1 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
3 * glib/gslice.h: Remove comma at the end of enum.
5 2005-12-04 Matthias Clasen <mclasen@redhat.com>
7 Handle multiple user names with the same UID better.
8 (#319535, Laszlo Peter)
10 * glib/gutils.c (g_get_any_init_do): When determining user
11 data, first look up $LOGNAME. If the UID doesn't match
12 getuid(), fall back to the current behaviour of looking
13 up the user data based on getuid().
15 2005-12-04 Matthias Clasen <mclasen@redhat.com>
17 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
18 to make gcc warn if a function result is ignored. (#145466,
19 Arjan van de Ven, Alex Larsson)
21 * glib/gmem.h: Add the new attribute to g_realloc and
24 2005-12-03 Matthias Clasen <mclasen@redhat.com>
27 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
30 * glib/gmappedfile.c: Make mapping of empty files
33 * glib/gfileutils.c: Don't fork a new process just to
34 fix the permissions of the created temp file. (#321318,
35 Alexis S. L. Carvalho)
37 2005-12-02 Matthias Clasen <mclasen@redhat.com>
39 * README.in: Add a note about Solaris threads.
42 * configure.in: Remove support for Solaris threads.
43 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
45 2005-12-02 Matthias Clasen <mclasen@redhat.com>
47 * configure.in: Define G_GUINT64_CONSTANT in analogy to
48 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
50 2005-12-02 Matthias Clasen <mclasen@redhat.com>
52 * glib/gslice.c: Win32 portability fixes and C99ism removal,
53 pointed out by Kazuki Iwamoto. (#323052)
55 2005-12-02 Matthias Clasen <mclasen@redhat.com>
57 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
58 the IA64 implementation.
60 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
62 * glib/gslice.c: some naming and type size fixups.
64 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
66 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
67 variant into something less verbose, but digestible for gcc-3.4.
69 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
71 * tests/slice-test.c: extended to perform the benchmarking on the old
72 memchunk code if 'O' is selected.
74 * tests/memchunks.c: new file which contains the old GLib mem chunks
75 implementation with prefix old_mem_chunk_.
77 * tests/Makefile.am: added memchunks.c
79 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
81 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
82 acceptable load distribution of the syscall.
83 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
84 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
86 2005-12-01 Matthias Clasen <mclasen@redhat.com>
88 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
89 double-free base interface members.
91 * tests/Makefile.am: Remove duplicate variable.
93 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
95 * glib/gslice.[hc]: new slice allocator implementation.
97 * tests/slice-test.c: added random slice allocation test.
99 * glib/gthread.[hc]: removed newly added private thread mem API.
101 * glib/gthreadinit.h:
104 * glib/gmem.c: divided glib threading initialisation into three phases,
105 initialisation where private keys and messaging are not available (only
106 needed by gmem.c), initialisation without messaging but private keys
107 available (gslice.c, gmessage.c), and full fledged initialisers that
108 server the rest of glib. initialisation functions got renamed to reflect
109 the limitations of their corresponding phases.
111 * glib/gmem.c: removed memchunk code, defer allocations to
114 * glib/gmem.[hc]: removed g_slice_* skeletons.
116 * glib/glib.symbols: added g_slice_* symbols.
118 * configure.in: check for availability of posix_memalign(3), memalign(3)
121 * glib/Makefile.am: added gslice.[hc].
123 2005-12-01 Tor Lillqvist <tml@novell.com>
125 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
126 trailing slash(es) for non-root folders. stat() fails if non-root
127 folders are specified with trailing slashes. It's too much hassle
128 to demand that callers strip such slashes themselves, especially
129 as it is easy to get it wrong and strip the slash of a root
131 (g_rename): On NT-based Windows, use MoveFileEx() with
132 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
134 2005-11-28 Matthias Clasen <mclasen@redhat.com>
136 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
139 * configure.in: Check whether do { } while (0) works.
141 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
142 G_STMT_END if it works.
144 2005-11-28 Matthias Clasen <mclasen@redhat.com>
146 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
147 * glib/gnode.c (g_node_depth_traverse_level):
148 * glib/gmem.c (g_allocator_new):
149 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
150 (g_hash_table_foreach_remove_or_steal): Silence compiler
153 2005-11-27 Matthias Clasen <mclasen@redhat.com>
155 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
156 Don't read beyond len. (#322520, Christian Persch)
158 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
160 * glib/gdataset.c: access datalist flags via atomic pointer access
161 functions, instead of acquiring the dataset lock. this is faster and
162 also matches the atomic pointer readouts in gdatalistprivate.h and
163 direct pointer modifications required by gobject.c.
165 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
167 * glib/gdataset.c: streamlined initialization code somewhat,
168 removed GData node cache.
