# Invocation command line: # /autobench/sources/speccpu2006/bin/runspec --config cpu2006.at6.128c.cfg --action run --rate 1 --copies 512 --reportable --nobuild --tune base --iterations 3 --define numcores=128 --define rhel=1 --define mybasearchp7=1 --define mybaseopt=3 --define libhuge=1 --define smartheap=1 --define lto=1 --extension AT-at600-rhel63-Saturn int # output_root was not used for this run ############################################################################ ####################################################################### # # SPEC CPU2006 Benchmark Config file IBM eServer pSeries Linux # ####################################################################### # THIS CONFIG FILE IS FOR BUILDING on P7 # # --define (-S) options that are passed to the runspec command # that dynamically affect this config file # # numcores=n The number of cores in the system # to determine how many copies to run in rate # # mybaseopt=x Allows override where x=3,4,5 (-O3, -O4, -O5) for base runs # # libhuge # smartheap # lto # libmass # # rhel # sles # yesL3 # noL3 # debug # # mymode64 # The following setting was inserted automatically as a result of # post-run basepeak application. basepeak = 1 %define P7 -mcpu=power7 -mtune=power7 %ifdef %{libhuge} %define LHBDT -B/usr/share/libhugetlbfs/ -Wl,--hugetlbfs-align %define LHLIB -lhugetlbfs %else %define LHBDT %define LHLIB %endif %if %{smartheap} %if %{mymode64} %define SMARTHEAPLIB -lsmartheap64 %else #%define SMARTHEAPLIB -lsmartheap %define SMARTHEAPLIB -ltcmalloc %endif %else %define SMARTHEAPLIB %endif %ifdef %{libmass} %define VEC_OPT -fvect-cost-model -mveclibabi=mass %define LIB_MATH -lm %if %{mymode64} %define LIB_MASS -L/opt/ibmcmp/xlmass/6.1/lib64 -lmass_64 -lmassvp7_64 -lmass_simdp7_64 %else %define LIB_MASS -L/opt/ibmcmp/xlmass/6.1/lib -lmass -lmassvp7 -lmass_simdp7 %endif %else %define VEC_OPT -fvect-cost-model %define LIB_MATH %define LIB_MASS %endif %ifdef %{lto} %define LTO -flto -fwhole-program -fuse-linker-plugin %else %define LTO %endif %define FDPRPRO /opt/ibm/fdprpro/bin/fdprpro %define FDPROPTS -O4 -omullX %define FDPRINT %ifdef %{numcores} %define make_j %{numcores} %else %define make_j 4 %endif #====================================================================== # Global Options apply to all runs #====================================================================== # # For just a report generated without a raw file or any runs required # Do a "runspec --reportonly" (also --action=report or --fakereport) # # Just set up the build or run directories, use --buildsetup or --setup # on the runspec command # Sets nobuild as the default - you will need -D to force a compile # nobuild = 1 # -D forces a build (recompile) # -N prohibits a build (recompile) #action = validate # -a #tune = base,peak # -T #iterations = 3 # -n output_format = raw,text,html,csv # -o #basepeak = 0 #reportable = 1 # -l turns off #ignore_errors = 0 # -I mean_anyway = 1 allow_extension_override = 1 #ext = LoP-ppc64 # -e #size = ref post_setup = sync parallel_test = 1 # # Include the run-time derived system information so that each # system run has the correct information #include: SUT.inc # ----- Begin inclusion of 'SUT.inc' ############################################################################ # company_name = IBM Corporation hw_vendor = IBM Corporation license_num = 11 sw_state = Multi-User hw_cpu_mhz = 3724 hw_cpu_name = POWER7 hw_memory = 1 TB (64 x 16 GB) DDR3 1066 MHz # # # hw_disk = 1x146 GB SAS 15K RPM # # sw_os000 = Red Hat Enterprise Linux Server release 6.3 sw_os001 = (ppc64) kernel 2.6.32-279.el6.ppc64 sw_file = ext4 # # sw_avail = Aug-2012 # essl.rte-5.1.0-0.ppc64 sw_compiler000 = gcc (GCC) 4.4.6 20120305 (Red Hat 4.4.6-4) sw_compiler001 =Advance Toolchain 6.0-0 gcc/g++ compiler essl.rte-5.1.0-0.ppc64 sw_other000= -IBM Advance Toolchain 6.0-0 essl.rte-5.1.0-0.ppc64 notes_000 = The following environment varibles were set before the runspec command: notes_005 = export HUGETLB_VERBOSE=0 # hw_model = IBM Power 780 (3.7 GHz, 128 core, RHEL, GCC) hw_avail =Oct-2012 # hw_ncpuorder =32,64,96,128 cores hw_nchips = hw_ncores = 128 hw_nthreadspercore = 4 hw_ncoresperchip =8 hw_pcache =32 KB I + 32 KB D on chip per core hw_fpu = Integrated hw_other =None # # hw_scache =256 KB I+D on chip per core hw_tcache =10 MB I+D on chip per core hw_ocache =None # # # notes_010 = export HUGETLB_MORECORE=yes notes_015 = export XLFRTEOPTS=intrinthds=1 # # # # ppc64_cpu output # ---- End inclusion of '/autobench/sources/speccpu2006/config/SUT.inc' # This tells runspec to set enviornment variables of the form # ENV_xxx if found in here. Forced to 0 for reportable runs env_vars = 0 # Control if the embedded md5 string of compiler options is checked for # recompile if the options have changed check_md5 = 0 # Controls if the log file info is echo'd to the screen teeout = yes # If 1, make copy of the config file before modifing with MD5 # signitures of the compile options. backup_config = 0 # verbose -> -v #verbose = 1 to 99, 5 is default, 91 is good for debugging %ifdef %{make_j} makeflags = -j%{make_j} %endif # We alway use submit to ensure there is a bindprocessor done. # use_submit_for_speed If set, use submit commands for speed as well as rate use_submit_for_speed = 1 # SPEED %if %{rhel} submit = numactl --membind=1 --physcpubind=\$SPECCOPYNUM $command #submit = echo "$command" > run.sh ; numactl --membind=\$SPECCOPYNUM --physcpubind=\$SPECCOPYNUM sh run.sh %endif %if %{sles} submit = echo "$command" > run.sh ; numactl --physcpubind=\$SPECCOPYNUM sh run.sh %endif bind = \ 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96 100 104 108 112 116 120 124 128 132 136 140 144 148 152 156 160 164 168 172 176 180 184 188 192 196 200 204 208 212 216 220 224 228 232 236 240 244 248 252 256 260 264 268 272 276 280 284 288 292 296 300 304 308 312 316 320 324 328 332 336 340 344 348 352 356 360 364 368 372 376 380 384 388 392 396 400 404 408 412 416 420 424 428 432 436 440 444 448 452 456 460 464 468 472 476 480 484 488 492 496 500 504 508 \ 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 101 105 109 113 117 121 125 129 133 137 141 145 149 153 157 161 165 169 173 177 181 185 189 193 197 201 205 209 213 217 221 225 229 233 237 241 245 249 253 257 261 265 269 273 277 281 285 289 293 297 301 305 309 313 317 321 325 329 333 337 341 345 349 353 357 361 365 369 373 377 381 385 389 393 397 401 405 409 413 417 421 425 429 433 437 441 445 449 453 457 461 465 469 473 477 481 485 489 493 497 501 505 509 \ 2 6 10 14 18 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 86 90 94 98 102 106 110 114 118 122 126 130 134 138 142 146 150 154 158 162 166 170 174 178 182 186 190 194 198 202 206 210 214 218 222 226 230 234 238 242 246 250 254 258 262 266 270 274 278 282 286 290 294 298 302 306 310 314 318 322 326 330 334 338 342 346 350 354 358 362 366 370 374 378 382 386 390 394 398 402 406 410 414 418 422 426 430 434 438 442 446 450 454 458 462 466 470 474 478 482 486 490 494 498 502 506 510 \ 3 7 11 15 19 23 27 31 35 39 43 47 51 55 59 63 67 71 75 79 83 87 91 95 99 103 107 111 115 119 123 127 131 135 139 143 147 151 155 159 163 167 171 175 179 183 187 191 195 199 203 207 211 215 219 223 227 231 235 239 243 247 251 255 259 263 267 271 275 279 283 287 291 295 299 303 307 311 315 319 323 327 331 335 339 343 347 351 355 359 363 367 371 375 379 383 387 391 395 399 403 407 411 415 419 423 427 431 435 439 443 447 451 455 459 