| $Id: runcpu-avoidance.html 6647 2021-09-14 21:26:51Z JohnHenning $ | Latest: www.spec.org/cpu2017/Docs/ |
|---|
Contents
Copy the build dir (triple only)
Place the binary in the run dir
This document describes how to use the SPEC CPU®2017 benchmarks while avoiding most of the tools. SPEC CPU 2017 is a product of the SPEC® non-profit corporation (about SPEC). You might be interested in this document if you have a need for more direct access to the benchmarks. For example:
Some users want to work directly with benchmark source code and compile by hand, rather than through the SPEC supplied tools. Perhaps an experimental compiler is under development, and it is more convenient to just issue "cc" commands in a sandbox. Perhaps a custom build process is needed in order to add instrumentation.
Some users want to run the benchmarks directly from the command line, rather than via the SPEC supplied tools. Perhaps this is part of a debugging effort, or is needed in order to collect a performance "trace".
If the above describes you, here is a suggested path which should lead quickly to your desired state. This document shows you how to use SPEC's tools for the minimal purpose of just generating work directories, for use as a private sandbox. Note, however, that you cannot do formal, "reportable" runs without using SPEC's toolset.
License reminder: Various commands below demonstrate copying benchmarks among systems. These examples assume that all the systems belong to licensed users of SPEC CPU 2017. For the SPEC CPU license, see www.spec.org/cpu2017/Docs/licenses/SPEC-License.pdf and for information about all the licensed software in SPEC CPU 2017, see SPEC CPU 2017 Licenses.
Three different environments are referenced in this document, using these labels:
"Unified": The SPEC toolset, the compilers, and the run environment are all on the same system.
"Cross compile": The SPEC toolset and the compilers are on one system; the run time environment is a different system.
"Triple": The SPEC toolset is on one system, the compiler is on a second, and the run time environment is a third.
Review one rule: Please read the rule on research/academic usage. It is understood that the suite may be used in ways other than the formal environment that the tools help to enforce. If you plan to publish your results, you must state how your usage of the suite differs from the standard usage.
So even if you skip over the tools and the run rules today, you should plan a time to come back and learn them later.
Install: Get through a successful installation, even if it is on a different system than the one that you care about. Yes, we are about to teach you how to mostly bypass the tools, but there will still be some minimal use. So you need a working toolset and a valid installation. If you have troubles with the install procedures described in install-guide-unix.html or install-guide-windows.html, please see techsupport.html and we'll try to help you.
Pick a benchmark: Pick a benchmark that will be your starting point.
Choose one benchmark from the CPU 2017 suite that you'd like to start with. For example, you might start with 503.bwaves_r (Fortran) or 519.lbm_r (C). These are two of the shortest benchmarks for lines of code, and therefore relatively easy to understand.
Pick a config file: Pick a config file for an environment that resembles your environment. You'll find a variety of config files in the directory $SPEC/config/ on Unix systems, or %SPEC%\config\ on Windows, or at www.spec.org/cpu2017 with the submitted CPU 2017 results. Don't worry if the config file you pick doesn't exactly match your environment; you're just looking for a somewhat reasonable starting point.
Fake it: Execute a "fake" run to set up run directories, including a build directory for source code, for the benchmark.
For example, let's suppose that you want to work with 503.bwaves_r and your environment is at least partially similar to the environment described in the comments for Example-gcc-macos-x86.cfg:
$ pwd /Users/reiner/spec/cpu2017/rc6 $ source shrc $ cd config $ cp Example-gcc-macos-x86.cfg my_test.cfg $ runcpu --fake --loose --size test --tune base --config my_test 519.lbm_r [warning: size=test] runcpu v5577 - Copyright 1999-2017 Standard Performance Evaluation Corporation . . . (lots of stuff goes by) . . Success: 1x519.lbm_r The log for this run is in /reiner/cpu2017/result/CPU2017.007.log runcpu finished at 2017-05-15 11:12:09; 4 total seconds elapsed $ (Notes about examples)
This command should report a success for the build, run and validation phases of the test case, but the actual commands have not been run. It is only a report of what would be run according to the config file that you have supplied.
Find the log: Near the bottom of the output from the previous step, notice the location of the log file for this run -- in the example above, log number 007. The log file contains a record of the commands as reported by the "fake" run. You can find the commands by searching for "%%".
Find the build dir: To find the build directory that was set up in the fake run, you can search for the string build/ in the log:
$ cd $SPEC/result $ grep build/ CPU2017.005.log Wrote to makefile '/reiner/cpu2017/benchspec/CPU/519.lbm_r/build/build_base_mytest-m64.0000/Makefile.deps': Wrote to makefile '/reiner/cpu2017/benchspec/CPU/519.lbm_r/build/build_base_mytest-m64.0000/Makefile.spec': $
Or, you can just go directly to the benchmark build directories and look for the most recent one. For example:,
$ go 519.lbm build /reiner/cpu2017/benchspec/CPU/519.lbm_r/build $ ls -gtd build* drwxrwxr-x 17 staff 578 May 15 11:12 build_base_mytest-m64.0000 $
In the example above, go is shorthand for getting us around the SPEC tree. The ls -gtd command prints the names of each build subdirectory, with the most recent first. If this is your first time here, there will be only one directory listed, as in the example above. (On Windows, the "go" command is not available; use cd to get to the analogous directory, which would be spelt with reversed slashes. The top of the SPEC tree is "%SPEC%", not "$SPEC". Instead of "ls -gtd", you would say something like "dir build*/o:d".)
You can work in this build directory, make source code changes, and try other build commands without affecting the original sources.
Copy the build dir (triple only): If you are using a unified or cross-compile environment, you can skip to the next step. But if you are using a triple environment, then you will want to package up the build directory with a program such as tar -- a handy copy is in the bin directory of your SPEC installation, as spectar. You can compress it with specxz. Then, you will move the package off to whatever system has compilers.
For example, you might say something like this:
$ spectar -cf - build_base_mytest-m64.0000/ | specxz > mybuild.tar.xz $ scp mybuild.tar.xz john@somesys: [reminder: copying] mybuild.tar.xz 100% 11KB 181.7KB/s 00:00 $
Note that the above example assumes that you have versions of xz and tar available on the system that has compilers, which you will use to unpack the compressed tarfile, typically with a command similar to this:
xz -dc mybuild.tar.xz | tar -xvf -
If you don't have xz available, you might try bzip2 or gzip on both the sending and receiving systems. If you use some other compression utility, be sure that it does not corrupt the files by destroying line endings, re-wrapping long lines, or otherwise subtracting value.
Build it: Generate an executable using the build directory. If you are using a unified or cross-compile environment, then you can say commands such as these:
$ cd build_base_mytest-m64.0000/ $ specmake clean rm -rf *.o lbm.out find . \( -name \*.o -o -name '*.fppized.f*' -o -name '*.i' -o -name '*.mod' \) -print | xargs rm -rf rm -rf lbm_r rm -rf lbm_r.exe rm -rf core rm -rf compiler-version.err options.err compiler-version.out make.out options.out $ specmake /SW/compilers/gcc-6.2.0/bin/gcc -std=c99 -m64 -c -o lbm.o -DSPEC -DNDEBUG -DSPEC_AUTO_SUPPRESS_OPENMP -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize -DSPEC_LP64 lbm.c specmake: /SW/compilers/gcc-6.2.0/bin/gcc: Command not found specmake: *** [/reiner/cpu2017/benchspec/Makefile.defaults:347: lbm.o] Error 127
Note above that the config file expected the compiler to be in a different location than where it is on
this system. At this point, we have three choices: redo the runcpu after editing the config file; or edit Makefile.spec; or override the location by setting the makevar SPECLANG on the command
line.
SPECLANG, despite its name, is just an ordinary user-created make variable, indicating the
directory where compilers can be found for use with our SPEC CPU tests.
