configure.in: Bump version
49 * === Released 2.9.3 ===
53 * glib/gthread.h (g_static_mutex_get_mutex_impl_shortcut):
54 Rework the strict aliasing fix to not break C++, pointed
55 out by Murray Cumming.
57 * glib/gasyncqueue.c (g_async_queue_push_sorted_unlocked):
58 Signal waiting threads, problem noticed by Christian Kellner.
60 2006-01-16 Matthias Clasen <mclasen@redhat.com>
62 Fix bug 326747, Alberto Ruiz:
64 * glib/gconvert.c (make_valid_utf8): Change this function to
65 replace unknown characters by the Unicode replacement character
66 instead of '?', and don't append "(invalid encoding)".
67 (g_filename_display_name, g_filename_display_basename): Document
68 how to determine if the filename was in an invalid encoding.
70 2006-01-14 Matthias Clasen <mclasen@redhat.com>
72 * glib/gtree.c: Replace the simple recursive implementation
73 by a nonrecursive, threaded implementation by Maurizio
76 2006-01-12 Matthias Clasen <mclasen@redhat.com>
78 * tests/asyncqueue-test.c (main): Fix the
79 assert to not always trigger. (#326558,
82 2006-01-11 Matthias Clasen <mclasen@redhat.com>
84 * tests/convert-test.c: Don't test items_read and
85 items_written in error cases where it is not set.
87 * tests/asyncqueue-test.c: Set error to NULL.
89 2006-01-08 Matthias Clasen <mclasen@redhat.com>
91 * glib/gtree.c: Break some long lines.
93 * tests/tree-test.c: Add more tests.
95 * glib/gslice.c: Include config.h first, pointed out
98 2005-12-20 Sven Herzberg <herzi@gnome-de.org>
100 * docs/reference/gobject/tmpl/gtype.sgml: explain that
101 G_TYPE_INSTANCE_GET_CLASS() does behave different during
104 2006-01-06 Matthias Clasen <mclasen@redhat.com>
106 * glib/gslice.c: Only define _XOPEN_SOURCE if we know
107 that we have posix_memalign(). (#323937, Bogdan Nicula)
109 2006-01-05 Matthias Clasen <mclasen@redhat.com>
111 * configure.in: Bump version
113 * === Released 2.9.2 ===
115 2006-01-05 Hans Breuer <hans@breuer.org>
117 * glib/makefile.msc.in : added gslice.obj
119 2006-01-05 Matthias Clasen <mclasen@redhat.com>
123 * glib/gmain.c: Clarify the documentation of source ids
124 a bit. (#325874, Dan Williams)
126 * configure.in: Fix another strict aliasing problem.
128 * glib/gthreadpool.c: Avoid use of varargs macro. (#325864,
131 2006-01-04 Manish Singh <yosh@gimp.org>
134 * glib/gstdio.h: don't macroized g_access, g_chdir, and g_unlink
135 either, since they have the same issue as g_rmdir. (Related to
138 Wed Jan 4 13:33:25 2006 Tim Janik <timj@gtk.org>
140 * glib/gslice.c (magazine_cache_pop_magazine): don't reverse chunk
141 order when creating magazines, so we hand out chunks with ascending
144 2006-01-04 Abel Cheung <maddog@linuxhall.org>
146 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
148 2006-01-03 Martyn Russell <martyn@imendio.com>
150 * docs/reference/glib/glib-sections.txt:
152 * glib/gthreadpool.[ch]:
153 - Added new API g_thread_pool_get_idle_time() and
154 g_thread_pool_set_idle_time(). (#324228).
156 * tests/threadpool-test.c:
157 - Updated test case to do thread pool sorting, thread pool with
158 no sorting and a thread pool with idle thread timeouts.
160 2006-01-03 Matthias Clasen <mclasen@redhat.com>
162 * glib/gmain.h: Add new functions here, too.
164 * glib/glib.symbols: Add new functions.
166 2005-12-20 Michael Meeks <michael.meeks@novell.com>
168 * glib/gmain.c (g_main_context_is_owner): new method
169 to determine if the current thread is the owner of the
172 2006-01-02 Matthias Clasen <mclasen@redhat.com>
176 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
177 expanding to rmdir, since rmdir is not declared in a portable
178 system header we can include in gstdio.h. (#325249, Jani Monoses)
180 2006-01-01 Tor Lillqvist <tml@novell.com>
182 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
183 stdout and stderr to unbuffered. Otherwise the giochannel layer
184 will try to read from them regardless whether the
185 g_io_channel_win32_poll() call here has indicated
186 readability or not. (#325310)
188 2005-12-29 Matthias Clasen <mclasen@redhat.com>
190 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
191 in the docs. (#325273, Declan Naughton)
193 2005-12-27 Manish Singh <yosh@gimp.org>
195 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
196 make sure we really override things.
198 2005-12-27 Matthias Clasen <mclasen@redhat.com>
200 Fix #316221, Michal Benes, Stanislav Brabec;
202 * configure.in: Fix a strict aliasing problem in
203 g_static_mutex_get_mutex().
204 * glib/gthread.h: ...and in
205 g_static_mutex_get_mutex_impl_shortcut().
207 * glib/gdatasetprivate.h: Add a cast to silence compiler
208 warnings. (#321978, Andrew Paprocki)
210 Partial fix for bug #323937, Bogdan Nicula.
212 * configure.in: Check for malloc.h
213 * glib/gslice.c: Don't include malloc.h unconditionally.
215 2005-12-27 Manish Singh <yosh@gimp.org>
217 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
218 (#324950, Dan Yefimov)
220 2005-12-26 Matthias Clasen <mclasen@redhat.com>
222 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
225 2005-12-25 Matthias Clasen <mclasen@redhat.com>
227 * AUTHORS: Update my email
229 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
231 2005-12-21 Matthias Clasen <mclasen@redhat.com>
233 * README.in: Add some notes on when 'make check' may fail.
235 2005-12-20 Matthias Clasen <mclasen@redhat.com>
238 * glib/gthreadpool.h:
239 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
240 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
242 * tests/threadpool-test.c: Test this.
244 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
246 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
247 handle errors without depending on gmessage.c which might not be
248 setup when the error occours.
249 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
250 always freeing can be achieved by adjusting the working set time to
251 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
252 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
253 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
254 minimize wastage if small amounts of differently sized structrues are
255 allocated, this does come at a performance cost of roughly 5% though.
256 fixed up block alignment calculation, so it works for varying
257 block sizes. only use strerror() not g_strerror() since the latter
258 depends on working GQuark and GSlice.
259 mem_error(): implemented in terms of fprintf and vfprintf.
261 * tests/slice-color.c: new program to test cache colorization effects.
263 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
264 G_SLICE_CONFIG_WORKING_SET_MSECS.
266 2005-12-17 Matthias Clasen <mclasen@redhat.com>
268 * glib/goption.c (parse_short_option): Set an error in all
269 failure cases. (#324332, Tim-Philipp Müller)
271 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
273 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
274 and g_atomic_pointer_get. Add g_atomic_int_set and
275 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
276 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
277 as functions (additionally to the macros in the header) for the
278 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
280 2005-12-16 Matthias Clasen <mclasen@redhat.com>
282 * glib/gmem.c (g_allocator_new): Don't return a pointer to
283 a const struct, since apps expect to be able to modify it.
284 (#324179, J. Ali Harlow)
286 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
288 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
289 for systems where the initialization of atomic variables requires a
290 write memory barrier.
292 2005-12-09 Matthias Clasen <mclasen@redhat.com>
294 * configure.in: Bump version
296 * === Released 2.9.1 ===
298 2005-12-09 Alexander Larsson <alexl@redhat.com>
302 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
305 2005-12-08 Matthias Clasen <mclasen@redhat.com>
309 2005-12-07 Martyn Russell <martyn@imendio.com>
311 * glib/gasyncqueue.c:
312 - Call g_queue_insert_sorted() instead of duplicating the code.
313 - Call g_queue_sort() instead of duplicating the code.
314 - Invert sort function results to make sure the same sort function
315 gives the same results across glist, gslist, gqueue and
318 * tests/asyncqueue-test.c:
319 - Updated the sort function to reflect the example in the
320 documentation for gasyncqueue.c.
322 2005-12-07 Martyn Russell <martyn@imendio.com>
324 * docs/reference/glib/glib-sections.txt:
325 * docs/reference/glib/tmpl/linked_lists_double.sgml:
326 * docs/reference/glib/tmpl/linked_lists_single.sgml:
329 - Added g_list_insert_sorted_with_data () and
330 g_slist_insert_sorted_with_data ().
331 - Removed the extra check in g_list_sort() and g_slist_sort() for
332 GCompareDataFunc vs. GCompareFunc.
334 2005-12-07 Tor Lillqvist <tml@novell.com>
336 * glib-gettextize.in: Look up prefix at run-time on Win32,
337 assuming the standard directory structure with glib-gettextize in
340 2005-12-06 Matthias Clasen <mclasen@redhat.com>
342 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
343 we need GTK+ 2.8 to compile against GLib 2.10.
345 2005-12-05 Manish Singh <yosh@gimp.org>
347 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
348 to override any user settings.
350 2005-12-05 Matthias Clasen <mclasen@redhat.com>
352 * configure.in: Revert an accidential commit
354 2005-12-05 Matthias Clasen <mclasen@redhat.com>
356 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
358 * tests/Makefile.am (TESTS_ENVIRONMENT):
359 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
360 MALLOC_CHECK_ and MALLOC_PERTURB_
362 * tests/run-collate-tests.sh: Run the collation tests explicitly
363 in en_US locale. (#320463)
365 * glib/gmem.h: Really deprecate GMemChunk.
372 * glib/gstrfuncs.c: Add versioned deprecation docs.
374 2005-12-05 Martyn Russell <martyn@imendio.com>
376 * docs/reference/glib/glib-sections.txt:
377 * glib/gasyncqueue.[ch]:
378 - Added support for sorting async queues by with _push_sorted(),
379 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
382 * tests/asyncqueue-test.c:
383 - Added test case for gasyncqueue.c
385 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
387 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
388 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
389 in bug 323178. moved time consuming logic from g_slice_free() out of
390 the inner loop, so g_slice_free_chain_with_offset() provides a real
391 performance benefit over g_slice_free1() now.
393 * glib/gslice.h: renamed g_slice_free_chain() to
394 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
395 a type-safe macro as suggested in bug 323178.
396 simplified the macro implementation of g_slice_free() and implemented
397 it in a type safe manner for all compliers as suggested by Morten
398 Welinder <mortenw@gnome.org>.
403 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
405 2005-12-05 Matthias Clasen <mclasen@redhat.com>
407 * glib/gasyncqueue.c: Add some docs.
409 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
411 * glib/gatomic.c: In the ia64 implementation, use
412 __sync builtin without _si or _di suffix. (#321229,
413 Stanislav Brabec, patch by Andreas Schwab)
415 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
417 * glib/gslice.h: Remove comma at the end of enum.
419 2005-12-04 Matthias Clasen <mclasen@redhat.com>
421 Handle multiple user names with the same UID better.
422 (#319535, Laszlo Peter)
424 * glib/gutils.c (g_get_any_init_do): When determining user
425 data, first look up $LOGNAME. If the UID doesn't match
426 getuid(), fall back to the current behaviour of looking
427 up the user data based on getuid().
429 2005-12-04 Matthias Clasen <mclasen@redhat.com>
431 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
432 to make gcc warn if a function result is ignored. (#145466,
433 Arjan van de Ven, Alex Larsson)
435 * glib/gmem.h: Add the new attribute to g_realloc and
438 2005-12-03 Matthias Clasen <mclasen@redhat.com>
441 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
444 * glib/gmappedfile.c: Make mapping of empty files
447 * glib/gfileutils.c: Don't fork a new process just to
448 fix the permissions of the created temp file. (#321318,
449 Alexis S. L. Carvalho)
451 2005-12-02 Matthias Clasen <mclasen@redhat.com>
453 * README.in: Add a note about Solaris threads.
456 * configure.in: Remove support for Solaris threads.
457 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
459 2005-12-02 Matthias Clasen <mclasen@redhat.com>
461 * configure.in: Define G_GUINT64_CONSTANT in analogy to
462 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
464 2005-12-02 Matthias Clasen <mclasen@redhat.com>
466 * glib/gslice.c: Win32 portability fixes and C99ism removal,
467 pointed out by Kazuki Iwamoto. (#323052)
469 2005-12-02 Matthias Clasen <mclasen@redhat.com>
471 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
472 the IA64 implementation.
474 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
476 * glib/gslice.c: some naming and type size fixups.
