1 2005-12-27 Manish Singh <yosh@gimp.org>
3 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
4 make sure we really override things.
6 2005-12-27 Matthias Clasen <mclasen@redhat.com>
8 Fix #316221, Michal Benes, Stanislav Brabec;
10 * configure.in: Fix a strict aliasing problem in
11 g_static_mutex_get_mutex().
12 * glib/gthread.h: ...and in
13 g_static_mutex_get_mutex_impl_shortcut().
15 * glib/gdatasetprivate.h: Add a cast to silence compiler
16 warnings. (#321978, Andrew Paprocki)
18 Partial fix for bug #32937, Bogdan Nicula.
20 * configure.in: Check for malloc.h
21 * glib/gslice.c: Don't include malloc.h unconditionally.
23 2005-12-27 Manish Singh <yosh@gimp.org>
25 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
26 (#324950, Dan Yefimov)
28 2005-12-26 Matthias Clasen <mclasen@redhat.com>
30 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
33 2005-12-25 Matthias Clasen <mclasen@redhat.com>
35 * AUTHORS: Update my email
37 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
39 2005-12-21 Matthias Clasen <mclasen@redhat.com>
41 * README.in: Add some notes on when 'make check' may fail.
43 2005-12-20 Matthias Clasen <mclasen@redhat.com>
47 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
48 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
50 * tests/threadpool-test.c: Test this.
52 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
54 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
55 handle errors without depending on gmessage.c which might not be
56 setup when the error occours.
57 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
58 always freeing can be achieved by adjusting the working set time to
59 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
60 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
61 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
62 minimize wastage if small amounts of differently sized structrues are
63 allocated, this does come at a performance cost of roughly 5% though.
64 fixed up block alignment calculation, so it works for varying
65 block sizes. only use strerror() not g_strerror() since the latter
66 depends on working GQuark and GSlice.
67 mem_error(): implemented in terms of fprintf and vfprintf.
69 * tests/slice-color.c: new program to test cache colorization effects.
71 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
72 G_SLICE_CONFIG_WORKING_SET_MSECS.
74 2005-12-17 Matthias Clasen <mclasen@redhat.com>
76 * glib/goption.c (parse_short_option): Set an error in all
77 failure cases. (#324332, Tim-Philipp Müller)
79 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
81 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
82 and g_atomic_pointer_get. Add g_atomic_int_set and
83 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
84 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
85 as functions (additionally to the macros in the header) for the
86 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
88 2005-12-16 Matthias Clasen <mclasen@redhat.com>
90 * glib/gmem.c (g_allocator_new): Don't return a pointer to
91 a const struct, since apps expect to be able to modify it.
92 (#324179, J. Ali Harlow)
94 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
96 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
97 for systems where the initialization of atomic variables requires a
100 2005-12-09 Matthias Clasen <mclasen@redhat.com>
102 * configure.in: Bump version
104 * === Released 2.9.1 ===
106 2005-12-09 Alexander Larsson <alexl@redhat.com>
110 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
113 2005-12-08 Matthias Clasen <mclasen@redhat.com>
117 2005-12-07 Martyn Russell <martyn@imendio.com>
119 * glib/gasyncqueue.c:
120 - Call g_queue_insert_sorted() instead of duplicating the code.
121 - Call g_queue_sort() instead of duplicating the code.
122 - Invert sort function results to make sure the same sort function
123 gives the same results across glist, gslist, gqueue and
126 * tests/asyncqueue-test.c:
127 - Updated the sort function to reflect the example in the
128 documentation for gasyncqueue.c.
130 2005-12-07 Martyn Russell <martyn@imendio.com>
132 * docs/reference/glib/glib-sections.txt:
133 * docs/reference/glib/tmpl/linked_lists_double.sgml:
134 * docs/reference/glib/tmpl/linked_lists_single.sgml:
137 - Added g_list_insert_sorted_with_data () and
138 g_slist_insert_sorted_with_data ().
139 - Removed the extra check in g_list_sort() and g_slist_sort() for
140 GCompareDataFunc vs. GCompareFunc.
142 2005-12-07 Tor Lillqvist <tml@novell.com>
144 * glib-gettextize.in: Look up prefix at run-time on Win32,
145 assuming the standard directory structure with glib-gettextize in
148 2005-12-06 Matthias Clasen <mclasen@redhat.com>
150 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
151 we need GTK+ 2.8 to compile against GLib 2.10.
153 2005-12-05 Manish Singh <yosh@gimp.org>
155 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
156 to override any user settings.
158 2005-12-05 Matthias Clasen <mclasen@redhat.com>
160 * configure.in: Revert an accidential commit
162 2005-12-05 Matthias Clasen <mclasen@redhat.com>
164 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
166 * tests/Makefile.am (TESTS_ENVIRONMENT):
167 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
168 MALLOC_CHECK_ and MALLOC_PERTURB_
170 * tests/run-collate-tests.sh: Run the collation tests explicitly
171 in en_US locale. (#320463)
173 * glib/gmem.h: Really deprecate GMemChunk.
180 * glib/gstrfuncs.c: Add versioned deprecation docs.
182 2005-12-05 Martyn Russell <martyn@imendio.com>
184 * docs/reference/glib/glib-sections.txt:
185 * glib/gasyncqueue.[ch]:
186 - Added support for sorting async queues by with _push_sorted(),
187 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
190 * tests/asyncqueue-test.c:
191 - Added test case for gasyncqueue.c
193 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
195 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
196 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
197 in bug 323178. moved time consuming logic from g_slice_free() out of
198 the inner loop, so g_slice_free_chain_with_offset() provides a real
199 performance benefit over g_slice_free1() now.
201 * glib/gslice.h: renamed g_slice_free_chain() to
202 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
203 a type-safe macro as suggested in bug 323178.
204 simplified the macro implementation of g_slice_free() and implemented
205 it in a type safe manner for all compliers as suggested by Morten
206 Welinder <mortenw@gnome.org>.
211 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
213 2005-12-05 Matthias Clasen <mclasen@redhat.com>
215 * glib/gasyncqueue.c: Add some docs.
217 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
219 * glib/gatomic.c: In the ia64 implementation, use
220 __sync builtin without _si or _di suffix. (#321229,
221 Stanislav Brabec, patch by Andreas Schwab)
223 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
225 * glib/gslice.h: Remove comma at the end of enum.
227 2005-12-04 Matthias Clasen <mclasen@redhat.com>
229 Handle multiple user names with the same UID better.
230 (#319535, Laszlo Peter)
232 * glib/gutils.c (g_get_any_init_do): When determining user
233 data, first look up $LOGNAME. If the UID doesn't match
234 getuid(), fall back to the current behaviour of looking
235 up the user data based on getuid().
237 2005-12-04 Matthias Clasen <mclasen@redhat.com>
239 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
240 to make gcc warn if a function result is ignored. (#145466,
241 Arjan van de Ven, Alex Larsson)
243 * glib/gmem.h: Add the new attribute to g_realloc and
246 2005-12-03 Matthias Clasen <mclasen@redhat.com>
249 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
252 * glib/gmappedfile.c: Make mapping of empty files
255 * glib/gfileutils.c: Don't fork a new process just to
256 fix the permissions of the created temp file. (#321318,
257 Alexis S. L. Carvalho)
259 2005-12-02 Matthias Clasen <mclasen@redhat.com>
261 * README.in: Add a note about Solaris threads.
264 * configure.in: Remove support for Solaris threads.
265 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
267 2005-12-02 Matthias Clasen <mclasen@redhat.com>
269 * configure.in: Define G_GUINT64_CONSTANT in analogy to
270 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
272 2005-12-02 Matthias Clasen <mclasen@redhat.com>
274 * glib/gslice.c: Win32 portability fixes and C99ism removal,
275 pointed out by Kazuki Iwamoto. (#323052)
277 2005-12-02 Matthias Clasen <mclasen@redhat.com>
279 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
280 the IA64 implementation.
282 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
284 * glib/gslice.c: some naming and type size fixups.
286 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
288 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
289 variant into something less verbose, but digestible for gcc-3.4.
291 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
293 * tests/slice-test.c: extended to perform the benchmarking on the old
294 memchunk code if 'O' is selected.
296 * tests/memchunks.c: new file which contains the old GLib mem chunks
297 implementation with prefix old_mem_chunk_.
