1 2008-10-16 Matthias Clasen <mclasen@redhat.com>
3 * === Released 2.19.0 ===
7 2008-10-14 Christian Persch <chpe@gnome.org>
9 Bug 556186 – gpoll.h breaks gmain.h inclusion
11 * glib/gpoll.h: Only disallow direct inclusion when
12 G_DISABLE_SINGLE_INCLUDES is defined.
14 2008-10-13 Sven Herzberg <sven@imendio.com>
16 Bug 556101 – static mutex yields warnings with g++
19 * configure.in: added an intermediate cast to gpointer
21 2008-10-13 Christian Persch <chpe@gnome.org>
23 Bug 555311 – format not a string literal and no format arguments
26 * glib/gshell.c: Use literal errors where appropriate.
28 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
30 Bug 551355 – [PATCH] Make glib build with libtool 2.2
32 * autogen.sh: Accept libtool 2.2. We are moving towards having
35 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
37 * configure.in: With libtool 2.x, regenerate libtool early.
38 Based on bug 552107, patch by Patryk Zawadzki
40 2008-10-10 Matthias Clasen <mclasen@redhat.com>
42 Bug 554557 – Patch to fix gcc warnings about missing format
47 * glib/grel.c: Fix gcc warnings. Patch by Kjartan Maraas
49 2008-10-10 Matthias Clasen <mclasen@redhat.com>
51 Bug 552861 – glib-2.0.m4 calls system(3) without storing its result
53 * m4macros/glib-2.0.m4: Cosmetic change to make -Werror happy.
54 Patch by Andreas Köhler
56 2008-10-08 Christian Persch <chpe@gnome.org>
58 Bug 555313 – GFileAttribute boxed type get_type function should
59 use usual get_type pattern
61 * gio/gfileattribute.c (g_file_attribute_info_list_get_type): Use
62 g_once_init_enter/leave.
64 2008-10-08 Christian Persch <chpe@gnome.org>
66 Bug 555314 – mem leak in gmarkup
68 * glib/gmarkup.c: (g_markup_parse_context_parse): Plug a mem leak.
70 2008-10-08 Tor Lillqvist <tml@novell.com>
72 Bug 554790 - g_convert() misbehaves with winiconv versions
74 * glib/win_iconv.c (kernel_mbtowc): If converting from ASCII,
75 explicitly check for and reject 8bit chars. MultiByteToWideChar()
76 doesn't, at least not on XP.
78 2008-10-06 Matthias Clasen <mclasen@redhat.com>
80 * glib/gtypes.h: Properly include gmacros.h
82 2008-10-06 Behdad Esfahbod <behdad@gnome.org>
84 Bug 555309 – giochannel breaks on error
85 Patch from Christian Persch
87 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
88 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags):
89 Don't shadow err. Oops!
91 2008-10-06 Christophe Fergeau <teuf@gnome.org>
93 Bug 555224 – Improve g_format_size_for_display doc
95 * glib/gfileutils.c: change g_format_size_for_display API doc to
96 explicitly say that the returned string has to be freed. Change
97 spelling of "newly allocated" to "newly-allocated" in g_file_read_link
98 API doc to be more consistent with what is done in that file.
100 2008-10-01 David Zeuthen <davidz@redhat.com>
102 * README.in: Add "Notes about glib 2.20" section detailing the
103 ramifications of the patch from bug #528670.
105 2008-09-30 Behdad Esfahbod <behdad@gnome.org>
107 Bug 554092 – glib doesn't return G_FILE_ERROR_NOENT et al on OS X
109 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
110 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags),
111 (g_io_unix_get_flags), (g_io_channel_new_file):
112 Like mclasen says: "well, thats the way errno works...,
113 save it or loose it". Save errno.
115 2008-09-30 Tor Lillqvist <tml@novell.com>
118 * glib/tests/Makefile.am: Bypass gtester related stuff on Windows.
120 2008-09-30 Tor Lillqvist <tml@novell.com>
123 * glib/gnulib/vasnprintf.c: Don't define _GNU_SOURCE on Windows,
124 as _GNU_SOURCE has unintended side effects when compiling against
125 newest mingw headers.
127 2008-09-26 Dan Winship <danw@gnome.org>
129 Bug 553447 – g_assert_no_error()
131 * glib/gtestutils.h (g_assert_no_error, g_assert_error): Macros to
132 assert that a GError is not set, or else is set to a particular
135 * glib/gtestutils.c (g_assertion_message_error): utility for
138 * glib/tests/keyfile.c:
139 * tests/asyncqueue-test.c:
140 * tests/bookmarkfile-test.c:
141 * tests/convert-test.c:
142 * tests/file-test.c: Use g_assert_error/g_assert_no_error
144 2008-09-26 Dan Winship <danw@gnome.org>
146 * glib/gthreadpool.c (wakeup_thread_marker): make this a "const
147 gpointer" rather than a gconstpointer to avoid warnings later
149 * glib/pcre/pcre_ucp_searchfuncs.c:
150 * glib/pcre/pcre_valid_utf8.c: #include "config.h"
152 * glib/tests/printf.c (test_d): fool gcc into not warning about
153 some printf format strings that we know are dubious
155 2008-09-26 Matthias Clasen <mclasen@redhat.com>
157 Bug 553857 – gbacktrace.h requires signal.h
159 * glib/gbacktrace.h: Include signal.h for raise().
