[Gems-users] A problem when running the Barnes benchmark

["Yu Jiaxin" <yjxhit@xxxxxxxxx>]

Hi all！

 

I want to run the Barnes benchmark in Opal + Ruby. Now I can run it well on 4 processors and 8 processors. But when on 16 and 32 processors, I got the Segmentation fault.

 

16 processors:

 

Segmentation fault (SIGSEGV) in main thread

The simulation state has been corrupted. Simulation cannot continue.

Please restart Simics.

Starting command line. (May have skipped commands in script files.)

[cpu13] v:0x000000000104878c p:0x00002c4878c  jmpl [%l0 + %g0], %o7

Setting new inspection cpu: cpu13

Traceback (most recent call last):

File "/root/gems/gen-scripts/mfacet.py", line 308, in console_branch_internal

    wait_for_string(get_console(), __prompt)

File "/root/gems/simics-3.0.30/x86-linux/lib/python/text_console_common.py", line 10, in wait_for_string

    wait_for_obj_hap("Xterm_Break_String", obj, break_id)

File "/root/gems/simics-3.0.30/x86-linux/lib/python/cli_impl.py", line 3374, in wait_for_obj_hap

    return wait_for_hap_common([hap_name, name, idx0])

File "/root/gems/simics-3.0.30/x86-linux/lib/python/cli_impl.py", line 3352, in wait_for_hap_common raise SimExc_Break, "Script branch interrupted"

sim_core.SimExc_Break: Script branch interrupted

Exception in python branch

 

32 processors:

 

Segmentation fault (SIGSEGV) in main thread

The simulation state has been corrupted. Simulation cannot continue.

Please restart Simics.

Starting command line. (May have skipped commands in script files.)

[cpu32] v:0x00000000000145f4 p:0x0001bc885f4  magic (sethi 0x1, %g0)

Setting new inspection cpu: cpu32

Traceback (most recent call last):

File "/root/gems/gen-scripts/mfacet.py", line 308, in console_branch_internal

    wait_for_string(get_console(), __prompt)

File "/root/gems/simics-3.0.30/x86-linux/lib/python/text_console_common.py", line 10, in wait_for_string

    wait_for_obj_hap("Xterm_Break_String", obj, break_id)

File "/root/gems/simics-3.0.30/x86-linux/lib/python/cli_impl.py", line 3374, in wait_for_obj_hap

    return wait_for_hap_common([hap_name, name, idx0])

File "/root/gems/simics-3.0.30/x86-linux/lib/python/cli_impl.py", line 3352, in wait_for_hap_common raise SimExc_Break, "Script branch interrupted"

sim_core.SimExc_Break: Script branch interrupted

Exception in python branch

 

The scripts is the following:

 

4 processors

@mfacet.run_sim_command('instruction-fetch-mode instruction-fetch-trace')

@mfacet.run_sim_command('istc-disable')

@mfacet.run_sim_command('dstc-disable')

@mfacet.run_sim_command('load-module ruby')

@mfacet.run_sim_command('load-module opal')

@mfacet.run_sim_command('ruby0.setparam g_NUM_PROCESSORS 2')

@mfacet.run_sim_command('ruby0.setparam g_NUM_SMT_THREADS 2')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_ASSOC 4')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_NUM_SETS_BITS 11')

@mfacet.run_sim_command('ruby0.setparam g_DATA_BLOCK_BYTES 64')

@mfacet.run_sim_command('ruby0.setparam NETWORK_LINK_LATENCY 4')

@mfacet.run_sim_command('ruby0.setparam g_FLIT_SIZE 4')

@mfacet.run_sim_command('ruby0.setparam_str g_NETWORK_TOPOLOGY TORUS_2D')

@mfacet.run_sim_command('ruby0.setparam g_endpoint_bandwidth 100')

@mfacet.run_sim_command('opal0.setparam CONFIG_LOGICAL_PER_PHY_PROC 2')

@mfacet.run_sim_command('opal0.setparam CONFIG_IREG_PHYSICAL 384')

@mfacet.run_sim_command('opal0.setparam CONFIG_FPREG_PHYSICAL 256')

@mfacet.run_sim_command('opal0.setparam CONFIG_CCREG_PHYSICAL 74')

@mfacet.run_sim_command('ruby0.init')

@mfacet.run_sim_command('opal0.init')

@mfacet.run_sim_command('opal0.sim-start result-8p.opal')

@mfacet.run_sim_command('opal0.sim-step 1000')

@mfacet.run_sim_command('ruby0.dump-stats %s.stats' % filename_prefix)

quit

 

 

8 processors

@mfacet.run_sim_command('instruction-fetch-mode instruction-fetch-trace')

@mfacet.run_sim_command('istc-disable')

@mfacet.run_sim_command('dstc-disable')

@mfacet.run_sim_command('load-module ruby')

@mfacet.run_sim_command('load-module opal')

@mfacet.run_sim_command('ruby0.setparam g_NUM_PROCESSORS 4')

@mfacet.run_sim_command('ruby0.setparam g_NUM_SMT_THREADS 2')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_ASSOC 4')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_NUM_SETS_BITS 11')