299 * tests/Makefile.am: added memchunks.c
301 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
303 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
304 acceptable load distribution of the syscall.
305 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
306 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
308 2005-12-01 Matthias Clasen <mclasen@redhat.com>
310 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
311 double-free base interface members.
313 * tests/Makefile.am: Remove duplicate variable.
315 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
317 * glib/gslice.[hc]: new slice allocator implementation.
319 * tests/slice-test.c: added random slice allocation test.
321 * glib/gthread.[hc]: removed newly added private thread mem API.
323 * glib/gthreadinit.h:
326 * glib/gmem.c: divided glib threading initialisation into three phases,
327 initialisation where private keys and messaging are not available (only
328 needed by gmem.c), initialisation without messaging but private keys
329 available (gslice.c, gmessage.c), and full fledged initialisers that
330 server the rest of glib. initialisation functions got renamed to reflect
331 the limitations of their corresponding phases.
333 * glib/gmem.c: removed memchunk code, defer allocations to
336 * glib/gmem.[hc]: removed g_slice_* skeletons.
338 * glib/glib.symbols: added g_slice_* symbols.
340 * configure.in: check for availability of posix_memalign(3), memalign(3)
343 * glib/Makefile.am: added gslice.[hc].
345 2005-12-01 Tor Lillqvist <tml@novell.com>
347 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
348 trailing slash(es) for non-root folders. stat() fails if non-root
349 folders are specified with trailing slashes. It's too much hassle
350 to demand that callers strip such slashes themselves, especially
351 as it is easy to get it wrong and strip the slash of a root
353 (g_rename): On NT-based Windows, use MoveFileEx() with
354 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
356 2005-11-28 Matthias Clasen <mclasen@redhat.com>
358 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
361 * configure.in: Check whether do { } while (0) works.
363 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
364 G_STMT_END if it works.
366 2005-11-28 Matthias Clasen <mclasen@redhat.com>
368 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
369 * glib/gnode.c (g_node_depth_traverse_level):
370 * glib/gmem.c (g_allocator_new):
371 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
372 (g_hash_table_foreach_remove_or_steal): Silence compiler
375 2005-11-27 Matthias Clasen <mclasen@redhat.com>
377 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
378 Don't read beyond len. (#322520, Christian Persch)
380 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
382 * glib/gdataset.c: access datalist flags via atomic pointer access
383 functions, instead of acquiring the dataset lock. this is faster and
384 also matches the atomic pointer readouts in gdatalistprivate.h and
385 direct pointer modifications required by gobject.c.
387 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
389 * glib/gdataset.c: streamlined initialization code somewhat,
390 removed GData node cache.
392 2005-11-22 Matthias Clasen <mclasen@redhat.com>
394 * glib/ghash.c (g_hash_table_ref):
395 (g_hash_table_unref): Mark these as new API.
397 Avoid double locking in g_intern_string (#322133,
400 * glib/gdataset.c (g_quark_from_string_internal): New
401 internal function which factors out common parts of
402 g_quark_from[_static]_string.
403 (g_quark_from_string, g_quark_from_static_string):
404 Use g_quark_from_string_internal.
405 (g_intern_string, g_intern_static_string): Use
406 g_quark_from_string_internal, and only take the
408 (g_quark_new): Don't store the strings shifted by -1
409 in the g_quarks array.
410 (g_quark_to_string): Adapt to the previous change.
412 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
416 g_hash_table_new_full(): create hash tables with a ref count of 1.
417 g_hash_table_ref(): atomically ref_count+=1
418 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
419 when refcount reaches 0.
420 g_hash_table_destroy(): just destroy keys and values, unref by 1.
421 g_hash_table_insert():
422 g_hash_table_replace(): assert ref_count>0.
425 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
428 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
430 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
431 variables. (#321984, Andrew Paprocki)
433 2005-11-18 Matthias Clasen <mclasen@redhat.com>
435 * configure.in: Bump version
437 2005-11-17 Matthias Clasen <mclasen@redhat.com>
439 * === Released 2.9.0 ===
441 * glib/glib.symbols: Updates
443 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
444 hints to the docs. (Paolo Borelli)
448 * glib/gthread.c (g_thread_foreach): Mark as new api.
450 * README.in: Updates.
454 * glib/gdate.c (g_date_set_time_t):
455 (g_date_set_time_val): Add functions to set a date from
456 a time_t and from a GTimeVal.
457 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
460 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
461 and g_utf8_offset_to_pointer.
463 * glib/gutf8.c (g_utf8_pointer_to_offset)
464 (g_utf8_offset_to_pointer): Handle negative offsets, and use
465 "stutter stepping" for going backwards. (#320638, Larry
470 * glib/gthread.c: const correctness fixes, found
471 by Arjan van de Ven and gcc.
473 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
475 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
476 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
478 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
480 * glib/guniprop.c: Use bit hacks instead when checking a general
481 category value against multiple values.
483 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
485 2005-11-10 Simos Xenitellis <simos@gnome.org>
487 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
489 2005-11-05 Tor Lillqvist <tml@novell.com>
491 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
492 (#320688) Instead set error and return failure.
494 2005-11-07 Matthias Clasen <mclasen@redhat.com>
496 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
497 while holding the context lock. (#320886, Andy Wingo)
499 2005-11-07 Billy Biggs <vektor@dumbterm.net>
501 * tests/refcount/Makefile.am: Add a missing reference to libglib
502 in the LDADD for this test.
504 2005-11-07 Matthias Clasen <mclasen@redhat.com>
506 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
507 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
508 GLib 2.9. (#320755, Luca Ferretti)
510 2005-11-04 Matthias Clasen <mclasen@redhat.com>
512 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
513 use a faster lookup table for g_unichar_get_mirror_char().
514 * glib/gmirroringtable.h: The new table.
515 * glib/gunichartables.h: Remove bidi_mirroring_table.
516 * glib/gen-unicode-tables.pl: Don't generate the mirroring
518 * glib/glib-mirroring-tab/*: Sources for the program
519 which generated gmirroringtable.h.
521 * glib/glist.c: Avoid some code duplication.
523 * glib/gscanner.h: Include gdataset.h. (#320322)
525 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
526 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
527 to help find critical warnings in applications. (#320017,
530 2005-11-02 Tor Lillqvist <tml@novell.com>
532 * glib/glib.symbols: Remove large amount of trailing whitespace
533 from one line. Remove #ifdef G_UNDEFINED from around
534 g_hash_table_new and g_hash_table_new_full symbols, what was the
535 purpose of that? Add the g_slice functions.
537 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
539 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
540 added g_slice_*() API to allocate and cache small bits of memory.
541 an actuall allocator implementation for g_slice_*() is still pending.
543 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
544 changed GRealThread list to use in-structure *next; fields instead
545 of GSList, in order for thread iteration to not depenend on g_slice_*()
547 _g_thread_mem_private_get():
548 _g_thread_mem_private_set(): added accessors for private memory,
549 needed because the ordinary GPrivate implementation relies on GArray
550 and GSList and therefore indirectly on working g_slice_*() allocations.
553 g_thread_foreach(): new public API function to loop over all existing threads.
567 removed GAllocator and free list usages and accompanying locks.
568 use g_slice_*() API to allocate and cache small bits of memory.
570 * glib/ghook.h: removed GMemChunk field from public API.
573 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
575 * glib/gnode.h: deprecate allocator API.
577 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
578 be used for poll rec lists.
580 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
581 g_relation_destroy(): free all tuples from the all_tuples hash table,
582 this effectively maintains the life time track keeping of tuples.
583 g_relation_delete_tuple(): free tuples which are removed from the
584 all_tuples hash table. this fixes a temporary leak that was present
585 in the memchunk code until the destruction of the relation.
587 2005-10-29 Matthias Clasen <mclasen@redhat.com>
589 * tests/convert-test.c: Add some tests for conversions between
590 UTF-8, UCS-4 and UTF-16.
592 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
593 of len == -1, noticed by Morten Welinder.
595 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
597 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
599 2005-10-26 Matthias Clasen <mclasen@redhat.com>
601 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
602 in the error case, and add some documentation. (#319806, Morten
605 2005-10-19 Manish Singh <yosh@gimp.org>
607 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
608 config.h symbol automatically. Fixes bug #313731.