170 2005-11-22 Matthias Clasen <mclasen@redhat.com>
172 * glib/ghash.c (g_hash_table_ref):
173 (g_hash_table_unref): Mark these as new API.
175 Avoid double locking in g_intern_string (#322133,
178 * glib/gdataset.c (g_quark_from_string_internal): New
179 internal function which factors out common parts of
180 g_quark_from[_static]_string.
181 (g_quark_from_string, g_quark_from_static_string):
182 Use g_quark_from_string_internal.
183 (g_intern_string, g_intern_static_string): Use
184 g_quark_from_string_internal, and only take the
186 (g_quark_new): Don't store the strings shifted by -1
187 in the g_quarks array.
188 (g_quark_to_string): Adapt to the previous change.
190 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
194 g_hash_table_new_full(): create hash tables with a ref count of 1.
195 g_hash_table_ref(): atomically ref_count+=1
196 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
197 when refcount reaches 0.
198 g_hash_table_destroy(): just destroy keys and values, unref by 1.
199 g_hash_table_insert():
200 g_hash_table_replace(): assert ref_count>0.
203 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
206 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
208 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
209 variables. (#321984, Andrew Paprocki)
211 2005-11-18 Matthias Clasen <mclasen@redhat.com>
213 * configure.in: Bump version
215 2005-11-17 Matthias Clasen <mclasen@redhat.com>
217 * === Released 2.9.0 ===
219 * glib/glib.symbols: Updates
221 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
222 hints to the docs. (Paolo Borelli)
226 * glib/gthread.c (g_thread_foreach): Mark as new api.
228 * README.in: Updates.
232 * glib/gdate.c (g_date_set_time_t):
233 (g_date_set_time_val): Add functions to set a date from
234 a time_t and from a GTimeVal.
235 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
238 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
239 and g_utf8_offset_to_pointer.
241 * glib/gutf8.c (g_utf8_pointer_to_offset)
242 (g_utf8_offset_to_pointer): Handle negative offsets, and use
243 "stutter stepping" for going backwards. (#320638, Larry
248 * glib/gthread.c: const correctness fixes, found
249 by Arjan van de Ven and gcc.
251 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
253 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
254 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
256 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
258 * glib/guniprop.c: Use bit hacks instead when checking a general
259 category value against multiple values.
261 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
263 2005-11-10 Simos Xenitellis <simos@gnome.org>
265 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
267 2005-11-05 Tor Lillqvist <tml@novell.com>
269 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
270 (#320688) Instead set error and return failure.
272 2005-11-07 Matthias Clasen <mclasen@redhat.com>
274 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
275 while holding the context lock. (#320886, Andy Wingo)
277 2005-11-07 Billy Biggs <vektor@dumbterm.net>
279 * tests/refcount/Makefile.am: Add a missing reference to libglib
280 in the LDADD for this test.
282 2005-11-07 Matthias Clasen <mclasen@redhat.com>
284 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
285 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
286 GLib 2.9. (#320755, Luca Ferretti)
288 2005-11-04 Matthias Clasen <mclasen@redhat.com>
290 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
291 use a faster lookup table for g_unichar_get_mirror_char().
292 * glib/gmirroringtable.h: The new table.
293 * glib/gunichartables.h: Remove bidi_mirroring_table.
294 * glib/gen-unicode-tables.pl: Don't generate the mirroring
296 * glib/glib-mirroring-tab/*: Sources for the program
297 which generated gmirroringtable.h.
299 * glib/glist.c: Avoid some code duplication.
301 * glib/gscanner.h: Include gdataset.h. (#320322)
303 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
304 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
305 to help find critical warnings in applications. (#320017,
308 2005-11-02 Tor Lillqvist <tml@novell.com>
310 * glib/glib.symbols: Remove large amount of trailing whitespace
311 from one line. Remove #ifdef G_UNDEFINED from around
312 g_hash_table_new and g_hash_table_new_full symbols, what was the
313 purpose of that? Add the g_slice functions.
315 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
317 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
318 added g_slice_*() API to allocate and cache small bits of memory.
319 an actuall allocator implementation for g_slice_*() is still pending.
321 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
322 changed GRealThread list to use in-structure *next; fields instead
323 of GSList, in order for thread iteration to not depenend on g_slice_*()
325 _g_thread_mem_private_get():
326 _g_thread_mem_private_set(): added accessors for private memory,
327 needed because the ordinary GPrivate implementation relies on GArray
328 and GSList and therefore indirectly on working g_slice_*() allocations.
331 g_thread_foreach(): new public API function to loop over all existing threads.
345 removed GAllocator and free list usages and accompanying locks.
346 use g_slice_*() API to allocate and cache small bits of memory.
348 * glib/ghook.h: removed GMemChunk field from public API.
351 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
353 * glib/gnode.h: deprecate allocator API.
355 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
356 be used for poll rec lists.
358 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
359 g_relation_destroy(): free all tuples from the all_tuples hash table,
360 this effectively maintains the life time track keeping of tuples.
361 g_relation_delete_tuple(): free tuples which are removed from the
362 all_tuples hash table. this fixes a temporary leak that was present
363 in the memchunk code until the destruction of the relation.
365 2005-10-29 Matthias Clasen <mclasen@redhat.com>
367 * tests/convert-test.c: Add some tests for conversions between
368 UTF-8, UCS-4 and UTF-16.
370 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
371 of len == -1, noticed by Morten Welinder.
373 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
375 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
377 2005-10-26 Matthias Clasen <mclasen@redhat.com>
379 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
380 in the error case, and add some documentation. (#319806, Morten
383 2005-10-19 Manish Singh <yosh@gimp.org>
385 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
386 config.h symbol automatically. Fixes bug #313731.
388 2005-10-19 Tor Lillqvist <tml@novell.com>
390 * glib/gwin32.c (g_win32_get_package_installation_directory):
391 Return a g_strdup()ed copy of the value stored in the hash table,
392 so that it can be g_free()d without leaving a dangling pointer in
393 the hash table. (#319232)
395 2005-10-06 Matthias Clasen <mclasen@redhat.com>
397 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
398 all-zero sequences correctly. (#317930, Sebastien Bacher)
400 2005-10-05 Matthias Clasen <mclasen@redhat.com>
402 * glib/gmarkup.c (g_markup_escape_text): Doc addition
404 * Makefile.am: Add Collation tests.
406 * tests/collate/*: Inputs and expected outputs for collation tests.
408 * tests/run-collate-tests.sh: Script to run collation tests.
410 * tests/unicode-collate.c (main): Rework slightly to make
411 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
413 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
415 * docs/reference/glib/tmpl/unicode.sgml:
416 * glib/gen-unicode-tables.pl:
418 * glib/gunichartables.h:
420 * tests/casefold.txt:
421 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
422 GUnicodeBreakType types. That may break some applications, like
425 2005-09-26 Matthias Clasen <mclasen@redhat.com>
427 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
428 docs. (#314393, Matthew F. Barnes)
431 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
432 they are already declared in glib.h. This doesn't break documented
433 use of gprintf.h, but should probably be pointed out in the
434 release notes for 2.10. (#314232, Behdad Esfahbod)
436 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
438 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
439 complexity optimization, based on a patch from Matthias Clasen.
441 * tests/patterntest.c: more tests, mostly from matthias.
443 2005-09-20 Matthias Clasen <mclasen@redhat.com>
445 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
446 (#316703, Mark Drago)
448 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
450 * glib/gpattern.c: applied a patch from matthias which checks on the
451 upper bound of GPatternSpec length to optimize matches.
454 * tests/patterntest.c: added more match cases.
456 2005-09-16 Tor Lillqvist <tml@novell.com>
458 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
459 warning to doc comment that these functions should not be used on
460 encodings like CP932.
462 2005-09-14 Matthias Clasen <mclasen@redhat.com>
464 * tests/keyfile-test.c: Add a test for grup names of length 1.
466 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
467 of length 1. (#316309)
469 2005-09-12 Matthias Clasen <mclasen@redhat.com>
471 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
473 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
475 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
476 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
479 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
481 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
482 (child_watch_helper_thread): Remove some dead code. Closes
485 2005-09-07 Tor Lillqvist <tml@novell.com>
487 * glib/Makefile.am: Create also a console version of the
488 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
489 It's otherwise identical to gspawn-win32-helper.exe, except marked
490 as a console application (linked without the -mwindows option).
492 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
493 Drop the dont_wait parameter. Its truth value correlated 100% with
494 the NULLness of the exit_status parameter anyway, so it's enough
495 to check whether exit_status is NULL. Invert the sense of the
496 dont_return_handle parameter and rename it to do_return_handle, to
497 make the code easier to read by avoiding double negations.
499 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
500 to do_spawn_with_pipes() accordingly.
502 (do_spawn_with_pipes): If we have a console, use the console
503 version of the helper program, otherwise use the GUI one. This
504 avoids extra console windows opening up in some situations. (In
505 case a console application uses the GUI gspawn-win32-helper.exe to
506 spawn another console application we would get a separate console
507 for the spawned console application).
509 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
511 2005-09-05 Matthias Clasen <mclasen@redhat.com>
513 * glib/gmappedfile.c (g_mapped_file_new): Report an error
514 if the file is too large. (#315275, Kjartan Maraas)
516 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
517 of read() is signed. (#315273, Kjartan Maraas)
519 2005-08-31 Tor Lillqvist <tml@novell.com>
521 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
522 the EXE and each DLL have their own atexit function chains.
