[jira] [Resolved] (TS-738) 'make check` fails on x86

["Leif Hedstrom (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/TS-738?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Leif Hedstrom resolved TS-738.
------------------------------

    Resolution: Fixed

Closing this, since John fixed it (thanks!)

> 'make check` fails on x86
> -------------------------
>
>                 Key: TS-738
>                 URL: https://issues.apache.org/jira/browse/TS-738
>             Project: Traffic Server
>          Issue Type: Bug
>          Components: Build
>    Affects Versions: 2.1.7
>         Environment: ATS 2.1.7 on x86, gcc 4.5.2
>            Reporter: Arno Toell
>            Assignee: John Plevyak
>            Priority: Minor
>             Fix For: 2.1.9
>
>
> The `make check' target of the Makefile fails to build on x86, AMD64 is fine. Output is (stripped to relevant parts only):
> {code}
> root@build:/home/arno/build-area/trafficserver-2.1.7-unstable# make check
> Making check in proxy/api/ts
> ...
> Create thread 59
> Create thread 60
> Create thread 61
> Create thread 62
> Create thread 63
> 0x08724a60   0x08724a60   0x08724a60
> FAIL: test_freelist
> PASS: test_arena
> test_List PASSED
> PASS: test_List
> =============================================
> 1 of 4 tests failed
> Please report to dev@trafficserver.apache.org
> =============================================
> {code}
> or 
> {code}
> /bin/bash: line 5: 29956 Segmentation fault      ${dir}$tst
> FAIL: test_freelist
> {code}
> or it even succeeds. This might be a race condition therefore. A backtrace from a core dump of a crashed {{LD_PRELOAD=./.libs/libtsutil.so.2 ./.libs/test_freelist}} run shows:
> {code}
> Core was generated by `./.libs/test_freelist'.
> Program terminated with signal 11, Segmentation fault.
> #0  ink_freelist_new (f=0x94ff018) at ink_queue.cc:322
> 322           SET_FREELIST_POINTER_VERSION(next, *ADDRESS_OF_NEXT(TO_PTR(FREELIST_POINTER(item)), f->offset),
> (gdb) bt
> #0  ink_freelist_new (f=0x94ff018) at ink_queue.cc:322
> #1  0x08048996 in test (d=0x1f) at test_freelist.cc:50
> #2  0xf772f7b0 in start_thread () from /lib/libpthread.so.0
> #3  0xf74688fe in clone () from /lib/libc.so.6
> {code}
> I'm not familiar with ATS' data structures, but gdb indicates the following values:
> {code}
> $2 = {s = {pointer = 0x1f1f1f1f, version = 165962584}, data = 712803871161786143}
> (gdb) print f
> $3 = (InkFreeList *) 0x94ff018
> (gdb) print f->offset
> $4 = 0
> {code}

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira