# Invocation command line: # /cpu2006/bin/runspec --verbose=6 -c linux64-pgi720b-barcelona-speed -F pgi72_flags.xml -T all fp # output_root was not used for this run ############################################################################ # PGI 7.1 CPU2006 Serial Configuration File for 64-bit Linux (Quad-Core) # # Compiler name/version: PGI 7.2 # Operating system version: 64-bit Linux # Hardware: AMD Opteron (Quad-core) # FP Base Pointer Size: 64-bit only # FP Peak Pointer Size: 64-bit only # INT Base Pointer Size: 32/64-bit # INT Peak Pointer Size: 32/64-bit # Auto Parallization: Not Used # # 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 # # Set your LD_LIBRARY_PATH to the location of the PGI and Smartheap runtime libraries # # Please adjust the SHL_DIR variable to the directory containing # the Smartheap library. ##################################################################### # Macro section #################################################################### # Modify this section to use the appropriate architecture flags %define pgi_tp64 -tp barcelona-64 %define pgi_tp32 -tp barcelona # Adjust the build jobs to the number of concurrent build processses %define build_jobs 8 ##################################################################### # Header Section ##################################################################### ext = proliant_amd ignore_errors = no tune = base,peak output_format = asc,pdf,raw,flags,cfg,html size = test,train,ref check_md5 = yes reportable = yes env_vars = no teeout = yes makeflags = -j %{build_jobs} #################################################################### # Include file containing the SUT hardware inforamtion # as well as the submit command, tester information and notes #################################################################### #include: pgi_sut.cfg # ----- Begin inclusion of 'pgi_sut.cfg' ############################################################################ ##################################################################### # Submit Section ##################################################################### # RATE #This config file is set to run these binaries on a 2P, 8 core system. If this #does not match the description of your system, change this config file, #specifically "bind0","bind1","bind2", etc. to match your system. Use #"man numactl" and "numactl --hardware" to better understand how to use #this command. #bind0 = numactl -l --physcpubind=0 #bind1 = numactl -l --physcpubind=1 #bind2 = numactl -l --physcpubind=2 #bind3 = numactl -l --physcpubind=3 #bind4 = numactl -l --physcpubind=4 #bind5 = numactl -l --physcpubind=5 #bind6 = numactl -l --physcpubind=6 #bind7 = numactl -l --physcpubind=7 # SPEED # Below is an example numactl command for a speed run using # 4 threads on a single Quad-core chip. # Please adjust as needed for your system # bind0 = numactl -l --physcpubind=0,1,2,3,4,5,6,7 use_submit_for_speed = 1 submit = echo "$command" > run.sh ; $BIND bash run.sh #################################################################### # Tester information #################################################################### default=default=default=default: license_num = 3 prepared_by = ISS Performance Engineering test_sponsor = Hewlett-Packard Company tester = Hewlett-Packard Company test_date = Mar-2008 #################################################################### # Hardware information #################################################################### hw_cpu_name = AMD Opteron 2356 hw_cpu_mhz = 2300 hw_disk = 1x146 GB 10 K SAS hw_fpu = Integrated hw_memory = 32 GB (8x4 GB, PC2-5300P CL5) hw_vendor = Hewlett-Packard Company hw_model000 = ProLiant BL465c G5 hw_model001 = (2.3 GHz AMD Opteron 2356) hw_avail = Mar-2008 hw_nchips = 2 hw_ncores = 8 hw_ncoresperchip= 4 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 = 2 MB I+D on chip per chip hw_ocache = None hw_other = None sw_file = ext2 sw_os000 = SUSE Linux Enterprise Server 10 (x86_64) SP1, sw_os001 = Kernel 2.6.16.46-0.12-smp sw_state = Run level 3 (multi-user) ##################################################################### # Notes ##################################################################### int=default=default=default: notes100 = Environment stack size set to 'unlimited' notes105 = Max locked memory set to 2097152 notes110 = PGI_HUGE_PAGES set to 896. notes120 = Total number of huge pages available is 7168. fp=default=default=default: # ---- End inclusion of '/cpu2006/config/pgi_sut.cfg' ############################ Software Info ############################ fp=default=default=default: sw_peak_ptrsize = 64-bit sw_base_ptrsize = 64-bit sw_auto_parallel = Yes sw_avail = May-2008 sw_other000 = binutils-2.18.50 sw_compiler000 = PGI Server Complete Version 7.2 int=default=default=default: sw_peak_ptrsize = 32/64-bit sw_base_ptrsize = 32/64-bit sw_auto_parallel = No sw_other1 = binutils 2.18.50 sw_other2 = SmartHeap 8.1 32-bit Library for Linux sw_avail = May-2008 sw_compiler1 = PGI Server Complete Version 7.2 default=default=default=default: ##################################################################### # Compiler selection # # CC = pgcc -w CXX = pgcpp -w FC = pgf95 -w SHL_DIR = /proj/qa/smartheap/SmartHeap_8.1/lib ##################################################################### # Optimization ##################################################################### default=default=default=default: EXTRA_LDFLAGS = -Bstatic_pgi fp=base=default=default: FOPTIMIZE = -fast -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mconcur -Msmartalloc=huge:896 %{pgi_tp64} COPTIMIZE = -fast -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mconcur -Msmartalloc=huge:896 %{pgi_tp64} CXXOPTIMIZE = -fast -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Msmartalloc=huge:896 --zc_eh %{pgi_tp64} int=base=default=default: COPTIMIZE = -fast -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mloop32 -Mconcur=innermost -Msmartalloc=huge:896 %{pgi_tp64} CXXOPTIMIZE = -fastsse -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mloop32 -Msmartalloc=huge:896 --zc_eh %{pgi_tp32} ##################################################################### ## FP Peak Flags ##################################################################### fp=peak=default=default: EXTRA_LIBS= basepeak=0 410.bwaves=peak=default=default: FOPTIMIZE = -fastsse -Mloop32 -Mfprelaxed -Msmartalloc -Mprefetch=distance:12,nta -Mconcur -Mipa=jobs:%{build_jobs},fast,inline %{pgi_tp64} feedback=0 basepeak=0 416.gamess=peak=default=default: FOPTIMIZE = -fast -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mvect=noaltcode -Msmartalloc=huge:896 %{pgi_tp64} submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 433.milc=peak=default=default: COPTIMIZE = -fastsse -Msmartalloc=huge:896 -Mconcur -Msafeptr -Mfprelaxed -Mipa=jobs:%{build_jobs},inline,arg,const,ptr,shape %{pgi_tp64} feedback=0 basepeak=0 434.zeusmp=peak=default=default: FOPTIMIZE = -fast -Mloop32 -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mconcur -Msmartalloc %{pgi_tp64} COPTIMIZE = -fast -Mloop32 -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mconcur -Msmartalloc %{pgi_tp64} basepeak=0 feedback=0 435.gromacs=peak=default=default: COPTIMIZE = -fast -Mconcur -Mfpapprox=rsqrt -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Msmartalloc=huge:896 %{pgi_tp64} FOPTIMIZE = -fast -Mconcur -Mfpapprox=rsqrt -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Msmartalloc=huge:896 %{pgi_tp64} feedback=0 basepeak=0 436.cactusADM=peak=default=default: FOPTIMIZE = -fastsse -Mfprelaxed -Mconcur -Msmartalloc -Mdse -Mipa=jobs:%{build_jobs},fast,inline %{pgi_tp64} COPTIMIZE = -fastsse -Mfprelaxed -Mconcur -Msmartalloc -Mdse -Mipa=jobs:%{build_jobs},fast,inline %{pgi_tp64} feedback=0 basepeak=0 437.leslie3d=peak=default=default: FOPTIMIZE = -fast -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Mconcur=noaltcode -Msmartalloc=huge:896 %{pgi_tp64} feedback=0 basepeak=0 444.namd=peak=default=default: CXXOPTIMIZE = -fast -Mfprelaxed -Msmartalloc=huge:896 --zc_eh -Mnodepchk -Munroll=n:4,m:8 %{pgi_tp64} PASS1_CXXFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CXXFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mconcur=noaltcode -Mpfo PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mconcur=noaltcode -Mpfo feedback=1 basepeak=0 447.dealII=peak=default=default: CXXOPTIMIZE = -fast -Mfprelaxed -Msmartalloc=huge:896 --zc_eh -Mnovect -alias=ansi -Mipa=jobs:%{build_jobs},fast,inline %{pgi_tp64} submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 450.soplex=peak=default=default: submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=1 453.povray=peak=default=default: CXXOPTIMIZE = -fast -Mfprelaxed -Msmartalloc %{pgi_tp64} PASS1_CXXFLAGS = -Mpfi=indirect PASS1_LDFLAGS = -Mpfi=indirect PASS2_CXXFLAGS = -Mipa=jobs:%{build_jobs},fast,inlinenopfo:3,staticfunc -Mpfo=indirect PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inlinenopfo:3,staticfunc -Mpfo=indirect submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=1 basepeak=0 454.calculix=peak=default=default: COPTIMIZE = -fastsse -Mloop32 -Mfprelaxed -Msmartalloc=huge:896 -Mipa=jobs:%{build_jobs},fast,inline %{pgi_tp64} FOPTIMIZE = -fastsse -Mloop32 -Mfprelaxed -Msmartalloc=huge:896 -Mipa=jobs:%{build_jobs},fast,inline %{pgi_tp64} submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 459.GemsFDTD=peak=default=default: FOPTIMIZE = -fast -Mfprelaxed -Msmartalloc=huge:896 -Mprefetch=nta %{pgi_tp64} PASS1_FFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_FFLAGS = -Mipa=jobs:%{build_jobs},fast,inlinenopfo -Mconcur -Mpfo PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inlinenopfo -Mconcur -Mpfo feedback=1 basepeak=0 465.tonto=peak=default=default: FOPTIMIZE = -fast -O4 -Mfprelaxed -Msmartalloc=huge:896 -Mprefetch=distance:8 -Mipa=jobs:%{build_jobs},fast,inline -Mvect=noaltcode %{pgi_tp64} submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 470.lbm=peak=default=default: COPTIMIZE = -fastsse -Mfprelaxed -Msmartalloc=huge:896 -Mipa=fast,inline,noarg -Mprefetch=distance:12,nta %{pgi_tp64} submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 481.wrf=peak=default=default: FOPTIMIZE = -fast -Mfprelaxed -Msmartalloc=huge:896 -Mconcur=noaltcode -Mvect=noaltcode -Mprefetch=distance:8 %{pgi_tp64} COPTIMIZE = -fast -Mfprelaxed -Msmartalloc=huge:896 -Mconcur=noaltcode -Mvect=noaltcode -Mprefetch=distance:8 %{pgi_tp64} feedback=0 basepeak=0 482.sphinx3=peak=default=default: COPTIMIZE = -fastsse -Mfprelaxed -Msmartalloc -Mprefetch=distance:12,nta %{pgi_tp64} PASS1_CFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mpfo PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mpfo submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=1 basepeak=0 ##################################################################### ## INT Peak Flags ##################################################################### int=peak=default=default: EXTRA_LIBS = basepeak=1 400.perlbench=peak=default=default: COPTIMIZE = -fast -O4 -Mfprelaxed -Msmartalloc=huge:896 -Mnounroll -Mnovect %{pgi_tp64} PASS1_CFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CFLAGS = -Mpfo -Mipa=jobs:%{build_jobs},inline PASS2_LDFLAGS = -Mpfo -Mipa=jobs:%{build_jobs},inline submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=1 basepeak=0 401.bzip2=peak=default=default: COPTIMIZE = -fast -O4 -Msmartalloc=huge:896 -Mnounroll %{pgi_tp64} PASS1_CFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CFLAGS = -Mconcur=innermost -Mpfo PASS2_LDFLAGS = -Mconcur=innermost -Mpfo feedback=1 basepeak=0 403.gcc=peak=default=default: COPTIMIZE = -fastsse -Mfprelaxed -Msmartalloc=huge:896 %{pgi_tp32} PASS1_CFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CFLAGS = -Mpfo -Mconcur -Mipa=jobs:%{build_jobs},fast,inline PASS2_LDFLAGS = -Mpfo -Mconcur -Mipa=jobs:%{build_jobs},fast,inline feedback=1 basepeak=0 429.mcf=peak=default=default: COPTIMIZE = -fastsse -Mconcur -Mipa=jobs:%{build_jobs},fast,inline:1 -Msmartalloc=huge:896 -Mloop32 %{pgi_tp32} feedback=0 basepeak=0 445.gobmk=peak=default=default: COPTIMIZE = -fast -O4 -Msmartalloc=huge:896 -Mfprelaxed -Mnovect %{pgi_tp64} PASS1_CFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CFLAGS = -Mpfo -Mconcur -Mipa=jobs:%{build_jobs},fast PASS2_LDFLAGS = -Mpfo -Mconcur -Mipa=jobs:%{build_jobs},fast feedback=1 basepeak=0 456.hmmer=peak=default=default: COPTIMIZE = -fastsse -Munroll=n:8 -Msmartalloc=huge:896 -Mfprelaxed -Mvect=partial -Msafeptr -Mipa=jobs:%{build_jobs},const,ptr,arg,inline %{pgi_tp64} submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 458.sjeng=peak=default=default: COPTIMIZE = -fastsse -Msmartalloc=huge:896 -Mfprelaxed %{pgi_tp64} PASS1_CFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CFLAGS = -Mipa=jobs:%{build_jobs},fast,inline:1,noarg -Mconcur -Mpfo PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inline:1,noarg -Mconcur -Mpfo feedback=1 basepeak=0 462.libquantum=peak=default=default: COPTIMIZE = -fastsse -Mfprelaxed -Msmartalloc -Mvect=nosse -Munroll=m:8 -Mconcur=innermost,noaltcode -Mipa=jobs:%{build_jobs},fast,noarg %{pgi_tp64} feedback=0 basepeak=0 464.h264ref=peak=default=default: COPTIMIZE = -fastsse -Msmartalloc=huge:896 -Mfprelaxed -Mloop32 %{pgi_tp64} PASS1_CFLAGS = -Mpfi=indirect PASS1_LDFLAGS = -Mpfi=indirect PASS2_CFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mpfo=indirect PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mpfo=indirect submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=1 basepeak=0 471.omnetpp=peak=default=default: submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=1 473.astar=peak=default=default: CXXOPTIMIZE = -fastsse -O4 -Msmartalloc=huge:896 -Mfprelaxed -Mloop32 --zc_eh %{pgi_tp32} PASS1_CXXFLAGS = -Mpfi PASS1_LDFLAGS = -Mpfi PASS2_CXXFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mpfo PASS2_LDFLAGS = -Mipa=jobs:%{build_jobs},fast,inline -Mpfo submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=1 basepeak=0 483.xalancbmk=peak=default=default: CXXOPTIMIZE = -fastsse -O4 -Mipa=jobs:%{build_jobs},fast,inline -Mfprelaxed -Msmartalloc -Mloop32 --zc_eh %{pgi_tp32} EXTRA_CXXLIBS = -L$(SHL_DIR) -lsmartheap submit = echo "$command" > run.sh ; numactl -l --physcpubind=1 bash run.sh feedback=0 basepeak=0 ##################################################################### # Portability ##################################################################### fp=default=default=default: PORTABILITY = -DSPEC_CPU_LP64 int=default=default=default: CPORTABILITY = -DSPEC_CPU_LP64 400.perlbench=default=default=default: CPORTABILITY = -DSPEC_CPU_LP64 -DSPEC_CPU_LINUX_X64 403.gcc=peak=default=default: CPORTABILITY = 429.mcf=peak=default=default: CPORTABILITY = 435.gromacs=default=default=default: LDPORTABILITY = -Mnomain 436.cactusADM=default=default=default: LDPORTABILITY = -Mnomain 454.calculix=default=default=default: LDPORTABILITY = -Mnomain 462.libquantum=default=default=default: CPORTABILITY = -DSPEC_CPU_LP64 -DSPEC_CPU_LINUX 481.wrf=default=default=default: CPORTABILITY = -DSPEC_CPU_CASE_FLAG -DSPEC_CPU_LINUX 483.xalancbmk=default=default=default: CXXPORTABILITY = -DSPEC_CPU_LINUX # 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. 454.calculix # The following setting was inserted automatically as a result of # post-run basepeak application. basepeak = 1 453.povray # The following setting was inserted automatically as a result of # post-run basepeak application. basepeak = 1 434.zeusmp # The following setting was inserted automatically as a result of # post-run basepeak application. basepeak = 1 416.gamess # The following setting was inserted automatically as a result of # post-run basepeak application. basepeak = 1 default: flagsurl000= http://www.spec.org/cpu2006/flags/hp-PGI72-PS32-flags.xml notes_plat_000 =BIOS configuration: notes_plat_005 = Power Regulator set to Static High Performance Mode notes_os_000 = Environment stack size set to 'unlimited' notes_os_005 = Max locked memory set to 2097152 notes_os_010 = PGI_HUGE_PAGES set to 896. notes_os_015 = Total number of huge pages available is 7168. notes_os_020 = NCPUS set to number of cores notes_os_025 = numactl used to bind processes to CPUs