463 467 471 475 479 483 487 491 495 499 503 507 511 # 16 NUMA nodes, 8-core per NUMA node # 128 cores, 4 threads per core (SMT4) # # Topology: # - Processors 0-31: NUMA node 0, # - Processors 32-63: NUMA node 1, # - Processors 64-95: NUMA node 2, # - Processors 96-127: NUMA node 3, # - Processors 128-159: NUMA node 4, # - Processors 160-191: NUMA node 5, # - Processors 192-223: NUMA node 6, # - Processors 224-255: NUMA node 7, # - Processors 256-287: NUMA node 8, # - Processors 288-319: NUMA node 9, # - Processors 320-351: NUMA node 10, # - Processors 352-383: NUMA node 11, # - Processors 384-415: NUMA node 12, # - Processors 416-447: NUMA node 13, # - Processors 448-479: NUMA node 14, # - Processors 480-511: NUMA node 15. %if %{numcores} == 128 copies = 512 submit = let "MYCPU=\$BIND"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM $command %endif #--prevent log file lines that are too long for vi log_line_width = 110 # Point to where the usr/compiler/OS Flags file is #flagsurl = IBM-Linux-gcc.xml #=============================================================================== # How the compilers are called # The settings used here are for IBM C/C++ and xl Fortran compilers #=============================================================================== default=default=default=default: %define MY_PATH /opt/at6.0/bin sw_compiler000 = C/C++: Version 4.7.2 of IBM CC = %{MY_PATH}/gcc CXX = %{MY_PATH}/g++ FC = %{MY_PATH}/gfortran default=default=default=default: #====================================================================== # Base Level Optimizations # No limit to number of flags, must be same for all of each Language. #====================================================================== default=base=default=default: %if %{mymode64} sw_base_ptrsize = 64-bit %else sw_base_ptrsize = 32-bit %endif %ifndef %{debug} %define DEBUG %else %define DEBUG -g %endif # Check for BASEARCH overrides # Default BASEARCH is P7 %define BASEARCH %{P7} %define MODE -m32 %if %{mymode64} %define MODE -m64 %endif # Note if set -mcpu=power7, then don't need to specify -mvsx ,-mpopcntd, -mcmpb, or -mhard-dfp. # But -maltivec is not set automatically. # Note when using -mcpu=power7, # - do not disable Altivec without also disabling VSX # this is illegal -mcpu=power7 -mno-altivec. # MAURICFO: update this according tothtat vect bug thing. # AT 5.0-1 Flag Issue: # # The fortran component 416.gamess fails to build # w/ -ftree-loop-linear (fortran compiler keeps looping). # # So, that flag was moved from GCCOPT to GCCOPT_LOOP. # Only C & CXX OPTIMIZE use it (INT); # F OPTIMIZE does _not_ (FP). %define GCCOPT -fpeel-loops -funroll-loops -ffast-math -ftree-vectorize -mvsx -maltivec -mpopcntd -mrecip=rsqrt %define GCCOPT_MATH -fno-associative-math %define GCCOPT_LOOP -ftree-loop-linear # Default mode is -O3 with p7 %define BASEOPT -O3 %{BASEARCH} %{MODE} %{GCCOPT} %{DEBUG} #Check for BASEOPT overrides %if %{mybaseopt} == 3 %define BASEOPT -O3 %{BASEARCH} %{MODE} %{GCCOPT} %{DEBUG} %endif %if %{mybaseopt} == 2 %define BASEOPT -O2 %{BASEARCH} %{MODE} %{GCCOPT} %{DEBUG} %endif %if %{mybaseopt} == 1 %define BASEOPT -O1 %{BASEARCH} %{MODE} %{GCCOPT} %{DEBUG} %endif int=base=default=default: COPTIMIZE = %{BASEOPT} -fno-strict-aliasing %if %{mybasearchp7} COPTIMIZE = %{BASEOPT} %{LTO} %endif EXTRA_CLIBS = %{LHLIB} CXXOPTIMIZE = %{BASEOPT} %{LTO} EXTRA_CXXLIBS = %{SMARTHEAPLIB} fp=base=default=default: COPTIMIZE = %{BASEOPT} %{VEC_OPT} CXXOPTIMIZE = %{BASEOPT} %{VEC_OPT} %{GCCOPT_MATH} FOPTIMIZE = %{BASEOPT} %{VEC_OPT} EXTRA_LIBS = %{LHLIB} %{LIB_MASS} %{LIB_MATH} default=base=default=default: %if %{mymode64} PORTABILITY = -DSPEC_CPU_LP64 %endif #====================================================================== # INT