Hi!
>>>>> "Sasha" == Sasha Pachev <sasha@stripped> writes:
Sasha> On Friday 02 February 2001 10:34, Chris Richmond MD6-FDC wrote:
>> Hi Folks,
>>
>> I've downloaded, compiled, and insatlled mysql-3.23.32.
>> I'm running on a Sparc Ultra10, Solaris 2.5.1, using gcc
>> gcc version cygnus-2.7-96q4.
>>
>> I'm trying to run the test suite and I had to disable dbd
>> since in I didn't include BerkleyDB, and innobase
>> since its also not available.
>>
>> The test that is failing is merge. The sdiff of the results
>> are here:
>>
>> host:/eng/db/MySQL/mysql-test/r> sdiff merge.reject merge.result
>> a b a b
>> 1 Testing 1
Sasha> Testing
>> 2 table 2 table
>> 3 t1 3 t1
>> 1 Testing 1
Sasha> Testing
>> 2 table 2 table
>> 3 t2 3 t2
>> a b a b
>> 1 Testing | 3 t1
>> 1 Testing | 3 t2
>> 1 Testing | 2 table
>> 1 Testing | 2 table
>> 1 Testing 1
Sasha> Testing
>> 1 Testing 1
Sasha> Testing
>>
>> The rest is identical. How do I interpret these results? Is something
>> broken, or should I skip this test also?
Sasha> This means there is either a bug in the test case, or in MySQL merge itself.
Sasha> Do not use merge feature until Monty and/or Serg resolve this issue. If other
Sasha> tests passed, and you are not using merge tables, you can use the binary you
Sasha> have built.
I did a check of this on our Solaris 2.7 sparc machine with
shell> gcc -v
Reading specs from /my/gnu/lib/gcc-lib/sparc-sun-solaris2.7/2.95/specs
and for me merge worked perfectly.
Could you try to either upgrade gcc to the above or configure mysqld
with --debug and run
cd mysql-test
mysql-test-run --local --debug merge
and mail me the file 'var/tmp/master.trace'
Regards.
Monty
