1 2005-12-02 Matthias Clasen <mclasen@redhat.com>
3 * glib/gslice.c: Win32 portability fixes and C99ism removal,
4 pointed out by Kazuki Iwamoto. (#323052)
6 2005-12-02 Matthias Clasen <mclasen@redhat.com>
8 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
9 the IA64 implementation.
11 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
13 * glib/gslice.c: some naming and type size fixups.
15 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
17 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
18 variant into something less verbose, but digestible for gcc-3.4.
20 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
22 * tests/slice-test.c: extended to perform the benchmarking on the old
23 memchunk code if 'O' is selected.
25 * tests/memchunks.c: new file which contains the old GLib mem chunks
26 implementation with prefix old_mem_chunk_.
28 * tests/Makefile.am: added memchunks.c
30 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
32 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
33 acceptable load distribution of the syscall.
34 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
35 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
37 2005-12-01 Matthias Clasen <mclasen@redhat.com>
39 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
40 double-free base interface members.
42 * tests/Makefile.am: Remove duplicate variable.
44 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
46 * glib/gslice.[hc]: new slice allocator implementation.
48 * tests/slice-test.c: added random slice allocation test.
50 * glib/gthread.[hc]: removed newly added private thread mem API.
55 * glib/gmem.c: divided glib threading initialisation into three phases,
56 initialisation where private keys and messaging are not available (only
57 needed by gmem.c), initialisation without messaging but private keys
58 available (gslice.c, gmessage.c), and full fledged initialisers that
59 server the rest of glib. initialisation functions got renamed to reflect
60 the limitations of their corresponding phases.
62 * glib/gmem.c: removed memchunk code, defer allocations to
65 * glib/gmem.[hc]: removed g_slice_* skeletons.
67 * glib/glib.symbols: added g_slice_* symbols.
69 * configure.in: check for availability of posix_memalign(3), memalign(3)
72 * glib/Makefile.am: added gslice.[hc].
74 2005-12-01 Tor Lillqvist <tml@novell.com>
76 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
77 trailing slash(es) for non-root folders. stat() fails if non-root
78 folders are specified with trailing slashes. It's too much hassle
79 to demand that callers strip such slashes themselves, especially
80 as it is easy to get it wrong and strip the slash of a root
82 (g_rename): On NT-based Windows, use MoveFileEx() with
83 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
85 2005-11-28 Matthias Clasen <mclasen@redhat.com>
87 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
90 * configure.in: Check whether do { } while (0) works.
92 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
93 G_STMT_END if it works.
95 2005-11-28 Matthias Clasen <mclasen@redhat.com>
97 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
98 * glib/gnode.c (g_node_depth_traverse_level):
99 * glib/gmem.c (g_allocator_new):
100 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
101 (g_hash_table_foreach_remove_or_steal): Silence compiler
104 2005-11-27 Matthias Clasen <mclasen@redhat.com>
106 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
107 Don't read beyond len. (#322520, Christian Persch)
109 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
111 * glib/gdataset.c: access datalist flags via atomic pointer access
112 functions, instead of acquiring the dataset lock. this is faster and
113 also matches the atomic pointer readouts in gdatalistprivate.h and
114 direct pointer modifications required by gobject.c.
116 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
118 * glib/gdataset.c: streamlined initialization code somewhat,
119 removed GData node cache.
121 2005-11-22 Matthias Clasen <mclasen@redhat.com>
123 * glib/ghash.c (g_hash_table_ref):
124 (g_hash_table_unref): Mark these as new API.
126 Avoid double locking in g_intern_string (#322133,
129 * glib/gdataset.c (g_quark_from_string_internal): New
130 internal function which factors out common parts of
131 g_quark_from[_static]_string.
132 (g_quark_from_string, g_quark_from_static_string):
133 Use g_quark_from_string_internal.
134 (g_intern_string, g_intern_static_string): Use
135 g_quark_from_string_internal, and only take the
137 (g_quark_new): Don't store the strings shifted by -1
138 in the g_quarks array.
139 (g_quark_to_string): Adapt to the previous change.
141 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
145 g_hash_table_new_full(): create hash tables with a ref count of 1.
146 g_hash_table_ref(): atomically ref_count+=1
147 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
148 when refcount reaches 0.
149 g_hash_table_destroy(): just destroy keys and values, unref by 1.
150 g_hash_table_insert():
151 g_hash_table_replace(): assert ref_count>0.
154 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
157 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
159 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
160 variables. (#321984, Andrew Paprocki)
162 2005-11-18 Matthias Clasen <mclasen@redhat.com>
164 * configure.in: Bump version
166 2005-11-17 Matthias Clasen <mclasen@redhat.com>
168 * === Released 2.9.0 ===
170 * glib/glib.symbols: Updates
172 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
173 hints to the docs. (Paolo Borelli)
177 * glib/gthread.c (g_thread_foreach): Mark as new api.
179 * README.in: Updates.
183 * glib/gdate.c (g_date_set_time_t):
184 (g_date_set_time_val): Add functions to set a date from
185 a time_t and from a GTimeVal.
186 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
189 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
190 and g_utf8_offset_to_pointer.
192 * glib/gutf8.c (g_utf8_pointer_to_offset)
193 (g_utf8_offset_to_pointer): Handle negative offsets, and use
194 "stutter stepping" for going backwards. (#320638, Larry
199 * glib/gthread.c: const correctness fixes, found
200 by Arjan van de Ven and gcc.
202 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
204 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
205 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
207 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
209 * glib/guniprop.c: Use bit hacks instead when checking a general
210 category value against multiple values.
212 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
214 2005-11-10 Simos Xenitellis <simos@gnome.org>
216 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
218 2005-11-05 Tor Lillqvist <tml@novell.com>
220 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
221 (#320688) Instead set error and return failure.
223 2005-11-07 Matthias Clasen <mclasen@redhat.com>
225 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
226 while holding the context lock. (#320886, Andy Wingo)
228 2005-11-07 Billy Biggs <vektor@dumbterm.net>
230 * tests/refcount/Makefile.am: Add a missing reference to libglib
231 in the LDADD for this test.
233 2005-11-07 Matthias Clasen <mclasen@redhat.com>
235 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
236 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
237 GLib 2.9. (#320755, Luca Ferretti)
239 2005-11-04 Matthias Clasen <mclasen@redhat.com>
241 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
242 use a faster lookup table for g_unichar_get_mirror_char().
243 * glib/gmirroringtable.h: The new table.
244 * glib/gunichartables.h: Remove bidi_mirroring_table.
245 * glib/gen-unicode-tables.pl: Don't generate the mirroring
247 * glib/glib-mirroring-tab/*: Sources for the program
248 which generated gmirroringtable.h.
250 * glib/glist.c: Avoid some code duplication.
252 * glib/gscanner.h: Include gdataset.h. (#320322)
254 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
255 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
256 to help find critical warnings in applications. (#320017,
259 2005-11-02 Tor Lillqvist <tml@novell.com>
261 * glib/glib.symbols: Remove large amount of trailing whitespace
262 from one line. Remove #ifdef G_UNDEFINED from around
263 g_hash_table_new and g_hash_table_new_full symbols, what was the
264 purpose of that? Add the g_slice functions.
266 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
268 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
269 added g_slice_*() API to allocate and cache small bits of memory.
270 an actuall allocator implementation for g_slice_*() is still pending.
272 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
273 changed GRealThread list to use in-structure *next; fields instead
274 of GSList, in order for thread iteration to not depenend on g_slice_*()
276 _g_thread_mem_private_get():
277 _g_thread_mem_private_set(): added accessors for private memory,
278 needed because the ordinary GPrivate implementation relies on GArray
279 and GSList and therefore indirectly on working g_slice_*() allocations.
282 g_thread_foreach(): new public API function to loop over all existing threads.
296 removed GAllocator and free list usages and accompanying locks.
297 use g_slice_*() API to allocate and cache small bits of memory.
299 * glib/ghook.h: removed GMemChunk field from public API.
302 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
304 * glib/gnode.h: deprecate allocator API.
306 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
307 be used for poll rec lists.
309 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
310 g_relation_destroy(): free all tuples from the all_tuples hash table,
311 this effectively maintains the life time track keeping of tuples.
312 g_relation_delete_tuple(): free tuples which are removed from the
313 all_tuples hash table. this fixes a temporary leak that was present
314 in the memchunk code until the destruction of the relation.
316 2005-10-29 Matthias Clasen <mclasen@redhat.com>
318 * tests/convert-test.c: Add some tests for conversions between
319 UTF-8, UCS-4 and UTF-16.
321 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
322 of len == -1, noticed by Morten Welinder.
324 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
326 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
328 2005-10-26 Matthias Clasen <mclasen@redhat.com>
330 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
331 in the error case, and add some documentation. (#319806, Morten
334 2005-10-19 Manish Singh <yosh@gimp.org>
336 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
337 config.h symbol automatically. Fixes bug #313731.
339 2005-10-19 Tor Lillqvist <tml@novell.com>
341 * glib/gwin32.c (g_win32_get_package_installation_directory):
342 Return a g_strdup()ed copy of the value stored in the hash table,
343 so that it can be g_free()d without leaving a dangling pointer in
344 the hash table. (#319232)
346 2005-10-06 Matthias Clasen <mclasen@redhat.com>
348 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
349 all-zero sequences correctly. (#317930, Sebastien Bacher)
351 2005-10-05 Matthias Clasen <mclasen@redhat.com>
353 * glib/gmarkup.c (g_markup_escape_text): Doc addition
355 * Makefile.am: Add Collation tests.
357 * tests/collate/*: Inputs and expected outputs for collation tests.
359 * tests/run-collate-tests.sh: Script to run collation tests.
361 * tests/unicode-collate.c (main): Rework slightly to make
362 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
364 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
366 * docs/reference/glib/tmpl/unicode.sgml:
367 * glib/gen-unicode-tables.pl:
369 * glib/gunichartables.h:
371 * tests/casefold.txt:
372 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
373 GUnicodeBreakType types. That may break some applications, like
376 2005-09-26 Matthias Clasen <mclasen@redhat.com>
378 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
379 docs. (#314393, Matthew F. Barnes)
382 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
383 they are already declared in glib.h. This doesn't break documented
384 use of gprintf.h, but should probably be pointed out in the
385 release notes for 2.10. (#314232, Behdad Esfahbod)
387 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
389 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
390 complexity optimization, based on a patch from Matthias Clasen.
392 * tests/patterntest.c: more tests, mostly from matthias.
394 2005-09-20 Matthias Clasen <mclasen@redhat.com>
396 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
397 (#316703, Mark Drago)
399 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
401 * glib/gpattern.c: applied a patch from matthias which checks on the
402 upper bound of GPatternSpec length to optimize matches.
405 * tests/patterntest.c: added more match cases.
407 2005-09-16 Tor Lillqvist <tml@novell.com>
409 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
410 warning to doc comment that these functions should not be used on
411 encodings like CP932.
413 2005-09-14 Matthias Clasen <mclasen@redhat.com>
415 * tests/keyfile-test.c: Add a test for grup names of length 1.
417 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
418 of length 1. (#316309)
420 2005-09-12 Matthias Clasen <mclasen@redhat.com>
422 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
424 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
426 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
427 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
430 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
432 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
433 (child_watch_helper_thread): Remove some dead code. Closes
436 2005-09-07 Tor Lillqvist <tml@novell.com>
438 * glib/Makefile.am: Create also a console version of the
439 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
440 It's otherwise identical to gspawn-win32-helper.exe, except marked
441 as a console application (linked without the -mwindows option).
443 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
444 Drop the dont_wait parameter. Its truth value correlated 100% with
445 the NULLness of the exit_status parameter anyway, so it's enough
446 to check whether exit_status is NULL. Invert the sense of the
447 dont_return_handle parameter and rename it to do_return_handle, to
448 make the code easier to read by avoiding double negations.
450 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
451 to do_spawn_with_pipes() accordingly.
453 (do_spawn_with_pipes): If we have a console, use the console
454 version of the helper program, otherwise use the GUI one. This
455 avoids extra console windows opening up in some situations. (In
456 case a console application uses the GUI gspawn-win32-helper.exe to
457 spawn another console application we would get a separate console
458 for the spawned console application).
460 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
462 2005-09-05 Matthias Clasen <mclasen@redhat.com>
464 * glib/gmappedfile.c (g_mapped_file_new): Report an error
465 if the file is too large. (#315275, Kjartan Maraas)
467 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
468 of read() is signed. (#315273, Kjartan Maraas)
470 2005-08-31 Tor Lillqvist <tml@novell.com>
472 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
473 the EXE and each DLL have their own atexit function chains.
