)]}'
{
  "log": [
    {
      "commit": "d9ab53ab40dcba66f34ca75695dc930e6093f789",
      "tree": "552f537b8f5decf1b74d8fc98b96ef3e65a01051",
      "parents": [
        "5bcd754ff8d7947978acb44e77dcab323973fb1e"
      ],
      "author": {
        "name": "Dinesh Dutt",
        "email": "ddutt@cumulusnetworks.com",
        "time": "Tue May 19 17:47:21 2015 -0700"
      },
      "committer": {
        "name": "Paul Jakma",
        "email": "paul.jakma@hpe.com",
        "time": "Tue Oct 04 13:07:52 2016 +0100"
      },
      "message": "bgpd, zebra: Use next hop tracking for connected routes too\n\nAllow next hop tracking to work with connected routes\nAnd cleanup obsolete code in bgp_scan and bgp_import.\n\nSigned-off-by: Dinesh Dutt \u003cddutt@cumulusnetworks.com\u003e\nSigned-off-by: Donald Sharp \u003csharpd@cumulusnetworks.com\u003e\nEdits: Paul Jakma \u003cpaul.jakma@hpe.com\u003e Rebase re-ordering conflicts with\n       NHT route-map, potential errors.\n"
    },
    {
      "commit": "774914f4223532256051bd6dd61cac20e8f9649f",
      "tree": "c6cecfa4c6c1eaff07638d2903e195571baa39b0",
      "parents": [
        "eeef0db2e9260fe76acb328a339025c432eb7c22"
      ],
      "author": {
        "name": "Donald Sharp",
        "email": "sharpd@cumulusnetworks.com",
        "time": "Wed Oct 14 08:50:39 2015 -0400"
      },
      "committer": {
        "name": "Paul Jakma",
        "email": "paul@quagga.net",
        "time": "Tue Oct 27 10:53:22 2015 +0000"
      },
      "message": "bgpd: fix using of two pointers for struct thread_master *\n\nbgp is using both bm-\u003emaster and master pointers interchangebly\nfor thread manipulation.  Since they are the same thing consolidate\nto one pointer.\n\nSigned-off-by: Donald Sharp \u003csharpd@cumulusnetworks.com\u003e\n"
    },
    {
      "commit": "dcdf399fbc7420d3edc230c9646f720825fcecda",
      "tree": "d4712072c844fff67e49e44fc9503d0b11f8062f",
      "parents": [
        "876b8be0ab24721e8f94d47dde022563f76db992"
      ],
      "author": {
        "name": "Paul Jakma",
        "email": "paul.jakma@sun.com",
        "time": "Sun Oct 15 23:39:59 2006 +0000"
      },
      "committer": {
        "name": "Paul Jakma",
        "email": "paul.jakma@sun.com",
        "time": "Sun Oct 15 23:39:59 2006 +0000"
      },
      "message": "[bgpd] Bug #302 fixes. ClearingCompleted event gets flushed, leaving peers stuck in Clearing.\n\n2006-10-14 Paul Jakma \u003cpaul.jakma@sun.com\u003e\n\n\t* bgp_fsm.h: Remove BGP_EVENT_FLUSH_ADD, dangerous and not\n\t  needed.\n\t* bgp_fsm.c: (bgp_stop) Move BGP_EVENT_FLUSH to the top of the\n\t  of the function, otherwise it could flush a ClearingCompleted\n          event, bug #302.\n\t* bgp_packet.c: Replace all BGP_EVENT_FLUSH_ADD with\n\t  BGP_EVENT_ADD, fixing bug #302.\n"
    },
    {
      "commit": "9fde6624fc480995449d8243fe85602d89927eb6",
      "tree": "6591e476329ec88bd9ba5fcc008c626d32b3fdb5",
      "parents": [
        "ca058a30b1ea57f83871ab4cf1c9a91ea4064d52"
      ],
      "author": {
        "name": "Paul Jakma",
        "email": "paul.jakma@sun.com",
        "time": "Thu Sep 14 03:02:02 2006 +0000"
      },
      "committer": {
        "name": "Paul Jakma",
        "email": "paul.jakma@sun.com",
        "time": "Thu Sep 14 03:02:02 2006 +0000"
      },
      "message": "[bgpd] simplify peer refcounts, squash slow peer leak\n\n2006-09-14 Paul Jakma \u003cpaul.jakma@sun.com\u003e\n\n\t* (general) fix the peer refcount issue exposed by previous, by\n\t  just removing refcounting of peer threads, which is mostly\n\t  senseless as they\u0027re references leading from struct peer,\n\t  which peer_free cancels anyway. No need to muck around..\n\t* bgp_fsm.h: Just remove the refcounting from the various\n\t  TIMER/READ/WRITE/EVENT ON/OFF/ADD macros.\n\t* bgp_fsm.c: (bgp_stop) use BGP_EVENT_FLUSH, no refcounts attached\n\t  to events anymore.\n\t  (bgp_event) remove peer_unlock, events not refcounted.\n\t* bgpd.c: (peer_free) flush events before free.\n"
    },
    {
      "commit": "ca058a30b1ea57f83871ab4cf1c9a91ea4064d52",
      "tree": "ab38ab59bad607c9b41a093cb8b35bec766f30b3",
      "parents": [
        "2815e61ffbbf9c362896f3912d925cf78e125ee1"
      ],
      "author": {
        "name": "Paul Jakma",
        "email": "paul.jakma@sun.com",
        "time": "Thu Sep 14 02:58:49 2006 +0000"
      },
      "committer": {
        "name": "Paul Jakma",
        "email": "paul.jakma@sun.com",
        "time": "Thu Sep 14 02:58:49 2006 +0000"
      },
      "message": "[bgpd] Fix 0.99 shutdown regression, introduce Clearing and Deleted states\n\n2006-09-14 Paul Jakma \u003cpaul.jakma@sun.com\u003e\n\n\t* (general) Fix some niggly issues around \u0027shutdown\u0027 and clearing\n\t  by adding a Clearing FSM wait-state and a hidden \u0027Deleted\u0027\n\t  FSM state, to allow deleted peers to \u0027cool off\u0027 and hit 0\n\t  references. This introduces a slow memory leak of struct peer,\n\t  however that\u0027s more a testament to the fragility of the\n\t  reference counting than a bug in this patch, cleanup of\n\t  reference counting to fix this is to follow.\n\t* bgpd.h: Add Clearing, Deleted states and Clearing_Completed\n\t  and event.\n\t* bgp_debug.c: (bgp_status_msg[]) Add strings for Clearing and\n\t  Deleted.\n\t* bgp_fsm.h: Don\u0027t allow timer/event threads to set anything\n\t  for Deleted peers.\n\t* bgp_fsm.c: (bgp_timer_set) Add Clearing and Deleted. Deleted\n\t  needs to stop everything.\n\t  (bgp_stop) Remove explicit fsm_change_status call, the\n\t  general framework handles the transition.\n\t  (bgp_start) Log a warning if a start is attempted on a peer\n\t  that should stay down, trying to start a peer.\n\t  (struct .. FSM) Add Clearing_Completed\n\t  events, has little influence except when in state\n\t  Clearing to signal wait-state can end.\n\t  Add Clearing and Deleted states, former is a wait-state,\n\t  latter is a placeholder state to allow peers to disappear\n\t  quietly once refcounts settle.\n\t  (bgp_event) Try reduce verbosity of FSM state-change debug,\n\t  changes to same state are not interesting (Established-\u003eEstablished)\n\t  Allow NULL action functions in FSM.\n\t* bgp_packet.c: (bgp_write) Use FSM events, rather than trying\n\t  to twiddle directly with FSM state behind the back of FSM.\n\t  (bgp_write_notify) ditto.\n\t  (bgp_read) Remove the vague ACCEPT_PEER peer_unlock, or else\n\t  this patch crashes, now it leaks instead.\n\t* bgp_route.c: (bgp_clear_node_complete) Clearing_Completed\n\t  event, to end clearing.\n\t  (bgp_clear_route) See extensive comments.\n\t* bgpd.c: (peer_free) should only be called while in Deleted,\n\t  peer refcounting controls when peer_free is called.\n\t  bgp_sync_delete should be here, not in peer_delete.\n\t  (peer_delete) Initiate delete.\n\t  Transition to Deleted state manually.\n\t  When removing peer from indices that provide visibility of it,\n\t  take great care to be idempotent wrt the reference counting\n\t  of struct peer through those indices.\n\t  Use bgp_timer_set, rather than replicating.\n\t  Call to bgp_sync_delete isn\u0027t appropriate here, sync can be\n\t  referenced while shutting down and finishing deletion.\n\t  (peer_group_bind) Take care to be idempotent wrt list references\n\t  indexing peers.\n"
    },
    {
      "commit": "200df115ea0ba6f54151d60fe5e9a7f6948de7c3",
      "tree": "2aedc0af80d2e805cdf2cefd50432423d2871216",
      "parents": [
        "b31b813671d96ea9e96801508442a90f0ef503fa"
      ],
      "author": {
        "name": "paul",
        "email": "paul",
        "time": "Wed Jun 01 11:17:05 2005 +0000"
      },
      "committer": {
        "name": "paul",
        "email": "paul",
        "time": "Wed Jun 01 11:17:05 2005 +0000"
      },
