[MDEV-18738] ASAN heap-use-after-free in copy_if_not_alloced / copy_fields Created: 2019-02-26  Updated: 2019-04-30  Resolved: 2019-04-30

Status: Closed
Project: MariaDB Server
Component/s: Data types, Server
Affects Version/s: 10.1, 10.2, 10.3, 10.4
Fix Version/s: 10.2.24, 10.1.39, 10.3.15, 10.4.5

Type: Bug Priority: Major
Reporter: Elena Stepanova Assignee: Alexander Barkov
Resolution: Fixed Votes: 0
Labels: None

Issue Links:
Duplicate
is duplicated by MDEV-19359 ASAN heap-use-after-free in copy_if_n... Closed
Relates
relates to MDEV-19359 ASAN heap-use-after-free in copy_if_n... Closed

 Description    

SELECT REPLACE( CAST( CURDATE() AS BINARY ), CURDATE(), REPEAT('a',32) ) AS f FROM mysql.user GROUP BY f;

10.1 28cb0417

 
==3123==ERROR: AddressSanitizer: heap-use-after-free on address 0x60e000075750 at pc 0x55e95f17c4e7 bp 0x7f930d1b0a80 sp 0x7f930d1b0a78
 
READ of size 10 at 0x60e000075750 thread T6
 
    #0 0x55e95f17c4e6 in copy_if_not_alloced(String*, String*, unsigned int) /data/src/10.1/sql/sql_string.cc:915
 
    #1 0x55e95f68d9f5 in Item_func_replace::val_str(String*) /data/src/10.1/sql/item_strfunc.cc:1197
 
    #2 0x55e95f536641 in Item_copy_string::copy() /data/src/10.1/sql/item.cc:4063
 
    #3 0x55e95f0c1fdb in copy_fields(TMP_TABLE_PARAM*) /data/src/10.1/sql/sql_select.cc:23163
 
    #4 0x55e95f0d0115 in end_send_group(JOIN*, st_join_table*, bool) /data/src/10.1/sql/sql_select.cc:19816
 
    #5 0x55e95f0657b2 in evaluate_join_record /data/src/10.1/sql/sql_select.cc:18702
 
    #6 0x55e95f0813fd in sub_select(JOIN*, st_join_table*, bool) /data/src/10.1/sql/sql_select.cc:18479
 
    #7 0x55e95f0ba9d0 in do_select /data/src/10.1/sql/sql_select.cc:18134
 
    #8 0x55e95f0ff54e in JOIN::exec_inner() /data/src/10.1/sql/sql_select.cc:3275
 
    #9 0x55e95f100836 in JOIN::exec() /data/src/10.1/sql/sql_select.cc:2562
 
    #10 0x55e95f0f5089 in mysql_select(THD*, Item***, TABLE_LIST*, unsigned int, List<Item>&, Item*, unsigned int, st_order*, st_order*, Item*, st_order*, unsigned long long, select_result*, st_select_lex_unit*, st_select_lex*) /data/src/10.1/sql/sql_select.cc:3499
 
    #11 0x55e95f0f5877 in handle_select(THD*, LEX*, select_result*, unsigned long) /data/src/10.1/sql/sql_select.cc:388
 
    #12 0x55e95efc4a59 in execute_sqlcom_select /data/src/10.1/sql/sql_parse.cc:5950
 
    #13 0x55e95efdec69 in mysql_execute_command(THD*) /data/src/10.1/sql/sql_parse.cc:2995
 
    #14 0x55e95eff6e7a in mysql_parse(THD*, char*, unsigned int, Parser_state*) /data/src/10.1/sql/sql_parse.cc:7468
 
    #15 0x55e95effdc5e in dispatch_command(enum_server_command, THD*, char*, unsigned int) /data/src/10.1/sql/sql_parse.cc:1496
 
    #16 0x55e95f0043ea in do_command(THD*) /data/src/10.1/sql/sql_parse.cc:1124
 
    #17 0x55e95f2ac8bf in do_handle_one_connection(THD*) /data/src/10.1/sql/sql_connect.cc:1330
 
    #18 0x55e95f2acdd0 in handle_one_connection /data/src/10.1/sql/sql_connect.cc:1242
 
    #19 0x55e95fb7bccc in pfs_spawn_thread /data/src/10.1/storage/perfschema/pfs.cc:1861
 
    #20 0x7f93198e7493 in start_thread (/lib/x86_64-linux-gnu/libpthread.so.0+0x7493)
 
    #21 0x7f9317ca093e in __clone (/lib/x86_64-linux-gnu/libc.so.6+0xe893e)
 
 
 
0x60e000075750 is located 112 bytes inside of 148-byte region [0x60e0000756e0,0x60e000075774)
 
freed by thread T6 here:
 
    #0 0x7f9319b51527 in __interceptor_free (/usr/lib/x86_64-linux-gnu/libasan.so.1+0x54527)
 
    #1 0x55e9603480e1 in free_memory /data/src/10.1/mysys/safemalloc.c:276
 
 
 
previously allocated by thread T6 here:
 
    #0 0x7f9319b5173f in malloc (/usr/lib/x86_64-linux-gnu/libasan.so.1+0x5473f)
 
    #1 0x55e9603481b2 in sf_malloc /data/src/10.1/mysys/safemalloc.c:115
 
    #2 0x55e960440c1a (/data/bld/10.1-asan/bin/mysqld+0x1d9ec1a)
 
 
 
Thread T6 created by T0 here:
 
    #0 0x7f9319b20bba in pthread_create (/usr/lib/x86_64-linux-gnu/libasan.so.1+0x23bba)
 
    #1 0x55e95fb87523 in spawn_thread_v1 /data/src/10.1/storage/perfschema/pfs.cc:1911
 
 
 
SUMMARY: AddressSanitizer: heap-use-after-free /data/src/10.1/sql/sql_string.cc:915 copy_if_not_alloced(String*, String*, unsigned int)
 
Shadow bytes around the buggy address:
 
  0x0c1c80006a90: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
 
  0x0c1c80006aa0: fa fa fa fa 00 00 00 00 00 00 00 00 00 00 00 00
 
  0x0c1c80006ab0: 00 00 00 00 00 00 00 04 fa fa fa fa fa fa fa fa
 
  0x0c1c80006ac0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 
  0x0c1c80006ad0: 00 00 04 fa fa fa fa fa fa fa fa fa fd fd fd fd
 
=>0x0c1c80006ae0: fd fd fd fd fd fd fd fd fd fd[fd]fd fd fd fd fa
 
  0x0c1c80006af0: fa fa fa fa fa fa fa fa 00 00 00 00 00 00 00 00
 
  0x0c1c80006b00: 00 00 00 00 00 00 00 00 00 00 02 04 fa fa fa fa
 
  0x0c1c80006b10: fa fa fa fa 00 00 00 00 00 00 00 00 00 00 00 00
 
  0x0c1c80006b20: 00 00 00 00 00 01 04 fa fa fa fa fa fa fa fa fa
 
  0x0c1c80006b30: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 
Shadow byte legend (one shadow byte represents 8 application bytes):
 
  Addressable:           00
 
  Partially addressable: 01 02 03 04 05 06 07 
  Heap left redzone:       fa
 
  Heap right redzone:      fb
 
  Freed heap region:       fd
 
  Stack left redzone:      f1
 
  Stack mid redzone:       f2
 
  Stack right redzone:     f3
 
  Stack partial redzone:   f4
 
  Stack after return:      f5
 
  Stack use after scope:   f8
 
  Global redzone:          f9
 
  Global init order:       f6
 
  Poisoned by user:        f7
 
  Contiguous container OOB:fc
 
  ASan internal:           fe
 
==3123==ABORTING



 Comments   
Comment by Alexander Barkov [ 2019-04-29 ]

mysql.user is not necessarily needed to reproduce this. It can be any other table:

DROP TABLE IF EXISTS t1;
 
CREATE TABLE t1 (a INT) ENGINE=MyISAM;
 
INSERT INTO t1 VALUES (1), (2);
 
SELECT REPLACE( CAST( CURDATE() AS BINARY ), CURDATE(), REPEAT('a',32) ) AS f FROM t1 GROUP BY f;

The problem is even repeatable with a single record:

DROP TABLE IF EXISTS t1;
 
CREATE TABLE t1 (a INT) ENGINE=MyISAM;
 
INSERT INTO t1 VALUES (1);
 
SELECT REPLACE( CAST( CURDATE() AS BINARY ), CURDATE(), REPEAT('a',32) ) AS f FROM t1 GROUP BY f;

Comment by Alexander Barkov [ 2019-04-29 ]

Also repeatable with these scripts:

DROP TABLE IF EXISTS t1;
 
CREATE TABLE t1 (a INT) ENGINE=MyISAM;
 
INSERT INTO t1 VALUES (1),(2);
 
SELECT REPLACE( LEFT( CURDATE(), 4), LEFT(CURDATE(),4), REPEAT('a',32) ) AS f FROM t1 GROUP BY f;


DROP TABLE IF EXISTS t1;
 
CREATE TABLE t1 (a INT) ENGINE=MyISAM;
 
INSERT INTO t1 VALUES (1),(2);
 
SELECT REPLACE(RIGHT(CURDATE(), 4), RIGHT(CURDATE(),4), REPEAT('a',32)) AS f FROM t1 GROUP BY f;


DROP TABLE IF EXISTS t1;
 
CREATE TABLE t1 (a INT) ENGINE=MyISAM;
 
INSERT INTO t1 VALUES (1),(2);
 
SELECT REPLACE(SUBSTR(CURDATE(),2,3), SUBSTR(CURDATE(),2,3), REPEAT('a',32)) AS f FROM t1 GROUP BY f;

Note, with these scripts the crash happens on a different line in Item_func_replace::val_str().

Generated at Thu Feb 08 08:46:22 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.