Mark Michelson [Fri, 9 Jan 2009 20:08:04 +0000 (20:08 +0000)]
Add check_via calls to more request handlers
INFO, NOTIFY, OPTIONS, REFER, and MESSAGE requests
were not checking the topmost Via to determine where
to send the response. Adding check_via calls to those
request handlers solves this.
When a SIP request or response arrives for a dialog with an associated Asterisk channel, and the lock on that channel cannot be obtained because it is held by another thread, instead of dropping the request/response, queue it for later processing when the channel lock becomes available.
Mark Michelson [Mon, 5 Jan 2009 16:51:59 +0000 (16:51 +0000)]
A couple of changes to T.38 SDP attribute handling
There are some boolean attributes for T.38 such
as T38FaxFillBitRemoval, T38FaxTranscodingMMR, and
T38FaxTranscodingJBIG. By simply being present, we
should treat these as a "true" value. The current
code, however, was requiring a 1 or 0 as the value
of the attribute in order to parse it. This is due
to the fact that there are some T.38 endpoints and
gateways that also transmit this information
incorrectly. This patch follows the "be liberal in
what you accept and strict in what you send"
philosophy by accepting both the correctly- and
incorrectly-formatted attributes, but only sending
information as it is supposed to be sent.
It was also discovered that a particular type of
T.38 gateway sends some non-standard T.38 SDP
attributes. Instead of using T38FaxMaxDatagram
and T38MaxBitRate, it used T38MaxDatagram and
T38FaxMaxRate respectively. We now will properly
accept these attributes as well.
Note that there are a lot of patches cited in
the below commit message template. This is
because the person who submitted these patches is
an awesome person and wrote 1.4, 1.6.0, and 1.6.1
variants.
(closes issue #13976)
Reported by: linulin
Patches:
chan_sip.c.1.4-update1.diff uploaded by arcivanov (license 648)
chan_sip.c.1.6.0-update1.diff uploaded by arcivanov (license 648)
chan_sip.c.1.6.1-update1.diff uploaded by arcivanov (license 648)
chan_sip.c.1.4-relaxedT38_update1.diff uploaded by arcivanov (license 648)
chan_sip.c.1.6.0-relaxedT38_update1.diff uploaded by arcivanov (license 648)
chan_sip.c.1.6.1-relaxedT38_update1.diff uploaded by arcivanov (license 648)
Tested by: arcivanov
Tilghman Lesher [Thu, 1 Jan 2009 00:01:22 +0000 (00:01 +0000)]
Repeat attempts to write when we receive -EAGAIN from the driver, as detailed
in the ALSA sample code (see http://www.alsa-project.org/alsa-doc/alsa-lib/_2test_2pcm_8c-example.html#a32)
Reported by: Jerry Geis (via the -users list)
Fixed by: me (license 14)
Mark Michelson [Mon, 22 Dec 2008 20:56:29 +0000 (20:56 +0000)]
Fix a deadlock relating to channel locks and autoservice
It has been discovered that if a channel is locked prior
to a call to ast_autoservice_stop, then it is likely that
a deadlock will occur. The reason is that the call to
ast_autoservice_stop has a check built into it to be sure
that the thread running autoservice is not currently trying
to manipulate the channel we are about to pull out of
autoservice.
The autoservice thread, however, cannot advance beyond where
it currently is, though, because it is trying to acquire
the lock of the channel for which autoservice is attempting
to be stopped.
The gist of all this is that a channel MUST NOT be locked
when attempting to stop autoservice on the channel.
In this particular case, the channel was locked by a call
to ast_read. A call to ast_exists_extension led to autoservice
being started and stopped due to the existence of dialplan
switches.
It may be that there are future commits which handle the same
symptoms but in a different location, but based on my looks through
the code, it is very rare to see a construct such as this one.
Steve Murphy [Fri, 19 Dec 2008 22:30:32 +0000 (22:30 +0000)]
This merges the masqpark branch into 1.4
These changes eliminate the need for (and use of)
the KEEPALIVE return code in res_features.c;
There are other places that use this result code
for similar purposes at a higher level, these appear
to be left alone in 1.4, but attacked in trunk.
The reason these changes are being made in 1.4, is
that parking ends a channel's life, in some situations,
and the code in the bridge (and some other places),
was not checking the result code properly, and dereferencing
the channel pointer, which could lead to memory corruption
and crashes.
Calling the masq_park function eliminates this danger
in higher levels.
A series of previous commits have replaced some parking calls
with masq_park, but this patch puts them ALL to rest,
(except one, purposely left alone because a masquerade
is done anyway), and gets rid of the code that tests
the KEEPALIVE result, and the NOHANGUP_PEER result codes.
While bug 13820 inspired this work, this patch does
not solve all the problems mentioned there.
I have tested this patch (again) to make sure I have
not introduced regressions.
Crashes that occurred when a parked party hung up
while the parking party was listening to the numbers
of the parking stall being assigned, is eliminated.
These are the cases where parking code may be activated:
1. Feature one touch (eg. *3)
2. Feature blind xfer to parking lot (eg ##700)
3. Run Park() app from dialplan (eg sip xfer to 700)
(eg. dahdi hookflash xfer to 700)
4. Run Park via manager.
The interesting testing cases for parking are:
I. A calls B, A parks B
a. B hangs up while A is getting the numbers announced.
b. B hangs up after A gets the announcement, but
before the parking time expires
c. B waits, time expires, A is redialed,
A answers, B and A are connected, after
which, B hangs up.
d. C picks up B while still in parking lot.
II. A calls B, B parks A
a. A hangs up while B is getting the numbers announced.
b. A hangs up after B gets the announcement, but
before the parking time expires
c. A waits, time expires, B is redialed,
B answers, A and B are connected, after
which, A hangs up.
d. C picks up A while still in parking lot.
Testing this throroughly involves acting all the permutations
of I and II, in situations 1,2,3, and 4.
Since I added a few more changes (ALL references to KEEPALIVE in the bridge
code eliimated (I missed one earlier), I retested
most of the above cases, and no crashes.
H-extension weirdness.
Current h-extension execution is not completely
correct for several of the cases.
For the case where A calls B, and A parks B, the
'h' exten is run on A's channel as soon as the park
is accomplished. This is expected behavior.
But when A calls B, and B parks A, this will be
current behavior:
After B parks A, B is hung up by the system, and
the 'h' (hangup) exten gets run, but the channel
mentioned will be a derivative of A's...
Thus, if A is DAHDI/1, and B is DAHDI/2,
the h-extension will be run on channel
Parked/DAHDI/1-1<ZOMBIE>, and the
start/answer/end info will be those
relating to Channel A.
And, in the case where A is reconnected to
B after the park time expires, when both parties
hang up after the joyful reunion, no h-exten
will be run at all.
In the case where C picks up A from the
parking lot, when either A or C hang up,
the h-exten will be run for the C channel.
CDR's are a separate issue, and not addressed
here.
As to WHY this strange behavior occurs,
the answer lies in the procedure followed
to accomplish handing over the channel
to the parking manager thread. This procedure
is called masquerading. In the process,
a duplicate copy of the channel is created,
and most of the active data is given to the
new copy. The original channel gets its name
changed to XXX<ZOMBIE> and keeps the PBX
information for the sake of the original
thread (preserving its role as a call
originator, if it had this role to begin
with), while the new channel is without
this info and becomes a call target (a
"peer").
In this case, the parking lot manager
thread is handed the new (masqueraded)
channel. It will not run an h-exten
on the channel if it hangs up while
in the parking lot. The h exten will
be run on the original channel instead,
in the original thread, after the bridge
completes.
See bug 13820 for our intentions as
to how to clean up the h exten behavior.
