Re: [PATCH] Assorted VMX patches

[Kip Macy <kip.macy@gmail.com>]

Core was generated by `gdbserver-xen curly:2345 --attach 5'.
Program terminated with signal 11, Segmentation fault.

warning: current_sos: Can't read pathname for load map: Input/output error

Reading symbols from /usr/lib/libxc.so.3.0...done.
Loaded symbols for /usr/lib/libxc.so.3.0
Reading symbols from /usr/lib/libxutil.so.3.0...done.
Loaded symbols for /usr/lib/libxutil.so.3.0
Reading symbols from /lib/i686/libc.so.6...done.
Loaded symbols for /lib/i686/libc.so.6
Reading symbols from /usr/lib/libz.so.1...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
#0  0xb7fc74e5 in map_domain_va (domid=0x5, cpu=0x0,
guest_va=0xc018e8bc, perm=0x1) at xc_ptrace.c:184
184         pde = page_array[pde >> PAGE_SHIFT] << PAGE_SHIFT;

     -Kip

On 4/25/05, Leendert van Doorn <leendert@watson.ibm.com> wrote:
> 
> # Have you tried the modified xc_ptrace on a paravirtualized VM? I don't
> # think that the machine address read out of the page tables would index
> # correctly into the array of pages which, presumably, are in physical
> # order. I think the changes in map_guest_va should probably be
> # conditional on the guest being vmx.
> 
> You are right, I didn't try the gdbserver patches on a paravirtualized VM,
> but I would expect xc_get_pfn_list to return an identity map in that case
> so that the dereferences still work.
> 
> If it doesn't, then you need to test the flags for ECF_VMX_GUEST and do
> the dereferences conditionally.
> 
>         Leendert
> 
>