Matthew Jordan [Fri, 18 Jan 2013 23:26:02 +0000 (23:26 +0000)]
Fix astcanary startup problem due to wrong pid value from before daemon call
When Asterisk forks itself into the background via a call to daemon, it must
re-set the pid value of the new process. Otherwise, astcanary gets the pid
value of the process before the fork, which prevents it from running. Asterisk
eventually starts lowering its priority, as it can no longer communicate
with the proverbial canary in the coal mine.
This patch ensures that the correct process identifier is used by astcanary.
(closes issue ASTERISK-20947)
Reported by: Jakob Hirsch
Tested by: mjordan
patches:
asterisk-10.12.0.astcanary_ppid.diff uploaded by Jakob Hirsch (license 6113)
David M. Lee [Fri, 18 Jan 2013 05:23:57 +0000 (05:23 +0000)]
Fix Record-Route parsing for large headers.
Record-Route parsing copied the header into a char[256] array, which can
be a problem if the header is longer than that. This patch parses the
header in place, without the copy, avoiding the issue.
In addition to the original patch, I added a unit test for the new
get_in_brackets_const function.
(closes issue ASTERISK-20837)
Reported by: Corey Farrell
Patches:
chan_sip-build_route-optimized-rev1.patch uploaded by Corey Farrell (license 5909)
(with minor changes by dlee)
Matthew Jordan [Thu, 17 Jan 2013 02:28:31 +0000 (02:28 +0000)]
Fix issue where chan_mobile fails to bind to first available port
Per the bluez API, in order to bind to the first available port, the rc_channel
field of the socket addressing structure used to bind the socket should be set
to 0. Previously, Asterisk had set the rc_channel field set to 1, causing it
to connect to whatever happens to be on port 1.
We could probably not explicitly set rc_channel to 0 since we memset the struct
earlier, but explicitly setting it will hopefully prevent someone from coming
in and setting it to some explicit port in the future.
(closes issue ASTERISK-16357)
Reported by: challado
Tested by: Alexander Heinz, Nikolay Ilduganov, benjamin, eliafino, David van Geyn
patches:
ASTERISK-16357.diff uploaded by Nikolay Ilduganov (license 6253)
Matthew Jordan [Wed, 16 Jan 2013 04:10:15 +0000 (04:10 +0000)]
Fix parsing SMSSRC for SMS messages
The parser for SMS messages would incorrectly parse out the from number.
The parsing would incorrectly start scanning for the from number at the
same index as the first double quote ("); this would inadvertently cause
it to treat the first double quote as the terminating double quote for
the from number as well.
The SMSSRC should now populate correctly.
(closes issue ASTERISK-16822)
Reported by: menschentier
Tested by: Jonas Falck
patches:
fixSMSSRC.patch uploaded by jonax (license 6320)
Matthew Jordan [Wed, 16 Jan 2013 00:10:37 +0000 (00:10 +0000)]
Set the INVALID_EXTEN channel variable when chan_misdn forces the 'i' extension
The chan_misdn channel driver will send a channel with an invalid destination
to the 'i' extension itself if said extension can be reached. It forgot,
however, to set the INVALID_EXTEN channel variable when it bounces the channel
to this extension. Dialplan writers everywhere moaned at yet another
inconsistency.
This is yet another example of why duplicating logic in multiple places results
in bugs that stick around in Jira for just under three years.
Yes: ASTERISK-15456 was created on January 18th, 2010. Patch committed on
January 15th, 2013. Ouch.
(closes issue ASTERISK-15456)
Reported by: Thomas Omerzu
patches:
chan_misdn_invalid.patch2 uploaded by Thomas Omerzu (license 5927)
Matthew Jordan [Tue, 15 Jan 2013 03:46:57 +0000 (03:46 +0000)]
Prevent crash in ConfBridge due to race condition when channels leave bridge
When a channel leaves a bridge, a race condition existed where the
bridge_channel's pvt structure would be accessed after it was disposed of.
This patch prevents that by setting the pointer to the pvt to NULL prior
to disposing of it.
Note that this patch is a backport from Asterisk 10. This particular race
condition was fixed as part of the larger code rework that occurred for that
release.
The solution to this problem was pointed out by Gunnar Harms in ASTERISK-16640.
David M. Lee [Mon, 14 Jan 2013 15:11:39 +0000 (15:11 +0000)]
Fix XML encoding of 'identity display' in NOTIFY messages, continued.
When r378933 was merged into 1.8, it should have also escaped
remote_display, since it will have the same XML encoding problem when
the caller/callee roles are reversed.
