Gluster should really block _all_ signals except the per-thread ones (SIGSEGV, SIGBUS, SIGILL, SIGSYS, SIGFPE).

> IIRC, not only the APIs we call creates threads, Gluster itself creates extra > threads. We can not mask them all in QEMU side.

Yes, but Gluster can create threads only from Gluster threads that were created before, or from APIs that we call ourselves.  So if all QEMU->Gluster entry points block signals, all thread created by Gluster will have the right mask.

https://code.engineering.redhat.com/gerrit/#/c/13711/

[root@dhcp201-162 ~]# /usr/libexec/qemu-kvm -M pc -cpu SandyBridge -m 2G -smp 1,sockets=2,cores=2,threads=1,maxcpus=2 -enable-kvm -name win2012 -drive file=gluster://10.65.201.191/vmstore/win2012r2-free-install.qcow2,if=none,id=drive-virtio-disk0,format=qcow2,cache=none,werror=stop,rerror=stop,aio=threads -drive file=/home/rhel6u5.qcow2,if=none,id=drive-scsi-disk,format=qcow2,cache=none,werror=stop,rerror=stop
VNC server running on `127.0.0.1:5901'


/usr/libexec/qemu-kvm -M pc -cpu SandyBridge -m 2G -smp
1,sockets=2,cores=2,threads=1,maxcpus=2 -enable-kvm -name win2012 -drive
file=gluster://10.65.201.191/vmstore/win2012r2-free-install.qcow2,if=none,id=drive-virtio-disk0,format=qcow2,cache=none,werror=stop,rerror=stop,aio=threads
-drive
file=/root/windows.iso,if=none,media=cdrom,id=drive-ide0,readonly=on,format=raw
-device ide-drive,bus=ide.1,unit=0,drive=drive-ide0,id=ide0,bootindex=1
VNC server running on `127.0.0.1:5900'


qemu-kvm launches the instance and no crash detected.

Version: 
* qemu-kvm-0.12.1.2-2.411.el6.x86_64
* glusterfs-3.4.0.34rhs-1.el6.x86_64
* kernel-2.6.32-421.el6.x86_64

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHBA-2013-1641.html