Bug 1000942

Summary: can't boot win2003-32bit guest with more than 1 vcpu
Product: Red Hat Enterprise Linux 7 Reporter: Xu Han <xuhan>
Component: qemu-kvmAssignee: Yvugenfi <yvugenfi>
Status: CLOSED CURRENTRELEASE QA Contact: Virtualization Bugs <virt-bugs>
Severity: medium Docs Contact:
Priority: medium    
Version: 7.0CC: acathrow, hhuang, juzhang, michen, rhod, virt-maint, xfu, xuhan, xwei, yvugenfi
Target Milestone: rc   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-12-30 02:52:08 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Xu Han 2013-08-26 07:29:57 UTC
Description of problem:
boot win2003-32bit guest with more than 1 vcpu can not success, just stall in black screen.

Version-Release number of selected component (if applicable):
kernel:3.10.0-11.el7.x86_64
qemu:qemu-kvm-1.5.2-4.el7.x86_64

How reproducible:
100%

Steps to Reproduce:
1.boot guest:
/usr/libexec/qemu-kvm -name vm1 -nodefaults -m 4G -smp 2,cores=2,threads=1,sockets=1 -M pc-i440fx-rhel7.0.0 -cpu host -rtc base=utc,clock=host,driftfix=slew -k en-us -boot menu=on -monitor stdio -vnc :1 -spice disable-ticketing,port=5931 -vga qxl -qmp tcp:0:5555,server,nowait -drive file=/home/guest-img/win2003-32-virtio.qcow2,if=none,id=drive-virtio-disk,format=qcow2,cache=none,werror=stop,rerror=stop -device virtio-blk-pci,scsi=off,drive=drive-virtio-disk,id=disk0,bootindex=1 -balloon virtio -monitor unix:/tmp/monitor2,server,nowait

Actual results:
guest stall in black screen

Expected results:
guest can be boot success

Additional info:
win2003-64bit won't hit this issue

Comment 2 Xiaoqing Wei 2013-08-26 08:29:51 UTC
is this a regression ? pls try downgrade and re-test

Comment 3 juzhang 2013-08-26 08:33:16 UTC
Can you have a try other emulated cpu module instead of -cpu host? I'm not very sure we official support -cpu host.

Comment 4 Yvugenfi@redhat.com 2013-08-26 13:45:40 UTC
Was the guest installed with 1 vCPU and then changed to 2 or you cannot make the initial boot for installation?

Keep in mind that Windows XP and Windows 2003 used different hals and kernels for SMP and single CPU. And those where deployed during installation.

Best regards,
Yan.

Comment 5 Xu Han 2013-08-27 05:14:10 UTC
(In reply to Yan Vugenfirer from comment #4)
> Was the guest installed with 1 vCPU and then changed to 2 or you cannot make
> the initial boot for installation?
> 
> Keep in mind that Windows XP and Windows 2003 used different hals and
> kernels for SMP and single CPU. And those where deployed during installation.
> 
> Best regards,
> Yan.

Hi Yan,

I have re-tested this issue. The win2003-32 bit guest started installation with 2 vCPUs.

During installation, copy file step would work normally. But after reboot, the guest would stall in black screen. Then I restarted qemu with 1 vCPU and saw installation went on. Base on this test result, win2k3 32bit guest still don't support multiple vcpus to boot. would you please help confirm this issue again?

Comment 6 Ronen Hod 2013-12-26 13:45:12 UTC
xuhan,
It is important to understand whether this is a regression. If it is, then we need to fix it, and if it is not, then we might defer it.
Thanks.

Comment 7 Xu Han 2013-12-27 06:36:01 UTC
Tested with 'qemu-kvm-1.3.0-6.el7.x86_64' and 'qemu-kvm-1.4.0-3.el7.x86_64' both hit this issue, so it isn't a regression. But can not reproduce this bug with 'qemu-kvm-1.5.3-30.el7.x86_64'.

Comment 8 Ronen Hod 2013-12-27 14:05:27 UTC
(In reply to xuhan from comment #7)
> Tested with 'qemu-kvm-1.3.0-6.el7.x86_64' and 'qemu-kvm-1.4.0-3.el7.x86_64'
> both hit this issue, so it isn't a regression. But can not reproduce this
> bug with 'qemu-kvm-1.5.3-30.el7.x86_64'.

Even better, so the current RHEL7 qemu does not have this bug?
If this is the case, then let's close this bug.
Does RHEL6 have this bug? If true, then please clone this BZ, and we will consider backporting the fix.

Comment 9 juzhang 2013-12-30 01:50:20 UTC
Hi Xuhan, 

Could you have a look and add your comment?

Best Regards,
Junyi

Comment 10 Xu Han 2013-12-30 02:52:08 UTC
(In reply to Ronen Hod from comment #8)
> Even better, so the current RHEL7 qemu does not have this bug?

Yes. According to comment 7, I'd close this bug.

> If this is the case, then let's close this bug.
> Does RHEL6 have this bug? If true, then please clone this BZ, and we will
> consider backporting the fix.

Tested on a RHEL6.5 host with 'qemu-kvm-0.12.1.2-2.415.el6.x86_64' by using following command lines:
# /usr/libexec/qemu-kvm ... \
-smp 2,cores=2,threads=1,sockets=1 \
...

# /usr/libexec/qemu-kvm ... \
-smp 2,cores=1,threads=2,sockets=1 \
...

# /usr/libexec/qemu-kvm ... \
-smp 2,cores=1,threads=1,sockets=2 \
...

All would not hit this issue.

Comment 11 Ronen Hod 2013-12-30 13:30:09 UTC
(In reply to xuhan from comment #10)

Thanks, Ronen.