KVM segfaults with a specific disk image.

[Evan]

Binary package hint: kvm

Using kvm 1:84+dfsg-0ubuntu11 (for what it's worth, it happens when using -hda as well). Curiously, the image works fine when written to a USB disk.

Starting program: /usr/bin/kvm -m 768 -drive file=ubuntu-usb-9.04.img,if=ide,boot=on
[Thread debugging using libthread_db enabled]
[New Thread 0x7f4e55167740 (LWP 21942)]
[New Thread 0x7f4e1f7e7950 (LWP 21945)]
[New Thread 0x7f4e1e283950 (LWP 21946)]
[Thread 0x7f4e1e283950 (LWP 21946) exited]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7f4e1f7e7950 (LWP 21945)]
qemu_paio_cancel (fd=<value optimized out>, aiocb=0x20ed270)
    at posix-aio-compat.c:184
184 posix-aio-compat.c: No such file or directory.
 in posix-aio-compat.c
(gdb) bt
#0 qemu_paio_cancel (fd=<value optimized out>, aiocb=0x20ed270)
    at posix-aio-compat.c:184
#1 0x000000000041c0d8 in raw_aio_cancel (blockacb=<value optimized out>)
    at block-raw-posix.c:681
#2 0x0000000000435380 in ide_dma_cancel (bm=0x217ee60)
    at /build/buildd/kvm-84+dfsg/qemu/hw/ide.c:2973
#3 0x00000000004353e5 in bmdma_cmd_writeb (opaque=0x217ee60, addr=0, val=0)
    at /build/buildd/kvm-84+dfsg/qemu/hw/ide.c:2987
#4 0x000000000052757d in kvm_outb (opaque=<value optimized out>, addr=0,
    data=0 '\0') at /build/buildd/kvm-84+dfsg/qemu/qemu-kvm.c:684
#5 0x0000000000553d45 in kvm_run (kvm=0x1da3040, vcpu=0, env=0x1da9980)
    at libkvm.c:722
#6 0x0000000000527d09 in kvm_cpu_exec (env=<value optimized out>)
    at /build/buildd/kvm-84+dfsg/qemu/qemu-kvm.c:205
#7 0x0000000000528008 in ap_main_loop (_env=<value optimized out>)
    at /build/buildd/kvm-84+dfsg/qemu/qemu-kvm.c:414
#8 0x00007f4e541283ba in start_thread () from /lib/libpthread.so.0
#9 0x00007f4e532adfcd in clone () from /lib/libc.so.6
#10 0x0000000000000000 in ?? ()

[Soren Hansen (soren)]

Can you give me the output of "file ubuntu-usb-9.04.img" and also tell me whether this happens immediately or after a while using the virtual machine?

[Anthony Liguori (anthony-codemonkey)]

I suspect this is a dup of 359447

[Dustin Kirkland  (kirkland)]

Evan-

Still an issue? Can you give us some feedback?

:-Dustin