va_vopy() is not universally available in all compilers, so make use of
the existing G_VA_COPY macro which either calls va_copy() if it is
available, or emulates it if otherwise.
This means we're not abusing the notifiers for meta_marshallres,
and we're able to later cleanly add other fields to GClosure.
We still have to leave the ABI intact for the GClosure->meta_marshal
bit, as old G_CLOSURE_N_NOTIFIERS macro instances still accesses it.
However, we always set it to zero to keep those macros working.
https://bugzilla.gnome.org/show_bug.cgi?id=661140
enums are stored in v_long but need to be marshalled as signed
integers. On platforms where int is 32 bits, taking the
address of v_long resulted in the wrong 32 bits being marshalled.
So we need to stuff the enum's int-sized value to a temporary
int-sized variable and marshall that instead.
Second, on return, libffi actually returns a pointer to a value
that's sized according to platform conventions, not according to
what the caller requested. ie if ffi_type_sint was requested, the
value can still be a 64-bit sign-extended long on a 64-bit
architecture like PPC64, thus the caller cannot simply cast
the return value as a pointer to the desired type, but must cast
as a pointer to an integral type and then cast to the desired
type to remove any sign extension complications.
For more information on how to correctly handle libffi return
values, see the following bug, specifically comment 35:
https://bugzilla.redhat.com/show_bug.cgi?id=736489
"For 64-bit ABIs that extend integral returns types to 64-bits, libffi always
returns full 64-bit values that you can truncate in the calling code. It's
just the way it is has always been. Please don't change libffi. I'll document
this clearly for the next version (perhaps there is a mention of this, I
haven't looked yet).
The same is true for returning 8-bit values, for instance, on 32-bit systems.
All ABIs extend those results to the full 32-bits so you need to provide a
properly aligned buffer that's big enough to hold the result."
https://bugzilla.gnome.org/show_bug.cgi?id=659881
The documentation for G_TYPE_CHAR says:
"The type designated by G_TYPE_CHAR is unconditionally an 8-bit signed
integer."
However the return value for g_value_get_char() was just "char" which
in C has an unspecified signedness; on e.g. x86 it's signed (which
matches the GType), but on e.g. PowerPC or ARM, it's not.
We can't break the old API, so we need to suck it up and add new API.
Port most internal users, but keep some tests of the old API too.
https://bugzilla.gnome.org/show_bug.cgi?id=659870
The g_object_bind_property_with_closures() function should set a
marshaller if the two GClosures don't have one already. This simplifies
the caller code and avoids duplication.
We need to add a new marshaller to the gmarshal.list matching the
signature of the GBindingTransformFunc function.
2008-06-22 Michael Natterer <mitch@imendio.com>
* *.c: remove trailing whitespace from newly added gtk-doc
comments and reformatted some where they contained overly long or
ill-formatted lines.
svn path=/trunk/; revision=7090
2008-06-22 Michael Natterer <mitch@imendio.com>
* *.c: moved includes back to the top of the files (before gtk-doc
SECTION comments). Add "config.h" in all files and move system
included before glib includes. Remove trailing whitespace from
SECTION comments and did some reformatting where lines were overly
long, no documentation content was changed.
svn path=/trunk/; revision=7089
* gobject/gboxed.c:
* gobject/gclosure.c:
* gobject/genums.c:
* gobject/gvalue.c:
Use file-names from the section file for SECTION: comments.
svn path=/trunk/; revision=7072
2005-09-08 Kjartan Maraas <kmaraas@gnome.org>
* gclosure.c: (g_closure_set_meta_marshal): Remove some unused
code.
* gsignal.c: (g_signal_emitv): Same here
* gtype.c: (g_type_register_fundamental): And here
* testgobject.c: Mark some functions static.
2005-08-02 Matthias Clasen <mclasen@redhat.com>
* gclosure.c (union ClosureInt): Mark the vint member as
volatile, which seems to be necessary to make atomic operations
work on s390.
