1 2008-10-16 Matthias Clasen <mclasen@redhat.com>
3 * configure.in: Bump version to 2.19.1
5 * === Released 2.19.0 ===
9 2008-10-14 Christian Persch <chpe@gnome.org>
11 Bug 556186 – gpoll.h breaks gmain.h inclusion
13 * glib/gpoll.h: Only disallow direct inclusion when
14 G_DISABLE_SINGLE_INCLUDES is defined.
16 2008-10-13 Sven Herzberg <sven@imendio.com>
18 Bug 556101 – static mutex yields warnings with g++
21 * configure.in: added an intermediate cast to gpointer
23 2008-10-13 Christian Persch <chpe@gnome.org>
25 Bug 555311 – format not a string literal and no format arguments
28 * glib/gshell.c: Use literal errors where appropriate.
30 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
32 Bug 551355 – [PATCH] Make glib build with libtool 2.2
34 * autogen.sh: Accept libtool 2.2. We are moving towards having
37 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
39 * configure.in: With libtool 2.x, regenerate libtool early.
40 Based on bug 552107, patch by Patryk Zawadzki
42 2008-10-10 Matthias Clasen <mclasen@redhat.com>
44 Bug 554557 – Patch to fix gcc warnings about missing format
49 * glib/grel.c: Fix gcc warnings. Patch by Kjartan Maraas
51 2008-10-10 Matthias Clasen <mclasen@redhat.com>
53 Bug 552861 – glib-2.0.m4 calls system(3) without storing its result
55 * m4macros/glib-2.0.m4: Cosmetic change to make -Werror happy.
56 Patch by Andreas Köhler
58 2008-10-08 Christian Persch <chpe@gnome.org>
60 Bug 555313 – GFileAttribute boxed type get_type function should
61 use usual get_type pattern
63 * gio/gfileattribute.c (g_file_attribute_info_list_get_type): Use
64 g_once_init_enter/leave.
66 2008-10-08 Christian Persch <chpe@gnome.org>
68 Bug 555314 – mem leak in gmarkup
70 * glib/gmarkup.c: (g_markup_parse_context_parse): Plug a mem leak.
72 2008-10-08 Tor Lillqvist <tml@novell.com>
74 Bug 554790 - g_convert() misbehaves with winiconv versions
76 * glib/win_iconv.c (kernel_mbtowc): If converting from ASCII,
77 explicitly check for and reject 8bit chars. MultiByteToWideChar()
78 doesn't, at least not on XP.
80 2008-10-06 Matthias Clasen <mclasen@redhat.com>
82 * glib/gtypes.h: Properly include gmacros.h
84 2008-10-06 Behdad Esfahbod <behdad@gnome.org>
86 Bug 555309 – giochannel breaks on error
87 Patch from Christian Persch
89 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
90 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags):
91 Don't shadow err. Oops!
93 2008-10-06 Christophe Fergeau <teuf@gnome.org>
95 Bug 555224 – Improve g_format_size_for_display doc
97 * glib/gfileutils.c: change g_format_size_for_display API doc to
98 explicitly say that the returned string has to be freed. Change
99 spelling of "newly allocated" to "newly-allocated" in g_file_read_link
100 API doc to be more consistent with what is done in that file.
102 2008-10-01 David Zeuthen <davidz@redhat.com>
104 * README.in: Add "Notes about glib 2.20" section detailing the
105 ramifications of the patch from bug #528670.
107 2008-09-30 Behdad Esfahbod <behdad@gnome.org>
109 Bug 554092 – glib doesn't return G_FILE_ERROR_NOENT et al on OS X
111 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
112 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags),
113 (g_io_unix_get_flags), (g_io_channel_new_file):
114 Like mclasen says: "well, thats the way errno works...,
115 save it or loose it". Save errno.
117 2008-09-30 Tor Lillqvist <tml@novell.com>
120 * glib/tests/Makefile.am: Bypass gtester related stuff on Windows.
122 2008-09-30 Tor Lillqvist <tml@novell.com>
125 * glib/gnulib/vasnprintf.c: Don't define _GNU_SOURCE on Windows,
126 as _GNU_SOURCE has unintended side effects when compiling against
127 newest mingw headers.
129 2008-09-26 Dan Winship <danw@gnome.org>
131 Bug 553447 – g_assert_no_error()
133 * glib/gtestutils.h (g_assert_no_error, g_assert_error): Macros to
134 assert that a GError is not set, or else is set to a particular
137 * glib/gtestutils.c (g_assertion_message_error): utility for
140 * glib/tests/keyfile.c:
141 * tests/asyncqueue-test.c:
142 * tests/bookmarkfile-test.c:
143 * tests/convert-test.c:
144 * tests/file-test.c: Use g_assert_error/g_assert_no_error
146 2008-09-26 Dan Winship <danw@gnome.org>
148 * glib/gthreadpool.c (wakeup_thread_marker): make this a "const
149 gpointer" rather than a gconstpointer to avoid warnings later
151 * glib/pcre/pcre_ucp_searchfuncs.c:
152 * glib/pcre/pcre_valid_utf8.c: #include "config.h"
154 * glib/tests/printf.c (test_d): fool gcc into not warning about
155 some printf format strings that we know are dubious
157 2008-09-26 Matthias Clasen <mclasen@redhat.com>
159 Bug 553857 – gbacktrace.h requires signal.h
161 * glib/gbacktrace.h: Include signal.h for raise().
