Details
-
Bug
-
Status: Closed (View Workflow)
-
Critical
-
Resolution: Fixed
-
10.1.4
-
None
-
10.2.1-5
Description
See buildbot, rpm-centos5-amd64 builder. Install step fails after installing galera.
I've repeated it easily — just build 10.1 in centos5-amd64-build VM, install galera rpm and start ./mtr --suite wsrep. This crashes immediately with the following stack trace:
#0 0x0000003dc520b9b2 in pthread_kill () from /lib64/libpthread.so.0
|
(gdb) bt
|
#0 0x0000003dc520b9b2 in pthread_kill () from /lib64/libpthread.so.0
|
#1 0x00000000007886b1 in handle_fatal_signal (sig=6)
|
at /mnt/bzr/git/sql/signal_handler.cc:266
|
#2 <signal handler called>
|
#3 0x0000003dc4a30265 in raise () from /lib64/libc.so.6
|
#4 0x0000003dc4a31d10 in abort () from /lib64/libc.so.6
|
#5 0x00002acda7232c44 in __gnu_cxx::__verbose_terminate_handler ()
|
from /usr/lib64/libstdc++.so.6
|
#6 0x00002acda7230db6 in ?? () from /usr/lib64/libstdc++.so.6
|
#7 0x00002acda7230de3 in std::terminate () from /usr/lib64/libstdc++.so.6
|
#8 0x00002acda7230eca in __cxa_throw () from /usr/lib64/libstdc++.so.6
|
#9 0x00002aaaab033a96 in gu::RegEx::Match::str ()
|
at galerautils/src/gu_regex.hpp:65
|
warning: (Internal error: pc 0x2aaaab033fd6 in read in psymtab, but not in symtab.)
|
|
|
warning: (Internal error: pc 0x2aaaab033fd6 in read in psymtab, but not in symtab.)
|
|
|
#10 0x00002aaaab033fd7 in gu::URI::get_port (this=Unhandled dwarf expression opcode 0xf3
|
)
|
at galerautils/src/gu_uri.hpp:170
|
warning: (Internal error: pc 0x2aaaab033fd6 in read in psymtab, but not in symtab.)
|
|
|
warning: (Internal error: pc 0x2aaaab033fb0 in read in psymtab, but not in symtab.)
|
|
|
#11 0x00007fff03d76560 in ?? ()
|
#12 0x00002aaaab0c6a2a in galera::ReplicatorSMM::InitConfig::InitConfig (this=Unhandled dwarf expression opcode 0xf3
|
)
|
at galera/src/replicator_smm_params.cpp:93
|
#13 0x77203a5045525357 in ?? ()
|
#14 0x00002acda7463e00 in ?? () from /usr/lib64/libstdc++.so.6
|
#15 0x6c6147203a292864 in ?? ()
|
#16 0x00007fff03d77910 in ?? ()
|
#17 0x373833337228392e in ?? ()
|
---Type <return> to continue, or q <return> to quit---
|
#18 0x646f432079622029 in ?? ()
|
#19 0x00002acdacbb8448 in ?? ()
|
#20 0x406f666e693c2079 in ?? ()
|
#21 0x00002acdacbaa5d8 in ?? ()
|
#22 0x6c203e6d6f632e01 in ?? ()
|
#23 0x7573206465646101 in ?? ()
|
#24 0x00002acdacbb8148 in ?? ()
|
#25 0x00002acda7463e18 in std::string::_Rep::_S_empty_rep_storage ()
|
from /usr/lib64/libstdc++.so.6
|
#26 0x00002aaaaaeb6a00 in ?? () from /usr/lib64/galera/libgalera_smm.so
|
#27 0x00002acdacbb8460 in ?? ()
|
#28 0x00002acdacbb8490 in ?? ()
|
#29 0x00002acdacbb8490 in ?? ()
|
#30 0x00002acda7463e18 in std::string::_Rep::_S_empty_rep_storage ()
|
from /usr/lib64/libstdc++.so.6
|
#31 0x00002aaaaaee3c01 in ?? () from /usr/lib64/galera/libgalera_smm.so
|
#32 0x00002acda7463e18 in std::string::_Rep::_S_empty_rep_storage ()
|
from /usr/lib64/libstdc++.so.6
|
#33 0x0000000200000000 in ?? ()
|
#34 0x0000000000001c36 in ?? ()
|
#35 0x0000000000000000 in ?? ()
|
Attachments
Issue Links
Activity
No, that's no enough as the failure is on the install VM. It doesn't matter how we build, it crashes when galera is installed. And even if we upgrade install VM, it won't help our users at all.
The fix is to record the dependency in the galera.rpm, to specify that it needs libgcc 4.1.2-55 or later. I've just verified — there is libgcc 4.1.2-55.el5 for CentOS 5, so this dependency will be satisfied and libgcc will be automatically upgraded, just as needed.
Raising priority since the issue is been around since 9 months already and affects bb.
Closing, now that Galera 25.3.19 packages are in staging repo.
Hi serg! I stumbled into this issue sometime back. It actually affects other
galera versions too. While its reproducible with the libgcc version 4.1.2-44.el5
currently installed on vm-centos5-amd64, after updating to 4.1.2-55.el5, its no
longer reproducible.
Should we just update gcc on vm-centos5-amd64 to the latest one available?