[Barrelfish-users] Kernel panic using qemu

[Calum McCall]

I get the following error message using qemu v0.10.5, GCC v4.3.5,  and GHCv6.10.4  and ld v2.17.50.0.6-14.el5

root (nd)
  Filesystem type is tftp, using whole disk
kernel /x86_64/sbin/elver loglevel=4
    [Multiboot-elf, <0x100000:0x21ea:0x0>, <0x1031ec:0x1c:0x6df8>, shtab=0x10a2f
8, entry=0x10000c]
module /x86_64/sbin/cpu loglevel=4
    [Multiboot-module @ 0x114000, 0xa9d24 bytes]
module /x86_64/sbin/init
    [Multiboot-module @ 0x1be000, 0x28b362 bytes]
module /x86_64/sbin/mem_serv
    [Multiboot-module @ 0x44a000, 0x25d678 bytes]
module /x86_64/sbin/monitor
    [Multiboot-module @ 0x6a8000, 0x421622 bytes]
module /x86_64/sbin/chips boot
    [Multiboot-module @ 0xaca000, 0x254432 bytes]
module  /x86_64/sbin/ramfsd boot
    [Multiboot-module @ 0xd1f000, 0x2bcf04 bytes]
module  /x86_64/sbin/skb boot
    [Multiboot-module @ 0xfdc000, 0x690828 bytes]
modulenounzip /skb_ramfs.cpio.gz nospawn
    [Multiboot-module @ 0x166d000, 0x9c338 bytes]
module  /x86_64/sbin/pci boot
    [Multiboot-module @ 0x170a000, 0x49d298 bytes]
module  /x86_64/sbin/spawnd boot
    [Multiboot-module @ 0x1ba8000, 0x3d101d bytes]
module /x86_64/sbin/serial
    [Multiboot-module @ 0x1f7a000, 0x2847ed bytes]
module /x86_64/sbin/fish
    [Multiboot-module @ 0x21ff000, 0x3b5b94 bytes]

Kernel starting at address 0xffffff80025b6000
elf_relocate: relocation 76 type 6011040
kernel 0 PANIC! kernel assertion "!"Unimplemented: Cannot handle relocation type"" failed at ../lib/elf/elf64.c:218
kernel 0: Cannot trap into debugger -- Interrupts not set up yet!

I had to remove the parameter "-Wno-packed-bitfield-compat" from hake/X86_64.hs as this is not supported
by GCC v4.3.5. Could this be causing the issue I'm seeing above? The readme does say that my version of GCC
is supported.

I also had to remove the "--build-id=none" flag from tools/elver/Hakefile as my version of ld does not
support it. But this seems to only add extra debugging info so I assume that's harmless.

Cheers,

-Calum