162 Pointed out by Sebastien Bacher
164 2008-09-26 Matthias Clasen <mclasen@redhat.com>
166 Bug 553724 – python interpretter path not patched in correctly
168 * glib/Makefile.am: Fix the sed magic to replace python.
170 2008-09-26 Matthias Clasen <mclasen@redhat.com>
172 * glib/gmain.c: Add some more docs.
174 * glib/giochannel.c: Move more docs inline, and improve them
177 2008-09-25 Tor Lillqvist <tml@novell.com>
179 Bug 553820 - gpoll.c: undeclared identifier
182 * glib/gpoll.c: Make the g_poll() function non-static also on
183 Windows. Prefix an underscore to the g_main_poll_debug variable
184 and make it non-static in gmain.c so that it can be used in
185 gpoll.c. Add back missing variable declaration.
187 2008-09-25 Tor Lillqvist <tml@novell.com>
189 * glib/gspawn-win32.c (do_spawn_with_pipes) (do_spawn_directly):
190 Just ignore the child_setup function, never call it. The is no
191 situation in which it could be useful on Windows. Do print a
192 warning, like before.
194 * glib/gspawn.c (g_spawn_async_with_pipes): Corresponding change
197 2008-09-24 Sven Herzberg <sven@imendio.com>
199 Be a little more explcit in the docs. Includes Owen's requested
202 * glib/gmain.c: improved documentation for g_source_attach() and
205 2008-09-23 Michael Natterer <mitch@imendio.com>
207 * glib/glib.h: #include <glib/gpoll.h>
209 * glib/gpoll.h: #error out if gpoll.h is included directly.
211 * glib/gpoll.c: remove trailing whitespace.
213 2008-09-23 Dan Winship <danw@gnome.org>
215 * glib/gpoll.[ch] (g_poll): Move this out of gmain.c and make it part
216 of the public API. (Part of Bug 505361 - gunixinputstream.c assumes
219 2008-09-23 Tor Lillqvist <tml@novell.com>
221 * glib/gmain.c (poll_rest) [Win32]: Fix embarrassing bug: I was
222 passing an incorrect third parameter to memmove(), had forgotten
223 to multiply by the size of the table entry. Just use a for loop
224 instead, clearer. Odd I didn't notice when testing this code.
226 2008-09-19 Hans Petter Jansson <hpj@novell.com>
228 Rewrite most of GHashTable to use open addressing with quadratic
229 probing instead of chaining. This has the potential to reduce memory
230 fragmentation significantly, while being slightly faster due to
231 better locality and no need to call alloc/free functions for nodes.
232 Benchmarks suggest it also uses less memory overall.
234 * glib/ghash.c (prime_mod): Table of suitable primes for
235 initial-probe distribution.
236 (g_hash_table_set_shift): New function.
237 (g_hash_table_find_closest_shift): New function.
238 (g_hash_table_set_shift_from_size): New function.
239 (g_hash_table_lookup_node_for_insertion): New function.
240 (g_hash_table_lookup_node): Rewritten to return node index instead of
241 pointer, use quadratic probe on flat table, and not return insertion
242 data. The latter saves some computation for read-only lookups.
243 (g_hash_table_remove_node): Rewrite to take a pointer directly to the
244 node structure to remove, and clear that. Remove unlinking code.
245 (g_hash_table_remove_all_nodes): Rewrite to not clear nodes
246 individually, but en masse using memset () after potentially calling
248 (iter_remove_or_steal): Use new data structure and algorithm. Vastly
249 simplified - now just a call to g_hash_table_remove_node ().
250 (g_hash_table_resize): New resize code, re-indexing with new prime
251 and cleaning up tombstones.
252 (g_hash_table_maybe_resize): Table may hold 8 buckets minimum, no less
253 than 1/4 load excluding tombstones, and no more than 15/16 load
254 including tombstones. These numbers are the results of a lot of
255 benchmarking with multiple complex applications, and should not be
257 (g_hash_table_iter_next)
258 (g_hash_table_lookup)
259 (g_hash_table_lookup_extended)
260 (g_hash_table_insert_internal)
261 (g_hash_table_remove_internal)
262 (g_hash_table_foreach_remove_or_steal)
263 (g_hash_table_foreach)
265 (g_hash_table_get_keys)
266 (g_hash_table_get_values): Use new data structure and algorithm,
267 fairly trivial changes.
269 2008-09-19 Tor Lillqvist <tml@novell.com>
271 * glib-zip.in: Look for man pages in share/man.
273 * glib/gutils.c (_glib_get_dll_directory)
274 * glib/gspawn-win32.c (do_spawn_with_pipes): Be a bit less
275 restrictive, look for the helper programs in the same folder where
276 the GLib DLL is, not necessarily in a "bin" subfolder of the top
277 GLib installation folder.
279 2008-09-18 Matthias Clasen <mclasen@redhat.com>
281 * configure.in: Bump version to 2.19.0
283 * ChangeLog.pre-2-18: rotate ChangeLog
285 * === branch for 2.18 ===