[
https://issues.apache.org/jira/browse/HDFS-4387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13551665#comment-13551665
]
Colin Patrick McCabe commented on HDFS-4387:
--------------------------------------------
even after making this change, JamVM crashes in nmdShutdown. This is a simple
function that does nothing but call {{MiniDFSCluster#shutdown}}.
{code}
2013-01-11 15:46:06,003 INFO hdfs.MiniDFSCluster
(MiniDFSCluster.java:shutdown(1411)) - Shutting down the Mini HDFS Cluster
2013-01-11 15:46:06,004 INFO hdfs.MiniDFSCluster
(MiniDFSCluster.java:shutdownDataNodes(1438)) - Shutting down DataNode 0
2013-01-11 15:46:06,004 WARN datanode.DirectoryScanner
(DirectoryScanner.java:shutdown(289)) - DirectoryScanner: shutdown has been
called
2013-01-11 15:46:06,015 INFO mortbay.log (Slf4jLog.java:info(67)) - Stopped
SelectChannelConnector@localhost:0
2013-01-11 15:46:06,016 INFO ipc.Server (Server.java:stop(2096)) - Stopping
server on 39373
2013-01-11 15:46:06,018 INFO ipc.Server (Server.java:run(829)) - Stopping IPC
Server Responder
2013-01-11 15:46:06,018 INFO ipc.Server (Server.java:run(687)) - Stopping IPC
Server listener on 39373
Segmentation fault (core dumped)
{code}
I really have no idea how to fix this, or even if it can be fixed. It looks
like something that shutdown itself is doing, rather than something in the C
code. Perhaps JamVM can't handle threads shutting down or the log source being
closed?
> test_libhdfs_threaded SEGV on OpenJDK 7
> ---------------------------------------
>
> Key: HDFS-4387
> URL: https://issues.apache.org/jira/browse/HDFS-4387
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: libhdfs
> Affects Versions: 3.0.0
> Reporter: Andy Isaacson
> Priority: Minor
>
> Building and running tests on OpenJDK 7 on Ubuntu 12.10 fails with {{mvn test
> -Pnative}}. The output is hard to decipher but the underlying issue is that
> {{test_libhdfs_native}} segfaults at startup.
> {noformat}
> (gdb) run
> Starting program:
> /mnt/trunk/hadoop-hdfs-project/hadoop-hdfs/target/native/test_libhdfs_threaded
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
> Program received signal SIGSEGV, Segmentation fault.
> 0x00007ffff739a897 in attachJNIThread (name=0x0, is_daemon=is_daemon@entry=0
> '\000', group=0x0) at thread.c:768
> 768 thread.c: No such file or directory.
> (gdb) where
> #0 0x00007ffff739a897 in attachJNIThread (name=0x0,
> is_daemon=is_daemon@entry=0 '\000', group=0x0) at thread.c:768
> #1 0x00007ffff7395020 in attachCurrentThread (is_daemon=0, args=0x0,
> penv=0x7fffffffddb8) at jni.c:1454
> #2 Jam_AttachCurrentThread (vm=<optimized out>, penv=0x7fffffffddb8,
> args=0x0) at jni.c:1466
> #3 0x00007ffff7bcf979 in getGlobalJNIEnv () at
> /mnt/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/native/libhdfs/jni_helper.c:527
> #4 getJNIEnv () at
> /mnt/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/native/libhdfs/jni_helper.c:585
> #5 0x0000000000402512 in nmdCreate (conf=conf@entry=0x7fffffffdeb0) at
> /mnt/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/native/libhdfs/native_mini_dfs.c:49
> #6 0x00000000004016e1 in main () at
> /mnt/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/native/libhdfs/test_libhdfs_threaded.c:283
> {noformat}
--
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
