Sorry, just to complete the previous context, I know that those two
files are inside the checkpoint directory, but it means that it made
just one checkpoint, although I'm calling m5.checkpoint() in a loop.
Does that mean it's over writing the same file every time?! If yes, how
can I change this allocation to have different checkpoint folders?
Thank you again
On 13/05/13 16:25, Negar Miralaei wrote:
Hi,
Thank you Ali, I could run it after passing --at-instruction.
I have another question, please. I'm using my own python code version,
and it doesn't have checkpointing and restoring from checkpoint
options. So, inside the code I call m5.checkpoint(dir). The checkpoint
output directory includes 2 files, "system.physmem.store0.pmem" and
"m5.cpt". There is no chekpoint folder like the ones I gave previously
by using se.py file (e.g. cpt.bzip2_source.179600). How can I restore
from the new checkpoints, and where are they (if any)?
Many thanks
Negar
On 11/05/13 15:20, Ali Saidi wrote:
Hi Negar,
When you restore you need to pass --at-instruction so the code that
is restoring from the checkpoint knows that the format should be by
instruction count and not tick count.
Thanks,
Ali
On May 10, 2013, at 8:14 AM, Negar Miralaei <nm...@cam.ac.uk
<mailto:nm...@cam.ac.uk>> wrote:
Hi,
I'm trying to create a checkpoint at a specific instruction number
and stop execution, then start running from the checkpoint for 100
number of instructions and then exit. Below are the commands I'm
using. Would anybody tell me if I'm running the correct commands?
Since I got the message "Exiting @ tick 281314000 because a thread
reached the max instruction count" for the first command, which I
guessed should be "... because reached the max number of checkpoints
"! And, an error message for restoring as follows.
Command line to create checkpoint:
/build/ARM/gem5.fast -d system/disks/CPU2000/output/
configs/example/se.py --bench bzip2_source --warmup-insts=1000
-cpu-type=atomic --at-instruction --take-checkpoints=179600
--max-checkpoints=1/
Command line to restore from checkpoint and run for 100 instructions:
/build/ARM/gem5.fast -d system/disks/CPU2000/output/
configs/example/se.py --bench bzip2_source -cpu-type=atomic -r
1////--maxinsts=100/
Error:
/Traceback (most recent call last)://
// File "<string>", line 1, in <module>//
// File "/GEM5/gem5/src/python/m5/main.py", line 387, in main//
// exec filecode in scope//
// File "configs/example/se-sim.py", line 235, in <module>//
// Simulation.run(options, root, system, FutureClass)//
// File "/GEM5/gem5/configs/common/Simulation.py", line 425, in run//
// maxtick, checkpoint_dir = findCptDir(options, maxtick, cptdir,
testsys)//
// File "/GEM5/gem5/configs/common/Simulation.py", line 158, in
findCptDir//
// maxtick = maxtick - int(cpts[cpt_num - 1])//
//ValueError: invalid literal for int() with base 10: ''/
Thanks
Negar
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org <mailto:gem5-users@gem5.org>
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
