[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] Testing status of fully virtualized guests (Intel VT) on 64bit XEN unstable

On 21 Jun 2006, at 13:51, Ed Smith wrote:

(XEN) Xen call trace:
(XEN)    [<ffff83000015c5d5>] __vmx_clear_vmcs+0x25/0x40
(XEN)    [<ffff83000015cb85>] vmx_destroy_vmcs+0x15/0x80
(XEN)    [<ffff83000015e2f9>] vmx_relinquish_guest_resources+0x29/0xa0
(XEN)    [<ffff83000011d3bf>] domain_relinquish_resources+0x18f/0x260
(XEN)    [<ffff830000107598>] domain_kill+0x28/0x60
(XEN)    [<ffff83000010625d>] do_dom0_op+0x57d/0xd20
(XEN)    [<ffff830000129702>] do_update_descriptor+0x352/0x3e0
(XEN)    [<ffff830000127ec9>] do_mmuext_op+0x719/0x7b0
(XEN)    [<ffff830000165d7b>] do_iret+0x7b/0x140
(XEN)    [<ffff830000164e62>] syscall_enter+0x62/0x67

I added error-return checking to all the low-level VMX instruction macros deliberately to flush out this kind of thing. This one is probably failing because VMCLEAR is being executed from the wrong CPU (actually the VMCS is active on another CPU). That failure type isn't documented, but it would make sense.

-- Keir

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.