Child sources are supposed to be blocked when their parents are, so
when adding a source to a blocked source, block the child too. Fixes a
warning when unblocking the parent.
Many (if not "almost all") programs that spawn other programs via
g_spawn_sync() or the like simply want to check whether or not the
child exited successfully, but doing so requires use of
platform-specific functionality and there's actually a fair amount of
boilerplate involved.
This new API will help drain a *lot* of mostly duplicated code in
GNOME, from gnome-session to gdm. And we can see that some bits even
inside GLib were doing it wrong; for example checking the exit status
on Unix, but ignoring it on Windows.
https://bugzilla.gnome.org/show_bug.cgi?id=679691
Rather than having a single priority-ordered list of GSources, store a
list of queues of each priority level. This means that adding a source
is now O(n) in the number of unique priority levels currently being
used, rather than O(n) in the total number of sources.
https://bugzilla.gnome.org/show_bug.cgi?id=619329
A child source does not have a priority of its own; it must have the
same priority as its parent. Enforce this in
g_source_set_priority_unlocked().
https://bugzilla.gnome.org/show_bug.cgi?id=619329
When blocking a source that has child sources, we need to consider the
children blocked as well. Otherwise they will still trigger repeatedly
in an inner loop started from the parent source's callback.
https://bugzilla.gnome.org/show_bug.cgi?id=669260
And remove a comment about Windows in the fallback implementation that
no longer applies, since there's now a separate Windows-specific
implementation.
Adding a child source to an already-attached parent source would
crash, because we were passing the parent's context when setting the
child's priority.
01ed78d525 introduced assertion checks for
creating a main context, forking, and attempting to use the main context
from the child side of the fork.
Some code (such as gnome-keyring-daemon) daemonise after calling
GMainContext. That's probably still mostly safe since we still only
have one side of the fork touching the context afterwards.
This use case is still troubling, however, since if any worker threads
have been created at the time of the fork(), we could end up in the
classic situation of leaving some mutexes in a locked state when the
other threads disappear from the copy of the image that the child gets.
This will require some deeper thinking...
Some code using GLib (gnome-keyring-daemon, for example) assumes that
they can catch signals by masking them out in the main thread and
calling sigwait() from a worker.
The problem is that our new worker thread catches the signals before
sigwait() has a chance and the default action occurs (typically
resulting in program termination).
If we mask all the signals in our worker, then this can't happen.
And remove the 'joinable' argument from g_thread_new() and
g_thread_new_full().
Change the wording in the docs. Clarify expectations for
(deprecated) g_thread_create().
Add g_main_context_ref_thread_default(), which always returns a
reffed GMainContext, rather than sometimes returning a (non-reffed)
GMainContext, and sometimes returning NULL. This simplifies the
bookkeeping in any code that needs to keep a reference to the
thread-default context for a while.
https://bugzilla.gnome.org/show_bug.cgi?id=660994
Deprecate both g_thread_create functions and add
g_thread_new() and g_thread_new_full(). The new functions
expect a name for the thread.
Change GThreadPool, GMainContext and GDBus to create named threads.
https://bugzilla.gnome.org/show_bug.cgi?id=660635
This variable, which is the cause of much grief, exists for two reasons:
- ensuring the the wakeup pipe doesn't fill up
- preventing the first poll() after adding a source from waking up
immediately
The first point is no longer an issue with GWakeup.
The second point is addressed by using different logic: we only signal a
wakeup in the case that the context is currently acquired by a thread
that is not us.
As an added bonus, we can now implement g_main_context_wakeup() without
taking a lock.
https://bugzilla.gnome.org/show_bug.cgi?id=583511https://bugzilla.gnome.org/show_bug.cgi?id=320888
Since GMainLoop is now purely monotonic, it really doesn't make sense to
cache the wallclock time just for the sake of making a deprecated call
more efficient.
The first time this is called, this creates a GMainContext * and a
thread to run it. Future calls return the same. There are a lot of
places that we could use this in GLib.