$ which gcc /davidz6.2/bin/gcc $ specmake SPECLANG=/davidz6.2/bin/ /davidz6.2/bin/gcc -std=c99 -m64 -c -o lbm.o -DSPEC -DNDEBUG -DSPEC_AUTO_SUPPRESS_OPENMP -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize -DSPEC_LP64 lbm.c /davidz6.2/bin/gcc -std=c99 -m64 -c -o main.o -DSPEC -DNDEBUG -DSPEC_AUTO_SUPPRESS_OPENMP -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize -DSPEC_LP64 main.c /davidz6.2/bin/gcc -std=c99 -m64 -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize lbm.o main.o -lm -o lbm_r $
You can also carry out a dry run of the build, which will display the build commands without attempting to run them, by adding -n to the specmake command line. You might find it useful to capture the output of specmake -n to a file, so it can easily be edited, and used as a script.
If you are trying to debug a new system, you can prototype changes to Makefile.spec or even to the benchmark sources.
If you are using a triple environment, then presumably it's because you don't have specmake working on the system where the compiler resides. But fear not: specmake is just GNU make under another name, so whatever make you have handy on the target system might work fine with the above commands. If not, then you'll need to extract the build commands from the log and try them on the system that has the compilers, using commands such as the following:
$ go result /reiner/cpu2017/result $ grep -n %% CPU2017.007.log | grep make | grep build 403:%% Fake commands from make (specmake -n --output-sync --jobs=4 build): 407:%% End of fake output from make (specmake -n --output-sync --jobs=4 build) $ head -407 CPU2017.007.log | tail -5 %% Fake commands from make (specmake -n --output-sync --jobs=4 build): /SW/compilers/gcc-6.2.0/bin/gcc -std=c99 -m64 -c -o lbm.o -DSPEC -DNDEBUG -DSPEC_AUTO_SUPPRESS_OPENMP -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize -DSPEC_LP64 lbm.c /SW/compilers/gcc-6.2.0/bin/gcc -std=c99 -m64 -c -o main.o -DSPEC -DNDEBUG -DSPEC_AUTO_SUPPRESS_OPENMP -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize -DSPEC_LP64 main.c /SW/compilers/gcc-6.2.0/bin/gcc -std=c99 -m64 -g -O3 -march=native -fno-strict-aliasing -fno-unsafe-math-optimizations -fno-tree-loop-vectorize lbm.o main.o -lm -o lbm_r %% End of fake output from make (specmake -n --output-sync --jobs=4 build) $
The first command above uses grep -n to find the line numbers of interest, and the second command prints them.
Find the run directory, and add the binary to it: Using techniques similar to those used to find the build directory, find the run directory established above, and place the binary into it. If you are using a unified or cross-compile environment, you can copy the binary directly into the run directory; if you are using a triple environment, then you'll have to retrieve the binary from the compilation system using whatever program you use to communicate between systems.
In a unified environment, the commands might look something like this:
$ go result /reiner/cpu2017/result $ grep 'Setting up' CPU2017.007.log Setting up environment for running 519.lbm_r... Setting up 519.lbm_r test base mytest-m64 (1 copy): run_base_test_mytest-m64.0000 $ go 519.lbm run /reiner/cpu2017/benchspec/CPU/519.lbm_r/run $ cd run_base_test_mytest-m64.0000/ $ cp ../../build/build_base_mytest-m64.0000/lbm_r . $
In the result directory, we search log 007 to find the correct name of the directory, go there, and copy the binary into it.
Copy the run dir: If you are using a unified environment, you can skip this step. Otherwise, you'll need to package up the run directory and transport it to the system where you want to run the benchmark. For example:
$ go 519.l run /reiner/cpu2017/benchspec/CPU/519.lbm_r/run $ spectar cf - run_base_test_mytest-m64.0000/ | specxz > myrun.tar.xz $ scp myrun.tar.xz john@mysys: [reminder: copying] myrun.tar.xz 100% 43KB 329.2KB/s 00:00 $
Note that the above example assumes that you have versions of xz and tar available on the run time system, which you will use to unpack the compressed tarfile, typically with something like this:
xz -dc myrun.tar.xz | tar -xvf -
If you don't have xz available, you might try bzip2 or gzip on both the sending and receiving systems. If you use some other compression utility, be sure that it does not corrupt the files by destroying line endings, re-wrapping long lines, or otherwise subtracting value.
