MySQL Bugs: #31089: InnoDB unsafe binlog test fails with wrong error

Bug #31089	InnoDB unsafe binlog test fails with wrong error
Submitted:	18 Sep 2007 21:20	Modified:	19 Sep 2007 1:04
Reporter:	Joshua Kugler (Basic Quality Contributor)	Email Updates:
Status:	Not a Bug	Impact on me:	None
Category:	MySQL Server: Compiling	Severity:	S3 (Non-critical)
Version:	5.0.45	OS:	Linux (Debian Sarge)
Assigned to:	Heikki Tuuri	CPU Architecture:	Any

Description:
When running 'make test' I get this:

Errors are (from /home/jkugler/src/mysql-5.0.45/mysql-test/var/log/mysqltest-time) :
mysqltest: At line 59: query 'select * from t1 where a = 2 and b = 2 for update' succeeded - should have failed with errno 1205...
(the last lines may be the most important ones)
Result from queries before failure can be found in r/innodb_unsafe_binlog.log

The log contains:

drop table if exists t1,t2;
create table t1 (id int not null, f_id int not null, f int not null,
primary key(f_id, id)) engine=innodb;
create table t2 (id int not null,s_id int not null,s varchar(200),
primary key(id)) engine=innodb;
INSERT INTO t1 VALUES (8, 1, 3);
INSERT INTO t1 VALUES (1, 2, 1);
INSERT INTO t2 VALUES (1, 0, '');
INSERT INTO t2 VALUES (8, 1, '');
commit;
DELETE ml.* FROM t1 AS ml LEFT JOIN t2 AS mm ON (mm.id=ml.id)
WHERE mm.id IS NULL;
select ml.* from t1 as ml left join t2 as mm on (mm.id=ml.id)
where mm.id is null lock in share mode;
id      f_id    f
drop table t1,t2;
create table t1 (id int not null, f_id int not null, f int not null,
primary key(id),key(f_id)) engine=innodb;
create table t2 (id int not null,s_id int not null,s varchar(200),
primary key(id),key(s_id)) engine=innodb;
INSERT INTO t1 VALUES (8, 1, 3);
INSERT INTO t1 VALUES (1, 2, 1);
INSERT INTO t2 VALUES (1, 0, '');
INSERT INTO t2 VALUES (8, 1, '');
commit;
delete ml.* from t1 as ml left join t2 as mm on (mm.s_id=ml.f_id) where mm.s is null;
select ml.* from t1 as ml left join t2 as mm on (mm.s_id=ml.f_id) where mm.s is null lock in share mode;
id      f_id    f
drop table t1,t2;
create table t1(a int not null, b int, primary key(a)) engine=innodb;
insert into t1 values(1,1),(2,2),(3,1),(4,2),(5,1),(6,2);
commit;
set autocommit = 0;
select * from t1 lock in share mode;
a       b
1       1
2       2
3       1
4       2
5       1
6       2
update t1 set b = 5 where b = 1;
set autocommit = 0;
select * from t1 where a = 2 and b = 2 for update;
a       b
2       2

My configure spec:

CFLAGS="-O3 -DPIC -fPIC -DUNDEF_HAVE_INITGROUPS -fno-strict-aliasing" CXXFLAGS="-O3 -fno-strict-aliasing -felide-constructors -fno-exceptions -fno-rtti -fPIC -DPIC -DUNDEF_HAVE_INITGROUPS" ./configure --prefix=/usr/local/ --enable-thread-safe-client --enable-assember --with-server-suffix=eei --with-unix-socket-path=/usr/local/tmp/mysql5.sock --with-tcp-port=3307 --with-mysqld-user=mysql5 --without-bench --without-debug --with-example-storage-engine --with-archive-storage-engine --with-csv-storage-engine --with-blackhole-storage-engine --with-federated-storage-engine --with-extra-charsets=all

How to repeat:
Use the above configure spec and make;make test.

Corrected Debian version

Upon further investigation, all tests pass if -O0 (no optimization) is used.  It seems optimization is breaking something.

$ gcc -v
Reading specs from /usr/lib/gcc-lib/i486-linux/3.3.5/specs
Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,pascal,objc,ada,treelang --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-gxx-include-dir=/usr/include/c++/3.3 --enable-shared --enable-__cxa_atexit --with-system-zlib --enable-nls --without-included-gettext --enable-clocale=gnu --enable-debug --enable-java-gc=boehm --enable-java-awt=xlib --enable-objc-gc i486-linux
Thread model: posix
gcc version 3.3.5 (Debian 1:3.3.5-13)

Hmm, it turns out this seems to be a gcc related bug.  Compiling on Ubuntu Dapper (gcc 4.0.3) with -O3 produces a binary that passes all tests.

I'm closing the bug as "not a bug." Sorry for the noise.

Maybe the test is timing-dependent.