[
https://issues.apache.org/jira/browse/TS-2665?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13948308#comment-13948308
]
Alexey Ivanov commented on TS-2665:
-----------------------------------
Just to be specific here is how deadlock looks like in RHEL6's version of glibc:
{code}
#0 0x00000039f2cf806e in __lll_lock_wait_private () from /lib64/libc.so.6
#1 0x00000039f2c7d31b in _L_lock_9495 () from /lib64/libc.so.6
#2 0x00000039f2c7a906 in malloc () from /lib64/libc.so.6
#3 0x00000039f240cb8d in _dl_map_object_deps () from
/lib64/ld-linux-x86-64.so.2
#4 0x00000039f2412a11 in dl_open_worker () from /lib64/ld-linux-x86-64.so.2
#5 0x00000039f240e196 in _dl_catch_error () from /lib64/ld-linux-x86-64.so.2
#6 0x00000039f241246a in _dl_open () from /lib64/ld-linux-x86-64.so.2
#7 0x00000039f2d26300 in do_dlopen () from /lib64/libc.so.6
#8 0x00000039f240e196 in _dl_catch_error () from /lib64/ld-linux-x86-64.so.2
#9 0x00000039f2d26457 in __libc_dlopen_mode () from /lib64/libc.so.6
#10 0x00000039f2cfe855 in init () from /lib64/libc.so.6
#11 0x00000039f340cb23 in pthread_once () from /lib64/libpthread.so.0
#12 0x00000039f2cfe954 in backtrace () from /lib64/libc.so.6
#13 0x0000003662a23d6e in ink_stack_trace_dump (sighandler_frame=2) at
ink_stack_trace.cc:53
#14 0x000000000052e138 in signal_handler (sig=11) at signals.cc:221
#15 <signal handler called>
#16 0x00000039f2c7617f in malloc_consolidate () from /lib64/libc.so.6
#17 0x00000039f2c79385 in _int_malloc () from /lib64/libc.so.6
#18 0x00000039f2c7a911 in malloc () from /lib64/libc.so.6
#19 0x0000003a91abd09d in operator new(unsigned long) () from
/usr/lib64/libstdc++.so.6
{code}
> Clean up ink_stack_trace_dump code
> ----------------------------------
>
> Key: TS-2665
> URL: https://issues.apache.org/jira/browse/TS-2665
> Project: Traffic Server
> Issue Type: Bug
> Components: Core
> Reporter: Brian Geffon
> Assignee: Brian Geffon
> Fix For: 5.0.0
>
>
> The code around backtrace is quite messy, it should be cleaned up. Also, in
> certain situations there exists a bug where malloc will block waiting for a
> lock that traffic_server held when it segfaulted, this should be resolved
> with an alarm.
> Thanks to Alexey Ivanov <[email protected]> for reporting this issue.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
