You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@trafficserver.apache.org by "weijin (JIRA)" <ji...@apache.org> on 2012/10/17 07:56:03 UTC

[jira] [Created] (TS-1533) make read_from_writer works in 3.0.x

weijin created TS-1533:
--------------------------

             Summary: make read_from_writer works in 3.0.x
                 Key: TS-1533
                 URL: https://issues.apache.org/jira/browse/TS-1533
             Project: Traffic Server
          Issue Type: Bug
          Components: Cache
    Affects Versions: 3.0.5
         Environment: all
            Reporter: weijin
            Assignee: weijin
             Fix For: 3.0.5


similar with TS-1433, this patch try to fix the bug that read_from_writer always failed even if we enable read_from_writer, and it also fix a race condition that may lead crash (to lock the write_vc after it was freed). need reviews.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (TS-1533) make read_from_writer works in 3.0.x

Posted by "weijin (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/TS-1533?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13477641#comment-13477641 ] 

weijin commented on TS-1533:
----------------------------

{code}
(gdb) bt
#0  Mutex_trylock (this=0x2ac3e263dca0, am=0x0, t=0x2ac3e1c35010) at ../iocore/eventsystem/I_Lock.h:387
#1  MutexTryLock::MutexTryLock (this=0x2ac3e263dca0, am=0x0, t=0x2ac3e1c35010) at ../iocore/eventsystem/I_Lock.h:607
#2  0x000000000061e4f4 in CacheVC::openReadFromWriter (this=0x362f7b0, event=2, e=0x2ac5b0006260) at CacheRead.cc:370
#3  0x0000000000660e44 in handleEvent (this=0x2ac3e1c35010, e=0x2ac5b0006260, calling_code=2) at I_Continuation.h:146
#4  EThread::process_event (this=0x2ac3e1c35010, e=0x2ac5b0006260, calling_code=2) at UnixEThread.cc:141
#5  0x0000000000661943 in EThread::execute (this=0x2ac3e1c35010) at UnixEThread.cc:218
#6  0x00000000006600d2 in spawn_thread_internal (a=0x21e08e0) at Thread.cc:88
#7  0x000000394ea077f1 in start_thread () from /lib64/libpthread.so.0
#8  0x000000394e6e570d in clone () from /lib64/libc.so.6
(gdb) f 2
#2  0x000000000061e4f4 in CacheVC::openReadFromWriter (this=0x362f7b0, event=2, e=0x2ac5b0006260) at CacheRead.cc:370
370	  CACHE_TRY_LOCK(writer_lock, write_vc->mutex, mutex->thread_holding);
(gdb) p write_vc->mutex 
$2 = {m_ptr = 0x0}
{code}
                
> make read_from_writer works in 3.0.x
> ------------------------------------
>
>                 Key: TS-1533
>                 URL: https://issues.apache.org/jira/browse/TS-1533
>             Project: Traffic Server
>          Issue Type: Bug
>          Components: Cache
>    Affects Versions: 3.0.5
>         Environment: all
>            Reporter: weijin
>            Assignee: weijin
>             Fix For: 3.0.5
>
>
> similar with TS-1433, this patch try to fix the bug that read_from_writer always failed even if we enable read_from_writer, and it also fix a race condition that may lead crash (to lock the write_vc after it was freed). need reviews.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira