| Bug #35877 | Update .. WHERE with function, constraint violation, crash | ||
|---|---|---|---|
| Submitted: | 7 Apr 2008 14:11 | Modified: | 12 Nov 21:54 |
| Reporter: | Matthias Leich | ||
| Status: | Closed | ||
| Category: | Server: SP | Severity: | S3 (Non-critical) |
| Version: | 5.1 | OS: | Any |
| Assigned to: | Jon Olav Hauglid | Target Version: | |
| Triage: | Triaged: D1 (Critical) | ||
[7 Apr 2008 14:11]
Matthias Leich
[7 Apr 2008 14:15]
Matthias Leich
testscript
Attachment: ml002.test (application/octet-stream, text), 993 bytes.
[7 Apr 2008 14:16]
Matthias Leich
Stack back trace
Attachment: stacktrace (application/octet-stream, text), 4.05 KiB.
[7 Apr 2008 14:45]
Matthias Leich
IMHO using a function within a WHERE qualification of
an UPDATE + the function itself modifies data is a very
rare case. The only case I can in the moment imagine
is that somebody has a DWH application. The updated table
is used for collecting some temporary results ("used on the
way to get the final report") and the function is
- much more complicated
- has to serve two purposes
- bring some aggregate tables up to date
- process aggregate tables for getting the return value
--> I4
I also guess that many users could find a different
solution - the function within the WHERE qualification-
does not modify data - which fits to their application
needs. --> W3
[9 Apr 2008 22:50]
Konstantin Osipov
Backtrace:
(gdb) bt
#0 0xffffe410 in __kernel_vsyscall ()
#1 0xb7d56875 in raise () from /lib/tls/i686/cmov/libc.so.6
#2 0xb7d58201 in abort () from /lib/tls/i686/cmov/libc.so.6
#3 0xb7d4fb6e in __assert_fail () from /lib/tls/i686/cmov/libc.so.6
#4 0x08246815 in Diagnostics_area::set_ok_status (this=0x8dd6678,
thd=0x8dd58e0, affected_rows_arg=39560943763462, last_insert_id_arg=0,
message_arg=0xb50de1f8 "Rows matched: 0 Changed: 0 Warnings: 0")
at sql_class.cc:397
#5 0x082fa9da in mysql_update (thd=0x8dd58e0, table_list=0x8e217d0,
fields=@0x8dd6cdc, values=@0x8dd6f28, conds=0x8e1a410, order_num=0,
order=0x0, limit=13046331721333608305, handle_duplicates=DUP_ERROR,
ignore=false) at sql_class.h:2146
#6 0x0826e719 in mysql_execute_command (thd=0x8dd58e0) at sql_parse.cc:2827
#7 0x0827479c in mysql_parse (thd=0x8dd58e0,
inBuf=0x8e207b0 "UPDATE t1_aux SET f2 = 2 WHERE f1 = f1_two_inserts()",
length=52, found_semicolon=0xb50deeec) at sql_parse.cc:5617
#8 0x08275b49 in dispatch_command (command=COM_QUERY, thd=0x8dd58e0,
packet=0x8e0e0f1 "UPDATE t1_aux SET f2 = 2 WHERE f1 = f1_two_inserts()",
packet_length=52) at sql_parse.cc:1121
#9 0x08276cf2 in do_command (thd=0x8dd58e0) at sql_parse.cc:781
#10 0x08265098 in handle_one_connection (arg=0x8dd58e0) at sql_connect.cc:1106
#11 0xb7eeb46b in start_thread () from /lib/tls/i686/cmov/libpthread.so.0
#12 0xb7dff6de in clone () from /lib/tls/i686/cmov/libc.so.6
[9 Apr 2008 22:51]
Konstantin Osipov
The assert was added as part of Bug#12713
[8 Aug 2008 19:08]
Evgeny Potemkin
Analysis of this bug: The f1_two_inserts function tries to insert NULL into a not null column and fails. This error isn't correctly propagated to the caller statement and causes failed assertion at setting return code of whole statement for the second time. See also bug#27563.
[15 May 11:49]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/74172 2777 Jon Olav Hauglid 2009-05-15 Bug #35877 Update .. WHERE with function, constraint violation, crash Unable to reproduce crash with current version of mysql-6.0-runtime. Test case for MyISAM/InnoDB based on the bug rapport added to sp.test.
[27 May 16:23]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/75058 2781 Jon Olav Hauglid 2009-05-27 Bug #35877 Update .. WHERE with function, constraint violation, crash Unable to reproduce crash with current version of mysql-6.0-runtime. Test case for MyISAM/InnoDB based on the bug rapport added to sp.test.
[29 Jul 10:28]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/79502 2833 Jon Olav Hauglid 2009-07-29 Bug #35877 Update .. WHERE with function, constraint violation, crash Unable to reproduce crash with current version of mysql-azalea. Test case for MyISAM/InnoDB based on the bug rapport added to sp_trans.test.
[29 Jul 10:31]
Jon Olav Hauglid
Pushed to azalea-bugfixing.
[4 Aug 21:51]
Bugs System
Pushed into 5.4.4-alpha (revid:alik@sun.com-20090804194615-h40sa098mx4z49qg) (version source revid:iggy@mysql.com-20090731204544-7nio1afvg0dmzs7g) (merge vers: 5.4.4-alpha) (pib:11)
[7 Aug 1:56]
Paul DuBois
Bug no longer present apparently; changes to test case only. No changelog entry needed.
[12 Oct 13:41]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/86554 2910 Jon Olav Hauglid 2009-10-12 Bug #35877 Update .. WHERE with function, constraint violation, crash Unable to reproduce crash with current version of the 5.5.0 codebase. Test case for MyISAM/InnoDB based on the bug rapport added to sp_trans.test. Backport of revno: 2617.65.9.
[12 Oct 13:43]
Jon Olav Hauglid
Pushed to mysql-next-mr-runtime (version 5.5.0).
[3 Nov 8:17]
Bugs System
Pushed into 6.0.14-alpha (revid:alik@sun.com-20091102151658-j9o4wgro47m5v84d) (version source revid:alik@sun.com-20091023064702-2f8jdmny61bdl94u) (merge vers: 6.0.14-alpha) (pib:13)
[3 Nov 16:39]
Paul DuBois
Changes to test case only. No changelog entry needed.
[12 Nov 9:20]
Bugs System
Pushed into 5.5.0-beta (revid:alik@sun.com-20091110093229-0bh5hix780cyeicl) (version source revid:mikael@mysql.com-20091103113702-p61dlwc6ml6fxg18) (merge vers: 5.5.0-beta) (pib:13)
[12 Nov 21:54]
Paul DuBois
Changes to test case only. No changelog entry needed.
