32 Commits (d60bcc56a7dbba38f5a3a72a8b2455f8de7cfdcd)

Author	SHA1	Message	Date
Joshua Colp	0231dd6ae7	stasis: Improve topic/subscription names and statistics. ... Topic names now follow: <subsystem>:<functionality>[/<object>] This ensures that they are all unique, and also provides better insight in to what each topic is for. Subscriber ids now also use the main topic name they are subscribed to and an incrementing integer as their identifier to make it easier to understand what the subscription is primarily responsible for. Both the CLI commands for listing topic and subscription statistics now sort to make it a bit easier to see what is going on. Subscriptions will now show all topics that they are receiving messages from, not just the main topic they were subscribed to. ASTERISK-28335 Change-Id: I484e971a38c3640f2bd156282e532eed84bf220d	6 years ago
Joshua Colp	50ac85cb40	stasis: Segment channel snapshot to reduce creation cost. ... When a channel snapshot was created it used to be done from scratch, copying all data (many strings). This incurs a cost when doing so. This change segments the channel snapshot into different components which can be reused if unchanged from the previous snapshot creation, reducing the cost. In normal cases this results in some pointers being copied with reference count being bumped, some integers being set, and a string or two copied. The other benefit is that it is now possible to determine if a channel snapshot update is redundant and thus stop it before a message is published to stasis. The specific segments in the channel snapshot were split up based on whether they are changed together, how often they are changed, and their general grouping. In practice only 1 (or 0) of the segments actually get changed in normal operation. Invalidation is done by setting a flag on the channel when the segment source is changed, forcing creation of a new segment when the channel snapshot is created. ASTERISK-28119 Change-Id: I5d7ef3df963a88ac47bc187d73c5225c315f8423	7 years ago
Joshua Colp	d0ccbb3377	stasis: Use an implementation specific channel snapshot cache. ... Channels no longer use the Stasis cache for channel snapshots. Instead they are stored in a hash table in stasis_channels which reduces the number of Stasis messages created and allows better storage. As a result the following APIs are no longer available since the stasis cache is no longer used: ast_channel_topic_cached() ast_channel_topic_all_cached() The ast_channel_cache_all() and ast_channel_cache_by_name() functions now return an ao2_container of ast_channel_snapshots rather than a container of stasis_messages therefore you can't (and don't need to) call stasis_cache functions on it. The ast_channel_topic_all() function now returns a normal topic not a cached one so you can't use stasis cache functions on it either. The ast_channel_snapshot_type() stasis message now has the ast_channel_snapshot_update structure as it's data. It contains the last snapshot and the new one. ast_channel_snapshot_get_latest() still returns the latest snapshot. The latest snapshot is now stored on the channel itself to eliminate cache hits when Stasis messages that have the snapshot as a payload are created. ASTERISK-28102 Change-Id: I9334febff60a82d7c39703e49059fa3a68825786	7 years ago
Corey Farrell	021ce938ca	astobj2: Remove legacy ao2_container_alloc routine. ... Replace usage of ao2_container_alloc with ao2_container_alloc_hash or ao2_container_alloc_list. Remove ao2_container_alloc macro. Change-Id: I0907d78bc66efc775672df37c8faad00f2f6c088	7 years ago
Joshua Colp	3077ad0c24	stasis: Add internal filtering of messages. ... This change adds the ability for subscriptions to indicate which message types they are interested in accepting. By doing so the filtering is done before being dispatched to the subscriber, reducing the amount of work that has to be done. This is optional and if a subscriber does not add message types they wish to accept and set the subscription to selective filtering the previous behavior is preserved and they receive all messages. There is also the ability to explicitly force the reception of all messages for cases such as AMI or ARI where a large number of messages are expected that are then generically converted into a different format. ASTERISK-28103 Change-Id: I99bee23895baa0a117985d51683f7963b77aa190	7 years ago
Corey Farrell	bc73337e07	core: Use macros to generate ao2_container callbacks where possible. ... This uses AO2_STRING_FIELD_HASH_FN and AO2_STRING_FIELD_CMP_FN where possible in the Asterisk core. This removes CMP_STOP from the result of CMP_FN callbacks for the following structure types: * ast_bucket_metadata * ast_bucket_scheme * generic_monitor_instance_list (ccss.c) * ast_bucket_file (media_cache.c) * named_acl Change-Id: Ide4c1449a894bce70dea1fef664dade9b57578f1	7 years ago
Corey Farrell	a6e5bae3ef	Remove ASTERISK_REGISTER_FILE. ... ASTERISK_REGISTER_FILE no longer has any purpose so this commit removes all traces of it. Previously exported symbols removed: * __ast_register_file * __ast_unregister_file * ast_complete_source_filename This also removes the mtx_prof static variable that was declared when MTX_PROFILE was enabled. This variable was only used in lock.c so it is now initialized in that file only. ASTERISK-26480 #close Change-Id: I1074af07d71f9e159c48ef36631aa432c86f9966	9 years ago
George Joseph	22db16fa81	endpoint/stasis: Eliminate duplicate events on endpoint status change ... When an endpoint is created, its messages are forwarded to both the tech endpoint topic and the all endpoints topic. This is done so that various parties interested in endpoint messages can subscribe to just the tech endpoint and receive all messages associated with that particular technology, as opposed to subscribing to the all endpoints topic. Unfortunately, when the tech endpoint is created, it also forwards all of its messages to the all topic. This results in duplicate messages whenever an endpoint publishes its messages. This patch resolves the duplicate message issue by creating a new function for Stasis caching topics, stasis_cp_sink_create. In most respects, this acts as a normal caching topic, save that it no longer forwards messages it receives to the all endpoints topic. This allows it to act as an aggregation "sink", while preserving the necessary caching behaviour. ASTERISK-25137 #close Reported-by: Vitezslav Novy ASTERISK-25116 #close Reported-by: George Joseph <george.joseph@fairview5.com> Tested-by: George Joseph <george.joseph@fairview5.com> Change-Id: Ie47784adfb973ab0063e59fc18f390d7dd26d17b	9 years ago
Scott Griepentrog	7691035312	endpoint snapshot: avoid second cleanup on alloc failure ... In ast_endpoint_snapshot_create(), a failure to init the string fields results in two attempts to ao2_cleanup the same pointer. Removed RAII_VAR to eliminate problem. ASTERISK-25375 #close Reported by: Scott Griepentrog Change-Id: If4d9dfb1bbe3836b623642ec690b6d49b25e8979	10 years ago
George Joseph	82716410a4	Revert "endpoint/stasis: Eliminate duplicate events on endpoint status change" ... This reverts commit 6fca75bb62. Change-Id: Ifee026cc63e22c5ac5717c37867a9f036373ae5a	10 years ago
George Joseph	6fca75bb62	endpoint/stasis: Eliminate duplicate events on endpoint status change ... When an endpoint was created, it's messages were being forwarded to both the tech endpoint topic and the all endpoints topic. Since the tech topic was also forwarded to all, this was resulting in duplicate messages whenever an endpoint published. This patch causes the endpoint to only forward to the tech topic and lets the tech topic forward to all. To accomplish this, the existing stasis_cp_single_create function (which both creates and forwards) was cloned and split into 2 functions, one that creates the topic and one that sets up the forwarding. This allows endpoint_internal_create to create the topic from the endpoint_all cache without forwarding it there, then allows it to do the forward to the tech's topic. ASTERISK-25137 #close Reported-by: Vitezslav Novy ASTERISK-25116 #close Reported-by: George Joseph <george.joseph@fairview5.com> Tested-by: George Joseph <george.joseph@fairview5.com> Change-Id: I26d7d4926a0861748fd3bdffe316b75b549a801c	10 years ago
George Joseph	51886c68dc	pjsip_options: Add qualify_timeout processing and eventing ... This is the second follow-on to https://reviewboard.asterisk.org/r/4572/ and the discussion at http://lists.digium.com/pipermail/asterisk-dev/2015-March/073921.html The basic issues are that changes in contact status don't cause events to be emitted for the associated endpoint. Only dynamic contact add/delete actions update the endpoint. Also, the qualify timeout is fixed by pjsip at 32 seconds which is a long time. This patch makes use of the new transaction timeout feature in r4585 and provides the following capabilities... 1. A new aor/contact variable 'qualify_timeout' has been added that allows the user to specify the maximum time in milliseconds to wait for a response to an OPTIONS message. The default is 3000ms. When the timer expires, the contact is marked unavailable. 2. Contact status changes are now propagated up to the endpoint as follows... When any contact is 'Available', the endpoint is marked as 'Reachable'. When all contacts are 'Unavailable', the endpoint is marked as 'Unreachable'. The existing endpoint events are generated appropriately. ASTERISK-24863 #close Change-Id: Id0ce0528e58014da1324856ea537e7765466044a Tested-by: Dmitriy Serov Tested-by: George Joseph <george.joseph@fairview5.com>	10 years ago
