Would changing --cpu-type=detailed to --cpu-type=timing solve your problem?



----- Original Message -----
From: "許順明" <[email protected]>
To: [email protected]
Sent: Tuesday, March 12, 2013 8:58:03 PM
Subject: [gem5-users] I want to use timing mode instead of atomic mode



Hello, after running the PARSEC benchmark, the stats file shows three parts of 
simulated result 





---------- Begin Simulation Statistics ---------- 
sim_seconds 0.026969 # Number of seconds simulated 
sim_ticks 26968606500 # Number of ticks simulated 
final_tick 2288234675000 # Number of ticks from beginning of simulation 
(restored from checkpoin\ 



---------- Begin Simulation Statistics ---------- 
sim_seconds 0.167776 # Number of seconds simulated 
sim_ticks 167776419500 # Number of ticks simulated 
final_tick 2456011094500 # Number of ticks from beginning of simulation 
(restored from checkpoin\ 



sim_seconds 35.613394 # Number of seconds simulated 
sim_ticks 35613394249000 # Number of ticks simulated 
final_tick 38081353711000 # Number of ticks from beginning of simulation 
(restored from checkpoi\ 

the command I used is ./build/ALPHA_MOESI_CMP_directory/gem5.opt 
-d=$benchmark_rcS/cmp_s --debug-flag=Cache --trace-file=cache.trc ./conf\ 
igs/example/fs.py --cpu-type=detailed --kernel=myK --script=$benchmark_rcS 
--caches --l2cache --num-l2caches=4 --l2_size=2MB --l2_assoc=8 -n 4 -F\ 
2346440080500 










I think the first and second parts are using atomic mode, and the third part is 
timing mode. 


I want to test the IPC value in different cache latency, but it seems only the 
timing mode can evaluate the change of IPC 


Could someone help me to using timing mode during whole simulation? 




_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users




CONFIDENTIALITY: Any information contained in this e-mail 
(including attachments) is the property of The State of Texas and 
unauthorized disclosure or use is prohibited. Sending, receiving or 
forwarding of confidential, proprietary and privileged information is 
prohibited under Lamar Policy. If you received this e-mail in error, 
please notify the sender and delete this e-mail from your system.
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to