1 2008-10-16 Matthias Clasen <mclasen@redhat.com>
5 2008-10-14 Christian Persch <chpe@gnome.org>
7 Bug 556186 – gpoll.h breaks gmain.h inclusion
9 * glib/gpoll.h: Only disallow direct inclusion when
10 G_DISABLE_SINGLE_INCLUDES is defined.
12 2008-10-13 Sven Herzberg <sven@imendio.com>
14 Bug 556101 – static mutex yields warnings with g++
17 * configure.in: added an intermediate cast to gpointer
19 2008-10-13 Christian Persch <chpe@gnome.org>
21 Bug 555311 – format not a string literal and no format arguments
24 * glib/gshell.c: Use literal errors where appropriate.
26 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
28 Bug 551355 – [PATCH] Make glib build with libtool 2.2
30 * autogen.sh: Accept libtool 2.2. We are moving towards having
33 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
35 * configure.in: With libtool 2.x, regenerate libtool early.
36 Based on bug 552107, patch by Patryk Zawadzki
38 2008-10-10 Matthias Clasen <mclasen@redhat.com>
40 Bug 554557 – Patch to fix gcc warnings about missing format
45 * glib/grel.c: Fix gcc warnings. Patch by Kjartan Maraas
47 2008-10-10 Matthias Clasen <mclasen@redhat.com>
49 Bug 552861 – glib-2.0.m4 calls system(3) without storing its result
51 * m4macros/glib-2.0.m4: Cosmetic change to make -Werror happy.
52 Patch by Andreas Köhler
54 2008-10-08 Christian Persch <chpe@gnome.org>
56 Bug 555313 – GFileAttribute boxed type get_type function should
57 use usual get_type pattern
59 * gio/gfileattribute.c (g_file_attribute_info_list_get_type): Use
60 g_once_init_enter/leave.
62 2008-10-08 Christian Persch <chpe@gnome.org>
64 Bug 555314 – mem leak in gmarkup
66 * glib/gmarkup.c: (g_markup_parse_context_parse): Plug a mem leak.
68 2008-10-08 Tor Lillqvist <tml@novell.com>
70 Bug 554790 - g_convert() misbehaves with winiconv versions
72 * glib/win_iconv.c (kernel_mbtowc): If converting from ASCII,
73 explicitly check for and reject 8bit chars. MultiByteToWideChar()
74 doesn't, at least not on XP.
76 2008-10-06 Matthias Clasen <mclasen@redhat.com>
78 * glib/gtypes.h: Properly include gmacros.h
80 2008-10-06 Behdad Esfahbod <behdad@gnome.org>
82 Bug 555309 – giochannel breaks on error
83 Patch from Christian Persch
85 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
86 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags):
87 Don't shadow err. Oops!
89 2008-10-06 Christophe Fergeau <teuf@gnome.org>
91 Bug 555224 – Improve g_format_size_for_display doc
93 * glib/gfileutils.c: change g_format_size_for_display API doc to
94 explicitly say that the returned string has to be freed. Change
95 spelling of "newly allocated" to "newly-allocated" in g_file_read_link
96 API doc to be more consistent with what is done in that file.
98 2008-10-01 David Zeuthen <davidz@redhat.com>
100 * README.in: Add "Notes about glib 2.20" section detailing the
101 ramifications of the patch from bug #528670.
103 2008-09-30 Behdad Esfahbod <behdad@gnome.org>
105 Bug 554092 – glib doesn't return G_FILE_ERROR_NOENT et al on OS X
107 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
108 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags),
109 (g_io_unix_get_flags), (g_io_channel_new_file):
110 Like mclasen says: "well, thats the way errno works...,
111 save it or loose it". Save errno.
113 2008-09-30 Tor Lillqvist <tml@novell.com>
116 * glib/tests/Makefile.am: Bypass gtester related stuff on Windows.
118 2008-09-30 Tor Lillqvist <tml@novell.com>
121 * glib/gnulib/vasnprintf.c: Don't define _GNU_SOURCE on Windows,
122 as _GNU_SOURCE has unintended side effects when compiling against
123 newest mingw headers.
125 2008-09-26 Dan Winship <danw@gnome.org>
127 Bug 553447 – g_assert_no_error()
129 * glib/gtestutils.h (g_assert_no_error, g_assert_error): Macros to
130 assert that a GError is not set, or else is set to a particular
133 * glib/gtestutils.c (g_assertion_message_error): utility for
136 * glib/tests/keyfile.c:
137 * tests/asyncqueue-test.c:
138 * tests/bookmarkfile-test.c:
139 * tests/convert-test.c:
140 * tests/file-test.c: Use g_assert_error/g_assert_no_error
142 2008-09-26 Dan Winship <danw@gnome.org>
144 * glib/gthreadpool.c (wakeup_thread_marker): make this a "const
145 gpointer" rather than a gconstpointer to avoid warnings later
147 * glib/pcre/pcre_ucp_searchfuncs.c:
148 * glib/pcre/pcre_valid_utf8.c: #include "config.h"
150 * glib/tests/printf.c (test_d): fool gcc into not warning about
151 some printf format strings that we know are dubious
153 2008-09-26 Matthias Clasen <mclasen@redhat.com>
155 Bug 553857 – gbacktrace.h requires signal.h
157 * glib/gbacktrace.h: Include signal.h for raise().
158 Pointed out by Sebastien Bacher
160 2008-09-26 Matthias Clasen <mclasen@redhat.com>
162 Bug 553724 – python interpretter path not patched in correctly
164 * glib/Makefile.am: Fix the sed magic to replace python.
166 2008-09-26 Matthias Clasen <mclasen@redhat.com>
168 * glib/gmain.c: Add some more docs.
170 * glib/giochannel.c: Move more docs inline, and improve them
173 2008-09-25 Tor Lillqvist <tml@novell.com>
175 Bug 553820 - gpoll.c: undeclared identifier
178 * glib/gpoll.c: Make the g_poll() function non-static also on
179 Windows. Prefix an underscore to the g_main_poll_debug variable
180 and make it non-static in gmain.c so that it can be used in
181 gpoll.c. Add back missing variable declaration.
183 2008-09-25 Tor Lillqvist <tml@novell.com>
185 * glib/gspawn-win32.c (do_spawn_with_pipes) (do_spawn_directly):
186 Just ignore the child_setup function, never call it. The is no
187 situation in which it could be useful on Windows. Do print a
188 warning, like before.
190 * glib/gspawn.c (g_spawn_async_with_pipes): Corresponding change
193 2008-09-24 Sven Herzberg <sven@imendio.com>
195 Be a little more explcit in the docs. Includes Owen's requested
198 * glib/gmain.c: improved documentation for g_source_attach() and
201 2008-09-23 Michael Natterer <mitch@imendio.com>
203 * glib/glib.h: #include <glib/gpoll.h>
205 * glib/gpoll.h: #error out if gpoll.h is included directly.
207 * glib/gpoll.c: remove trailing whitespace.
209 2008-09-23 Dan Winship <danw@gnome.org>
211 * glib/gpoll.[ch] (g_poll): Move this out of gmain.c and make it part
212 of the public API. (Part of Bug 505361 - gunixinputstream.c assumes
215 2008-09-23 Tor Lillqvist <tml@novell.com>
217 * glib/gmain.c (poll_rest) [Win32]: Fix embarrassing bug: I was
218 passing an incorrect third parameter to memmove(), had forgotten
219 to multiply by the size of the table entry. Just use a for loop
220 instead, clearer. Odd I didn't notice when testing this code.
222 2008-09-19 Hans Petter Jansson <hpj@novell.com>
224 Rewrite most of GHashTable to use open addressing with quadratic
225 probing instead of chaining. This has the potential to reduce memory
226 fragmentation significantly, while being slightly faster due to
227 better locality and no need to call alloc/free functions for nodes.
228 Benchmarks suggest it also uses less memory overall.
230 * glib/ghash.c (prime_mod): Table of suitable primes for
231 initial-probe distribution.
232 (g_hash_table_set_shift): New function.
233 (g_hash_table_find_closest_shift): New function.
234 (g_hash_table_set_shift_from_size): New function.
235 (g_hash_table_lookup_node_for_insertion): New function.
236 (g_hash_table_lookup_node): Rewritten to return node index instead of
237 pointer, use quadratic probe on flat table, and not return insertion
238 data. The latter saves some computation for read-only lookups.
239 (g_hash_table_remove_node): Rewrite to take a pointer directly to the
240 node structure to remove, and clear that. Remove unlinking code.
241 (g_hash_table_remove_all_nodes): Rewrite to not clear nodes
242 individually, but en masse using memset () after potentially calling
244 (iter_remove_or_steal): Use new data structure and algorithm. Vastly
245 simplified - now just a call to g_hash_table_remove_node ().
246 (g_hash_table_resize): New resize code, re-indexing with new prime
247 and cleaning up tombstones.
248 (g_hash_table_maybe_resize): Table may hold 8 buckets minimum, no less
249 than 1/4 load excluding tombstones, and no more than 15/16 load
250 including tombstones. These numbers are the results of a lot of
251 benchmarking with multiple complex applications, and should not be
253 (g_hash_table_iter_next)
254 (g_hash_table_lookup)
255 (g_hash_table_lookup_extended)
256 (g_hash_table_insert_internal)
257 (g_hash_table_remove_internal)
258 (g_hash_table_foreach_remove_or_steal)
259 (g_hash_table_foreach)
261 (g_hash_table_get_keys)
262 (g_hash_table_get_values): Use new data structure and algorithm,
263 fairly trivial changes.
265 2008-09-19 Tor Lillqvist <tml@novell.com>
267 * glib-zip.in: Look for man pages in share/man.
269 * glib/gutils.c (_glib_get_dll_directory)
270 * glib/gspawn-win32.c (do_spawn_with_pipes): Be a bit less
271 restrictive, look for the helper programs in the same folder where
272 the GLib DLL is, not necessarily in a "bin" subfolder of the top
273 GLib installation folder.
275 2008-09-18 Matthias Clasen <mclasen@redhat.com>
277 * configure.in: Bump version to 2.19.0
279 * ChangeLog.pre-2-18: rotate ChangeLog
281 * === branch for 2.18 ===