- 1 to 2 GB of main memory to run SPECrate (per copy); 16 GB for SPECspeed.
- 250 GB disk space is recommended; a minimal installation needs 10 GB.
- C, C++, and Fortran compilers (or a set of pre-compiled binaries from another CPU 2017 user). A variety of chips and operating systems are supported.
SPEC CPU running process need some tools, like spectar,specperl etc. These tools are just wrappers for open sources tools. SPEC doesn't support RISCV officially, so we need to build this tool set by hand. There is an officail help document can be referred.
https://www.spec.org/cpu2017/Docs/tools-build.html
The issues that we met are listed in the following.
In order for packagetools script to package toolset after building, we'd better untar $SPEC/install_archives/tools-src.tar to $SPEC/tools directory.
The config.guess and config.sub files of the following tools need to be replaced. These files are too old。
./specinvoke/config.guess
./specsum/build-aux/config.guess
./tar-1.28/build-aux/config.guess
./make-4.2.1/config/config.guess
./rxp-1.5.0/config.guess
./expat-2.1.0/conftools/config.guess
./xz-5.2.2/build-aux/config.guess
The two files can be replaced with files in these links:
http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess
http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub
Error message
make[2]: Entering directory '/home/riscv/benchmarks/tools/src/make-4.2.1'^M
cd tests && perl ./run_make_tests.pl -srcdir /home/riscv/benchmarks/tools/src/maa
ke-4.2.1 -make ../make ^M
Can't locate FindBin.pm in @INC (you may need to install the FindBin module) (@II
NC contains: /usr/local/lib64/perl5/5.32 /usr/local/share/perl5/5.32 /usr/lib64//
perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl55
) at ./run_make_tests.pl line 32.^M
BEGIN failed--compilation aborted at ./run_make_tests.pl line 32.^M
Solution ：sudo dnf install perl
Due to compile config files bug of perl on gcc10, we need to modify Configure and cflags.SH files. Replace 1* with 1.* in sentence case "$gccversion"
This is a bug. This test will fail after 2020. Need to modify TimeDate-2.30/t/getdate.t
require Time::Local;
-my $offset = Time::Local::timegm(0,0,0,1,0,70);
+my $offset = Time::Local::timegm(0,0,0,1,0,1970);
Solution: Copy all from cpu_mount to new directory, and add write permission to the new directory.
$ mkdir cpu2017
$ cp -r cpu_mount/* cpu2017/
$ chmod -R +w cpu2017
sudo dnf install texinfo
Before running buildtools, we can set some enviroment variables, like set MAKEFLAGS to use multithreaded compilation.
$ export MAKEFLAGS=-j4
$ ./buildtools
After building toolset, we need to use packagetools to generate directory $SPEC/tools/bin/linux-riscv64.
If buildtools finished successfully, there will be tools like specperl, spectar, specxz under $SPEC/bin directory.
Then we can make $SPEC/tools/bin/linux-riscv64 by hand, and create a file named description under it. The content of description looks like:
For riscv64 Linux systems
Built on Qemu Fedora 5.10.6-200.0.riscv64.fc33.rii
scv64 with GCC 10.3.1 20210422 (Red Hat 10.3.1-1)Finally, back to $SPEC, and run this command under $SPEC:
$ ./bin/packagetools linux-riscv64When finished, there will be this prompt.
And some files are generated under $SPEC/tools/bin/linux-riscv64:
$ ./install.sh -u linux-riscv64 -d /home/riscv/benchmarks/cpu2017_install
Before running SPEC CPU, we need to set stack size to unlimited:
$ ulimit -s unlimited
INT SPEED
runcpu -c gcc-test.cfg --noreportable -n 1 -I -T base intspeed
INT RATE
runcpu -c gcc-test.cfg --noreportable -n 1 -I -T base -C 4 intrate
FP SPEED
runcpu -c gcc-test.cfg --noreportable -n 1 -I -T base fpspeed
FP RATE
runcpu -c gcc-test.cfg --noreportable -n 1 -I -T base -C 4 fprate
Notice line 160 in config file:
%define GCCge10 # EDIT: remove the '#' from column 1 if using GCC 10 or later
IF GCC version is before 10, this line should be commented. Otherwise, will cause this error:
gfortran: error: unrecognized command line option '-fallow-argument-mismatch'; dd id you mean '-Wno-argument-mismatch'?
Becuase flag -fallow-argument-mismatch is newly introduced in GCC10.