610 2005-10-19 Tor Lillqvist <tml@novell.com>
612 * glib/gwin32.c (g_win32_get_package_installation_directory):
613 Return a g_strdup()ed copy of the value stored in the hash table,
614 so that it can be g_free()d without leaving a dangling pointer in
615 the hash table. (#319232)
617 2005-10-06 Matthias Clasen <mclasen@redhat.com>
619 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
620 all-zero sequences correctly. (#317930, Sebastien Bacher)
622 2005-10-05 Matthias Clasen <mclasen@redhat.com>
624 * glib/gmarkup.c (g_markup_escape_text): Doc addition
626 * Makefile.am: Add Collation tests.
628 * tests/collate/*: Inputs and expected outputs for collation tests.
630 * tests/run-collate-tests.sh: Script to run collation tests.
632 * tests/unicode-collate.c (main): Rework slightly to make
633 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
635 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
637 * docs/reference/glib/tmpl/unicode.sgml:
638 * glib/gen-unicode-tables.pl:
640 * glib/gunichartables.h:
642 * tests/casefold.txt:
643 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
644 GUnicodeBreakType types. That may break some applications, like
647 2005-09-26 Matthias Clasen <mclasen@redhat.com>
649 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
650 docs. (#314393, Matthew F. Barnes)
653 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
654 they are already declared in glib.h. This doesn't break documented
655 use of gprintf.h, but should probably be pointed out in the
656 release notes for 2.10. (#314232, Behdad Esfahbod)
658 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
660 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
661 complexity optimization, based on a patch from Matthias Clasen.
663 * tests/patterntest.c: more tests, mostly from matthias.
665 2005-09-20 Matthias Clasen <mclasen@redhat.com>
667 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
668 (#316703, Mark Drago)
670 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
672 * glib/gpattern.c: applied a patch from matthias which checks on the
673 upper bound of GPatternSpec length to optimize matches.
676 * tests/patterntest.c: added more match cases.
678 2005-09-16 Tor Lillqvist <tml@novell.com>
680 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
681 warning to doc comment that these functions should not be used on
682 encodings like CP932.
684 2005-09-14 Matthias Clasen <mclasen@redhat.com>
686 * tests/keyfile-test.c: Add a test for grup names of length 1.
688 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
689 of length 1. (#316309)
691 2005-09-12 Matthias Clasen <mclasen@redhat.com>
693 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
695 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
697 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
698 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
701 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
703 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
704 (child_watch_helper_thread): Remove some dead code. Closes
707 2005-09-07 Tor Lillqvist <tml@novell.com>
709 * glib/Makefile.am: Create also a console version of the
710 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
711 It's otherwise identical to gspawn-win32-helper.exe, except marked
712 as a console application (linked without the -mwindows option).
714 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
715 Drop the dont_wait parameter. Its truth value correlated 100% with
716 the NULLness of the exit_status parameter anyway, so it's enough
717 to check whether exit_status is NULL. Invert the sense of the
718 dont_return_handle parameter and rename it to do_return_handle, to
719 make the code easier to read by avoiding double negations.
721 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
722 to do_spawn_with_pipes() accordingly.
724 (do_spawn_with_pipes): If we have a console, use the console
725 version of the helper program, otherwise use the GUI one. This
726 avoids extra console windows opening up in some situations. (In
727 case a console application uses the GUI gspawn-win32-helper.exe to
728 spawn another console application we would get a separate console
729 for the spawned console application).
731 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
733 2005-09-05 Matthias Clasen <mclasen@redhat.com>
735 * glib/gmappedfile.c (g_mapped_file_new): Report an error
736 if the file is too large. (#315275, Kjartan Maraas)
738 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
739 of read() is signed. (#315273, Kjartan Maraas)
741 2005-08-31 Tor Lillqvist <tml@novell.com>
743 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
744 the EXE and each DLL have their own atexit function chains.
746 #define g_atexit as atexit instead. This means it has a
747 better chance of doing what the caller wants. For instance,
748 gtkhtml calls g_atexit() registering a function in gtkhtml
749 itself. This caused a crash when g_atexit() was implemented as a
750 function in the GLib DLL. The gtkhtml DLL was already unloaded by
751 the time the GLib DLL got unloaded.
