[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Libunwind-devel] Plan for future releases?
From: |
David Abdurachmanov |
Subject: |
Re: [Libunwind-devel] Plan for future releases? |
Date: |
Sun, 21 Feb 2016 16:28:22 +0200 |
> On 14 Feb 2016, at 10:03, David Abdurachmanov <address@hidden> wrote:
>
> I could run test suite on x86_64/CentOS 6 & 7, aarch64/CentOS 7,
> ppc64le/Fedora 22 and ppc64/Fedora 22 once you have release candidate.
Hi,
Here are results based on c56fb8f99ed88bea006baada12b8577bf771ad14
(Fri Nov 13 14:31:57 2015 +0100) from master.
I tested:
x86_64 : Scientific Linux 6.7, CentOS 7.2, Fedora 23
aarch64 : CentOS 7.2
ppc64le : Fedora 22
ppc64 : Fedora 22
## Fedora 23 -- x86_64
http://davidlt.web.cern.ch/davidlt/vault/libunwind/fedora_23_x86_64.log //
make check
http://davidlt.web.cern.ch/davidlt/vault/libunwind/test-suite_fedora_23_x86_64.log
============================================================================
Testsuite summary for libunwind 1.1
============================================================================
# TOTAL: 33
# PASS: 30
# SKIP: 0
# XFAIL: 2
# FAIL: 1
# XPASS: 0
# ERROR: 0
Hmm...
FAIL: run-coredump-unwind
=========================
lt-test-coredump-unwind: _UCD_create('/tmp/libunwind-test-BTqjysULP5/core*')
failed
FAIL run-coredump-unwind (exit status: 1)
I see this was brought up on the list.
After looking in similar crashes on further tests, I found that the culprit
here is /proc/sys/kernel/core_pattern
I looked at run-coredump-unwind and it looks that crasher is suppose to create
core dump in , i.e. /tmp/libunwind-test-BTqjysULP5
From CentOS 7.2:
% cat /proc/sys/kernel/core_pattern
/var/tmp/cores/core.%e.%p.%h.%t
So, all the core dump automatically go into /var/tmp/cores, not to as tests
expect.
From Fedora:
cat /proc/sys/kernel/core_pattern
|/usr/lib/systemd/systemd-coredump %p %u %g %s %t %e
From Scientific Linux 6:
% cat /proc/sys/kernel/core_pattern
core
It looks like from user perspective I cannot override it. Thus these are
unstable -- system configuration dependable (in my case all machines were
configured differently)
## Fedora 22 -- ppc64le
http://davidlt.web.cern.ch/davidlt/vault/libunwind/fedora_22_ppc64le.log //
make check
http://davidlt.web.cern.ch/davidlt/vault/libunwind/test-suite_fedora_22_ppc64le.log
============================================================================
Testsuite summary for libunwind 1.1
============================================================================
# TOTAL: 33
# PASS: 21
# SKIP: 0
# XFAIL: 2
# FAIL: 10
# XPASS: 0
# ERROR: 0
## Fedora 22 -- ppc64
http://davidlt.web.cern.ch/davidlt/vault/libunwind/fedora_22_ppc64.log // make
check
http://davidlt.web.cern.ch/davidlt/vault/libunwind/test-suite_fedora_22_ppc64.log
============================================================================
Testsuite summary for libunwind 1.1
============================================================================
# TOTAL: 33
# PASS: 21
# SKIP: 0
# XFAIL: 0
# FAIL: 10
# XPASS: 2
# ERROR: 0
## CentOS 7.2.1511 -- aarch64
http://davidlt.web.cern.ch/davidlt/vault/libunwind/centos_7.2.1511_aarch64.log
// make check
http://davidlt.web.cern.ch/davidlt/vault/libunwind/test-suite_centos_7.2.1511_aarch64.log
============================================================================
Testsuite summary for libunwind 1.1
============================================================================
# TOTAL: 33
# PASS: 26
# SKIP: 0
# XFAIL: 2
# FAIL: 5
# XPASS: 0
# ERROR: 0
## CentOS 7.2.1511 -- x86_64
http://davidlt.web.cern.ch/davidlt/vault/libunwind/centos_7.2.1511_x86_64.log
// make check
http://davidlt.web.cern.ch/davidlt/vault/libunwind/test-suite_centos_7.2.1511_x86_64.log
============================================================================
Testsuite summary for libunwind 1.1
============================================================================
# TOTAL: 34
# PASS: 30
# SKIP: 0
# XFAIL: 2
# FAIL: 2
# XPASS: 0
# ERROR: 0
## Scientific Linux 6.7 -- x86_64
http://davidlt.web.cern.ch/davidlt/vault/libunwind/ScientificLinux_6.7_x86_64.log
// make check
Nothing failed.
======================================================
All 33 tests behaved as expected (2 expected failures)
======================================================
Machine had
% cat /proc/sys/kernel/core_pattern
core
Thus core files were generated where expected.
Cheers,
david
- [Libunwind-devel] Plan for future releases?, Yichao Yu, 2016/02/06
- Re: [Libunwind-devel] Plan for future releases?, Arun Sharma, 2016/02/13
- Re: [Libunwind-devel] Plan for future releases?, David Abdurachmanov, 2016/02/14
- Re: [Libunwind-devel] Plan for future releases?,
David Abdurachmanov <=
- Re: [Libunwind-devel] Plan for future releases?, Arun Sharma, 2016/02/22
- Re: [Libunwind-devel] Plan for future releases?, David Abdurachmanov, 2016/02/23
- Re: [Libunwind-devel] Plan for future releases?, David Abdurachmanov, 2016/02/23
- Re: [Libunwind-devel] Plan for future releases?, Arun Sharma, 2016/02/28
- Re: [Libunwind-devel] Plan for future releases?, Harald Servat, 2016/02/29
Re: [Libunwind-devel] Plan for future releases?, Konstantin Belousov, 2016/02/17
Re: [Libunwind-devel] Plan for future releases?, Michael Daniels, 2016/02/17