Re: [Qemu-ppc] qemu-system-ppc64 -M ppce500 always booting to qemu monitor console plus no output on serial console

[Alexander Graf]

On 12/16/2015 05:07 PM, Do Vincenzo wrote:
> On 16.12.15  16:42, Alexander Graf wrote:
> > On 12/16/2015 03:26 PM, Do Vincenzo wrote:
> > > On 16.12.15  14:38, Alexander Graf wrote:
> > > > Please don't top post on this mailing list :). Just write your replies
> > > > below the previous ones at the place the original question was.
> > > Ok got it, I'm just not that used to sending e-mails on these lists.
> > >
> > > > Yes, the option is correct. So that's not the problem. One thing you
> > > > could try is to pass the uImage into -kernel rather than the kernel.
> > > So I've already tried to use the uImage but still no output.
> > >
> > > > If that doesn't help, we need to dive into some debugging :). Boot the
> > > > VM with the -s -S parameters. Then on a different shell do
> > > > $ gdb vmlinux -ex 'target remote localhost:1234'
> > > > In the following gdb shell, type
> > > > (gdb) c
> > > > after a bit, Ctrl-C and checl where you are
> > > > (gdb) bt
> > > > Also try to print the log buffer:
> > > > (gdb) x /200c __log_buf
> > > > That should give some hints on what's going wrong.
> > > > Alex
> > > Here's the log buffer after the steps you described: 
> > > http://paste.ubuntu.com/14050259/
> > This shows all zeros, so we didn't get far. Which made me realize
> > another thing missing from your command line: the CPU type.
> >
> > Please add -cpu e5500 to the QEMU parameters. If that doesn't help, run
> > without KVM and add -D log -d in_asm,cpu,int and look at the "log" file.
> > Up to which point does booting get?
> >
> > Alex
> I've added the cpu parameter as you said, still no output.
> The logfile created with the -D option keeps getting bigger, I've stopped it 
> after ~30s, you can find here the first 3000 lines.
>
> log: http://paste.ubuntu.com/14051480/
>
> I couldn't understand most of these lines but I see lots of zeros which doesn't 
> seem to be a good thing.
> Let me know if you need more details.

It's doing a lot of things - all the way to the end of the paste above. 
Check for the last IN: occurence and check with gdb

  (gdb) l *0x1234
  (gdb) x /i 0x1234

on the address you see in the IN: block where you're getting stuck. 
Also, try the backtrace again with the cpu parameter passed in.


Alex