| Bug #54744 | valgrind reports leak for mysqlbinlog | ||
|---|---|---|---|
| Submitted: | 23 Jun 2010 15:34 | Modified: | 6 Aug 2010 14:58 |
| Reporter: | Luis Soares | Email Updates: | |
| Status: | Closed | Impact on me: | |
| Category: | MySQL Server: Replication | Severity: | S3 (Non-critical) |
| Version: | trunk | OS: | Any |
| Assigned to: | Luis Soares | CPU Architecture: | Any |
[23 Jun 2010 20:02]
Valeriy Kravchuk
Verified just as described on Mac OS X: ... ==98366== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0) ==98366== malloc/free: in use at exit: 14,537 bytes in 17 blocks. ==98366== malloc/free: 19 allocs, 2 frees, 15,053 bytes allocated. ==98366== ==98366== searching for pointers to 17 not-freed blocks. ......................==98366== checked 6,241,840 bytes. ... ==98366== ERROR ==98366== ==98366== ==98366== 7,900 bytes in 1 blocks are still reachable in loss record 10 of 10 ==98366== at 0x4CE53: malloc+99 (in /usr/local/lib/valgrind/x86-darwin/vgpreload_memcheck.so) ==98366== by 0x7CF695: tzsetwall_basic+65 (in /usr/lib/libSystem.B.dylib) ==98366== by 0x7EF999: localtime_r+54 (in /usr/lib/libSystem.B.dylib) ==98366== by 0x41A806: my_init_time+44 (in /Users/openxs/dbs/trunk/lib/mysql/libmysqlclient.16.dylib) ==98366== by 0x1B941: main+95 (in bin/mysqlbinlog) ==98366== ==98366== LEAK SUMMARY: ==98366== definitely lost: 0 bytes in 0 blocks. ==98366== possibly lost: 0 bytes in 0 blocks. ==98366== still reachable: 14,537 bytes in 17 blocks. ==98366== suppressed: 0 bytes in 0 blocks. ...
[6 Jul 2010 22:33]
Luis Soares
On my local environment I get: valgrind --tool=memcheck --leak-check=full --show-reachable=yes -v client/mysqlbinlog ==21158== HEAP SUMMARY: ==21158== in use at exit: 368 bytes in 2 blocks ==21158== total heap usage: 36 allocs, 34 frees, 47,524 bytes allocated ==21158== ==21158== Searching for pointers to 2 not-freed blocks ==21158== Checked 2,638,152 bytes ==21158== ==21158== 88 bytes in 1 blocks are still reachable in loss record 1 of 2 ==21158== at 0x4C284A8: malloc (vg_replace_malloc.c:236) ==21158== by 0x4420C1: DbugMalloc (dbug.c:2183) ==21158== by 0x43E525: code_state (dbug.c:397) ==21158== by 0x4408D6: _db_enter_ (dbug.c:1152) ==21158== by 0x46AF4C: dirname_part (mf_dirname.c:70) ==21158== by 0x460C9E: intern_filename (mf_pack.c:525) ==21158== by 0x45A090: my_basic_init (my_init.c:123) ==21158== by 0x45A0D0: my_init (my_init.c:146) ==21158== by 0x40CDFB: main (mysqlbinlog.cc:2020) ==21158== ==21158== 280 bytes in 1 blocks are still reachable in loss record 2 of 2 ==21158== at 0x4C277CC: calloc (vg_replace_malloc.c:467) ==21158== by 0x46815F: my_thread_init (my_thr_init.c:352) ==21158== by 0x467A61: my_thread_basic_global_init (my_thr_init.c:126) ==21158== by 0x467CAD: my_thread_global_init (my_thr_init.c:185) ==21158== by 0x45A02F: my_basic_init (my_init.c:103) ==21158== by 0x45A0D0: my_init (my_init.c:146) ==21158== by 0x40CDFB: main (mysqlbinlog.cc:2020) ==21158== ==21158== LEAK SUMMARY: ==21158== definitely lost: 0 bytes in 0 blocks ==21158== indirectly lost: 0 bytes in 0 blocks ==21158== possibly lost: 0 bytes in 0 blocks ==21158== still reachable: 368 bytes in 2 blocks ==21158== suppressed: 0 bytes in 0 blocks ==21158== ==21158== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 4 from 4) --21158-- --21158-- used_suppression: 2 dl-hack3-cond-1 --21158-- used_suppression: 2 glibc-2.5.x-on-SUSE-10.2-(PPC)-2a ==21158==
[6 Jul 2010 22:42]
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/112981 3085 Luis Soares 2010-07-06 BUG#54744: valgrind reports leak for mysqlbinlog The server was not cleaning up dbug allocated memory before exiting. This is not a real problem, as this memory would be deallocated anyway. Nonetheless, we improve the mysqlbinlog exit procedure, wrt to memory book-keeping, when no parameter is given. To fix this, we deploy a call to my_thread_end() before the thread exits, which will also free pending dbug related allocated blocks.