Jeff Peeler [Fri, 19 Dec 2008 19:48:00 +0000 (19:48 +0000)]
(closes issue #13480)
Reported by: tzafrir
Replace a bunch of if defined checks for Zaptel/DAHDI through several new defines in dahdi_compat.h. This removes a lot of code duplication. Example from bug:
Russell Bryant [Thu, 18 Dec 2008 21:39:25 +0000 (21:39 +0000)]
Make ast_carefulwrite() be more careful.
This patch handles some additional cases that could result in partial writes
to the file description. This was done to address complaints about partial
writes on AMI.
(issue #13546) (more changes needed to address potential problems in 1.6)
Reported by: srt
Tested by: russell
Review: http://reviewboard.digium.com/r/99/
Tilghman Lesher [Thu, 18 Dec 2008 21:14:47 +0000 (21:14 +0000)]
Add mutexes around accesses to the IMAP library interface. This prevents
certain crashes, especially when shared mailboxes are used.
(closes issue #13653)
Reported by: howardwilkinson
Patches:
asterisk-1.4.21.2-appvoicemail-sharedimap-lock.patch uploaded by howardwilkinson (license 590)
Tested by: jpeeler
Joshua Colp [Thu, 18 Dec 2008 17:11:42 +0000 (17:11 +0000)]
Only care about a compatible codec for early bridging if we are actually bridging to another channel. If we are not we actually want to bring the audio back to us.
(closes issue #13545)
Reported by: davidw
Mark Michelson [Tue, 16 Dec 2008 23:04:27 +0000 (23:04 +0000)]
After looking through SIP registration code most of the day, this
is one of the few things I could find that was just plain wrong.
Even though it probably isn't possible for it to happen, it seems weird
to have code that checks if a pointer is NULL and then immediately dereferences
that pointer if it was NULL.
Russell Bryant [Tue, 16 Dec 2008 21:10:44 +0000 (21:10 +0000)]
Do not dereference the channel if AST_PBX_KEEPALIVE has been returned.
This is a bug I noticed while looking at the code for app_macro. This return code
means that another thread has assumed ownership of the channel and it can no longer
be touched. (I hate this return code with a passion, by the way.)
Russell Bryant [Tue, 16 Dec 2008 20:35:25 +0000 (20:35 +0000)]
Add "restart gracefully" to the AMI blacklist of CLI commands.
"module unload" was already identified as a command that can not be used
from the AMI. "restart gracefully" effectively unloads all modules, and will
run in to the same problems.
Russell Bryant [Tue, 16 Dec 2008 17:06:29 +0000 (17:06 +0000)]
Fix memory leak and invalid reporting issues with DEBUG_THREADLOCALS.
One issue was that the ast_mutex_* API was being used within the context of the
thread local data destructors. We would go off and allocate more thread local data
while the pthread lib was in the middle of destroying it all. This led to a memory
leak.
Another issue was an invalid argument being provided to the the object_add
API call.
(closes issue #13678)
Reported by: ys
Tested by: russell
Russell Bryant [Tue, 16 Dec 2008 15:56:37 +0000 (15:56 +0000)]
Fix a memory leak related to the use of the "setvar" configuration option.
The problem was that these variables were being appended to the list of vars
on the sip_pvt every time a re-registration or re-subscription came in.
Since it's just a waste of memory to put them there unless the request was an
INVITE, then the fix is to check the request type before copying the vars.
(closes issue #14037)
Reported by: marvinek
Tested by: russell
Steve Murphy [Tue, 16 Dec 2008 15:15:58 +0000 (15:15 +0000)]
I added a sentence to clarify why - and ' ' are ignored in patterns
as per bug 14076. Leif says he'll put some stuff about it in the
extensions.conf sample, etc.
Joshua Colp [Mon, 15 Dec 2008 18:11:21 +0000 (18:11 +0000)]
Do not try to unlock a non-existant channel if the transfer fails.
