1 2005-12-04 Matthias Clasen <mclasen@redhat.com>
3 Handle multiple user names with the same UID better.
4 (#319535, Laszlo Peter)
6 * glib/gutils.c (g_get_any_init_do): When determining user
7 data, first look up $LOGNAME. If the UID doesn't match
8 getuid(), fall back to the current behaviour of looking
9 up the user data based on getuid().
11 2005-12-04 Matthias Clasen <mclasen@redhat.com>
13 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
14 to make gcc warn if a function result is ignored. (#145466,
15 Arjan van de Ven, Alex Larsson)
17 * glib/gmem.h: Add the new attribute to g_realloc and
20 2005-12-03 Matthias Clasen <mclasen@redhat.com>
23 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
26 * glib/gmappedfile.c: Make mapping of empty files
29 * glib/gfileutils.c: Don't fork a new process just to
30 fix the permissions of the created temp file. (#321318,
31 Alexis S. L. Carvalho)
33 2005-12-02 Matthias Clasen <mclasen@redhat.com>
35 * README.in: Add a note about Solaris threads.
38 * configure.in: Remove support for Solaris threads.
39 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
41 2005-12-02 Matthias Clasen <mclasen@redhat.com>
43 * configure.in: Define G_GUINT64_CONSTANT in analogy to
44 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
46 2005-12-02 Matthias Clasen <mclasen@redhat.com>
48 * glib/gslice.c: Win32 portability fixes and C99ism removal,
49 pointed out by Kazuki Iwamoto. (#323052)
51 2005-12-02 Matthias Clasen <mclasen@redhat.com>
53 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
54 the IA64 implementation.
56 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
58 * glib/gslice.c: some naming and type size fixups.
60 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
62 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
63 variant into something less verbose, but digestible for gcc-3.4.
65 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
67 * tests/slice-test.c: extended to perform the benchmarking on the old
68 memchunk code if 'O' is selected.
70 * tests/memchunks.c: new file which contains the old GLib mem chunks
71 implementation with prefix old_mem_chunk_.
73 * tests/Makefile.am: added memchunks.c
75 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
77 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
78 acceptable load distribution of the syscall.
79 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
80 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
82 2005-12-01 Matthias Clasen <mclasen@redhat.com>
84 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
85 double-free base interface members.
87 * tests/Makefile.am: Remove duplicate variable.
89 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
91 * glib/gslice.[hc]: new slice allocator implementation.
93 * tests/slice-test.c: added random slice allocation test.
95 * glib/gthread.[hc]: removed newly added private thread mem API.
100 * glib/gmem.c: divided glib threading initialisation into three phases,
101 initialisation where private keys and messaging are not available (only
102 needed by gmem.c), initialisation without messaging but private keys
103 available (gslice.c, gmessage.c), and full fledged initialisers that
104 server the rest of glib. initialisation functions got renamed to reflect
105 the limitations of their corresponding phases.
107 * glib/gmem.c: removed memchunk code, defer allocations to
110 * glib/gmem.[hc]: removed g_slice_* skeletons.
112 * glib/glib.symbols: added g_slice_* symbols.
114 * configure.in: check for availability of posix_memalign(3), memalign(3)
117 * glib/Makefile.am: added gslice.[hc].
119 2005-12-01 Tor Lillqvist <tml@novell.com>
121 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
122 trailing slash(es) for non-root folders. stat() fails if non-root
123 folders are specified with trailing slashes. It's too much hassle
124 to demand that callers strip such slashes themselves, especially
125 as it is easy to get it wrong and strip the slash of a root
127 (g_rename): On NT-based Windows, use MoveFileEx() with
128 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
130 2005-11-28 Matthias Clasen <mclasen@redhat.com>
132 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
135 * configure.in: Check whether do { } while (0) works.
137 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
138 G_STMT_END if it works.
140 2005-11-28 Matthias Clasen <mclasen@redhat.com>
142 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
143 * glib/gnode.c (g_node_depth_traverse_level):
144 * glib/gmem.c (g_allocator_new):
145 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
146 (g_hash_table_foreach_remove_or_steal): Silence compiler
149 2005-11-27 Matthias Clasen <mclasen@redhat.com>
151 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
152 Don't read beyond len. (#322520, Christian Persch)
154 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
156 * glib/gdataset.c: access datalist flags via atomic pointer access
157 functions, instead of acquiring the dataset lock. this is faster and
158 also matches the atomic pointer readouts in gdatalistprivate.h and
159 direct pointer modifications required by gobject.c.
161 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
163 * glib/gdataset.c: streamlined initialization code somewhat,
164 removed GData node cache.
166 2005-11-22 Matthias Clasen <mclasen@redhat.com>
168 * glib/ghash.c (g_hash_table_ref):
169 (g_hash_table_unref): Mark these as new API.
171 Avoid double locking in g_intern_string (#322133,
174 * glib/gdataset.c (g_quark_from_string_internal): New
175 internal function which factors out common parts of
176 g_quark_from[_static]_string.
177 (g_quark_from_string, g_quark_from_static_string):
178 Use g_quark_from_string_internal.
179 (g_intern_string, g_intern_static_string): Use
180 g_quark_from_string_internal, and only take the
182 (g_quark_new): Don't store the strings shifted by -1
183 in the g_quarks array.
184 (g_quark_to_string): Adapt to the previous change.
186 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
190 g_hash_table_new_full(): create hash tables with a ref count of 1.
191 g_hash_table_ref(): atomically ref_count+=1
192 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
193 when refcount reaches 0.
194 g_hash_table_destroy(): just destroy keys and values, unref by 1.
195 g_hash_table_insert():
196 g_hash_table_replace(): assert ref_count>0.
199 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
202 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
204 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
205 variables. (#321984, Andrew Paprocki)
207 2005-11-18 Matthias Clasen <mclasen@redhat.com>
209 * configure.in: Bump version
211 2005-11-17 Matthias Clasen <mclasen@redhat.com>
213 * === Released 2.9.0 ===
215 * glib/glib.symbols: Updates
217 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
218 hints to the docs. (Paolo Borelli)
222 * glib/gthread.c (g_thread_foreach): Mark as new api.
224 * README.in: Updates.
228 * glib/gdate.c (g_date_set_time_t):
229 (g_date_set_time_val): Add functions to set a date from
230 a time_t and from a GTimeVal.
231 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
234 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
235 and g_utf8_offset_to_pointer.
237 * glib/gutf8.c (g_utf8_pointer_to_offset)
238 (g_utf8_offset_to_pointer): Handle negative offsets, and use
239 "stutter stepping" for going backwards. (#320638, Larry
244 * glib/gthread.c: const correctness fixes, found
245 by Arjan van de Ven and gcc.
247 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
249 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
250 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
252 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
254 * glib/guniprop.c: Use bit hacks instead when checking a general
255 category value against multiple values.
257 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
259 2005-11-10 Simos Xenitellis <simos@gnome.org>
261 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
263 2005-11-05 Tor Lillqvist <tml@novell.com>
265 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
266 (#320688) Instead set error and return failure.
268 2005-11-07 Matthias Clasen <mclasen@redhat.com>
270 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
271 while holding the context lock. (#320886, Andy Wingo)
273 2005-11-07 Billy Biggs <vektor@dumbterm.net>
275 * tests/refcount/Makefile.am: Add a missing reference to libglib
276 in the LDADD for this test.
278 2005-11-07 Matthias Clasen <mclasen@redhat.com>
280 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
281 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
282 GLib 2.9. (#320755, Luca Ferretti)
284 2005-11-04 Matthias Clasen <mclasen@redhat.com>
286 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
287 use a faster lookup table for g_unichar_get_mirror_char().
288 * glib/gmirroringtable.h: The new table.
289 * glib/gunichartables.h: Remove bidi_mirroring_table.
290 * glib/gen-unicode-tables.pl: Don't generate the mirroring
292 * glib/glib-mirroring-tab/*: Sources for the program
293 which generated gmirroringtable.h.
295 * glib/glist.c: Avoid some code duplication.
297 * glib/gscanner.h: Include gdataset.h. (#320322)
299 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
300 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
301 to help find critical warnings in applications. (#320017,
304 2005-11-02 Tor Lillqvist <tml@novell.com>
306 * glib/glib.symbols: Remove large amount of trailing whitespace
307 from one line. Remove #ifdef G_UNDEFINED from around
308 g_hash_table_new and g_hash_table_new_full symbols, what was the
309 purpose of that? Add the g_slice functions.
311 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
313 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
314 added g_slice_*() API to allocate and cache small bits of memory.
315 an actuall allocator implementation for g_slice_*() is still pending.
317 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
318 changed GRealThread list to use in-structure *next; fields instead
319 of GSList, in order for thread iteration to not depenend on g_slice_*()
321 _g_thread_mem_private_get():
322 _g_thread_mem_private_set(): added accessors for private memory,
323 needed because the ordinary GPrivate implementation relies on GArray
324 and GSList and therefore indirectly on working g_slice_*() allocations.
327 g_thread_foreach(): new public API function to loop over all existing threads.
341 removed GAllocator and free list usages and accompanying locks.
342 use g_slice_*() API to allocate and cache small bits of memory.
344 * glib/ghook.h: removed GMemChunk field from public API.