Portability Flags # no optimizations allowed here #====================================================================== 400.perlbench=default=default=default: %if %{mymode64} CPORTABILITY = -DSPEC_CPU_LINUX_PPC -DSPEC_CPU_LP64 %else CPORTABILITY = -DSPEC_CPU_LINUX_PPC %endif 403.gcc=default=default=default: %if %{mymode64} CPORTABILITY = -DSPEC_CPU_LP64 %endif 462.libquantum=default=default=default: CPORTABILITY = -DSPEC_CPU_LINUX 464.h264ref=default=default=default: CPORTABILITY = -fsigned-char 483.xalancbmk=default=default=default: CXXPORTABILITY = -DSPEC_CPU_LINUX #====================================================================== # FP Portability Flags # no optimizations allowed here #====================================================================== 436.cactusADM=peak=default=default: %if %{mymode64} CPORTABILITY = -DSPEC_CPU_LP64 %endif 447.dealII=default=default=default: #CXXOPTIMIZE = -O3 %{BASEARCH} %{MODE} -fpeel-loops -funroll-loops -ffast-math -ftree-vectorize -mvsx -maltivec -mpopcntd -mrecip=rsqrt -fno-associative-math %{DEBUG} %{VEC_OPT} #EXTRA_LIBS = %{LHLIB} %{LIB_MASS} %{LIB_MATH} %if %{mymode64} PORTABILITY = -DSPEC_CPU_LP64 %endif CXXPORTABILITY = -DSPEC_CPU_LINUX 481.wrf=default=default=default: CPORTABILITY = -DSPEC_CPU_CASE_FLAG -DSPEC_CPU_LINUX_PPC 482.sphinx3=default=default=default: CPORTABILITY = -fsigned-char #====================================================================== # # Default peak flags. # #====================================================================== default=peak=default=default: sw_peak_ptrsize = 32-bit %define O2 -O2 %{BASEARCH} %{MODE} %{GCCOPT} %{DEBUG} %define O3 -O3 %{BASEARCH} %{MODE} %{GCCOPT} %{DEBUG} #====================================================================== # INT peak Flags #====================================================================== int=peak=default=default: 400.perlbench=peak=default=default: # C only Benchmark OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = %{SMARTHEAPLIB} PASS1_CFLAGS = -fprofile-generate PASS1_LDFLAGS = -fprofile-generate PASS2_CFLAGS = -fprofile-use PASS2_LDFLAGS = -fprofile-use %ifdef %{64} CPORTABILITY = -DSPEC_CPU_LINUX_PPC -DSPEC_CPU_LP64 COPTIMIZE = -m64 %else CPORTABILITY = -DSPEC_CPU_LINUX_PPC %endif 401.bzip2=peak=default=default: # C only Benchmark %define FDPROPTS_BZIP2 -O4 -vrox OPTIMIZE = %{O3} %{LTO} -fno-branch-count-reg -fprefetch-loop-arrays EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q fdo_pre1 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run1 = $command fdo_post1 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_BZIP2} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 403.gcc=peak=default=default: # C only Benchmark %define FDPROPTS_GCC -O4 -nodp -rtb -m power7 -A 1 OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q #PASS1_CFLAGS = -fprofile-generate #PASS1_LDFLAGS = -fprofile-generate -fwhole-program -fuse-linker-plugin #PASS2_CFLAGS = -fprofile-use #PASS2_LDFLAGS = -fprofile-use -fwhole-program -fuse-linker-plugin fdo_pre2 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run2 = $command fdo_post2 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_GCC} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} CPORTABILITY = -DSPEC_CPU_LP64 COPTIMIZE = -m64 %endif 429.mcf=peak=default=default: # C only Benchmark %if %{numcores} == 8 copies = 8 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %if %{numcores} == 12 copies = 12 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/24))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %if %{numcores} == 16 copies = 16 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %if %{numcores} == 32 copies = 32 