On Monday, March 26, 2012 05:51:37 PM Vadim Rozenfeld wrote: > On Monday, March 26, 2012 02:45:34 PM Paul Fisher wrote: > > On 26 March 2012 09:50, Vadim Rozenfeld <vroze...@redhat.com> wrote: > > > On Sunday, March 25, 2012 07:01:54 PM Yan Vugenfirer wrote: > > > Hi Paul, > > > Could you try reproducing this problem on "-smp 2" guest, with > > > small memory dump option turned on, instead of kernel memory dump. > > > > > > Thanks, > > > Vadim. > > > > Hi Vadim. > > > > Thank you kindly for getting back to me. > > > > I can certainly do that: > > > > Dump and BSOD are in a zip at > > http://dl.dropbox.com/u/12332019/2200-VirtioBSOD-2.zip > > > > This time running with 2 cores: > > > > qemu-kvm -m 1024 -smp sockets=1,cores=2 -cpu host -nodefaults -vga cirrus > > -vnc :1 -drive if=none,id=block.0,format=raw, > > cache=writeback,file=drive.img -device > > virtio-blk-pci,bootindex=1,drive=block.0 -monitor stdio > > > > Same procedure to get it to break - run crystal disk mark. Passed the > > stream read/write and 512k block read/write - broke at the seek intensive > > 4k level. > > > > Regards, > > Paul > > Hi Paul, > Managed to reproduce this problem, thank you. Going to fix it in the build. > Meanwhile, instead of writeback, could you try cache=off ? > > Best regards, > Vadim.
should be fixed in build25 http://download.devel.redhat.com/brewroot/packages/virtio-win- prewhql/0.1/25/win/virtio-win-prewhql-0.1.zip http://people.redhat.com/vrozenfe/build-25/virtio-win-prewhql-0.1.zip Note that it is a pre-WHQL build, so use it on your own risk. Use https://bugzilla.redhat.com/show_bug.cgi?id=810694 for the bug tracking. Cheers, Vadim.