Sean Bright [Thu, 5 May 2011 21:20:00 +0000 (21:20 +0000)]
Add some new editline bindings by default, and allow for user specified configuration.
I excluded the part of this patch that used the HOME environment variable since
the built-in editline library goes to great lengths to disallow that. Instead
only settings the EDITRC environment variable will use a user specified file.
Also, the default environment variable use to determine the edit more is
AST_EDITMODE instead of AST_EDITOR (although the latter is still supported).
This is based on an uncommitted patch by jpeeler for the issue. Instead of
relocking and then unlocking the channel though, we keep the lock on the channel
until we are finished doing what we need to the channel.
Set SO_KEEPALIVE on SIP TCP sockets so that they eventually go away when a peer
abruptly disappears. This mostly occurs after a successful registration.
(closes issue #17544)
Reported by: marcelloceschia
Patches:
(modified) tcptls.patch uploaded by st (license 907)
........
Damien Wedhorn [Thu, 5 May 2011 09:03:32 +0000 (09:03 +0000)]
Add setsubstate_congestion and setsubstate_progress.
Move handling of both state handling from skinny_indicate to it's own sub.
Also, modified behaviour to not hangup the sub and let the dialplan
have a chance in doing what it wants for congestion. Added various states to
substate2str and added these states where applicable for other set_substate_
procs.
Damien Wedhorn [Thu, 5 May 2011 08:10:14 +0000 (08:10 +0000)]
Add setsubstate_busy.
Move handling of setting busy state from skinny_indicate to it's own sub.
Also, modified behaviour to not hangup the sub and let the dialplan
have a chance in doing what it wants (eg busy(10); hangup() in the dialplan
now gives a busy indication for 10 secs and then hangs up.
Stefan Schmidt [Thu, 5 May 2011 07:09:20 +0000 (07:09 +0000)]
Adding the Move to Front Hash functionality
Moving a found object to the front of its bucket to reduce the necessary traversal steps to find an object. This change improves the search time on large system with many data or in link lists.
Use the correct HTTP method when generating our digest, otherwise we always fail.
When calculating the 'A2' portion of our digest for verification, we need the
HTTP method that is currently in use. Unfortunately our mapping function was
incorrect, resulting in invalid hashes being generated and, in turn, failures
in authentication.
Damien Wedhorn [Wed, 4 May 2011 21:44:06 +0000 (21:44 +0000)]
Add setsubstate_ringout (equivalent to AST_STATE ringing).
Renamed previous setsubstate_ringout to setsubstate_dialing for a state
when attempting to dial a number, substate ringout now for when core
has indicated that the channel is actually ringing on the other end.
Also added substate2str for debugging purposes.
Fixes session-timers=refuse not being enforced for *caller*
During handle_request_invite, the session timer mode was retrieved from
a cached variable. This patch forces a peer lookup of the session timer
mode in the case of an incoming invite.
Damien Wedhorn [Wed, 4 May 2011 08:25:47 +0000 (08:25 +0000)]
Add setsubstate_ringin.
Added setsubstate_ringin. skinny_call now calls sss_ringin rather than inline.
Fixed previous issue so that setsubstate_connected now use SUBSTATE_RINGIN
to determine is an AST_CONTROL_ANSWER should be queued.
Damien Wedhorn [Wed, 4 May 2011 07:10:04 +0000 (07:10 +0000)]
Cleanup skinny callinfo.
Cosolidated the working out of the callinfo to be sent into
transmit_callinfo. Replaced ambiguous sub->outgoing with calldirection
which can be SKINNY_INCOMING or SKINNY_OUTGOING (same value as the
skinny protocol).
Tilghman Lesher [Tue, 3 May 2011 23:36:35 +0000 (23:36 +0000)]
If multiple [general] contexts occur from sip.conf (usually due to external includes), merge them.
The original implementation of this did the merging of all contexts with the
same name in the realtime layer, but that implementation severely breaks
drivers which use the same context name (e.g. iax.conf, type={peer,user}).
Therefore, the implementation needs to do the merging for particular entries
only, based upon what contexts would allow that in the channel driver itself.
This implementation is for chan_sip only, but others could be added in the
future.
The dahdi_hangup() call does not clean up the channel fully.
After dahdi_hangup() has supposedly hungup an ISDN channel there is still
traffic on the S0-bus because the channel was not cleaned up fully.
Shuffled the hangup code to include some missing cleanup. Also fixed some
code formatting in the area. I think the primary missing clean up code
was the call to tone_zone_play_tone() to turn off any active tones on the
channel.
This has already been discussed and should have been resolved earlier. View
revsion 285565's log for more information about why it is important to not
put timer in the Require header.
This change optimizes the free_via() function and removes some redundant null
checking. It also fixes compliance with RFC 3261 section 18.2.2 by always using
the port specified in the Via header for routing responses (even when maddr is
not set). Also the htons() function is now used when setting the port.
Additional documentation comments have been added in various places to make the
logic in the code clearer.
Allow transfer loops without allowing forwarding loops
We try to avoid the situation where two phones may be forwarded to each other
causing an infinite loop by storing each dialed interface in a channel
datastore and checking the list before dialing out. This works, but currently
breaks situations like A calls B, A transfers B to C, B transfers C to A, and A
transfers C to B. Since human interaction is happening here and not an
automated forwarding loop, it should be allowed.
This patch removes the dialed_interfaces datastore when a call is bridged (a
suggestion from the brilliant mmichelson). If a call is being bridged, it
should be safe to assume that we aren't stuck in a loop.
Since we are now handling this is the bridge code, the previous attempts at
handling it in app_dial and app_queue are removed.
The 'e' special extension fails to trigger in at least two cases.
The 'e' extension is a fall back for the 'i', 't', or 'T' extensions if
any of them do not exist. Many of the places the 'e' extension was
supposed to be invoked fail because the priority was set wrong. There
were two places where the 'e' extension was not even checked for fall
back.
* Made invoke the 'e' extension similarly to the previous 'i', 't', or 'T'
extension check and added the 'e' extension as a fall back to the two
missing locations.
* Prioritized and optimized some hangup tests associated with the 'e'
extension.
The code that set the bit within the select bitfield was correct, but the
bounds-checking code was not. The change to that line uses the new _bitsize
macro for clarity. Also, FD_ZERO macro did not zero-out anything but the
first word of the bitfield, so this could have caused problems with modules
using that macro with the expanded bitfield.
This patch resolves a fairly complex deadlock that can occur with the
combination of chan_local and a dialplan switch, such as dynamic realtime
extensions, which pulls autoservice into the picture when doing a dialplan
lookup.
(closes issue #18818)
Reported by: nic
Patches:
issue18818.patch uploaded by jthurman (license 614)
18818.v1.txt uploaded by russell (license 2)
Tested by: nic, jthurman, kterzi, steve-howes, sysreq, IshMalik
........
Initialize the buffers used to hold the result from GET VARIABLE or
GET VARIABLE FULL. The bug report shows func_read returning garbage in
the result. It assumed that the buffer passed in was initialized, like many
other functions do. In the more common code path (through the dialplan), it
is initialized, so just initialize it here too.
Added limits to the number of unauthenticated sessions TCP based protocols are allowed to have open simultaneously. Also added timeouts for unauthenticated sessions where it made sense to do so.
Unrelated, the manager interface now properly checks if the user has the "system" privilege before executing shell commands via the Originate action.
AST-2011-005
AST-2011-006
(closes issue #18787)
Reported by: kobaz
(related to issue #18996)
Reported by: tzafrir
........
................
................
Don't allocate more space than necessary for a sip_pkt
This extra allocation is a hold-over from when pkt->data was a
character array. Now that it is an allocated string, just allocate
enough for the sip_pkt.
........
................
codec_dahdi: DAHDI still advertises formats using the old bitfields.
Previously, the DAHDI format bit fields matched up with the Asterisk
bitfields. Since the Asterisk codec bit fields were replaced in r306010,
codec_dahdi needs to contain the formats itself. In the future, the DAHDI
formats should either change to something other than bitfields, or the
bitfields need to move from include/dahdi/kernel.h to
include/dahdi/user.h.
Richard Mudgett [Mon, 18 Apr 2011 19:48:00 +0000 (19:48 +0000)]
Problems with ISDN MWI to phones.
The "controlling user number" is always the number of the voice mail box
which is identical with the subscriber number itself. This number which
is listed in the ISDN phone MWI menu cannot be called back to contact the
voice mail box. The controlling user number should be made configurable.
The AsyncAGI command loop is lax in the value it returns for the return status.
* Return correct status: SUCCESS/FAILED/HANGUP. Previously, abnormal
exits from the command loop such as hangup would return SUCCESS.
* The "asyncagi break" command now returns SUCCESS and is now the only way
to break the command loop with that status. Previously, it returned
FAILED.
* The AMI event AsyncAGI End is no longer sent if the AsyncAGI Start event
is not sent. Previously, this happened because of an error setting up the
AGI pipes.
* All executed AGI commands now get an AsyncAGI Exec result event.
Previously, if the command returned failure (because of hangup), the
command loop just exited with FAILURE and did not send the AsyncAGI Exec
result event.
* Makes sure that the channel frame queue is empty on hangup.
Remove the need for deadlock avoidance in chan_sip do_monitor.
Deadlock avoidance between the sip pvt and the pvt->owner is
very difficult. Now that channel's are ao2 objects, this complication
is no longer necessary. It turns out the pvt's msg queue only
exists because of deadlock avoidance (when deadlock avoidance fails
msgs were added to a queue to be processed later), so this goes away as well.
The technique used in the new sip_lock_pvt_full() function should
be used as a template for replacing all locations where deadlock
avoidance occurs between a channel tech_pvt and the pvt's owner.
My hope is that this will begin a reversal of the invalid channel
driver locking architecture we have been using for so long.
This patch also resolves an issue where the pvt->owner gets
unlocked during processing the msg queue.
sip codec negotiation of dynamic rtp payloads error fix
This patch fixes how chan_sip handles dynamic rtp payload types
it does not understand. At the moment if a dynamic payload's mime
type does not match one we understand, the payload does not get
removed from our payload table. As a result of this, the payload
is set to whatever dynamic codec we use internally for that payload
number on outgoing INVITES. This is incorrect.
This patch fixes this by properly checking the rtpmap set function's
return code to make sure it was found. The function can return both
-1 and -2 depending on the source of the mismatch. We were just
checking -1 explicitly.
Terry Wilson [Thu, 14 Apr 2011 21:53:01 +0000 (21:53 +0000)]
Sets video mark bit on format field correctly
This fixes a regression in the media architecture change
where video frames did not have their video mark set
correctly. dvossel wrote this. twilson kindly committed
this, mmichelson found the bug.
Richard Mudgett [Thu, 14 Apr 2011 18:22:35 +0000 (18:22 +0000)]
Add Device State Information CCSS for Generic Devices.
Add Asterisk Device State information and callbacks to the Call Completion
Supplemental Services for generic agents.
There are currently not many devices that have native support for CCSS.
Even as the devices become available there may be other reasons why one
may choose to not take advantage of the native abilities and stick with
the generic implementation. The generic implementation is quite capable
and could be greatly enhanced by adding device state capabilities. A
phone could then subscribe to the device state with a BLF key in
conjunction with Asterisk hints.
The advantages of the device state information would allow a single button
to: request CCSS, cancel a CCSS request, and display the current state of
a CCSS request.
For example, you may have a single button that when not lit, there is no
active CCSS request. When you press that button, the dialplan can query
the DEVICE_STATE() associated with that caller to determine whether they
should be calling CallCompletionRequest() or CallCompletionCancel(). If
there is currently a pending request, then the dialplan would cancel it.
This also has the advantage of showing the true state of a request, which
is an asynchronous call, even when CallCompletionRequest() thinks it was
successful. The actual request could ultimately fail. Once lit, further
feedback can be provided to the caller about the current state of their
request since it will be updated by the CCSS State Machine as appropriate.
The DEVICE_STATE mapping is configurable since the BLF being used on a
given phone type may vary. The idea is to allow some level of
customization as to the phone's behavior.
As an example, you may want the BLF key to go solid once you have
requested a callback. You may then want the LED to blink (typically
ringing) when either the callback is in process, which is a visual
indication that the incoming call is the desired callback. You may want
it to blink when the callee is ready but you are busy, giving you a visual
indication that the target is available as you may want to get off the
line so that the callback can be successful.