G_THREADS_ENABLED still exists, but is always defined. It is still
possible to use libglib without threads, but gobject (and everything
above it) is now guaranteed to be using threads (as, in fact, it was
before, since it was accidentally impossible to compile with
--disable-threads).
https://bugzilla.gnome.org/show_bug.cgi?id=616754
Since there was nothing guaranteeing synchronization of the
assignments to checked and clockid, it would be possible for one
thread to set clockid = CLOCK_MONOTONIC, and for another thread to see
checked = TRUE but still clockid = CLOCK_REALTIME.
https://bugzilla.gnome.org/show_bug.cgi?id=655129
The Linux eventfd() call is basically tailor made for the main loop
wake up pipe - all we want is a threadsafe way to write to a file
descriptor, and wake up the context on the other end; we don't care
about the content at all.
The eventfd manual page basically explains the benefits:
Applications can use an eventfd file descriptor instead of a
pipe (see pipe(2)) in all cases where a pipe is used simply to
signal events. The kernel overhead of an eventfd file
descriptor is much lower than that of a pipe, and only one file
descriptor is required (versus the two required for a pipe).
When writing my multithreaded spawn test case I actually hit the 1024
file descriptor limit quickly, because we used 2 fds per main context.
This brings that down to 1.
https://bugzilla.gnome.org/show_bug.cgi?id=653140
Running gthread/tests/spawn-multithreaded in a loop, I very easily hit:
GLib-CRITICAL **: g_main_context_wakeup: assertion `g_atomic_int_get (&context->ref_count) > 0' failed
Testing the refcount still left a window where we would fall into the
assertion. Fix this by just locking the context.
Unify it more with the rest of the signal handling code. There's
no reason not to specify SA_RESTART and SA_NOCLDSTOP for flags
always, so just do it.
Remove unnecessary initialization, and push the internal API
towards just ensure_unix_signal_handler_installed_unlocked().
https://bugzilla.gnome.org/show_bug.cgi?id=651725
_GNU_SOURCE must be defined before including any other (system)
header, so defining it in glib-unix.h (and hoping no one has included
anything else before that) is wrong. And the "#define _USE_GNU"
workaround for this problem in gnetworkingprivate.h is even wronger
(and still prone to failure anyway due to single-include guards).
Fix this by defining _GNU_SOURCE in config.h when building against
glibc. In theory this is bad because new releases of glibc may include
symbols that conflict with glib symbols, which could then cause
compile failures. However, most people only see new releases of glibc
when they upgrade their distro, at which point they also generally get
new releases of gcc, which have new warnings/errors to clean up
anyway.
https://bugzilla.gnome.org/show_bug.cgi?id=649201
This new API allows watching a few select Unix signals;
looking through the list on my system, I didn't see anything
else that I think it'd reasonable to watch.
We build on the previous patch to make the child watch helper thread
that existed on Unix handle these signals in the threaded case.
In the non-threaded case, they're just global variables.
https://bugzilla.gnome.org/show_bug.cgi?id=644941
GLib historically has been designed to be "mostly" portable; there
are some functions only available on Unix like g_io_channel_unix_new(),
but these are typically paired with obvious counterparts for Win32.
However, as GLib is used not only by portable software, but components
targeting Unix (or even just Linux), there are a few cases where it
would be very convenient if GLib shipped built-in functionality.
This initial patch is a basic wrapper around pipe2(), including
fallbacks for older kernels. This pairs well with the
existing g_spawn_*() API and its child_setup functionality.
However, in the future, I want to add a signal() wrapper here,
complete with proxying the signal to a mainloop. I have initial code
for this, but doing it sanely (including factoring out gmain.c's
private worker thread), is a complex task, and I don't want to block
on that.
See also gwin32.h for Win32 specific functionality.
https://bugzilla.gnome.org/show_bug.cgi?id=644941
g_tls_certificate_list_new_from_file() was leaking the file contents,
and GSource was leaking the GSourcePrivate structure that got
created when using child sources.
This adds "child source" support to GSource. A child source behaves
basically like a GPollFD; when you add a source to a context, all of
its child sources are added with the same priority; when you destroy a
source, all of its child sources are destroyed; and when a child
source triggers, its parent source's dispatch function is run.
Use cases include:
- adding a GTimeoutSource to another source to cause the source to
automatically trigger after a certain timeout.
- wrapping an existing source type with a new type that has
a different callback signature
- creating a source that triggers based on different conditions
at different times.
https://bugzilla.gnome.org/show_bug.cgi?id=634239
Previously if a source got finalized while still attached to a
context, it would warn and re-ref the source. But then it just freed
it anyway... So keep the warning but drop the re-ref.
https://bugzilla.gnome.org/show_bug.cgi?id=634239
glib is trying to move toward using microseconds-in-gint64 as its
universal time format.
No real API breaks here since GTimeSpec is new this unstable release
series.
The code was designed to deal with any granularity of timer and due to
the use of GTimeVal/GTimeSpec, the math for this gets extremely
confusing.
From a practical standpoint, we only ever have a granularity of seconds.
Take advantage of that fact in the code and vastly simplify the math.
Adds a new function g_main_context_invoke() (and _full() variant).
This function takes a main context, a function and a user_data. If the
main context is already acquired in the current thread, the function is
invoked directly. If the main context is the default main context of
the current thread and it can be acquired then the function is invoked
directly while the context is owned. Otherwise, the function is
scheduled as an idle on the context.
These allow applications to give meaningful names to their sources.
Source names can then be used for debugging and profiling, for
example with systemtap or gdb.
https://bugzilla.gnome.org/show_bug.cgi?id=606044
These allow applications to give meaningful names to their sources.
Source names can then be used for debugging and profiling, for
example with systemtap or gdb.
https://bugzilla.gnome.org/show_bug.cgi?id=606044
2008-09-25 Tor Lillqvist <tml@novell.com>
Bug 553820 - gpoll.c: undeclared identifier
* glib/gmain.c
* glib/gpoll.c: Make the g_poll() function non-static also on
Windows. Prefix an underscore to the g_main_poll_debug variable
and make it non-static in gmain.c so that it can be used in
gpoll.c. Add back missing variable declaration.
svn path=/trunk/; revision=7542
2008-09-24 Sven Herzberg <sven@imendio.com>
Be a little more explcit in the docs. Includes Owen's requested
changes.
* glib/gmain.c: improved documentation for g_source_attach() and
g_source_destroy()
svn path=/trunk/; revision=7539
* glib/gpoll.c (g_poll): Move this out of gmain.c and make it part
of the public API. (Part of Bug 505361 - gunixinputstream.c
assumes poll() available.)
svn path=/trunk/; revision=7535
2008-09-23 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (poll_rest) [Win32]: Fix embarrassing bug: I was
passing an incorrect third parameter to memmove(), had forgotten
to multiply by the size of the table entry. Just use a for loop
instead, clearer. Odd I didn't notice when testing this code.
svn path=/trunk/; revision=7533
2008-09-09 Matthias Clasen <mclasen@redhat.com>
Bug 523463 – Core dump in gmain.c:2482:IA__g_main_context_check()
* glib/gmain.c (g_main_context_check): Be robust against setting
event fields on the fly, as e.g. happens in linc. Tracked down
by Paul Smith, fix proposed by Owen Taylor.
svn path=/trunk/; revision=7447
2008-08-21 Tor Lillqvist <tml@novell.com>
* glib/gmain.c: Rework the g_poll() implementation on Windows to
match poll() semantics more closely. This makes the test program
in bug #468910 behave better and doesn't seem to break anything
else.
If polling several GPollFDs, i.e. messages and/or waitable
handles, first check if one or several of them are in the
signalled state right away, and return indication for all that are
in that case.
If not, then poll with timeout and indicate only the single one
that the Win32 wait function tells us as before.
Remove unnecessary ifdefs, as we always have G_MAIN_POLL_DEBUG
defined on Windows.
Initialise g_main_poll_debug in g_main_context_new() so we have it
before testing it in one case.
Don't add several copies of a handle in the array of handles to
wait for. The documentation says this is not allowed, although it
did seem to work fine in practise. But do as the documentations
says anyway.
svn path=/trunk/; revision=7375
2008-08-04 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (g_get_current_time): MSDN says: "Do not cast a
pointer to a FILETIME structure to either a LARGE_INTEGER* or
__int64* value because it can cause alignment faults on 64-bit
Windows." So don't do that then. Indeed the code did work randomly
on Win64 when compiled with optimisation.
svn path=/trunk/; revision=7308
2008-08-04 Tor Lillqvist <tml@novell.com>
* glib/giowin32.c
* glib/gmain.c
* glib/gspawn-win32.c
* glib/gspawn-win32-helper.c: Change gssize casts introduced on
2008-07-28 to gintptr casts now that we have that. gssize is as
such the same as gintptr on both 32- and 64-bit Windows, but the
gintptr name indicates that it is used to hold pointers, i.e. also
HANDLEs.
svn path=/trunk/; revision=7306
2008-07-28 Tor Lillqvist <tml@novell.com>
Fix problems on 64-bit Windows. Avoid warnings, some of which
indicated actual problems, some which were just annoyances.
Where casts to an integer type are needed for pointers, use
gssize. Technically intptr_t would be the more proper type, but we
still want to be compilable with MSVS6 and 7 which don't have
intptr_t. MSVS8 and 9 do have intptr_t, but in <crtdefs.h>, not
<stdint.h>.
Use %p to print out handles. Use gssize casts when assigning
GPollFD::fd fields.
Use G_GSIZE_FORMAT when printing size_t values.
* configure.in: Define automake conditional G_OS_WIN32_X64 which
is true on Win64.
* glib/giochannel.h: Use slightly different prototype for
g_io_channel_win32_new_messages() on Win64 with gsize instead of
guint.
* glib/giowin32.c
* glib/gmain.c
* glib/gspawn-win32.c
* tests/testglib.c: Generic changes as described above.
* glib/gmain.h: Don't bother mentioning GIMP in comment.
* glib/grel.c (tuple_hash_2): Use all bits of pointer.
* glib/gspawn-win32.c
* glib/gspawn-win32-helper.c: Use gssize types in the
communication between parent and helper process, so that we can
pass process handles, which are pointers, also on Win64.
* glib/gtimer.c (g_time_val_to_iso8601): time_t is 64 bits on
Win64 so we can't pass the address of a GTimeVal::tv_sec which is
a long directly to gmtime(). On the other hand, changing
GTimeVal::tv_sec to be a gint64 on Win64 is not really feasible
either, as that would then require changes in much code that uses
GTimeVals.
* glib/gspawn-win32.c
* glib/Makefile.am: Call the helper programs
gspawn-win64-helper.exe and gspawn-win64-helper-console.exe on
Win64, to avoid potential risk of running a 32-bit version of the
helper.
svn path=/trunk/; revision=7260
* glib/gmain.c (g_get_current_time): Add a g_return_val_if_fail
check in both versions. Proposed by Patrik Olsson in bug 540545.
svn path=/trunk/; revision=7142
2008-06-11 Tor Lillqvist <tml@novell.com>
* glib/gmain.c
* glib/gspawn.c: Clarify what a "child pid" is in the doc
comments.
svn path=/trunk/; revision=6995
2008-05-17 Tor Lillqvist <tml@novell.com>
* glib/gmain.c: Improve debugging printouts. When
G_MAIN_POLL_DEBUG is defined, we check for an environment variable
of the same name to decide whether to print out debugging
information or not. G_MAIN_POLL_DEBUG is always defined on Windows
as there is more often a need to debug this stuff there. On Unix
the definition has to be uncommented (or done on the compile
command line).
svn path=/trunk/; revision=6897
2008-05-02 13:24:13 Tim Janik <timj@imendio.com>
* glib/gmain.c: renamed GMainDispatch.dispatching_sources to be more
descriptive and distinguishable within the source file.
svn path=/trunk/; revision=6874
2008-03-31 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (g_poll): Improve fix for #525192 below: Use
SleepEx() so that the sleep is alertable. Thanks to John
Ehresman.
svn path=/trunk/; revision=6794
2008-03-31 Tor Lillqvist <tml@novell.com>
Bug 525192 - 100% CPU if run main loop with no IO sources
* glib/gmain.c (g_poll) [Win32]: Patch by Neil Roberts.
svn path=/trunk/; revision=6787
2008-02-26 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (g_poll): Further patch by Vlad Grecescu: Drop
the code path that called WaitMessage(), as WaitMessage()
doesn't offer any chance for APCs to run. Instead just use the
code path with MsgWaitForMultipleObjectsEx() even for the
wait-only-for-messages case. (#517484)
svn path=/trunk/; revision=6597
2008-02-24 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (g_poll) [Win32]: Use alertable wait functions so
that I/O completion routines or user-mode Asynchronous Procedure
Calls can be run. (#517484, Vlad Grecescu)
svn path=/trunk/; revision=6571
The g_timeout_add_seconds() API lacks a _full() counterpart, allowing the
setting of a destroy notification function to be invoked when the timeout
source is removed.
This patch adds g_timeout_add_seconds_full() to the public API and
reimplements g_timeout_add_seconds() as a call to g_timeout_add_seconds_full().
svn path=/trunk/; revision=5575
2007-03-15 Chris Wilson <chris@chris-wilson.co.uk>
* glib/gmain.c (g_main_dispatch): Replace a
g_slist_prepend/g_slist_remove pair with an on-stack link
and open coding. (#416094)
svn path=/trunk/; revision=5407
* glib/gmain.c (child_watch_helper_thread): Readd a
return which was removed as dead code a while ago.
icc may consider it dead, but gcc doesn't like non-void
functions without a return... (#354707)
2006-12-12 Matthias Clasen <mclasen@redhat.com>
* configure.in: Add a check for broken poll on Mac OS X.
* glib/gmain.c: Use poll emulation on OS X. (#302672, Toby Peterson,
patch by Dave Vasilevsky)
2006-06-01 Matthias Clasen <mclasen@redhat.com>
* glib/glib.symbols:
* glib/gmain.h:
* glib/gmain.c: Add two new functions,
g_source_set_funcs and g_source_is_destroyed,
that will be necessary to solve thread-safety
issues with idles in GTK+. (#321886, Chris Wilson)
2006-05-09 Sebastian Wilhelmi <wilhelmi@google.com>
* glib/gthreadinit.h: Renamed to glib/gthreadprivate.h and moved
system thread identifier comparision and assignment macros from
glib/gthread.c to glib/gthreadprivate.h.
* glib/Makefile.am, glib/gatomic.c, glib/gconvert.c, glib/gmain.c,
glib/gmem.c, glib/gmessages.c, glib/grand.c, glib/gslice.c,
glib/gthread.c, glib/gutils.c, gthread/gthread-impl.c: Use
glib/gthreadprivate.h instead of glib/gthreadinit.h.
* gthread/gthread-impl.c: Use GSystemThread instead of GThread for
owner determination. This fixes#311043 and is mostly modeled
after the patch from jylefort@FreeBSD.org.
2006-03-20 Matthias Clasen <mclasen@redhat.com>
* glib/gmem.c (profiler_log): use standard_calloc to allocate
the profile_data. (#335209, Chris Wilson)
* glib/gmain.c (g_main_context_unref): Avoid a deadlock.
(#335207, Chris Wilson)
Minor optimizations (#335216, Chris Wilson):
* glib/gasyncqueue.c (g_async_queue_pop_intern_unlocked): Use
g_queue_peek_tail_link instead of g_queue_peek_tail.
* glib/glist.c:
* glib/gslist.c: Avoid some memset calls.
2005-12-20 Michael Meeks <michael.meeks@novell.com>
* glib/gmain.c (g_main_context_is_owner): new method
to determine if the current thread is the owner of the
context.
Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
* glib/gslice.c: implement chain walking for arbitrary ->next pointer
offsets in g_slice_free_chain_with_offset() based on a patch by behdad
in bug 323178. moved time consuming logic from g_slice_free() out of
the inner loop, so g_slice_free_chain_with_offset() provides a real
performance benefit over g_slice_free1() now.
* glib/gslice.h: renamed g_slice_free_chain() to
g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
a type-safe macro as suggested in bug 323178.
simplified the macro implementation of g_slice_free() and implemented
it in a type safe manner for all compliers as suggested by Morten
Welinder <mortenw@gnome.org>.
* glib/gmain.c:
* glib/glist.c:
* glib/gslist.c:
* glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
* glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
added g_slice_*() API to allocate and cache small bits of memory.
an actuall allocator implementation for g_slice_*() is still pending.
* glib/gthread.[hc]: changes from a patch by Matthias Clasen.
changed GRealThread list to use in-structure *next; fields instead
of GSList, in order for thread iteration to not depenend on g_slice_*()
indirectly.
_g_thread_mem_private_get():
_g_thread_mem_private_set(): added accessors for private memory,
needed because the ordinary GPrivate implementation relies on GArray
and GSList and therefore indirectly on working g_slice_*() allocations.
* glib/gthread.[hc]:
g_thread_foreach(): new public API function to loop over all existing threads.
* glib/gdataset.c:
* glib/gstring.c:
* glib/gcache.c:
* glib/garray.c:
* glib/gqueue.c:
* glib/gslist.c:
* glib/glist.c:
* glib/ghash.c:
* glib/gtree.c:
* glib/ghook.c:
* glib/gmain.c:
* glib/gnode.c:
removed GAllocator and free list usages and accompanying locks.
use g_slice_*() API to allocate and cache small bits of memory.
* glib/ghook.h: removed GMemChunk field from public API.
* glib/gslist.h:
* glib/glist.h: deprecate allocator API, provide _free1() for consistency.
* glib/gnode.h: deprecate allocator API.
* glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
be used for poll rec lists.
* glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
g_relation_destroy(): free all tuples from the all_tuples hash table,
this effectively maintains the life time track keeping of tuples.
g_relation_delete_tuple(): free tuples which are removed from the
all_tuples hash table. this fixes a temporary leak that was present
in the memchunk code until the destruction of the relation.
2005-06-30 Matthias Clasen <mclasen@redhat.com>
Apply a patch which may make GLib work on BeOS again.
(#309157, Kian Duffy)
* glib/gmain.c: Update the BeOS includes.
* glib/gstdio.c:
* glib/gutils.c (g_find_program_in_path):
* glib/gbacktrace.c (g_on_error_stack_trace): Use the UNIX
implementation on BeOS, as well.
* configure.in: Don't put glib 1.0 into G_MODULE_LIBS, even
on BeOS.
2005-04-01 Matthias Clasen <mclasen@redhat.com>
* glib/gmain.c (g_child_watch_source_new): Add a note regarding
waitpid(-1).
(g_child_watch_source_init_multi_threaded):
(g_child_watch_source_init_single): Don't use SA_RESTART,
since it causes problems on at least one platform. (#168352)
2005-03-29 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (g_get_current_time): [Win32] Use
GetSystemTimeAsFileTime() instead of time() and
GetTickCount(). Much simpler.
2005-03-29 Tor Lillqvist <tml@novell.com>
* glib/gmain.c (g_poll): If the event fired, assign
f->revents=f->events. We can't know whether the upper layer using
the event actually is readable, writeable or what, so say that all
the conditions hold. Remove the ResetEvent() call that has been
ifdeffed out anyway for a long time. Remove an "#ifdef 1" and
#endif pair of lines, that code is not optional.
2005-03-13 Matthias Clasen <mclasen@redhat.com>
Make PLT-reduction work with gcc4, and don't include
everything in galias.h:
* glib/glib.symbols: Group symbols by header and source file.
* glib/makegalias.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
* glib/Makefile.am (galiasdef.c): Add a rule to generate this
file.
* glib/*.c: Include galias.h after the other GLib headers,
include galiasdef.c at the bottom.
Mon Nov 8 10:45:50 2004 Manish Singh <yosh@gimp.org>
* glib/gasyncqueue.c
* glib/ghook.c
* glib/giochannel.c: g_return_if_fail -> g_return_val_if_fail
* glib/gmain.c: Ditto, plus also make g_main_context_ref() actually
return the passed in pointer.
2004-11-08 Matthias Clasen <mclasen@redhat.com>
* glib/gmain.c: Initialize child_watch_count to 1, so
that we don't miss the very first child if it exits
before we set up the child watch. In that case we had
previously source->count == child_watch_count == 0,
causing g_child_watch_check() to skip the waitpid()
call. (#154827, Gustavo Carneiro)
* glib/gmain.c (g_child_watch_source_init_single)
(g_child_watch_source_init_multi_threaded): Use sigaction()
instead of signal(). (#136867, Jonas Jonsson, patch by
Archana Shah)
2004-10-08 Matthias Clasen <mclasen@redhat.com>
* glib/gmain.c (g_child_watch_add_full):
* glib/gmain.c (g_child_watch_add): Document that GLib supports only
a single callback per pid. (#154828, Gustavo Carneiro)
2004-04-22 Matthias Clasen <mclasen@redhat.com>
* glib/gmain.c (g_main_context_query): Only set time_is_current to
FALSE if context->timeout is not zero. (#137795, Christian Krause)
2004-04-10 Tor Lillqvist <tml@iki.fi>
* glib/giowin32.c
* glib/gmain.c
* glib/gstrfuncs.c
* glib/gthread.c: Decorating variable definitions with
__declspec(dllexport) causes problems on Cygwin build, and isn't
really needed for a native Win32 build with mingw or MSVC, so
remove. (#138402, Roger Leigh)
* glib/libcharset/localcharset.c: Use Win32-specific code also on
Cygwin.
* tests/uri-test.c: Don't assume that local filenames are in UTF-8
on Cygwin, either. (#138412, Roger Leigh)
Fri Mar 19 15:21:09 2004 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c: Fix the accidental revert of the
fixes from #112222 that happened when the GChildWatch
code was added.
Sun Mar 14 12:58:30 2004 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c: if _POLL_EMUL_H is defined, undefine
HAVE_POLL to prefer our own poll() emulation to the
lame OS/X one. (#136956, Manish Singh)
2004-03-04 Sebastian Wilhelmi <seppi@seppi.de>
* glib/gmain.c: Use the atomic integer operations for GMainContext
and GMainLoop reference counting.
* glib/gmain.c: Hold the main_context_list lock, when iterating
the list. Only call g_main_context_wakeup for positive reference
count.
Mon Mar 1 16:49:51 2004 Owen Taylor <otaylor@redhat.com>
* === Released 2.3.5 ===
* configure.in: Version 2.3.5, interface ago 0.
* NEWS: Some further updates.
Mon Mar 1 15:49:09 2004 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c (check_for_child_exited): Don't
call waitpid() on a source that has already exited.
* glib/gmain.c (g_child_watch_check): Return TRUE
only if the child actually exited.
Mon Mar 1 15:39:57 2004 Owen Taylor <otaylor@redhat.com>
Patch from J. Ali Harlow
* configure.in: Use void * not HANDLE for GPid on win32.
* glib/gspawn.[ch] glib/gspawn-win32.[ch] glib/glib.def:
Add g_spawn_close_pid().
* glib/gspawn.[ch]: Make g_spawn functions take
GPid * instead if int * (GPid == int on unix, will
produce compile warnings until fixed on Win32.)
* tests/child-test.c: Make the test a little more
inappropriately verbose.
* glib/gmain.c: Add some documentation warnings about
not closing @pid while the source is active.
2003-02-28 Hans Breuer <hans@breuer.org>
* glib/glib.def : updated externals, including those
from bug #135386
* glib/makefile.msc.in : build gatomic.c
* glibconfig.h.win32.in : removed duplicate definition
of G_MAXSIZE, typedef void* GPid instead of int
* tests/child-test.c glib/gmain.c :
applied patch from J. Ali Harlow <ali@juiblex.co.uk> to fix
g_child_watch implementation on win32, bug #50296
2003-02-26 Hans Breuer <hans@breuer.org>
* glib/glib.def : added g_hash_table_find and a
bunch of g_queue_*
* glib/gmain.c : make it compile on win32,
child_wake_up_pipe replaced by semaphore like it is done
for the other wake_up_pipe
* config.h.win32.in : added HAVE_INT64_AND_I64
* glibconfig.h.win32.in : G_MAXSIZE .. G_M??INT64,
and typedef for GPid
* test/env-test.c : don't let the local log function
collide in namespace with standard C
Sat Feb 14 01:21:34 2004 Matthias Clasen <maclas@gmx.de>
* glib/gmain.h:
* glib/gmain.c (g_child_watch_source_new):
* glib/gmain.c (g_child_watch_add):
* glib/gmain.c (g_child_watch_add_full): Wrap waitpid() as a
GSource. This is a partial implementation of the "Unix signal
source". (#50296, Jonathan R. Blandford)
* configure.in: Add the necessary configury to typedef GPid
appropriately.