@mfacet.run_sim_command('ruby0.setparam g_DATA_BLOCK_BYTES 64')

@mfacet.run_sim_command('ruby0.setparam NETWORK_LINK_LATENCY 4')

@mfacet.run_sim_command('ruby0.setparam g_FLIT_SIZE 4')

@mfacet.run_sim_command('ruby0.setparam_str g_NETWORK_TOPOLOGY TORUS_2D')

@mfacet.run_sim_command('ruby0.setparam g_endpoint_bandwidth 100')

@mfacet.run_sim_command('opal0.setparam CONFIG_LOGICAL_PER_PHY_PROC 2')

@mfacet.run_sim_command('opal0.setparam CONFIG_IREG_PHYSICAL 384')

@mfacet.run_sim_command('opal0.setparam CONFIG_FPREG_PHYSICAL 256')

@mfacet.run_sim_command('opal0.setparam CONFIG_CCREG_PHYSICAL 74')

@mfacet.run_sim_command('ruby0.init')

@mfacet.run_sim_command('opal0.init')

@mfacet.run_sim_command('opal0.sim-start result-8p.opal')

@mfacet.run_sim_command('opal0.sim-step 1000')

@mfacet.run_sim_command('ruby0.dump-stats %s.stats' % filename_prefix)

quit

 

16 processors

@mfacet.run_sim_command('instruction-fetch-mode instruction-fetch-trace')

@mfacet.run_sim_command('istc-disable')

@mfacet.run_sim_command('dstc-disable')

@mfacet.run_sim_command('load-module ruby')

@mfacet.run_sim_command('load-module opal')

@mfacet.run_sim_command('ruby0.setparam g_NUM_PROCESSORS 8')

#@mfacet.run_sim_command('ruby0.setparam g_NUM_SMT_THREADS 2')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_ASSOC 4')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_NUM_SETS_BITS 11')

@mfacet.run_sim_command('ruby0.setparam g_DATA_BLOCK_BYTES 64')

@mfacet.run_sim_command('ruby0.setparam NETWORK_LINK_LATENCY 4')

@mfacet.run_sim_command('ruby0.setparam g_FLIT_SIZE 4')

@mfacet.run_sim_command('ruby0.setparam_str g_NETWORK_TOPOLOGY TORUS_2D')

@mfacet.run_sim_command('ruby0.setparam g_endpoint_bandwidth 100')

@mfacet.run_sim_command('opal0.setparam CONFIG_LOGICAL_PER_PHY_PROC 2')

@mfacet.run_sim_command('opal0.setparam CONFIG_IREG_PHYSICAL 384')

@mfacet.run_sim_command('opal0.setparam CONFIG_FPREG_PHYSICAL 256')

@mfacet.run_sim_command('opal0.setparam CONFIG_CCREG_PHYSICAL 74')

@mfacet.run_sim_command('ruby0.init')

@mfacet.run_sim_command('opal0.init')

@mfacet.run_sim_command('opal0.sim-start result-16p.opal')

@mfacet.run_sim_command('opal0.sim-step 1000')

@mfacet.run_sim_command('ruby0.dump-stats %s.stats' % filename_prefix)

quit

 

32 processors

@mfacet.run_sim_command('instruction-fetch-mode instruction-fetch-trace')

@mfacet.run_sim_command('istc-disable')

@mfacet.run_sim_command('dstc-disable')

@mfacet.run_sim_command('load-module ruby')

@mfacet.run_sim_command('load-module opal')

@mfacet.run_sim_command('ruby0.setparam g_NUM_PROCESSORS 16')

#@mfacet.run_sim_command('ruby0.setparam g_NUM_SMT_THREADS 2')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_ASSOC 4')

@mfacet.run_sim_command('ruby0.setparam L2_CACHE_NUM_SETS_BITS 11')

@mfacet.run_sim_command('ruby0.setparam g_DATA_BLOCK_BYTES 64')

@mfacet.run_sim_command('ruby0.setparam NETWORK_LINK_LATENCY 4')

@mfacet.run_sim_command('ruby0.setparam g_FLIT_SIZE 4')

@mfacet.run_sim_command('ruby0.setparam_str g_NETWORK_TOPOLOGY TORUS_2D')

@mfacet.run_sim_command('ruby0.setparam g_endpoint_bandwidth 100')

@mfacet.run_sim_command('opal0.setparam CONFIG_LOGICAL_PER_PHY_PROC 2')

@mfacet.run_sim_command('opal0.setparam CONFIG_IREG_PHYSICAL 384')

@mfacet.run_sim_command('opal0.setparam CONFIG_FPREG_PHYSICAL 256')

@mfacet.run_sim_command('opal0.setparam CONFIG_CCREG_PHYSICAL 74')

@mfacet.run_sim_command('ruby0.init')

@mfacet.run_sim_command('opal0.init')

@mfacet.run_sim_command('opal0.sim-start result-16p.opal')

@mfacet.run_sim_command('opal0.sim-step 1000')

@mfacet.run_sim_command('ruby0.dump-stats %s.stats' % filename_prefix)

quit

 

Thank you very much!

2011-05-23

---

Yu Jiaxin