Hi,

I have seen in the mailing list (and I have tried it myself) that
checkpointing in Ruby is only possible with ALPHA_MOESI_hammer protocol.
However I don't know about the other protocols.

Best,



On Thu, Jun 27, 2013 at 6:48 PM, tejasi pimpalkhute <[email protected]>wrote:

> Hi,
>
> Actually, even I get the same error as Amina but while running in se mode
> when I use --fast-forward option.
> The command line I am using is:
>
> ./build/ALPHA_MESI_CMP_directory/gem5.opt configs/example/se_modular.py -n
> 16 --cpu-type=detailed --clock=1GHz --topology=MeshDirSolo --caches
> --cacheline_size=128 --l2cache --num-l2caches=16 --l1d_size=16kB
> --l1i_size=16kB --l2_size=128kB --l1d_assoc=2 --l1i_assoc=2 --l2_assoc=2
> --num-dirs=1 --mesh-rows=4 --ruby --garnet-network=fixed --rootdir="spec2k"
> --benchmark="gcc;mgrid;gcc;mgrid;gcc;mgrid;gcc;mgrid;gcc;mgrid;gcc;mgrid;gcc;mgrid;gcc;mgrid"
> -I 40000000 -F 10000000
>
> I get the error:
>  panic: RubyPort::M5Port::recvAtomic() not implemented!
>
> Is this because cpu switching happens from atomic to detailed in case of
> fast forwarding or checkpointing and ruby does not support atomic accesses?
> If this is correct, is there any workaround for this? I need to use fast
> forwarding in my simulations.
> Please throw some light on this.
>
>
> On Thu, Jun 27, 2013 at 10:10 AM, Erfan Azarkhish 
> <[email protected]>wrote:
>
>> Hi,
>>
>> Can you try building the system with this configuration:
>> ALPHA_MOESI_hammer? it may fix the problem
>>
>> Best,
>>
>>
>> On Thu, Jun 27, 2013 at 1:07 PM, amina belhaj messaoud <
>> [email protected]> wrote:
>>
>>> hello
>>> I am running this command :
>>> build/ALPHA/gem5.opt  configs/example/ruby_fs.py   --topology=Mesh
>>> --garnet-network=fixed
>>> --kernel=/home/amina/Gem5/FSfILES/m5_system_2.0b3/binaries/vmlinux
>>> --disk-image=/home/amina/Gem5/FSfILES/m5_system_2.0b3/disks/linux-latest.img
>>> --script=configs/boot/hello.rcS  --num-cpu=2 --num-dirs=2 --cpu-type=timing
>>> --mesh-rows=2 --caches --l2cache --num-l2cache=2
>>> --checkpoint-dir=/home/amina/gemo/checks
>>>
>>> hello.rcS contains :
>>>
>>> /sbin/m5 checkpoint
>>> echo hello
>>> ls
>>>
>>> echo hello again
>>>
>>> ls
>>> /sbin/m5 exit
>>>  but when  after that the checkpoint is benn writen  , the system
>>> aborted
>>> with message : panic  : possible deadlock detected
>>>
>>> when I want to restore from the checkpoint :
>>>  build/ALPHA/gem5.opt configs/example/ruby_fs.py --topology=Mesh
>>> --garnet-network=fixed
>>> --kernel=/home/amina/Gem5/FSfILES/m5_system_2.0b3/binaries/vmlinux
>>> --disk-image=/home/amina/Gem5/FSfILES/m5_system_2.0b3/disks/linux-latest.img
>>> --script=configs/boot/hello.rcS --num-cpu=2 --num-dirs=2 --mesh-rows=2
>>> --caches --l2cache --num-l2cache=2 --checkpoint-dir=/home/amina/gemo/checks
>>> -r 1
>>>
>>> I have this :
>>>  panic: RubyPort::M5Port::recvAtomic() not implemented!
>>>  @ cycle 2438047479500
>>> [recvAtomic:build/ALPHA/mem/ruby/system/RubyPort.cc, line 139]
>>> Memory Usage: 703072 KBytes
>>> Program aborted at cycle 2438047479500
>>> Aborted
>>>
>>> PLEASE HELP !!
>>>
>>> amina
>>>
>>>
>>>
>>> _______________________________________________
>>> gem5-users mailing list
>>> [email protected]
>>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>>>
>>
>>
>>
>> --
>> Erfan Azarkhish
>> Micrel Lab - Viale Carlo Pepoli 3/2 - 40123, Bologna
>> DEIS - University of Bologna, Italy
>>
>> _______________________________________________
>> gem5-users mailing list
>> [email protected]
>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>>
>
>
>
> --
> Thanks and Regards,
> Tejasi
>
> _______________________________________________
> gem5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>



-- 
Erfan Azarkhish
Micrel Lab - Viale Carlo Pepoli 3/2 - 40123, Bologna
DEIS - University of Bologna, Italy
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to