Re: [Xen-devel] Fwd: Debugging "Xen BUG at ... list.h" on ARM

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Mon, 2014-04-14 at 10:30 +0100, Thomas Leonard wrote:
> Sorry, you're right, I added an extra assert there. Here's a stack
> trace with a clean build (stable-4.4 branch + debug=y):
> 
> (XEN) Assertion 'entry->next->prev == entry' failed, line 175, file
> /home/tal/work/cubieboard/xen/xen/include/xen/list.h
> (XEN) Xen BUG at /home/tal/work/cubieboard/xen/xen/include/xen/list.h:175
> (XEN) CPU1: Unexpected Trap: Undefined Instruction
> (XEN) ----[ Xen-4.4.1-pre  arm32  debug=y  Not tainted ]----
> (XEN) CPU:    1
> (XEN) PC:     002414cc __bug+0x28/0x44
> (XEN) CPSR:   200e01da MODE:Hypervisor
> (XEN)      R0: 002656dc R1: 00000001 R2: 3fd2bd80 R3: 200e01da
> (XEN)      R4: 000000af R5: 0025d238 R6: 002e4020 R7: 002e6380
> (XEN)      R8: 002e6380 R9: 600e01da R10:4003a940 R11:40037e34 R12:00000002
> (XEN) HYP: SP: 40037e2c LR: 002414cc
> (XEN)
> (XEN)   VTCR_EL2: 80002558
> (XEN)  VTTBR_EL2: 000100007ec06000
> (XEN)
> (XEN)  SCTLR_EL2: 30cd187f
> (XEN)    HCR_EL2: 0000000000282835
> (XEN)  TTBR0_EL2: 0000000076011000
> (XEN)
> (XEN)    ESR_EL2: 00000000
> (XEN)  HPFAR_EL2: 000000000001c810
> (XEN)      HDFAR: c8800f00
> (XEN)      HIFAR: c87400a0
> (XEN)
> (XEN) Xen stack trace from sp=40037e2c:
> (XEN)    00000000 40037e64 0022e32c 15882183 00000007 062e43e6 062e43e6 
> 00000000
> (XEN)    4003a000 40004000 00000002 002e62ac 00000000 40037e7c 00250480 
> 077e6000
> (XEN)    40004000 4003a000 4003b000 40037e94 00243e70 002e6500 40004000 
> 4003b000
> (XEN)    4003b000 40037ea4 002445b8 14f1129d 00000007 40037f0c 002266e8 
> 00000000
> (XEN)    4003b7c8 40037efc 002e62a8 ffffffff ffffffff 002e6588 0024f9d0 
> 40037eec
> (XEN)    0022884c 40004000 00000000 ffffffff ffffffff 40037f00 c07063f8 
> 40037f54
> (XEN)    00000000 002b1ff4 002ae000 002e7614 00265b80 002b1ff0 002e4254 
> 40037f3c
> (XEN)    00229b28 00000000 ffffffff 002e7614 002ae000 002e7614 c07063f8 
> c07414d7
> (XEN)    c050078c c06fe000 c07414d7 40037f44 00229bc0 40037f54 0024ee00 
> c0706474
> (XEN)    c06fe000 40037f58 00251150 ffffffed 077de000 c070746c 00000000 
> c0706474
> (XEN)    c06fe000 c07063f8 c07414d7 c050078c c06fe000 c07414d7 c06fe000 
> c070d900
> (XEN)    ffffffff 0000d43b c0019d28 600e0093 00000000 be9695f4 c0742080 
> c0012400
> (XEN)    c06fffa8 c000f600 c074208c c0012760 c0742098 c00125e0 00000000 
> 00000000
> (XEN)    00000000 00000000 00000000 00000000 00000000 600e0013 800f0193 
> 800f0093
> (XEN)    600e0193 00000000 00000000 030c3080 451a8c04
> (XEN) Xen call trace:
> (XEN)    [<002414cc>] __bug+0x28/0x44 (PC)
> (XEN)    [<002414cc>] __bug+0x28/0x44 (LR)
> (XEN)    [<0022e32c>] set_timer+0x1d0/0x278
> (XEN)    [<00250480>] virt_timer_save+0x88/0x90


This presumably implies that the timer is already active, which it
really shouldn't be while the vcpu is scheduled.

The guest here is the same stuff as
https://github.com/talex5/xen-paratest ?

Ian.

> (XEN)    [<00243e70>] schedule_tail+0x48/0x2f4
> (XEN)    [<002445b8>] context_switch+0x114/0x118
> (XEN)    [<002266e8>] schedule+0x8a0/0x8c8
> (XEN)    [<00229b28>] __do_softirq+0xf4/0x100
> (XEN)    [<00229bc0>] do_softirq+0x14/0x18
> (XEN)    [<0024ee00>] leave_hypervisor_tail+0x50/0x6c
> (XEN)    [<00251150>] return_to_guest+0xc/0xb8



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel