Windows 7 x64 HVM and error 07B - how to change scsi controller type - 4.0RC3

[happymeal00]

Qubes OS version:

4.0 RC3

Affected TemplateVMs:

StandaloneVM (Windows 7 HVM)

---

Steps to reproduce the behavior:

when booting an existing Windows 7 x64 systems exported from VirtualBox, the system hangs with BSOD error 0x0000007B.
I have tried to follow suggested changes by enabling other scsi controllers by booting back into Virtual box and as explained here: http://www.justandrew.net/2009/10/stop-0x0000007b-on-p2vd-windows-7.html but nothing has changed.

While looking at xen logs written in /var/log/xen/console/guest--dm.log i can see that passed scsi controller is a mptsas1068 controller for which my Windows system doesn't have any 64 bit driver.

Expected behavior:

normal boot of the Windows x64 HVM machine and recognition of the associated storage controller/ attached disks

Actual behavior:

BSOD just after boot before loading all *.sys system drivers.
BSOD error number is 0x0000007B

General notes:

I have tried to understand how to pass a diffrent scsi controller (lsi for instance) but it's unclear for me where to set this in the libvirt template. Directly in the xen.xml template file or in the block.xml template file? under what section and which syntax to be used?

---

Related issues:

[happymeal00]

I found a bypass solution for this problem by booting back the vm into virtualbox and by enabling the following registry keyword responsible for the load of the scsi mptsas1060 controller:

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\LSI_SAS key:Start value:0 (0 meaning to be loaded).

As a general solution i believe it would make sense to provide a way to edit (either temporary or permanently) a template configuration file allowing the choiche between available qemu emulated controller.

[awokd]

@alcreator, what's the easiest way to "[switch] r/w disks to IDE", like you described in #3651?

[alcreator]

When I was testing the ide disks, I used the same method that I've described in #4321 (comment).

The argument replacement commands were:

#Convert SCSI rw disks to IDE
dm_args=$(echo "$dm_args" | sed "s/-drive\\${SP}file=\/dev\/xvd\(.\),if=none,id=disk\(.\),format=host_device,cache=writeback,readonly=off\\${SP}-device\\${SP}scsi-hd,bus=scsi0.0,drive=disk.,wwn=0x352540005175626./-drive\\${SP}if=ide,media=disk,file=\/dev\/xvd\1,id=disk\2,format=host_device,cache=writeback,readonly=off/g")
#Convert IDE cdrom to SCSI
dm_args=$(echo "$dm_args" | sed "s/-drive\\${SP}if=ide,readonly=on,media=cdrom,id=ide-51840,file=\/dev\/xvdi,format=host_device/-drive\\${SP}if=none,readonly=on,file=\/dev\/xvdi,format=host_device,id=ide-51840\\${SP}-device\\${SP}scsi-cd,bus=scsi0.0,drive=ide-51840,wwn=0x352540005175626f/g")

To get xen to provide these commands be default instead, take a look at the tools/libxl/libxl_dm.c file in xen and the patch-stubdom-linux-0005.patch from qubes-vmm-xen. The code to set the disks as SCSI is added by this patch; it can be modified to only attach ro disks as SCSI.

[github-actions]

This issue is being closed because:

• This issue is on the "Release 4.0 updates" milestone.
• Qubes OS 4.0 reached EOL (end-of-life) over one year ago.
• There has not been any activity on this issue in over one year.

If anyone believes that this issue should be reopened and reassigned to an active milestone, please leave a brief comment.
(For example, if a bug still affects Qubes OS 4.1, then the comment "Affects 4.1" will suffice.)