Run it: If you are using a unified environment, you can use specinvoke to see the command lines that run the benchmark, and/or capture them to a shell script. You can also run them using judicious(*) cut and paste:
$ go 519.lbm run run_base_test_mytest-m64.0000 /reiner/cpu2017/benchspec/CPU/519.lbm_r/run/run_base_test_mytest-m64.0000 $ cp ../../build/build_base_mytest-m64.0000/lbm_r . $ specinvoke -n # specinvoke r<dev> # Invoked as: specinvoke -n # timer ticks over every 1000 ns # Use another -n on the command line to see chdir commands and env dump # Starting run for copy #0 ../run_base_test_mytest-m64.0000/lbm_r_base.mytest-m64 20 reference.dat 0 1 100_100_130_cf_a.of 0<&- > lbm.out 2>> lbm.err specinvoke exit: rc=0
(*) Note above that the lbm_r binary to include additional identifiers (_base.mytest-m64) - which we simply ignore in the command that is cut-and-pasted, because the binary built by hand is just lbm_r.
$ ./lbm_r 20 reference.dat 0 1 100_100_130_cf_a.of 0<&- > lbm.out 2>> lbm.err [warning: size=test] $ cat lbm.out MAIN_printInfo: grid size : 100 x 100 x 130 = 1.30 * 10^6 Cells nTimeSteps : 20 result file : reference.dat action : nothing simulation type: channel flow obstacle file : 100_100_130_cf_a.of LBM_showGridStatistics: nObstacleCells: 498440 nAccelCells: 0 nFluidCells: 801560 minRho: 1.000000 maxRho: 1.000000 Mass: 1.300000e+06 minU : 0.000000 maxU : 0.000000 LBM_showGridStatistics: nObstacleCells: 498440 nAccelCells: 0 nFluidCells: 801560 minRho: 1.000000 maxRho: 1.040865 Mass: 1.300963e+06 minU : 0.000000 maxU : 0.012647 $
If you are using a cross-compile or triple environment, you can capture the commands to a file and execute that. Be sure to follow the instructions carefully for how to do that, noting in particular the items above your environment, at the specinvoke chapter of SPEC CPU 2017 Utilities.
Alternatively, you can extract the run commands from the log.
$ go result /reiner/cpu2017/result $ grep -n %% CPU2017.007.log | grep benchmark_run 498:%% Fake commands from benchmark_run (/reiner/cpu2017/bin/spe...): 605:%% End of fake output from benchmark_run (/reiner/cpu2017/bin/spe...) $ head -605 CPU2017.007.log | tail -4 cd /reiner/cpu2017/benchspec/CPU/519.lbm_r/run/run_base_test_mytest-m64.0000 ../run_base_test_mytest-m64.0000/lbm_r_base.mytest-m64 20 reference.dat 0 1 100_100_130_cf_a.of 0<&- > lbm.out 2>> lbm.err specinvoke exit: rc=0 %% End of fake output from benchmark_run (/reiner/cpu2017/bin/spe...) $
Save your work: Important: if you are at all interested in saving your work, move the build/build* and run/run* directories to some safer location. That way, your work areas will not be accidentally deleted the next time someone comes along and uses one of runcpu cleanup actions..
Repeat: Admittedly, the large number of steps that it took to get here may seem like a lot of trouble. But that's why you started with a simple benchmark and the simplest workload (--size test in the fake step). Now that you've got the pattern down, it is hoped that it will be straightforward to repeat the process for the other available workloads: --size=train and --size=ref, and then for additional benchmarks.
But if you're finding it tedious... then maybe this is an opportunity to sell you on the notion of using runcpu after all, which automates all this tedium. If the reason you came here was because runcpu doesn't work on your brand-new environment, then perhaps you'll want to try to get it built, using the hints in tools-build.html.
Note that this document has only discussed getting the benchmarks built and running. Presumably at some point you'd like to know whether your system got the correct answer. At that point, you can use specdiff, which is explained in utility.html.
Avoiding runcpu: Using the SPEC CPU®2017 benchmarks while making minimal use of the SPEC tool set: Copyright © 2017-2021 Standard Performance Evaluation Corporation (SPEC®)