* Copyright © 2008 codethink
* Copyright © 2009 Red Hat, Inc.
*
+ * SPDX-License-Identifier: LGPL-2.1-or-later
+ *
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
- * version 2 of the License, or (at your option) any later version.
+ * version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General
- * Public License along with this library; if not, write to the
- * Free Software Foundation, Inc., 59 Temple Place, Suite 330,
- * Boston, MA 02111-1307, USA.
+ * Public License along with this library; if not, see <http://www.gnu.org/licenses/>.
*
* Authors: Ryan Lortie <desrt@desrt.ca>
* Alexander Larsson <alexl@redhat.com>
#include "giostream.h"
#include "gasyncresult.h"
+#include "gioprivate.h"
#include "gtask.h"
/**
* @see_also: #GInputStream, #GOutputStream
*
* GIOStream represents an object that has both read and write streams.
- * Generally the two streams acts as separate input and output streams,
+ * Generally the two streams act as separate input and output streams,
* but they share some common resources and state. For instance, for
- * seekable streams they may use the same position in both streams.
+ * seekable streams, both streams may use the same position.
*
- * Examples of #GIOStream objects are #GSocketConnection which represents
- * a two-way network connection, and #GFileIOStream which represent a
+ * Examples of #GIOStream objects are #GSocketConnection, which represents
+ * a two-way network connection; and #GFileIOStream, which represents a
* file handle opened in read-write mode.
*
* To do the actual reading and writing you need to get the substreams
* The #GIOStream object owns the input and the output streams, not the other
* way around, so keeping the substreams alive will not keep the #GIOStream
* object alive. If the #GIOStream object is freed it will be closed, thus
- * closing the substream, so even if the substreams stay alive they will
- * always just return a %G_IO_ERROR_CLOSED for all operations.
+ * closing the substreams, so even if the substreams stay alive they will
+ * always return %G_IO_ERROR_CLOSED for all operations.
*
* To close a stream use g_io_stream_close() which will close the common
* stream object and also the individual substreams. You can also close
* the substreams themselves. In most cases this only marks the
- * substream as closed, so further I/O on it fails. However, some streams
- * may support "half-closed" states where one direction of the stream
- * is actually shut down.
+ * substream as closed, so further I/O on it fails but common state in the
+ * #GIOStream may still be open. However, some streams may support
+ * "half-closed" states where one direction of the stream is actually shut down.
+ *
+ * Operations on #GIOStreams cannot be started while another operation on the
+ * #GIOStream or its substreams is in progress. Specifically, an application can
+ * read from the #GInputStream and write to the #GOutputStream simultaneously
+ * (either in separate threads, or as asynchronous operations in the same
+ * thread), but an application cannot start any #GIOStream operation while there
+ * is a #GIOStream, #GInputStream or #GOutputStream operation in progress, and
+ * an application can’t start any #GInputStream or #GOutputStream operation
+ * while there is a #GIOStream operation in progress.
+ *
+ * This is a product of individual stream operations being associated with a
+ * given #GMainContext (the thread-default context at the time the operation was
+ * started), rather than entire streams being associated with a single
+ * #GMainContext.
+ *
+ * GIO may run operations on #GIOStreams from other (worker) threads, and this
+ * may be exposed to application code in the behaviour of wrapper streams, such
+ * as #GBufferedInputStream or #GTlsConnection. With such wrapper APIs,
+ * application code may only run operations on the base (wrapped) stream when
+ * the wrapper stream is idle. Note that the semantics of such operations may
+ * not be well-defined due to the state the wrapper stream leaves the base
+ * stream in (though they are guaranteed not to crash).
*
* Since: 2.22
*/
struct _GIOStreamPrivate {
guint closed : 1;
guint pending : 1;
- GAsyncReadyCallback outstanding_callback;
};
static gboolean g_io_stream_real_close (GIOStream *stream,
* already set or @stream is closed, it will return %FALSE and set
* @error.
*
- * Return value: %TRUE if pending was previously unset and is now set.
+ * Returns: %TRUE if pending was previously unset and is now set.
*
* Since: 2.22
*/
/**
* g_io_stream_close:
* @stream: a #GIOStream
- * @cancellable: (allow-none): optional #GCancellable object, %NULL to ignore
+ * @cancellable: (nullable): optional #GCancellable object, %NULL to ignore
* @error: location to store the error occurring, or %NULL to ignore
*
* Closes the stream, releasing resources related to it. This will also
- * closes the individual input and output streams, if they are not already
+ * close the individual input and output streams, if they are not already
* closed.
*
* Once the stream is closed, all other operations will return
* The default implementation of this method just calls close on the
* individual input/output streams.