478 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
480 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
481 variant into something less verbose, but digestible for gcc-3.4.
483 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
485 * tests/slice-test.c: extended to perform the benchmarking on the old
486 memchunk code if 'O' is selected.
488 * tests/memchunks.c: new file which contains the old GLib mem chunks
489 implementation with prefix old_mem_chunk_.
491 * tests/Makefile.am: added memchunks.c
493 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
495 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
496 acceptable load distribution of the syscall.
497 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
498 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
500 2005-12-01 Matthias Clasen <mclasen@redhat.com>
502 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
503 double-free base interface members.
505 * tests/Makefile.am: Remove duplicate variable.
507 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
509 * glib/gslice.[hc]: new slice allocator implementation.
511 * tests/slice-test.c: added random slice allocation test.
513 * glib/gthread.[hc]: removed newly added private thread mem API.
515 * glib/gthreadinit.h:
518 * glib/gmem.c: divided glib threading initialisation into three phases,
519 initialisation where private keys and messaging are not available (only
520 needed by gmem.c), initialisation without messaging but private keys
521 available (gslice.c, gmessage.c), and full fledged initialisers that
522 server the rest of glib. initialisation functions got renamed to reflect
523 the limitations of their corresponding phases.
525 * glib/gmem.c: removed memchunk code, defer allocations to
528 * glib/gmem.[hc]: removed g_slice_* skeletons.
530 * glib/glib.symbols: added g_slice_* symbols.
532 * configure.in: check for availability of posix_memalign(3), memalign(3)
535 * glib/Makefile.am: added gslice.[hc].
537 2005-12-01 Tor Lillqvist <tml@novell.com>
539 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
540 trailing slash(es) for non-root folders. stat() fails if non-root
541 folders are specified with trailing slashes. It's too much hassle
542 to demand that callers strip such slashes themselves, especially
543 as it is easy to get it wrong and strip the slash of a root
545 (g_rename): On NT-based Windows, use MoveFileEx() with
546 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
548 2005-11-28 Matthias Clasen <mclasen@redhat.com>
550 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
553 * configure.in: Check whether do { } while (0) works.
555 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
556 G_STMT_END if it works.
558 2005-11-28 Matthias Clasen <mclasen@redhat.com>
560 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
561 * glib/gnode.c (g_node_depth_traverse_level):
562 * glib/gmem.c (g_allocator_new):
563 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
564 (g_hash_table_foreach_remove_or_steal): Silence compiler
567 2005-11-27 Matthias Clasen <mclasen@redhat.com>
569 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
570 Don't read beyond len. (#322520, Christian Persch)
572 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
574 * glib/gdataset.c: access datalist flags via atomic pointer access
575 functions, instead of acquiring the dataset lock. this is faster and
576 also matches the atomic pointer readouts in gdatalistprivate.h and
577 direct pointer modifications required by gobject.c.
579 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
581 * glib/gdataset.c: streamlined initialization code somewhat,
582 removed GData node cache.
584 2005-11-22 Matthias Clasen <mclasen@redhat.com>
586 * glib/ghash.c (g_hash_table_ref):
587 (g_hash_table_unref): Mark these as new API.
589 Avoid double locking in g_intern_string (#322133,
592 * glib/gdataset.c (g_quark_from_string_internal): New
593 internal function which factors out common parts of
594 g_quark_from[_static]_string.
595 (g_quark_from_string, g_quark_from_static_string):
596 Use g_quark_from_string_internal.
597 (g_intern_string, g_intern_static_string): Use
598 g_quark_from_string_internal, and only take the
600 (g_quark_new): Don't store the strings shifted by -1
601 in the g_quarks array.
602 (g_quark_to_string): Adapt to the previous change.
604 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
608 g_hash_table_new_full(): create hash tables with a ref count of 1.
609 g_hash_table_ref(): atomically ref_count+=1
610 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
611 when refcount reaches 0.
612 g_hash_table_destroy(): just destroy keys and values, unref by 1.
613 g_hash_table_insert():
614 g_hash_table_replace(): assert ref_count>0.
