Bug 812705

reproduced on:
2.6.32-262.el6.x86_64
qemu-kvm-0.12.1.2-2.275.el6.x86_64

using virtio-blk

Sorry for report this bug on a private tree, i can reproduce this on the latest qemu-kvm and using virtio-scsi disk :

# uname -r;rpm -q qemu-kvm
2.6.32-262.el6.x86_64
qemu-kvm-0.12.1.2-2.275.el6.x86_64

Kevin, can you look whether its a symptom of a critical issue?

The problem could in theory occur even with the default cluster size, even though it's rather unlikely. It happens when allocating requests to more than 16 different L2 tables are queued because they depend on other requests.

It causes a qemu abort(), but image consistency is not harmed. Critical enough to be fixed in 6.3, I'd say.

I found a reproducer, added it to qemu-iotests and sent upstream patches. RHEL code is different, so I'm working on a different fix there.

reporduce this issue with steps and  environment as follows:
# uname -r
2.6.32-269.el6.x86_64
# rpm -q qemu-kvm
qemu-kvm-0.12.1.2-2.287.el6.x86_64

steps:
1)# qemu-img create -f qcow2 sysdisk.qcow2 20G -o cluster_size=4096
2)boot guest with virtio-scsi interface and cluster_size=4096,
 /usr/libexec/qemu-kvm -cpu cpu64-rhel6 -rtc base=localtime,clock=host,driftfix=slew -M rhel6.3.0 -enable-kvm -name rhel6.3 -smp 4,cores=2,threads=1,sockets=2 -m 4G -uuid a3d13230-f1c1-4dc9-95de-bb92b2017674 -boot menu=on -drive file=/home/sysdisk.qcow2,if=none,id=drive-virtio-disk0,format=qcow2,cache=none,aio=native,media=disk,werror=stop,rerror=stop -device virtio-scsi-pci,id=bus1 -device scsi-hd,bus=bus1.0,drive=drive-virtio-disk0,id=virtio-scsi-pci0 -netdev tap,id=hostnet0,vhost=on,script=/etc/qemu-ifup -device virtio-net-pci,netdev=hostnet0,id=net0,mac=44:37:E6:97:58:89 -spice port=9000,disable-ticketing -vga qxl -global qxl-vga.vram_size=67108864 -monitor stdio -usb -device usb-tablet,id=input1 -drive file=/home/RHEL6.3-20120426.2-Server-x86_64-DVD1.iso,if=none,id=drive-virtio-disk1,format=raw,cache=none,aio=native,media=cdrom -device virtio-scsi-pci,id=bus2 -device scsi-cd,bus=bus2.0,drive=drive-virtio-disk1,id=virtio-scsi-pci1,bootindex=1

results : when formating
disk qemu-kvm gets Aborted.
bt
#0  0x00007ffff57798a5 in raise () from /lib64/libc.so.6
#1  0x00007ffff577b085 in abort () from /lib64/libc.so.6
#2  0x00007ffff7e3d7ae in qcow2_cache_find_entry_to_replace (bs=0x7ffff86ef010, c=0x7ffff86d8df0, offset=70295552, table=0x7fffda5a4108, 
    read_from_disk=false) at block/qcow2-cache.c:209
#3  qcow2_cache_do_get (bs=0x7ffff86ef010, c=0x7ffff86d8df0, offset=70295552, table=0x7fffda5a4108, read_from_disk=false)
    at block/qcow2-cache.c:229
#4  0x00007ffff7e3b619 in l2_allocate (bs=0x7ffff86ef010, offset=2693144576, new_l2_table=0x7fffda5a4198, new_l2_offset=0x7fffda5a41a0, 
    new_l2_index=0x7fffda5a41ac) at block/qcow2-cluster.c:180
#5  get_cluster_table (bs=0x7ffff86ef010, offset=2693144576, new_l2_table=0x7fffda5a4198, new_l2_offset=0x7fffda5a41a0, 
    new_l2_index=0x7fffda5a41ac) at block/qcow2-cluster.c:512
#6  0x00007ffff7e3ba66 in qcow2_alloc_cluster_offset (bs=0x7ffff86ef010, offset=2693144576, n_start=0, n_end=1008, num=0x7fffda5a42bc, 
    m=0x7fffda5a4250) at block/qcow2-cluster.c:714
#7  0x00007ffff7e3773f in qcow2_co_writev (bs=0x7ffff86ef010, sector_num=<value optimized out>, remaining_sectors=1008, 
    qiov=0x7fffda4a4088) at block/qcow2.c:555
#8  0x00007ffff7e2293a in bdrv_co_do_writev (bs=0x7ffff86ef010, sector_num=5260048, nb_sectors=1008, qiov=0x7fffda4a4088, 
    flags=<value optimized out>) at block.c:1741
#9  0x00007ffff7e229e1 in bdrv_co_do_rw (opaque=0x7fffda4a4290) at block.c:3039
#10 0x00007ffff7e27eeb in coroutine_trampoline (i0=<value optimized out>, i1=<value optimized out>) at coroutine-ucontext.c:129
#11 0x00007ffff578a630 in ?? () from /lib64/libc.so.6
#12 0x00007fffed148530 in ?? ()
#13 0x0000000000000000 in ?? ()


verify this issue with steps and  environment as follows:
version
# uname -r
2.6.32-262.el6.x86_64
rpm -q  qemu-kvm
qemu-kvm-0.12.1.2-2.290.el6.x86_64
 
the steps as same as reproduce.

results:

qemu-kvm work well ,not abort on the point formating.so this issue has been fixed.

    Technical note added. If any revisions are required, please edit the "Technical Notes" field
    accordingly. All revisions will be proofread by the Engineering Content Services team.
    
    New Contents:
No Documentation Needed

    Technical note updated. If any revisions are required, please edit the "Technical Notes" field
    accordingly. All revisions will be proofread by the Engineering Content Services team.
    
    Diffed Contents:
@@ -1 +1 @@
-No Documentation Needed+NEEDINFO

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-2012-0746.html