1 2005-12-29 Matthias Clasen <mclasen@redhat.com>
3 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
4 in the docs. (#325273, Declan Naughton)
6 2005-12-27 Manish Singh <yosh@gimp.org>
8 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
9 make sure we really override things.
11 2005-12-27 Matthias Clasen <mclasen@redhat.com>
13 Fix #316221, Michal Benes, Stanislav Brabec;
15 * configure.in: Fix a strict aliasing problem in
16 g_static_mutex_get_mutex().
17 * glib/gthread.h: ...and in
18 g_static_mutex_get_mutex_impl_shortcut().
20 * glib/gdatasetprivate.h: Add a cast to silence compiler
21 warnings. (#321978, Andrew Paprocki)
23 Partial fix for bug #32937, Bogdan Nicula.
25 * configure.in: Check for malloc.h
26 * glib/gslice.c: Don't include malloc.h unconditionally.
28 2005-12-27 Manish Singh <yosh@gimp.org>
30 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
31 (#324950, Dan Yefimov)
33 2005-12-26 Matthias Clasen <mclasen@redhat.com>
35 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
38 2005-12-25 Matthias Clasen <mclasen@redhat.com>
40 * AUTHORS: Update my email
42 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
44 2005-12-21 Matthias Clasen <mclasen@redhat.com>
46 * README.in: Add some notes on when 'make check' may fail.
48 2005-12-20 Matthias Clasen <mclasen@redhat.com>
52 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
53 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
55 * tests/threadpool-test.c: Test this.
57 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
59 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
60 handle errors without depending on gmessage.c which might not be
61 setup when the error occours.
62 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
63 always freeing can be achieved by adjusting the working set time to
64 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
65 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
66 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
67 minimize wastage if small amounts of differently sized structrues are
68 allocated, this does come at a performance cost of roughly 5% though.
69 fixed up block alignment calculation, so it works for varying
70 block sizes. only use strerror() not g_strerror() since the latter
71 depends on working GQuark and GSlice.
72 mem_error(): implemented in terms of fprintf and vfprintf.
74 * tests/slice-color.c: new program to test cache colorization effects.
76 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
77 G_SLICE_CONFIG_WORKING_SET_MSECS.
79 2005-12-17 Matthias Clasen <mclasen@redhat.com>
81 * glib/goption.c (parse_short_option): Set an error in all
82 failure cases. (#324332, Tim-Philipp Müller)
84 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
86 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
87 and g_atomic_pointer_get. Add g_atomic_int_set and
88 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
89 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
90 as functions (additionally to the macros in the header) for the
91 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
93 2005-12-16 Matthias Clasen <mclasen@redhat.com>
95 * glib/gmem.c (g_allocator_new): Don't return a pointer to
96 a const struct, since apps expect to be able to modify it.
97 (#324179, J. Ali Harlow)
99 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
101 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
102 for systems where the initialization of atomic variables requires a
103 write memory barrier.
105 2005-12-09 Matthias Clasen <mclasen@redhat.com>
107 * configure.in: Bump version
109 * === Released 2.9.1 ===
111 2005-12-09 Alexander Larsson <alexl@redhat.com>
115 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
118 2005-12-08 Matthias Clasen <mclasen@redhat.com>
122 2005-12-07 Martyn Russell <martyn@imendio.com>
124 * glib/gasyncqueue.c:
125 - Call g_queue_insert_sorted() instead of duplicating the code.
126 - Call g_queue_sort() instead of duplicating the code.
127 - Invert sort function results to make sure the same sort function
128 gives the same results across glist, gslist, gqueue and
131 * tests/asyncqueue-test.c:
132 - Updated the sort function to reflect the example in the
133 documentation for gasyncqueue.c.
135 2005-12-07 Martyn Russell <martyn@imendio.com>
137 * docs/reference/glib/glib-sections.txt:
138 * docs/reference/glib/tmpl/linked_lists_double.sgml:
139 * docs/reference/glib/tmpl/linked_lists_single.sgml:
142 - Added g_list_insert_sorted_with_data () and
143 g_slist_insert_sorted_with_data ().
144 - Removed the extra check in g_list_sort() and g_slist_sort() for
145 GCompareDataFunc vs. GCompareFunc.
147 2005-12-07 Tor Lillqvist <tml@novell.com>
149 * glib-gettextize.in: Look up prefix at run-time on Win32,
150 assuming the standard directory structure with glib-gettextize in
153 2005-12-06 Matthias Clasen <mclasen@redhat.com>
155 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
156 we need GTK+ 2.8 to compile against GLib 2.10.
158 2005-12-05 Manish Singh <yosh@gimp.org>
160 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
161 to override any user settings.
163 2005-12-05 Matthias Clasen <mclasen@redhat.com>
165 * configure.in: Revert an accidential commit
167 2005-12-05 Matthias Clasen <mclasen@redhat.com>
169 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
171 * tests/Makefile.am (TESTS_ENVIRONMENT):
172 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
173 MALLOC_CHECK_ and MALLOC_PERTURB_
175 * tests/run-collate-tests.sh: Run the collation tests explicitly
176 in en_US locale. (#320463)
178 * glib/gmem.h: Really deprecate GMemChunk.
185 * glib/gstrfuncs.c: Add versioned deprecation docs.
187 2005-12-05 Martyn Russell <martyn@imendio.com>
189 * docs/reference/glib/glib-sections.txt:
190 * glib/gasyncqueue.[ch]:
191 - Added support for sorting async queues by with _push_sorted(),
192 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
195 * tests/asyncqueue-test.c:
196 - Added test case for gasyncqueue.c
198 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
200 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
201 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
202 in bug 323178. moved time consuming logic from g_slice_free() out of
203 the inner loop, so g_slice_free_chain_with_offset() provides a real
204 performance benefit over g_slice_free1() now.
206 * glib/gslice.h: renamed g_slice_free_chain() to
207 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
208 a type-safe macro as suggested in bug 323178.
209 simplified the macro implementation of g_slice_free() and implemented
210 it in a type safe manner for all compliers as suggested by Morten
211 Welinder <mortenw@gnome.org>.
216 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
218 2005-12-05 Matthias Clasen <mclasen@redhat.com>
220 * glib/gasyncqueue.c: Add some docs.
222 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
224 * glib/gatomic.c: In the ia64 implementation, use
225 __sync builtin without _si or _di suffix. (#321229,
226 Stanislav Brabec, patch by Andreas Schwab)
228 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
230 * glib/gslice.h: Remove comma at the end of enum.
232 2005-12-04 Matthias Clasen <mclasen@redhat.com>
234 Handle multiple user names with the same UID better.
235 (#319535, Laszlo Peter)
237 * glib/gutils.c (g_get_any_init_do): When determining user
238 data, first look up $LOGNAME. If the UID doesn't match
239 getuid(), fall back to the current behaviour of looking
240 up the user data based on getuid().
242 2005-12-04 Matthias Clasen <mclasen@redhat.com>
244 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
245 to make gcc warn if a function result is ignored. (#145466,
246 Arjan van de Ven, Alex Larsson)
248 * glib/gmem.h: Add the new attribute to g_realloc and
251 2005-12-03 Matthias Clasen <mclasen@redhat.com>
254 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
257 * glib/gmappedfile.c: Make mapping of empty files
260 * glib/gfileutils.c: Don't fork a new process just to
261 fix the permissions of the created temp file. (#321318,
262 Alexis S. L. Carvalho)
264 2005-12-02 Matthias Clasen <mclasen@redhat.com>
266 * README.in: Add a note about Solaris threads.
269 * configure.in: Remove support for Solaris threads.
270 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
272 2005-12-02 Matthias Clasen <mclasen@redhat.com>
274 * configure.in: Define G_GUINT64_CONSTANT in analogy to
275 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
277 2005-12-02 Matthias Clasen <mclasen@redhat.com>
279 * glib/gslice.c: Win32 portability fixes and C99ism removal,
280 pointed out by Kazuki Iwamoto. (#323052)
282 2005-12-02 Matthias Clasen <mclasen@redhat.com>
284 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
285 the IA64 implementation.
287 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
289 * glib/gslice.c: some naming and type size fixups.
291 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
293 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
294 variant into something less verbose, but digestible for gcc-3.4.
296 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
298 * tests/slice-test.c: extended to perform the benchmarking on the old
299 memchunk code if 'O' is selected.
