Description of problem:
Balloon a Win7-64 guest very slowly.

Version-Release number of selected component (if applicable):
kernel-2.6.32-411.el6.x86_64
qemu-kvm-rhev-0.12.1.2-2.389.el6.x86_64
virtio-win-prewhql-0.1-67.iso

How reproducible:
100%

Steps to Reproduce:
1. Boot a Win7-64 guest with mem=8192 :
/usr/libexec/qemu-kvm \
    -S \
    -name 'virt-tests-vm1' \
    -nodefaults \
    -chardev socket,id=qmp_id_qmpmonitor1,path=/tmp/monitor-qmpmonitor1-20130809-150153-w7QnGFsU,server,nowait \
    -mon chardev=qmp_id_qmpmonitor1,mode=control \
    -chardev socket,id=serial_id_serial1,path=/tmp/serial-serial1-20130809-150153-w7QnGFsU,server,nowait \
    -device isa-serial,chardev=serial_id_serial1 \
    -chardev socket,id=seabioslog_id_20130809-150153-w7QnGFsU,path=/tmp/seabios-20130809-150153-w7QnGFsU,server,nowait \
    -device isa-debugcon,chardev=seabioslog_id_20130809-150153-w7QnGFsU,iobase=0x402 \
    -device ich9-usb-uhci1,id=usb1,bus=pci.0,addr=0x4 \
    -drive file='/home/win7-64-virtio.qcow2',index=0,if=none,id=drive-virtio-disk1,media=disk,cache=none,snapshot=off,format=qcow2,aio=native \
    -device virtio-blk-pci,bus=pci.0,addr=0x5,drive=drive-virtio-disk1,bootindex=0 \
    -device virtio-net-pci,netdev=idN4f4DY,mac='9a:eb:ec:ed:ee:ef',bus=pci.0,addr=0x3,id='idYpIz37' \
    -netdev tap,id=idN4f4DY,vhost=on \
    -m 8192 \
    -smp 16,maxcpus=16,cores=4,threads=2,sockets=2 \
    -cpu 'Opteron_G4',,hv_relaxed \
    -M pc \
    -device usb-tablet,id=usb-tablet1,bus=usb1.0,port=1 \
    -vnc :0 \
    -vga std \
    -rtc base=localtime,clock=host,driftfix=slew  \
    -boot order=cdn,once=c,menu=off   \
    -balloon virtio \
    -enable-kvm \
    -monitor stdio \
    -drive file='/home/virtio-win-prewhql-0.1-67.iso',index=1,if=none,id=drive-ide0-0-0,media=cdrom,format=raw \
    -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0 \

2. (qemu) info balloon
   balloon: actual=8192

3. (qemu) balloon 4096

Actual results:
Balloon from 8192MB to 4096MB, needs about 11mins, about 6.2MB/s.

Expected results:
Balloon should be faster.
Tested with Linux.6.4-64 guest, balloon evict just needs few seconds.

Additional info:
1. There is no load on guest & host.

guest:
Before balloon:
cpu: 0%
mem: 690MB

Balloon ing:
cpu: 6%
mem: 690MB -> 4.74GB

After balloon:
cpu: 0%
mem: 4.74GB

When balloon, cpu & mem increase smoothly.

2. Tested with Linux.6.4-64 guest, balloon evict just needs few seconds.

3. Tested with virtio-win-1.6.5-6.el6_4, also slowly.

4. cpuinfo:
processor	: 31
vendor_id	: AuthenticAMD
cpu family	: 21
model		: 1
model name	: AMD Opteron(TM) Processor 6272                 
stepping	: 2
cpu MHz		: 2099.977
cache size	: 2048 KB
physical id	: 1
siblings	: 16
core id		: 7
cpu cores	: 8
apicid		: 79
initial apicid	: 47
fpu		: yes
f2.6.32-411.el6.x86_64pu_exception	: yes
cpuid level	: 13
wp		: yes
flags		: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nonstop_tsc extd_apicid amd_dcm aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 nodeid_msr topoext perfctr_core cpb npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
bogomips	: 4199.76
TLB size	: 1536 4K pages
clflush size	: 64
cache_alignment	: 64
address sizes	: 48 bits physical, 48 bits virtual
power management: ts ttp tm 100mhzsteps hwpstate cpb

5. mem:
# free -m
             total       used       free     shared    buffers     cached
Mem:         64374      24727      39647          0         45      23742
-/+ buffers/cache:        939      63435
Swap:        57343          0      57343