Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Won't Fix
-
10.1(EOL)
-
None
Description
Started happening recently. Maybe it's related to sid upgrade (gcc 6 and such).
http://buildbot.askmonty.org/buildbot/builders/kvm-deb-sid-x86/builds/2068/steps/test_5/logs/stdio
encryption.innodb_first_page 'cbc,xtradb' w1 [ fail ] Found warnings/errors in server log file!
|
Test ended at 2016-08-16 07:48:02
|
line
|
160816 7:47:58 [ERROR] mysqld got signal 11 ;
|
Attempting backtrace. You can use the following information to find out
|
^ Found warnings in /dev/shm/var/1/log/mysqld.1.err
|
ok
|
|
|
- skipping '/dev/shm/var/1/log/encryption.innodb_first_page-cbc,xtradb/'
|
|
|
Retrying test encryption.innodb_first_page, attempt(2/3)...
|
Attachments
Issue Links
- relates to
-
MDEV-7069 Fix buildbot failures in main server trees
-
- Closed
-
Activity
It only happens on sid.
Maybe cvicentiu knows something about it? (in case it's related to gcc 6).
The only thing that I know for certain about gcc 6 is that they always assume "this" pointer is not null. Any if conditions that check for this are automatically set to true.
We previously had something like:
bool SomeClass::method() { |
if (!this) return true; |
return false; |
}
|
|
|
SomeClass *p = NULL;
|
p->method(); // Before GCC v6 this returned true all the time. Now it returns false all the time. |
There are some places in the code where we do this. I plan to fix them, but strapped for time till the dev meeting.
Failed to repeat this on both my laptop or perro (used --repeat=60 --parallel=4 few times). Need a stack trace if fails again.