(closes issue #13800)
Reported by: dwagner
Patches:
asterisk-1.4.22-chan-sip-nullp.patch uploaded by tweety (license 608)
Russell Bryant [Mon, 15 Dec 2008 14:31:37 +0000 (14:31 +0000)]
Handle a case where a call can be bridged to a channel that is still ringing.
The issue that was reported was about a case where a RINGING channel got
redirected to an extension to pick up a call from parking. Once the parked
call got taken out of parking, it heard silence until the other side answered.
Ideally, the caller that was parked would get a ringing indication. This patch
fixes this case so that the caller receives ringback once it comes out of
parking until the other side answers.
The fixes are:
- Make sure we remember that a channel was an outgoing channel when doing
a masquerade. This prevents an erroneous ast_answer() call on the channel,
which causes a bogus 200 OK to be sent in the case of SIP.
- Add some additional comments to explain related parts of code.
- Update the handling of the ast_channel visible_indication field. Storing
values that are not stateful is pointless. Control frames that are events
or commands should be ignored.
- When a bridge first starts, check to see if the peer channel needs to be
given ringing indication because the calling side is still ringing.
- Rework ast_indicate_data() a bit for the sake of readability.
Tilghman Lesher [Sat, 13 Dec 2008 23:22:02 +0000 (23:22 +0000)]
Change the default calldurationlimit from the special value 0 to -1, so we
can better detect an exceptional case. This follows on to the changes made
in revision 156386. Related to issue #13851.
(closes issue #13974)
Reported by: paradise
Patches:
20081208__bug13974.diff.txt uploaded by Corydon76 (license 14)
Tested by: file, blitzrage, ZX81
Tilghman Lesher [Fri, 12 Dec 2008 22:03:10 +0000 (22:03 +0000)]
Simple fix for Ctrl-C not immediately exiting Asterisk, but also add a
pointer inside editline to look back to asterisk.c, so others don't spend
as much time as I did looking (in the wrong place) for the appropriate
function.
Reported by: ZX81, via the #asterisk-users channel
Fixed by: me (license 14)
Russell Bryant [Fri, 12 Dec 2008 13:44:08 +0000 (13:44 +0000)]
Resolve issues that could cause DTMF to be processed out of order.
These changes come from team/russell/issue_12658
1) Change autoservice to put digits on the head of the channel's frame readq
instead of the tail. If there were frames on the readq that autoservice
had not yet read, the previous code would have resulted in out of order
processing. This required a new API call to queue a frame to the head
of the queue instead of the tail.
2) Change up the processing of DTMF in ast_read(). Some of the problems
were the result of having two sources of pending DTMF frames. There
was the dtmfq and the more generic readq. Both were used for pending
DTMF in various scenarios. Simplifying things to only use the frame
readq avoids some of the problems.
3) Fix a bug where a DTMF END frame could get passed through when it
shouldn't have. If code set END_DTMF_ONLY in the middle of digit emulation,
and a digit arrived before emulation was complete, digits would get
processed out of order.
Tilghman Lesher [Thu, 11 Dec 2008 23:35:55 +0000 (23:35 +0000)]
When a Ctrl-C or Ctrl-D ends a remote console, on certain shells, the terminal
is messed up. By intercepting those events with a signal handler in the remote
console, we can avoid those issues.
(closes issue #13464)
Reported by: tzafrir
Patches:
20081110__bug13464.diff.txt uploaded by Corydon76 (license 14)
Tested by: blitzrage
Russell Bryant [Thu, 11 Dec 2008 16:54:51 +0000 (16:54 +0000)]
Fix an issue that made it so you could only have a single caller executing
a custom feature at a time. This was especially problematic when custom
features ran for any appreciable amount of time.
The fix turned out to be quite simple. The dynamic features are now stored
in a read/write list instead of a list using a mutex.
