[Xen-devel] Qemu-dm crash with 4.9-unstable

[Razvan Cojocaru <rcojocaru@xxxxxxxxxxxxxxx>]

Hello,

I've installed Xen 4.9-unstable (master branch, commit
1e88db4701d6e2d00c04795e6aacaea942b617e6) on an old Ubuntu workhorse
machine, and when using quemu-xen-traditional, quemu-dm crashes with
this backtrace:

# gdb /usr/lib/xen/bin/qemu-dm core
Reading symbols from /usr/lib/xen/bin/qemu-dm...done.
[New LWP 16753]
[New LWP 16758]
[New LWP 16759]
[New LWP 16834]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `/usr/lib/xen/bin/qemu-dm -d 19 -domain-name
win7_32 -vnc 0.0.0.0:0 -vncunused -'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  __memcpy_sse2_unaligned () at
../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S:36
36      ../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S: No such
file or directory.
(gdb) bt
#0  __memcpy_sse2_unaligned () at
../sysdeps/x86_64/multiarch/memcpy-sse2-unaligned.S:36
#1  0x00007f6ed1857e0d in do_dm_op (xch=0x23bb160, domid=19, nr_bufs=2)
at xc_private.c:808
#2  0x00007f6ed1851f85 in xc_hvm_track_dirty_vram (xch=0x23bb160,
dom=19, first_pfn=983040, nr=576, dirty_bitmap=0x7ffcd33ad730) at
xc_misc.c:562
#3  0x0000000000454183 in vga_draw_graphic (s=0x23cef18, full_update=0)
at ~/work/xen.git/tools/qemu-xen-traditional-dir/hw/vga.c:1794
#4  0x0000000000454a2d in vga_update_display (opaque=0x23cef18) at
~/work/xen.git/tools/qemu-xen-traditional-dir/hw/vga.c:1952
#5  0x00000000004cae9c in vga_hw_update () at console.c:165
#6  0x00000000004febf9 in vnc_update_client (opaque=0x23e9920) at vnc.c:794
#7  0x00000000004089da in qemu_run_timers (ptimer_head=0xb25960
<active_timers>, current_time=1629566) at
~/work/xen.git/tools/qemu-xen-traditional-dir/vl.c:1235
#8  0x000000000040d16e in main_loop_wait (timeout=3600000) at
~/work/xen.git/tools/qemu-xen-traditional-dir/vl.c:3884
#9  0x00000000004bae56 in main_loop () at helper2.c:619
#10 0x0000000000410650 in main (argc=24, argv=0x7ffcd33ae0e8,
envp=0x7ffcd33ae1b0) at
~/work/xen.git/tools/qemu-xen-traditional-dir/vl.c:6216
(gdb) f 1
#1  0x00007f6ed1857e0d in do_dm_op (xch=0x23bb160, domid=19, nr_bufs=2)
at xc_private.c:808
808             memcpy(bounce[idx].h, u, size);

xl dmesg doesn't report anything unusual:

(d19) Invoking ROMBIOS ...
(XEN) stdvga.c:174:d19v0 entering stdvga mode
(d19) VGABios $Id: vgabios.c,v 1.67 2008/01/27 09:44:12 vruppert Exp $
(d19) Bochs BIOS - build: 06/23/99
(d19) $Revision: 1.221 $ $Date: 2008/12/07 17:32:29 $
(d19) Options: apmbios pcibios eltorito PMM
(d19)
(d19) ata0 master: QEMU HARDDISK ATA-7 Hard-Disk (10000 MBytes)
(d19)
(d19)
(d19)
(d19) Press F12 for boot menu.
(d19)
(d19) Booting from Hard Disk...
(XEN) stdvga.c:179:d19v0 leaving stdvga mode
(XEN) irq.c:276: Dom19 PCI link 0 changed 5 -> 0
(XEN) irq.c:276: Dom19 PCI link 1 changed 10 -> 0
(XEN) irq.c:276: Dom19 PCI link 2 changed 11 -> 0
(XEN) irq.c:276: Dom19 PCI link 3 changed 5 -> 0

but qemu-dm-win7_32.log has this at the end of it:

ACPI:debug: write addr=0xb044, val=0xfe.
ACPI:debug: write addr=0xb045, val=0x89.
ACPI:debug: write addr=0xb044, val=0xff.
ACPI:debug: write addr=0xb045, val=0x89.
xencall: error: alloc_pages: mmap failed: Cannot allocate memory
track_dirty_vram(f0000000, 240) failed (-1, 12)

I should say that the guest does start booting, I get to the "Loading
Windows" logo (it's a Windows 7) and past that, and just when the mouse
cursor shows up, qemu dies and obviously takes my VNC connection with
it. This is not introspection-related in any way, it happens with a
plain Xen install with no fiddling going on whatsoever.

I've searched in all the FAQs I could think of - perhaps I've missed
something?


Thanks,
Razvan

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