      "message": "2005-06-01 Paul Jakma \u003cpaul.jakma@sun.com\u003e\n\n\t* bgpd/(general) refcount struct peer and bgp_info, hence allowing us\n\t  add work_queues for bgp_process.\n\t* bgpd/bgp_route.h: (struct bgp_info) Add \u0027lock\u0027 field for refcount.\n\t  Add bgp_info_{lock,unlock} helper functions.\n\t  Add bgp_info_{add,delete} helpers, to remove need for\n\t  users managing locking/freeing of bgp_info and bgp_node\u0027s.\n\t* bgpd/bgp_table.h: (struct bgp_node) Add a flags field, and\n\t  BGP_NODE_PROCESS_SCHEDULED to merge redundant processing of\n\t  nodes.\n\t* bgpd/bgp_fsm.h: Make the ON/OFF/ADD/REMOVE macros lock and unlock\n\t  peer reference as appropriate.\n\t* bgpd/bgp_damp.c: Remove its internal prototypes for\n\t  bgp_info_delete/free. Just use bgp_info_delete.\n\t* bgpd/bgpd.h: (struct bgp_master) Add work_queue pointers.\n\t  (struct peer) Add reference count \u0027lock\u0027\n\t  (peer_lock,peer_unlock) New helpers to take/release reference\n\t  on struct peer.\n\t* bgpd/bgp_advertise.c: (general) Add peer and bgp_info refcounting\n\t  and balance how references are taken and released.\n\t  (bgp_advertise_free) release bgp_info reference, if appropriate\n\t  (bgp_adj_out_free) unlock peer\n\t  (bgp_advertise_clean) leave the adv references alone, or else\n\t  call bgp_advertise_free cant unlock them.\n\t  (bgp_adj_out_set) lock the peer on new adj\u0027s, leave the reference\n\t  alone otherwise. lock the new bgp_info reference.\n\t  (bgp_adj_in_set) lock the peer reference\n\t  (bgp_adj_in_remove) and unlock it here\n\t  (bgp_sync_delete) make hash_free on peer conditional, just in\n\t  case.\n\t* bgpd/bgp_fsm.c: (general) document that the timers depend on\n\t  bgp_event to release a peer reference.\n\t  (bgp_fsm_change_status) moved up the file, unchanged.\n\t  (bgp_stop) Decrement peer lock as many times as cancel_event\n\t  canceled - shouldnt be needed but just in case.\n\t  stream_fifo_clean of obuf made conditional, just in case.\n\t  (bgp_event) always unlock the peer, regardless of return value\n\t  of bgp_fsm_change_status.\n\t* bgpd/bgp_packet.c: (general) change several bgp_stop\u0027s to BGP_EVENT\u0027s.\n\t  (bgp_read) Add a mysterious extra peer_unlock for ACCEPT_PEERs\n\t  along with a comment on it.\n\t* bgpd/bgp_route.c: (general) Add refcounting of bgp_info, cleanup\n\t  some of the resource management around bgp_info. Refcount peer.\n\t  Add workqueues for bgp_process and clear_table.\n\t  (bgp_info_new) make static\n\t  (bgp_info_free) Ditto, and unlock the peer reference.\n\t  (bgp_info_lock,bgp_info_unlock) new exported functions\n\t  (bgp_info_add) Add a bgp_info to a bgp_node in correct fashion,\n\t  taking care of reference counts.\n\t  (bgp_info_delete) do the opposite of bgp_info_add.\n\t  (bgp_process_rsclient) Converted into a work_queue work function.\n\t  (bgp_process_main) ditto.\n\t  (bgp_processq_del) process work queue item deconstructor\n\t  (bgp_process_queue_init) process work queue init\n\t  (bgp_process) call init function if required, set up queue item\n\t  and add to queue, rather than calling process functions directly.\n\t  (bgp_rib_remove) let bgp_info_delete manage bgp_info refcounts\n\t  (bgp_rib_withdraw) ditto\n\t  (bgp_update_rsclient) let bgp_info_add manage refcounts\n\t  (bgp_update_main) ditto\n\t  (bgp_clear_route_node) clear_node_queue work function, does\n\t  per-node aspects of what bgp_clear_route_table did previously\n\t  (bgp_clear_node_queue_del) clear_node_queue item delete function\n\t  (bgp_clear_node_complete) clear_node_queue completion function,\n\t  it unplugs the process queues, which have to be blocked while\n\t  clear_node_queue is being processed to prevent a race.\n\t  (bgp_clear_node_queue_init) init function for clear_node_queue\n\t  work queues\n\t  (bgp_clear_route_table) Sets up items onto a workqueue now, rather\n\t  than clearing each node directly. Plugs both process queues to\n\t  avoid potential race.