Re: [Gems-users] TM Simulation (SIMICS SEG FAULT)

[BYONG WU CHONG <bernard.chong@xxxxxxxx>]

I got pass this error by 
1. moving start of ruby after a barrier by using Barrier_break() function.
2. moving set_transaction_registers() after starting ruby.
   So this is strictly same as http://www.cs.wisc.edu/gems/doc/gems-wiki/moin.cgi/TM_Workload_Setup
   I should not have taken this steps lightly.
3. Using "psradm -i 0 1 2 .. n" after setting processor cabinets.

I am having additional error of 
1. Segmentation Fault on barnes LL TM simulation
2. Bus Error on barnes_debug LL TM simulation
3. Assertion error on barnes and barnes_debug EE TM simulation

The assertion error log is following
-------------------- CLI ERROR LOG START --------------------
memoryBits: 3 memorySizeBits: 32 Address: [0x8796d1040, line 0x8796d1040]error: limit exceeded.  dataBlockBits: 6 memoryModuleBlocks: 8388608 index: 71087752
failed assertion 'index < RubyConfig::memoryModuleBlocks()' at fn integer_t Address::memoryModuleIndex() const in common/Address.h:210
failed assertion 'index < RubyConfig::memoryModuleBlocks()' at fn integer_t Address::memoryModuleIndex() const in common/Address.h:210
At this point you might want to attach a debug to the running and get to the
crash site; otherwise press enter to continue
PID: 8176
-------------------- CLI ERROR LOG END --------------------

However, this error seems independent of the problems that I had and it seems that above three solutions solved this initial problem, so I will start a new post.

Thanks for reading this.


-----Original Message-----
From: gems-users-bounces@xxxxxxxxxxx [mailto:gems-users-bounces@xxxxxxxxxxx] On Behalf Of BYONG WU CHONG
Sent: Friday, March 27, 2009 8:17 PM
To: Gems Users
Subject: [Gems-users] TM Simulation (SIMICS SEG FAULT)

Hi, I was running LL TM from GEMS 2.1 and I got this SIMIC SEG FAULT error.

...
5145618 3 [3,0] ISOLATE XACT LOAD VA: [0xfee7b980, line 0xfee7b980] PA: [0x7690f980, line 0x7690f980] XACT LEVEL: 1 PC = [0xff235590, line 0xff235580]
5145618 3 [3,0] REQUESTING DATA FROM BUFFER [0x7690f9a8, line 0x7690f980]
 5145618 3 [3,0] FORWARD FROM WRITE BUFFER [0x7690f9a8, line 0x7690f980] 0x0
 5145618 3 [3,0] FORWARD FROM WRITE BUFFER [0x7690f9a9, line 0x7690f980] 0x1
 5145618 3 [3,0] FORWARD FROM WRITE BUFFER [0x7690f9aa, line 0x7690f980] 0x9f
 5145618 3 [3,0] FORWARD FROM WRITE BUFFER [0x7690f9ab, line 0x7690f980] 0x1c
 5145618 3 [3,0] FROM WRITE BUFFER [0x7690f9a8, line 0x7690f980] [ 0x0 0x1 0x9f 0x1c  ] 1
SIMICS SEG FAULT  for thread 0 of proc 3
SIMICS SEG FAULT

This is the whole dump of the CLI window with XACT_DEBUG_LEVEL = 9
http://www.cs.utah.edu/~bchong/CA/barnes-LL-04-debug-CLIout.txt

This is simulation's output, which I don't think it helps much.
http://www.cs.utah.edu/~bchong/CA/SPLASH2_barnes_LLTM_04_debug.out

This is the LL TM python script I copied and modified from ATMTP's (but I disabled all ATMTP definitions).
http://www.cs.utah.edu/~bchong/CA/tm_ll.py

This is the actual Simics script I used.
http://www.cs.utah.edu/~bchong/CA/barnes-LL-04-debug.simics


I got the bindings right, and I think I got the initializations right according to this thread,
https://lists.cs.wisc.edu/archive/gems-users/2008-March/msg00030.shtml
but it seems something is wrong.


Please help me to bug this error.

_______________________________________________
Gems-users mailing list
Gems-users@xxxxxxxxxxx
https://lists.cs.wisc.edu/mailman/listinfo/gems-users
Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/"; to your search.