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; hugectl --heap numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %define FDPROPTS_MCF -O3 -m power7 -A 1 -omullX OPTIMIZE = %{O3} %{LTO} --param max-goto-duplication-insns=16 EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q EXTRA_LDFLAGS = -fwhole-program fdo_pre1 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run1 = $command fdo_post1 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_MCF} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 445.gobmk=peak=default=default: # C only Benchmark %define FDPROPTS_GOBMK -m power7 -A 1 -q -O3 -omullX OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q #PASS1_CFLAGS = -fprofile-generate #PASS1_LDFLAGS = -fprofile-generate -fwhole-program -fuse-linker-plugin #PASS2_CFLAGS = -fprofile-use #PASS2_LDFLAGS = -fprofile-use -fwhole-program -fuse-linker-plugin fdo_pre2 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run2 = $command fdo_post2 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_GOBMK} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 456.hmmer=peak=default=default: # C only Benchmark %define FDPROPTS_HMMER -O4 -omullX #OPTIMIZE = %{O3} %{LTO} -finline-limit=11880 OPTIMIZE = %{O3} -finline-limit=11880 EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q #PASS1_CFLAGS = -fprofile-generate #PASS1_LDFLAGS = -fprofile-generate -fwhole-program -fuse-linker-plugin #PASS2_CFLAGS = -fprofile-use #PASS2_LDFLAGS = -fprofile-use -fwhole-program -fuse-linker-plugin fdo_pre2 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run2 = $command fdo_post2 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_HMMER} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 458.sjeng=peak=default=default: # C only Benchmark %define FDPROPTS_SJENG -O4 -omullX OPTIMIZE = %{O3} %{LTO} -misel EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q PASS1_CFLAGS = -fprofile-generate PASS1_LDFLAGS = -fprofile-generate PASS2_CFLAGS = -fprofile-use PASS2_LDFLAGS = -fprofile-use fdo_pre2 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run2 = $command fdo_post2 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_SJENG} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 462.libquantum=peak=default=default: # C only Benchmark # We want just 1 copy per core, not SMT, when we have L3cache %if %{numcores} == 8 copies = 8 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %if %{numcores} == 12 copies = 12 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/24))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %if %{numcores} == 16 copies = 16 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %if %{numcores} == 32 copies = 32 submit = echo "$command" > run.sh; let "MYCPU=4*\$SPECCOPYNUM"; MYMEM=`printf '%d' \$((\$MYCPU/32))` ; hugectl --heap numactl --physcpubind=\$MYCPU --membind=\$MYMEM sh run.sh %endif %define FDPROPTS_LIBQUANTUM -m power7 -A 1 -bf -dp -lro -nop -RC -RD -tb -tlo -vro -isf 88 -lu 8 -hrf 0.10 -sdp 4 -lun 27 -omullX %if %{noL3} %define FDPROPTS_LIBQUANTUM -ece -pca -RC -RD -tb -tlo -lu 5 -rt 0.75 -hrf 0.75 -sdpla 4 -rcctf 0.55 -rccrf 0.10 -rcaf 2 -omullX %endif OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q PASS1_CFLAGS = -fprofile-generate PASS1_LDFLAGS = -fprofile-generate PASS2_CFLAGS = -fprofile-use PASS2_LDFLAGS = -fprofile-use fdo_pre2 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run2 = $command fdo_post2 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_LIBQUANTUM} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 464.h264ref=peak=default=default: # C only Benchmark #OPTIMIZE = %{O3} %{LTO} -fno-branch-count-reg -misel -fprefetch-loop-arrays OPTIMIZE = %{O3} -fno-branch-count-reg -misel -fprefetch-loop-arrays EXTRA_LIBS = -lhugetlbfs LDCFLAGS = -Wl,-q PASS1_CFLAGS = -fprofile-generate PASS1_LDFLAGS = -fprofile-generate PASS2_CFLAGS = -fprofile-use PASS2_LDFLAGS = -fprofile-use #fdo_pre1 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr ${baseexe}; cp ${baseexe}.instr ${baseexe} #fdo_run1 = $command #fdo_post1 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof -O4 -vrox -nodp ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} COPTIMIZE = -m64 %endif 471.omnetpp=peak=default=default: # C++ Benchmark #%define FDPROPTS_OMNETPP -O3 -lu -1 -nodp -sdp 9 OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = %{SMARTHEAPLIB} LDCXXFLAGS = -Wl,-q #EXTRA_LDFLAGS = -fwhole-program -fuse-linker-plugin #fdo_pre1 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} #fdo_run1 = $command #fdo_post1 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_OMNETPP} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} CXXOPTIMIZE = -m64 %endif 473.astar=peak=default=default: # C++ Benchmark %define FDPROPTS_ASTAR -O4 -m power7 -A 1 OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = %{SMARTHEAPLIB} LDCXXFLAGS = -Wl,-q #PASS1_CFLAGS = -fprofile-generate #PASS1_LDFLAGS = -fprofile-generate #PASS2_CFLAGS = -fprofile-use #PASS2_LDFLAGS = -fprofile-use fdo_pre1 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} fdo_run1 = $command fdo_post1 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_ASTAR} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} CXXOPTIMIZE = -m64 %endif 483.xalancbmk=peak=default=default: # C++ Benchmark #%define FDPROPTS_XALAN -O3 -m power7 OPTIMIZE = %{O3} %{LTO} EXTRA_LIBS = %{SMARTHEAPLIB} LDCXXFLAGS = -Wl,-q #fdo_pre1 = cp ${baseexe} ${baseexe}.orig ; %{FDPRPRO} -a instr %{FDPRINT} ${baseexe}; cp ${baseexe}.instr ${baseexe} #fdo_run1 = $command #fdo_post1 = cp ${baseexe}.orig ${baseexe} ; %{FDPRPRO} -a opt -f ${baseexe}.nprof %{FDPROPTS_XALAN} ${baseexe} ; cp ${baseexe}.fdpr ${baseexe} %ifdef %{64} CXXOPTIMIZE = -m64 %endif 999.specrand=peak=default=default: # C only Benchmark - Must run, but not reported unless fails OPTIMIZE = %{O3} #============================================================ # FP peak Flags #============================================================ fp=peak=default=default: 998.specrand=peak=default=default: # C only Benchmark - Must run, but not reported unless fails OPTIMIZE = %{O3} # The following settings were obtained by running 'specperl $[top]/Docs/sysinfo' default: hw_cpu_name = POWER7+ hw_disk = 12x146.8 GB SAS SFF 15K RPM hw_nchips = 16 prepared_by = -- sw_file = ext4 sw_state = Run level 3 (multi-user) # 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: test_sponsor = IBM Corporation tester = IBM Corporation flagsurl000 = http://www.spec.org/cpu2006/flags/IBM-Power.xml flagsurl001 = http://www.spec.org/cpu2006/flags/IBM-Linux-AT.xml hw_cpu_char000 = Intelligent Energy Optimization enabled, hw_cpu_char001 = up to 4.144 GHz notes_submit_000 = The config file option 'submit' was used notes_submit_005 = to assign benchmark copy to specific kernel thread using notes_submit_010 = the "numactl" command (see flags file for details). notes_plat_000 = Service processor memory mirroring property disabled. notes_comp_iler_invocation_000 =For more information about IBM Advance Toolchain, including support, see notes_comp_iler_invocation_005 =ftp://linuxpatch.ncsa.uiuc.edu/toolchain/at/redhat/RHEL6/at6.0/release_notes.at6.0-6.0-0.html notes_os_000 = ulimit -s (stack) set to 1048576. notes_os_005 = ulimit -n (open files) set to 500000. notes_os_010 = notes_os_015 = Large pages reserved as follows by root user: notes_os_020 = echo 30000 > /proc/sys/vm/nr_hugepages notes_os_025 = echo 6000 > /proc/sys/vm/nr_overcommit_hugepages