Re: bzr commit into mysql-next-mr-bugfixing branch (guilhem.bichot:3260)

[Davi Arnaut]

On 1/19/11 7:53 PM, Bjorn Munch wrote:
> On 19/01 21.14, Guilhem Bichot wrote:
>>
>> Maybe we could do a sql_print_warning() instead of DBUG_ASSERT(0); mtr
>> warns when it sees an unexpected warning, it marks the test as failed,
>> prints the warning on stderr, and returns 1 as exit code. I guess this
>> would be good enough?
>
> I'm not familiar with all the context here, but just let me add a
> comment from the perspective of mtr: yes, mtr will see an unexpected
> warning in the server log and report the test as failed, but it has no
> way of knowing what action from the test caused that warning. The test
> will have run to its end.

For those warnings we could generate a "dump trace" without crashing the 
server, it would require some minimal changes to handle_segfault. This 
would take care of printing the query, a stack trace, etc.

Regards,

Davi