301 * tests/memchunks.c: new file which contains the old GLib mem chunks
302 implementation with prefix old_mem_chunk_.
304 * tests/Makefile.am: added memchunks.c
306 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
308 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
309 acceptable load distribution of the syscall.
310 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
311 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
313 2005-12-01 Matthias Clasen <mclasen@redhat.com>
315 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
316 double-free base interface members.
318 * tests/Makefile.am: Remove duplicate variable.
320 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
322 * glib/gslice.[hc]: new slice allocator implementation.
324 * tests/slice-test.c: added random slice allocation test.
326 * glib/gthread.[hc]: removed newly added private thread mem API.
328 * glib/gthreadinit.h:
331 * glib/gmem.c: divided glib threading initialisation into three phases,
332 initialisation where private keys and messaging are not available (only
333 needed by gmem.c), initialisation without messaging but private keys
334 available (gslice.c, gmessage.c), and full fledged initialisers that
335 server the rest of glib. initialisation functions got renamed to reflect
336 the limitations of their corresponding phases.
338 * glib/gmem.c: removed memchunk code, defer allocations to
341 * glib/gmem.[hc]: removed g_slice_* skeletons.
343 * glib/glib.symbols: added g_slice_* symbols.
345 * configure.in: check for availability of posix_memalign(3), memalign(3)
348 * glib/Makefile.am: added gslice.[hc].
350 2005-12-01 Tor Lillqvist <tml@novell.com>
352 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
353 trailing slash(es) for non-root folders. stat() fails if non-root
354 folders are specified with trailing slashes. It's too much hassle
355 to demand that callers strip such slashes themselves, especially
356 as it is easy to get it wrong and strip the slash of a root
358 (g_rename): On NT-based Windows, use MoveFileEx() with
359 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
361 2005-11-28 Matthias Clasen <mclasen@redhat.com>
363 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
366 * configure.in: Check whether do { } while (0) works.
368 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
369 G_STMT_END if it works.
371 2005-11-28 Matthias Clasen <mclasen@redhat.com>
373 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
374 * glib/gnode.c (g_node_depth_traverse_level):
375 * glib/gmem.c (g_allocator_new):
376 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
377 (g_hash_table_foreach_remove_or_steal): Silence compiler
380 2005-11-27 Matthias Clasen <mclasen@redhat.com>
382 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
383 Don't read beyond len. (#322520, Christian Persch)
385 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
387 * glib/gdataset.c: access datalist flags via atomic pointer access
388 functions, instead of acquiring the dataset lock. this is faster and
389 also matches the atomic pointer readouts in gdatalistprivate.h and
390 direct pointer modifications required by gobject.c.
392 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
394 * glib/gdataset.c: streamlined initialization code somewhat,
395 removed GData node cache.
397 2005-11-22 Matthias Clasen <mclasen@redhat.com>
399 * glib/ghash.c (g_hash_table_ref):
400 (g_hash_table_unref): Mark these as new API.
402 Avoid double locking in g_intern_string (#322133,
405 * glib/gdataset.c (g_quark_from_string_internal): New
406 internal function which factors out common parts of
407 g_quark_from[_static]_string.
408 (g_quark_from_string, g_quark_from_static_string):
409 Use g_quark_from_string_internal.
410 (g_intern_string, g_intern_static_string): Use
411 g_quark_from_string_internal, and only take the
413 (g_quark_new): Don't store the strings shifted by -1
414 in the g_quarks array.
415 (g_quark_to_string): Adapt to the previous change.
417 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
421 g_hash_table_new_full(): create hash tables with a ref count of 1.
422 g_hash_table_ref(): atomically ref_count+=1
423 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
424 when refcount reaches 0.
425 g_hash_table_destroy(): just destroy keys and values, unref by 1.
426 g_hash_table_insert():
427 g_hash_table_replace(): assert ref_count>0.
430 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
433 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
435 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
436 variables. (#321984, Andrew Paprocki)
438 2005-11-18 Matthias Clasen <mclasen@redhat.com>
440 * configure.in: Bump version
442 2005-11-17 Matthias Clasen <mclasen@redhat.com>
444 * === Released 2.9.0 ===
446 * glib/glib.symbols: Updates
448 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
449 hints to the docs. (Paolo Borelli)
453 * glib/gthread.c (g_thread_foreach): Mark as new api.
455 * README.in: Updates.
459 * glib/gdate.c (g_date_set_time_t):
460 (g_date_set_time_val): Add functions to set a date from
461 a time_t and from a GTimeVal.
462 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
465 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
466 and g_utf8_offset_to_pointer.
468 * glib/gutf8.c (g_utf8_pointer_to_offset)
469 (g_utf8_offset_to_pointer): Handle negative offsets, and use
470 "stutter stepping" for going backwards. (#320638, Larry
475 * glib/gthread.c: const correctness fixes, found
476 by Arjan van de Ven and gcc.
478 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
480 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
481 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
483 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
485 * glib/guniprop.c: Use bit hacks instead when checking a general
486 category value against multiple values.
488 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
490 2005-11-10 Simos Xenitellis <simos@gnome.org>
492 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
494 2005-11-05 Tor Lillqvist <tml@novell.com>
496 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
497 (#320688) Instead set error and return failure.
499 2005-11-07 Matthias Clasen <mclasen@redhat.com>
501 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
502 while holding the context lock. (#320886, Andy Wingo)
504 2005-11-07 Billy Biggs <vektor@dumbterm.net>
506 * tests/refcount/Makefile.am: Add a missing reference to libglib
507 in the LDADD for this test.
509 2005-11-07 Matthias Clasen <mclasen@redhat.com>
511 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
512 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
513 GLib 2.9. (#320755, Luca Ferretti)
515 2005-11-04 Matthias Clasen <mclasen@redhat.com>
517 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
518 use a faster lookup table for g_unichar_get_mirror_char().
519 * glib/gmirroringtable.h: The new table.
520 * glib/gunichartables.h: Remove bidi_mirroring_table.
521 * glib/gen-unicode-tables.pl: Don't generate the mirroring
523 * glib/glib-mirroring-tab/*: Sources for the program
524 which generated gmirroringtable.h.
526 * glib/glist.c: Avoid some code duplication.
528 * glib/gscanner.h: Include gdataset.h. (#320322)
530 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
531 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
532 to help find critical warnings in applications. (#320017,
535 2005-11-02 Tor Lillqvist <tml@novell.com>
537 * glib/glib.symbols: Remove large amount of trailing whitespace
538 from one line. Remove #ifdef G_UNDEFINED from around
539 g_hash_table_new and g_hash_table_new_full symbols, what was the
540 purpose of that? Add the g_slice functions.
542 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
544 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
545 added g_slice_*() API to allocate and cache small bits of memory.
546 an actuall allocator implementation for g_slice_*() is still pending.
548 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
549 changed GRealThread list to use in-structure *next; fields instead
550 of GSList, in order for thread iteration to not depenend on g_slice_*()
552 _g_thread_mem_private_get():
553 _g_thread_mem_private_set(): added accessors for private memory,
554 needed because the ordinary GPrivate implementation relies on GArray
555 and GSList and therefore indirectly on working g_slice_*() allocations.
558 g_thread_foreach(): new public API function to loop over all existing threads.
572 removed GAllocator and free list usages and accompanying locks.
573 use g_slice_*() API to allocate and cache small bits of memory.
575 * glib/ghook.h: removed GMemChunk field from public API.
578 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
580 * glib/gnode.h: deprecate allocator API.
582 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
583 be used for poll rec lists.
585 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
586 g_relation_destroy(): free all tuples from the all_tuples hash table,
587 this effectively maintains the life time track keeping of tuples.
588 g_relation_delete_tuple(): free tuples which are removed from the
589 all_tuples hash table. this fixes a temporary leak that was present
590 in the memchunk code until the destruction of the relation.
592 2005-10-29 Matthias Clasen <mclasen@redhat.com>
594 * tests/convert-test.c: Add some tests for conversions between
595 UTF-8, UCS-4 and UTF-16.
597 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
598 of len == -1, noticed by Morten Welinder.
600 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
602 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
604 2005-10-26 Matthias Clasen <mclasen@redhat.com>
606 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
607 in the error case, and add some documentation. (#319806, Morten
610 2005-10-19 Manish Singh <yosh@gimp.org>
612 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
613 config.h symbol automatically. Fixes bug #313731.
