Re: [Qemu-devel] Fedora FC21 - Bug: 100% CPU and hangs in gettimeofday(&tp, NULL); forever

[Paolo Bonzini <pbonzini@redhat.com>]

On 15/02/2015 09:18, Gerhard Wiesinger wrote:
> 
> ppoll([{fd=98<socket:[4660317]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=89<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=88<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=87<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=86<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=85<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=84<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=83<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=82<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=81<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=80<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=79<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=78<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=77<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=76<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=75<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=74<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=73<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=72<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=71<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=70<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=69<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=68<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=67<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=66<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=65<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=64<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=63<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=62<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=61<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=60<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP},
> {fd=59<anon_inode:[eventfd]>, events=POLLIN|POLLERR|POLLHUP}, ...], 71,
> {0, 124112228}, NULL, 8) = 1 ([...], left {0, 124090614})
> write(7<anon_inode:[eventfd]>, "\1\0\0\0\0\0\0\0", 8) = 8
> read(6<anon_inode:[eventfd]>, "\1\0\0\0\0\0\0\0", 512) = 8
> write(7<anon_inode:[eventfd]>, "\1\0\0\0\0\0\0\0", 8) = 8

Can you grab some random backtraces ("thread apply all bt full") with gdb?

What is the libvirt XML or qemu command line?

Paolo

> Kernel (host/guest):  3.18.6-200.fc21.x86_64 #1 SMP
> qemu-kvm-2.2.0-5.fc21.x86_64
> 
> Bug 1178975 - endless loop in clock_gettime() on a kvm-based VM
> https://bugzilla.redhat.com/show_bug.cgi?id=1178975
> is fixed (didn't occour with the test program posted at
> https://bugzilla.redhat.com/show_bug.cgi?id=1178975#c28 in 30min,
> happened before reproduceable in 2min, still running)
> 
> So I guess there is another problem in the kernel with volatile and gcc
> optimizations (or maybe in qemu-KVM)

No, this doesn't look like volatile.