ICODE-ADC/lammps
4
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@12592 f3b2605a-c512-4ea7-a41b-209d697bcdaa

Browse Source
This commit is contained in:
sjplimp
2014-10-07 15:08:33 +00:00
parent fcad656d92
commit 16dcead2d2
16 changed files with 1085 additions and 583 deletions
Download Patch File Download Diff File Expand all files Collapse all files

68
doc/Section_accelerate.html
View File

@ -165,8 +165,8 @@ coprocessors.
</P> </P>
<P>All of these commands are in packages provided with LAMMPS. An <P>All of these commands are in packages provided with LAMMPS. An
overview of packages is give in <A HREF = "Section_packages.html">Section overview of packages is give in <A HREF = "Section_packages.html">Section
packages</A>. Currently, there are 6 accelerator packages</A>. These are the accelerator packages
packages in LAMMPS, either as standard or user packages: currently in LAMMPS, either as standard or user packages:
</P> </P>
<DIV ALIGN=center><TABLE BORDER=1 > <DIV ALIGN=center><TABLE BORDER=1 >
<TR><TD ><A HREF = "accelerate_cuda.html">USER-CUDA</A> </TD><TD > for NVIDIA GPUs</TD></TR> <TR><TD ><A HREF = "accelerate_cuda.html">USER-CUDA</A> </TD><TD > for NVIDIA GPUs</TD></TR>
@ -201,8 +201,8 @@ for that style.
</P> </P>
<P>To use an accelerator package in LAMMPS, and one or more of the styles <P>To use an accelerator package in LAMMPS, and one or more of the styles
it provides, follow these general steps. Details vary from package to it provides, follow these general steps. Details vary from package to
package and are explained in the individual accelerator sub-section package and are explained in the individual accelerator doc pages,
doc pages, listed above: listed above:
</P> </P>
<DIV ALIGN=center><TABLE BORDER=1 > <DIV ALIGN=center><TABLE BORDER=1 >
<TR><TD >build the accelerator library </TD><TD > only for USER-CUDA and GPU packages </TD></TR> <TR><TD >build the accelerator library </TD><TD > only for USER-CUDA and GPU packages </TD></TR>
@ -215,26 +215,46 @@ doc pages, listed above:
<TR><TD >use accelerated styles in your input script </TD><TD > via "-sf" <A HREF = "Section_start.html#start_7">command-line switch</A> or <A HREF = "suffix.html">suffix</A> command <TR><TD >use accelerated styles in your input script </TD><TD > via "-sf" <A HREF = "Section_start.html#start_7">command-line switch</A> or <A HREF = "suffix.html">suffix</A> command
</TD></TR></TABLE></DIV> </TD></TR></TABLE></DIV>
<P>The first 4 steps typically only need to be done once, to create an <P>The first 4 steps can be done as a single command, using the
executable that uses one or more accelerator packages. We are working src/Make.py tool. The Make.py tool is discussed in <A HREF = "Section_start.html#start_4">Section
to create a "make" tool that will perform all these 4 steps in a 2.4</A> of the manual, and its use is
single command. illustrated in the individual accelerator sections. Typically these
steps only need to be done once, to create an executable that uses one
or more accelerator packages.
</P> </P>
<P>The last 4 steps can all be done from the command-line when LAMMPS is <P>The last 4 steps can all be done from the command-line when LAMMPS is
launched, without changing your input script. Or you can add launched, without changing your input script, as illustrated in the
individual accelerator sections. Or you can add
<A HREF = "package.html">package</A> and <A HREF = "suffix.html">suffix</A> commands to your input <A HREF = "package.html">package</A> and <A HREF = "suffix.html">suffix</A> commands to your input
script. script.
</P> </P>
<P>The examples directory has several sub-directories with scripts and <P>IMPORTANT NOTE: With a few exceptions, you can build a single LAMMPS
README files for how to use the following accelerator packages: executable with all its accelerator packages installed. Note that the
USER-INTEL and KOKKOS packages require you to choose one of their
options when building. I.e. CPU or Phi for USER-INTEL. OpenMP, Cuda,
or Phi for KOKKOS. Here are the exceptions; you cannot build a single
executable with:
</P> </P>
<UL><LI>examples/cuda for USER-CUDA package <UL><LI>both the USER-INTEL Phi and KOKKOS Phi options
<LI>examples/gpu for GPU package <LI>the USER-INTEL Phi or Kokkos Phi option, and either the USER-CUDA or GPU packages
<LI>examples/intel for USER-INTEL package
<LI>examples/kokkos for KOKKOS package
</UL> </UL>
<P>Likewise, the bench directory has FERMI and KEPLER sub-directories <P>See the examples/accelerate/README and make.list files for sample
with scripts and README files for using all the accelerator packages. Make.py commands that build LAMMPS with any or all of the accelerator
packages. As an example, here is a command that builds with all the
GPU related packages installed (USER-CUDA, GPU, KOKKOS with Cuda),
including settings to build the needed auxiliary USER-CUDA and GPU
libraries for Kepler GPUs:
</P>
<PRE>Make.py -j 16 -p omp gpu cuda kokkos -cc nvcc wrap=mpi -cuda mode=double arch=35 -gpu mode=double arch=35 \ -kokkos cuda arch=35 lib-all file mpi
</PRE>
<P>The examples/accelerate directory also has input scripts that can be
used with all of the accelerator packages. See its README file for
details.
</P>
<P>Likewise, the bench directory has FERMI and KEPLER and PHI
sub-directories with Make.py commands and input scripts for using all
the accelerator packages on various machines. See the README files in
those dirs.
</P> </P>
<P>As mentioned above, the <A HREF = "http://lammps.sandia.gov/bench.html">Benchmark <P>As mentioned above, the <A HREF = "http://lammps.sandia.gov/bench.html">Benchmark
page</A> of the LAMMPS web site gives page</A> of the LAMMPS web site gives
@ -243,23 +263,25 @@ of the standard LAMMPS benchmark problems, as a function of problem
size and number of compute nodes, on different hardware platforms. size and number of compute nodes, on different hardware platforms.
</P> </P>
<P>Here is a brief summary of what the various packages provide. Details <P>Here is a brief summary of what the various packages provide. Details
are in the individual package sub-sections listed above. are in the individual accelerator sections.
</P> </P>
<UL><LI>Styles with a "cuda" or "gpu" suffix are part of the USER-CUDA or GPU <UL><LI>Styles with a "cuda" or "gpu" suffix are part of the USER-CUDA or GPU
packages, and can be run on NVIDIA GPUs. The speed-up on a GPU packages, and can be run on NVIDIA GPUs. The speed-up on a GPU
depends on a variety of factors, as discussed below. depends on a variety of factors, discussed in the accelerator
sections.
<LI>Styles with an "intel" suffix are part of the USER-INTEL <LI>Styles with an "intel" suffix are part of the USER-INTEL
package. These styles support vectorized single and mixed precision package. These styles support vectorized single and mixed precision
calculations, in addition to full double precision. In extreme cases, calculations, in addition to full double precision. In extreme cases,
this can provide speedups over 3.5x on CPUs. The package also this can provide speedups over 3.5x on CPUs. The package also
supports acceleration with offload to Intel(R) Xeon Phi(TM) supports acceleration in "offload" mode to Intel(R) Xeon Phi(TM)
coprocessors. This can result in additional speedup over 2x depending coprocessors. This can result in additional speedup over 2x depending
on the hardware configuration. on the hardware configuration.
<LI>Styles with a "kk" suffix are part of the KOKKOS package, and can be <LI>Styles with a "kk" suffix are part of the KOKKOS package, and can be
run using OpenMP, on an NVIDIA GPU, or on an Intel Xeon Phi. The run using OpenMP on multicore CPUs, on an NVIDIA GPU, or on an Intel
speed-up depends on a variety of factors, as discussed below. Xeon Phi in "native" mode. The speed-up depends on a variety of
factors, as discussed on the KOKKOS accelerator page.
<LI>Styles with an "omp" suffix are part of the USER-OMP package and allow <LI>Styles with an "omp" suffix are part of the USER-OMP package and allow
a pair-style to be run in multi-threaded mode using OpenMP. This can a pair-style to be run in multi-threaded mode using OpenMP. This can
@ -272,7 +294,7 @@ overload the available bandwidth for communication.
speed-up the pairwise calculations of your simulation by 5-25% on a speed-up the pairwise calculations of your simulation by 5-25% on a
CPU. CPU.
</UL> </UL>
<P>The individual accelerator package sub-sections explain: <P>The individual accelerator package doc pages explain:
</P> </P>
<UL><LI>what hardware and software the accelerated package requires <UL><LI>what hardware and software the accelerated package requires
<LI>how to build LAMMPS with the accelerated package <LI>how to build LAMMPS with the accelerated package

70
doc/Section_accelerate.txt
View File

@ -152,8 +152,8 @@ coprocessors.
All of these commands are in packages provided with LAMMPS. An All of these commands are in packages provided with LAMMPS. An
overview of packages is give in "Section overview of packages is give in "Section
packages"_Section_packages.html. Currently, there are 6 accelerator packages"_Section_packages.html. These are the accelerator packages
packages in LAMMPS, either as standard or user packages: currently in LAMMPS, either as standard or user packages:
"USER-CUDA"_accelerate_cuda.html : for NVIDIA GPUs "USER-CUDA"_accelerate_cuda.html : for NVIDIA GPUs
"GPU"_accelerate_gpu.html : for NVIDIA GPUs as well as OpenCL support "GPU"_accelerate_gpu.html : for NVIDIA GPUs as well as OpenCL support
@ -186,8 +186,8 @@ for that style.