617 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
620 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
622 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
623 variables. (#321984, Andrew Paprocki)
625 2005-11-18 Matthias Clasen <mclasen@redhat.com>
627 * configure.in: Bump version
629 2005-11-17 Matthias Clasen <mclasen@redhat.com>
631 * === Released 2.9.0 ===
633 * glib/glib.symbols: Updates
635 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
636 hints to the docs. (Paolo Borelli)
640 * glib/gthread.c (g_thread_foreach): Mark as new api.
642 * README.in: Updates.
646 * glib/gdate.c (g_date_set_time_t):
647 (g_date_set_time_val): Add functions to set a date from
648 a time_t and from a GTimeVal.
649 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
652 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
653 and g_utf8_offset_to_pointer.
655 * glib/gutf8.c (g_utf8_pointer_to_offset)
656 (g_utf8_offset_to_pointer): Handle negative offsets, and use
657 "stutter stepping" for going backwards. (#320638, Larry
662 * glib/gthread.c: const correctness fixes, found
663 by Arjan van de Ven and gcc.
665 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
667 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
668 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
670 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
672 * glib/guniprop.c: Use bit hacks instead when checking a general
673 category value against multiple values.
675 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
677 2005-11-10 Simos Xenitellis <simos@gnome.org>
679 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
681 2005-11-05 Tor Lillqvist <tml@novell.com>
683 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
684 (#320688) Instead set error and return failure.
686 2005-11-07 Matthias Clasen <mclasen@redhat.com>
688 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
689 while holding the context lock. (#320886, Andy Wingo)
691 2005-11-07 Billy Biggs <vektor@dumbterm.net>
693 * tests/refcount/Makefile.am: Add a missing reference to libglib
694 in the LDADD for this test.
696 2005-11-07 Matthias Clasen <mclasen@redhat.com>
698 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
699 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
700 GLib 2.9. (#320755, Luca Ferretti)
702 2005-11-04 Matthias Clasen <mclasen@redhat.com>
704 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
705 use a faster lookup table for g_unichar_get_mirror_char().
706 * glib/gmirroringtable.h: The new table.
707 * glib/gunichartables.h: Remove bidi_mirroring_table.
708 * glib/gen-unicode-tables.pl: Don't generate the mirroring
710 * glib/glib-mirroring-tab/*: Sources for the program
711 which generated gmirroringtable.h.
713 * glib/glist.c: Avoid some code duplication.
715 * glib/gscanner.h: Include gdataset.h. (#320322)
717 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
718 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
719 to help find critical warnings in applications. (#320017,
722 2005-11-02 Tor Lillqvist <tml@novell.com>
724 * glib/glib.symbols: Remove large amount of trailing whitespace
725 from one line. Remove #ifdef G_UNDEFINED from around
726 g_hash_table_new and g_hash_table_new_full symbols, what was the
727 purpose of that? Add the g_slice functions.
729 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
731 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
732 added g_slice_*() API to allocate and cache small bits of memory.
733 an actuall allocator implementation for g_slice_*() is still pending.
735 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
736 changed GRealThread list to use in-structure *next; fields instead
737 of GSList, in order for thread iteration to not depenend on g_slice_*()
739 _g_thread_mem_private_get():
740 _g_thread_mem_private_set(): added accessors for private memory,
741 needed because the ordinary GPrivate implementation relies on GArray
742 and GSList and therefore indirectly on working g_slice_*() allocations.
745 g_thread_foreach(): new public API function to loop over all existing threads.
759 removed GAllocator and free list usages and accompanying locks.
760 use g_slice_*() API to allocate and cache small bits of memory.
762 * glib/ghook.h: removed GMemChunk field from public API.
765 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
767 * glib/gnode.h: deprecate allocator API.
769 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
770 be used for poll rec lists.
772 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
773 g_relation_destroy(): free all tuples from the all_tuples hash table,
774 this effectively maintains the life time track keeping of tuples.
775 g_relation_delete_tuple(): free tuples which are removed from the
776 all_tuples hash table. this fixes a temporary leak that was present
777 in the memchunk code until the destruction of the relation.
779 2005-10-29 Matthias Clasen <mclasen@redhat.com>
781 * tests/convert-test.c: Add some tests for conversions between
782 UTF-8, UCS-4 and UTF-16.
784 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
785 of len == -1, noticed by Morten Welinder.
