| Summary: | QXL WHQL test 1150 is running and qemu-kvm hangs during test of a Windows7x64 guest. | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 6 | Reporter: | Bill Sanford <bsanford> |
| Component: | qemu-kvm | Assignee: | Alon Levy <alevy> |
| Status: | CLOSED NOTABUG | QA Contact: | Virtualization Bugs <virt-bugs> |
| Severity: | high | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 6.1 | CC: | acathrow, bcao, cpelland, dawu, dblechte, juzhang, marcandre.lureau, mdeng, michen, mkenneth, pmatouse, pvine, qzhang, tburke, uril, virt-maint |
| 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: | 2011-10-18 14:15:08 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Bug Depends On: | |||
| Bug Blocks: | 613169, 682015 | ||
|
Description
Bill Sanford
2011-07-11 19:21:17 UTC
There is a huge (4.2 gig) corefile with this bug. Is there a central location I can put it instead of attaching it to BZ? Hi, Bill Thanks for reporting this issue. Could you tell which virtio-win driver you used? It's seems like the qxl device (In reply to comment #4) > It's seems like the qxl device Yes, this is a bug related with qxl driver, I'v sent mail to devel for the detailed information of qxl driver, and will try to reproduce it with whql after I get the reply. Best Regards, Dawn In DTM, how do you look up for "WHQL test 1151 (Child test is 1150)" Why do you have: "-cpu Nehalem,+rdtscp,+xtpr,+tm2,+est,+vmx,+ds_cpl,+monitor,+pbe,+tm,+ht,+ss,+acpi,+ds,+vme" Did you setup the machine with libvirt? Is this 100% reproducible? thanks Why are you using all the above cpu flags? Only Nehalem should be fine Marc-Andre, I installed a Windows VM on a RHEL6.1 machine with the hybrid repo and brought up a basic SPICE client to install the QXL drivers in a SPICE client window on the same machine. I didn't set all of those flags. I would think that setting up the client in VM Manager would use libvirt and set all of those flags. Hi Bill, I have no test 1151 - I have test 281 with the same title. I've run it 2/2 so far with the untagged qxl-win-0.1-10 driver (it's probably been cleaned of build artifacts so I'll start a new one). The title is "Device Path Exerciser". Probably a difference in versions of DTM. I'm using 1.6.8367.000 from 2011/02/03-17:00 (according to About box). Same dates for all Kits (Devices_LK, DisplayLK, and the rest). I've been using Windows 7 x64. What version DTM are you using? I missed the note about the core dump, but 4 Gigabyte is a problem - can you try to reproduce with recent qemu and provide a stack trace? Alon Fixed in qxl-win-0.1-12. See https://bugzilla.redhat.com/show_bug.cgi?id=688883#c30 for note on skipped versions. Put the qxl-win version only in the "Fixed In Version", correcting. This may be already fixed with 179, but 183 contains an additional fix that might be relevant. See https://bugzilla.redhat.com/show_bug.cgi?id=721118#c14 From Description: qemu-kvm command line contains "-vga qxl -global qxl-vga.vram_size=9437184" Only 9MB ? Looking at the command line again, I see an rtl8139 is used - this is probably the reason for the test failure. The tests need to be run using either e1000 with the intel provided newest driver, or with virtio-net-pci. Please try rerunning with one of those network device + driver combinations and let me know if you still see the failure. I'm moving back to assigned meanwhile, I don't think this has anything to do with a new driver or qemu-kvm. (In reply to comment #13) > Looking at the command line again, I see an rtl8139 is used - this is probably > the reason for the test failure. The tests need to be run using either e1000 > with the intel provided newest driver, or with virtio-net-pci. > > Please try rerunning with one of those network device + driver combinations and > let me know if you still see the failure. > > I'm moving back to assigned meanwhile, I don't think this has anything to do > with a new driver or qemu-kvm. Currently,we only use e1000 for window 7 guest.Test 281 with the same title in DTM with version 1.6.8367.The following is our WHQL ordinary testing command. Use the e1000 and qxl-win-0.1-12,QE didn't see the failure. ------------------------------------------------------------------------ /usr/libexec/qemu-kvm -m 2G -smp 2 -cpu cpu64-rhel6,+x2apic -usb -device usb-tablet -drive file=win7-64-balloon.raw,if=none,id=drive-ide0-0-0,werror=stop,rerror=stop,cache=none,format=raw -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -netdev tap,id=hostnet0,script=/etc/qemu-ifup -device e1000,netdev=hostnet0,mac=00:12:13:18:2F:14,bus=pci.0,addr=0x4,id=net0 -boot c -uuid 8e09f9fa-bd05-4b86-9abf-ce0f0fe8ec0e -no-kvm-pit-reinjection -chardev socket,id=111a,path=/tmp/monitor-win7-64-balloon,server,nowait -mon chardev=111a,mode=readline -name win7-64-balloon -spice port=5931,disable-ticketing -vga qxl -device virtio-balloon-pci,addr=0x6,bus=pci.0 -rtc base=localtime (In reply to comment #12) > From Description: > qemu-kvm command line contains "-vga qxl -global qxl-vga.vram_size=9437184" > > Only 9MB ? starting VM via rhev-m will make it . (In reply to comment #13) > Looking at the command line again, I see an rtl8139 is used - this is probably > the reason for the test failure. The tests need to be run using either e1000 > with the intel provided newest driver, or with virtio-net-pci. Tried on qemu-kvm-0.12.1.2-2.195.el6.x86_64 with r8139 nic ,still can not reproduce. steps: 1.start VM: /usr/libexec/qemu-kvm -M rhel6.2.0 -cpu cpu64-rhel6,family=0x6,+x2apic -enable-kvm -m 2G -smp 2G -name Win7x64-qxlw7 -uuid bc6e979b-83e2-4895-8188-96e0b3721470 -monitor unix:/tmp/tt,server,nowait -rtc base=utc -boot c -drive file=/home/win7-64-0920.raw,if=none,id=drive-ide0-0-0,format=raw,cache=none,werror=stop,rerror=stop -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -netdev tap,script=/etc/qemu-ifup,id=hostnet0 -device rtl8139,netdev=hostnet0,id=net0,mac=52:54:00:d0:4d:60,bus=pci.0,addr=0x3 -chardev pty,id=charserial0 -device isa-serial,chardev=charserial0,id=serial0 -usb -spice port=8001,disable-ticketing -vga qxl -global qxl-vga.vram_size=9437184 -balloon none 2.install qxl-win-1.0.9 3.install dtm client in the guest ,and run job 281 "Device Path Exerciser" Actual Results: Tried 2 times ,both can not reproduce . The vm used rtl8139, which caused interrupt handler timeout during my testing, and failed some of the DTM tests, not sure if this one as well. Regardless, just like comment 14 by degmin, I did not see any problems with the DevPathExer test on DTM 1.6.8367.000, in which this subtest is called DevPathExer V3.3 and numbered 279 - Device Path Excerciser Library Job, on both win7 32 bit and 64. Closing as not a bug. Alon |