Matthew Jordan [Sun, 13 Jan 2013 21:15:06 +0000 (21:15 +0000)]
Reset RTP timestamp; sequence number on SSRC change
In r370252 for ASTERISK-18404, Asterisk's handling of RTP was modified to
better account for out of order RTP packets. This was accomplished by using the
RTP timestamp and sequence number to check for out of order packets. However,
when a SSRC change occurs, the timestamp and sequence number will no longer
have any relation to the previously received packets. The variables tracking
the timestamp and sequence number therefore have to be reset.
David M. Lee [Sat, 12 Jan 2013 06:26:34 +0000 (06:26 +0000)]
Fix XML encoding of 'identity display' in NOTIFY messages.
XML encoding in chan_sip is accomplished by naively building the XML
directly from strings. While this usually works, it fails to take into
account escaping the reserved characters in XML.
This patch adds an 'ast_xml_escape' function, which works similarly to
'ast_uri_encode'. This is used to properly escape the local_display
attribute in XML formatted NOTIFY messages.
Several things to note:
* The Right Thing(TM) to do would probably be to replace the
ast_build_string stuff with building an ast_xml_doc. That's a much
bigger change, and out of scope for the original ticket, so I
refrained myself.
* It is with great sadness that I wrote my own ast_xml_escape
function. There's one in libxml2, but it's knee-deep in
libxml2-ness, and not easily used to one-off escape a
string.
* I only escaped the string we know is causing problems
(local_display). At least some of the other strings are
URI-encoded, which should be XML safe. Rather than figuring out
what's safe and escaping what's not, it would be much cleaner to
simply build an ast_xml_doc for the messages and let the XML
library do the XML escaping. Like I said, that's out of scope.
Jonathan Rose [Fri, 4 Jan 2013 22:54:31 +0000 (22:54 +0000)]
res_srtp: Prevent a crash from occurring due to srtp_create failures in srtp_create
Under some circumstances, libsrtp's srtp_create function deallocates memory that
it wasn't initially responsible for allocating. Because we weren't initially
aware of this behavior, this memory was still used in spite of being unallocated
during the course of the srtp_unprotect function. A while back I made a patch
which would set this value to NULL, but that exposed a possible condition where
we would then try to check a member of the struct which would cause a segfault.
In order to address these problems, ast_srtp_unprotect will now set an error value
when it ends without a valid SRTP session which will result in the caller of
srtp_unprotect observing this error and hanging up the relevant channel instead of
trying to keep using the invalid session address.
Fix SIP Notify Messages To Have The Proper IP Address In The FROM Field
On a multihomed server when sending a NOTIFY message, we were not figuring out
which network should be used to contact the peer.
This patch fixes the problem by calling ast_sip_ouraddrfor() and then
build_via() so that our NOTIFY message contains the correct IP address.
Also, a debug message is being added to help follow the call-id changes that
occur. This was helpful for confirming that the IP address was set properly
since the call-id contains the IP address. It also will be helpful for
troubleshooting purposes when following a call in the debug logs.
(closes issue ASTERISK-20805)
Reported by: Bryan Hunt
Tested by: Bryan Hunt, Michael L. Young
Patches:
asterisk-20805-notify-ip-v2.diff uploaded by Michael L. Young (license 5026)
Fix Queue Log Reporting Every Call COMPLETECALLER With "h" Extension Present
When the "h" extension is present within the context of the queue, all calls
are being reported COMPLETECALLER even when the agent is hanging up the call.
This patch checks to see if the agent hung-up or not instead of only relying on
checking if the queue (caller) channel hung-up or not. It would appear that
having the h extension in the mix, the pbx goes to the h extension,
"hanging-up" the queue channel and triggering the reporting of COMPLETECALLER.
(closes issue ASTERISK-20743)
Reported by: call
Tested by: call, Michael L. Young
Patches:
asterisk-20743-q-cmplt-caller.diff
uploaded by Michael L. Young (license 5026)
Richard Mudgett [Thu, 3 Jan 2013 19:40:55 +0000 (19:40 +0000)]
chan_agent: Fix wrapup time wait response.
* Made agent_cont_sleep() and agent_ack_sleep() stop waiting if the wrapup
time expires. agent_cont_sleep() had tried but returned the wrong value
to stop waiting.
* Made agent_ack_sleep() take a struct agent_pvt pointer instead of a void
pointer for better type safety.
Kinsey Moore [Thu, 3 Jan 2013 18:35:38 +0000 (18:35 +0000)]
Add missing test event
This test event was missing from channel.c causing the dial_LS_options
test to fail intermittently because of a race condition where most code
paths emitted the test event but this one did not. The dial_LS_options
test should stop bouncing now.
Matthew Jordan [Wed, 2 Jan 2013 21:48:57 +0000 (21:48 +0000)]
Prevent crashes from occurring when reading from data sources with large values
When reading configuration data from an Asterisk .conf file or when pulling
data from an Asterisk RealTime backend, Asterisk was copying the data on the
stack for manipulation. Unfortunately, it is possible to read configuration
data or realtime data from some data source that provides a large blob of
characters. This could potentially cause a crash via a stack overflow.
This patch prevents large sets of data from being read from an ARA backend or
from an Asterisk conf file.
Richard Mudgett [Wed, 2 Jan 2013 21:08:15 +0000 (21:08 +0000)]
Fix AMI redirect action with two channels failing to redirect both channels.
The AMI redirect action can fail to redirect two channels that are bridged
together. There is a race between the AMI thread redirecting the two
channels and the bridge thread noticing that a channel is hungup from the
redirects.
* Made the bridge wait for both channels to be redirected before exiting.
* Made the AMI redirect check that all required headers are present before
proceeding with the redirection.
* Made the AMI redirect require that any supplied ExtraChannel exist
before proceeding. Previously the code fell back to a single channel
redirect operation.
(closes issue ASTERISK-18975)
Reported by: Ben Klang
Matthew Jordan [Wed, 2 Jan 2013 16:54:20 +0000 (16:54 +0000)]
Prevent exhaustion of system resources through exploitation of event cache
Asterisk maintains an internal cache for devices in the event subsystem. The
device state cache holds the state of each device known to Asterisk, such that
consumers of device state information can query for the last known state for
a particular device, even if it is not part of an active call. The concept of
a device in Asterisk can include entities that do not have a physical
representation. One way that this occurred was when anonymous calls are allowed
in Asterisk. A device was automatically created and stored in the cache for
each anonymous call that occurred; this was possible in the SIP and IAX2
channel drivers and through channel drivers that utilized the
res_jabber/res_xmpp resource modules (Gtalk, Jingle, and Motif). These devices
are never removed from the system, allowing anonymous calls to potentially
exhaust a system's resources.
This patch changes the event cache subsystem and device state management to
no longer cache devices that are not associated with a physical entity.
(issue ASTERISK-20175)
Reported by: Russell Bryant, Leif Madsen, Joshua Colp
Tested by: kmoore
patches:
event-cachability-3.diff uploaded by jcolp (license 5000)
Matthew Jordan [Wed, 2 Jan 2013 15:16:10 +0000 (15:16 +0000)]
Resolve crashes due to large stack allocations when using TCP
Asterisk had several places where messages received over various network
transports may be copied in a single stack allocation. In the case of TCP,
since multiple packets in a stream may be concatenated together, this can
lead to large allocations that overflow the stack.
This patch modifies those portions of Asterisk using TCP to either
favor heap allocations or use an upper bound to ensure that the stack will not
overflow:
* For SIP, the allocation now has an upper limit
* For HTTP, the allocation is now a heap allocation instead of a stack
allocation
* For XMPP (in res_jabber), the allocation has been eliminated since it was
unnecesary.
Note that the HTTP portion of this issue was independently found by Brandon
Edwards of Exodus Intelligence.
(issue ASTERISK-20658)
Reported by: wdoekes, Brandon Edwards
Tested by: mmichelson, wdoekes
patches:
ASTERISK-20658_res_jabber.c.patch uploaded by mmichelson (license 5049)
issueA20658_http_postvars_use_malloc2.patch uploaded by wdoekes (license 5674)
issueA20658_limit_sip_packet_size3.patch uploaded by wdoekes (license 5674)
Kinsey Moore [Mon, 31 Dec 2012 14:41:37 +0000 (14:41 +0000)]
Ensure chan_sip rejects encrypted streams without crypto info
This ensures that Asterisk rejects encrypted media streams (RTP/SAVP
audio and video) that are missing cryptographic keys and ensures that
the incoming SDP is consistent with RFC4568 as far as having a crypto
attribute present for any SAVP streams.
Mark Michelson [Thu, 20 Dec 2012 20:26:47 +0000 (20:26 +0000)]
Adjust RTP instance's available_formats callback to return the correct type.
The RTP engine public function that gets the available formats expects a
format_t to be returned; however when calling into an RTP instance's
callback to get the available formats, the callback returned an int.
This never was noticed in Asterisk because the two RTP engines included
do not provide an available_formats callback.
This introduces an API change, and the proposal for this change was brought
up on the Asterisk developers mailing list [1]. There was no public objection
to this change, so it is now being put in.
Richard Mudgett [Fri, 14 Dec 2012 21:23:15 +0000 (21:23 +0000)]
app_queue: Revert bad ringinuse=no patch.
With the option ringinuse=no set, the patch committed for ASTERISK-16115
causes non-SIP queue members to never be called because the device state
is checked after a channel is created to determine if the member is busy.
These queue members always get the "Member %s is busy, cannot dial"
message.
Most channel drivers other than chan_sip use the default device state
handling. The default device-state state is considered in use or unknown
if the channel exists or not respectively.
Kinsey Moore [Thu, 13 Dec 2012 13:43:41 +0000 (13:43 +0000)]
Ensure Min-SE is included in outbound INVITEs
Asterisk now includes Min-SE in outbound INVITEs when the value is not
90 (the default) and session timers are not disabled. This has the
effect of Asterisk following RFC4028 more closely with regard to 422
responses and preventing situations in which Asterisk would be forced
to temporarily accept a call to tear it down based on a Session-Expires
below the locally configured Min-SE.
(issue SWP-5051)
Review: https://reviewboard.asterisk.org/r/2222/ Reported-by: Kinsey Moore Patch-by: Kinsey Moore
git-svn-id: https://origsvn.digium.com/svn/asterisk/branches/1.8@377946 65c4cc65-6c06-0410-ace0-fbb531ad65f3
Mark Michelson [Tue, 11 Dec 2012 20:45:17 +0000 (20:45 +0000)]
Fix crash that can occur if CLI registration fails for an aliased command.
A recent memory leak fix in main/cli.c causes an ast_cli_entry's command
field to be freed and NULLed if ast_cli_register() fails. res_clialiases
was ignoring the return value of ast_cli_register() and was then passing
the NULL command off to a a hash function. This resulted in a crash.
The fix is not to ignore the erroneous return value. If ast_cli_register()
fails, then we do not continue trying to process the current alias.
Richard Mudgett [Tue, 11 Dec 2012 02:09:33 +0000 (02:09 +0000)]
Cleanup indications on exit.
* Made ast_unregister_indication_country() unlink the found tone zone
before selecting a new default_tone_zone to make it impossible to select
the tone zone being unregistered again.
* Ringcadence is no longer parsed twice in store_config_tone_zone().
* Cleanup CLI commands and destroy default_tone_zone on exit.
Kinsey Moore [Mon, 10 Dec 2012 16:51:35 +0000 (16:51 +0000)]
Ensure ReceiveFax provides a CED tone via T.38
When using res_fax_digium, the T.38 CED tone was not being provided
properly which would cause some incoming faxes to fail. This was not an
issue with res_fax_spandsp since it does not strictly honor the
send_ced flag and sends the CED tone whenever receiving a T.38 fax.
(closes issue FAX-343) Reported-by: Benjamin Tietz Patch-by: Kinsey Moore
git-svn-id: https://origsvn.digium.com/svn/asterisk/branches/1.8@377655 65c4cc65-6c06-0410-ace0-fbb531ad65f3
Kinsey Moore [Mon, 10 Dec 2012 14:39:07 +0000 (14:39 +0000)]
Handle Session-Expires less than local Min-SE in 200 OK
Ensure that a call is immediately torn down if a Session-Expires value
received in a 200 OK is less than the local Min-SE. This also prevents
Asterisk from allowing calls with Session-Expires below the
RFC4028-mandated minimum (90s).
Fix code to send in both rx and tx open stream messages correct codecs. Found that on phase 0/1 phones wrong codecs cause to no audio in some situations.
Tilghman Lesher [Mon, 10 Dec 2012 01:38:52 +0000 (01:38 +0000)]
Improve documentation by making all of the colors used readable,
no matter what the background color is.
Dark blue on a black background is unreadable, as is yellow on a
light background. This patch turns on the bright attribute for
colors when on a dark background and turns *off* the bright
attribute when the -W command line option is used (indicating a
_light_ background). This ensures that text is readable in both
cases.
Richard Mudgett [Sat, 8 Dec 2012 00:28:51 +0000 (00:28 +0000)]
Fix order of SIP allow/disallow in MySQL contrib script.
Using the contrib sippeers.sql script to create the sippeers MySQL table
would result in being unable to place calls if you set the disallow value
to all.
(closes issue ASTERISK-20756)
Reported by: Andre Luis
Patches:
sippeers.patch patch uploaded by Andre Luis
Joshua Colp [Wed, 5 Dec 2012 16:48:01 +0000 (16:48 +0000)]
Fix a SIP request memory leak with TLS connections.
During the TLS re-work in chan_sip some TLS specific code was moved
into a separate function. This function operates on a copy of the
incoming SIP request. This copy was never deinitialized causing a
memory leak for each request processed.
