[Android-virt] Corrupt kernel loading with SEMIHOSTING?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 20 February 2012 06:27, Rusty Russell <rusty at rustcorp.com.au> wrote:
> Hi Peter,
>
> ? ? ? ?I've spent a fun few hours, where booting under the
> modeldebugger works fine, but the isim_system fails uncompressing the
> kernel. ?I assumed I'd broken something in boot_wrapper and hyp mode,
> but reverting that gave the same thing.

By 'works under model debugger' do you mean:
 * works if all single stepped
 * works if you just hit run
 * works when model debugger is debugging the isim_system
 * works only if model debugger is debugging a DLL model
?

If all the parameters are the same then it should be equivalent.
The only thing that springs to mind is that host load addresses will
end up different and maybe I messed something up in the semihosting
code with a signed/unsigned comparison?

There's some tracing plugin support for the models which you should
be able to use to dump an instruction trace in both cases and
compare, which is probably just about tractable for a problem this
early in boot.

-- PMM



[Index of Archives]     [Linux KVM]     [Spice Development]     [Libvirt]     [Libvirt Users]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux