Hi Ali, all,

I am running the streamcluster ( a multi-threaded application) on a
single-core, in FS mode, AtomicCPU. And, i configure the application (using
NTHREAD environment variable) to run using a single thread.

The application does a pthread_barrier_wait. Since there is a single thread
all over, do you see the pthread_barrier_wait invocation to lock-up or so?
In which case, should i be running the application with NTHREAD=2 (any
value greater than 1)? Kindly help me with pointers.

BR/Nizam


On Mon, Jan 11, 2016 at 8:49 AM, Nizamudheen Ahmed <[email protected]>
wrote:

> Hi Ali,
>
> I am running on Atomic CPU mode. I am running the streamcluster kernel (in
> PARSEC). It has been running for the past 12 hours and not completed (may
> be already in the ROI stuff). I wanted to check if that kind of
> simulation-time is expected in GEM5.
>
> BR/Nizam
>
> On Mon, Jan 11, 2016 at 3:00 AM, Ali Saidi <[email protected]> wrote:
>
>> What model are you using to get to the ROI? Typically people get to the
>> ROI with the atomic cpu and then simulate the ROI with the o3 CPU.
>>
>> Ali
>>
>> On Jan 10, 2016, at 7:06 AM, Nizamudheen Ahmed <[email protected]>
>> wrote:
>>
>> Hi champs
>>
>> I am trying to run parsec benchmarks for 'simlarge' configuration on gem5
>> in FS mode on ARM ISA. It seems to take awful lot of time to even enter the
>> ROI. I ran it for 7 hours or so and I am yet to encounter the ROI for
>> streamcluster kernel. Is this expected? What is the typical time expected
>> to run this kernel? Should I explore simpoints?
>> Thanks in advance.
>> BR/Nizam
>> _______________________________________________
>> gem5-users mailing list
>> [email protected]
>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>>
>>
>>
>> _______________________________________________
>> gem5-users mailing list
>> [email protected]
>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>>
>
>
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to