787 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
789 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
791 2005-10-26 Matthias Clasen <mclasen@redhat.com>
793 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
794 in the error case, and add some documentation. (#319806, Morten
797 2005-10-19 Manish Singh <yosh@gimp.org>
799 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
800 config.h symbol automatically. Fixes bug #313731.
802 2005-10-19 Tor Lillqvist <tml@novell.com>
804 * glib/gwin32.c (g_win32_get_package_installation_directory):
805 Return a g_strdup()ed copy of the value stored in the hash table,
806 so that it can be g_free()d without leaving a dangling pointer in
807 the hash table. (#319232)
809 2005-10-06 Matthias Clasen <mclasen@redhat.com>
811 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
812 all-zero sequences correctly. (#317930, Sebastien Bacher)
814 2005-10-05 Matthias Clasen <mclasen@redhat.com>
816 * glib/gmarkup.c (g_markup_escape_text): Doc addition
818 * Makefile.am: Add Collation tests.
820 * tests/collate/*: Inputs and expected outputs for collation tests.
822 * tests/run-collate-tests.sh: Script to run collation tests.
824 * tests/unicode-collate.c (main): Rework slightly to make
825 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
827 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
829 * docs/reference/glib/tmpl/unicode.sgml:
830 * glib/gen-unicode-tables.pl:
832 * glib/gunichartables.h:
834 * tests/casefold.txt:
835 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
836 GUnicodeBreakType types. That may break some applications, like
839 2005-09-26 Matthias Clasen <mclasen@redhat.com>
841 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
842 docs. (#314393, Matthew F. Barnes)
845 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
846 they are already declared in glib.h. This doesn't break documented
847 use of gprintf.h, but should probably be pointed out in the
848 release notes for 2.10. (#314232, Behdad Esfahbod)
850 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
852 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
853 complexity optimization, based on a patch from Matthias Clasen.
855 * tests/patterntest.c: more tests, mostly from matthias.
857 2005-09-20 Matthias Clasen <mclasen@redhat.com>
859 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
860 (#316703, Mark Drago)
862 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
864 * glib/gpattern.c: applied a patch from matthias which checks on the
865 upper bound of GPatternSpec length to optimize matches.
868 * tests/patterntest.c: added more match cases.
870 2005-09-16 Tor Lillqvist <tml@novell.com>
872 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
873 warning to doc comment that these functions should not be used on
874 encodings like CP932.
876 2005-09-14 Matthias Clasen <mclasen@redhat.com>
878 * tests/keyfile-test.c: Add a test for grup names of length 1.
880 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
881 of length 1. (#316309)
883 2005-09-12 Matthias Clasen <mclasen@redhat.com>
885 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
887 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
889 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
890 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
893 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
895 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
896 (child_watch_helper_thread): Remove some dead code. Closes
899 2005-09-07 Tor Lillqvist <tml@novell.com>
901 * glib/Makefile.am: Create also a console version of the
902 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
903 It's otherwise identical to gspawn-win32-helper.exe, except marked
904 as a console application (linked without the -mwindows option).
906 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
907 Drop the dont_wait parameter. Its truth value correlated 100% with
908 the NULLness of the exit_status parameter anyway, so it's enough
909 to check whether exit_status is NULL. Invert the sense of the
910 dont_return_handle parameter and rename it to do_return_handle, to
911 make the code easier to read by avoiding double negations.
913 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
914 to do_spawn_with_pipes() accordingly.
916 (do_spawn_with_pipes): If we have a console, use the console
917 version of the helper program, otherwise use the GUI one. This
918 avoids extra console windows opening up in some situations. (In
919 case a console application uses the GUI gspawn-win32-helper.exe to
920 spawn another console application we would get a separate console
921 for the spawned console application).
923 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
925 2005-09-05 Matthias Clasen <mclasen@redhat.com>
927 * glib/gmappedfile.c (g_mapped_file_new): Report an error
928 if the file is too large. (#315275, Kjartan Maraas)
930 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
931 of read() is signed. (#315273, Kjartan Maraas)
933 2005-08-31 Tor Lillqvist <tml@novell.com>
935 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
936 the EXE and each DLL have their own atexit function chains.
938 #define g_atexit as atexit instead. This means it has a
939 better chance of doing what the caller wants. For instance,
940 gtkhtml calls g_atexit() registering a function in gtkhtml
941 itself. This caused a crash when g_atexit() was implemented as a
942 function in the GLib DLL. The gtkhtml DLL was already unloaded by
943 the time the GLib DLL got unloaded.
