Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Not a Bug
-
None
-
None
Description
if Maxscale is built with `-DWITH_ASAN=Y` and Maxscale crashes the core dump file is broken.
GBD reports something like:
bq. BFD: Warning: /home/vagrant/core-maxscale-6-997-995-8590-1550622059 is truncated: expected core file size >= 15393782775808, found: 2232320.
|
bq. [New LWP 8590]
|
bq. [New LWP 8591]
|
bq. [New LWP 8592]
|
bq. [New LWP 8593]
|
bq. [New LWP 8594]
|
bq. [New LWP 8595]
|
bq. [New LWP 8596]
|
bq. [New LWP 8597]
|
bq. [New LWP 8598]
|
bq. [New LWP 8599]
|
bq. [New LWP 8600]
|
bq. Cannot access memory at address 0x7fc916c35128
|
bq. Cannot access memory at address 0x7fc916c35120
|
bq. Failed to read a valid object file image from memory.
|
bq. Core was generated by `/usr/bin/maxscale'.
|
bq. Program terminated with signal 6, Aborted.
|
bq. #0 0x00007fc9066f02d2 in ?? ()
|
bq. (gdb) bt
|
bq. Python Exception <class 'gdb.MemoryError'> Cannot access memory at address 0x7ffdb960e5c0:
|
bq. (gdb)
|
bq.
|
examples:
http://max-tst-01.mariadb.com/LOGS/run_test-926/LOGS/mxs2057_systemd_watchdog/000/core-maxscale-6-997-995-8590-1550622059
http://max-tst-01.mariadb.com/LOGS/run_test-926/LOGS/mxs1719/000/core-maxscale-6-997-995-5736-1550621476
http://max-tst-01.mariadb.com/LOGS/run_test-926/LOGS/mxs813_long_hostname/000/core-maxscale-6-997-995-5161-1550621321
(binary for these coredumps is here =http://max-tst-01.mariadb.com/ci-repository/./2.3-buildbot-Feb19-20:00:07/mariadb-maxscale/centos/7/x86_64//)