To use an accelerator package in LAMMPS, and one or more of the styles To use an accelerator package in LAMMPS, and one or more of the styles
it provides, follow these general steps. Details vary from package to it provides, follow these general steps. Details vary from package to
package and are explained in the individual accelerator sub-section package and are explained in the individual accelerator doc pages,
doc pages, listed above: listed above:
build the accelerator library | build the accelerator library |
only for USER-CUDA and GPU packages | only for USER-CUDA and GPU packages |
@ -211,26 +211,48 @@ use accelerated styles in your input script |
via "-sf" "command-line switch"_Section_start.html#start_7 or via "-sf" "command-line switch"_Section_start.html#start_7 or
"suffix"_suffix.html command :tb(c=2,s=|) "suffix"_suffix.html command :tb(c=2,s=|)
The first 4 steps typically only need to be done once, to create an The first 4 steps can be done as a single command, using the
executable that uses one or more accelerator packages. We are working src/Make.py tool. The Make.py tool is discussed in "Section
to create a "make" tool that will perform all these 4 steps in a 2.4"_Section_start.html#start_4 of the manual, and its use is
single command. illustrated in the individual accelerator sections. Typically these
steps only need to be done once, to create an executable that uses one
or more accelerator packages.
The last 4 steps can all be done from the command-line when LAMMPS is The last 4 steps can all be done from the command-line when LAMMPS is
launched, without changing your input script. Or you can add launched, without changing your input script, as illustrated in the
individual accelerator sections. Or you can add
"package"_package.html and "suffix"_suffix.html commands to your input "package"_package.html and "suffix"_suffix.html commands to your input
script. script.
The examples directory has several sub-directories with scripts and IMPORTANT NOTE: With a few exceptions, you can build a single LAMMPS
README files for how to use the following accelerator packages: executable with all its accelerator packages installed. Note that the
USER-INTEL and KOKKOS packages require you to choose one of their
options when building. I.e. CPU or Phi for USER-INTEL. OpenMP, Cuda,
or Phi for KOKKOS. Here are the exceptions; you cannot build a single
executable with:
examples/cuda for USER-CUDA package both the USER-INTEL Phi and KOKKOS Phi options
examples/gpu for GPU package the USER-INTEL Phi or Kokkos Phi option, and either the USER-CUDA or GPU packages :ul
examples/intel for USER-INTEL package
examples/kokkos for KOKKOS package :ul
Likewise, the bench directory has FERMI and KEPLER sub-directories See the examples/accelerate/README and make.list files for sample
with scripts and README files for using all the accelerator packages. Make.py commands that build LAMMPS with any or all of the accelerator
packages. As an example, here is a command that builds with all the
GPU related packages installed (USER-CUDA, GPU, KOKKOS with Cuda),
including settings to build the needed auxiliary USER-CUDA and GPU
libraries for Kepler GPUs:
Make.py -j 16 -p omp gpu cuda kokkos -cc nvcc wrap=mpi \
-cuda mode=double arch=35 -gpu mode=double arch=35 \\
-kokkos cuda arch=35 lib-all file mpi :pre
The examples/accelerate directory also has input scripts that can be
used with all of the accelerator packages. See its README file for
details.
Likewise, the bench directory has FERMI and KEPLER and PHI
sub-directories with Make.py commands and input scripts for using all
the accelerator packages on various machines. See the README files in
those dirs.
As mentioned above, the "Benchmark As mentioned above, the "Benchmark
page"_http://lammps.sandia.gov/bench.html of the LAMMPS web site gives page"_http://lammps.sandia.gov/bench.html of the LAMMPS web site gives
@ -239,23 +261,25 @@ of the standard LAMMPS benchmark problems, as a function of problem
size and number of compute nodes, on different hardware platforms. size and number of compute nodes, on different hardware platforms.
Here is a brief summary of what the various packages provide. Details Here is a brief summary of what the various packages provide. Details
are in the individual package sub-sections listed above. are in the individual accelerator sections.
Styles with a "cuda" or "gpu" suffix are part of the USER-CUDA or GPU Styles with a "cuda" or "gpu" suffix are part of the USER-CUDA or GPU
packages, and can be run on NVIDIA GPUs. The speed-up on a GPU packages, and can be run on NVIDIA GPUs. The speed-up on a GPU
depends on a variety of factors, as discussed below. :ulb,l depends on a variety of factors, discussed in the accelerator
sections. :ulb,l
Styles with an "intel" suffix are part of the USER-INTEL Styles with an "intel" suffix are part of the USER-INTEL
package. These styles support vectorized single and mixed precision package. These styles support vectorized single and mixed precision
calculations, in addition to full double precision. In extreme cases, calculations, in addition to full double precision. In extreme cases,
this can provide speedups over 3.5x on CPUs. The package also this can provide speedups over 3.5x on CPUs. The package also
supports acceleration with offload to Intel(R) Xeon Phi(TM) supports acceleration in "offload" mode to Intel(R) Xeon Phi(TM)
coprocessors. This can result in additional speedup over 2x depending coprocessors. This can result in additional speedup over 2x depending
on the hardware configuration. :l on the hardware configuration. :l
Styles with a "kk" suffix are part of the KOKKOS package, and can be Styles with a "kk" suffix are part of the KOKKOS package, and can be
run using OpenMP, on an NVIDIA GPU, or on an Intel Xeon Phi. The run using OpenMP on multicore CPUs, on an NVIDIA GPU, or on an Intel
speed-up depends on a variety of factors, as discussed below. :l Xeon Phi in "native" mode. The speed-up depends on a variety of
factors, as discussed on the KOKKOS accelerator page. :l
Styles with an "omp" suffix are part of the USER-OMP package and allow Styles with an "omp" suffix are part of the USER-OMP package and allow
a pair-style to be run in multi-threaded mode using OpenMP. This can a pair-style to be run in multi-threaded mode using OpenMP. This can
@ -268,7 +292,7 @@ Styles with an "opt" suffix are part of the OPT package and typically
speed-up the pairwise calculations of your simulation by 5-25% on a speed-up the pairwise calculations of your simulation by 5-25% on a
CPU. :l,ule CPU. :l,ule
The individual accelerator package sub-sections explain: The individual accelerator package doc pages explain:
what hardware and software the accelerated package requires what hardware and software the accelerated package requires
how to build LAMMPS with the accelerated package how to build LAMMPS with the accelerated package

534
doc/Section_start.html
View File

@ -85,14 +85,27 @@ launch a LAMMPS Windows executable on a Windows box.
<A NAME = "start_2_1"></A><B><I>Read this first:</I></B> <A NAME = "start_2_1"></A><B><I>Read this first:</I></B>
<P>If you want to avoid building LAMMPS, read the preceeding section <P>If you want to avoid building LAMMPS yourself, read the preceeding
about options available for downloading and installing executables. section about options available for downloading and installing
Details are discussed on the <A HREF = "download">download</A> page. executables. Details are discussed on the <A HREF = "download">download</A> page.
</P> </P>
<P>Building LAMMPS can be simple or not-so-simple. If MPI is already <P>Building LAMMPS can be simple or not-so-simple. If all you need are
installed on your machine (or you just want to run LAMMPS in serial) the default packages installed in LAMMPS, and MPI is already installed
and you can use one of the provided machine Makefiles and the build on your machine, or you just want to run LAMMPS in serial, then you
works on your platform, then it's simple. can typically use the Makefile.mpi or Makefile.serial files in
src/MAKE and type one of these lines (from the src dir):
</P>
<PRE>make mpi
make serial
</PRE>
<P>Or if one of the other Makefile.machine files in the src/MAKE
sub-directories matches your system (type "make" to see a list), you
can use it as-is by typing (for example):
</P>
<PRE>make stampede
</PRE>
<P>If any of these builds with an existing Makefile.machine works on your
system, then you're done!
</P> </P>
<P>If you want to do one of these: <P>If you want to do one of these:
</P> </P>
@ -105,7 +118,14 @@ works on your platform, then it's simple.
auxiliary libraries exist on your machine or install them if they auxiliary libraries exist on your machine or install them if they
don't. You may need to build additional libraries that are part of don't. You may need to build additional libraries that are part of
the LAMMPS package, before building LAMMPS. You may need to edit a the LAMMPS package, before building LAMMPS. You may need to edit a
machine Makefile to make it compatible with your system. Makefile.machine file to make it compatible with your system.
</P>
<P>Note that there is a Make.py tool in the src directory that automates
several of these steps, but you still have to know what you are doing.
<A HREF = "#start_4">Section 2.4</A> below describes the tool. It is a convenient
way to work with installing/un-installing various packages, the
Makefile.machine changes required by some packages, and the auxiliary
libraries some of them use.
</P> </P>
<P>Please read the following sections carefully. If you are not <P>Please read the following sections carefully. If you are not
comfortable with makefiles, or building codes on a Unix platform, or comfortable with makefiles, or building codes on a Unix platform, or
@ -121,7 +141,7 @@ please post the issue to the <A HREF = "http://lammps.sandia.gov/mail.html">LAMM
list</A>. list</A>.
</P> </P>
<P>If you succeed in building LAMMPS on a new kind of machine, for which <P>If you succeed in building LAMMPS on a new kind of machine, for which
there isn't a similar machine Makefile included in the src/MAKE there isn't a similar machine Makefile included in the src/MAKE/MORE
directory, then send it to the developers and we can include it in the directory, then send it to the developers and we can include it in the
LAMMPS distribution. LAMMPS distribution.
</P> </P>
@ -133,21 +153,43 @@ LAMMPS distribution.
</P> </P>
<P>The src directory contains the C++ source and header files for LAMMPS. <P>The src directory contains the C++ source and header files for LAMMPS.
It also contains a top-level Makefile and a MAKE sub-directory with It also contains a top-level Makefile and a MAKE sub-directory with
low-level Makefile.* files for many machines. From within the src low-level Makefile.* files for many systems and machines. See the
directory, type "make" or "gmake". You should see a list of available src/MAKE/README file for a quick overview of what files are available
choices. If one of those is the machine and options you want, you can and what sub-directories they are in.
type a command like:
</P> </P>
<PRE>make linux <P>The src/MAKE dir has a few files that should work as-is on many
platforms. The src/MAKE/OPTIONS dir has more that inovke additional
compiler, MPI, and other setting options commonly used by LAMMPS, to
illustrate their syntax. The src/MAKE/MACHINES dir has many more that
have been tweaked or optimized for specific machines. These files are
all good starting points if you find you need to change them for your
machine. Put any file you edit into the src/MAKE/MINE directory and
it will be never be touched by any LAMMPS updates.
</P>
<P>From within the src directory, type "make" or "gmake". You should see
a list of available choices from src/MAKE and all of its
sub-directories. If one of those has the options you want or is the
machine you want, you can type a command like:
</P>
<PRE>make mpi
or
make serial_icc
or or
gmake mac gmake mac
</PRE> </PRE>
<P>Note that the corresponding Makefile.machine can exist in src/MAKE or
any of its sub-directories. If a file with the same name appears in
multiple places (not a good idea), the order they are used is as
follows: src/MAKE/MINE, src/MAKE, src/MAKE/OPTIONS, src/MAKE/MACHINES.
This gives preference to a file you have created/edited and put in
src/MAKE/MINE.
</P>
<P>Note that on a multi-processor or multi-core platform you can launch a <P>Note that on a multi-processor or multi-core platform you can launch a
parallel make, by using the "-j" switch with the make command, which parallel make, by using the "-j" switch with the make command, which
will build LAMMPS more quickly. will build LAMMPS more quickly.
</P> </P>
<P>If you get no errors and an executable like lmp_linux or lmp_mac is <P>If you get no errors and an executable like lmp_mpi or lmp_g++_serial
produced, you're done; it's your lucky day. or lmp_mac is produced, then you're done; it's your lucky day.
</P> </P>
<P>Note that by default only a few of LAMMPS optional packages are <P>Note that by default only a few of LAMMPS optional packages are
installed. To build LAMMPS with optional packages, see <A HREF = "#start_3">this installed. To build LAMMPS with optional packages, see <A HREF = "#start_3">this
@ -157,43 +199,47 @@ section</A> below.
</P> </P>
<P>If Step 0 did not work, you will need to create a low-level Makefile <P>If Step 0 did not work, you will need to create a low-level Makefile
for your machine, like Makefile.foo. You should make a copy of an for your machine, like Makefile.foo. You should make a copy of an
existing src/MAKE/Makefile.* as a starting point. The only portions existing Makefile.* in src/MAKE or one of its sub-directories as a
of the file you need to edit are the first line, the "compiler/linker starting point. The only portions of the file you need to edit are
settings" section, and the "LAMMPS-specific settings" section. the first line, the "compiler/linker settings" section, and the
"LAMMPS-specific settings" section. When it works, put the edited
file in src/MAKE/MINE and it will not be altered by any future LAMMPS
updates.
</P> </P>
<P><B>Step 2</B> <P><B>Step 2</B>
</P> </P>
<P>Change the first line of src/MAKE/Makefile.foo to list the word "foo" <P>Change the first line of Makefile.foo to list the word "foo" after the
after the "#", and whatever other options it will set. This is the "#", and whatever other options it will set. This is the line you
line you will see if you just type "make". will see if you just type "make".
</P> </P>
<P><B>Step 3</B> <P><B>Step 3</B>
</P> </P>
<P>The "compiler/linker settings" section lists compiler and linker <P>The "compiler/linker settings" section lists compiler and linker
settings for your C++ compiler, including optimization flags. You can settings for your C++ compiler, including optimization flags. You can
use g++, the open-source GNU compiler, which is available on all Unix use g++, the open-source GNU compiler, which is available on all Unix
systems. You can also use mpicc which will typically be available if systems. You can also use mpicxx which will typically be available if
MPI is installed on your system, though you should check which actual MPI is installed on your system, though you should check which actual
compiler it wraps. Vendor compilers often produce faster code. On compiler it wraps. Vendor compilers often produce faster code. On
boxes with Intel CPUs, we suggest using the commercial Intel icc boxes with Intel CPUs, we suggest using the Intel icc compiler, which
compiler, which can be downloaded from <A HREF = "http://www.intel.com/software/products/noncom">Intel's compiler site</A>. can be downloaded from <A HREF = "http://www.intel.com/software/products/noncom">Intel's compiler site</A>.
</P> </P>
<P>If building a C++ code on your machine requires additional libraries, <P>If building a C++ code on your machine requires additional libraries,
then you should list them as part of the LIB variable. then you should list them as part of the LIB variable. You should
not need to do this if you use mpicxx.
</P> </P>
<P>The DEPFLAGS setting is what triggers the C++ compiler to create a <P>The DEPFLAGS setting is what triggers the C++ compiler to create a
dependency list for a source file. This speeds re-compilation when dependency list for a source file. This speeds re-compilation when
source (*.cpp) or header (*.h) files are edited. Some compilers do source (*.cpp) or header (*.h) files are edited. Some compilers do
not support dependency file creation, or may use a different switch not support dependency file creation, or may use a different switch
than -D. GNU g++ works with -D. If your compiler can't create than -D. GNU g++ and Intel icc works with -D. If your compiler can't
dependency files, then you'll need to create a Makefile.foo patterned create dependency files, then you'll need to create a Makefile.foo
after Makefile.storm, which uses different rules that do not involve patterned after Makefile.storm, which uses different rules that do not
dependency files. Note that when you build LAMMPS for the first time involve dependency files. Note that when you build LAMMPS for the
on a new platform, a long list of *.d files will be printed out first time on a new platform, a long list of *.d files will be printed
rapidly. This is not an error; it is the Makefile doing its normal out rapidly. This is not an error; it is the Makefile doing its
creation of dependencies. normal creation of dependencies.
</P> </P>
<P><B>Step 4</B> <P><B>Step 4</B>
</P> </P>
@ -277,20 +323,21 @@ Step 6 below for info about building LAMMPS with an FFT library.
<P><B>Step 5</B> <P><B>Step 5</B>
</P> </P>
<P>The 3 MPI variables are used to specify an MPI library to build LAMMPS <P>The 3 MPI variables are used to specify an MPI library to build LAMMPS
with. with. Note that you do not need to set these if you use the MPI
compiler mpicxx for your CC and LINK setting in the section above.
The MPI wrapper knows where to find the needed files.
</P> </P>
<P>If you want LAMMPS to run in parallel, you must have an MPI library <P>If you want LAMMPS to run in parallel, you must have an MPI library
installed on your platform. If you use an MPI-wrapped compiler, such installed on your platform. If MPI is installed on your system in the
as "mpicc" to build LAMMPS, you should be able to leave these 3 usual place (under /usr/local), you also may not need to specify these
variables blank; the MPI wrapper knows where to find the needed files. 3 variables, assuming /usr/local is in your path. On some large
If not, and MPI is installed on your system in the usual place (under parallel machines which use "modules" for their compile/link
/usr/local), you also may not need to specify these 3 variables. On environements, you may simply need to include the correct module in
some large parallel machines which use "modules" for their your build environment, before building LAMMPS. Or the parallel
compile/link environements, you may simply need to include the correct machine may have a vendor-provided MPI which the compiler has no
module in your build environment. Or the parallel machine may have a trouble finding.
vendor-provided MPI which the compiler has no trouble finding.
</P> </P>
<P>Failing this, with these 3 variables you can specify where the mpi.h <P>Failing this, these 3 variables can be used to specify where the mpi.h
file (MPI_INC) and the MPI library file (MPI_PATH) are found and the file (MPI_INC) and the MPI library file (MPI_PATH) are found and the
name of the library file (MPI_LIB). name of the library file (MPI_LIB).
</P> </P>
@ -310,20 +357,22 @@ arise when linking LAMMPS to the MPI library.
</P> </P>
<P>If you just want to run LAMMPS on a single processor, you can use the <P>If you just want to run LAMMPS on a single processor, you can use the
dummy MPI library provided in src/STUBS, since you don't need a true dummy MPI library provided in src/STUBS, since you don't need a true
MPI library installed on your system. See the MPI library installed on your system. See src/MAKE/Makefile.serial
src/MAKE/Makefile.serial file for how to specify the 3 MPI variables for how to specify the 3 MPI variables in this case. You will also
in this case. You will also need to build the STUBS library for your need to build the STUBS library for your platform before making LAMMPS
platform before making LAMMPS itself. To build from the src itself. Note that if you are building with src/MAKE/Makefile.serial,
directory, type "make stubs", or from the STUBS dir, type "make". e.g. by typing "make serial", then the STUBS library is built for you.
This should create a libmpi_stubs.a file suitable for linking to
LAMMPS. If the build fails, you will need to edit the STUBS/Makefile
for your platform.
</P> </P>
<P>The file STUBS/mpi.c provides a CPU timer function called <P>To build the STUBS library from the src directory, type "make stubs",
MPI_Wtime() that calls gettimeofday() . If your system doesn't or from the src/STUBS dir, type "make". This should create a
support gettimeofday() , you'll need to insert code to call another libmpi_stubs.a file suitable for linking to LAMMPS. If the build
timer. Note that the ANSI-standard function clock() rolls over after fails, you will need to edit the STUBS/Makefile for your platform.
an hour or so, and is therefore insufficient for timing long LAMMPS </P>
<P>The file STUBS/mpi.c provides a CPU timer function called MPI_Wtime()
that calls gettimeofday() . If your system doesn't support
gettimeofday() , you'll need to insert code to call another timer.