Device state information is sent back via the ast_devstate_prov_add()
callback for any generic CCSS device as it traverses through the state
machine. You simply provide a map between CC_STATE values and the
corresponding AST_DEVICE state values.
You could then generate hints against these states similar to what is
possible today with Custom Devstates or MeetMe states. For example, you
may have an extension 3000 that is currently associated with device
SIP/3000. You could then create a feature code for that extension that
may look something like:
exten => *823000,hint,ccss:sip/3000
You would then subscribe a BLF button to *823000 which would point to the
dialplan that handled CCSS requests/cancels using the available
DEVICE_STATE() information about ccss:sip/3000 to make the decision about
what to do.
(closes issue #18788)
Reported by: p_lindheimer
Patches:
ccss.trunk.18788.patch uploaded by p lindheimer (license 558)
Modified with final reviewboard comments.
Tested by: p_lindheimer, loloski
Asterisk does not hangup a channel after endpoint hangs up.
If the call that the dialplan started an AGI script for is hungup while
the AGI script is in the middle of a command then the AGI script is not
notified of the hangup. There are many AGI Exec commands that this can
happen with. The reported applications have been: Background, Wait, Read,
and Dial. Also the AGI Get Data command.
* Don't wait on the Asterisk channel after it has hung up. The channel is
likely to never need servicing again.
* Restored the AGI script's ability to return the AGI_RESULT_HANGUP value
in run_agi(). It previously only could return AGI_RESULT_SUCCESS or
AGI_RESULT_FAILURE after the DeadAGI and AGI applications were merged.
(closes issue #17954)
Reported by: mn3250
Patches:
issue17954_v1.8.patch uploaded by rmudgett (license 664)
issue17954_v1.6.2.patch uploaded by rmudgett (license 664)
issue17954_v1.4.patch uploaded by rmudgett (license 664)
Tested by: rmudgett
JIRA SWP-2171
Alexandr Anikin [Tue, 12 Apr 2011 21:59:18 +0000 (21:59 +0000)]
IPv6 support for chan_ooh323
IPv6 support for ooh323,
bindaddr, peers and users ip can be IPv4 or IPv6 addr
correction for multi-homed mode (0.0.0.0 or :: bindaddr)
can work in dual 6/4 mode with :: bindaddr
gatekeeper mode isn't supported in v6 mode while
chan_dahdi reloading with variables set via setvar in chan_dahdi.conf would
stay in the dahdi_pvt structs for individual channels (causing them to just
continue adding the new ones to the list) and also there was a memory leak
causes by the conf objects. This patch resolves both of these by using
ast_variables_destroy during the loading process.
Backport a restructuring change from trunk to make the next change stand out.
........
r313369 | rmudgett | 2011-04-11 18:08:02 -0500 (Mon, 11 Apr 2011) | 13 lines
Frames from the inbound channel should go to all outbound channels in app_dial.c.
In app_dial.c:wait_for_answer() frames from the inbound channel should be
sent to all outbound channels instead of only if there is just one
outbound channel.
Control frames like AST_CONTROL_CONNECTED_LINE need to be passed to all of
the the outbound channels. This can happen if a blond transfer is done by
a remote switch on the inbound channel.
Stuck channel using FEATD_MF if caller hangs up at the right time.
The cause was actually a caller hanging up just at the end of the Feature
Group D DTMF tones that setup the call. The reason for this is a "guard
timer" that's implemented using ast_safe_sleep(100). If the caller
happens to hang up AFTER the final tone of the DTMF string but BEFORE the
end of that ast_safe_sleep(), then ast_safe_sleep() will return non-zero.
This causes the code to bounce to the end of ss_thread(), but it does NOT
tear down the call properly.
This should be a rare occurrence because the caller has to hang up at
EXACTLY the right time. Nonetheless, it was happening quite regularly on
the reporter's system. It's not easily reproducible, unless you purposely
increase the guard-time to 2000 or more. Once you do that, you can
reproduce it every time by watching the DTMF debug and hanging up just as
it ends.
projects / thirdparty / asterisk.git / log