#------------------------------------------------------------------------------
# SPEC CPU(R) 2017 config for gcc/g++/gfortran on Linux x86
#------------------------------------------------------------------------------
#
# Usage: (1) Copy this to a new name
# cd $SPEC/config
# cp Example-x.cfg myname.cfg
# (2) Change items that are marked 'EDIT' (search for it)
#
# SPEC tested this config file with:
# Compiler version(s): Various. See note "Older GCC" below.
# Operating system(s): Oracle Linux Server 6, 7, 8 /
# Red Hat Enterprise Linux Server 6, 7, 8
# SUSE Linux Enterprise Server 15
# Ubuntu 19.04
# Hardware: Xeon, EPYC
#
# If your system differs, this config file might not work.
# You might find a better config file at https://www.spec.org/cpu2017/results
#
# Note: Older GCC
#
# Please use the newest GCC that you can. The default version packaged with
# your operating system may be very old; look for alternate packages with a
# newer version.
#
# If you have no choice and must use an old version, here is what to expect:
#
# - "peak" tuning: Several benchmarks will fail at peak tuning if you use
# compilers older than GCC 7.
# In that case, please use base only.
# See: https://www.spec.org/cpu2017/Docs/overview.html#Q16
# https://www.spec.org/cpu2017/Docs/config.html#tune
# Peak tuning is expected to work for all or nearly all
# benchmarks as of GCC 7 or later.
# Exception:
# - See topic "628.pop2_s basepeak", below.
#
# - "base" tuning: This config file is expected to work for base tuning with
# GCC 4.8.5 or later
# Exception:
# - Compilers vintage about 4.9 may need to turn off the
# tree vectorizer, by adding to the base OPTIMIZE flags:
# -fno-tree-loop-vectorize
#
# Unexpected errors? Try reducing the optimization level, or try removing:
# -march=native
#
# Compiler issues: Contact your compiler vendor, not SPEC.
# For SPEC help: https://www.spec.org/cpu2017/Docs/techsupport.html
#------------------------------------------------------------------------------
#--------- Label --------------------------------------------------------------
# Arbitrary string to tag binaries (no spaces allowed)
# Two Suggestions: # (1) EDIT this label as you try new ideas.
%ifndef %{label}
% define label "gcc-test" # (2) Use a label meaningful to *you*.
%endif
#--------- Preprocessor -------------------------------------------------------
%ifndef %{bits} # EDIT to control 32 or 64 bit compilation. Or,
% define bits 64 # you can set it on the command line using:
%endif # 'runcpu --define bits=nn'
%ifndef %{build_ncpus} # EDIT to adjust number of simultaneous compiles.
% define build_ncpus 4 # Or, you can set it on the command line:
%endif # 'runcpu --define build_ncpus=nn'
# Don't change this part.
%if %{bits} == 64
% define model -march=rv64imafdc
%elif %{bits} == 32
% define model -march=rv32
%else
% error Please define number of bits - see instructions in config file
%endif
%if %{label} =~ m/ /
% error Your label "%{label}" contains spaces. Please try underscores instead.
%endif
%if %{label} !~ m/^[a-zA-Z0-9._-]+$/
% error Illegal character in label "%{label}". Please use only alphanumerics, underscore, hyphen, and period.
%endif
#--------- Global Settings ----------------------------------------------------
# For info, see:
# https://www.spec.org/cpu2017/Docs/config.html#fieldname
# Example: https://www.spec.org/cpu2017/Docs/config.html#tune
command_add_redirect = 1
flagsurl = $[top]/config/flags/gcc.xml
ignore_errors = 1
iterations = 1
label = %{label}-m%{bits}
line_width = 1020
log_line_width = 1020
makeflags = --jobs=%{build_ncpus}
mean_anyway = 1
output_format = txt,html,cfg,pdf,csv
preenv = 1
reportable = 0
tune = base,peak # EDIT if needed: set to "base" for old GCC.
# See note "Older GCC" above.
#--------- How Many CPUs? -----------------------------------------------------
# Both SPECrate and SPECspeed can test multiple chips / cores / hw threads
# - For SPECrate, you set the number of copies.
# - For SPECspeed, you set the number of threads.
# See: https://www.spec.org/cpu2017/Docs/system-requirements.html#MultipleCPUs
#
# q. How many should I set?
# a. Unknown, you will have to try it and see!
#
# To get you started, some suggestions:
#
# copies - This config file defaults to testing only 1 copy. You might
# try changing it to match the number of cores on your system,
# or perhaps the number of virtual CPUs as reported by:
# grep -c processor /proc/cpuinfo
# Be sure you have enough memory. See:
# https://www.spec.org/cpu2017/Docs/system-requirements.html#memory
#
# threads - This config file sets a starting point. You could try raising
# it. A higher thread count is much more likely to be useful for
# fpspeed than for intspeed.
#
intrate,fprate:
copies = 4 # EDIT to change number of copies (see above)
intspeed,fpspeed:
threads = 4 # EDIT to change number of OpenMP threads (see above)
#------- Compilers ------------------------------------------------------------
default:
# EDIT: The parent directory for your compiler.
# Do not include the trailing /bin/
# Do not include a trailing slash
# Examples:
# 1 On a Red Hat system, you said:
# 'yum install devtoolset-9'
# Use: % define gcc_dir "/opt/rh/devtoolset-9/root/usr"
#
# 2 You built GCC in: /disk1/mybuild/gcc-10.1.0/bin/gcc
# Use: % define gcc_dir "/disk1/mybuild/gcc-10.1.0"
#
# 3 You want: /usr/bin/gcc
# Use: % define gcc_dir "/usr"
# WARNING: See section "Older GCC" above.
#
%ifndef %{gcc_dir}
% define gcc_dir "/usr" # EDIT above)
%endif
# EDIT: If your compiler version is 10 or greater, you must enable the next
# line to avoid compile errors for several FP benchmarks
#
%define GCCge10 # EDIT: remove the '#' from column 1 if using GCC 10 or later
# EDIT if needed: the preENV line adds library directories to the runtime
# path. You can adjust it, or add lines for other environment variables.
# See: https://www.spec.org/cpu2017/Docs/config.html#preenv
# and: https://gcc.gnu.org/onlinedocs/gcc/Environment-Variables.html
#preENV_LD_LIBRARY_PATH = %{gcc_dir}/lib64/:%{gcc_dir}/lib/:/lib64
#preENV_LD_LIBRARY_PATH = %{gcc_dir}/lib64/:%{gcc_dir}/lib/:/lib64:%{ENV_LD_LIBRARY_PATH}
SPECLANG = %{gcc_dir}/bin/
CC = $(SPECLANG)gcc -std=c99 %{model}
CXX = $(SPECLANG)g++ -std=c++03 %{model}
FC = $(SPECLANG)gfortran %{model}
# How to say "Show me your version, please"
CC_VERSION_OPTION = --version
CXX_VERSION_OPTION = --version
FC_VERSION_OPTION = --version
default:
%if %{bits} == 64
sw_base_ptrsize = 64-bit
sw_peak_ptrsize = 64-bit
%else
sw_base_ptrsize = 32-bit
sw_peak_ptrsize = 32-bit
%endif
#--------- Portability --------------------------------------------------------
default: # data model applies to all benchmarks
%if %{bits} == 32
# Strongly recommended because at run-time, operations using modern file
# systems may fail spectacularly and frequently (or, worse, quietly and
# randomly) if a program does not accommodate 64-bit metadata.
EXTRA_PORTABILITY = -D_FILE_OFFSET_BITS=64
%else
EXTRA_PORTABILITY = -DSPEC_LP64
%endif
# Benchmark-specific portability (ordered by last 2 digits of bmark number)
500.perlbench_r,600.perlbench_s: #lang='C'
%if %{bits} == 32
% define suffix IA32
%else
% define suffix X64
%endif
PORTABILITY = -DSPEC_LINUX_%{suffix}
521.wrf_r,621.wrf_s: #lang='F,C'
CPORTABILITY = -DSPEC_CASE_FLAG
FPORTABILITY = -fconvert=big-endian
523.xalancbmk_r,623.xalancbmk_s: #lang='CXX'
PORTABILITY = -DSPEC_LINUX
526.blender_r: #lang='CXX,C'
PORTABILITY = -funsigned-char -DSPEC_LINUX
527.cam4_r,627.cam4_s: #lang='F,C'
PORTABILITY = -DSPEC_CASE_FLAG
628.pop2_s: #lang='F,C'
CPORTABILITY = -DSPEC_CASE_FLAG
FPORTABILITY = -fconvert=big-endian
#----------------------------------------------------------------------
# GCC workarounds that do not count as PORTABILITY
#----------------------------------------------------------------------
# The workarounds in this section would not qualify under the SPEC CPU
# PORTABILITY rule.
# - In peak, they can be set as needed for individual benchmarks.
# - In base, individual settings are not allowed; set for whole suite.
# See:
# https://www.spec.org/cpu2017/Docs/runrules.html#portability
# https://www.spec.org/cpu2017/Docs/runrules.html#BaseFlags
#
# Integer workarounds - peak
#
500.perlbench_r,600.perlbench_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/500.perlbench_r.html
EXTRA_CFLAGS = -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-finite-math-only
502.gcc_r,602.gcc_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/502.gcc_r.html
EXTRA_CFLAGS = -fno-strict-aliasing -fgnu89-inline
505.mcf_r,605.mcf_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/505.mcf_r.html
EXTRA_CFLAGS = -fno-strict-aliasing
525.x264_r,625.x264_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/525.x264_r.html
EXTRA_CFLAGS = -fcommon
#
# Integer workarounds - base - combine the above - https://www.spec.org/cpu2017/Docs/runrules.html#BaseFlags
#
intrate,intspeed=base:
EXTRA_CFLAGS = -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-finite-math-only -fgnu89-inline -fcommon
#
# Floating Point workarounds - peak
#
511.povray_r=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/511.povray_r.html
EXTRA_CFLAGS = -fno-strict-aliasing
521.wrf_r,621.wrf_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/521.wrf_r.html
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
527.cam4_r,627.cam4_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/527.cam4_r.html
EXTRA_CFLAGS = -fno-strict-aliasing
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
# See also topic "628.pop2_s basepeak" below
628.pop2_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/628.pop2_s.html
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
#
# FP workarounds - base - combine the above - https://www.spec.org/cpu2017/Docs/runrules.html#BaseFlags
#
fprate,fpspeed=base:
EXTRA_CFLAGS = -fno-strict-aliasing
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
#-------- Tuning Flags common to Base and Peak --------------------------------
#
# Speed (OpenMP and Autopar allowed)
#
%if %{bits} == 32
intspeed,fpspeed:
#
# Many of the speed benchmarks (6nn.benchmark_s) do not fit in 32 bits
# If you wish to run SPECint2017_speed or SPECfp2017_speed, please use
#
# runcpu --define bits=64
#
fail_build = 1
%else
intspeed,fpspeed:
EXTRA_OPTIMIZE = -fopenmp -DSPEC_OPENMP
fpspeed:
#
# 627.cam4 needs a big stack; the preENV will apply it to all
# benchmarks in the set, as required by the rules.
#
preENV_OMP_STACKSIZE = 120M
%endif
#-------- Base Tuning Flags ----------------------------------------------
# EDIT if needed -- If you run into errors, you may need to adjust the
# optimization - for example you may need to remove
# the -march=native. See topic "Older GCC" above.
#
default=base: # flags for all base
OPTIMIZE = -g -O3 -march=rv64imafdc -flto
COPTIMIZE = -ffast-math
#-------- Peak Tuning Flags ----------------------------------------------
default=peak:
OPTIMIZE = -g -Ofast -march=rv64imafdc -flto
PASS1_FLAGS = -fprofile-generate
PASS2_FLAGS = -fprofile-use
# 628.pop2_s basepeak: Depending on the interplay of several optimizations,
# 628.pop2_s might not validate with peak tuning. Use the base
# version instead. See:
# https:// www.spec.org/cpu2017/Docs/benchmarks/628.pop2_s.html
628.pop2_s=peak:
basepeak = yes
#------------------------------------------------------------------------------
# Tester and System Descriptions - EDIT all sections below this point
#------------------------------------------------------------------------------
# For info about any field, see
# https://www.spec.org/cpu2017/Docs/config.html#fieldname
# Example: https://www.spec.org/cpu2017/Docs/config.html#hw_memory
#-------------------------------------------------------------------------------
#--------- EDIT to match your version -----------------------------------------
default:
sw_compiler001 = C/C++/Fortran: Version 10.3.0 of GCC, the
sw_compiler002 = GNU Compiler Collection
#--------- EDIT info about you ------------------------------------------------
# To understand the difference between hw_vendor/sponsor/tester, see:
# https://www.spec.org/cpu2017/Docs/config.html#test_sponsor
intrate,intspeed,fprate,fpspeed: # Important: keep this line
hw_vendor = SiFive
tester = PLCT
test_sponsor = PLCT
license_num = 0
# prepared_by = # Ima Pseudonym # Whatever you like: is never output
#--------- EDIT system availability dates -------------------------------------
intrate,intspeed,fprate,fpspeed: # Important: keep this line
# Example # Brief info about field
hw_avail = Sep-2021 # Date of LAST hardware component to ship
sw_avail = Apr-20i21 # Date of LAST software component to ship
fw_bios = 1.0.0 released Sep-2021 # Firmware information
#--------- EDIT system information --------------------------------------------
intrate,intspeed,fprate,fpspeed: # Important: keep this line
# Example # Brief info about field
hw_cpu_name = SiFive U74 # chip name
hw_cpu_nominal_mhz = 1433MHz # Nominal chip frequency, in MHz
hw_cpu_max_mhz = 1536MHz # Max chip frequency, in MHz
# hw_disk = # 9 x 9 TB SATA III 9999 RPM # Size, type, other perf-relevant info
hw_model = hifive unmatched # system model name
hw_nchips = 1 # number chips enabled
hw_ncores = 4 # number cores enabled
# hw_ncpuorder = 1 # Ordering options
hw_nthreadspercore = 1 # number threads enabled per core
hw_other = None # Other perf-relevant hw, or "None"
# hw_memory001 = # 999 GB (99 x 9 GB 2Rx4 PC4-2133P-R, # The 'PCn-etc' is from the JEDEC
# hw_memory002 = # running at 1600 MHz) # label on the DIMM.
hw_pcache = 32 KB I + 32 KB D on chip per core # Primary cache size, type, location
hw_scache = None # Second cache or "None"
hw_tcache = None # Third cache or "None"
hw_ocache = None # Other cache or "None"
# sw_file = # ext99 # File system
# sw_os001 = # Linux Sailboat # Operating system
# sw_os002 = # Distribution 7.2 SP1 # and version
sw_other = None # Other perf-relevant sw, or "None"