615 2005-10-19 Tor Lillqvist <tml@novell.com>
617 * glib/gwin32.c (g_win32_get_package_installation_directory):
618 Return a g_strdup()ed copy of the value stored in the hash table,
619 so that it can be g_free()d without leaving a dangling pointer in
620 the hash table. (#319232)
622 2005-10-06 Matthias Clasen <mclasen@redhat.com>
624 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
625 all-zero sequences correctly. (#317930, Sebastien Bacher)
627 2005-10-05 Matthias Clasen <mclasen@redhat.com>
629 * glib/gmarkup.c (g_markup_escape_text): Doc addition
631 * Makefile.am: Add Collation tests.
633 * tests/collate/*: Inputs and expected outputs for collation tests.
635 * tests/run-collate-tests.sh: Script to run collation tests.
637 * tests/unicode-collate.c (main): Rework slightly to make
638 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
640 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
642 * docs/reference/glib/tmpl/unicode.sgml:
643 * glib/gen-unicode-tables.pl:
645 * glib/gunichartables.h:
647 * tests/casefold.txt:
648 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
649 GUnicodeBreakType types. That may break some applications, like
652 2005-09-26 Matthias Clasen <mclasen@redhat.com>
654 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
655 docs. (#314393, Matthew F. Barnes)
658 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
659 they are already declared in glib.h. This doesn't break documented
660 use of gprintf.h, but should probably be pointed out in the
661 release notes for 2.10. (#314232, Behdad Esfahbod)
663 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
665 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
666 complexity optimization, based on a patch from Matthias Clasen.
668 * tests/patterntest.c: more tests, mostly from matthias.
670 2005-09-20 Matthias Clasen <mclasen@redhat.com>
672 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
673 (#316703, Mark Drago)
675 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
677 * glib/gpattern.c: applied a patch from matthias which checks on the
678 upper bound of GPatternSpec length to optimize matches.
681 * tests/patterntest.c: added more match cases.
683 2005-09-16 Tor Lillqvist <tml@novell.com>
685 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
686 warning to doc comment that these functions should not be used on
687 encodings like CP932.
689 2005-09-14 Matthias Clasen <mclasen@redhat.com>
691 * tests/keyfile-test.c: Add a test for grup names of length 1.
693 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
694 of length 1. (#316309)
696 2005-09-12 Matthias Clasen <mclasen@redhat.com>
698 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
700 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
702 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
703 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
706 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
708 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
709 (child_watch_helper_thread): Remove some dead code. Closes
712 2005-09-07 Tor Lillqvist <tml@novell.com>
714 * glib/Makefile.am: Create also a console version of the
715 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
716 It's otherwise identical to gspawn-win32-helper.exe, except marked
717 as a console application (linked without the -mwindows option).
719 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
720 Drop the dont_wait parameter. Its truth value correlated 100% with
721 the NULLness of the exit_status parameter anyway, so it's enough
722 to check whether exit_status is NULL. Invert the sense of the
723 dont_return_handle parameter and rename it to do_return_handle, to
724 make the code easier to read by avoiding double negations.
726 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
727 to do_spawn_with_pipes() accordingly.
729 (do_spawn_with_pipes): If we have a console, use the console
730 version of the helper program, otherwise use the GUI one. This
731 avoids extra console windows opening up in some situations. (In
732 case a console application uses the GUI gspawn-win32-helper.exe to
733 spawn another console application we would get a separate console
734 for the spawned console application).
736 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
738 2005-09-05 Matthias Clasen <mclasen@redhat.com>
740 * glib/gmappedfile.c (g_mapped_file_new): Report an error
741 if the file is too large. (#315275, Kjartan Maraas)
743 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
744 of read() is signed. (#315273, Kjartan Maraas)
746 2005-08-31 Tor Lillqvist <tml@novell.com>
748 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
749 the EXE and each DLL have their own atexit function chains.
751 #define g_atexit as atexit instead. This means it has a
752 better chance of doing what the caller wants. For instance,
753 gtkhtml calls g_atexit() registering a function in gtkhtml
754 itself. This caused a crash when g_atexit() was implemented as a
755 function in the GLib DLL. The gtkhtml DLL was already unloaded by
756 the time the GLib DLL got unloaded.
758 * glib/gutils.c: #undef the #define mentioned above, to also get a
759 real g_atexit() into the DLL for backward compatibility. Document
760 the Windows behaviour of g_atexit(), and document the varying ways
761 atexit() can behave in the context of dynamically loaded modules
764 2005-08-31 Matthias Clasen <mclasen@redhat.com>
768 * glib/gdataset.c: Add string interning functions.
770 2005-08-28 Matthias Clasen <mclasen@redhat.com>
772 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
775 2005-08-26 Matthias Clasen <mclasen@redhat.com>
777 * configure.in: Bump version to 2.9.0
779 2005-08-25 Tor Lillqvist <tml@novell.com>
781 Make also the g_spawn*() functions take parameters in the GLib
782 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
783 API or ABI. Like the other GLib API that was earlier changed to
784 use UTF-8 on Windows, the names of the functions that take UTF-8
785 have _utf8 suffixes added by using preprocessor macros in the
786 header file. The old names are kept for functions with the old
787 behaviour, taking parameters in the system codepage, for DLL ABI
790 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
791 g_spawn*() functions.
793 * glib/gspawn-win32.c: Use wide-char API on NT-based
794 Windows. Convert parameters from UTF-8 to wide chars (NT) or
795 system codepage (Win9x) and call the C library _wspawn*() or
796 spawn*() functions respectvely. Add DLL ABI stability versions
797 that take parameters in the system codepage.
799 * glib/gspawn-win32-helper.c: On NT-based Windows use the
800 wide-char versions of argv and envp, and use wide-char API to
801 change directory and spawn the program to run. Remove the verbose
802 debugging output, it was too complex to modify for the wide-char
803 features. (Just add temporary debugging printouts if needed, no
804 need to have them permanently in the source.)
806 * glib/gspawn.c: Corresponding documentation updates.
808 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
809 symbols as PRIVATE, add the new _utf8-suffixed ones.
811 2005-08-24 Stepan Kasal <kasal@ucw.cz>
813 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
814 guint64 otherwise and that can produce warnings about comparison
815 between signed and unsigned.
817 2005-08-23 Matthias Clasen <mclasen@redhat.com>
819 * glib/gutils.c: Fix the crt_externs.h include.
821 2005-08-23 Stepan Kasal <kasal@ucw.cz>
823 * NEWS: Fix spelling of my first name.
825 2005-08-23 Matthias Clasen <mclasen@redhat.com>
829 * === Released 2.8.1 ===
833 2005-08-20 Hans Breuer <hans@breuer.org>
835 * glib/makefile.msc.in : link with ws2_32.lib
837 2005-08-18 Tor Lillqvist <tml@novell.com>
839 * configure.in: Check for <sys/wait.h>
841 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
843 2005-08-18 Ross Burton <ross@burtonini.com>
846 Optimise single-character insertions.
851 * tests/string-test.c:
852 Add tests for new optimisation, and fix a leak.
854 2005-08-17 Matthias Clasen <mclasen@redhat.com>
856 * configure.in: Check for crt_externs.h and _NSGetEnviron.
858 * glib/gutils.c: On Darwin, include crt-externs.h and
859 define environ using _NSGetEnviron(). (#313731)
861 2005-08-16 Stepan Kasal <kasal@ucw.cz>
863 * glib/gutils.c (g_get_any_init): Move the body of the big if...
864 (g_get_any_init_do): ... to this new function.
865 (g_get_any_init): Declare as inline.
866 (g_get_any_init_locked): New inline function, does the locking.
867 Make use of these two throughout the code.
869 2005-08-15 Matthias Clasen <mclasen@redhat.com>
871 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
872 the child process and then simply return. This makes
873 The "S" option work as documented in g_on_error_query().
874 (#313125, Matthew F. Barnes)
876 * glib/gunicode.h: Update the link to Unicode category
877 values. (#313369, Behnam Esfahbod)
879 * glib/gqueue.c (g_queue_find_custom): Clarify docs
880 a little. (#311727, Tristan van Berkom)
882 * glib/abicheck.sh, gobject/abicheck.sh: Make the
883 check work on ia64 too, where some symbols ended up
884 in yet another section.