Mon Aug 1 23:00:42 2005 Tim Janik <timj@imendio.com>
* gclosure.c: turned all modifications to the first 32 integer bits in a
closure into atomic accesses. wrapped write accesses into special macros
to keep the atomic modification logic in a single place. comment cleanups.
* gclosure.h: made all atomicly accessed closure fields volatile.
* gobject.h: made ref_count field volatile.
Sat Jul 30 21:10:26 2005 Tim Janik <timj@gtk.org>
* gobject.c: reverted notify_mutex introduction, since this prevents
parallelized setting of object properties on different objects, and
serves no apparent purpose (to me at least).
g_object_real_dispose(): removed non-atomic reference count
modifications.
g_object_unref(): make sure the closures array is destroyed when
destroying signal handlers.
* gparam.c: cosmetic changes.
* gsignal.c: comment fixup. allow 16bit blocking count.
* gsignal.c: reverted GClosure related changes.
* gclosure.c: reverted premature commit of atomic reference
counting attempt.
2005-07-15 Matthias Clasen <mclasen@redhat.com>
Make refcounting threadsafe by using atomic
operations. (#166020, Wim Taymans)
* gobject.c: Use a recursive lock to protect the
notify queue.
(g_object_unref): Get rid of g_object_last_unref and
do the last unref handling in g_object_unref.
(g_object_ref, g_object_unref): Use atomic operations.
* gsignal.c (struct _HandlerMatch): Use a full integer
for the ref_count field.
(handler_ref, handler_unref_R): Use atomic operations.
* gparam.c (g_param_spec_ref, g_param_spec_unref):
Use atomic operations instead of a lock to make the
refcounting threadsafe.
* gclosure.c (g_closure_ref, g_closure_unref): Use atomic
operations. This is more complicated here, since the
refcount is stored in a bitfield, so we also have
to access all other bitfield members atomically.
* gsignal.c (handlers_find): Read the meta_marshal flag
of the closure atomically.
* tests/Makefile.am (SUBDIRS): Add tests/refcount
* configure.in: Add tests/refcount
* tests/refcount/properties.c: Test property changes
from multiple threads.
* tests/refcount/signals.c: Test signal emission from
multiple threads.
* tests/refcount/objects.c: Test refcounting from
multiple threads.
* tests/refcount/objects2.c:
* tests/refcount/properties2.c: Tests to measure the
overhead of threadsafe refcounting.
* glib/giochannel.c (g_io_channel_ref, g_io_channel_unref):
Use atomic operations to make refcounting
threadsafe. (#166020, Wim Taymans)
2005-03-14 Matthias Clasen <mclasen@redhat.com>
* abicheck.sh: Make work again.
* gsourceclosure.c: Fix a typo.
Make PLT-reduction work with gcc4, and don't include
everything in gobjectalias.h:
* gobject.symbols: Group symbols by header and source
file.
* makegobjectalias.pl: Protect definitions by the same
preprocessor symbols used to guard the headers. Move
the alias declarations to a separate file which is
produced when calling makegalias.pl -def
* Makefile.am (gobjectaliasdef.c): Add a rule to
build this file.
* *.c: Include gobjectalias.h after the other i
GLib headers, include gobjectaliasdef.c at the bottom.
2004-09-16 Matthias Clasen <mclasen@redhat.com>
Implement the same PLT reduction technique used in GTK+:
* Makefile.am: Generate gobjectalias.h from gobject.symbols.
(BUILT_SOURCES): Add gobjectalias.h.
* makegobjectalias.pl: Script to generate gobjectalias.h.
* *.c: Include gobjectalias.h
Wed Nov 7 00:56:00 2001 Tim Janik <timj@gtk.org>
* gclosure.h: provide G_CLOSURE_N_NOTIFIERS() for people that need
to walk the notifier list (puhh, black magic, stay-away-warning ;).
Mon Apr 9 18:57:44 2001 Tim Janik <timj@gtk.org>
* configure.in: increment version to 1.3.4 (binary 0, interface 0).
Mon Apr 9 18:56:15 2001 Tim Janik <timj@gtk.org>
* gclosure.c (g_closure_invoke): only require marshal/meta_marshal if
we're valid (about to actually do marshalling).
Wed Mar 14 18:46:54 2001 Tim Janik <timj@gtk.org>
* gscanner.[hc]: removed archaic gpointer derived_data; relict and
added a GData member instead.
* glist.[hc]: added g_list_remove_all().
* gslist.[hc]: added g_slist_remove_all().
Sat Mar 17 23:18:36 2001 Tim Janik <timj@gtk.org>
* gobject.c (g_object_get_property): minor bug-fix.
* gbsearcharray.[hc]: provide a macro for static initialization and
functions g_bsearch_array_new() and g_bsearch_array_destroy() for
dynamic allocations.
* gboxed.c: introduce G_TYPE_GSTRING, boxed type for GString.
* gclosure.[hc]: naming corrections.
Fri Mar 9 16:42:08 2001 Tim Janik <timj@gtk.org>
* gvaluetypes.[hc]: moved g_strdup_value_contents() into this file as
a public function (was static in gobject.c before). it's a bit odd
to have that function here, especially since it requires extra includes,
but then it doesn't very well fit somewhere else either.
* gparamspecs.c: added default/max/min checks to param spec creation
functions.
Wed Mar 7 09:36:33 2001 Tim Janik <timj@gtk.org>
* gboxed.[hc]: changed prototype of g_boxed_type_register_static()
to contain an optional init function and a hint at whether the
boxed structure uses ref counting internally.
added g_value_set_boxed_take_ownership().
made G_TYPE_BOXED an abstract value type.
* genums.[hc]: made G_TYPE_ENUM and G_TYPE_FLAGS abstract value
types.
* glib-genmarshal.c: argument type changes, preparation for third-party
arg specification.
* gobject.[hc]: cleaned up get/set property code.
added g_strdup_value_contents() to improve warnings.
* gparam.[hc]: added g_param_value_convert(), taking over responsibility
of the old g_value_convert(). added G_PARAM_LAX_VALIDATION flag so
validation alterations may be valid a part of the property setting
process.
* gparamspecs.[hc]: made value comparisons stable (for sort applications).
added GParamSpecValueArray, a param spec for value arrays and
GParamSpecClosure. nuked the value exchange functions and
GParamSpecCCallback.
* gtype.[hc]: catch unintialized usages of the type system with
g_return_val_if_uninitialized(). introduced G_TYPE_FLAG_VALUE_ABSTRACT
to flag types that introduce a value table, but can't be used for
g_value_init(). cleaned up reserved type ids.
* gvalue.[hc]: code cleanups and saner checking.
nuked the value exchange API. implemented value transformations, we
can't really "convert" values, rather transforms are an anylogy to
C casts, real conversions need a param spec for validation, which is
why g_param_value_convert() does real conversions now.
* gvaluearray.[hc]: new files that implement a GValueArray, a struct
that can hold inhomogeneous arrays of value (to that extend that it
also allowes undefined values, i.e. G_VALUE_TYPE(value)==0).
this is exposed to the type system as a boxed type.
* gvaluetransform.c: new file implementing most of the former value
exchange functions as single-sided transformations.
* gvaluetypes.[hc]: nuked G_TYPE_CCALLBACK, added
g_value_set_string_take_ownership().
* *.h: s/G_IS_VALUE_/G_VALUE_HOLDS_/.
* *.[hc]: many fixes and cleanups.
* many warning improvements.
Tue Feb 27 18:35:15 2001 Tim Janik <timj@gtk.org>
* gobject.c (g_object_get_valist): urg, pass G_VALUE_NOCOPY_CONTENTS
into G_VALUE_LCOPY(), this needs proper documenting.
* gparam.c: fixed G_PARAM_USER_MASK.
* gtype.c (type_data_make_W):
(type_data_last_unref_Wm): fixed invalid memory freeing.
* gobject.c (g_object_last_unref): destroy signal handlers associated
with object, right before finalization.
* gsignal.c (g_signal_parse_name): catch destroyed nodes or signals
that don't actually support details.
* gobject.[hc]: got rid of property trailers. nuked GObject
properties "data" and the "signal" variants.
(g_object_connect): new convenience function to do multiple
signal connections at once.
(g_object_disconnect): likewise, for disconnections.
* gparam.[hc] (g_param_spec_pool_lookup): took out trailer support.
* gvalue.[hc]: marked g_value_fits_pointer() and g_value_peek_pointer()
as private (the latter got renamed from g_value_get_as_pointer()).
Wed Mar 7 09:32:06 2001 Tim Janik <timj@gtk.org>
* glib-object.h: add gvaluearray.h.
* gstring.[hc]: fixup naming of g_string_sprint*.
* gtypes.h: fixed GCompareDataFunc naming.
Wed Mar 7 09:33:27 2001 Tim Janik <timj@gtk.org>
* gobject/Makefile.am: shuffled rules to avoid excessive
rebuilds.
* gobject/gobject-sections.txt: updates.
* gobject/tmpl/*: bunch of updates, added another patch
from Eric Lemings <eric.b.lemings@lmco.com>.
Fri Feb 16 06:52:20 2001 Tim Janik <timj@gtk.org>
* gobject/tmpl/types.sgml: incorporated huge docu patch from Eric
Lemings <eric.b.lemings@lmco.com> with a bunch of editing on my part.
Fri Feb 16 07:10:44 2001 Tim Janik <timj@gtk.org>
* gclosure.c:
(g_closure_ref):
(g_closure_sink): make closure sinking explicit.
* gsignal.c:
(g_signal_connect_data):
(g_signal_connect_closure):
(g_signal_connect_closure_by_id):
(g_signal_newv): perform explicit closure sinking.
Thu Feb 8 00:31:45 2001 Tim Janik <timj@gtk.org>
* gtype.h: added G_TYPE_DEBUG_NONE for/from Eric Lemings ;)
Wed Jan 31 07:14:22 2001 Tim Janik <timj@gtk.org>
* gobject/Makefile.am: adapt to work with new CVS gtk-doc, leaving the old
rules in place caused bogus recursions. main changes:
- add to conditionalized section:
all-local:
$(MAKE) scan
$(MAKE) templates
$(MAKE) sgml
$(MAKE) html.stamp
html.stamp: sgml.stamp $(EXTRA_SGML_FILES)
$(MAKE) html
DOC_STAMPS= html.stamp sgml.stamp
- change:
maintainer-clean-local: clean
- cd $(srcdir) && rm -rf sgml html $(DOC_MODULE)-decl-list.txt $(DOC_MODULE)-decl.txt
+ cd $(srcdir) && rm -rf sgml html $(DOC_MODULE)-decl-list.txt $(DOC_MODULE)-decl.txt $(DOC_STAMPS)
* glib/Makefile.am (maintainer-clean-local): dito.
Wed Jan 31 06:21:32 2001 Tim Janik <timj@gtk.org>
* gobject/tmpl/types.sgml: applied docu patch from Eric Lemings
<eric.b.lemings@lmco.com>, did some more editing.
Wed Jan 31 06:19:49 2001 Tim Janik <timj@gtk.org>
* gparam.h: gtk-doc sucks for not dealing with #define inside enums.
* gtype.[hc]: added G_TYPE_FLAG_RESERVED_ID_BIT, a bit in the type
number that's supposed to be left untouched (preserved mainly
for the signal code).
* *.c: added thread safety code, based on an old patch from sebastian.
the remaining thread safety issues are now datalists on pspecs (to be
solved im gdataset.c) and gvalue.c where locking concerns value exchange
functionality only, and that's soon to be revised.