* tests/Makefile.am:
* tests/child-test.c: Test child_watch sources.
Sat Nov 1 08:45:38 2003 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c (g_main_context_iterate): Set the
return value from the result of g_main_context_check()
(after we poll) rather than g_main_context_prepare.
(#121675, Padraig O'Briain)
Mon Aug 25 12:17:20 2003 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c (g_main_context_unref_and_unlock):
When DISABLE_MEM_POOLS is set, loop through and free
the poll records explicitely, since g_mem_chunk_destroy()
won't do it. (#118121, Morten Welinder)
2003-08-15 Tor Lillqvist <tml@iki.fi>
* glib/gmain.c (g_poll): [Win32] Don't exceed handle array
bounds. Warn if there would be too many handles to wait
for. (WaitForMultipleObjects() has a relatively low limit of 64
handles. The Win32 IO channel code should be fixed not to need to
wait for one handle per file or socket being watched. Later.)
Thu Aug 7 14:15:44 2003 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c (g_idle_source_new): Make the default priority
for idle sources G_PRIORITY_DEFAULT_IDLE as anybody would
expect and document that. (#114461, reported by Andy Wingo)
Wed Jul 9 16:27:26 2003 Manish Singh <yosh@gimp.org>
* glib/gmain.c (g_main_context_unref_and_unlock): check if
context->poll_chunk is set before trying to destroy it. For example,
This can happen if no sources are added to a context before it is
freed.
Thu Jun 5 23:40:31 2003 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c: When dispatching a source that is
!CAN_RECURSE, temporarily remove any file descriptors
that that source has registered from the main loop, to keep
recursive main loops from busy-waiting if input
becomes available on one of those file descriptors.
(#112222, Christian Krause)
* glib/gmain.c (g_source_set_priority): Properly
remove the source from the context's source list
and reinsert it sorted, rather than simply setting
source->next/prev to NULL! (#114274)
2003-06-05 Tor Lillqvist <tml@iki.fi>
* glib/giochannel.h
* glib/gmain.h: Remove bogus (Win32-only) declaration of
g_main_poll_win32_msg_add(). No such function exists.
* glib/gmain.c (g_poll) [Win32]: Use g_win32_error_message() for
better warning messages.
Mon Apr 7 13:40:28 2003 Owen Taylor <otaylor@redhat.com>
* glib/gmain.c (g_main_loop_run): When waiting for
the main loop to be freed up, wait on either
!loop->is_running or got_ownership, not both.
(Caused gtk_dialog_run() not to work in other
threads, reported by Jean-Yves Lefort)
2003-02-14 Sebastian Wilhelmi <seppi@seppi.de>
Fixes for #101264 and #99372:
* glib/gconvert.h, glib/gmain.c, glib/gmem.c, glib/gmessages.c,
glib/grand.c: Include gthreadinit.h and rename the thread
initialization functions a bit and let them start with _, so that
later we can stop exporting them.
* glib/gmem.c, glib/gmessages.c: Move the g_private_new() calls to
new functions. They have to be called after setting
g_threads_got_initialized to TRUE (see #101264).
* glib/gthread.c: Include gthreadinit.h. Renamed g_mutex_init() to
g_thread_init_glib(). Call the thread initialization functions
(which are not allowed to call g_private_new), then set
g_threads_got_initialized to TRUE, then call the other thread
initialization functions (which must not call anything but
g_private_new()).
* glib/gthreadinit.h: New private header to cleanly declare all
thread initialization functions.
* gthread/gthread-impl.c: Include gthreadinit.h. In
g_thread_init() just call g_thread_init_glib(), which in turn calls the
other functions (see #99372).
* glib/Makefile.am: Added gthreadinit.h.
