[MCOL-1722] Regression in semi_join Created: 2018-09-14  Updated: 2019-01-02  Resolved: 2019-01-02

Status: Closed
Project: MariaDB ColumnStore
Component/s: None
Affects Version/s: 1.1.7
Fix Version/s: 1.1.7

Type: Bug Priority: Critical
Reporter: Andrew Hutchings (Inactive) Assignee: Daniel Lee (Inactive)
Resolution: Fixed Votes: 0
Labels: None

Issue Links:
Problem/Incident
is caused by MCOL-1188 assertion 'fColumn.get() && fSub && f... Closed
Sprint: 2018-17, 2018-18, 2018-19, 2018-20, 2018-21

 Description   

There is a regression found in g3_simple_semi_join.sql

Will analyse and add more details in the comments

 Comments   
Comment by Andrew Hutchings (Inactive) [ 2018-09-14 ]

Diff of failure:

--- g3_simple_semi_join.sql.ref.log	2016-08-16 11:18:48.784463914 +0100
 
+++ g3_simple_semi_join.sql.log	2018-09-14 10:39:41.198690560 +0100
 
@@ -775,4 +775,4 @@
 
 count(o_orderkey)	sum(o_totalprice)
 
 1500000	226829306447.46
 
 count(o_orderkey)	sum(o_totalprice)
 
-1500000	226829306447.46
 
+0	NULL

Git bisect shows regression caused by:

58bc867ca17b85d2447fe25835787d58bea2c73c is the first bad commit
 
commit 58bc867ca17b85d2447fe25835787d58bea2c73c
 
Author: Ravi Prakash <ravi.prakash@mariadb.com>
 
Date:   Thu Aug 30 20:24:24 2018 -0700
 
 
 
    MCOL-1188 assertion 'fColumn.get() && fSub && fFunc' failed, ... MySQL server crashed.
 
    
    The problem was in processing a subquery in the where clause that was categorized as a
 
    CACHE_ITEM in the parse tree. The fix involved how we walk the parse tree in gp_walk()

Comment by Andrew Hutchings (Inactive) [ 2018-09-18 ]

Problem was due to a bad test result file. The new behaviour is now correct. Ravi's code is now in develop-1.1 of the regression test.

For QA: regression suite should now pass

Comment by Daniel Lee (Inactive) [ 2019-01-02 ]

Build verified: 1.1.7-1

Git Commits
server=8220579
engine=88da1b6

Regression passed

Generated at Thu Feb 08 02:30:53 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.