cleanup more on att-xfer

git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@15134 d0543943-73ff-0310-b7d9-9358b9ac24b2
This commit is contained in:
Anthony Minessale
2009-10-09 22:28:31 +00:00
parent 7681f944b4
commit 06c1349b08
4 changed files with 80 additions and 25 deletions

View File

@@ -1664,6 +1664,22 @@ SWITCH_DECLARE(void) switch_channel_clear_state_handler(switch_channel_t *channe
switch_mutex_unlock(channel->state_mutex);
}
SWITCH_DECLARE(void) switch_channel_restart(switch_channel_t *channel)
{
switch_channel_set_state(channel, CS_RESET);
switch_channel_wait_for_state_timeout(channel, CS_RESET, 5000);
switch_channel_set_state(channel, CS_EXECUTE);
}
/* XXX This is a somewhat simple operation. Were essentially taking the extension that one channel
was executing and generating a new extension for another channel that starts out where the
original one left off with an optional forward offset. Since all we are really doing is
copying a few basic pool-allocated structures from one channel to another there really is
not much to worry about here in terms of threading since we use read-write locks.
While the features are nice, they only really are needed in one specific crazy attended
transfer scenario where one channel was in the middle of calling a particular extension
when it was rudely cut off by a transfer key press. XXX */
SWITCH_DECLARE(switch_status_t) switch_channel_caller_extension_masquerade(switch_channel_t *orig_channel, switch_channel_t *new_channel, uint32_t offset)
{
switch_caller_profile_t *caller_profile;