(closes issue #13478)
Reported by: neutrino88
Fix suggested by file
Ensure that moh_generate does not return prematurely before local_ast_moh_stop is called. Also, the sleep in mp3_spawn now only occurs for http locations since it seems to have been added originally only for failing media streams.
Joshua Colp [Wed, 10 Dec 2008 19:01:17 +0000 (19:01 +0000)]
Fix subscription based MWI up a bit. We only want to put sip: at the beginning of the URI if it is not already there and revert code to ignore destination check if subscribing for MWI.
(closes issue #12560)
Reported by: vsauer
Patches:
patch001.diff uploaded by ramonpeek (license 266)
Joshua Colp [Wed, 10 Dec 2008 17:50:43 +0000 (17:50 +0000)]
When a SIP peer unregisters set the expiry time back to 0 so that the 200 OK contains an expires of 0.
(closes issue #13599)
Reported by: hjourdain
Patches:
chan_sip.c.diff uploaded by hjourdain (license 583)
I checked, and I added a mod to the trunk version
of Asterisk that would make it 8-bit transparent
on 27 Nov 2007, but I made no such updates to
1.4. My best guess is that 1.4 was released, and
it was not appropriate to commit an enhancement.
But I'm going to add the same fixes to 1.4 now,
for the following reasons:
1. wetwired is correct; 1.4 is **mostly** 8-bit
transparent now. This is because the lexical
token forming rules use . in most 'word'
state continuances. It's just the beginning
of a 'word' that is picky.
2. Accepting 8-bit chars in some places and
not others leads to bug reports like this.
This patch was originally uploaded to issue #13783 by
jamessan. While the issue was closed for other reasons, this
patch is valid and fixes a separate problem, and is thus
being committed.
Joshua Colp [Wed, 10 Dec 2008 16:05:29 +0000 (16:05 +0000)]
Increment the sequence number on the end packets for RFC2833. After reading the RFC some more and doing some testing I agree with this change.
(closes issue #12983)
Reported by: vt
Patches:
dtmf_inc_seqnum_on_end_pkts.diff uploaded by vt (license 520)
Russell Bryant [Tue, 9 Dec 2008 22:17:39 +0000 (22:17 +0000)]
Remove the test_for_thread_safety() function completely.
The test is not valid. Besides, if we actually suspected that recursive
mutexes were not working, we would get a ton of LOG_ERROR messages when
DEBUG_THREADS is turned on.
(inspired by a discussion on the asterisk-dev list)
Tilghman Lesher [Tue, 9 Dec 2008 21:53:25 +0000 (21:53 +0000)]
We appear to have documented tz= in the [general] section of voicemail.conf,
without actually having implemented it. Oops.
(Reported by Olivier on the -users list)
Joshua Colp [Tue, 9 Dec 2008 21:14:29 +0000 (21:14 +0000)]
Add 'down' as a valid state for directed call pickup. This creeps up when we receive session progress when dialing a device and not ringing.
(closes issue #14005)
Reported by: ddl
Russell Bryant [Tue, 9 Dec 2008 20:57:35 +0000 (20:57 +0000)]
Fix an issue where callers on an incoming call on an SLA trunk would not hear ringback.
We need to make sure that we don't start writing audio to the trunk channel until we're
actually ready to answer it. Otherwise, the channel driver will treat it as inband
progress, even though all they are getting is silence.
Steve Murphy [Tue, 9 Dec 2008 20:20:54 +0000 (20:20 +0000)]
In discussion with seanbright on #asterisk-dev, I have added a default rule, and an option to suppress the default rule from being generated in the flex output, for the sake of those OS's where they didn't tweak flex's ECHO macro, and the compiler doesn't like it. The regressions are OK with this.
Joshua Colp [Tue, 9 Dec 2008 19:47:07 +0000 (19:47 +0000)]
Make sure that the timestamp for DTMF is not the same as the previous voice frame and do not send audio when transmitting DTMF as this confuses some equipment.
