| Bug #81868 | pfs_instr-oom unit test fails on MySQL Server 5.7.13 | ||
|---|---|---|---|
| Submitted: | 15 Jun 2016 17:06 | Modified: | 17 Jun 2016 13:00 |
| Reporter: | Yura Sorokin (OCA) | Email Updates: | |
| Status: | Closed | Impact on me: | |
| Category: | MySQL Server: Tests | Severity: | S3 (Non-critical) |
| Version: | 5.7.13 | OS: | Any |
| Assigned to: | CPU Architecture: | Any | |
[15 Jun 2016 17:11]
Yura Sorokin
Debugging with GDB reveals readable call stack *************************************************************** [yura@localhost mysql-build]$ gdb ctest GNU gdb (GDB) Red Hat Enterprise Linux 7.6.1-80.el7 Copyright (C) 2013 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-redhat-linux-gnu". For bug reporting instructions, please see: <http://www.gnu.org/software/gdb/bugs/>... Reading symbols from /usr/bin/ctest...Reading symbols from /usr/bin/ctest...(no debugging symbols found)...done. (no debugging symbols found)...done. Missing separate debuginfos, use: debuginfo-install cmake-2.8.11-5.el7.x86_64 (gdb) set follow-fork-mode child (gdb) run -R pfs_instr-oom Starting program: /usr/bin/ctest -R pfs_instr-oom [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Test project /home/yura/ws/mysql-build Start 6: pfs_instr-oom [New process 18047] [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". process 18047 is executing new program: /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7ffff7fdb840 (LWP 18047)] 0x00007ffff6d1df7b in __memcmp_sse4_1 () from /lib64/libc.so.6 Missing separate debuginfos, use: debuginfo-install glibc-2.17-106.el7_2.6.x86_64 libgcc-4.8.5-4.el7.x86_64 libstdc++-4.8.5-4.el7.x86_64 zlib-1.2.7-15.el7.x86_64 (gdb) bt #0 0x00007ffff6d1df7b in __memcmp_sse4_1 () from /lib64/libc.so.6 #1 0x000000000044d31a in my_strnncoll_binary (cs=0x954900 <my_charset_bin>, s=0x7ffff7f3a028 <Address 0x7ffff7f3a028 out of bounds>, slen=59, t=0x7fffffffc410 "/home/yura/ws/mysql-build/storage/perfschema/unittest/dummy", tlen=59, t_is_prefix=0 '\000') at /home/yura/ws/mysql-server/strings/ctype-bin.c:84 #2 0x000000000043bed8 in my_lfind (head=0x9f6628, cs=0x954900 <my_charset_bin>, hashnr=1930419929, key=0x7fffffffc410 "/home/yura/ws/mysql-build/storage/perfschema/unittest/dummy", keylen=59, cursor=0x7fffffffc170, pins=0x9f2600) at /home/yura/ws/mysql-server/mysys/lf_hash.c:113 #3 0x000000000043c43e in my_lsearch (head=0x9f6628, cs=0x954900 <my_charset_bin>, hashnr=1930419929, key=0x7fffffffc410 "/home/yura/ws/mysql-build/storage/perfschema/unittest/dummy", keylen=59, pins=0x9f2600) at /home/yura/ws/mysql-server/mysys/lf_hash.c:364 #4 0x000000000043cba6 in lf_hash_search (hash=0x9e50a0 <filename_hash>, pins=0x9f2600, key=0x7fffffffc410, keylen=59) at /home/yura/ws/mysql-server/mysys/lf_hash.c:589 #5 0x0000000000420f5d in find_or_create_file (thread=0x7fffffffd480, klass=0x7fffffffcc40, filename=0x4741f7 "dummy", len=5, create=true) at /home/yura/ws/mysql-server/storage/perfschema/pfs_instr.cc:844 #6 0x00000000004039c6 in test_oom () at /home/yura/ws/mysql-server/storage/perfschema/unittest/pfs_instr-oom-t.cc:249 #7 0x0000000000404451 in do_all_tests () at /home/yura/ws/mysql-server/storage/perfschema/unittest/pfs_instr-oom-t.cc:420 #8 0x0000000000404487 in main () at /home/yura/ws/mysql-server/storage/perfschema/unittest/pfs_instr-oom-t.cc:427 ***************************************************************
[16 Jun 2016 4:21]
MySQL Verification Team
Hello Yura Sorokin, Thank you for the report. Thanks, Umesh
[16 Jun 2016 4:22]
MySQL Verification Team
test results
Attachment: 81868.results (application/octet-stream, text), 194.14 KiB.
