# Invocation command line: # /root/work/cpu2006/bin/runspec -c amd1002li-rate-revC.cfg --rate --define autopar -T all fp --nopreenv --note-preenv # output_root was not used for this run ############################################################################ # AMD SPEC CPU2006 V1.1 Rate Configuration File for 64-bit Linux # # CPU2006 Version 1.1 # Compiler name/version: Open64 4.2.4 # Operating system version: 64-bit Linux # Hardware: AMD Opteron (Lisbon) # FP Base Pointer Size: 64-bit only # FP Peak Pointer Size: 32/64-bit # INT Base Pointer Size: 32/64-bit # INT Peak Pointer Size: 32/64-bit # Auto Parallization: Yes # # Important! Please run with your stack size set to 'unlimited'. # Failure to do so may cause 483.xalancbmk to get a stack overflow during # execution. # Using csh: unlimit # Using bash: ulimit -s unlimited # # Your LD_LIBRARY_PATH must be set to the location of the dependent # runtime libraries. (Though this should be taken care of by the # preENV_LD_LIBRARY_PATH found in the AMD1002li-rate-revX.inc file.) # This line should be commented out during builds. # ##################################################################### # Although many settings are different, Lisbon uses the same binaries # as Magny-Cours ext = amd1002mc-rate-revC ignore_errors = no tune = base,peak output_format = all flagsurl000 = http://www.spec.org/cpu2006/flags/x86-open64-424-flags-rate-revC.20101109.xml $[top]/amd-platform-rate-revC.xml size = test,train,ref check_md5 = yes reportable = yes env_vars = no teeout = yes mean_anyway = yes teerunout = yes post_setup = sync #################################################################### # Include file containing the commonly changed fields #################################################################### %define inc_file_name AMD1002li-rate-revC.inc #include: %{inc_file_name} # ----- Begin inclusion of 'AMD1002li-rate-revC.inc' ############################################################################ ##################################################################### # Paths ##################################################################### # Set location of run time libraries for runs. preENV_LD_LIBRARY_PATH = $[top]/amd1002-rate-libs-revC/64:$[top]/amd1002-rate-libs-revC/32 # Set this path to your SmartHeap Library for builds. SMARTHEAP_DIR = /root/work/libraries/SmartHeap-8.1/lib ##################################################################### # Machine Specific Settings ##################################################################### # If your machine orders nodes sequentially, you have 16GB/socket # memory using 8GB DIMMS, and your system has huge pages configured, # most settings will be correctly configured based on the below two # settings. See README.amd1002li-rate for more information. # How many sockets does your test system have? (1 or 2) %define num_sockets 1 # How many cores per socket does your system have? (4 or 6) %define cores_per_socket 6 ######################################################################## # You should not need to change anything in this section unless you have # a non-standard configuration. See README.amd1002li-rate for more # information. ######################################################################## %if ('%{cores_per_socket}' eq '6') && ('%{num_sockets}' eq '1') % define num_system_huge_pages 2700 % define cores_per_node 6 % define mem_size 16 % define mem_sticks 2 %elif ('%{cores_per_socket}' eq '6') && ('%{num_sockets}' eq '2') % define num_system_huge_pages 5400 % define cores_per_node 6 % define mem_size 32 % define mem_sticks 4 %elif ('%{cores_per_socket}' eq '4') && ('%{num_sockets}' eq '1') % define num_system_huge_pages 3584 % define cores_per_node 4 % define mem_size 16 % define mem_sticks 2 %elif ('%{cores_per_socket}' eq '4') && ('%{num_sockets}' eq '2') % define num_system_huge_pages 7168 % define cores_per_node 4 % define mem_size 32 % define mem_sticks 4 %endif #################################################################### # Tester information #################################################################### license_num = 49 prepared_by = AMD Performance Lab tester = Advanced Micro Devices test_sponsor = Advanced Micro Devices hw_vendor = Tyan hw_model000 = Tyan YR190B8228, hw_model001 = AMD Opteron 4170 HE #################################################################### # Hardware information #################################################################### hw_avail = Aug-2010 hw_cpu_name = AMD Opteron 4170 HE hw_cpu_mhz = 2100 hw_fpu = Integrated hw_nchips = 1 hw_ncores = 6 hw_ncoresperchip = 6 hw_nthreadspercore = 1 hw_ncpuorder = 1,2 chips hw_pcache = 64 KB I + 64 KB D on chip per core hw_scache = 512 KB I+D on chip per core hw_tcache = 6 MB I+D on chip per chip hw_ocache = None hw_disk000 = 1 x 128 GB SATA SSD hw_disk001 = Crucial RealSSD C300 CTFDDAC128MAG-1G1 hw_memory = 16 GB (2 x 8 GB 2Rx4 PC3-10600R-9, ECC) hw_other = None sw_file = ext3 #sw_os000 = Red Hat Enterprise Linux Server release 5.4, #sw_os001 = Kernel 2.6.18-164.9.1.el5 sw_os000 = SUSE Linux Enterprise Server 11 (x86_64), sw_os001 = Kernel 2.6.27.19-5-default sw_state = Run level 3 (Full multiuser with network) ##################################################################### # Notes ##################################################################### notes_025 =The x86 Open64 Compiler Suite is only available from (and supported by) AMD at notes_030 =http://developer.amd.com/cpu/open64 notes_submit_000 ='numactl' was used to bind copies to the cores. notes_submit_005 =See the configuration file for details. notes_os_000 ='ulimit -s unlimited' was used to set environment stack size notes_os_005 ='ulimit -l 2097152' was used to set environment locked pages in memory limit notes_os_010 = notes_os_015 =Set vm/nr_hugepages=2700 in /etc/sysctl.conf notes_os_020 =mount -t hugetlbfs nodev /mnt/hugepages # Only for SLES 11 and later notes_os_025 = notes_os_030 =Binaries were compiled on SLES10 SP2 with binutils 2.18 # ---- End inclusion of '/root/work/cpu2006/config/AMD1002li-rate-revC.inc' #################################################### # Auto-configured settings based on your .inc file # #################################################### %if '%{cores_per_socket}' eq '6' # Config settings for 6-core systems % ifdef %{autopar} preENV_OMP_NUM_THREADS = 6 % endif #autopar preENV_HUGETLB_LIMIT = 450 bind0 = numactl -m 0 --physcpubind=0 bind1 = numactl -m 0 --physcpubind=1 bind2 = numactl -m 0 --physcpubind=2 bind3 = numactl -m 0 --physcpubind=3 bind4 = numactl -m 0 --physcpubind=4 bind5 = numactl -m 0 --physcpubind=5 bind6 = numactl -m 1 --physcpubind=6 bind7 = numactl -m 1 --physcpubind=7 bind8 = numactl -m 1 --physcpubind=8 bind9 = numactl -m 1 --physcpubind=9 bind10 = numactl -m 1 --physcpubind=10 bind11 = numactl -m 1 --physcpubind=11 % if '%{num_sockets}' eq '1' # Config settings for 1P 6-core system copies = 6 % define num_nodes 1 % elif '%{num_sockets}' eq '2' # Config settings for 2P 6-core system copies = 12 % define num_nodes 2 % else #num_sockets % error Please define "num_sockets" to either '1' or '2' in %{inc_file_name}. This error was generated % endif #num_sockets %elif '%{cores_per_socket}' eq '4' # Config settings for 4-core systems % ifdef %{autopar} preENV_OMP_NUM_THREADS = 4 % endif #autopar preENV_HUGETLB_LIMIT = 896 bind0 = numactl -m 0 --physcpubind=0 bind1 = numactl -m 0 --physcpubind=1 bind2 = numactl -m 0 --physcpubind=2 bind3 = numactl -m 0 --physcpubind=3 bind4 = numactl -m 1 --physcpubind=4 bind5 = numactl -m 1 --physcpubind=5 bind6 = numactl -m 1 --physcpubind=6 bind7 = numactl -m 1 --physcpubind=7 % if '%{num_sockets}' eq '1' # Config settings for 1P 4-core system copies = 4 % define num_nodes 1 % elif '%{num_sockets}' eq '2' # Config settings for 2P 4-core system copies = 8 % define num_nodes 2 % else #num_sockets % error Please define "num_sockets" to either '1' or '2' in %{inc_file_name}. This error was generated % endif #num_sockets %else #cores_per_socket % error Please define "cores_per_socket" to either '4' or '8' in %{inc_file_name}. This error was generated %endif #cores_per_socket ################################ # End auto-configured settings # ################################ submit = echo "$command" > run.sh ; $BIND bash run.sh ############################ Software Info ############################ fp=default=default=default: sw_base_ptrsize = 64-bit sw_peak_ptrsize = 32/64-bit sw_other = None sw_avail = Jul-2010 sw_compiler = x86 Open64 4.2.4 Compiler Suite (from AMD) #sw_other1 = binutils 2.18 int=default=default=default: sw_base_ptrsize = 32/64-bit sw_peak_ptrsize = 32/64-bit sw_avail = July-2010 sw_compiler = x86 Open64 4.2.4 Compiler Suite (from AMD) sw_other1 = SmartHeap 8.1 32-bit Library for Linux #sw_other2 = binutils 2.18 default=default=default=default: ##################################################################### # Compiler selection # # CC = opencc -march=barcelona -mso CXX = openCC -march=barcelona -mso FC = openf95 -march=barcelona -mso ##################################################################### # Portability for default 64-bit code generation ##################################################################### default=default=default: PORTABILITY = -DSPEC_CPU_LP64 ##################################################################### # INT Portability ##################################################################### 400.perlbench=default=default: CPORTABILITY = -DSPEC_CPU_LINUX_X64 403.gcc=peak=default: PORTABILITY = 429.mcf=peak=default: PORTABILITY = 462.libquantum=default=default: CPORTABILITY = -DSPEC_CPU_LINUX 471.omnetpp=default=default: # Needed to avoid -DSPEC_CPU_LP64 on -m32 C++ base codes PORTABILITY = 473.astar=default=default: # Needed to avoid -DSPEC_CPU_LP64 on -m32 C++ base codes PORTABILITY = 483.xalancbmk=default=default: CXXPORTABILITY = -DSPEC_CPU_LINUX # Needed to avoid -DSPEC_CPU_LP64 on -m32 C++ base codes PORTABILITY = ################################################################# # # # SPECint Tuning & Notes # # # ################################################################# int=base: COPTIMIZE = -Ofast -CG:local_sched_alg=1 -INLINE:aggressive=on -IPA:plimit=8000:small_pu=100 -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -m32 -INLINE:aggressive=on -CG:cmp_peep=on EXTRA_CXXLIBS = -L$(SMARTHEAP_DIR) -lsmartheap ##################################################################### # INT Peak Tuning ##################################################################### int=peak=default: COPTIMIZE = -Ofast -CG:local_sched_alg=1 -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -m32 -HP:bdt=2m:heap=2m PASS1_CFLAGS = -fb_create fbdata PASS1_CXXFLAGS = -fb_create fbdata PASS1_LDFLAGS = -fb_create fbdata PASS2_CFLAGS = -fb_opt fbdata PASS2_CXXFLAGS = -fb_opt fbdata PASS2_LDFLAGS = -fb_opt fbdata 400.perlbench=peak=default: COPTIMIZE = -Ofast -IPA:plimit=20000 -LNO:opt=0 -OPT:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -WOPT:if_conv=0 -CG:local_sched_alg=1:unroll_fb_req=on -HP:bdt=2m:heap=2m 401.bzip2=peak=default: COPTIMIZE = -O3 -OPT:alias=disjoint:goto=off -CG:local_sched_alg=1 -HP:bdt=2m:heap=2m 403.gcc=peak=default: COPTIMIZE = -Ofast -LNO:trip_count=256:prefetch_ahead=10 -CG:cmp_peep=on -m32 -HP:bdt=2m:heap=2m -GRA:unspill=on -IPA:small_pu=200 429.mcf=peak=default: COPTIMIZE = -O3 -ipa -INLINE:aggressive=on -CG:gcm=off -GRA:prioritize_by_density=on -m32 -HP:bdt=2m:heap=2m feedback = 0 445.gobmk=peak=default: COPTIMIZE = -O3 -OPT:alias=restrict:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -ipa -IPA:plimit=750:min_hotness=300:pu_reorder=1 -LNO:prefetch=1:ignore_feedback=off -CG:p2align=on:unroll_fb_req=on -HP:bdt=2m:heap=2m 456.hmmer=peak=default: COPTIMIZE = -Ofast -LNO:prefetch=0 -OPT:alias=disjoint:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -CG:local_sched_alg=1:cflow=0:push_pop_int_saved_regs=off:cmp_peep=on -HP:bdt=2m:heap=2m 458.sjeng=peak=default: COPTIMIZE = -O3 -ipa -LNO:ignore_feedback=off:full_unroll=10:fusion=0:fission=2 -IPA:pu_reorder=2 -CG:ptr_load_use=0 -OPT:unroll_times_max=8 -INLINE:aggressive=on 462.libquantum=peak=default: COPTIMIZE = -Ofast -LNO:pf2=0 -CG:gcm=off:use_prefetchnta=on:cmp_peep=on -WOPT:aggstr=0 -HP:bdt=2m:heap=2m -OPT:alias=disjoint -INLINE:aggressive=on -IPA:space=1000:plimit=20000 feedback = 0 464.h264ref=peak=default: COPTIMIZE = -O3 -IPA:plimit=20000 -OPT:alias=disjoint -LNO:prefetch=0 -CG:ptr_load_use=0:push_pop_int_saved_regs=off 471.omnetpp=peak=default: basepeak = 1 473.astar=peak=default: CXXOPTIMIZE = -Ofast -TENV:frame_pointer=off -WOPT:if_conv=0 -GRA:optimize_boundary=on -OPT:alias=disjoint -INLINE:aggressive=on -IPA:small_pu=3000:plimit=3000 -m32 -HP:bdt=2m:heap=2m 483.xalancbmk=peak=default: CXXOPTIMIZE = -Ofast -INLINE:aggressive=on -m32 -CG:cmp_peep=on -GRA:unspill=on -TENV:frame_pointer=off -fno-emit-exceptions EXTRA_CXXLIBS = -L$(SMARTHEAP_DIR) -lsmartheap feedback = 0 ##################################################################### # FP Portability ##################################################################### 436.cactusADM=default=default: FPORTABILITY = -fno-second-underscore 447.dealII=peak=default: PORTABILITY = 450.soplex=peak=default: PORTABILITY = 481.wrf=default=default: CPORTABILITY = -DSPEC_CPU_LINUX -DSPEC_CPU_CASE_FLAG FPORTABILITY = -fno-second-underscore ################################################################# # # # SPECfp Tuning # # # ################################################################# fp=base: COPTIMIZE = -Ofast -OPT:malloc_alg=1 -HP:bdt=2m CXXOPTIMIZE = -Ofast -static -INLINE:aggressive=on -OPT:malloc_alg=1 -HP:bdt=2m FOPTIMIZE = -Ofast -HP ##################################################################### # FP Peak Tuning ##################################################################### fp=peak=default: COPTIMIZE = -Ofast -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -INLINE:aggressive=on -HP:bdt=2m:heap=2m FOPTIMIZE = -Ofast -HP:bdt=2m:heap=2m PASS1_CFLAGS = -fb_create fbdata PASS2_CFLAGS = -fb_opt fbdata PASS1_CXXFLAGS = -fb_create fbdata PASS2_CXXFLAGS = -fb_opt fbdata PASS1_FFLAGS = -fb_create fbdata PASS2_FFLAGS = -fb_opt fbdata PASS1_LDFLAGS = -fb_create fbdata PASS2_LDFLAGS = -fb_opt fbdata 410.bwaves=peak=default: FOPTIMIZE = -O3 -OPT:Ofast:treeheight=on -LNO:blocking=off:prefetch_ahead=5:ignore_feedback=off -WOPT:aggstr=0 -HP:bdt=2m:heap=2m -CG:cmp_peep=on feedback = 0 416.gamess=peak=default: FOPTIMIZE = -O3 -LNO:fu=6:blocking=0:prefetch=0 -OPT:Ofast:ro=3:unroll_size=256 -HP:bdt=2m:heap=2m 433.milc=peak=default: COPTIMIZE = -Ofast -CG:movnti=1:local_sched_alg=1 -CG:locs_shallow_depth=1 -HP:bdt=2m:heap=2m -LNO:prefetch=3 feedback = 0 434.zeusmp=peak=default: FOPTIMIZE = -Ofast -LNO:blocking=off:interchange=off -OPT:treeheight=on:unroll_size=256 -CG:cmp_peep=on -GRA:prioritize_by_density=on -HP feedback = 0 435.gromacs=peak=default: FOPTIMIZE = -Ofast -OPT:rsqrt=2 -HP:bdt=2m:heap=2m COPTIMIZE = $(FOPTIMIZE) feedback = 0 436.cactusADM=peak=default: FOPTIMIZE = -Ofast -apo -LNO:prefetch_ahead=1 -HP:bdt=2m:heap=2m -LANG:heap_allocation_threshold=100 COPTIMIZE = $(FOPTIMIZE) copies = %{num_nodes} %if '%{cores_per_socket}' eq '6' # Settings for 6-core systems bind0 = numactl -m 0 --physcpubind=0,1,2,3,4,5 bind1 = numactl -m 1 --physcpubind=6,7,8,9,10,11 %elif '%{cores_per_socket}' eq '4' # Settings for 4-core systems bind0 = numactl -m 0 --physcpubind=0,1,2,3 bind1 = numactl -m 1 --physcpubind=4,5,6,7 %else #cores_per_socket % error Please define "cores_per_socket" to either '4' or '6'. %endif #cores_per_socket 437.leslie3d=peak=default: FOPTIMIZE = -Ofast -HP:bdt=2m:heap=2m feedback = 0 444.namd=peak=default: CXXOPTIMIZE = -Ofast -LNO:ignore_feedback=off -CG:local_sched_alg=2:load_exe=0:compute_to=on -OPT:unroll_size=256 -fno-exceptions -HP:bdt=2m:heap=2m 447.dealII=peak=default: CXXOPTIMIZE = -Ofast -static -INLINE:aggressive=on -LNO:opt=0 -fno-emit-exceptions -m32 -OPT:unroll_times_max=8:unroll_size=256:unroll_level=2 -HP:bdt=2m:heap=2m -GRA:unspill=on -CG:cmp_peep=on -TENV:frame_pointer=off feedback = 0 450.soplex=peak=default: CXXOPTIMIZE = -O3 -INLINE:aggressive=on -OPT:IEEE_arith=3:IEEE_NaN_Inf=off:fold_unsigned_relops=on:malloc_alg=1 -CG:load_exe=0 -fno-exceptions -m32 -HP:bdt=2m 453.povray=peak=default: CXXOPTIMIZE = -Ofast -INLINE:aggressive=on 454.calculix=peak=default: FOPTIMIZE = -Ofast -CG:load_exe=0:ptr_load_use=0:local_sched_alg=2:compute_to=on -LNO:prefetch_ahead=30 -WOPT:unroll=2 -GRA:optimize_boundary=on -HP:bdt=2m:heap=2m COPTIMIZE = $(FOPTIMIZE) feedback = 0 459.GemsFDTD=peak=default: FOPTIMIZE = -Ofast -LNO:fission=2:prefetch_ahead=1 -CG:load_exe=0:local_sched_alg=1 -HP feedback = 0 465.tonto=peak=default: FOPTIMIZE = -Ofast -OPT:alias=no_f90_pointer_alias -LNO:blocking=off -CG:load_exe=1 -IPA:plimit=525 -HP feedback = 0 470.lbm=peak=default: basepeak = 1 481.wrf=peak=default: FOPTIMIZE = -Ofast -LNO:blocking=off:prefetch_ahead=10 -LANG:copyinout=off -IPA:callee_limit=5000 -GRA:prioritize_by_density=on -m3dnow -HP COPTIMIZE = $(FOPTIMIZE) feedback = 0 482.sphinx3=peak=default: COPTIMIZE = -Ofast -OPT:malloc_alg=2 -CG:sse_cse_regs=0:locs_shallow_depth=1:cmp_peep=on:local_sched_alg=1 -INLINE:aggressive=on # The following section was added automatically, and contains settings that # did not appear in the original configuration file, but were added to the # raw file after the run. default: flagsurl001 = http://www.spec.org/cpu2006/flags/amd-platform-rate-revC.20110119.xml notes_000 =Environment variables set by runspec before the start of the run: notes_005 =HUGETLB_LIMIT = "450" notes_010 =LD_LIBRARY_PATH = "/root/work/cpu2006/amd1002-rate-libs-revC/64:/root/work/cpu2006/amd1002-rate-libs-revC/32" notes_015 =OMP_NUM_THREADS = "6" notes_020 =