Re: New x86-Setup code breaks HVM-XEN boot

From: H. Peter Anvin
Date: Fri Aug 31 2007 - 10:22:00 EST

Christian Ehrhardt wrote:
(XEN) Failed vm entry (exit reason 0x80000021) caused by invalid guest state (0).
(XEN) ************* VMCS Area **************
(XEN) 16-bit Guest-State Fields
(XEN) 0x00000800: 0x0018 0x0010 0x0018 0x0018 (XEN) 0x00000808: 0x0018 0x0018 0x0000 0x0008 (XEN) 16-bit Host-State Fields
(XEN) 0x00000c00: 0xe010 0xe008 0xe010 0xe010 (XEN) 0x00000c08: 0xe010 0xe010 0xe050 (XEN) 64-bit Control Fields
(XEN) 0x00002000: 0x0000000000bd5000 0x0000000000000000 0x0000000000bd4000 0x0000000000000000 (XEN) 0x00002004: 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 (XEN) 0x00002008: 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 (XEN) 0x0000200c: 0x0000000000000000 0x0000000000000000 ------------------ ------------------ (XEN) 0x00002010: 0x00000000d8ba12e5 0x00000000ffd5e176 0x0000000000000000 0x0000000000000000 (XEN) 64-bit Guest-State Fields
(XEN) 0x00002800: 0x00000000ffffffff 0x00000000ffffffff 0x0000000000000000 0x0000000000000000 (XEN) 32-bit Control Fields
(XEN) 0x00004000: 0x0000001f 0x0681e7fa 0x00004008 0x00000000 (XEN) 0x00004008: 0x00000000 0x00000000 0x0003edff 0x00000000 (XEN) 0x00004010: 0x00000000 0x000011ff 0x00000000 0x00000020 (XEN) 0x00004018: 0x00000000 0x00000000 0x00000000 (XEN) 32-bit RO Data Fields
(XEN) 0x00004400: 0x0000000c 0x80000021 0x00000000 0x00050033 (XEN) 0x00004408: 0x00000000 0x00000000 0x00000003 0x00000000 (XEN) 32-bit Guest-State Fields
(XEN) 0x00004800: 0xffffffff 0xffffffff 0xffffffff 0xffffffff (XEN) 0x00004808: 0xffffffff 0xffffffff 0x00000000 0x00002067 (XEN) 0x00004810: 0x00006fb5 0x00000000 0x0000d0ff 0x0000d0ff (XEN) 0x00004818: 0x0000d0ff 0x0000d0ff 0x0000d0ff 0x0000d0ff (XEN) 0x00004820: 0x00000082 0x0000008b 0x00000000 0x00000000 (XEN) 0x00004828: 0x00000000 0x00000000 (XEN) 32-bit Host-State Fields
(XEN) 0x00004c00: 0x00000000 (XEN) Natural 64-bit Control Fields
(XEN) 0x00006000: 0x00000000ffffffff 0x00000000ffffffff 0x0000000000050033 0x0000000000000651 (XEN) 0x00006008: 0x0000000000000000 0x0000000000000000 0x0000000000000000 0x0000000000000000 (XEN) 64-bit RO Data Fields
(XEN) 0x00006400: 0x0000000000000000 0x0000000000010002 0x00000000003f7fd5 0x00000000001a2000 (XEN) 0x00006408: 0x00000000000d11b2 0x0000000000000000 (XEN) Natural 64-bit Guest-State Fields
(XEN) 0x00006800: 0x0000000080050033 0x0000000000bbfda0 0x0000000000002671 0x00000000ffffffff (XEN) 0x00006808: 0x00000000ffffffff 0x00000000ffffffff 0x00000000ffffffff 0x00000000ffffffff (XEN) 0x00006810: 0x00000000ffffffff 0x0000000000000000 0x00000000000d7264 0x00000000190df0bc (XEN) 0x00006818: 0x0000000000000000 0x0000000000000400 0x000000000000f9e8 0x0000000000100000 (XEN) 0x00006820: 0x0000000000010002 0x0000000000000000 0x0000000000000000 0x0000000000000000 (XEN) Natural 64-bit Host-State Fields
(XEN) 0x00006c00: 0x000000008005003b 0x00000000a1f47000 0x00000000000026f0 0x0000000000000000 (XEN) 0x00006c08: 0x0000000000000000 0x00000000ff1cf380 0x00000000fe800000 0x00000000ffbeb080 (XEN) 0x00006c10: 0x0000000000000000 0x0000000000000000 0x00000000ffbf3fe8 0x00000000ff14ab10 (XEN) **************************************
(XEN) domain_crash_sync called from vmx.c:2154
(XEN) Domain 353 (vcpu#0) crashed on cpu#1:
(XEN) ----[ Xen-3.0.3-1 x86_32p debug=n Not tainted ]----
(XEN) CPU: 1
(XEN) EIP: 0010:[<00100000>]
(XEN) EFLAGS: 00010002 CONTEXT: hvm
(XEN) eax: 00100000 ebx: 00000000 ecx: f0000018 edx: 00050013
(XEN) esi: 000932a0 edi: 000042d0 ebp: 000d0000 esp: 0000f9e8
(XEN) cr0: 00050033 cr4: 00000651 cr3: 00bbfda0 cr2: 00000000
(XEN) ds: 0018 es: 0018 fs: 0018 gs: 0018 ss: 0018 cs: 0010


Xen crashes because it thinks VMX should be handling this, but VMX doesn't think so (the exit reason is "invalid state".)

At this point, paging is not yet enabled (CR0.PG = 0), but that is not natively supported by VMX. From a cursory look it doesn't appear as though that Xen has recognized that it's supposed to emulate this in one way or another.

I'm on the road, so I don't have time for a more detailed analysis just at the moment.

-hpa
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/