475 #define g_atexit as atexit instead. This means it has a
476 better chance of doing what the caller wants. For instance,
477 gtkhtml calls g_atexit() registering a function in gtkhtml
478 itself. This caused a crash when g_atexit() was implemented as a
479 function in the GLib DLL. The gtkhtml DLL was already unloaded by
480 the time the GLib DLL got unloaded.
482 * glib/gutils.c: #undef the #define mentioned above, to also get a
483 real g_atexit() into the DLL for backward compatibility. Document
484 the Windows behaviour of g_atexit(), and document the varying ways
485 atexit() can behave in the context of dynamically loaded modules
488 2005-08-31 Matthias Clasen <mclasen@redhat.com>
492 * glib/gdataset.c: Add string interning functions.
494 2005-08-28 Matthias Clasen <mclasen@redhat.com>
496 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
499 2005-08-26 Matthias Clasen <mclasen@redhat.com>
501 * configure.in: Bump version to 2.9.0
503 2005-08-25 Tor Lillqvist <tml@novell.com>
505 Make also the g_spawn*() functions take parameters in the GLib
506 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
507 API or ABI. Like the other GLib API that was earlier changed to
508 use UTF-8 on Windows, the names of the functions that take UTF-8
509 have _utf8 suffixes added by using preprocessor macros in the
510 header file. The old names are kept for functions with the old
511 behaviour, taking parameters in the system codepage, for DLL ABI
514 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
515 g_spawn*() functions.
517 * glib/gspawn-win32.c: Use wide-char API on NT-based
518 Windows. Convert parameters from UTF-8 to wide chars (NT) or
519 system codepage (Win9x) and call the C library _wspawn*() or
520 spawn*() functions respectvely. Add DLL ABI stability versions
521 that take parameters in the system codepage.
523 * glib/gspawn-win32-helper.c: On NT-based Windows use the
524 wide-char versions of argv and envp, and use wide-char API to
525 change directory and spawn the program to run. Remove the verbose
526 debugging output, it was too complex to modify for the wide-char
527 features. (Just add temporary debugging printouts if needed, no
528 need to have them permanently in the source.)
530 * glib/gspawn.c: Corresponding documentation updates.
532 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
533 symbols as PRIVATE, add the new _utf8-suffixed ones.
535 2005-08-24 Stepan Kasal <kasal@ucw.cz>
537 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
538 guint64 otherwise and that can produce warnings about comparison
539 between signed and unsigned.
541 2005-08-23 Matthias Clasen <mclasen@redhat.com>
543 * glib/gutils.c: Fix the crt_externs.h include.
545 2005-08-23 Stepan Kasal <kasal@ucw.cz>
547 * NEWS: Fix spelling of my first name.
549 2005-08-23 Matthias Clasen <mclasen@redhat.com>
553 * === Released 2.8.1 ===
557 2005-08-20 Hans Breuer <hans@breuer.org>
559 * glib/makefile.msc.in : link with ws2_32.lib
561 2005-08-18 Tor Lillqvist <tml@novell.com>
563 * configure.in: Check for <sys/wait.h>
565 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
567 2005-08-18 Ross Burton <ross@burtonini.com>
570 Optimise single-character insertions.
575 * tests/string-test.c:
576 Add tests for new optimisation, and fix a leak.
578 2005-08-17 Matthias Clasen <mclasen@redhat.com>
580 * configure.in: Check for crt_externs.h and _NSGetEnviron.
582 * glib/gutils.c: On Darwin, include crt-externs.h and
583 define environ using _NSGetEnviron(). (#313731)
585 2005-08-16 Stepan Kasal <kasal@ucw.cz>
587 * glib/gutils.c (g_get_any_init): Move the body of the big if...
588 (g_get_any_init_do): ... to this new function.
589 (g_get_any_init): Declare as inline.
590 (g_get_any_init_locked): New inline function, does the locking.
591 Make use of these two throughout the code.
593 2005-08-15 Matthias Clasen <mclasen@redhat.com>
595 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
596 the child process and then simply return. This makes
597 The "S" option work as documented in g_on_error_query().
598 (#313125, Matthew F. Barnes)
600 * glib/gunicode.h: Update the link to Unicode category
601 values. (#313369, Behnam Esfahbod)
603 * glib/gqueue.c (g_queue_find_custom): Clarify docs
604 a little. (#311727, Tristan van Berkom)
606 * glib/abicheck.sh, gobject/abicheck.sh: Make the
607 check work on ia64 too, where some symbols ended up
608 in yet another section.