(closes issue #13209)
Reported by: ip-rob
Patches:
13209.diff uploaded by file (license 11)
Tested by: ip-rob, bujones
This crash was the result of a few small errors that
would combine in 64-bit land to result in a crash.
32-bit land might have seen these combine to mysteriously
drop the args to an application call, in certain
circumstances.
Also, in trying to find this bug, I spotted
a situation in the flex input, where, in passing
back a 'word' to the parser, it would allocate
a buffer larger than necessary. I changed the
usage in such situations, so that strdup was
not used, but rather, an ast_malloc, followed
by ast_copy_string.
I removed a field from the pval struct, in
u2, that was never getting used, and set in
one spot in the code. I believe it was an
artifact of a previous fix to make switch
cases work invisibly with extens.
And, for goto's I removed a '!' from
before a strcmp, that has been there
since the initial merging of AEL2, that
might prevent the proper target of a
goto from being found. This was pretty
harmless on its own, as it would just
louse up a consistency check for users.
Many thanks to ckjohnsonme for providing
a simplified and complete set of information
about the bug, that helped considerably in
finding and fixing the problem.
Now, to get aelparse up and running again
in trunk, and out of its "horribly broken" state,
so I can run the regression suite!
Russell Bryant [Tue, 9 Dec 2008 14:52:25 +0000 (14:52 +0000)]
Fix a problem with GROUP() settings on a masquerade.
The previous code carried over group settings from the old channel to the new
one. However, it did nothing with the group settings that were already on the
new channel. This patch removes all group settings that already existed on the
new channel.
I have a more complicated version of this patch which addresses only the most
blatant problem with this, which is that a channel can end up with multiple
group settings in the same category. However, I could not think of a use case
for keeping any of the group settings from the old channel, so I went this route
for now.
This mirrors a check that was present in ast_rtp_read to also be in ast_rtp_raw_write to not schedule sending the receiver report if the remote RTCP endpoint address isn't present in the RTCP structure.
Mark Michelson [Thu, 4 Dec 2008 16:44:18 +0000 (16:44 +0000)]
Fix a callerid parsing issue. If someone formatted callerid like the
following: "name <number>" (including the quotation marks), then the parts
would be parsed as
name: "name
number: number
This is because the closing quotation mark was not discovered since the number
and everything after was parsed out of the string earlier. Now, there is a check
to see if the closing quote occurs after the number, so that we can know if we
should strip off the opening quote on the name.
Tilghman Lesher [Wed, 3 Dec 2008 17:55:12 +0000 (17:55 +0000)]
During bridge code, the channel bridge may return a retry code, if a transfer
was initiated but not yet completed. If the bridge is immediately retried,
then we may send a storm of TXREQ packets, even though the first set is sent
reliably (retransmitted). Fixes AST-137.
Tilghman Lesher [Wed, 3 Dec 2008 17:34:34 +0000 (17:34 +0000)]
If an entry is added to the directory during a scan when another entry expires,
then that new entry will not be processed promptly, but must wait for either a
future entry to start or a current entry's retry to occur. If no other entries
exist in the directory (other than the new entries) when a bunch expire, then
the new entries must wait until another new entry is added to be processed.
This was a rather weird race condition, really. Fixes AST-147.
Tilghman Lesher [Wed, 3 Dec 2008 16:58:34 +0000 (16:58 +0000)]
Don't start scanning the directory until all modules are loaded, because some
required modules (channels, apps, functions) may not yet be in memory yet.
Fixes AST-149.
Tilghman Lesher [Wed, 3 Dec 2008 14:09:35 +0000 (14:09 +0000)]
Jon Bonilla (Manwe) pointed out on the -dev list:
"I guess that having only ip-phones in mind is not a good approach. Since it is
possible to have a sip proxy connected to asterisk we could receive a 407
(unauthorized) or 483 (too many hops) as response and dialog ending would not be
a good behavior."
So modified.
projects / thirdparty / asterisk.git / log
