1 2008-11-21 Christophe Fergeau <teuf@gnome.org>
3 Bug 561212 – GFileReadMoreCallback API doc refers to non-existant function
5 * gio/giotypes.h: fix name of function referred to in
6 GFileReadMoreCallback API documentation
8 2008-11-21 Matthias Clasen <mclasen@redhat.com>
10 * configure.in: Bump gtk-doc dependency to 1.11 for
11 nicer index-generation.
13 2008-11-14 Matthias Clasen <mclasen@redhat.com>
17 2008-11-04 Christian Dywan <christian@imendio.com>
19 Bug 558672 – NULL key lookup using g_hash_table_lookup_extended()
22 Clarify g_hash_table_lookup_extended
24 2008-11-02 Tor Lillqvist <tml@novell.com>
26 * configure.in: Expand also build/win32/vs9/Makefile.
28 2008-10-31 Christian Dywan <christian@imendio.com>
30 Bug 558513 – g_warn_if_fail FIXME in gtestutils
32 * glib/gtestutils.c (g_test_log_send):
33 Turn g_assert into g_warn_if_fail as advised
35 2008-10-31 Grahame Bowland <grahame@angrygoats.net>
37 Bug 558185 – 'parent' variable in g_local_file_get_child_for_display_name()
38 hits g_object_unref(NULL) assertion
41 - remove unused variable. Patch by Matt Johnston <matt@ucc.asn.au>
43 2008-10-29 16:11:14 Tim Janik <timj@gtk.org>
45 * glib/gmacros.h: added G_PASTE() and G_STATIC_ASSERT(), based on
46 patches by Christian Persch and Christian Dywan. Fixes:
47 Bug 558381 - Add support for compile time assertions
49 2008-10-27 Joseph Pingenot <gnomesvn@digitasaru.net>
51 * glib/ghash.c: changed "#GHashTableIterator" to "#GHashTableIter"
52 in the documentation-comments above g_hash_table_foreach_remove()
53 and g_hash_table_foreach_steal() to correctly name and link to
54 GHashTableIter. This affects the GNOME documentation website
55 as well, so further steps are likely necessary beyond this subproject..
57 2008-10-23 Matthias Clasen <mclasen@redhat.com>
59 Bug 557210 – g_compute_checksum_for_* asserts with less than 2 bytes
61 * glib/gchecksum.c (g_compute_checksum_for_data): Accept
62 lengths < 2. Patch by Tommi Komulainen
64 * tests/checksum-test.c: Add a test for this
66 2008-10-23 Matthias Clasen <mclasen@redhat.com>
68 Bug 556921 – gpoll.h breaks hal compilation
70 * glib/gpoll.h: Include gtypes.h. Pointed out by Anis Elleuch
72 2008-10-20 Christian Persch <chpe@gnome.org>
74 Bug 557087 – mem leak in g_content_types_get_registered
76 * gio/gcontenttype.c: Plug a mem leak.
78 2008-10-19 Michael Natterer <mitch@imendio.com>
80 Bug 556186 – gpoll.h breaks gmain.h inclusion
82 * glib/gpoll.h: different fix: disallow direct inclusion
83 unconditionally again but make an exception if included from
84 gmain.h to maintain compatibility.
86 2008-10-16 Matthias Clasen <mclasen@redhat.com>
88 * configure.in: Bump version to 2.19.1
90 * === Released 2.19.0 ===
94 2008-10-14 Christian Persch <chpe@gnome.org>
96 Bug 556186 – gpoll.h breaks gmain.h inclusion
98 * glib/gpoll.h: Only disallow direct inclusion when
99 G_DISABLE_SINGLE_INCLUDES is defined.
101 2008-10-13 Sven Herzberg <sven@imendio.com>
103 Bug 556101 – static mutex yields warnings with g++
104 Reviewed by Tim Janik
106 * configure.in: added an intermediate cast to gpointer
108 2008-10-13 Christian Persch <chpe@gnome.org>
110 Bug 555311 – format not a string literal and no format arguments
113 * glib/gshell.c: Use literal errors where appropriate.
115 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
117 Bug 551355 – [PATCH] Make glib build with libtool 2.2
119 * autogen.sh: Accept libtool 2.2. We are moving towards having
122 2008-10-10 Behdad Esfahbod <behdad@gnome.org>
124 * configure.in: With libtool 2.x, regenerate libtool early.
125 Based on bug 552107, patch by Patryk Zawadzki
127 2008-10-10 Matthias Clasen <mclasen@redhat.com>
129 Bug 554557 – Patch to fix gcc warnings about missing format
134 * glib/grel.c: Fix gcc warnings. Patch by Kjartan Maraas
136 2008-10-10 Matthias Clasen <mclasen@redhat.com>
138 Bug 552861 – glib-2.0.m4 calls system(3) without storing its result
140 * m4macros/glib-2.0.m4: Cosmetic change to make -Werror happy.