# sw_state = # Run level 99 # Software state.
# power_management = # briefly summarize power settings
# Note: Some commented-out fields above are automatically set to preliminary
# values by sysinfo
# https://www.spec.org/cpu2017/Docs/config.html#sysinfo
# Uncomment lines for which you already know a better answer than sysinfo
#------------------------------------------------------------------------------
# SPEC CPU(R) 2017 config for clang on riscv64
#------------------------------------------------------------------------------
#
# Usage: (1) Copy this to a new name
# cd $SPEC/config
# cp Example-x.cfg myname.cfg
# (2) Change items that are marked 'EDIT' (search for it)
#
# SPEC tested this config file with:
# Compiler version(s): Various. See note "Older GCC" below.
# Operating system(s): Oracle Linux Server 6, 7, 8 /
# Red Hat Enterprise Linux Server 6, 7, 8
# SUSE Linux Enterprise Server 15
# Ubuntu 19.04
# Hardware: Xeon, EPYC
#
# If your system differs, this config file might not work.
# You might find a better config file at https://www.spec.org/cpu2017/results
#
# Note: Older GCC
#
# Please use the newest GCC that you can. The default version packaged with
# your operating system may be very old; look for alternate packages with a
# newer version.
#
# If you have no choice and must use an old version, here is what to expect:
#
# - "peak" tuning: Several benchmarks will fail at peak tuning if you use
# compilers older than GCC 7.
# In that case, please use base only.
# See: https://www.spec.org/cpu2017/Docs/overview.html#Q16
# https://www.spec.org/cpu2017/Docs/config.html#tune
# Peak tuning is expected to work for all or nearly all
# benchmarks as of GCC 7 or later.
# Exception:
# - See topic "628.pop2_s basepeak", below.
#
# - "base" tuning: This config file is expected to work for base tuning with
# GCC 4.8.5 or later
# Exception:
# - Compilers vintage about 4.9 may need to turn off the
# tree vectorizer, by adding to the base OPTIMIZE flags:
# -fno-tree-loop-vectorize
#
# Unexpected errors? Try reducing the optimization level, or try removing:
# -march=native
#
# Compiler issues: Contact your compiler vendor, not SPEC.
# For SPEC help: https://www.spec.org/cpu2017/Docs/techsupport.html
#------------------------------------------------------------------------------
#--------- Label --------------------------------------------------------------
# Arbitrary string to tag binaries (no spaces allowed)
# Two Suggestions: # (1) EDIT this label as you try new ideas.
%ifndef %{label}
% define label "clang12-riscv64" # (2) Use a label meaningful to *you*.
%endif
#--------- Preprocessor -------------------------------------------------------
%ifndef %{bits} # EDIT to control 32 or 64 bit compilation. Or,
% define bits 64 # you can set it on the command line using:
%endif # 'runcpu --define bits=nn'
%ifndef %{build_ncpus} # EDIT to adjust number of simultaneous compiles.
% define build_ncpus 4 # Or, you can set it on the command line:
%endif # 'runcpu --define build_ncpus=nn'
# Don't change this part.
%if %{bits} == 64
% define model -march=rv64gc
%elif %{bits} == 32
% define model -march=rv32gc
%else
% error Please define number of bits - see instructions in config file
%endif
%if %{label} =~ m/ /
% error Your label "%{label}" contains spaces. Please try underscores instead.
%endif
%if %{label} !~ m/^[a-zA-Z0-9._-]+$/
% error Illegal character in label "%{label}". Please use only alphanumerics, underscore, hyphen, and period.
%endif
#--------- Global Settings ----------------------------------------------------
# For info, see:
# https://www.spec.org/cpu2017/Docs/config.html#fieldname
# Example: https://www.spec.org/cpu2017/Docs/config.html#tune
command_add_redirect = 1
flagsurl = $[top]/config/flags/clang.xml
ignore_errors = 1
iterations = 1
label = %{label}-%{bits}
line_width = 1020
log_line_width = 1020
makeflags = --jobs=%{build_ncpus}
mean_anyway = 1
output_format = txt,html,cfg,pdf,csv
preenv = 1
reportable = 0
tune = base,peak # EDIT if needed: set to "base" for old GCC.
