[Gems-users] Simulation time of GEMS2.0


Date: Mon, 12 Nov 2007 18:45:37 +0900
From: Susumu Takeda <takeda@xxxxxxxxxxxxxxxxxxxxx>
Subject: [Gems-users] Simulation time of GEMS2.0
Dear list,

I am executing deque with GEMS2.0. 
(The protocol is MESI_CMP_filter_directory.)
Execution ended at one hour when USER_MODE_DATA_ONLY was set to true. 
However, execution doesn't end after one week when making it to false. 
Does it take time like this in GEMS2.0?
Or, do I have the problem?

The execution environment is as follows. 

	CPU : Intel Xeon 3GHz
	MEMORY : 8GB
	OS : CentOS 4

	simics version 2.2.19
	
	ruby0.setparam_str PROFILE_EXCEPTIONS true
	ruby0.setparam_str PROFILE_XACT true
	ruby0.setparam_str PROFILE_NONXACT false
	ruby0.setparam_str XACT_DEBUG true
	ruby0.setparam XACT_DEBUG_LEVEL 3
	ruby0.setparam_str XACT_MEMORY true
	ruby0.setparam_str XACT_ENABLE_TOURMALINE false
	ruby0.setparam_str XACT_ISOLATION_CHECK false
	ruby0.setparam_str PERFECT_FILTER true
	ruby0.setparam_str READ_WRITE_FILTER Perfect_
	ruby0.setparam_str PERFECT_VIRTUAL_FILTER true
	ruby0.setparam_str VIRTUAL_READ_WRITE_FILTER Perfect_
	ruby0.setparam_str PERFECT_SUMMARY_FILTER true
	ruby0.setparam_str SUMMARY_READ_WRITE_FILTER Perfect_
	ruby0.setparam_str XACT_EAGER_CD true
	ruby0.setparam_str XACT_LAZY_VM false
	ruby0.setparam_str XACT_CONFLICT_RES TIMESTAMP
	ruby0.setparam XACT_COMMIT_TOKEN_LATENCY 0
	ruby0.setparam_str XACT_VISUALIZER true
	ruby0.setparam_str XACT_NO_BACKOFF false
	ruby0.setparam XACT_LOG_BUFFER_SIZE 0
	ruby0.setparam XACT_STORE_PREDICTOR_ENTRIES 256
	ruby0.setparam XACT_STORE_PREDICTOR_HISTORY 256
	ruby0.setparam XACT_STORE_PREDICTOR_THRESHOLD 4
	ruby0.setparam XACT_FIRST_ACCESS_COST 0
	ruby0.setparam XACT_FIRST_PAGE_ACCESS_COST 0
	ruby0.setparam_str ENABLE_MAGIC_WAITING false
	ruby0.setparam_str ENABLE_WATCHPOINT false
	ruby0.setparam_str XACT_ENABLE_VIRTUALIZATION_LOGTM_SE true


Thanks in advance,

-- 
*****************************************
Susumu TAKEDA
Kyushu University
E-mail:takeda@xxxxxxxxxxxxxxxxxxxxx
*****************************************
[← Prev in Thread] Current Thread [Next in Thread→]
  • [Gems-users] Simulation time of GEMS2.0, Susumu Takeda <=