945 * glib/gutils.c: #undef the #define mentioned above, to also get a
946 real g_atexit() into the DLL for backward compatibility. Document
947 the Windows behaviour of g_atexit(), and document the varying ways
948 atexit() can behave in the context of dynamically loaded modules
951 2005-08-31 Matthias Clasen <mclasen@redhat.com>
955 * glib/gdataset.c: Add string interning functions.
957 2005-08-28 Matthias Clasen <mclasen@redhat.com>
959 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
962 2005-08-26 Matthias Clasen <mclasen@redhat.com>
964 * configure.in: Bump version to 2.9.0
966 2005-08-25 Tor Lillqvist <tml@novell.com>
968 Make also the g_spawn*() functions take parameters in the GLib
969 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
970 API or ABI. Like the other GLib API that was earlier changed to
971 use UTF-8 on Windows, the names of the functions that take UTF-8
972 have _utf8 suffixes added by using preprocessor macros in the
973 header file. The old names are kept for functions with the old
974 behaviour, taking parameters in the system codepage, for DLL ABI
977 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
978 g_spawn*() functions.
980 * glib/gspawn-win32.c: Use wide-char API on NT-based
981 Windows. Convert parameters from UTF-8 to wide chars (NT) or
982 system codepage (Win9x) and call the C library _wspawn*() or
983 spawn*() functions respectvely. Add DLL ABI stability versions
984 that take parameters in the system codepage.
986 * glib/gspawn-win32-helper.c: On NT-based Windows use the
987 wide-char versions of argv and envp, and use wide-char API to
988 change directory and spawn the program to run. Remove the verbose
989 debugging output, it was too complex to modify for the wide-char
990 features. (Just add temporary debugging printouts if needed, no
991 need to have them permanently in the source.)
993 * glib/gspawn.c: Corresponding documentation updates.
995 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
996 symbols as PRIVATE, add the new _utf8-suffixed ones.
998 2005-08-24 Stepan Kasal <kasal@ucw.cz>
1000 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
1001 guint64 otherwise and that can produce warnings about comparison
1002 between signed and unsigned.
1004 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1006 * glib/gutils.c: Fix the crt_externs.h include.
1008 2005-08-23 Stepan Kasal <kasal@ucw.cz>
1010 * NEWS: Fix spelling of my first name.
1012 2005-08-23 Matthias Clasen <mclasen@redhat.com>
1016 * === Released 2.8.1 ===
1020 2005-08-20 Hans Breuer <hans@breuer.org>
1022 * glib/makefile.msc.in : link with ws2_32.lib
1024 2005-08-18 Tor Lillqvist <tml@novell.com>
1026 * configure.in: Check for <sys/wait.h>
1028 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
1030 2005-08-18 Ross Burton <ross@burtonini.com>
1033 Optimise single-character insertions.
1038 * tests/string-test.c:
1039 Add tests for new optimisation, and fix a leak.
1041 2005-08-17 Matthias Clasen <mclasen@redhat.com>
1043 * configure.in: Check for crt_externs.h and _NSGetEnviron.
1045 * glib/gutils.c: On Darwin, include crt-externs.h and
1046 define environ using _NSGetEnviron(). (#313731)
1048 2005-08-16 Stepan Kasal <kasal@ucw.cz>
1050 * glib/gutils.c (g_get_any_init): Move the body of the big if...
1051 (g_get_any_init_do): ... to this new function.
1052 (g_get_any_init): Declare as inline.
1053 (g_get_any_init_locked): New inline function, does the locking.
1054 Make use of these two throughout the code.
1056 2005-08-15 Matthias Clasen <mclasen@redhat.com>
1058 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
1059 the child process and then simply return. This makes
1060 The "S" option work as documented in g_on_error_query().
1061 (#313125, Matthew F. Barnes)
1063 * glib/gunicode.h: Update the link to Unicode category
1064 values. (#313369, Behnam Esfahbod)
1066 * glib/gqueue.c (g_queue_find_custom): Clarify docs
1067 a little. (#311727, Tristan van Berkom)
1069 * glib/abicheck.sh, gobject/abicheck.sh: Make the
1070 check work on ia64 too, where some symbols ended up
1071 in yet another section.