347 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
349 * glib/gnode.h: deprecate allocator API.
351 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
352 be used for poll rec lists.
354 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
355 g_relation_destroy(): free all tuples from the all_tuples hash table,
356 this effectively maintains the life time track keeping of tuples.
357 g_relation_delete_tuple(): free tuples which are removed from the
358 all_tuples hash table. this fixes a temporary leak that was present
359 in the memchunk code until the destruction of the relation.
361 2005-10-29 Matthias Clasen <mclasen@redhat.com>
363 * tests/convert-test.c: Add some tests for conversions between
364 UTF-8, UCS-4 and UTF-16.
366 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
367 of len == -1, noticed by Morten Welinder.
369 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
371 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
373 2005-10-26 Matthias Clasen <mclasen@redhat.com>
375 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
376 in the error case, and add some documentation. (#319806, Morten
379 2005-10-19 Manish Singh <yosh@gimp.org>
381 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
382 config.h symbol automatically. Fixes bug #313731.
384 2005-10-19 Tor Lillqvist <tml@novell.com>
386 * glib/gwin32.c (g_win32_get_package_installation_directory):
387 Return a g_strdup()ed copy of the value stored in the hash table,
388 so that it can be g_free()d without leaving a dangling pointer in
389 the hash table. (#319232)
391 2005-10-06 Matthias Clasen <mclasen@redhat.com>
393 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
394 all-zero sequences correctly. (#317930, Sebastien Bacher)
396 2005-10-05 Matthias Clasen <mclasen@redhat.com>
398 * glib/gmarkup.c (g_markup_escape_text): Doc addition
400 * Makefile.am: Add Collation tests.
402 * tests/collate/*: Inputs and expected outputs for collation tests.
404 * tests/run-collate-tests.sh: Script to run collation tests.
406 * tests/unicode-collate.c (main): Rework slightly to make
407 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
409 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
411 * docs/reference/glib/tmpl/unicode.sgml:
412 * glib/gen-unicode-tables.pl:
414 * glib/gunichartables.h:
416 * tests/casefold.txt:
417 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
418 GUnicodeBreakType types. That may break some applications, like
421 2005-09-26 Matthias Clasen <mclasen@redhat.com>
423 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
424 docs. (#314393, Matthew F. Barnes)
427 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
428 they are already declared in glib.h. This doesn't break documented
429 use of gprintf.h, but should probably be pointed out in the
430 release notes for 2.10. (#314232, Behdad Esfahbod)
432 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
434 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
435 complexity optimization, based on a patch from Matthias Clasen.
437 * tests/patterntest.c: more tests, mostly from matthias.
439 2005-09-20 Matthias Clasen <mclasen@redhat.com>
441 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
442 (#316703, Mark Drago)
444 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
446 * glib/gpattern.c: applied a patch from matthias which checks on the
447 upper bound of GPatternSpec length to optimize matches.
450 * tests/patterntest.c: added more match cases.
452 2005-09-16 Tor Lillqvist <tml@novell.com>
454 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
455 warning to doc comment that these functions should not be used on
456 encodings like CP932.
458 2005-09-14 Matthias Clasen <mclasen@redhat.com>
460 * tests/keyfile-test.c: Add a test for grup names of length 1.
462 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
463 of length 1. (#316309)
465 2005-09-12 Matthias Clasen <mclasen@redhat.com>
467 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
469 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
471 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
472 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
475 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
477 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
478 (child_watch_helper_thread): Remove some dead code. Closes
481 2005-09-07 Tor Lillqvist <tml@novell.com>
483 * glib/Makefile.am: Create also a console version of the
484 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
485 It's otherwise identical to gspawn-win32-helper.exe, except marked
486 as a console application (linked without the -mwindows option).
488 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
489 Drop the dont_wait parameter. Its truth value correlated 100% with
490 the NULLness of the exit_status parameter anyway, so it's enough
491 to check whether exit_status is NULL. Invert the sense of the
492 dont_return_handle parameter and rename it to do_return_handle, to
493 make the code easier to read by avoiding double negations.
495 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
496 to do_spawn_with_pipes() accordingly.
498 (do_spawn_with_pipes): If we have a console, use the console
499 version of the helper program, otherwise use the GUI one. This
500 avoids extra console windows opening up in some situations. (In
501 case a console application uses the GUI gspawn-win32-helper.exe to
502 spawn another console application we would get a separate console
503 for the spawned console application).
505 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
507 2005-09-05 Matthias Clasen <mclasen@redhat.com>
509 * glib/gmappedfile.c (g_mapped_file_new): Report an error
510 if the file is too large. (#315275, Kjartan Maraas)
512 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
513 of read() is signed. (#315273, Kjartan Maraas)
515 2005-08-31 Tor Lillqvist <tml@novell.com>
517 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
518 the EXE and each DLL have their own atexit function chains.