# See note "Older GCC" above.
#--------- How Many CPUs? -----------------------------------------------------
# Both SPECrate and SPECspeed can test multiple chips / cores / hw threads
# - For SPECrate, you set the number of copies.
# - For SPECspeed, you set the number of threads.
# See: https://www.spec.org/cpu2017/Docs/system-requirements.html#MultipleCPUs
#
# q. How many should I set?
# a. Unknown, you will have to try it and see!
#
# To get you started, some suggestions:
#
# copies - This config file defaults to testing only 1 copy. You might
# try changing it to match the number of cores on your system,
# or perhaps the number of virtual CPUs as reported by:
# grep -c processor /proc/cpuinfo
# Be sure you have enough memory. See:
# https://www.spec.org/cpu2017/Docs/system-requirements.html#memory
#
# threads - This config file sets a starting point. You could try raising
# it. A higher thread count is much more likely to be useful for
# fpspeed than for intspeed.
#
intrate,fprate:
copies = 4 # EDIT to change number of copies (see above)
intspeed,fpspeed:
threads = 16 # EDIT to change number of OpenMP threads (see above)
#------- Compilers ------------------------------------------------------------
default:
# EDIT: The parent directory for your compiler.
# Do not include the trailing /bin/
# Do not include a trailing slash
# Examples:
# 1 On a Red Hat system, you said:
# 'yum install devtoolset-9'
# Use: % define gcc_dir "/opt/rh/devtoolset-9/root/usr"
#
# 2 You built GCC in: /disk1/mybuild/gcc-10.1.0/bin/gcc
# Use: % define gcc_dir "/disk1/mybuild/gcc-10.1.0"
#
# 3 You want: /usr/bin/gcc
# Use: % define gcc_dir "/usr"
# WARNING: See section "Older GCC" above.
#
%ifndef %{gcc_dir}
% define gcc_dir "/usr/local" # EDIT above)
%endif
# EDIT: If your compiler version is 10 or greater, you must enable the next
# line to avoid compile errors for several FP benchmarks
#
%define GCCge10 # EDIT: remove the '#' from column 1 if using GCC 10 or later
# EDIT if needed: the preENV line adds library directories to the runtime
# path. You can adjust it, or add lines for other environment variables.
# See: https://www.spec.org/cpu2017/Docs/config.html#preenv
# and: https://gcc.gnu.org/onlinedocs/gcc/Environment-Variables.html
preENV_LD_LIBRARY_PATH = %{gcc_dir}/lib64/:%{gcc_dir}/lib/:/lib64
#preENV_LD_LIBRARY_PATH = %{gcc_dir}/lib64/:%{gcc_dir}/lib/:/lib64:%{ENV_LD_LIBRARY_PATH}
SPECLANG = %{gcc_dir}/bin/
CC = $(SPECLANG)clang $(model)
CXX = $(SPECLANG)clang++ -std=c++98 $(model)
FC = $(SPECLANG)flang $(model)
# How to say "Show me your version, please"
CC_VERSION_OPTION = --version
CXX_VERSION_OPTION = --version
FC_VERSION_OPTION = --version
default:
%if %{bits} == 64
sw_base_ptrsize = 64-bit
sw_peak_ptrsize = 64-bit
%else
sw_base_ptrsize = 32-bit
sw_peak_ptrsize = 32-bit
%endif
#--- No Fortran -------
# Fortran benchmarks are not expected to work with this config file.
# If you wish, you can run the other benchmarks using:
# runcpu no_fortran - all CPU 2017 benchmarks that do not use Fortran
# runcpu intrate_no_fortran - integer rate benchmarks that do not use Fortran
# runcpu intspeed_no_fortran - integer speed benchmarks that do not use Fortran
# runcpu fprate_no_fortran - floating point rate benchmarks that do not use Fortran
# runcpu fpspeed_no_fortran - floating point speed benchmarks that do not use Fortran
#
# If you *do* have a Fortran compiler, then you will need to set the correct options for
# 'FC' and 'FC_VERSION_OPTION' just above -- see
# http://www.spec.org/cpu2017/Docs/config.html#makeCompiler
# You must also remove the 2 lines right after this comment.
#any_fortran:
# fail_build = 1
#--------- Portability --------------------------------------------------------
default: # data model applies to all benchmarks
%if %{bits} == 32
# Strongly recommended because at run-time, operations using modern file
# systems may fail spectacularly and frequently (or, worse, quietly and
# randomly) if a program does not accommodate 64-bit metadata.
EXTRA_PORTABILITY = -D_FILE_OFFSET_BITS=64
%else
EXTRA_PORTABILITY = -DSPEC_LP64
%endif
# Benchmark-specific portability (ordered by last 2 digits of bmark number)
500.perlbench_r,600.perlbench_s: #lang='C'
%if %{bits} == 32
% define suffix IA32
%else
% define suffix X64
%endif
PORTABILITY = -DSPEC_LINUX_%{suffix}
521.wrf_r,621.wrf_s: #lang='F,C'
CPORTABILITY = -DSPEC_CASE_FLAG -DSPEC_LP64
# FPORTABILITY = -fconvert=big-endian
523.xalancbmk_r,623.xalancbmk_s: #lang='CXX'
PORTABILITY = -DSPEC_LINUX
526.blender_r: #lang='CXX,C'
PORTABILITY = -funsigned-char -DSPEC_LINUX
527.cam4_r,627.cam4_s: #lang='F,C'
PORTABILITY = -DSPEC_CASE_FLAG -DSPEC_LP64
619.lbm_s,627.cam4_s,628.pop2_s,638.imagick_s,644.nab_s,657.xz_s:
PORTABILITY = -fopenmp -I/usr/lib/gcc/riscv64-linux-gnu/10/include
628.pop2_s: #lang='F,C'
CPORTABILITY = -DSPEC_CASE_FLAG -DSPEC_LP64
EXTRA_OPTIMIZE = -fopenmp
# FPORTABILITY = -fconvert=big-endian
#----------------------------------------------------------------------
# GCC workarounds that do not count as PORTABILITY
#----------------------------------------------------------------------
# The workarounds in this section would not qualify under the SPEC CPU
# PORTABILITY rule.
# - In peak, they can be set as needed for individual benchmarks.
# - In base, individual settings are not allowed; set for whole suite.
# See:
# https://www.spec.org/cpu2017/Docs/runrules.html#portability
# https://www.spec.org/cpu2017/Docs/runrules.html#BaseFlags
#
# Integer workarounds - peak
#
500.perlbench_r,600.perlbench_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/500.perlbench_r.html
EXTRA_CFLAGS = -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-finite-math-only
502.gcc_r,602.gcc_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/502.gcc_r.html
EXTRA_CFLAGS = -fno-strict-aliasing -fgnu89-inline
505.mcf_r,605.mcf_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/505.mcf_r.html
EXTRA_CFLAGS = -fno-strict-aliasing
525.x264_r,625.x264_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/525.x264_r.html
EXTRA_CFLAGS = -fcommon -I/usr/lib/gcc/riscv64-linux-gnu/10/include
638.imagick_s=peak:
EXTRA_OPTIMIZE = -fopenmp
644.nab_s=peak:
EXTRA_OPTIMIZE = -fopenmp
654.roms_s=peak:
EXTRA_OPTIMIZE = -fopenmp
#
# Integer workarounds - base - combine the above - https://www.spec.org/cpu2017/Docs/runrules.html#BaseFlags
#
intrate,intspeed=base:
EXTRA_COPTIMIZE = -fgnu89-inline -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-finite-math-only -fcommon
525.x264_r,625.x264_s=base: # https://www.spec.org/cpu2017/Docs/benchmarks/525.x264_r.html
EXTRA_CFLAGS = -fcommon -I/usr/lib/gcc/riscv64-linux-gnu/10/include
557.xz_r,657.xz_s=base:
EXTRA_CFLAGS = -fopenmp -I/usr/lib/gcc/riscv64-linux-gnu/10/include
fpspeed=base:
# EXTRA_LDFLAGS = -Wl,-stack_size,0x10000000,-fgnu89-inline
EXTRA_CFLAGS = -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-finite-math-only -fgnu89-inline -fcommon
#
# Floating Point workarounds - peak
#
511.povray_r=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/511.povray_r.html
EXTRA_CFLAGS = -fno-strict-aliasing
521.wrf_r,621.wrf_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/521.wrf_r.html
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
527.cam4_r,627.cam4_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/527.cam4_r.html
EXTRA_CFLAGS = -fno-strict-aliasing
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
# See also topic "628.pop2_s basepeak" below
628.pop2_s=peak: # https://www.spec.org/cpu2017/Docs/benchmarks/628.pop2_s.html
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
#
# FP workarounds - base - combine the above - https://www.spec.org/cpu2017/Docs/runrules.html#BaseFlags
#
fprate,fpspeed=base:
EXTRA_CFLAGS = -fno-strict-aliasing
% ifdef %{GCCge10} # workaround for GCC v10 (and presumably later)
EXTRA_FFLAGS = -fallow-argument-mismatch
% endif
#-------- Tuning Flags common to Base and Peak --------------------------------
#
# Speed (O#penMP and Autopar allowed)
#
%if %{bits} == 32
intspeed,fpspeed:
#
# Many of the speed benchmarks (6nn.benchmark_s) do not fit in 32 bits
# If you wish to run SPECint2017_speed or SPECfp2017_speed, please use
#
# runcpu --define bits=64
#
fail_build = 1
%else
intspeed,fpspeed:
EXTRA_OPTIMIZE = -DSPEC_OPENMP -DUSE_OPENMP
fpspeed:
#
# 627.cam4 needs a big stack; the preENV will apply it to all
# benchmarks in the set, as required by the rules.
#
preENV_OMP_STACKSIZE = 120M
%endif
#-------- Base Tuning Flags ----------------------------------------------
# EDIT if needed -- If you run into errors, you may need to adjust the
# optimization - for example you may need to remove
# the -march=native. See topic "Older GCC" above.
#
default=base: # flags for all base
COPTIMIZE = -g -O3 -Wl,-mllvm -Wl,-reduce-array-computations=3
CXXOPTIMIZE = -g -O3 -Wl,-mllvm -Wl,-reduce-array-computations=3
FOPTIMIZE = -g -O3 -Wl,-mllvm -Wl,-reduce-array-computations=3 -fno-finite-math-only
#-------- Peak Tuning Flags ----------------------------------------------
default=peak:
OPTIMIZE = -g -Ofast
#PASS1_FLAGS = -fprofile-instr-generate=pgo_data
#PASS2_FLAGS = -fprofile-instr-use
#fdo_post1 = xcrun llvm-profdata merge -output=default.profdata pgo_data
# 628.pop2_s basepeak: iDepending on the interplay of several optimizations,
# 628.pop2_s might not validate with peak tuning. Use the base
# version instead. See:
# https:// www.spec.org/cpu2017/Docs/benchmarks/628.pop2_s.html
628.pop2_s=peak:
basepeak = yes
#------------------------------------------------------------------------------
# Tester and System Descriptions - EDIT all sections below this point
#------------------------------------------------------------------------------
# For info about any field, see
# https://www.spec.org/cpu2017/Docs/config.html#fieldname
# Example: https://www.spec.org/cpu2017/Docs/config.html#hw_memory
#-------------------------------------------------------------------------------
#--------- EDIT to match your version -----------------------------------------
default:
sw_compiler001 = Ubuntu clang version 12.0.0-3ubuntu1~21.04.2
#--------- EDIT info about you ------------------------------------------------
# To understand the difference between hw_vendor/sponsor/tester, see:
# https://www.spec.org/cpu2017/Docs/config.html#test_sponsor
intrate,intspeed,fprate,fpspeed: # Important: keep this line
hw_vendor = SiFive
tester = PLCT
test_sponsor = PLCT
license_num = 0
# prepared_by = # Ima Pseudonym # Whatever you like: is never output
#--------- EDIT system availability dates -------------------------------------
intrate,intspeed,fprate,fpspeed: # Important: keep this line
# Example # Brief info about field
hw_avail = Sep-2021 # Date of LAST hardware component to ship
sw_avail = Apr-20i21 # Date of LAST software component to ship
fw_bios = 1.0.0 released Sep-2021 # Firmware information
#--------- EDIT system information --------------------------------------------
intrate,intspeed,fprate,fpspeed: # Important: keep this line
# Example # Brief info about field
hw_cpu_name = SiFive U74 # chip name
hw_cpu_nominal_mhz = 1433MHz # Nominal chip frequency, in MHz
hw_cpu_max_mhz = 1536MHz # Max chip frequency, in MHz
# hw_disk = # 9 x 9 TB SATA III 9999 RPM # Size, type, other perf-relevant info
hw_model = hifive unmatched # system model name
hw_nchips = 1 # number chips enabled
hw_ncores = 4 # number cores enabled
# hw_ncpuorder = 1 # Ordering options
hw_nthreadspercore = 4 # number threads enabled per core
hw_other = None # Other perf-relevant hw, or "None"
# hw_memory001 = # 999 GB (99 x 9 GB 2Rx4 PC4-2133P-R, # The 'PCn-etc' is from the JEDEC
# hw_memory002 = # running at 1600 MHz) # label on the DIMM.
hw_pcache = 32 KB I + 32 KB D on chip per core # Primary cache size, type, location
hw_scache = None # Second cache or "None"
hw_tcache = None # Third cache or "None"
hw_ocache = None # Other cache or "None"
# sw_file = # ext99 # File system
# sw_os001 = # Linux Sailboat # Operating system
# sw_os002 = # Distribution 7.2 SP1 # and version
sw_other = None # Other perf-relevant sw, or "None"
# sw_state = # Run level 99 # Software state.
# power_management = # briefly summarize power settings
# Note: Some commented-out fields above are automatically set to preliminary
# values by sysinfo
# https://www.spec.org/cpu2017/Docs/config.html#sysinfo
# Uncomment lines for which you already know a better answer than sysinfo