524 #define g_atexit as atexit instead. This means it has a
525 better chance of doing what the caller wants. For instance,
526 gtkhtml calls g_atexit() registering a function in gtkhtml
527 itself. This caused a crash when g_atexit() was implemented as a
528 function in the GLib DLL. The gtkhtml DLL was already unloaded by
529 the time the GLib DLL got unloaded.
531 * glib/gutils.c: #undef the #define mentioned above, to also get a
532 real g_atexit() into the DLL for backward compatibility. Document
533 the Windows behaviour of g_atexit(), and document the varying ways
534 atexit() can behave in the context of dynamically loaded modules
537 2005-08-31 Matthias Clasen <mclasen@redhat.com>
541 * glib/gdataset.c: Add string interning functions.
543 2005-08-28 Matthias Clasen <mclasen@redhat.com>
545 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
548 2005-08-26 Matthias Clasen <mclasen@redhat.com>
550 * configure.in: Bump version to 2.9.0
552 2005-08-25 Tor Lillqvist <tml@novell.com>
554 Make also the g_spawn*() functions take parameters in the GLib
555 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
556 API or ABI. Like the other GLib API that was earlier changed to
557 use UTF-8 on Windows, the names of the functions that take UTF-8
558 have _utf8 suffixes added by using preprocessor macros in the
559 header file. The old names are kept for functions with the old
560 behaviour, taking parameters in the system codepage, for DLL ABI
563 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
564 g_spawn*() functions.
566 * glib/gspawn-win32.c: Use wide-char API on NT-based
567 Windows. Convert parameters from UTF-8 to wide chars (NT) or
568 system codepage (Win9x) and call the C library _wspawn*() or
569 spawn*() functions respectvely. Add DLL ABI stability versions
570 that take parameters in the system codepage.
572 * glib/gspawn-win32-helper.c: On NT-based Windows use the
573 wide-char versions of argv and envp, and use wide-char API to
574 change directory and spawn the program to run. Remove the verbose
575 debugging output, it was too complex to modify for the wide-char
576 features. (Just add temporary debugging printouts if needed, no
577 need to have them permanently in the source.)
579 * glib/gspawn.c: Corresponding documentation updates.
581 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
582 symbols as PRIVATE, add the new _utf8-suffixed ones.
584 2005-08-24 Stepan Kasal <kasal@ucw.cz>
586 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
587 guint64 otherwise and that can produce warnings about comparison
588 between signed and unsigned.
590 2005-08-23 Matthias Clasen <mclasen@redhat.com>
592 * glib/gutils.c: Fix the crt_externs.h include.
594 2005-08-23 Stepan Kasal <kasal@ucw.cz>
596 * NEWS: Fix spelling of my first name.
598 2005-08-23 Matthias Clasen <mclasen@redhat.com>
602 * === Released 2.8.1 ===
606 2005-08-20 Hans Breuer <hans@breuer.org>
608 * glib/makefile.msc.in : link with ws2_32.lib
610 2005-08-18 Tor Lillqvist <tml@novell.com>
612 * configure.in: Check for <sys/wait.h>
614 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
616 2005-08-18 Ross Burton <ross@burtonini.com>
619 Optimise single-character insertions.
624 * tests/string-test.c:
625 Add tests for new optimisation, and fix a leak.
627 2005-08-17 Matthias Clasen <mclasen@redhat.com>
629 * configure.in: Check for crt_externs.h and _NSGetEnviron.
631 * glib/gutils.c: On Darwin, include crt-externs.h and
632 define environ using _NSGetEnviron(). (#313731)
634 2005-08-16 Stepan Kasal <kasal@ucw.cz>
636 * glib/gutils.c (g_get_any_init): Move the body of the big if...
637 (g_get_any_init_do): ... to this new function.
638 (g_get_any_init): Declare as inline.
639 (g_get_any_init_locked): New inline function, does the locking.
640 Make use of these two throughout the code.
642 2005-08-15 Matthias Clasen <mclasen@redhat.com>
644 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
645 the child process and then simply return. This makes
646 The "S" option work as documented in g_on_error_query().
647 (#313125, Matthew F. Barnes)
649 * glib/gunicode.h: Update the link to Unicode category
650 values. (#313369, Behnam Esfahbod)
652 * glib/gqueue.c (g_queue_find_custom): Clarify docs
653 a little. (#311727, Tristan van Berkom)
655 * glib/abicheck.sh, gobject/abicheck.sh: Make the
656 check work on ia64 too, where some symbols ended up
657 in yet another section.