This function is now given a SIP request structure which it can use
to copy the incoming request into. This reduces the amount of memory
allocations done since the internal allocated components are reused
between packets and also ensures the SIP request structure is
deinitialized when the TLS connection is torn down.
Jonathan Rose [Wed, 5 Dec 2012 16:46:28 +0000 (16:46 +0000)]
res_srtp: Fix a crash caused by srtp_dealloc on an already dealloced session
When srtp_create fails, the session may be dealloced or just not alloced. At
the same time though, the session pointer might not be set to NULL in this
process and attempting to srtp_dealloc it again will cause a segfault. This
patch checks for failure of srtp_create and sets the session pointer to NULL
if it fails.
Richard Mudgett [Mon, 3 Dec 2012 22:51:41 +0000 (22:51 +0000)]
Cleanup ast_run_atexits() atexits list.
* Convert atexits list to a mutex instead of a rd/wr lock. The lock is
only write locked.
* Move CLI verbose Asterisk ending message to where AMI message is output
in really_quit() to avoid further surprises about using stuff already
shutdown.
Mark Michelson [Fri, 30 Nov 2012 16:12:07 +0000 (16:12 +0000)]
Fix potential crashes during SIP attended transfers.
The principal behind this patch is simple. During a transfer,
we manipulate channels that are owned by a separate thread than
the one we currently are running in, so it makes sense that we
need to grab a reference to the channels so that they cannot
disappear out from under us.
In the wild, crashes were sometimes seen when the transferring
party would hang up the call before the transfer target answered
the call. The most common place to see the crash occur was when
attempting to send a connected line update to the transferer
channel.
(closes issue ASTERISK-20226)
Reported by Jared Smith
Patches:
ASTERISK-20226.patch uploaded by Mark Michelson (License #5049)
Tested by: Jared Smith
Michael L. Young [Thu, 29 Nov 2012 21:49:40 +0000 (21:49 +0000)]
Improve Code Readability And Fix Setting natdetected Flag
For 1.8, 10, 11 and trunk we are are improving the code readability.
For 11 and trunk, auto nat detection was added. The natdetected flag was being
set to 1 when the host address in the VIA header did not specifiy a port. This
patch fixes this by setting the port on the temporary sock address used to
SIP_STANDARD_PORT in order for the sock address comparison to work properly.
(closes issue ASTERISK-20724)
Reported by: Michael L. Young
Patches:
asterisk-20724-set-port-v2.diff uploaded by Michael L. Young (license 5026)
* Adds the following CLI commands to control MALLOC_DEBUG reporting of
unreleased malloc memory when Asterisk is shut down.
memory atexit list on
memory atexit list off
memory atexit summary byline
memory atexit summary byfunc
memory atexit summary byfile
memory atexit summary off
* Made check all remaining allocated region blocks atexit for fence
violations.
* Increased the allocated region hash table size by about three times. It
still isn't large enough considering the number of malloced blocks
Asterisk uses.
* Made CLI "memory show allocations anomalies" use
regions_check_all_fences().
Jonathan Rose [Wed, 28 Nov 2012 16:23:47 +0000 (16:23 +0000)]
manager: Make challenge work with allowmultiplelogin=no
Prior to this patch, challenge would yield a multiple logins error if used
without providing the username (which isn't really supposed to be an argument
to challenge) if allowmultiplelogin was set to no because allowmultiplelogin
finds a user with a zero length login name. This check is simply disabled for
the challenge action when the username is empty by this patch.
(closes issue ASTERISK-20677)
Reported by: Vladimir
Patches:
challenge_action_nomultiplelogin.diff uploaded by Jonathan Rose (license 6182)
Richard Mudgett [Tue, 27 Nov 2012 23:47:27 +0000 (23:47 +0000)]
Fix extension matching with the '-' char.
The '-' char is supposed to be ignored by the dialplan extension matching.
Unfortunately, it's treatment is not handled consistently throughout the
extension matching code.
* Made the old exten matching code consistently ignore '-' chars.
* Made the old exten matching code consistently handle case in the
matching.
* Made ignore empty character sets.
* Fixed ast_extension_cmp() to return -1, 0, or 1 as documented. The only
user of it in pbx_lua.c was testing for -1. It was originally returning
the strcmp() value for less than which is not usually going to be -1.
* Fix character set sorting if the sets have the same number of characters
and start with the same character. Character set [0-9] now sorts before
[02-9a] as originally intended.
* Updated some extension label and priority already in use warnings to
also indicate if the extension is aliased.
Richard Mudgett [Tue, 27 Nov 2012 20:32:31 +0000 (20:32 +0000)]
Remove unnecessary channel module references.
* Removed call to ast_module_user_hangup_all() in res_config_mysql.c since
it is effectively a noop. No channels can attach a reference to that
module.
* Removed call to ast_module_user_hangup_all() in app_celgenuserevent.c.
The caller of unload_module() has already called it.
* Removed redundant channel module references in pbx_dundi.c. The
registered dialplan function callback dispatchers for the read/read2/write
callbacks already reference the module before calling.
* pbx_dundi: Moved unregistering CLI commands, DUNDi switch, and dialplan
functions to the first thing the unload_module() does. This will reduce
the chance of new channels using DUNDi services while the module is being
torn down.
Matthew Jordan [Thu, 22 Nov 2012 23:51:02 +0000 (23:51 +0000)]
Re-initialize logmsgs mutex upon logger initialization to prevent lock errors
Similar to the patch that moved the fork earlier in the startup sequence to
prevent mutex errors in the recursive mutex surrounding the read/write thread
registration lock, this patch re-initializes the logmsgs mutex. Part of the
start up sequence before forking the process into the background includes
reading asterisk.conf; this has to occur prior to the call to daemon in order
to read startup parameters. When reading in a conf file, log statements can
be generated. Since this can't be avoided, the mutex instead is
re-initialized to ensure a reset of any thread tracking information.
This patch also includes some additional debugging to catch errors when
locking or unlocking the recursive mutex that surrounds locks when the
DEBUG_THREADS build option is enabled. DO_CRASH or THREAD_CRASH will
cause an abort() if a mutex error is detected.
Matthew Jordan [Sun, 18 Nov 2012 20:11:18 +0000 (20:11 +0000)]
Reorder startup sequence to prevent lockups when process is sent to background
Although it is very rare and timing dependent, the potential exists for the
call to 'daemon' to cause what appears to be a deadlock in Asterisk during
startup. This can occur when a recursive mutex is obtained prior to the
daemon call executing. Since daemon uses fork to send the process into the
background, any threading primitives are unsafe to re-use after the call.
Implementations of pthread recursive mutexes are highly likely to store the
thread identifier of the thread that previously obtained the mutex. If
the mutex was locked prior to the fork, a subsequent unlock operation will
potentially fail as the thread identifier is no longer valid. Since the
mutex is still locked, all subsequent attempts to grab the mutex by other
threads will block.
This behavior exhibited itself most often when DEBUG_THREADS was enabled, as
this compile time option surrounds the mutexes in Asterisk with another
recursive mutex that protects the storage of thread related information. This
made it much more likely that a recursive mutex would be obtained prior to
daemon and unlocked after the call.
This patch does the following:
a) It backports a patch from Asterisk 11 that prevents the spawning of the
localtime monitoring thread. This thread is now spawned after Asterisk has
fully booted.
b) It re-orders the startup sequence to call daemon earlier during Asterisk
startup. This limits the potential of threading primitives being accessed
by initialization calls before daemon is called.
c) It removes calls to ast_verbose/ast_log/etc. prior to daemon being called.
Developers should send error messages directly to stderr prior to daemon,
as calls to ast_log may access recursive mutexes that store thread related
information.
d) It reorganizes when thread local storage is created for storing lock
information during the creation of threads. Prior to this patch, the
read/write lock protecting the list of threads in ast_register_thread would
utilize the lock in the thread local storage prior to it being initialized;
this patch prevents that.
On a very related note, this patch will *greatly* improve the stability of the
Asterisk Test Suite.
Jonathan Rose [Fri, 16 Nov 2012 19:31:48 +0000 (19:31 +0000)]
monitor: prevent attempts to move/remove recordings skipped with 'i' and 'o'.
The i and o options for monitor skip the input and output sides of a recording
respectively. This patch addresses a problem in those options when monitor is
called without specifying a specific filename where monitor will try to move
the recording that was skipped. Since this usually doesn't exist when these
options are used, it would produce a warning when it does this in most cases,
but it is conceivable that there are use cases where this could result in
moving/removing a file unintentionally.
(closes issue ASTERISK-20641)
Reported by: Jonathan Rose
Review: https://reviewboard.asterisk.org/r/2190/
Jonathan Rose [Thu, 15 Nov 2012 22:40:50 +0000 (22:40 +0000)]
app_meetme: Fix channels lingering when hung up under certain conditions
Channels would get stuck and MeetMe would repeatedly display an Unable
to write frame to channel error in the conf_run function if hung up
during certain sound prompts such as during user count announcements.
This patch fixes that by reintroducing a hangup check in the meetme's
main loop (also in conf_run).
(closes issue ASTERISK-20486)
Reported by: Michael Cargile
Review: https://reviewboard.asterisk.org/r/2187/
Patches:
meetme_hangup_patch_ASTERISK-20486_v3.diff uploaded by Jonathan Rose (license 6182)
David M. Lee [Thu, 15 Nov 2012 22:27:37 +0000 (22:27 +0000)]
Migrate hashtest/hashtest2 to be unit tests.
Both hashtest and hashtest2 are manual testing apps that thrash hash
tables (hashtab and ao2 containers, respectively), by spinning up
several threads that randomly insert, delete, lookup and iterate over
the hash table. If the app doesn't crash, the hash table probably passes
the test. Those utils are not a part of the typical Asterisk build, so
they do not usually get compiled. This all makes them less that useful.
This patch removes those manual test programs and replaces them with
Asterisk unit test modules (test_{hashtab,astobj2}_thrash.so). It also
attempts to make the tests more deterministic.
* Rather than spinning up some number of threads that operate on the
hash table randomly, spin up four threads that concurrenly add,
remove, lookup and iterate over the hash table.
* Each thread checks the state of the hash table both during and after
execution, and indicates a test failure if things are not as expected.
* Each thread times out after 60 seconds to prevent deadlocking the unit
test run.
(closes issue ASTERISK-20505)
Reported by: Matt Jordan
Review: https://reviewboard.asterisk.org/r/2189/
Rusty Newton [Thu, 15 Nov 2012 01:43:43 +0000 (01:43 +0000)]
Patch to play correct sound file when a voicemail's urgent status is removed
We were attempting to play "vm-urgent-removed", which didn't exist. Now we play "vm-marked-nonurgent" which exists
and is the correct sound file. Previous behavior was silence and a warning on the CLI.
(issue ASTERISK-20280)
(closes issue ASTERISK-20280)
Reported by: Tomo Takebe
Tested by: Rusty Newton
Patches:
asterisk20280.patch uploaded by Rusty Newton (license 5829)
Richard Mudgett [Wed, 14 Nov 2012 19:48:13 +0000 (19:48 +0000)]
Fix call files when astspooldir is relative.
Future dated call files are ignored when astspooldir is relative to the
current directory. The queue_file() assumed that the qdir needed to be
prepended if the given filename did not start with a '/'. If astspooldir
is relative it is not going to start from the root directory obviously so
it will not start with a '/'. The filename used in queue_file()
ultimately results in qdir prepended multiple times.
* Made queue_file() not prepend qdir if the filename contains a '/'.
(closes issue ASTERISK-20593)
Reported by: James Le Cuirot
Patches:
0004-Fix-future-call-files-from-relative-directories.patch (license #6439) patch uploaded by James Le Cuirot
Brent Eagles [Tue, 13 Nov 2012 18:10:22 +0000 (18:10 +0000)]
Patch to prevent stopping the active generator when it is not the silence
generator.
This patch introduces an internal helper function to safely check whether the
current generator is the one that is expected before deactivating it. The
current externally accessible ast_channel_stop_generator() function has been
modified to be implemented in terms of the new function.
(closes issue ASTERISK-19918)
Reported by: Eduardo Abad
Joshua Colp [Mon, 12 Nov 2012 20:44:00 +0000 (20:44 +0000)]
Properly check if the "Context" and "Extension" headers are empty in a ShowDialPlan action.
The code which handles the ShowDialPlan action wrongly assumed that a non-NULL return value
from the function which retrieves headers from an action indicates that the header has a
value. This is incorrect and the contents must be checked to see if they are blank.
Michael L. Young [Mon, 12 Nov 2012 20:13:53 +0000 (20:13 +0000)]
Fix Dynamic Hints Variable Substition - Underscore Problem
When adding a dynamic hint, if an extension contains an underscore no variable
subsitution is being performed.
This patch changes from checking if the extension contains an underscore to
checking if the extension begins with an underscore.
(closes issue ASTERISK-20639)
Reported by: Steven T. Wheeler
Tested by: Steven T. Wheeler, Michael L. Young
Patches:
asterisk-20639-dynamic-hint-underscore.diff
uploaded by Michael L. Young (license 5026)
Richard Mudgett [Thu, 8 Nov 2012 21:05:40 +0000 (21:05 +0000)]
chan_dahdi/SS7: Made reject incoming call for an in-alarm or blocked channel.
If a SS7 call comes in requesting a CIC that is in-alarm, the call is
accepted and connects if the extension exists in the dialplan. The call
does not have any audio.
