errors in debug builds when maintainer mode is on
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
| Percona Server moved to https://jira.percona.com/projects/PS |
Fix Released
|
High
|
Tomislav Plavcic | |
| 5.5 |
Fix Released
|
High
|
Tomislav Plavcic | |
| 5.6 |
Fix Released
|
High
|
Tomislav Plavcic |
Bug Description
After 5.6.22 and 5.5.41 merges mysql maintainer mode is turned on by default for debug builds on linux which currently reports some warnings as errors.
Those errors need to be resolved but as current workaround we will disable mysql maintainer mode in debug builds for releases.
Some of the errors seen:
PS 5.5 on ubuntu lucid:
13:34:33 /mnt/workspace/
13:34:33 /mnt/workspace/
13:34:33 /mnt/workspace/
PS 5.5 on ubuntu trusty:
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 for (i= (prec2 - 1) % DIG_PER_DEC1; *buf2 < powers10[i--]; prec2--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 for (i=(prec1-1) % DIG_PER_DEC1; *buf1 < powers10[i--]; prec1--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 for (i= (prec2 - 1) % DIG_PER_DEC1; *buf2 < powers10[i--]; prec2--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 for (i=(prec1-1) % DIG_PER_DEC1; *buf1 < powers10[i--]; prec1--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 *buf0 % powers10[i++] == 0;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 for (i= (intg - 1) % DIG_PER_DEC1; *buf0 < powers10[i--]; intg--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 for (i= (intg - 1) % DIG_PER_DEC1; *buf0 < powers10[i--]; intg--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 for (; *buf_beg < powers10[i--]; start++) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 for (; *buf_end % powers10[i++] == 0; stop--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 /mnt/workspace/
13:39:24 for (i= (intg - 1) % DIG_PER_DEC1; *buf0 < powers10[i--]; intg--) ;
13:39:24 ^
13:39:24 /mnt/workspace/
PS 5.6 on ubuntu trusty:
07:35:59 /mnt/workspace/
07:35:59 /mnt/workspace/
07:35:59 DBUG_ASSERT(index + PFS_SIZE_OF_A_TOKEN <= digest-
07:35:59 ^
07:35:59 /mnt/workspace/
07:35:59 DBUG_ASSERT(index + PFS_SIZE_OF_A_TOKEN <= digest-
07:35:59 ^
07:35:59 /mnt/workspace/
07:35:59 DBUG_ASSERT(index + PFS_SIZE_OF_A_TOKEN <= digest-
07:35:59 ^
07:35:59 /mnt/workspace/
07:35:59 DBUG_ASSERT(index + PFS_SIZE_OF_A_TOKEN <= digest-
07:35:59 ^
07:35:59 /mnt/workspace/
07:35:59 DBUG_ASSERT(index + PFS_SIZE_OF_A_TOKEN <= digest-
PS 5.6 on ubuntu precise:
07:36:28 /mnt/workspace/
07:36:28 /mnt/workspace/
07:36:28 if (!new_field)
07:36:28 ^
07:36:28 /mnt/workspace/
07:36:28 Create_field *new_field;
07:36:28 ^
07:36:28 /mnt/workspace/
07:36:28 if (!new_field)
07:36:28 ^
07:36:28 /mnt/workspace/
07:36:28 Create_field *new_field;
PS 5.6 on debian6-32:
07:37:58 /home/jenkins/
07:37:58 /home/jenkins/
PS 5.6 on ubuntu lucid:
07:36:41 /mnt/workspace/
07:36:41 /mnt/workspace/
Changed in percona-server: | |
status: | New → Triaged |
Tomislav Plavcic (tplavcic) wrote : | #2 |
Can't seem to find similar report on bugs.mysql.com, but need more time since I think maybe the errors depend on our other debug build flags. So I'll create report if needed after further investigation.
07:37:58 /home/jenkins/
07:37:58 /home/jenkins/
is an actual PS bug:
int
mysql_execute_
{
...
struct system_variables *per_query_
...
if (this or that) goto error;
...
if (lex->set_statement && !lex->var_
per_
...
error:
...
if (lex->set_statement && !lex->var_
...
free_
thd->variables= *per_query_
my_
...
}
So, in case of "this or that" we end up with corrupted thd->variables going forward, and this might be the cause of numerous SET STATEMENT crashes: bug 1387951, bug 1417028, bug 1416975, bug 1416956, etc, etc, etc, see the set-statement bug tag.
All the relevant SET STATEMENT bugs were collected into bug 1418049, bug 1387951, and bug 1412423, each corresponding to one goto error in the above.
Shahriyar Rzayev (rzayev-sehriyar) wrote : | #5 |
Percona now uses JIRA for bug reports so this bug report is migrated to: https:/
The majority of these warnings are not Percona Server-specific, is there a bugs.mysql.com report or reports? Can you file one if there isn't?