141 Patch by Andreas Köhler
143 2008-10-08 Christian Persch <chpe@gnome.org>
145 Bug 555313 – GFileAttribute boxed type get_type function should
146 use usual get_type pattern
148 * gio/gfileattribute.c (g_file_attribute_info_list_get_type): Use
149 g_once_init_enter/leave.
151 2008-10-08 Christian Persch <chpe@gnome.org>
153 Bug 555314 – mem leak in gmarkup
155 * glib/gmarkup.c: (g_markup_parse_context_parse): Plug a mem leak.
157 2008-10-08 Tor Lillqvist <tml@novell.com>
159 Bug 554790 - g_convert() misbehaves with winiconv versions
161 * glib/win_iconv.c (kernel_mbtowc): If converting from ASCII,
162 explicitly check for and reject 8bit chars. MultiByteToWideChar()
163 doesn't, at least not on XP.
165 2008-10-06 Matthias Clasen <mclasen@redhat.com>
167 * glib/gtypes.h: Properly include gmacros.h
169 2008-10-06 Behdad Esfahbod <behdad@gnome.org>
171 Bug 555309 – giochannel breaks on error
172 Patch from Christian Persch
174 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
175 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags):
176 Don't shadow err. Oops!
178 2008-10-06 Christophe Fergeau <teuf@gnome.org>
180 Bug 555224 – Improve g_format_size_for_display doc
182 * glib/gfileutils.c: change g_format_size_for_display API doc to
183 explicitly say that the returned string has to be freed. Change
184 spelling of "newly allocated" to "newly-allocated" in g_file_read_link
185 API doc to be more consistent with what is done in that file.
187 2008-10-01 David Zeuthen <davidz@redhat.com>
189 * README.in: Add "Notes about glib 2.20" section detailing the
190 ramifications of the patch from bug #528670.
192 2008-09-30 Behdad Esfahbod <behdad@gnome.org>
194 Bug 554092 – glib doesn't return G_FILE_ERROR_NOENT et al on OS X
196 * glib/giounix.c (g_io_unix_read), (g_io_unix_write),
197 (g_io_unix_seek), (g_io_unix_close), (g_io_unix_set_flags),
198 (g_io_unix_get_flags), (g_io_channel_new_file):
199 Like mclasen says: "well, thats the way errno works...,
200 save it or loose it". Save errno.
202 2008-09-30 Tor Lillqvist <tml@novell.com>
205 * glib/tests/Makefile.am: Bypass gtester related stuff on Windows.
207 2008-09-30 Tor Lillqvist <tml@novell.com>
210 * glib/gnulib/vasnprintf.c: Don't define _GNU_SOURCE on Windows,
211 as _GNU_SOURCE has unintended side effects when compiling against
212 newest mingw headers.
214 2008-09-26 Dan Winship <danw@gnome.org>
216 Bug 553447 – g_assert_no_error()
218 * glib/gtestutils.h (g_assert_no_error, g_assert_error): Macros to
219 assert that a GError is not set, or else is set to a particular
222 * glib/gtestutils.c (g_assertion_message_error): utility for
225 * glib/tests/keyfile.c:
226 * tests/asyncqueue-test.c:
227 * tests/bookmarkfile-test.c:
228 * tests/convert-test.c:
229 * tests/file-test.c: Use g_assert_error/g_assert_no_error
231 2008-09-26 Dan Winship <danw@gnome.org>
233 * glib/gthreadpool.c (wakeup_thread_marker): make this a "const
234 gpointer" rather than a gconstpointer to avoid warnings later
236 * glib/pcre/pcre_ucp_searchfuncs.c:
237 * glib/pcre/pcre_valid_utf8.c: #include "config.h"
239 * glib/tests/printf.c (test_d): fool gcc into not warning about
240 some printf format strings that we know are dubious
242 2008-09-26 Matthias Clasen <mclasen@redhat.com>
244 Bug 553857 – gbacktrace.h requires signal.h
246 * glib/gbacktrace.h: Include signal.h for raise().
247 Pointed out by Sebastien Bacher
249 2008-09-26 Matthias Clasen <mclasen@redhat.com>
251 Bug 553724 – python interpretter path not patched in correctly
253 * glib/Makefile.am: Fix the sed magic to replace python.
255 2008-09-26 Matthias Clasen <mclasen@redhat.com>
257 * glib/gmain.c: Add some more docs.
