[jira] [Commented] (TS-3896) Crash in OpenDir::close_write

["Alan M. Carroll (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TS-3896?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14732140#comment-14732140 ] 

Alan M. Carroll commented on TS-3896:
-------------------------------------

This was observed at Yahoo! and traced back to an unlocked mutex. The cache logic presumes it is locked and uses that to get an {{EThread}} pointer. Because the mutex is not locked that pointer is {{NULL}}.

> Crash in OpenDir::close_write
> -----------------------------
>
>                 Key: TS-3896
>                 URL: https://issues.apache.org/jira/browse/TS-3896
>             Project: Traffic Server
>          Issue Type: Bug
>          Components: Cache
>            Reporter: Alan M. Carroll
>
> gdb) bt full
> #0 0x00000000007269ca in OpenDir::close_write (this=0x2acc5c0171d0, cont=0x2acce09a6970) at CacheDir.cc:144
> h = 2872149697
> b = 193
> #1 0x0000000000724970 in Vol::close_write (this=0x2acc5c017010, cont=0x2acce09a6970) at P_CacheInternal.h:808
> No locals.
> #2 0x0000000000745c6e in CacheVC::openWriteCloseDir (this=0x2acce09a6970) at CacheWrite.cc:1081
> lock = {m =
> {m_ptr = 0x333e1c0}
> , lock_acquired = true}
> _func_ = "openWriteCloseDir"
> #3 0x0000000000746db0 in CacheVC::openWriteClose (this=0x2acce09a6970, event=0, e=0x0) at CacheWrite.cc:1289
> No locals.
> #4 0x00000000007248ab in CacheVC::die (this=0x2acce09a6970) at P_CacheInternal.h:734
> No locals.
> #5 0x000000000071694b in CacheVC::do_io_close (this=0x2acce09a6970, alerrno=0) at Cache.cc:376
> previous_closed = 0
> #6 0x0000000000519ebc in VConnection::do_io (this=0x2acce09a6970, op=4, c=0x0, nbytes=9223372036854775807, cb=0x0, data=0) at ../iocore/eventsystem/P_VConnection.h:105
> No locals.
> #7 0x00000000005fe7ee in HttpCacheSM::abort_write (this=0x2acccc6929c8) at HttpCacheSM.h:114
> No locals.
> #8 0x00000000005fe946 in HttpCacheSM::end_both (this=0x2acccc6929c8) at HttpCacheSM.h:142
> No locals.
> #9 0x00000000005f91fe in HttpSM::kill_this (this=0x2acccc690f90) at HttpSM.cc:6576
> _func_ = "kill_this"
> #10 0x00000000005eb8b3 in HttpSM::main_handler (this=0x2acccc690f90, event=3, data=0x2acd18008898) at HttpSM.cc:2631
> jump_point = (int (HttpSM::)(HttpSM *, int, void *)) 0x5e6012 <HttpSM::state_watch_for_client_abort(int, void)>
> _func_ = "main_handler"
> vc_entry = 0x2acccc692858
> #11 0x00000000005106f0 in Continuation::handleEvent (this=0x2acccc690f90, event=3, data=0x2acd18008898) at ../iocore/eventsystem/I_Continuation.h:145
> No locals.
> #12 0x0000000000776b53 in read_signal_and_update (event=3, vc=0x2acd18008780) at UnixNetVConnection.cc:141
> _func_ = "read_signal_and_update"
> #13 0x0000000000776e8e in read_signal_done (event=3, nh=0x2acb84a0d8a0, vc=0x2acd18008780) at UnixNetVConnection.cc:202
> No locals.
> #14 0x0000000000776f46 in read_signal_error (nh=0x2acb84a0d8a0, vc=0x2acd18008780, lerrno=0) at UnixNetVConnection.cc:226
> No locals.
> #15 0x0000000000779929 in UnixNetVConnection::readSignalError (this=0x2acd18008780, nh=0x2acb84a0d8a0, err=0) at UnixNetVConnection.cc:992
> No locals.
> #16 0x000000000075e529 in SSLNetVConnection::net_read_io (this=0x2acd18008780, nh=0x2acb84a0d8a0, lthread=0x2acb84a0a010) at SSLNetVConnection.cc:422
> ret = 10955
> bytes = 0
> s = 0x2acd18008890
> --Type <return> to continue, or q <return> to quit--
> _func_ = "net_read_io"
> buf = @0x2acb86e2cba0
> r = 0
> lock = {m =
> {m_ptr = 0x2acb86e2cb70}
> , lock_acquired = 198}
> ntodo = 7804412
> #17 0x00000000007706ec in NetHandler::mainNetEvent (this=0x2acb84a0d8a0, event=5, e=0x2b91630) at UnixNet.cc:546
> epd = 0x2acd180089a0
> pd = 0x2acb88c11010
> _func_ = "mainNetEvent"
> poll_timeout = 10
> vc = 0x2acd18008780
> #18 0x00000000005106f0 in Continuation::handleEvent (this=0x2acb84a0d8a0, event=5, data=0x2b91630) at ../iocore/eventsystem/I_Continuation.h:145
> No locals.
> #19 0x0000000000797442 in EThread::process_event (this=0x2acb84a0a010, e=0x2b91630, calling_code=5) at UnixEThread.cc:128
> c_temp = 0x2acb84a0d8a0
> lock = {m =
> {m_ptr = 0x2a3ed90}
> , lock_acquired = true}
> #20 0x000000000079794c in EThread::execute (this=0x2acb84a0a010) at UnixEThread.cc:252
> done_one = false
> e = 0x2b91630
> NegativeQueue = {<DLL<Event, Event::Link_link>> =
> {head = 0x0}
> , tail = 0x0}
> next_time = 1439415094080030928
> #21 0x00000000007969ed in spawn_thread_internal (a=0x2a5eef0) at Thread.cc:85
> p = 0x2a5eef0
> #22 0x00002acb7e9029d1 in start_thread () from /lib64/libpthread.so.0



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)