753 * glib/gutils.c: #undef the #define mentioned above, to also get a
754 real g_atexit() into the DLL for backward compatibility. Document
755 the Windows behaviour of g_atexit(), and document the varying ways
756 atexit() can behave in the context of dynamically loaded modules
759 2005-08-31 Matthias Clasen <mclasen@redhat.com>
763 * glib/gdataset.c: Add string interning functions.
765 2005-08-28 Matthias Clasen <mclasen@redhat.com>
767 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
770 2005-08-26 Matthias Clasen <mclasen@redhat.com>
772 * configure.in: Bump version to 2.9.0
774 2005-08-25 Tor Lillqvist <tml@novell.com>
776 Make also the g_spawn*() functions take parameters in the GLib
777 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
778 API or ABI. Like the other GLib API that was earlier changed to
779 use UTF-8 on Windows, the names of the functions that take UTF-8
780 have _utf8 suffixes added by using preprocessor macros in the
781 header file. The old names are kept for functions with the old
782 behaviour, taking parameters in the system codepage, for DLL ABI
785 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
786 g_spawn*() functions.
788 * glib/gspawn-win32.c: Use wide-char API on NT-based
789 Windows. Convert parameters from UTF-8 to wide chars (NT) or
790 system codepage (Win9x) and call the C library _wspawn*() or
791 spawn*() functions respectvely. Add DLL ABI stability versions
792 that take parameters in the system codepage.
794 * glib/gspawn-win32-helper.c: On NT-based Windows use the
795 wide-char versions of argv and envp, and use wide-char API to
796 change directory and spawn the program to run. Remove the verbose
797 debugging output, it was too complex to modify for the wide-char
798 features. (Just add temporary debugging printouts if needed, no
799 need to have them permanently in the source.)
801 * glib/gspawn.c: Corresponding documentation updates.
803 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
804 symbols as PRIVATE, add the new _utf8-suffixed ones.
806 2005-08-24 Stepan Kasal <kasal@ucw.cz>
808 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
809 guint64 otherwise and that can produce warnings about comparison
810 between signed and unsigned.
812 2005-08-23 Matthias Clasen <mclasen@redhat.com>
814 * glib/gutils.c: Fix the crt_externs.h include.
816 2005-08-23 Stepan Kasal <kasal@ucw.cz>
818 * NEWS: Fix spelling of my first name.
820 2005-08-23 Matthias Clasen <mclasen@redhat.com>
824 * === Released 2.8.1 ===
828 2005-08-20 Hans Breuer <hans@breuer.org>
830 * glib/makefile.msc.in : link with ws2_32.lib
832 2005-08-18 Tor Lillqvist <tml@novell.com>
834 * configure.in: Check for <sys/wait.h>
836 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
838 2005-08-18 Ross Burton <ross@burtonini.com>
841 Optimise single-character insertions.
846 * tests/string-test.c:
847 Add tests for new optimisation, and fix a leak.
849 2005-08-17 Matthias Clasen <mclasen@redhat.com>
851 * configure.in: Check for crt_externs.h and _NSGetEnviron.
853 * glib/gutils.c: On Darwin, include crt-externs.h and
854 define environ using _NSGetEnviron(). (#313731)
856 2005-08-16 Stepan Kasal <kasal@ucw.cz>
858 * glib/gutils.c (g_get_any_init): Move the body of the big if...
859 (g_get_any_init_do): ... to this new function.
860 (g_get_any_init): Declare as inline.
861 (g_get_any_init_locked): New inline function, does the locking.
862 Make use of these two throughout the code.
864 2005-08-15 Matthias Clasen <mclasen@redhat.com>
866 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
867 the child process and then simply return. This makes
868 The "S" option work as documented in g_on_error_query().
869 (#313125, Matthew F. Barnes)
871 * glib/gunicode.h: Update the link to Unicode category
872 values. (#313369, Behnam Esfahbod)
874 * glib/gqueue.c (g_queue_find_custom): Clarify docs
875 a little. (#311727, Tristan van Berkom)
877 * glib/abicheck.sh, gobject/abicheck.sh: Make the
878 check work on ia64 too, where some symbols ended up
879 in yet another section.