160 Pointed out by Sebastien Bacher
162 2008-09-26 Matthias Clasen <mclasen@redhat.com>
164 Bug 553724 – python interpretter path not patched in correctly
166 * glib/Makefile.am: Fix the sed magic to replace python.
168 2008-09-26 Matthias Clasen <mclasen@redhat.com>
170 * glib/gmain.c: Add some more docs.
172 * glib/giochannel.c: Move more docs inline, and improve them
175 2008-09-25 Tor Lillqvist <tml@novell.com>
177 Bug 553820 - gpoll.c: undeclared identifier
180 * glib/gpoll.c: Make the g_poll() function non-static also on
181 Windows. Prefix an underscore to the g_main_poll_debug variable
182 and make it non-static in gmain.c so that it can be used in
183 gpoll.c. Add back missing variable declaration.
185 2008-09-25 Tor Lillqvist <tml@novell.com>
187 * glib/gspawn-win32.c (do_spawn_with_pipes) (do_spawn_directly):
188 Just ignore the child_setup function, never call it. The is no
189 situation in which it could be useful on Windows. Do print a
190 warning, like before.
192 * glib/gspawn.c (g_spawn_async_with_pipes): Corresponding change
195 2008-09-24 Sven Herzberg <sven@imendio.com>
197 Be a little more explcit in the docs. Includes Owen's requested
200 * glib/gmain.c: improved documentation for g_source_attach() and
203 2008-09-23 Michael Natterer <mitch@imendio.com>
205 * glib/glib.h: #include <glib/gpoll.h>
207 * glib/gpoll.h: #error out if gpoll.h is included directly.
209 * glib/gpoll.c: remove trailing whitespace.
211 2008-09-23 Dan Winship <danw@gnome.org>
213 * glib/gpoll.[ch] (g_poll): Move this out of gmain.c and make it part
214 of the public API. (Part of Bug 505361 - gunixinputstream.c assumes
217 2008-09-23 Tor Lillqvist <tml@novell.com>
219 * glib/gmain.c (poll_rest) [Win32]: Fix embarrassing bug: I was
220 passing an incorrect third parameter to memmove(), had forgotten
221 to multiply by the size of the table entry. Just use a for loop
222 instead, clearer. Odd I didn't notice when testing this code.
224 2008-09-19 Hans Petter Jansson <hpj@novell.com>
226 Rewrite most of GHashTable to use open addressing with quadratic
227 probing instead of chaining. This has the potential to reduce memory
228 fragmentation significantly, while being slightly faster due to
229 better locality and no need to call alloc/free functions for nodes.
230 Benchmarks suggest it also uses less memory overall.
232 * glib/ghash.c (prime_mod): Table of suitable primes for
233 initial-probe distribution.
234 (g_hash_table_set_shift): New function.
235 (g_hash_table_find_closest_shift): New function.
236 (g_hash_table_set_shift_from_size): New function.
237 (g_hash_table_lookup_node_for_insertion): New function.
238 (g_hash_table_lookup_node): Rewritten to return node index instead of
239 pointer, use quadratic probe on flat table, and not return insertion
240 data. The latter saves some computation for read-only lookups.
241 (g_hash_table_remove_node): Rewrite to take a pointer directly to the
242 node structure to remove, and clear that. Remove unlinking code.
243 (g_hash_table_remove_all_nodes): Rewrite to not clear nodes
244 individually, but en masse using memset () after potentially calling
246 (iter_remove_or_steal): Use new data structure and algorithm. Vastly
247 simplified - now just a call to g_hash_table_remove_node ().
248 (g_hash_table_resize): New resize code, re-indexing with new prime
249 and cleaning up tombstones.
250 (g_hash_table_maybe_resize): Table may hold 8 buckets minimum, no less
251 than 1/4 load excluding tombstones, and no more than 15/16 load
252 including tombstones. These numbers are the results of a lot of
253 benchmarking with multiple complex applications, and should not be
255 (g_hash_table_iter_next)
256 (g_hash_table_lookup)
257 (g_hash_table_lookup_extended)
258 (g_hash_table_insert_internal)
259 (g_hash_table_remove_internal)
260 (g_hash_table_foreach_remove_or_steal)
261 (g_hash_table_foreach)
263 (g_hash_table_get_keys)
264 (g_hash_table_get_values): Use new data structure and algorithm,
265 fairly trivial changes.
267 2008-09-19 Tor Lillqvist <tml@novell.com>
269 * glib-zip.in: Look for man pages in share/man.
271 * glib/gutils.c (_glib_get_dll_directory)
272 * glib/gspawn-win32.c (do_spawn_with_pipes): Be a bit less
273 restrictive, look for the helper programs in the same folder where
274 the GLib DLL is, not necessarily in a "bin" subfolder of the top
275 GLib installation folder.
277 2008-09-18 Matthias Clasen <mclasen@redhat.com>
279 * configure.in: Bump version to 2.19.0
281 * ChangeLog.pre-2-18: rotate ChangeLog
283 * === branch for 2.18 ===