Hide Forgot
Created attachment 490510 [details] snapshots Description of problem: ping pong migrate a win2003-64 guest with qxl video adapter will result bsod. Version-Release number of selected component (if applicable): qemu-kvm-0.12.1.2-2.153.el6.x86_64 qxl-win-driver:qxl-win-0.1-4 How reproducible: always Steps to Reproduce: 1.boot a win2003-64-qxl guest : ###### qemu-kvm -drive file='win2003-64-virtio.qcow2',index=0,if=none,id=drive-virtio-disk1,media=disk,cache=none,format=qcow2,aio=native -device virtio-blk-pci,bus=pci.0,addr=0x4,drive=drive-virtio-disk1,id=virtio-disk1 -device rtl8139,netdev=idDxyhum,mac=9a:f3:19:18:b6:09,netdev=idDxyhum,id=ndev00idDxyhum,bus=pci.0,addr=0x3 -netdev tap,id=idDxyhum,ifname='t0-171835-YiMT',script='qemu-ifup-switch',downscript='no' -m 4096 -smp 2,cores=1,threads=1,sockets=2 -drive file='/winutils.iso',index=1,if=none,id=drive-ide0-0-0,media=cdrom,readonly=on,format=raw -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -cpu cpu64-rhel6,+sse2,+x2apic -rtc base=localtime,clock=host,driftfix=none -boot order=cdn,once=c,menu=off -usbdevice tablet -enable-kvm \ -spice port=8000,disable-ticketing \ \ -vga qxl ######### 2.boot another guest for migrate: cmd as step 1...-incoming tcp:0:5200 3. do several rounds of ping pong migration,guest bsod and memdump. Actual results: guest bsod and memdump Expected results: guest running normally Additional info: 1. host info: kernel :kernel-2.6.32-125.el6.x86_64 (8G 4cpu) cpu processor : 3 vendor_id : GenuineIntel cpu family : 6 model : 23 model name : Intel(R) Core(TM)2 Quad CPU Q9400 @ 2.66GHz stepping : 10 cpu MHz : 2660.513 cache size : 3072 KB physical id : 0 siblings : 4 core id : 3 2. qxl qxl-win-0.1-4 from brew 3. vgabios vgabios-0.6b-3.5.el6.noarch 4. dump info memdump analyzed result attached. 5. guest win2003-64
Created attachment 490512 [details] memdump analyzed result.
Since RHEL 6.1 External Beta has begun, and this bug remains unresolved, it has been rejected as it is not proposed as exception or blocker. Red Hat invites you to ask your support representative to propose this request, if appropriate and relevant, in the next release of Red Hat Enterprise Linux.
Created attachment 491195 [details] migrate using vnc instead of spice. this bug is also triggered when using vnc instead of spice when migrate. attached the memory dump analyzed result
Created attachment 491197 [details] might this bug is no related to display protocol,they shows the same BSOD code when using diff display protocol
Created attachment 491200 [details] might this bug is no related to display protocol,they shows the same BSOD code when using diff display protocol
(In reply to comment #6) > Created attachment 491200 [details] > might this bug is no related to display protocol,they shows the same BSOD code > when using diff display protocol Hello Xiaoqing, Can you post dump files? Best regards, Vadim.
Created attachment 502702 [details] the memory dump itself,2GB after unzip
Created attachment 502703 [details] yet another dump file.part1
Created attachment 502704 [details] yet another dump file.part2 these two parts need to be concatenated before unzip, cat 2k3... > zipped_file.zip
(In reply to comment #10) > (In reply to comment #6) > > Created attachment 491200 [details] > > might this bug is no related to display protocol,they shows the same BSOD code > > when using diff display protocol > > Hello Xiaoqing, > Can you post dump files? Hi Vadim, dump files uploaded, one kernel_dump,one full memory dump. Hope will help, Best Regards, Xiaoqing.
Hi Xiaoqing, Could you try reproducing this bug with the latest qxl driver. I'm asking because the qxl driver used in this test comes without PDB symbols. Best regards, Vadim. From my conversation with Arnon: > I'm looking into > https://bugzilla.redhat.com/show_bug.cgi?id=694431 > bug, and I see that it crashes inside of qxl driver, > but w/o symbols I can say nothing concrete. > > The qxl driver is from > https://brewweb.devel.redhat.com/buildinfo?buildID=159581 > but there are no symbols there. > > Best regards, > Vadim. > > Unfortunately we added PDBs to the brew build only later, at qxl-win-0.1-7 https://brewweb.devel.redhat.com/buildinfo?buildID=170864 So if it's possible to reproduce with these drivers, life will be easier
Created attachment 514978 [details] win7 bsod 0x8e mini dump
Created attachment 514979 [details] win7 mini analyzed
(In reply to comment #23) > Created attachment 514979 [details] > win7 mini analyzed Thank you Xiaoqing! I'm trying to debug the following crash dump https://bugzilla.redhat.com/attachment.cgi?id=514978 but unfortunately still have some problems. Will it be possible to install debug version of qxl from http://download.devel.redhat.com/brewroot/packages/qxl-win/0.1/7/win/qxl_w7_x86_debug.zip and try reproducing it one more time? Best regards, Vadim.
(In reply to comment #24) Hi Vadim, Actually , I tried to trigger this bug using debug-driver before , but it doesn't generate full dump :( , so I use not-debug-driver , also not full dump :( :( . I will try to reproduce this on a new win7,it's installing. and will use debug driver . maybe a fresh guest can generate full dump :) . you need the dump itself or just the analyzed_result enough ? Kind Regards, Xiaoqing.
(In reply to comment #25) > (In reply to comment #24) > Hi Vadim, > > Actually , I tried to trigger this bug using debug-driver before , but it > doesn't generate full dump :( , so I use not-debug-driver , also not full dump > :( :( . > > I will try to reproduce this on a new win7,it's installing. > > and will use debug driver . maybe a fresh guest can generate full dump :) . > > you need the dump itself or just the analyzed_result enough ? > > Kind Regards, > Xiaoqing. Hi Xiaoqing. Actually I even don't need a full dump. Having kernel or compact will be absolutely fine. Best regards, Vadim.
*** This bug has been marked as a duplicate of bug 714801 ***