On Mon, Jun 3, 2013 at 12:59 PM, Nilay Vaish <ni...@cs.wisc.edu> wrote:

> On Mon, 3 Jun 2013, Joel Hestness wrote:
>
>  Hi Nilay,
>>  You're correct about revision 9603.  I've tested this with more revisions
>> and I've found that the system hang exists at least back to changeset
>> 8969.
>> Earlier revisions fail for other reasons that have since been fixed.
>>
>> *@All:* I'm pretty surprised by this... Has anyone been successfully using
>>
>> this configuration (x86 FS with the classic memory model, multicore O3
>> CPUs
>> and checkpoint restore)?
>>
>>  Thanks,
>>  Joel
>>
>>
>> On Fri, May 31, 2013 at 4:37 PM, Nilay Vaish <ni...@cs.wisc.edu> wrote:
>>
>>  To start with, I suggest that you compare the executions of 9603 and 9606
>>> since I expect them to produce the same output if you are not using ruby,
>>> which is true in your case. The checkpoints produced for the two versions
>>> should be the same and the execution afterwords should also proceed along
>>> the same lines.
>>>
>>>
> Are you saying that 9606 works correctly?
>

No, 9606 also hangs.  I've tested 9 different revisions, evenly spread from
8969 to 9703, and they all witness the system hang.  As such, I don't have
a baseline revision that actually works correctly.

My surprise is that this bug has been around for at least a year, but it
appears that no one is using these system configurations or inquired about
the bug before.

  Joel

-- 
  Joel Hestness
  PhD Student, Computer Architecture
  Dept. of Computer Science, University of Wisconsin - Madison
  http://pages.cs.wisc.edu/~hestness/
_______________________________________________
gem5-dev mailing list
gem5-dev@gem5.org
http://m5sim.org/mailman/listinfo/gem5-dev

Reply via email to