\n\t  (bgp_static_withdraw_rsclient) let bgp_info_{add,delete} manage\n\t  bgp_info refcounts.\n\t  (bgp_static_update_rsclient) ditto\n\t  (bgp_static_update_main) ditto\n\t  (bgp_static_update_vpnv4) ditto, remove unneeded cast.\n\t  (bgp_static_withdraw) see bgp_static_withdraw_rsclient\n\t  (bgp_static_withdraw_vpnv4) ditto\n\t  (bgp_aggregate_{route,add,delete}) ditto\n\t  (bgp_redistribute_{add,delete,withdraw}) ditto\n\t* bgpd/bgp_vty.c: (peer_rsclient_set_vty) lock rsclient list peer\n\t  reference\n\t  (peer_rsclient_unset_vty) ditto, but unlock same reference\n\t* bgpd/bgpd.c: (peer_free) handle frees of info to be kept for lifetime\n\t  of struct peer.\n\t  (peer_lock,peer_unlock) peer refcount helpers\n\t  (peer_new) add initial refcounts\n\t  (peer_create,peer_create_accept) lock peer as appropriate\n\t  (peer_delete) unlock as appropriate, move out some free\u0027s to\n\t  peer_free.\n\t  (peer_group_bind,peer_group_unbind) peer refcounting as\n\t  appropriate.\n\t  (bgp_create) check CALLOC return value.\n\t  (bgp_terminate) free workqueues too.\n\t* lib/memtypes.c: Add MTYPE_BGP_PROCESS_QUEUE and\n\t  MTYPE_BGP_CLEAR_NODE_QUEUE\n"
    },
    {
      "commit": "00d252cb5ff21ecc537ab40385316b41622a95e2",
      "tree": "c0132f4be53d8ae251ecb1dd8c39816f8734d08f",
      "parents": [
        "269d74fdc39a612da8f627bf97628c68d25a16ab"
      ],
      "author": {
        "name": "paul",
        "email": "paul",
        "time": "Mon May 23 14:19:54 2005 +0000"
      },
      "committer": {
        "name": "paul",
        "email": "paul",
        "time": "Mon May 23 14:19:54 2005 +0000"
      },
      "message": "2005-05-23 Paul Jakma \u003cpaul@dishone.st\u003e\n\n\t* bgp_fsm.h: Add extern qualifier to exported functions\n\t* bgp_nexthop.c: add static to nexthop specific globals\n\t* *.h: Add guard defines\n"
    },
    {
      "commit": "fd79ac918b8feaacebe9719adaac97dffb69137a",
      "tree": "d0665eb68e60da9d6e364414cdb61830f19f33d3",
      "parents": [
        "39db97e4e02eae08a1e18528367b6e9b07eb6a93"
      ],
      "author": {
        "name": "paul",
        "email": "paul",
        "time": "Wed Oct 13 05:06:08 2004 +0000"
      },
      "committer": {
        "name": "paul",
        "email": "paul",
        "time": "Wed Oct 13 05:06:08 2004 +0000"
      },
      "message": "2004-10-13 Paul Jakma \u003cpaul@dishone.st\u003e\n\n\t* (global) more const\u0027ification and fixups of types to clean up code.\n\t* bgp_mplsvpn.{c,h}: (str2tag) fix abuse. Still not perfect,\n          should use something like the VTY_GET_INTEGER macro, but without\n          the vty_out bits..\n        * bgp_routemap.c: (set_aggregator_as) use VTY_GET_INTEGER_RANGE\n          (no_set_aggregator_as) ditto.\n        * bgpd.c: (peer_uptime) fix unlikely bug, where no buffer is\n          returned, add comments about troublesome return value.\n"
    },
    {
      "commit": "e0701b7955b883c5437269a382e1afc76ee71e5c",
      "tree": "d00034c1b7f369fc8214d401b7a44ef586d73e15",
      "parents": [
        "62843e433d8f85bce381edc6202c4af6c5cec31b"
      ],
      "author": {
        "name": "hasso",
        "email": "hasso",
        "time": "Thu May 20 09:19:34 2004 +0000"
      },
      "committer": {
        "name": "hasso",
        "email": "hasso",
        "time": "Thu May 20 09:19:34 2004 +0000"
      },
      "message": "Merge bgpd changeset 1176 from Zebra repository by Rivo Nurges.\n"
    },
    {
      "commit": "718e3744195351130f4ce7dbe0613f4b3e23df93",
      "tree": "bac2ad39971cd43f31241ef123bd4e470f695ac9",
      "parents": [],
      "author": {
        "name": "paul",
        "email": "paul",
        "time": "Fri Dec 13 20:15:29 2002 +0000"
      },
      "committer": {
        "name": "paul",
        "email": "paul",
        "time": "Fri Dec 13 20:15:29 2002 +0000"
      },
      "message": "Initial revision\n"
    }
  ]
}