*
- * Return value: %TRUE on success, %FALSE on failure
+ * Returns: %TRUE on success, %FALSE on failure
*
* Since: 2.22
*/
gpointer user_data)
{
GIOStream *stream = G_IO_STREAM (source_object);
+ GIOStreamClass *klass = G_IO_STREAM_GET_CLASS (stream);
+ GTask *task = user_data;
+ GError *error = NULL;
+ gboolean success;
stream->priv->closed = TRUE;
g_io_stream_clear_pending (stream);
- if (stream->priv->outstanding_callback)
- (*stream->priv->outstanding_callback) (source_object, res, user_data);
- g_object_unref (stream);
+
+ if (g_async_result_legacy_propagate_error (res, &error))
+ success = FALSE;
+ else
+ success = klass->close_finish (stream, res, &error);
+
+ if (error)
+ g_task_return_error (task, error);
+ else
+ g_task_return_boolean (task, success);
+
+ g_object_unref (task);
}
/**
* g_io_stream_close_async:
* @stream: a #GIOStream
* @io_priority: the io priority of the request
- * @cancellable: (allow-none): optional cancellable object
+ * @cancellable: (nullable): optional cancellable object
* @callback: (scope async): callback to call when the request is satisfied
* @user_data: (closure): the data to pass to callback function
*
{
GIOStreamClass *class;
GError *error = NULL;
+ GTask *task;
g_return_if_fail (G_IS_IO_STREAM (stream));
+ task = g_task_new (stream, cancellable, callback, user_data);
+ g_task_set_source_tag (task, g_io_stream_close_async);
+
if (stream->priv->closed)
{
- GTask *task;
-
- task = g_task_new (stream, cancellable, callback, user_data);
- g_task_set_source_tag (task, g_io_stream_close_async);
g_task_return_boolean (task, TRUE);
g_object_unref (task);
return;
if (!g_io_stream_set_pending (stream, &error))
{
- g_task_report_error (stream, callback, user_data,
- g_io_stream_close_async,
- error);
+ g_task_return_error (task, error);
+ g_object_unref (task);
return;
}
class = G_IO_STREAM_GET_CLASS (stream);
- stream->priv->outstanding_callback = callback;
- g_object_ref (stream);
+
class->close_async (stream, io_priority, cancellable,
- async_ready_close_callback_wrapper, user_data);
+ async_ready_close_callback_wrapper, task);
}
/**
GAsyncResult *result,
GError **error)
{
- GIOStreamClass *class;
-
g_return_val_if_fail (G_IS_IO_STREAM (stream), FALSE);
- g_return_val_if_fail (G_IS_ASYNC_RESULT (result), FALSE);
-
- if (g_async_result_legacy_propagate_error (result, error))
- return FALSE;
- else if (g_async_result_is_tagged (result, g_io_stream_close_async))
- return g_task_propagate_boolean (G_TASK (result), error);
+ g_return_val_if_fail (g_task_is_valid (result, stream), FALSE);
- class = G_IO_STREAM_GET_CLASS (stream);
- return class->close_finish (stream, result, error);
+ return g_task_propagate_boolean (G_TASK (result), error);
}
g_task_return_boolean (task, TRUE);
}
+typedef struct
+{
+ GError *error;
+ gint pending;
+} CloseAsyncData;
+
+static void
+stream_close_complete (GObject *source,
+ GAsyncResult *result,
+ gpointer user_data)
+{
+ GTask *task = user_data;
+ CloseAsyncData *data;
+
+ data = g_task_get_task_data (task);
+ data->pending--;
+
+ if (G_IS_OUTPUT_STREAM (source))
+ {
+ GError *error = NULL;
+
+ /* Match behaviour with the sync route and give precedent to the
+ * error returned from closing the output stream.
+ */
+ g_output_stream_close_finish (G_OUTPUT_STREAM (source), result, &error);
+ if (error)
+ {
+ if (data->error)
+ g_error_free (data->error);
+ data->error = error;
+ }
+ }
+ else
+ g_input_stream_close_finish (G_INPUT_STREAM (source), result, data->error ? NULL : &data->error);
+
+ if (data->pending == 0)
+ {
+ if (data->error)
+ g_task_return_error (task, data->error);
+ else
+ g_task_return_boolean (task, TRUE);
+
+ g_slice_free (CloseAsyncData, data);
+ g_object_unref (task);
+ }
+}
+
static void
g_io_stream_real_close_async (GIOStream *stream,
int io_priority,
GAsyncReadyCallback callback,
gpointer user_data)
{
+ GInputStream *input;
+ GOutputStream *output;
GTask *task;
task = g_task_new (stream, cancellable, callback, user_data);
+ g_task_set_source_tag (task, g_io_stream_real_close_async);
g_task_set_check_cancellable (task, FALSE);
g_task_set_priority (task, io_priority);
-
- g_task_run_in_thread (task, close_async_thread);
- g_object_unref (task);
+
+ input = g_io_stream_get_input_stream (stream);
+ output = g_io_stream_get_output_stream (stream);
+
+ if (g_input_stream_async_close_is_via_threads (input) && g_output_stream_async_close_is_via_threads (output))
+ {
+ /* No sense in dispatching to the thread twice -- just do it all
+ * in one go.
+ */
+ g_task_run_in_thread (task, close_async_thread);
+ g_object_unref (task);
+ }
+ else
+ {
+ CloseAsyncData *data;
+
+ /* We should avoid dispatching to another thread in case either
+ * object that would not do it for itself because it may not be
+ * threadsafe.
+ */
+ data = g_slice_new (CloseAsyncData);
+ data->error = NULL;
+ data->pending = 2;
+
+ g_task_set_task_data (task, data, NULL);
+ g_input_stream_close_async (input, io_priority, cancellable, stream_close_complete, task);
+ g_output_stream_close_async (output, io_priority, cancellable, stream_close_complete, task);
+ }
}
static gboolean
* @stream2: a #GIOStream.
* @flags: a set of #GIOStreamSpliceFlags.
* @io_priority: the io priority of the request.
- * @cancellable: (allow-none): optional #GCancellable object, %NULL to ignore.
+ * @cancellable: (nullable): optional #GCancellable object, %NULL to ignore.
* @callback: (scope async): a #GAsyncReadyCallback.
* @user_data: (closure): user data passed to @callback.
*
- * Asyncronously splice the output stream of @stream1 to the input stream of
+ * Asynchronously splice the output stream of @stream1 to the input stream of
* @stream2, and splice the output stream of @stream2 to the input stream of
* @stream1.
*
ctx->completed = 0;
task = g_task_new (NULL, cancellable, callback, user_data);
+ g_task_set_source_tag (task, g_io_stream_splice_async);
g_task_set_task_data (task, ctx, (GDestroyNotify) splice_context_free);
if (cancellable != NULL)
projects / platform / upstream / glib.git / blobdiff