259 * glib/giochannel.c: Move more docs inline, and improve them
262 2008-09-25 Tor Lillqvist <tml@novell.com>
264 Bug 553820 - gpoll.c: undeclared identifier
267 * glib/gpoll.c: Make the g_poll() function non-static also on
268 Windows. Prefix an underscore to the g_main_poll_debug variable
269 and make it non-static in gmain.c so that it can be used in
270 gpoll.c. Add back missing variable declaration.
272 2008-09-25 Tor Lillqvist <tml@novell.com>
274 * glib/gspawn-win32.c (do_spawn_with_pipes) (do_spawn_directly):
275 Just ignore the child_setup function, never call it. The is no
276 situation in which it could be useful on Windows. Do print a
277 warning, like before.
279 * glib/gspawn.c (g_spawn_async_with_pipes): Corresponding change
282 2008-09-24 Sven Herzberg <sven@imendio.com>
284 Be a little more explcit in the docs. Includes Owen's requested
287 * glib/gmain.c: improved documentation for g_source_attach() and
290 2008-09-23 Michael Natterer <mitch@imendio.com>
292 * glib/glib.h: #include <glib/gpoll.h>
294 * glib/gpoll.h: #error out if gpoll.h is included directly.
296 * glib/gpoll.c: remove trailing whitespace.
298 2008-09-23 Dan Winship <danw@gnome.org>
300 * glib/gpoll.[ch] (g_poll): Move this out of gmain.c and make it part
301 of the public API. (Part of Bug 505361 - gunixinputstream.c assumes
304 2008-09-23 Tor Lillqvist <tml@novell.com>
306 * glib/gmain.c (poll_rest) [Win32]: Fix embarrassing bug: I was
307 passing an incorrect third parameter to memmove(), had forgotten
308 to multiply by the size of the table entry. Just use a for loop
309 instead, clearer. Odd I didn't notice when testing this code.
311 2008-09-19 Hans Petter Jansson <hpj@novell.com>
313 Rewrite most of GHashTable to use open addressing with quadratic
314 probing instead of chaining. This has the potential to reduce memory
315 fragmentation significantly, while being slightly faster due to
316 better locality and no need to call alloc/free functions for nodes.
317 Benchmarks suggest it also uses less memory overall.
319 * glib/ghash.c (prime_mod): Table of suitable primes for
320 initial-probe distribution.
321 (g_hash_table_set_shift): New function.
322 (g_hash_table_find_closest_shift): New function.
323 (g_hash_table_set_shift_from_size): New function.
324 (g_hash_table_lookup_node_for_insertion): New function.
325 (g_hash_table_lookup_node): Rewritten to return node index instead of
326 pointer, use quadratic probe on flat table, and not return insertion
327 data. The latter saves some computation for read-only lookups.
328 (g_hash_table_remove_node): Rewrite to take a pointer directly to the
329 node structure to remove, and clear that. Remove unlinking code.
330 (g_hash_table_remove_all_nodes): Rewrite to not clear nodes
331 individually, but en masse using memset () after potentially calling
333 (iter_remove_or_steal): Use new data structure and algorithm. Vastly
334 simplified - now just a call to g_hash_table_remove_node ().
335 (g_hash_table_resize): New resize code, re-indexing with new prime
336 and cleaning up tombstones.
337 (g_hash_table_maybe_resize): Table may hold 8 buckets minimum, no less
338 than 1/4 load excluding tombstones, and no more than 15/16 load
339 including tombstones. These numbers are the results of a lot of
340 benchmarking with multiple complex applications, and should not be
342 (g_hash_table_iter_next)
343 (g_hash_table_lookup)
344 (g_hash_table_lookup_extended)
345 (g_hash_table_insert_internal)
346 (g_hash_table_remove_internal)
347 (g_hash_table_foreach_remove_or_steal)
348 (g_hash_table_foreach)
350 (g_hash_table_get_keys)
351 (g_hash_table_get_values): Use new data structure and algorithm,
352 fairly trivial changes.
354 2008-09-19 Tor Lillqvist <tml@novell.com>
356 * glib-zip.in: Look for man pages in share/man.
358 * glib/gutils.c (_glib_get_dll_directory)
359 * glib/gspawn-win32.c (do_spawn_with_pipes): Be a bit less
360 restrictive, look for the helper programs in the same folder where
361 the GLib DLL is, not necessarily in a "bin" subfolder of the top
362 GLib installation folder.
364 2008-09-18 Matthias Clasen <mclasen@redhat.com>
366 * configure.in: Bump version to 2.19.0
368 * ChangeLog.pre-2-18: rotate ChangeLog
370 * === branch for 2.18 ===