* Made release the call immediately with circuit congestion cause.
(closes issue ASTERISK-20204)
Reported by: Tuan Le
Patches:
jira_asterisk_20204_v1.8.patch (license #5621) patch uploaded by rmudgett
Richard Mudgett [Thu, 8 Nov 2012 16:58:52 +0000 (16:58 +0000)]
Add MALLOC_DEBUG enhancements.
* Makes malloc() behave like calloc(). It will return a memory block
filled with 0x55. A nonzero value.
* Makes free() fill the released memory block and boundary fence's with
0xdeaddead. Any pointer use after free is going to have a pointer
pointing to 0xdeaddead. The 0xdeaddead pointer is usually an invalid
memory address so a crash is expected.
* Puts the freed memory block into a circular array so it is not reused
immediately.
* When the circular array rotates out a memory block to the heap it checks
that the memory has not been altered from 0xdeaddead.
* Made the astmm_log message wording better.
* Made crash if the DO_CRASH menuselect option is enabled and something is
found.
* Fixed a potential alignment issue on 64 bit systems.
struct ast_region.data[] should now be aligned correctly for all
platforms.
* Extracted region_check_fences() from __ast_free_region() and
handle_memory_show().
Mark Michelson [Wed, 7 Nov 2012 17:01:13 +0000 (17:01 +0000)]
Fix misuses of timeouts throughout the code.
Prior to this change, a common method for determining if a timeout
was reached was to call a function such as ast_waitfor_n() and inspect
the out parameter that told how many milliseconds were left, then use
that as the input to ast_waitfor_n() on the next go-around.
The problem with this is that in some cases, submillisecond timeouts
can occur, resulting in the out parameter not decreasing any. When this
happens thousands of times, the result is that the timeout takes much
longer than intended to be reached. As an example, I had a situation where
a 3 second timeout took multiple days to finally end since most wakeups
from ast_waitfor_n() were under a millisecond.
This patch seeks to fix this pattern throughout the code. Now we log the
time when an operation began and find the difference in wall clock time
between now and when the event started. This means that sub-millisecond timeouts
now cannot play havoc when trying to determine if something has timed out.
Part of this fix also includes changing the function ast_waitfor() so that it
is possible for it to return less than zero when a negative timeout is given
to it. This makes it actually possible to detect errors in ast_waitfor() when
there is no timeout.
(closes issue ASTERISK-20414)
reported by David M. Lee
Matthew Jordan [Mon, 5 Nov 2012 22:50:21 +0000 (22:50 +0000)]
Refactor ast_timer_ack to return an error and handle the error in timer users
Currently, if an acknowledgement of a timer fails Asterisk will not realize
that a serious error occurred and will continue attempting to use the timer's
file descriptor. This can lead to situations where errors stream to the
CLI/log file. This consumes significant resources, masks the actual problem
that occurred (whatever caused the timer to fail in the first place), and
can leave channels in odd states.
This patch propagates the errors in the timing resource modules up through
the timer core, and makes users of these timers handle acknowledgement
failures. It also adds some defensive coding around the use of timers
to prevent using bad file descriptors in off nominal code paths.
Note that the patch created by the issue reporter was modified slightly for
this commit and backported to 1.8, as it was originally written for
Asterisk 10.
Matthew Jordan [Sun, 4 Nov 2012 03:06:49 +0000 (03:06 +0000)]
Don't attempt to purge sessions when no sessions exist
Manager's tcp/tls objects have a periodic function that purge old manager
sessions periodically. During shutdown, the underlying container holding
those sessions can be disposed of and set to NULL before the tcp/tls periodic
function is stopped. If the periodic function fires, it will attempt to
iterate over a NULL container.
This patch checks for whether or not the sessions container exists before
attempting to purge sessions out of it. If the sessions container is NULL,
we simply return.
Note that this error was also caught by the Asterisk Test Suite.
Matthew Jordan [Sun, 4 Nov 2012 02:25:48 +0000 (02:25 +0000)]
Properly clean up manager resources on exit
This patch does two things:
1) It properly unregisters the manager CLI commands
2) It cleans up AMI users on exit. Prior to this patch, the AMI users
were not being disposed of properly, resulting in a memory leak.
Matthew Jordan [Sun, 4 Nov 2012 00:54:19 +0000 (00:54 +0000)]
Fix memory leak when unloading XML documentation
This patch is a modified version of a patch originally committed for the
Asterisk 11 branch in r375756. A portion of that patch, that fixed the
memory leak during unloading XML documentation, applies to branches 1.8
and 10 as well.
The patch for this issue was modified for these two branches.
projects / thirdparty / asterisk.git / log