Matt Jordan	4a58261694	git migration: Refactor the ASTERISK_FILE_VERSION macro ... Git does not support the ability to replace a token with a version string during check-in. While it does have support for replacing a token on clone, this is somewhat sub-optimal: the token is replaced with the object hash, which is not particularly easy for human consumption. What's more, in practice, the source file version was often not terribly useful. Generally, when triaging bugs, the overall version of Asterisk is far more useful than an individual SVN version of a file. As a result, this patch removes Asterisk's support for showing source file versions. Specifically, it does the following: * Rename ASTERISK_FILE_VERSION macro to ASTERISK_REGISTER_FILE, and remove passing the version in with the macro. Other facilities than 'core show file version' make use of the file names, such as setting a debug level only on a specific file. As such, the act of registering source files with the Asterisk core still has use. The macro rename now reflects the new macro purpose. * main/asterisk: - Refactor the file_version structure to reflect that it no longer tracks a version field. - Remove the "core show file version" CLI command. Without the file version, it is no longer useful. - Remove the ast_file_version_find function. The file version is no longer tracked. - Rename ast_register_file_version/ast_unregister_file_version to ast_register_file/ast_unregister_file, respectively. * main/manager: Remove value from the Version key of the ModuleCheck Action. The actual key itself has not been removed, as doing so would absolutely constitute a backwards incompatible change. However, since the file version is no longer tracked, there is no need to attempt to include it in the Version key. * UPGRADE: Add notes for: - Modification to the ModuleCheck AMI Action - Removal of the "core show file version" CLI command Change-Id: I6cf0ff280e1668bf4957dc21f32a5ff43444a40e	10 years ago
Scott Griepentrog	8c65c9167e	Various: bugfixes found via chaos ... Using DEBUG_CHAOS several instances of a null pointer crash, and one uninitialized variable were uncovered and fixed. Also added details on why Asterisk failed to initialize. Review: https://reviewboard.asterisk.org/r/4468/ ........ Merged revisions 433064 from http://svn.asterisk.org/svn/asterisk/branches/13 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@433065 65c4cc65-6c06-0410-ace0-fbb531ad65f3	10 years ago
Joshua Colp	2889f074a0	core: Don't create snapshots with locks. ... Snapshots are immutable and are never changed. Allocating them with a lock is wasteful. Review: https://reviewboard.asterisk.org/r/4469/ ........ Merged revisions 432742 from http://svn.asterisk.org/svn/asterisk/branches/13 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@432743 65c4cc65-6c06-0410-ace0-fbb531ad65f3	10 years ago
Matthew Jordan	1106e8fd0f	main/stasis: Allow subscriptions to use a threadpool for message delivery ... Prior to this patch, all Stasis subscriptions would receive a dedicated thread for servicing published messages. In contrast, prior to r400178 (see review https://reviewboard.asterisk.org/r/2881/), the subscriptions shared a thread pool. It was discovered during some initial work on Stasis that, for a low subscription count with high message throughput, the threadpool was not as performant as simply having a dedicated thread per subscriber. For situations where a subscriber receives a substantial number of messages and is always present, the model of having a dedicated thread per subscriber makes sense. While we still have plenty of subscriptions that would follow this model, e.g., AMI, CDRs, CEL, etc., there are plenty that also fall into the following two categories: * Large number of subscriptions, specifically those tied to endpoints/peers. * Low number of messages. Some subscriptions exist specifically to coordinate a single message - the subscription is created, a message is published, the delivery is synchronized, and the subscription is destroyed. In both of the latter two cases, creating a dedicated thread is wasteful (and in the case of a large number of peers/endpoints, harmful). In those cases, having shared delivery threads is far more performant. This patch adds the ability of a subscriber to Stasis to choose whether or not their messages are dispatched on a dedicated thread or on a threadpool. The threadpool is configurable through stasis.conf. Review: https://reviewboard.asterisk.org/r/4193 ASTERISK-24533 #close Reported by: xrobau Tested by: xrobau ........ Merged revisions 428681 from http://svn.asterisk.org/svn/asterisk/branches/12 ........ Merged revisions 428687 from http://svn.asterisk.org/svn/asterisk/branches/13 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@428688 65c4cc65-6c06-0410-ace0-fbb531ad65f3	11 years ago
Kinsey Moore	f1036f40dc	Stasis: Allow message types to be blocked ... This introduces stasis.conf and a mechanism to prevent certain message types from being published. Internally, this works by preventing the chosen message types from being created which ensures that those message types can never be published. This patch also adjusts message publishers such that message payloads are not created if the related message type is not available. ASTERISK-23943 #close Review: https://reviewboard.asterisk.org/r/3823/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@420124 65c4cc65-6c06-0410-ace0-fbb531ad65f3	11 years ago
Matthew Jordan	ccc6e8bd17	endpoints: Fix failing unit tests from r419196 ... This patch does two things: (1) It updates the unit tests to expect additional stasis messages. More messages are now sent to the endpoint topic, due to forwarding all channel messages and the forwarding relationship set up between endpoints themselves. (2) Remove the technology forwarding subscription during ast_endpoint_shutdown. This prevents an improper double shutdown of an endpoint from occurring. ........ Merged revisions 419318 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419319 65c4cc65-6c06-0410-ace0-fbb531ad65f3	11 years ago
Matthew Jordan	bb87796f67	ARI: Fix endpoint/channel subscription issues; allow for subscriptions to tech ... This patch serves two purposes: (1) It fixes some bugs with endpoint subscriptions not reporting all of the channel events (2) It serves as the preliminary work needed for ASTERISK-23692, which allows for sending/receiving arbitrary out of call text messages through ARI in a technology agnostic fashion. The messaging functionality described on ASTERISK-23692 requires two things: (1) The ability to send/receive messages associated with an endpoint. This is relatively straight forwards with the endpoint core in Asterisk now. (2) The ability to send/receive messages associated with a technology and an arbitrary technology defined URI. This is less straight forward, as endpoints are formed from a tech + resource pair. We don't have a mechanism to note that a technology that *may* have endpoints exists. This patch provides such a mechanism, and fixes a few bugs along the way. The first major bug this patch fixes is the forwarding of channel messages to their respective endpoints. Prior to this patch, there were two problems: (1) Channel caching messages weren't forwarded. Thus, the endpoints missed most of the interesting bits (such as channel creation, destruction, state changes, etc.) (2) Channels weren't associated with their endpoint until after creation. This resulted in endpoints missing the channel creation message, which limited the usefulness of the subscription in the first place (a major use case being 'tell me when this endpoint has a channel'). Unfortunately, this meant another parameter to ast_channel_alloc. Since not all channel technologies support an ast_endpoint, this patch makes such a call optional and opts for a new function, ast_channel_alloc_with_endpoint. When endpoints are created, they will implicitly create a technology endpoint for their technology (if one does not already exist). A technology endpoint is special in that it has no state, cannot have channels created for it, cannot be created explicitly, and cannot be destroyed except on shutdown. It does, however, have all messages from other endpoints in its technology forwarded to it. Combined with the bug fixes, we now have Stasis messages being properly forwarded. Consider the following scenario: two PJSIP endpoints (foo and bar), where bar has a single channel associated with it and foo has two channels associated with it. The messages would be forwarded as follows: channel PJSIP/foo-1 -- \ --> endpoint PJSIP/foo -- / \ channel PJSIP/foo-2 -- \ ---- > endpoint PJSIP / channel PJSIP/bar-1 -----> endpoint PJSIP/bar -- ARI, through the applications resource, can: - subscribe to endpoint:PJSIP/foo and get notifications for channels PJSIP/foo-1,PJSIP/foo-2 and endpoint PJSIP/foo - subscribe to endpoint:PJSIP/bar and get notifications for channels PJSIP/bar-1 and endpoint PJSIP/bar - subscribe to endpoint:PJSIP and get notifications for channels PJSIP/foo-1,PJSIP/foo-2,PJSIP/bar-1 and endpoints PJSIP/foo,PJSIP/bar Note that since endpoint PJSIP never changes, it never has events itself. It merely provides an aggregation point for all other endpoints in its technology (which in turn aggregate all channel messages associated with that endpoint). This patch also adds endpoints to res_xmpp and chan_motif, because the actual messaging work will need it (messaging without XMPP is just sad). Review: https://reviewboard.asterisk.org/r/3760/ ASTERISK-23692 ........ Merged revisions 419196 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419203 65c4cc65-6c06-0410-ace0-fbb531ad65f3	11 years ago
Matthew Jordan	9cc1a8e893	stasis: Reduce creation of channel snapshots to improve performance ... During some performance testing of Asterisk with AGI, ARI, and lots of Local channels, we noticed that there's quite a hit in performance during channel creation and releasing to the dialplan (ARI continue). After investigating the performance spike that occurs during channel creation, we discovered that we create a lot of channel snapshots that are technically unnecessary. This includes creating snapshots during: * AGI execution * Returning objects for ARI commands * During some Local channel operations * During some dialling operations * During variable setting * During some bridging operations And more. This patch does the following: - It removes a number of fields from channel snapshots. These fields were rarely used, were expensive to have on the snapshot, and hurt performance. This included formats, translation paths, Log Call ID, callgroup, pickup group, and all channel variables. As a result, AMI Status, "core show channel", "core show channelvar", and "pjsip show channel" were modified to either hit the live channel or not show certain pieces of data. While this is unfortunate, the performance gain from this patch is worth the loss in behaviour. - It adds a mechanism to publish a cached snapshot + blob. A large number of publications were changed to use this, including: - During Dial begin - During Variable assignment (if no AMI variables are emitted - if AMI variables are set, we have to make snapshots when a variable is changed) - During channel pickup - When a channel is put on hold/unhold - When a DTMF digit is begun/ended - When creating a bridge snapshot - When an AOC event is raised - During Local channel optimization/Local bridging - When endpoint snapshots are generated - All AGI events - All ARI responses that return a channel - Events in the AgentPool, MeetMe, and some in Queue - Additionally, some extraneous channel snapshots were being made that were unnecessary. These were removed. - The result of ast_hashtab_hash_string is now cached in stasis_cache. This reduces a large number of calls to ast_hashtab_hash_string, which reduced the amount of time spent in this function in gprof by around 50%. #ASTERISK-23811 #close Reported by: Matt Jordan Review: https://reviewboard.asterisk.org/r/3568/ ........ Merged revisions 416211 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@416216 65c4cc65-6c06-0410-ace0-fbb531ad65f3	11 years ago
Kevin Harwell	28c0cb28d0	channel locking: Add locking for channel snapshot creation ... Original commit message by mmichelson (asterisk 12 r403311): "This adds channel locks around calls to create channel snapshots as well as other functions which operate on a channel and then end up creating a channel snapshot. Functions that expect the channel to be locked prior to being called have had their documentation updated to indicate such." The above was initially committed and then reverted at r403398. The problem was found to be in core_local.c in the publish_local_bridge_message function. The ast_unreal_lock_all function locks and adds a reference to the returned channels and while they were being unlocked they were not being unreffed when no longer needed. Fixed by unreffing the channels. Also in bridge.c a lock was obtained on "other->chan", but then an attempt was made to unlock "other" and not the previously locked channel. Fixed by unlocking "other->chan" (closes issue ASTERISK-22709) Reported by: John Bigelow ........ Merged revisions 404237 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@404260 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
Joshua Colp	dcb642e2da	endpoints: Keep a reference to channel ids when creating snapshot. ... The snapshot process for endpoints uses the channel ids present on the endpoint itself. Without keeping a reference it was possible for the strings to be freed underneath any consumer of an endpoint snapshot. A reference is now held by the snapshot to the channel ids and released when the snapshot is destroyed. (issue ASTERISK-22801) Reported by: Matt Jordan ........ Merged revisions 403542 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@403543 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
David M. Lee	1212906351	Reverting r403311. It's causing ARI tests to hang. ... ........ Merged revisions 403398 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@403404 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
Mark Michelson	8e8b329e14	Add channel locking for channel snapshot creation. ... This adds channel locks around calls to create channel snapshots as well as other functions which operate on a channel and then end up creating a channel snapshot. Functions that expect the channel to be locked prior to being called have had their documentation updated to indicate such. ........ Merged revisions 403311 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@403314 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
Matthew Jordan	8d7873b836	ARI: Add subscription support ... This patch adds an /applications API to ARI, allowing explicit management of Stasis applications. * GET /applications - list current applications * GET /applications/{applicationName} - get details of a specific application * POST /applications/{applicationName}/subscription - explicitly subscribe to a channel, bridge or endpoint * DELETE /applications/{applicationName}/subscription - explicitly unsubscribe from a channel, bridge or endpoint Subscriptions work by a reference counting mechanism: if you subscript to an event source X number of times, you must unsubscribe X number of times to stop receiveing events for that event source. Review: https://reviewboard.asterisk.org/r/2862 (issue ASTERISK-22451) Reported by: Matt Jordan ........ Merged revisions 400522 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@400523 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
David M. Lee	2de42c2a25	Multiple revisions 399887,400138,400178,400180-400181 ... ........ r399887 | dlee | 2013-09-26 10:41:47 -0500 (Thu, 26 Sep 2013) | 1 line Minor performance bump by not allocate manager variable struct if we don't need it ........ r400138 | dlee | 2013-09-30 10:24:00 -0500 (Mon, 30 Sep 2013) | 23 lines Stasis performance improvements This patch addresses several performance problems that were found in the initial performance testing of Asterisk 12. The Stasis dispatch object was allocated as an AO2 object, even though it has a very confined lifecycle. This was replaced with a straight ast_malloc(). The Stasis message router was spending an inordinate amount of time searching hash tables. In this case, most of our routers had 6 or fewer routes in them to begin with. This was replaced with an array that's searched linearly for the route. We more heavily rely on AO2 objects in Asterisk 12, and the memset() in ao2_ref() actually became noticeable on the profile. This was #ifdef'ed to only run when AO2_DEBUG was enabled. After being misled by an erroneous comment in taskprocessor.c during profiling, the wrong comment was removed. Review: https://reviewboard.asterisk.org/r/2873/ ........ r400178 | dlee | 2013-09-30 13:26:27 -0500 (Mon, 30 Sep 2013) | 24 lines Taskprocessor optimization; switch Stasis to use taskprocessors This patch optimizes taskprocessor to use a semaphore for signaling, which the OS can do a better job at managing contention and waiting that we can with a mutex and condition. The taskprocessor execution was also slightly optimized to reduce the number of locks taken. The only observable difference in the taskprocessor implementation is that when the final reference to the taskprocessor goes away, it will execute all tasks to completion instead of discarding the unexecuted tasks. For systems where unnamed semaphores are not supported, a really simple semaphore implementation is provided. (Which gives identical performance as the original taskprocessor implementation). The way we ended up implementing Stasis caused the threadpool to be a burden instead of a boost to performance. This was switched to just use taskprocessors directly for subscriptions. Review: https://reviewboard.asterisk.org/r/2881/ ........ r400180 | dlee | 2013-09-30 13:39:34 -0500 (Mon, 30 Sep 2013) | 28 lines Optimize how Stasis forwards are dispatched This patch optimizes how forwards are dispatched in Stasis. Originally, forwards were dispatched as subscriptions that are invoked on the publishing thread. This did not account for the vast number of forwards we would end up having in the system, and the amount of work it would take to walk though the forward subscriptions. This patch modifies Stasis so that rather than walking the tree of forwards on every dispatch, when forwards and subscriptions are changed, the subscriber list for every topic in the tree is changed. This has a couple of benefits. First, this reduces the workload of dispatching messages. It also reduces contention when dispatching to different topics that happen to forward to the same aggregation topic (as happens with all of the channel, bridge and endpoint topics). Since forwards are no longer subscriptions, the bulk of this patch is simply changing stasis_subscription objects to stasis_forward objects (which, admittedly, I should have done in the first place.) Since this required me to yet again put in a growing array, I finally abstracted that out into a set of ast_vector macros in asterisk/vector.h. Review: https://reviewboard.asterisk.org/r/2883/ ........ r400181 | dlee | 2013-09-30 13:48:57 -0500 (Mon, 30 Sep 2013) | 28 lines Remove dispatch object allocation from Stasis publishing While looking for areas for performance improvement, I realized that an unused feature in Stasis was negatively impacting performance. When a message is sent to a subscriber, a dispatch object is allocated for the dispatch, containing the topic the message was published to, the subscriber the message is being sent to, and the message itself. The topic is actually unused by any subscriber in Asterisk today. And the subscriber is associated with the taskprocessor the message is being dispatched to. First, this patch removes the unused topic parameter from Stasis subscription callbacks. Second, this patch introduces the concept of taskprocessor local data, data that may be set on a taskprocessor and provided along with the data pointer when a task is pushed using the ast_taskprocessor_push_local() call. This allows the task to have both data specific to that taskprocessor, in addition to data specific to that invocation. With those two changes, the dispatch object can be removed completely, and the message is simply refcounted and sent directly to the taskprocessor. Review: https://reviewboard.asterisk.org/r/2884/ ........ Merged revisions 399887,400138,400178,400180-400181 from http://svn.asterisk.org/svn/asterisk/branches/12 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@400186 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
David M. Lee	e1b959ccbb	Split caching out from the stasis_caching_topic. ... In working with res_stasis, I discovered a significant limitation to the current structure of stasis_caching_topics: you cannot subscribe to cache updates for a single channel/bridge/endpoint/etc. To address this, this patch splits the cache away from the stasis_caching_topic, making it a first class object. The stasis_cache object is shared amongst individual stasis_caching_topics that are created per channel/endpoint/etc. These are still forwarded to global whatever_all_cached topics, so their use from most of the code does not change. In making these changes, I noticed that we frequently used a similar pattern for bridges, endpoints and channels: single_topic ----------------> all_topic ^ | single_topic_cached ----+----> all_topic_cached | +----> cache This pattern was extracted as the 'Stasis Caching Pattern', defined in stasis_caching_pattern.h. This avoids a lot of duplicate code between the different domain objects. Since the cache is now disassociated from its upstream caching topics, this also necessitated a change to how the 'guaranteed' flag worked for retrieving from a cache. The code for handling the caching guarantee was extracted into a 'stasis_topic_wait' function, which works for any stasis_topic. (closes issue ASTERISK-22002) Review: https://reviewboard.asterisk.org/r/2672/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@395954 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
Joshua Colp	65c492e851	Add support for requiring that all queued messages on a caching topic have been handled before ... retrieving from the cache and also change adding channels to an endpoint to be an immediate operation. Review: https://reviewboard.asterisk.org/r/2599/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@391596 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
Matthew Jordan	272dd008d0	Fix memory leak while loading modules, adding formats, and destroying endpoints ... This patch fixes three memory leaks * When we load a module with the LOAD_PRIORITY flag, we remove its entry from the load order list. Unfortunately, we don't free the memory associated with entry in the list. This patch corrects that and properly frees the memory for the module in the list. * When adding a custom format (such as SILK or CELT), the routine for adding the format was leaking a reference. RAII_VAR cleans this up properly. * We now de-ref the channel_snapshot appropriately when an endpoint is disposed of ........ Merged revisions 391489 from http://svn.asterisk.org/svn/asterisk/branches/1.8 ........ Merged revisions 391507 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@391521 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
Kinsey Moore	759a7e4a30	Rework stasis cache clear events ... Stasis cache clear message payloads now consist of a stasis_message representative of the message to be cleared from the cache. This allows multiple parallel caches to coexist and be cleared properly by the same cache clear message even when keyed on different fields. This change fixes a bug where multiple cache clears could be posted for channels. The cache clear is now produced in the destructor instead of ast_hangup. Additionally, dummy channels are no longer capable of producing channel snapshots. Review: https://reviewboard.asterisk.org/r/2596 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@390830 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
David M. Lee	b97c71bb11	Fix shutdown assertions in stasis-core ... In r388005, macros were introduced to consistently define message types. This added an assert if a message type was used either before it was initialized or after it had been cleaned up. It turns out that this assertion fires during shutdown. This actually exposed a hidden shutdown ordering problem. Since unsubscribing is asynchronous, it's possible that the message types used by the subscription could be freed before the final message of the subscription was processed. This patch adds stasis_subscription_join(), which blocks until the last message has been processed by the subscription. Since joining was most commonly done right after an unsubscribe, a stasis_unsubscribe_and_join() convenience function was also added. Similar functions were also added to the stasis_caching_topic and stasis_message_router, since they wrap subscriptions and have similar problems. Other code in trunk was refactored to join() where appropriate, or at least verify that the subscription was complete before being destroyed. Review: https://reviewboard.asterisk.org/r/2540 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@389011 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago
David M. Lee	e06e519a90	Initial support for endpoints. ... An endpoint is an external device/system that may offer/accept channels to/from Asterisk. While this is a very useful concept for end users, it is surprisingly not a core concept within Asterisk itself. This patch defines ast_endpoint as a separate object, which channel drivers may use to expose their concept of an endpoint. As the channel driver creates channels, it can use ast_endpoint_add_channel() to associate channels to the endpoint. This updated the endpoint appropriately, and forwards all of the channel's events to the endpoint's topic. In order to avoid excessive locking on the endpoint object itself, the mutable state is not accessible via getters. Instead, you can create a snapshot using ast_endpoint_snapshot_create() to get a consistent snapshot of the internal state. This patch also includes a set of topics and messages associated with endpoints, and implementations of the endpoint-related RESTful API. chan_sip was updated to create endpoints with SIP peers, but the state of the endpoints is not updated with the state of the peer. Along for the ride in this patch is a Stasis test API. This is a stasis_message_sink object, which can be subscribed to a Stasis topic. It has functions for blocking while waiting for conditions in the message sink to be fulfilled. (closes issue ASTERISK-21421) Review: https://reviewboard.asterisk.org/r/2492/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@387932 65c4cc65-6c06-0410-ace0-fbb531ad65f3	12 years ago