| Bug #59941 | cluster/j - columns with varbinary(>255) fails | ||
|---|---|---|---|
| Submitted: | 4 Feb 2011 10:50 | Modified: | 31 Mar 2011 11:10 |
| Reporter: | Johan Andersson | Email Updates: | |
| Status: | Closed | Impact on me: | |
| Category: | MySQL Cluster: Cluster/J | Severity: | S3 (Non-critical) |
| Version: | 7.1.9a | OS: | Any |
| Assigned to: | Craig Russell | CPU Architecture: | Any |
| Tags: | cluster, ClusterJ | ||
[4 Feb 2011 22:18]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/130484 4083 Craig L Russell 2011-02-04 Bug #59941 VM crash invalid memory access with varbinary(256) column clusterj was incorrectly asking for the charset for a longvarbinary column.
[4 Feb 2011 23:54]
Craig Russell
The important part of the patch is the file storage/ndb/clusterj/clusterj-tie/src/main/java/com/mysql/clusterj/tie/ColumnImpl.java The others are test files.
[31 Mar 2011 11:10]
Jon Stephens
Documented fix in the NDB-7.1.12 changelog as follows:
ClusterJ incorrectly asked for the character set for a LONGVARBINARY
column with VARBINARY(256) columns, leading to a VM crash caused by
invalid memory access.
Closed.
Description: When having a column with VARBINARY(256) (or larger) the following happens: XXX xxx=s.newInstance(XXX.class); [2/4/11 11:06:37 AM] Johan Andersson: Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 5 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 8 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 15 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 31 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 47 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 48 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 49 Feb 4, 2011 11:05:01 AM com.mysql.clusterj.tie.Utility addCollation INFO: Adding charset converter latin1 for collation 94 Invalid memory access of location 0x0 rip=0x11a01184b How to repeat: create table t1(a integer primary key, b varbinary(256))engine=ndb; ... myT1 m1=s.newInstance(myT1.class); gives the above myT1.java import com.mysql.clusterj.annotation.PersistenceCapable; import com.mysql.clusterj.annotation.PrimaryKey; @PersistenceCapable(table="t1") public interface myT1 { @PrimaryKey int getA(); void setA(int id); byte [] getB(); void setB(byte [] data); } Main.java import java.util.Properties; import com.mysql.clusterj.*; public class Main { public static void main(String[] args) { // TODO Auto-generated method stub try { Properties p = new Properties(); p.setProperty("com.mysql.clusterj.connectstring", "localhost:1186"); p.setProperty("com.mysql.clusterj.database", "test"); SessionFactory sessionFactory=ClusterJHelper.getSessionFactory(p); Session s = sessionFactory.getSession(); System.err.println("here"); myT1 m1=s.newInstance(myT1.class); } catch(Exception e) { e.printStackTrace(); } } }