Update of /cvsroot/boost/boost/boost/interprocess/ipc
In directory sc8-pr-cvs3.sourceforge.net:/tmp/cvs-serv19387/ipc
Modified Files:
message_queue.hpp
Log Message:
Changes to correct regression tests for intel-win-9.1 & cw-9.4
Index: message_queue.hpp
===================================================================
RCS file: /cvsroot/boost/boost/boost/interprocess/ipc/message_queue.hpp,v
retrieving revision 1.5
retrieving revision 1.6
diff -u -d -r1.5 -r1.6
--- message_queue.hpp 4 May 2007 20:53:10 -0000 1.5
+++ message_queue.hpp 12 May 2007 12:51:21 -0000 1.6
@@ -1,6 +1,6 @@
//////////////////////////////////////////////////////////////////////////////
//
-// (C) Copyright Ion Gaztañaga 2005-2007. Distributed under the Boost
+// (C) Copyright Ion Gaztanaga 2005-2007. Distributed under the Boost
// Software License, Version 1.0. (See accompanying file
// LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
//
@@ -154,29 +154,16 @@
/*!Returns the needed memory size for the shared message queue. Never
throws*/
static std::size_t get_mem_size(std::size_t max_msg_size, std::size_t
max_num_msg);
- /*!This header is the prefix of each message in the queue*/
- class msg_hdr_t;
- friend class msg_hdr_t;
-
- /*!This functor is the predicate to order stored messages by priority*/
- class priority_functor;
- friend class priority_functor;
-
- class mq_hdr_t;
- friend class mq_hdr_t;
-
- /*!This is the atomic functor to be executed when creating or opening
- shared memory. Never throws*/
- class initialization_func_t;
- friend class initialization_func_t;
-
detail::managed_open_or_create_impl<shared_memory_object> m_shmem;
/// @endcond
};
/// @cond
+
+namespace detail {
+
/*!This header is the prefix of each message in the queue*/
-class message_queue::msg_hdr_t
+class msg_hdr_t
{
public:
std::size_t len; // Message length
@@ -186,7 +173,7 @@
};
/*!This functor is the predicate to order stored messages by priority*/
-class message_queue::priority_functor
+class priority_functor
{
public:
bool operator()(const offset_ptr<msg_hdr_t> &msg1,
@@ -227,11 +214,11 @@
msg_hdr_t structure. Each of this message is pointed by one pointer of
the index structure.
*/
-class message_queue::mq_hdr_t
- : public priority_functor
+class mq_hdr_t
+ : public detail::priority_functor
{
typedef offset_ptr<msg_hdr_t> msg_hdr_ptr_t;
-public:
+ public:
/*!Constructor. This object must be constructed in the beginning of the
shared memory of the size returned by the function "get_mem_size".
This constructor initializes the needed resources and creates
@@ -285,11 +272,11 @@
(std::size_t max_msg_size, std::size_t max_num_msg)
{
const std::size_t
- msg_hdr_align = boost::alignment_of<msg_hdr_t>::value,
+ msg_hdr_align = boost::alignment_of<detail::msg_hdr_t>::value,
index_align = boost::alignment_of<msg_hdr_ptr_t>::value,
r_hdr_size = detail::ct_rounded_size<sizeof(mq_hdr_t),
index_align>::value,
r_index_size =
detail::get_rounded_size(sizeof(msg_hdr_ptr_t)*max_num_msg, msg_hdr_align),
- r_max_msg_size = detail::get_rounded_size(max_msg_size,
msg_hdr_align) + sizeof(msg_hdr_t);
+ r_max_msg_size = detail::get_rounded_size(max_msg_size,
msg_hdr_align) + sizeof(detail::msg_hdr_t);
return r_hdr_size + r_index_size + (max_num_msg*r_max_msg_size) +
detail::managed_open_or_create_impl<shared_memory_object>::ManagedOpenOrCreateUserOffset;
}
@@ -299,18 +286,18 @@
void initialize_memory()
{
const std::size_t
- msg_hdr_align = boost::alignment_of<msg_hdr_t>::value,
+ msg_hdr_align = boost::alignment_of<detail::msg_hdr_t>::value,
index_align = boost::alignment_of<msg_hdr_ptr_t>::value,
r_hdr_size = detail::ct_rounded_size<sizeof(mq_hdr_t),
index_align>::value,
r_index_size =
detail::get_rounded_size(sizeof(msg_hdr_ptr_t)*m_max_num_msg, msg_hdr_align),
- r_max_msg_size = detail::get_rounded_size(m_max_msg_size,
msg_hdr_align) + sizeof(msg_hdr_t);
+ r_max_msg_size = detail::get_rounded_size(m_max_msg_size,
msg_hdr_align) + sizeof(detail::msg_hdr_t);
//Pointer to the index
msg_hdr_ptr_t *index = reinterpret_cast<msg_hdr_ptr_t*>
(detail::char_ptr_cast(this)+r_hdr_size);
//Pointer to the first message header
- msg_hdr_t *msg_hdr = reinterpret_cast<msg_hdr_t*>
+ detail::msg_hdr_t *msg_hdr = reinterpret_cast<detail::msg_hdr_t*>
(detail::char_ptr_cast(this)+r_hdr_size+r_index_size);
//Initialize the pointer to the index
@@ -319,7 +306,7 @@
//Initialize the index so each slot points to a preallocated message
for(std::size_t i = 0; i < m_max_num_msg; ++i){
index[i] = msg_hdr;
- msg_hdr = reinterpret_cast<msg_hdr_t*>
+ msg_hdr = reinterpret_cast<detail::msg_hdr_t*>
(detail::char_ptr_cast(msg_hdr)+r_max_msg_size);
}
}
@@ -344,14 +331,14 @@
/*!This is the atomic functor to be executed when creating or opening
shared memory. Never throws*/
-class message_queue::initialization_func_t
+class initialization_func_t
{
public:
initialization_func_t(std::size_t maxmsg = 0,
std::size_t maxmsgsize = 0)
: m_maxmsg (maxmsg), m_maxmsgsize(maxmsgsize) {}
- bool operator()(void *address, std::size_t size, bool created)
+ bool operator()(void *address, std::size_t, bool created)
{
char *mptr;
@@ -371,6 +358,8 @@
const std::size_t m_maxmsg;
const std::size_t m_maxmsgsize;
};
+
+} //namespace detail {
/// @endcond
inline message_queue::~message_queue()
@@ -378,7 +367,7 @@
inline std::size_t message_queue::get_mem_size
(std::size_t max_msg_size, std::size_t max_num_msg)
-{ return mq_hdr_t::get_mem_size(max_msg_size, max_num_msg); }
+{ return detail::mq_hdr_t::get_mem_size(max_msg_size, max_num_msg); }
inline message_queue::message_queue(detail::create_only_t create_only,
const char *name,
@@ -391,7 +380,7 @@
read_write,
(void*)0,
//Prepare initialization functor
- initialization_func_t (max_num_msg, max_msg_size))
+ detail::initialization_func_t (max_num_msg, max_msg_size))
{}
/*!Opens or creates a process shared message queue with name "name".
@@ -410,7 +399,7 @@
read_write,
(void*)0,
//Prepare initialization functor
- initialization_func_t (max_num_msg, max_msg_size))
+ detail::initialization_func_t (max_num_msg, max_msg_size))
{}
/*!Opens a previously created process shared message queue with name
"name".
@@ -424,7 +413,7 @@
read_write,
(void*)0,
//Prepare initialization functor
- initialization_func_t ())
+ detail::initialization_func_t ())
{}
inline void message_queue::send
@@ -444,7 +433,7 @@
const void *buffer, std::size_t
buffer_size,
unsigned int priority, const
boost::posix_time::ptime &abs_time)
{
- mq_hdr_t *p_hdr = static_cast<mq_hdr_t*>(m_shmem.get_address());
+ detail::mq_hdr_t *p_hdr =
static_cast<detail::mq_hdr_t*>(m_shmem.get_address());
//Check if buffer is smaller than maximum allowed
if (buffer_size > p_hdr->m_max_msg_size) {
throw interprocess_exception(size_error);
@@ -482,7 +471,7 @@
}
//Get the first free message from free message queue
- msg_hdr_t *free_msg = p_hdr->free_msg();
+ detail::msg_hdr_t *free_msg = p_hdr->free_msg();
if (free_msg == 0) {
throw interprocess_exception();
}
@@ -528,7 +517,7 @@
std::size_t &recvd_size, unsigned int &priority,
const boost::posix_time::ptime &abs_time)
{
- mq_hdr_t *p_hdr = static_cast<mq_hdr_t*>(m_shmem.get_address());
+ detail::mq_hdr_t *p_hdr =
static_cast<detail::mq_hdr_t*>(m_shmem.get_address());
//Check if buffer is big enough for any message
if (buffer_size < p_hdr->m_max_msg_size) {
throw interprocess_exception(size_error);
@@ -567,7 +556,7 @@
}
//Thre is at least message ready to pick, get the top one
- msg_hdr_t *top_msg = p_hdr->top_msg();
+ detail::msg_hdr_t *top_msg = p_hdr->top_msg();
//Paranoia check
if (top_msg == 0) {
@@ -597,18 +586,18 @@
inline std::size_t message_queue::get_max_msg() const
{
- mq_hdr_t *p_hdr = static_cast<mq_hdr_t*>(m_shmem.get_address());
+ detail::mq_hdr_t *p_hdr =
static_cast<detail::mq_hdr_t*>(m_shmem.get_address());
return p_hdr ? p_hdr->m_max_num_msg : 0; }
inline std::size_t message_queue::get_max_msg_size() const
{
- mq_hdr_t *p_hdr = static_cast<mq_hdr_t*>(m_shmem.get_address());
+ detail::mq_hdr_t *p_hdr =
static_cast<detail::mq_hdr_t*>(m_shmem.get_address());
return p_hdr ? p_hdr->m_max_msg_size : 0;
}
inline std::size_t message_queue::get_num_msg()
{
- mq_hdr_t *p_hdr = static_cast<mq_hdr_t*>(m_shmem.get_address());
+ detail::mq_hdr_t *p_hdr =
static_cast<detail::mq_hdr_t*>(m_shmem.get_address());
if(p_hdr){
//---------------------------------------------
scoped_lock<interprocess_mutex> lock(p_hdr->m_mutex);
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Boost-cvs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/boost-cvs