Note that the ANSI-standard function clock() rolls over after an hour
or so, and is therefore insufficient for timing long LAMMPS
simulations. simulations.
</P> </P>
<P><B>Step 6</B> <P><B>Step 6</B>
@ -410,11 +459,9 @@ section</A> below, before proceeding to Step 9.
</P> </P>
<P><B>Step 9</B> <P><B>Step 9</B>
</P> </P>
<P>That's it. Once you have a correct Makefile.foo, you have installed <P>That's it. Once you have a correct Makefile.foo, and you have
the optional LAMMPS packages you want to include in your build, and pre-built any other needed libraries (e.g. MPI, FFT, etc) all you need
you have pre-built any other needed libraries (e.g. MPI, FFT, package to do from the src directory is type something like this:
libraries), all you need to do from the src directory is type
something like this:
</P> </P>
<PRE>make foo <PRE>make foo
or or
@ -530,7 +577,7 @@ neighbor lists and would run very slowly in terms of CPU secs/timestep.
<A NAME = "start_2_5"></A><B><I>Building for a Mac:</I></B> <A NAME = "start_2_5"></A><B><I>Building for a Mac:</I></B>
<P>OS X is BSD Unix, so it should just work. See the <P>OS X is BSD Unix, so it should just work. See the
src/MAKE/Makefile.mac file. src/MAKE/MACHINES/Makefile.mac and Makefile.mac_mpi files.
</P> </P>
<HR> <HR>
@ -559,7 +606,7 @@ excluded, you can build it yourself.
</P> </P>
<P>One way to do this is install and use cygwin to build LAMMPS with a <P>One way to do this is install and use cygwin to build LAMMPS with a
standard unix style make program, just as you would on a Linux box; standard unix style make program, just as you would on a Linux box;
see src/MAKE/Makefile.cygwin. see src/MAKE/MACHINES/Makefile.cygwin.
</P> </P>
<P>The other way to do this is using Visual Studio and project files. <P>The other way to do this is using Visual Studio and project files.
See the src/WINDOWS directory and its README.txt file for instructions See the src/WINDOWS directory and its README.txt file for instructions
@ -574,8 +621,13 @@ on both a basic build and a customized build with pacakges you select.
<UL><LI><A HREF = "#start_3_1">Package basics</A> <UL><LI><A HREF = "#start_3_1">Package basics</A>
<LI><A HREF = "#start_3_2">Including/excluding packages</A> <LI><A HREF = "#start_3_2">Including/excluding packages</A>
<LI><A HREF = "#start_3_3">Packages that require extra libraries</A> <LI><A HREF = "#start_3_3">Packages that require extra libraries</A>
<LI><A HREF = "#start_3_4">Packages that use make variable settings</A> <LI><A HREF = "#start_3_4">Packages that require Makefile.machine settings</A>
</UL> </UL>
<P>Note that the following <A HREF = "#start_4">Section 2.4</A> describes the Make.py
tool which can be used to install/un-install packages and build the
auxiliary libraries which some of them use. It can also auto-edit a
Makefile.machine to add settings needed by some packages.
</P>
<HR> <HR>
<A NAME = "start_3_1"></A><B><I>Package basics:</I></B> <A NAME = "start_3_1"></A><B><I>Package basics:</I></B>
@ -583,9 +635,11 @@ on both a basic build and a customized build with pacakges you select.
<P>The source code for LAMMPS is structured as a set of core files which <P>The source code for LAMMPS is structured as a set of core files which
are always included, plus optional packages. Packages are groups of are always included, plus optional packages. Packages are groups of
files that enable a specific set of features. For example, force files that enable a specific set of features. For example, force
fields for molecular systems or granular systems are in packages. You fields for molecular systems or granular systems are in packages.
can see the list of all packages by typing "make package" from within </P>
the src directory of the LAMMPS distribution. <P>You can see the list of all packages by typing "make package" from
within the src directory of the LAMMPS distribution. This also lists
various make commands that can be used to manipulate packages.
</P> </P>
<P>If you use a command in a LAMMPS input script that is specific to a <P>If you use a command in a LAMMPS input script that is specific to a
particular package, you must have built LAMMPS with that package, else particular package, you must have built LAMMPS with that package, else
@ -652,10 +706,11 @@ I.e. individual files are only included if their dependencies are
already included. Likewise, if a package is excluded, other files already included. Likewise, if a package is excluded, other files
dependent on that package are also excluded. dependent on that package are also excluded.
</P> </P>
<P>The reason to exclude packages is if you will never run certain kinds <P>If you will never run simulations that use the features in a
of simulations. For some packages, this will keep you from having to particular packages, there is no reason to include it in your build.
build auxiliary libraries (see below), and will also produce a smaller For some packages, this will keep you from having to build auxiliary
executable which may run a bit faster. libraries (see below), and will also produce a smaller executable
which may run a bit faster.
</P> </P>
<P>When you download a LAMMPS tarball, these packages are pre-installed <P>When you download a LAMMPS tarball, these packages are pre-installed
in the src directory: KSPACE, MANYBODY,MOLECULE. When you download in the src directory: KSPACE, MANYBODY,MOLECULE. When you download
@ -666,9 +721,10 @@ pre-installed.
no-name", where "name" is the name of the package in lower-case, e.g. no-name", where "name" is the name of the package in lower-case, e.g.
name = kspace for the KSPACE package or name = user-atc for the name = kspace for the KSPACE package or name = user-atc for the
USER-ATC package. You can also type "make yes-standard", "make USER-ATC package. You can also type "make yes-standard", "make
no-standard", "make yes-user", "make no-user", "make yes-all" or "make no-standard", "make yes-std", "make no-std", "make yes-user", "make
no-all" to include/exclude various sets of packages. Type "make no-user", "make yes-all" or "make no-all" to include/exclude various
package" to see the all of the package-related make options. sets of packages. Type "make package" to see the all of the
package-related make options.
</P> </P>
<P>IMPORTANT NOTE: Inclusion/exclusion of a package works by simply <P>IMPORTANT NOTE: Inclusion/exclusion of a package works by simply
moving files back and forth between the main src directory and moving files back and forth between the main src directory and
@ -682,18 +738,19 @@ sub-directories. You do not normally need to use these commands
unless you are editing LAMMPS files or have downloaded a patch from unless you are editing LAMMPS files or have downloaded a patch from
the LAMMPS WWW site. the LAMMPS WWW site.
</P> </P>
<P>Typing "make package-update" will overwrite src files with files from <P>Typing "make package-update" or "make pu" will overwrite src files
the package sub-directories if the package has been included. It with files from the package sub-directories if the package has been
should be used after a patch is installed, since patches only update included. It should be used after a patch is installed, since patches
the files in the package sub-directory, but not the src files. Typing only update the files in the package sub-directory, but not the src
"make package-overwrite" will overwrite files in the package files. Typing "make package-overwrite" will overwrite files in the
sub-directories with src files. package sub-directories with src files.
</P> </P>
<P>Typing "make package-status" will show which packages are currently <P>Typing "make package-status" or "make ps" will show which packages are
included. Of those that are included, it will list files that are currently included. Of those that are included, it will list files
different in the src directory and package sub-directory. Typing that are different in the src directory and package sub-directory.
"make package-diff" lists all differences between these files. Again, Typing "make package-diff" lists all differences between these files.
type "make package" to see all of the package-related make options. Again, type "make package" to see all of the package-related make
options.
</P> </P>
<HR> <HR>
@ -705,16 +762,16 @@ you get a LAMMPS build error about a missing library, this is likely
the reason. See the <A HREF = "Section_packages.html">Section_packages</A> doc page the reason. See the <A HREF = "Section_packages.html">Section_packages</A> doc page
for a list of packages that have auxiliary libraries. for a list of packages that have auxiliary libraries.
</P> </P>
<P>Code for some of these auxiliary libraries is included in the LAMMPS <P>Code for most of these auxiliary libraries is included in the LAMMPS
distribution under the lib directory. Examples are the USER-ATC and distribution under the lib directory. Examples are the USER-ATC and
MEAM packages. Some auxiliary libraries are NOT included with LAMMPS; MEAM packages. A few auxiliary libraries are NOT included with
to use the associated package you must download and install the LAMMPS; to use the associated package you must download and install
auxiliary library yourself. Examples are the KIM and VORONOI and the auxiliary library yourself. Examples are the KIM and VORONOI and
USER-MOLFILE packages. USER-MOLFILE packages.
</P> </P>
<P>For libraries with provided source code, each lib directory has a <P>For provided libraries, each lib directory has a README file
README file (e.g. lib/reax/README) with instructions on how to build (e.g. lib/reax/README) with instructions on how to build that library.
that library. Typically this is done by typing something like: Typically this is done by typing something like:
</P> </P>
<PRE>make -f Makefile.g++ <PRE>make -f Makefile.g++
</PRE> </PRE>
@ -746,168 +803,205 @@ is built with, typically requires additional Fortran-to-C libraries be
included in the link. Another example are the BLAS and LAPACK included in the link. Another example are the BLAS and LAPACK
libraries needed to use the USER-ATC or USER-AWPMD packages. libraries needed to use the USER-ATC or USER-AWPMD packages.
</P> </P>
<P>For libraries without provided source code, see the <P>For libraries without provided source code, the file
src/package/Makefile.lammps file for information on where to find the src/package/README has information on where to find the library and
library and how to build it. E.g. the file src/KIM/Makefile.lammps or how to build it, e.g. src/VORONOI/README. There is also a
src/VORONOI/Makefile.lammps or src/UESR-MOLFILE/Makefile.lammps. Makefile.lammps file in the src/package directory. E.g. files
These files serve the same purpose as the lib/package/Makefile.lammps src/KIM/Makefile.lammps or src/VORONOI/Makefile.lammps or
files described above. The files have settings needed when LAMMPS is src/UESR-MOLFILE/Makefile.lammps. These files serve the same purpose
built to link with the corresponding auxiliary library. as the lib/package/Makefile.lammps files described above. The files
have settings needed when LAMMPS is built to link with the
corresponding auxiliary library.
</P> </P>
<P>Again, you must insure that the settings in <P>Again, you must insure that the settings in
src/package/Makefile.lammps are appropriate for your system and where src/package/Makefile.lammps are appropriate for your system and where
you installed the auxiliary library. If they are not, the LAMMPS you installed the auxiliary library. If they are not, the LAMMPS
build will fail. build will typically fail.
</P> </P>
<HR> <HR>
<A NAME = "start_3_4"></A><B><I>Packages that use make variable settings</I></B> <A NAME = "start_3_4"></A><B><I>Packages that require Makefile.machine settings</I></B>
<P>One package, the KOKKOS package, allows its build options to be <P>A few packages require specific settings in Makefile.machine, to
specified by setting variables via the "make" command, rather than by either build or use the package effectively. These are the
first building an auxiliary library and editing a Makefile.lammps USER-INTEL, KOKKOS, USER-OMP, and OPT packages. The details of what
file, as discussed in the previous sub-section for other packages. flags to add or what variables to define are given on the doc pages
This is for convenience since it is common to want to experiment with that describe each of these accelerator packages in detail:
different Kokkos library options. Using variables enables a direct
re-build of LAMMPS and its Kokkos dependencies, so that a benchmark
test with different Kokkos options can be quickly performed.
</P> </P>
<P>The syntax for setting make variables is as follows. You must <UL><LI><A HREF = "accelerate_intel.html">USER-INTEL package</A>
use a GNU-compatible make command for this to work. Try "gmake" <LI><A HREF = "accelerate_kokkos.html">KOKKOS package</A>
if your system's standard make complains. <LI><A HREF = "accelerate_omp.html">USER-OMP package</A>
</P> <LI><A HREF = "accelerate_opt.html">OPT package</A>
<PRE>make yes-kokkos
make g++ VAR1=value VAR2=value ...
</PRE>
<P>The first line installs the KOKKOS package, which only needs to be
done once. The second line builds LAMMPS with src/MAKE/Makefile.g++
and optionally sets one or more variables that affect the build. Each
variable is specified in upper-case; its value follows an equal sign
with no spaces. The second line can be repeated with different
variable settings, though a "clean" must be done before the rebuild.
Type "make clean" to see options for this operation.
</P>
<P>These are the variables that can be specified. Each takes a value of
<I>yes</I> or <I>no</I>. The default value is listed, which is set in the
lib/kokkos/Makefile.lammps file. See <A HREF = "Section_accelerate.html#acc_8">this
section</A> for a discussion of what is
meant by "host" and "device" in the Kokkos context.
</P>
<UL><LI>OMP, default = <I>yes</I>
<LI>CUDA, default = <I>no</I>
<LI>HWLOC, default = <I>no</I>
<LI>AVX, default = <I>no</I>
<LI>MIC, default = <I>no</I>
<LI>LIBRT, default = <I>no</I>
<LI>DEBUG, default = <I>no</I>
</UL> </UL>
<P>OMP sets the parallelization method used for Kokkos code (within <P>Here is a brief summary of what Makefile.machine changes are needed.
LAMMPS) that runs on the host. OMP=yes means that OpenMP will be Note that the Make.py tool, described in the next <A HREF = "#start_4">Section
used. OMP=no means that pthreads will be used. 2.4</A> can automatically add the needed info to an existing
machine Makefile, using simple command-line arguments.
</P> </P>
<P>CUDA sets the parallelization method used for Kokkos code (within <P>In src/MAKE/OPTIONS see the following Makefiles for examples of the
LAMMPS) that runs on the device. CUDA=yes means an NVIDIA GPU running changes described below:
CUDA will be used. CUDA=no means that the OMP=yes or OMP=no setting
will be used for the device as well as the host.
</P> </P>
<P>If CUDA=yes, then the lo-level Makefile in the src/MAKE directory must <UL><LI>Makefile.intel_cpu
use "nvcc" as its compiler, via its CC setting. For best performance <LI>Makefile.intel_phi
its CCFLAGS setting should use -O3 and have an -arch setting that <LI>Makefile.kokkos_omp
matches the compute capability of your NVIDIA hardware and software <LI>Makefile.kokkos_cuda
installation, e.g. -arch=sm_20. Generally Fermi Generation GPUs are <LI>Makefile.kokkos_phi
sm_20, while Kepler generation GPUs are sm_30 or sm_35 and Maxwell <LI>Makefile.omp
cards are sm_50. A complete list can be found on </UL>
<A HREF = "http://en.wikipedia.org/wiki/CUDA#Supported_GPUs">wikipedia</A>. You can <P>For the USER-INTEL package, you have 2 choices when building. You can
also use the deviceQuery tool that comes with the CUDA samples. Note build with CPU or Phi support. The latter uses Xeon Phi chips in
the minimal required compute capability is 2.0, but this will give "offload" mode. Each of these modes requires additional settings in
signicantly reduced performance compared to Kepler generation GPUs your Makefile.machine for CCFLAGS and LINKFLAGS.
with compute capability 3.x. For the LINK setting, "nvcc" should not
be used; instead use g++ or another compiler suitable for linking C++
applications. Often you will want to use your MPI compiler wrapper
for this setting (i.e. mpicxx). Finally, the lo-level Makefile must
also have a "Compilation rule" for creating *.o files from *.cu files.
See src/Makefile.cuda for an example of a lo-level Makefile with all
of these settings.
</P> </P>
<P>HWLOC binds threads to hardware cores, so they do not migrate during a <P>For CPU mode (if using an Intel compiler):
simulation. HWLOC=yes should always be used if running with OMP=no
for pthreads. It is not necessary for OMP=yes for OpenMP, because
OpenMP provides alternative methods via environment variables for
binding threads to hardware cores. More info on binding threads to
cores is given in <A HREF = "Section_accelerate.html#acc_8">this section</A>.
</P> </P>
<P>AVX enables Intel advanced vector extensions when compiling for an <UL><LI>CCFLAGS: add -fopenmp, -DLAMMPS_MEMALIGN=64, -restrict, -xHost, -fno-alias, -ansi-alias, -override-limits
Intel-compatible chip. AVX=yes should only be set if your host <LI>LINKFLAGS: add -fopenmp
hardware supports AVX. If it does not support it, this will cause a </UL>
run-time crash. <P>For Phi mode add the following in addition to the CPU mode flags:
</P> </P>
<P>MIC enables compiler switches needed when compling for an Intel Phi <UL><LI>CCFLAGS: add -DLMP_INTEL_OFFLOAD and
processor. <LI>LINKFLAGS: add -offload
</UL>
<P>And also add this to CCFLAGS:
</P> </P>
<P>LIBRT enables use of a more accurate timer mechanism on most Unix <PRE>-offload-option,mic,compiler,"-fp-model fast=2 -mGLOB_default_function_attrs=\"gather_scatter_loop_unroll=4\""
platforms. This library is not available on all platforms. </PRE>
<P>For the KOKKOS package, you have 3 choices when building. You can
build with OMP or Cuda or Phi support. Phi support uses Xeon Phi
chips in "native" mode. This can be done by setting the following
variables in your Makefile.machine:
</P> </P>
<P>DEBUG is only useful when developing a Kokkos-enabled style within <UL><LI>for OMP support, set OMP = yes
LAMMPS. DEBUG=yes enables printing of run-time debugging information <LI>for Cuda support, set OMP = yes and CUDA = yes
that can be useful. It also enables runtime bounds checking on Kokkos <LI>for Phi support, set OMP = yes and MIC = yes
data structures. </UL>
<P>These can also be set as additional arguments to the make command, e.g.
</P> </P>
<PRE>make g++ OMP=yes MIC=yes
</PRE>
<P>Building the KOKKOS package with CUDA support requires a Makefile
machine that uses the NVIDIA "nvcc" compiler, as well as an
appropriate "arch" setting appropriate to the GPU hardware and NVIDIA
software you have on your machine. See
src/MAKE/OPTIONS/Makefile.kokkos_cuda for an example of such a machine
Makefile.
</P>
<P>For the USER-OMP package, your Makefile.machine needs additional
settings for CCFLAGS and LINKFLAGS.
</P>
<UL><LI>CCFLAGS: add -fopenmp and -restrict
<LI>LINKFLAGS: add -fopenmp
</UL>
<P>For the OPT package, your Makefile.machine needs an additional
settings for CCFLAGS.
</P>
<UL><LI>CCFLAGS: add -restrict
</UL>
<HR> <HR>
<H4><A NAME = "start_4"></A>2.4 Building LAMMPS via the Make.py script <H4><A NAME = "start_4"></A>2.4 Building LAMMPS via the Make.py script
</H4> </H4>
<P>The src directory includes a Make.py script, written <P>The src directory includes a Make.py script, written in Python, which
in Python, which can be used to automate various steps can be used to automate various steps of the build process. It is
of the build process. particularly useful for working with the accelerator packages, as well
as other packages which require auxiliary libraries to be built.