[6 Jul 2010 22:48]
Luis Soares
My env: - ubuntu 10.04 - valgrind: valgrind-3.6.0.SVN-Debian - arch: x86_64 - gcc: gcc (Ubuntu 4.4.3-4ubuntu5) 4.4.3 After patching mysqlbinlog.cc with the patch above, I get: $ valgrind --tool=memcheck --leak-check=full --show-reachable=yes -v client/mysqlbinlog (...) ==23685== ==23685== HEAP SUMMARY: ==23685== in use at exit: 0 bytes in 0 blocks ==23685== total heap usage: 36 allocs, 36 frees, 47,524 bytes allocated ==23685== ==23685== All heap blocks were freed -- no leaks are possible ==23685== ==23685== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 4 from 4) --23685-- --23685-- used_suppression: 2 dl-hack3-cond-1 --23685-- used_suppression: 2 glibc-2.5.x-on-SUSE-10.2-(PPC)-2a ==23685== ==23685== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 4 from 4)
[8 Jul 2010 9:29]
Luis Soares
Queued in mysql-trunk-bugfixing and merged to mysql-next-mr-bugfixing: - http://pb2.norway.sun.com/web.py?template=push_details&push=1402801 - http://pb2.norway.sun.com/web.py?template=push_details&push=1402747
[8 Jul 2010 16:26]
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/113161 3100 Luis Soares 2010-07-08 Revert patch for BUG#54744.
[8 Jul 2010 22:41]
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/113186 3102 Luis Soares 2010-07-08 BUG#54744: valgrind reports leak for mysqlbinlog The server was not cleaning up some dbug allocated memory before exiting. This is not a real problem, as this memory would be deallocated anyway. Nonetheless, we improve the mysqlbinlog exit procedure, wrt to memory book-keeping, when no parameter is given. To fix this, we deploy a call to my_end() before the thread exits.
[9 Jul 2010 11:45]
Luis Soares
Queued in mysql-trunk-bugfixing and merged to mysql-next-mr-bugfixing: - http://pb2.norway.sun.com/web.py?template=push_details&push=1406089 - http://pb2.norway.sun.com/web.py?template=push_details&push=1406084
[23 Jul 2010 12:28]
Bugs System
Pushed into mysql-trunk 5.5.6-m3 (revid:alik@sun.com-20100723121820-jryu2fuw3pc53q9w) (version source revid:vasil.dimov@oracle.com-20100531152341-x2d4hma644icamh1) (merge vers: 5.5.5-m3) (pib:18)
[23 Jul 2010 12:35]
Bugs System
Pushed into mysql-next-mr (revid:alik@sun.com-20100723121929-90e9zemk3jkr2ocy) (version source revid:vasil.dimov@oracle.com-20100531152341-x2d4hma644icamh1) (pib:18)
[6 Aug 2010 14:58]
Jon Stephens
No user-visible changes to document. Closed without further action.
Description: valgrind issues the following warnings if mysqlbinlog is called without any parameters: ==20983== ==20983== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 5 from 1) ==20983== malloc/free: in use at exit: 1,912 bytes in 7 blocks. ==20983== malloc/free: 25 allocs, 18 frees, 5,572 bytes allocated. ==20983== For counts of detected errors, rerun with: -v ==20983== searching for pointers to 7 not-freed blocks. ==20983== checked 2,648,104 bytes. ==20983== ==20983== LEAK SUMMARY: ==20983== definitely lost: 0 bytes in 0 blocks. ==20983== possibly lost: 0 bytes in 0 blocks. ==20983== still reachable: 1,912 bytes in 7 blocks. ==20983== suppressed: 0 bytes in 0 blocks. Environment details: tree: mysql-trunk-bugfixing revid: revision-id: joerg@mysql.com-20100623143203-btlp7s33lq2fqizm valgrind: valgrind-3.2.1 host: fimafeng11 OS: Linux gcc: gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14) How to repeat: 1. bzr clone -r revid:joerg@mysql.com-20100623143203-btlp7s33lq2fqizm mysql-trunk-bugfixing BZR_REPO/mysql-trunk-bugfixing 2. cd mysql-trunk-bugfixing 3. bash BUILD/compile-amd64-valgrind-max 4. valgrind client/mysqlbinlog (for additional info) 5. valgrind --tool=memcheck --leak-check=full --show-reachable=yes -v client/mysqlbinlog Suggested fix: .