520 #define g_atexit as atexit instead. This means it has a
521 better chance of doing what the caller wants. For instance,
522 gtkhtml calls g_atexit() registering a function in gtkhtml
523 itself. This caused a crash when g_atexit() was implemented as a
524 function in the GLib DLL. The gtkhtml DLL was already unloaded by
525 the time the GLib DLL got unloaded.
527 * glib/gutils.c: #undef the #define mentioned above, to also get a
528 real g_atexit() into the DLL for backward compatibility. Document
529 the Windows behaviour of g_atexit(), and document the varying ways
530 atexit() can behave in the context of dynamically loaded modules
533 2005-08-31 Matthias Clasen <mclasen@redhat.com>
537 * glib/gdataset.c: Add string interning functions.
539 2005-08-28 Matthias Clasen <mclasen@redhat.com>
541 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
544 2005-08-26 Matthias Clasen <mclasen@redhat.com>
546 * configure.in: Bump version to 2.9.0
548 2005-08-25 Tor Lillqvist <tml@novell.com>
550 Make also the g_spawn*() functions take parameters in the GLib
551 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
552 API or ABI. Like the other GLib API that was earlier changed to
553 use UTF-8 on Windows, the names of the functions that take UTF-8
554 have _utf8 suffixes added by using preprocessor macros in the
555 header file. The old names are kept for functions with the old
556 behaviour, taking parameters in the system codepage, for DLL ABI
559 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
560 g_spawn*() functions.
562 * glib/gspawn-win32.c: Use wide-char API on NT-based
563 Windows. Convert parameters from UTF-8 to wide chars (NT) or
564 system codepage (Win9x) and call the C library _wspawn*() or
565 spawn*() functions respectvely. Add DLL ABI stability versions
566 that take parameters in the system codepage.
568 * glib/gspawn-win32-helper.c: On NT-based Windows use the
569 wide-char versions of argv and envp, and use wide-char API to
570 change directory and spawn the program to run. Remove the verbose
571 debugging output, it was too complex to modify for the wide-char
572 features. (Just add temporary debugging printouts if needed, no
573 need to have them permanently in the source.)
575 * glib/gspawn.c: Corresponding documentation updates.
577 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
578 symbols as PRIVATE, add the new _utf8-suffixed ones.
580 2005-08-24 Stepan Kasal <kasal@ucw.cz>
582 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
583 guint64 otherwise and that can produce warnings about comparison
584 between signed and unsigned.
586 2005-08-23 Matthias Clasen <mclasen@redhat.com>
588 * glib/gutils.c: Fix the crt_externs.h include.
590 2005-08-23 Stepan Kasal <kasal@ucw.cz>
592 * NEWS: Fix spelling of my first name.
594 2005-08-23 Matthias Clasen <mclasen@redhat.com>
598 * === Released 2.8.1 ===
602 2005-08-20 Hans Breuer <hans@breuer.org>
604 * glib/makefile.msc.in : link with ws2_32.lib
606 2005-08-18 Tor Lillqvist <tml@novell.com>
608 * configure.in: Check for <sys/wait.h>
610 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
612 2005-08-18 Ross Burton <ross@burtonini.com>
615 Optimise single-character insertions.
620 * tests/string-test.c:
621 Add tests for new optimisation, and fix a leak.
623 2005-08-17 Matthias Clasen <mclasen@redhat.com>
625 * configure.in: Check for crt_externs.h and _NSGetEnviron.
627 * glib/gutils.c: On Darwin, include crt-externs.h and
628 define environ using _NSGetEnviron(). (#313731)
630 2005-08-16 Stepan Kasal <kasal@ucw.cz>
632 * glib/gutils.c (g_get_any_init): Move the body of the big if...
633 (g_get_any_init_do): ... to this new function.
634 (g_get_any_init): Declare as inline.
635 (g_get_any_init_locked): New inline function, does the locking.
636 Make use of these two throughout the code.
638 2005-08-15 Matthias Clasen <mclasen@redhat.com>
640 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
641 the child process and then simply return. This makes
642 The "S" option work as documented in g_on_error_query().
643 (#313125, Matthew F. Barnes)
645 * glib/gunicode.h: Update the link to Unicode category
646 values. (#313369, Behnam Esfahbod)
648 * glib/gqueue.c (g_queue_find_custom): Clarify docs
649 a little. (#311727, Tristan van Berkom)
651 * glib/abicheck.sh, gobject/abicheck.sh: Make the
652 check work on ia64 too, where some symbols ended up
653 in yet another section.