[17 Jun 2016 12:11]
Yura Sorokin
Re-initializing 'file_hash' for the second 'find_or_create_file()' call (*) I confirm the code being submitted is offered under the terms of the OCA, and that I am authorized to contribute it.
Contribution: pfs_instr-oom-5.7.13_crash_fix.patch (application/octet-stream, text), 957 bytes.
[17 Jun 2016 13:00]
Erlend Dahl
This has been fixed in the upcoming 5.7.14 release under the heading of Bug#23186653 PERFORMANCE SCHEMA UNIT TESTS PFS_INSTR-OOM & PFS_INSTR_CLASS FAIL REGULARLY
Description: 'pfs_instr-oom' unit test fails with the following output ************************************************************************* ./mysql-test/mtr --debug-server 1st --unit-tests Logging: /home/yura/ws/mysql-server/mysql-test/mysql-test-run.pl --debug-server 1st --unit-tests MySQL Version 5.7.13 Checking supported features... - SSL connections supported - binaries are debug compiled Collecting tests... Checking leftover processes... Removing old var directory... Creating var directory '/home/yura/ws/mysql-build/mysql-test/var'... Installing system database... ============================================================================== TEST RESULT TIME (ms) or COMMENT -------------------------------------------------------------------------- worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 13000..13009 main.1st [ pass ] 5 unit_tests [ fail ] Test ended at 2016-06-15 19:47:06 ctest failed with exit code 2048, see result below -------------------------------------------------------------------------- Unit tests: 98% tests passed, 1 tests failed out of 64 The following tests FAILED: 6 - pfs_instr-oom (Failed) Report from unit tests in /home/yura/ws/mysql-build/mysql-test/var/ctest.log -------------------------------------------------------------------------- The servers were restarted 0 times Spent 0.005 of 27 seconds executing testcases Completed: Failed 1/2 tests, 50.00% were successful. Failing test(s): unit_tests The log files in var/log may give you some hint of what went wrong. If you want to report this error, please read first the documentation at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html mysql-test-run: *** ERROR: there were failing test cases ************************************************************************* Here is an excerpt from 'ctest.log' ************************************************************************* 6/64 Test #6: pfs_instr-oom ................................................................................................***Failed 0.06 sec 1..28 ok 1 - instances init ok 2 - create mutex ok 3 - oom (create mutex) ok 4 - instances init ok 5 - create rwlock ok 6 - oom (create rwlock) ok 7 - instances init ok 8 - create cond ok 9 - oom (create cond) ok 10 - instances init ok 11 - create file Signal 11 thrown, attempting backtrace. stack_bottom = 0 thread_stack 0x0 /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x439681] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x4049cc] /lib64/libc.so.6(+0x35670)[0x7f4f48e69670] /lib64/libc.so.6(+0x15ef7b)[0x7f4f48f92f7b] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x44d31a] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x43bed8] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x43c43e] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x43cba6] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x420f5d] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x4039c6] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x404451] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x404487] /lib64/libc.so.6(__libc_start_main+0xf5)[0x7f4f48e55b15] /home/yura/ws/mysql-build/storage/perfschema/unittest/pfs_instr-oom-t[0x402d59] Start 7: pfs_account-oom 7/64 Test #7: pfs_account-oom .............................................................................................. Passed 0.05 sec Start 8: pfs_host-oom ************************************************************************* Tested on CentOS 7.2 x64, GCC 4.8.5 with default debug build. How to repeat: git clone -b 5.7 https://github.com/mysql/mysql-server.git mkdir mysql-build cd mysql-build cmake \ ../mysql-server \ -DCMAKE_BUILD_TYPE=Debug \ -DMYSQL_MAINTAINER_MODE=ON \ -DBUILD_CONFIG=mysql_release \ -DFEATURE_SET=community \ -DENABLE_DTRACE=OFF \ -DWITH_SSL=system \ -DWITH_ZLIB=system \ -DENABLE_DOWNLOADS=1 \ -DDOWNLOAD_BOOST=1 \ -DWITH_BOOST=../deps ./mysql-test/mtr --debug-server 1st --unit-tests Suggested fix: Most probably this failure is connected with the fix to Bug #22662621 "UNIT TEST PFS_INSTR_OOM-T LEAKS MEMORY" commit e90a0e3 (https://github.com/mysql/mysql-server/commit/e90a0e3)