MySQL Bugs: #4967: Disable innodb at startup, and then any innodb based query crashes the server

Bug #4967	Disable innodb at startup, and then any innodb based query crashes the server
Submitted:	9 Aug 2004 22:32	Modified:	20 Aug 2004 12:10
Reporter:	Nathaniel Blanchard	Email Updates:
Status:	Can't repeat	Impact on me:	None
Category:	MySQL Server: InnoDB storage engine	Severity:	S2 (Serious)
Version:	4.1.4	OS:	Linux (Linux)
Assigned to:		CPU Architecture:	Any

Description:
We wanted to have our application that requires Innodb test that the MySQL server it was connecting to supported InnoDB.  To do that, we created a table using the C API specifying the ENGINE = InnoDB at the end of the create table query.  When executed, this query crashes the server.  Interestingly enough, a 2nd execution of the same test program complains that the table already exists.  If I try and manually drop the table using the mysql client, it too crashes the server.

How to repeat:
    iResult = mysql_query( conn, "CREATE TABLE test (test1 INT) ENGINE=InnoDB" );

Execute that line while connected to a mysql server that has the skip-innodb flag set in its my.cnf.

Couldn't reproduce it so far with 4.1.3 and command line client,
tables are created as MyISAM when server started with --skip-innodb,
will try 4.1 bitkeeper sources and C API tomorrow ...

In case it helps, here is the crash log.

Number of processes running now: 0
040809 16:05:14 mysqld restarted
/usr/local/mysql/libexec/mysqld: ready for connections.
Version: '4.1.4-beta-log' socket: '/tmp/mysql.sock' port: 3306
mysqld got signal 11;
This could be because you hit a bug. It is also possible that this binary
or one of the libraries it was linked against is corrupt, improperly built,
or misconfigured. This error can also be caused by malfunctioning hardware.
We will try our best to scrape up some info that will hopefully help diagnose
the problem, but since we have already crashed, something is definitely wrong
and this may fail.

key_buffer_size=1048576
read_buffer_size=20967424
max_used_connections=1
max_connections=100
threads_connected=1
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_connections = 1131119 K
bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

thd=0x858a7e0
Attempting backtrace. You can use the following information to find out
where mysqld died. If you see no messages after this, something went
terribly wrong...
Cannot determine thread, fp=0xbf5fdc58, backtrace may not be correct.
Stack range sanity check OK, backtrace follows:
0x80b3247
0x8317306
0x83154d0
0x8214e07
0x8132777
0x8136583
0x812c47d
0x813a92a
0x8141701
0x813a12b
0x8139ba1
0x80c7c73
0x80c98c1
0x80c3d5c
0x80c37d6
0x80c2fbf
0x8314aaf
0x834b06a
New value of fp=(nil) failed sanity check, terminating stack trace!
Please read http://www.mysql.com/doc/en/Using_stack_trace.html and follow instructions on how to resolve the stack trace. Resolved
stack trace is much more helpful in diagnosing the problem, so please do
resolve it
Trying to get some variables.
Some pointers may be invalid and cause the dump to abort...
thd->query at 0x85a55e8 = drop database test3
thd->thread_id=1
The manual page at http://www.mysql.com/doc/en/Crashing.html contains
information that should help you find out what is causing the crash.

Also, lets assume that it doesn't crash.  Is defaulting to a MyISAM table type even when the user specifies a different type the correct action?  I would want/expect an error message to be returned.  Obviously, I the developer know what I am doing in saying I want an InnoDB table.  

What is the correct way to test that InnoDB is enabled?

Regarding your last comment:

Starting with version 4.1.1 you will get a warning if you try 
to create a table using a handler that is not enabled.

You can see (starting with 4.1.2) which table handlers are
available and which is the default using 

  SHOW STORAGE ENGINES;

  ( http://dev.mysql.com/doc/mysql/en/SHOW_ENGINES.html )

So four your purpose SHOW STORAGE ENGINES would
be the best solution. But if you plan to also support mysql 4.0.x
i'm afraid you have to create a table and use SHOW CREATE
TABLE to verify that it was really created as requested and
not with the default storage engine type.

Even with 4.1 bitkeeper source i cannot repeat the crash.

(test source file is attached)

test file that creates a table using C API

Attachment: create_test.c (text/plain), 1.53 KiB.