libvirt users - Sep 2014 - Re: 1.2.7 and 1.2.8 fail to start container: libvirt_lxc[4904]: segfault at 0 ip ...error 4 in libc-2.17.so[

HI all

Centos 7,  3.10.0-123.6.3.el7.x86_64
libvirt 1.27, libvirt 1.2.8 builded from source with
./configure --prefix=/usr
make && make install
LXC with direct network failed to start:

Sep 16 19:19:38 node01 kernel: device br502 entered promiscuous mode
Sep 16 19:19:39 node01 kernel: device br502 left promiscuous mode
Sep 16 19:19:39 node01 avahi-daemon[1532]: Withdrawing workstation
service for macvlan0.
Sep 16 19:19:39 node01 kernel: XFS (dm-16): Mounting Filesystem
Sep 16 19:19:39 node01 kernel: XFS (dm-16): Ending clean mount
Sep 16 19:19:39 node01 kernel: libvirt_lxc[4904]: segfault at 0 ip
00007ffe3cbf0df6 sp 00007ffe3fa03c98 error 4 in
libc-2.17.so[7ffe3cabf000+1b6000]
Sep 16 19:19:39 node01 abrt-hook-ccpp: Saved core dump of pid 1
(/usr/lib/systemd/systemd) to /var/tmp/abrt/ccpp-2014-09-16-19:19:39-1
(716800 bytes)
Sep 16 19:19:39 node01 journal: Cannot recv data: Connection reset by peer
Sep 16 19:19:39 node01 journal: internal error: guest failed to start:

In libvirt-1.2.6 work fine.

LXC config:
 <domain type='lxc' id='5933'>
  <name>ce7-t1</name>
  <uuid>f80ad54d-6560-4bd0-aa6d-df3e29888914</uuid>
  <memory unit='KiB'>2097152</memory>
  <currentMemory unit='KiB'>2097152</currentMemory>
  <memtune>
    <hard_limit unit='KiB'>2097152</hard_limit>
    <soft_limit unit='KiB'>2097152</soft_limit>
    <swap_hard_limit unit='KiB'>3145728</swap_hard_limit>
  </memtune>
  <vcpu placement='static'>2</vcpu>
  <resource>
    <partition>/machine</partition>
  </resource>
  <os>
    <type arch='x86_64'>exe</type>
    <init>/sbin/init</init>
  </os>
  <features>
    <privnet/>
  </features>
  <clock offset='utc'/>
  <on_poweroff>destroy</on_poweroff>
  <on_reboot>restart</on_reboot>
  <on_crash>restart</on_crash>
  <devices>
    <emulator>/usr/libexec/libvirt_lxc</emulator>
    <filesystem type='block' accessmode='passthrough'>
      <source dev='/dev/data/ce7-t1'/>
      <target dir='/'/>
    </filesystem>
    <filesystem type='ram' accessmode='passthrough'>
      <source usage='524288' units='KiB'/>
      <target dir='/dev/shm'/>
    </filesystem>
    <interface type='direct'>
      <mac address='02:00:00:58:d8:15'/>
      <source dev='br502' mode='bridge'/>
    </interface>
    <console type='pty' tty='/dev/pts/2'>
      <source path='/dev/pts/2'/>
      <target type='lxc' port='0'/>
      <alias name='console0'/>
    </console>
  </devices>
  <seclabel type='none'/>
</domain>

Where br502 attached to vlan interface :
#brctl  show br502
bridge name     bridge id               STP enabled     interfaces
br502           8000.002590e2da34       no              eno1.502


In 1.2.7 and 1.2.8  lxc with nat network work fine.

b.r.
 Maxim Kozin

On 16.09.2014 17:40, mxs kolo wrote:
> HI all
>
> Centos 7,  3.10.0-123.6.3.el7.x86_64
> libvirt 1.27, libvirt 1.2.8 builded from source with
> ./configure --prefix=/usr
> make && make install
> LXC with direct network failed to start:
>
> Sep 16 19:19:38 node01 kernel: device br502 entered promiscuous mode
> Sep 16 19:19:39 node01 kernel: device br502 left promiscuous mode
> Sep 16 19:19:39 node01 avahi-daemon[1532]: Withdrawing workstation
> service for macvlan0.
> Sep 16 19:19:39 node01 kernel: XFS (dm-16): Mounting Filesystem
> Sep 16 19:19:39 node01 kernel: XFS (dm-16): Ending clean mount
> Sep 16 19:19:39 node01 kernel: libvirt_lxc[4904]: segfault at 0 ip
> 00007ffe3cbf0df6 sp 00007ffe3fa03c98 error 4 in
> libc-2.17.so[7ffe3cabf000+1b6000]
> Sep 16 19:19:39 node01 abrt-hook-ccpp: Saved core dump of pid 1
> (/usr/lib/systemd/systemd) to /var/tmp/abrt/ccpp-2014-09-16-19:19:39-1
> (716800 bytes)
This file would be interesting, can you:

gdb /usr/libexec/libvirt_lxc -c /var/tmp/abrt/ccpp-2014-09-16-19:19:39-1
(gdb shell should appear here)
(gdb) bt

and paste the backtrace here? That way we can identify where the lxc 
helper is crashing.

Michal

Hi all
>
> gdb /usr/libexec/libvirt_lxc -c /var/tmp/abrt/ccpp-2014-09-16-19:19:39-1
> (gdb shell should appear here)
> (gdb) bt
>
> and paste the backtrace here? That way we can identify where the lxc
> helper is crashing.
Another node with  1.2.8, same case with direct+vlan:

[root@node02 libvirt-1.2.8]# gdb /usr/libexec/libvirt_lxc -c
/var/tmp/abrt/ccpp-2014-09-18-19\:48\:04-1/coredump
...

#0  0x00007f4ca4604df6 in __strcmp_sse42 () from /lib64/libc.so.6
#1  0x00007f4ca63b3809 in lxcContainerGetNetDef (vmDef=<optimized
out>, vmDef=<optimized out>, devName=0x7f4ca6f94770
"macvlan0") at
lxc/lxc_container.c:475
#2  lxcContainerRenameAndEnableInterfaces (veths=0x7f4ca6f94750,
nveths=1, vmDef=0x7f4ca6f9b160) at lxc/lxc_container.c:505
#3  lxcContainerChild (data=0x7fffd6f685c0) at lxc/lxc_container.c:2122
#4  0x00007f4ca45c901d in clone () from /lib64/libc.so.6

b.r.
 Maxim Kozin