</P> </P>
<P>You can run the script from the src directory by typing either: <P>You can run Make.py from the src directory by typing either:
</P> </P>
<PRE>Make.py <PRE>Make.py -h
python Make.py python Make.py -h
</PRE> </PRE>
<P>which will give you info about the tool. For the former to work, you <P>which will give you help info about the tool. For the former to work,
may need to edit the 1st line of the script to point to your local you may need to edit the first line of Make.py to point to your local
Python. And you may need to insure the script is executable: Python. And you may need to insure the script is executable:
</P> </P>
<PRE>chmod +x Make.py <PRE>chmod +x Make.py
</PRE> </PRE>
<P>The following options are supported as switches: <P>Here are examples of build tasks you can perform with Make.py:
</P> </P>
<UL><LI>-i file1 file2 ... <DIV ALIGN=center><TABLE BORDER=1 >
<LI>-p package1 package2 ... <TR><TD >Install/uninstall packages</TD><TD > Make.py -p no-lib kokkos omp intel</TD></TR>
<LI>-u package1 package2 ... <TR><TD >Build specific auxiliary libs</TD><TD > Make.py lib-atc lib-meam</TD></TR>
<LI>-e package1 arg1 arg2 package2 ... <TR><TD >Build libs for all installed packages</TD><TD > Make.py -p cuda gpu -gpu mode=double arch=31 lib-all</TD></TR>
<LI>-o dir <TR><TD >Create a Makefile from scratch with a compiler and MPI</TD><TD > Make.py -m none -cc g++ -mpi mpich file</TD></TR>
<LI>-b machine <TR><TD >Augment Makefile.serial with settings for installed packages</TD><TD > Make.py -p intel -intel cpu -m serial file</TD></TR>
<LI>-s suffix1 suffix2 ... <TR><TD >Add JPG and FFTW support to Makefile.mpi</TD><TD > Make.py -m mpi -jpg -fft fftw file</TD></TR>
<LI>-l dir <TR><TD >Build LAMMPS with a parallel make using Makefile.mpi</TD><TD > Make.py -j 16 -m mpi exe</TD></TR>
<LI>-j N <TR><TD >Build LAMMPS and libs it needs using Makefile.serial with accelerator settings</TD><TD > Make.py -p gpu intel -intel cpu lib-all file serial
<LI>-h switch1 switch2 ... </TD></TR></TABLE></DIV>
<P>The bench and examples directories give Make.py commands that can be
used to build LAMMPS with the various packages and options needed to
run all the benchmark and example input scripts. See these files for
more details:
</P>
<UL><LI>bench/README
<LI>bench/FERMI/README
<LI>bench/KEPLER/README
<LI>bench/PHI/README
<LI>examples/README
<LI>examples/accelerate/README
<LI>examples/accelerate/make.list
</UL> </UL>
<P>Help on any switch can be listed by using -h, e.g. <P>All of the Make.py options and syntax help can be accessed by using
the "-h" switch.
</P> </P>
<PRE>Make.py -h -i -p <P>E.g. typing "Make.py -h" gives
</P>
<PRE>Syntax: Make.py switch args ... <I>action1</I> <I>action2</I> ...
actions:
lib-all, lib-dir, clean, file, exe or machine
zero or more actions, in any order (machine must be last)
switches:
-d (dir), -j (jmake), -m (makefile), -o (output),
-p (packages), -r (redo), -s (settings), -v (verbose)
switches for libs:
-atc, -awpmd, -colvars, -cuda
-gpu, -meam, -poems, -qmmm, -reax
switches for build and makefile options:
-intel, -kokkos, -cc, -mpi, -fft, -jpg, -png
</PRE> </PRE>
<P>At a hi-level, these are the kinds of package management <P>Using the "-h" switch with other switches and actions gives additional
and build tasks that can be performed easily, using info on all the other specified switches or actions. The "-h" can be
the Make.py tool: anywhere in the command-line and the other switches do not need their
arguments. E.g. type "Make.py -h -d -atc -intel" will print:
</P> </P>
<UL><LI>install/uninstall packages and build the associated external libs (use -p and -u and -e) <PRE>-d dir
<LI>install packages needed for one or more input scripts (use -i and -p) dir = LAMMPS home dir
<LI>build LAMMPS, either in the src dir or new dir (use -b) if -d not specified, working dir must be lammps/src
<LI>create a new dir with only the source code needed for one or more input scripts (use -i and -o) </PRE>
</UL> <PRE>-atc make=suffix lammps=suffix2
<P>The last bullet can be useful when you wish to build a stripped-down all args are optional and can be in any order
version of LAMMPS to run a specific script(s). Or when you wish to make = use Makefile.suffix (def = g++)
move the minimal amount of files to another platform for a remote lammps = use Makefile.lammps.suffix2 (def = EXTRAMAKE in makefile)
LAMMPS build. </PRE>
<PRE>-intel mode
mode = cpu or phi (def = cpu)
build Intel package for CPU or Xeon Phi
</PRE>
<P>Note that Make.py never overwrites an existing Makefile.machine.
Instead, it creates src/MAKE/MINE/Makefile.auto, which you can save or
rename if desired. Likewise it creates an executable named
src/lmp_auto, which you can rename using the -o switch if desired.
</P> </P>
<P>Note that using Make.py is not a substitute for insuring you have a <P>The most recently executed Make.py commmand is saved in
valid src/MAKE/Makefile.foo for your system, or that external library src/Make.py.last. You can use the "-r" switch (for redo) to re-invoke
Makefiles in any lib/* directories you use are also valid for your the last command, or you can save a sequence of one or more Make.py
system. But once you have done that, you can use Make.py to quickly commands to a file and invoke the file of commands using "-r". You
include/exclude the packages and external libraries needed by your can also label the commands in the file and invoke one or more of them
input scripts. by name.
</P>
<P>A typical use of Make.py is to start with a valid Makefile.machine for
your system, that works for a vanilla LAMMPS build, i.e. when optional
packages are not installed. You can then use Make.py to add various
settings (FFT, JPG, PNG) to the Makefile.machine as well as change its
compiler and MPI options. You can also add additional packages to the
build, as well as build the needed supporting libraries.
</P>
<P>You can also use Make.py to create a new Makefile.machine from
scratch, using the "-m none" switch, if you also specify what compiler
and MPI options to use, via the "-cc" and "-mpi" switches.
</P> </P>
<HR> <HR>

522
doc/Section_start.txt
View File

@ -79,14 +79,27 @@ This section has the following sub-sections:
[{Read this first:}] :link(start_2_1) [{Read this first:}] :link(start_2_1)
If you want to avoid building LAMMPS, read the preceeding section If you want to avoid building LAMMPS yourself, read the preceeding
about options available for downloading and installing executables. section about options available for downloading and installing
Details are discussed on the "download"_download page. executables. Details are discussed on the "download"_download page.
Building LAMMPS can be simple or not-so-simple. If MPI is already Building LAMMPS can be simple or not-so-simple. If all you need are
installed on your machine (or you just want to run LAMMPS in serial) the default packages installed in LAMMPS, and MPI is already installed
and you can use one of the provided machine Makefiles and the build on your machine, or you just want to run LAMMPS in serial, then you
works on your platform, then it's simple. can typically use the Makefile.mpi or Makefile.serial files in
src/MAKE and type one of these lines (from the src dir):
make mpi
make serial :pre
Or if one of the other Makefile.machine files in the src/MAKE
sub-directories matches your system (type "make" to see a list), you
can use it as-is by typing (for example):
make stampede :pre
If any of these builds with an existing Makefile.machine works on your
system, then you're done!
If you want to do one of these: If you want to do one of these:
@ -99,7 +112,14 @@ then building LAMMPS is more complicated. You may need to find where
auxiliary libraries exist on your machine or install them if they auxiliary libraries exist on your machine or install them if they
don't. You may need to build additional libraries that are part of don't. You may need to build additional libraries that are part of
the LAMMPS package, before building LAMMPS. You may need to edit a the LAMMPS package, before building LAMMPS. You may need to edit a
machine Makefile to make it compatible with your system. Makefile.machine file to make it compatible with your system.
Note that there is a Make.py tool in the src directory that automates
several of these steps, but you still have to know what you are doing.
"Section 2.4"_#start_4 below describes the tool. It is a convenient
way to work with installing/un-installing various packages, the
Makefile.machine changes required by some packages, and the auxiliary
libraries some of them use.
Please read the following sections carefully. If you are not Please read the following sections carefully. If you are not
comfortable with makefiles, or building codes on a Unix platform, or comfortable with makefiles, or building codes on a Unix platform, or
@ -115,7 +135,7 @@ please post the issue to the "LAMMPS mail
list"_http://lammps.sandia.gov/mail.html. list"_http://lammps.sandia.gov/mail.html.
If you succeed in building LAMMPS on a new kind of machine, for which If you succeed in building LAMMPS on a new kind of machine, for which
there isn't a similar machine Makefile included in the src/MAKE there isn't a similar machine Makefile included in the src/MAKE/MORE
directory, then send it to the developers and we can include it in the directory, then send it to the developers and we can include it in the
LAMMPS distribution. LAMMPS distribution.
@ -127,21 +147,43 @@ LAMMPS distribution.
The src directory contains the C++ source and header files for LAMMPS. The src directory contains the C++ source and header files for LAMMPS.
It also contains a top-level Makefile and a MAKE sub-directory with It also contains a top-level Makefile and a MAKE sub-directory with
low-level Makefile.* files for many machines. From within the src low-level Makefile.* files for many systems and machines. See the
directory, type "make" or "gmake". You should see a list of available src/MAKE/README file for a quick overview of what files are available
choices. If one of those is the machine and options you want, you can and what sub-directories they are in.
type a command like:
make linux The src/MAKE dir has a few files that should work as-is on many
platforms. The src/MAKE/OPTIONS dir has more that inovke additional
compiler, MPI, and other setting options commonly used by LAMMPS, to
illustrate their syntax. The src/MAKE/MACHINES dir has many more that
have been tweaked or optimized for specific machines. These files are
all good starting points if you find you need to change them for your
machine. Put any file you edit into the src/MAKE/MINE directory and
it will be never be touched by any LAMMPS updates.
From within the src directory, type "make" or "gmake". You should see
a list of available choices from src/MAKE and all of its
sub-directories. If one of those has the options you want or is the
machine you want, you can type a command like:
make mpi
or
make serial_icc
or or
gmake mac :pre gmake mac :pre
Note that the corresponding Makefile.machine can exist in src/MAKE or
any of its sub-directories. If a file with the same name appears in
multiple places (not a good idea), the order they are used is as
follows: src/MAKE/MINE, src/MAKE, src/MAKE/OPTIONS, src/MAKE/MACHINES.
This gives preference to a file you have created/edited and put in
src/MAKE/MINE.
Note that on a multi-processor or multi-core platform you can launch a Note that on a multi-processor or multi-core platform you can launch a
parallel make, by using the "-j" switch with the make command, which parallel make, by using the "-j" switch with the make command, which
will build LAMMPS more quickly. will build LAMMPS more quickly.
If you get no errors and an executable like lmp_linux or lmp_mac is If you get no errors and an executable like lmp_mpi or lmp_g++_serial
produced, you're done; it's your lucky day. or lmp_mac is produced, then you're done; it's your lucky day.
Note that by default only a few of LAMMPS optional packages are Note that by default only a few of LAMMPS optional packages are
installed. To build LAMMPS with optional packages, see "this installed. To build LAMMPS with optional packages, see "this
@ -151,43 +193,47 @@ section"_#start_3 below.
If Step 0 did not work, you will need to create a low-level Makefile If Step 0 did not work, you will need to create a low-level Makefile
for your machine, like Makefile.foo. You should make a copy of an for your machine, like Makefile.foo. You should make a copy of an
existing src/MAKE/Makefile.* as a starting point. The only portions existing Makefile.* in src/MAKE or one of its sub-directories as a
of the file you need to edit are the first line, the "compiler/linker starting point. The only portions of the file you need to edit are
settings" section, and the "LAMMPS-specific settings" section. the first line, the "compiler/linker settings" section, and the
"LAMMPS-specific settings" section. When it works, put the edited
file in src/MAKE/MINE and it will not be altered by any future LAMMPS
updates.
[Step 2] [Step 2]
Change the first line of src/MAKE/Makefile.foo to list the word "foo" Change the first line of Makefile.foo to list the word "foo" after the
after the "#", and whatever other options it will set. This is the "#", and whatever other options it will set. This is the line you
line you will see if you just type "make". will see if you just type "make".
[Step 3] [Step 3]
The "compiler/linker settings" section lists compiler and linker The "compiler/linker settings" section lists compiler and linker
settings for your C++ compiler, including optimization flags. You can settings for your C++ compiler, including optimization flags. You can
use g++, the open-source GNU compiler, which is available on all Unix use g++, the open-source GNU compiler, which is available on all Unix
systems. You can also use mpicc which will typically be available if systems. You can also use mpicxx which will typically be available if
MPI is installed on your system, though you should check which actual MPI is installed on your system, though you should check which actual
compiler it wraps. Vendor compilers often produce faster code. On compiler it wraps. Vendor compilers often produce faster code. On
boxes with Intel CPUs, we suggest using the commercial Intel icc boxes with Intel CPUs, we suggest using the Intel icc compiler, which
compiler, which can be downloaded from "Intel's compiler site"_intel. can be downloaded from "Intel's compiler site"_intel.
:link(intel,http://www.intel.com/software/products/noncom) :link(intel,http://www.intel.com/software/products/noncom)
If building a C++ code on your machine requires additional libraries, If building a C++ code on your machine requires additional libraries,
then you should list them as part of the LIB variable. then you should list them as part of the LIB variable. You should
not need to do this if you use mpicxx.
The DEPFLAGS setting is what triggers the C++ compiler to create a The DEPFLAGS setting is what triggers the C++ compiler to create a
dependency list for a source file. This speeds re-compilation when dependency list for a source file. This speeds re-compilation when
source (*.cpp) or header (*.h) files are edited. Some compilers do source (*.cpp) or header (*.h) files are edited. Some compilers do
not support dependency file creation, or may use a different switch not support dependency file creation, or may use a different switch
than -D. GNU g++ works with -D. If your compiler can't create than -D. GNU g++ and Intel icc works with -D. If your compiler can't
dependency files, then you'll need to create a Makefile.foo patterned create dependency files, then you'll need to create a Makefile.foo
after Makefile.storm, which uses different rules that do not involve patterned after Makefile.storm, which uses different rules that do not
dependency files. Note that when you build LAMMPS for the first time involve dependency files. Note that when you build LAMMPS for the
on a new platform, a long list of *.d files will be printed out first time on a new platform, a long list of *.d files will be printed
rapidly. This is not an error; it is the Makefile doing its normal out rapidly. This is not an error; it is the Makefile doing its
creation of dependencies. normal creation of dependencies.
[Step 4] [Step 4]
@ -271,20 +317,21 @@ Step 6 below for info about building LAMMPS with an FFT library.
[Step 5] [Step 5]
The 3 MPI variables are used to specify an MPI library to build LAMMPS The 3 MPI variables are used to specify an MPI library to build LAMMPS
with. with. Note that you do not need to set these if you use the MPI
compiler mpicxx for your CC and LINK setting in the section above.
The MPI wrapper knows where to find the needed files.
If you want LAMMPS to run in parallel, you must have an MPI library If you want LAMMPS to run in parallel, you must have an MPI library
installed on your platform. If you use an MPI-wrapped compiler, such installed on your platform. If MPI is installed on your system in the
as "mpicc" to build LAMMPS, you should be able to leave these 3 usual place (under /usr/local), you also may not need to specify these
variables blank; the MPI wrapper knows where to find the needed files. 3 variables, assuming /usr/local is in your path. On some large
If not, and MPI is installed on your system in the usual place (under parallel machines which use "modules" for their compile/link
/usr/local), you also may not need to specify these 3 variables. On environements, you may simply need to include the correct module in
some large parallel machines which use "modules" for their your build environment, before building LAMMPS. Or the parallel
compile/link environements, you may simply need to include the correct machine may have a vendor-provided MPI which the compiler has no
module in your build environment. Or the parallel machine may have a trouble finding.
vendor-provided MPI which the compiler has no trouble finding.
Failing this, with these 3 variables you can specify where the mpi.h Failing this, these 3 variables can be used to specify where the mpi.h
file (MPI_INC) and the MPI library file (MPI_PATH) are found and the file (MPI_INC) and the MPI library file (MPI_PATH) are found and the
name of the library file (MPI_LIB). name of the library file (MPI_LIB).
@ -304,20 +351,22 @@ arise when linking LAMMPS to the MPI library.
If you just want to run LAMMPS on a single processor, you can use the If you just want to run LAMMPS on a single processor, you can use the
dummy MPI library provided in src/STUBS, since you don't need a true dummy MPI library provided in src/STUBS, since you don't need a true
MPI library installed on your system. See the MPI library installed on your system. See src/MAKE/Makefile.serial
src/MAKE/Makefile.serial file for how to specify the 3 MPI variables for how to specify the 3 MPI variables in this case. You will also
in this case. You will also need to build the STUBS library for your need to build the STUBS library for your platform before making LAMMPS
platform before making LAMMPS itself. To build from the src itself. Note that if you are building with src/MAKE/Makefile.serial,
directory, type "make stubs", or from the STUBS dir, type "make". e.g. by typing "make serial", then the STUBS library is built for you.
This should create a libmpi_stubs.a file suitable for linking to
LAMMPS. If the build fails, you will need to edit the STUBS/Makefile
for your platform.
The file STUBS/mpi.c provides a CPU timer function called To build the STUBS library from the src directory, type "make stubs",
MPI_Wtime() that calls gettimeofday() . If your system doesn't or from the src/STUBS dir, type "make". This should create a
support gettimeofday() , you'll need to insert code to call another libmpi_stubs.a file suitable for linking to LAMMPS. If the build
timer. Note that the ANSI-standard function clock() rolls over after fails, you will need to edit the STUBS/Makefile for your platform.
an hour or so, and is therefore insufficient for timing long LAMMPS
The file STUBS/mpi.c provides a CPU timer function called MPI_Wtime()
that calls gettimeofday() . If your system doesn't support
gettimeofday() , you'll need to insert code to call another timer.
Note that the ANSI-standard function clock() rolls over after an hour
or so, and is therefore insufficient for timing long LAMMPS
simulations. simulations.
[Step 6] [Step 6]
@ -404,11 +453,9 @@ section"_#start_3 below, before proceeding to Step 9.
[Step 9] [Step 9]
That's it. Once you have a correct Makefile.foo, you have installed That's it. Once you have a correct Makefile.foo, and you have
the optional LAMMPS packages you want to include in your build, and pre-built any other needed libraries (e.g. MPI, FFT, etc) all you need
you have pre-built any other needed libraries (e.g. MPI, FFT, package to do from the src directory is type something like this:
libraries), all you need to do from the src directory is type
something like this:
make foo make foo
or or
@ -524,7 +571,7 @@ neighbor lists and would run very slowly in terms of CPU secs/timestep.
[{Building for a Mac:}] :link(start_2_5) [{Building for a Mac:}] :link(start_2_5)
OS X is BSD Unix, so it should just work. See the OS X is BSD Unix, so it should just work. See the
src/MAKE/Makefile.mac file. src/MAKE/MACHINES/Makefile.mac and Makefile.mac_mpi files.
:line :line
@ -553,7 +600,7 @@ excluded, you can build it yourself.
One way to do this is install and use cygwin to build LAMMPS with a One way to do this is install and use cygwin to build LAMMPS with a
standard unix style make program, just as you would on a Linux box; standard unix style make program, just as you would on a Linux box;
see src/MAKE/Makefile.cygwin. see src/MAKE/MACHINES/Makefile.cygwin.
The other way to do this is using Visual Studio and project files. The other way to do this is using Visual Studio and project files.
See the src/WINDOWS directory and its README.txt file for instructions See the src/WINDOWS directory and its README.txt file for instructions
@ -568,7 +615,12 @@ This section has the following sub-sections:
"Package basics"_#start_3_1 "Package basics"_#start_3_1
"Including/excluding packages"_#start_3_2 "Including/excluding packages"_#start_3_2
"Packages that require extra libraries"_#start_3_3 "Packages that require extra libraries"_#start_3_3
"Packages that use make variable settings"_#start_3_4 :ul "Packages that require Makefile.machine settings"_#start_3_4 :ul
Note that the following "Section 2.4"_#start_4 describes the Make.py
tool which can be used to install/un-install packages and build the
auxiliary libraries which some of them use. It can also auto-edit a
Makefile.machine to add settings needed by some packages.
:line :line
@ -577,9 +629,11 @@ This section has the following sub-sections:
The source code for LAMMPS is structured as a set of core files which The source code for LAMMPS is structured as a set of core files which
are always included, plus optional packages. Packages are groups of are always included, plus optional packages. Packages are groups of
files that enable a specific set of features. For example, force files that enable a specific set of features. For example, force
fields for molecular systems or granular systems are in packages. You fields for molecular systems or granular systems are in packages.
can see the list of all packages by typing "make package" from within
the src directory of the LAMMPS distribution. You can see the list of all packages by typing "make package" from
within the src directory of the LAMMPS distribution. This also lists
various make commands that can be used to manipulate packages.
If you use a command in a LAMMPS input script that is specific to a If you use a command in a LAMMPS input script that is specific to a
particular package, you must have built LAMMPS with that package, else particular package, you must have built LAMMPS with that package, else
@ -646,10 +700,11 @@ I.e. individual files are only included if their dependencies are
already included. Likewise, if a package is excluded, other files already included. Likewise, if a package is excluded, other files
dependent on that package are also excluded. dependent on that package are also excluded.
The reason to exclude packages is if you will never run certain kinds If you will never run simulations that use the features in a
of simulations. For some packages, this will keep you from having to particular packages, there is no reason to include it in your build.
build auxiliary libraries (see below), and will also produce a smaller For some packages, this will keep you from having to build auxiliary
executable which may run a bit faster. libraries (see below), and will also produce a smaller executable
which may run a bit faster.
When you download a LAMMPS tarball, these packages are pre-installed When you download a LAMMPS tarball, these packages are pre-installed
in the src directory: KSPACE, MANYBODY,MOLECULE. When you download in the src directory: KSPACE, MANYBODY,MOLECULE. When you download
@ -660,9 +715,10 @@ Packages are included or excluded by typing "make yes-name" or "make
no-name", where "name" is the name of the package in lower-case, e.g. no-name", where "name" is the name of the package in lower-case, e.g.
name = kspace for the KSPACE package or name = user-atc for the name = kspace for the KSPACE package or name = user-atc for the
USER-ATC package. You can also type "make yes-standard", "make USER-ATC package. You can also type "make yes-standard", "make
no-standard", "make yes-user", "make no-user", "make yes-all" or "make no-standard", "make yes-std", "make no-std", "make yes-user", "make
no-all" to include/exclude various sets of packages. Type "make no-user", "make yes-all" or "make no-all" to include/exclude various
package" to see the all of the package-related make options. sets of packages. Type "make package" to see the all of the
package-related make options.
IMPORTANT NOTE: Inclusion/exclusion of a package works by simply IMPORTANT NOTE: Inclusion/exclusion of a package works by simply
moving files back and forth between the main src directory and moving files back and forth between the main src directory and
@ -676,18 +732,19 @@ sub-directories. You do not normally need to use these commands
unless you are editing LAMMPS files or have downloaded a patch from unless you are editing LAMMPS files or have downloaded a patch from
the LAMMPS WWW site. the LAMMPS WWW site.
Typing "make package-update" will overwrite src files with files from Typing "make package-update" or "make pu" will overwrite src files
the package sub-directories if the package has been included. It with files from the package sub-directories if the package has been
should be used after a patch is installed, since patches only update included. It should be used after a patch is installed, since patches
the files in the package sub-directory, but not the src files. Typing only update the files in the package sub-directory, but not the src
"make package-overwrite" will overwrite files in the package files. Typing "make package-overwrite" will overwrite files in the
sub-directories with src files. package sub-directories with src files.
Typing "make package-status" will show which packages are currently Typing "make package-status" or "make ps" will show which packages are
included. Of those that are included, it will list files that are currently included. Of those that are included, it will list files
different in the src directory and package sub-directory. Typing that are different in the src directory and package sub-directory.
"make package-diff" lists all differences between these files. Again, Typing "make package-diff" lists all differences between these files.
type "make package" to see all of the package-related make options. Again, type "make package" to see all of the package-related make
options.
:line :line
@ -699,16 +756,16 @@ you get a LAMMPS build error about a missing library, this is likely
the reason. See the "Section_packages"_Section_packages.html doc page the reason. See the "Section_packages"_Section_packages.html doc page
for a list of packages that have auxiliary libraries. for a list of packages that have auxiliary libraries.
Code for some of these auxiliary libraries is included in the LAMMPS Code for most of these auxiliary libraries is included in the LAMMPS
distribution under the lib directory. Examples are the USER-ATC and distribution under the lib directory. Examples are the USER-ATC and
MEAM packages. Some auxiliary libraries are NOT included with LAMMPS; MEAM packages. A few auxiliary libraries are NOT included with
to use the associated package you must download and install the LAMMPS; to use the associated package you must download and install
auxiliary library yourself. Examples are the KIM and VORONOI and the auxiliary library yourself. Examples are the KIM and VORONOI and
USER-MOLFILE packages. USER-MOLFILE packages.
For libraries with provided source code, each lib directory has a For provided libraries, each lib directory has a README file
README file (e.g. lib/reax/README) with instructions on how to build (e.g. lib/reax/README) with instructions on how to build that library.
that library. Typically this is done by typing something like: Typically this is done by typing something like:
make -f Makefile.g++ :pre make -f Makefile.g++ :pre
@ -740,168 +797,203 @@ is built with, typically requires additional Fortran-to-C libraries be
included in the link. Another example are the BLAS and LAPACK included in the link. Another example are the BLAS and LAPACK
libraries needed to use the USER-ATC or USER-AWPMD packages. libraries needed to use the USER-ATC or USER-AWPMD packages.
For libraries without provided source code, see the For libraries without provided source code, the file
src/package/Makefile.lammps file for information on where to find the src/package/README has information on where to find the library and
library and how to build it. E.g. the file src/KIM/Makefile.lammps or how to build it, e.g. src/VORONOI/README. There is also a
src/VORONOI/Makefile.lammps or src/UESR-MOLFILE/Makefile.lammps. Makefile.lammps file in the src/package directory. E.g. files
These files serve the same purpose as the lib/package/Makefile.lammps src/KIM/Makefile.lammps or src/VORONOI/Makefile.lammps or
files described above. The files have settings needed when LAMMPS is src/UESR-MOLFILE/Makefile.lammps. These files serve the same purpose
built to link with the corresponding auxiliary library. as the lib/package/Makefile.lammps files described above. The files
have settings needed when LAMMPS is built to link with the
corresponding auxiliary library.
Again, you must insure that the settings in Again, you must insure that the settings in
src/package/Makefile.lammps are appropriate for your system and where src/package/Makefile.lammps are appropriate for your system and where
you installed the auxiliary library. If they are not, the LAMMPS you installed the auxiliary library. If they are not, the LAMMPS
build will fail. build will typically fail.
:line :line
[{Packages that use make variable settings}] :link(start_3_4) [{Packages that require Makefile.machine settings}] :link(start_3_4)
One package, the KOKKOS package, allows its build options to be A few packages require specific settings in Makefile.machine, to
specified by setting variables via the "make" command, rather than by either build or use the package effectively. These are the
first building an auxiliary library and editing a Makefile.lammps USER-INTEL, KOKKOS, USER-OMP, and OPT packages. The details of what
file, as discussed in the previous sub-section for other packages. flags to add or what variables to define are given on the doc pages
This is for convenience since it is common to want to experiment with that describe each of these accelerator packages in detail:
different Kokkos library options. Using variables enables a direct
re-build of LAMMPS and its Kokkos dependencies, so that a benchmark
test with different Kokkos options can be quickly performed.
The syntax for setting make variables is as follows. You must "USER-INTEL package"_accelerate_intel.html
use a GNU-compatible make command for this to work. Try "gmake" "KOKKOS package"_accelerate_kokkos.html
if your system's standard make complains. "USER-OMP package"_accelerate_omp.html
"OPT package"_accelerate_opt.html :ul
make yes-kokkos Here is a brief summary of what Makefile.machine changes are needed.
make g++ VAR1=value VAR2=value ... :pre Note that the Make.py tool, described in the next "Section
2.4"_#start_4 can automatically add the needed info to an existing
machine Makefile, using simple command-line arguments.
The first line installs the KOKKOS package, which only needs to be In src/MAKE/OPTIONS see the following Makefiles for examples of the
done once. The second line builds LAMMPS with src/MAKE/Makefile.g++ changes described below:
and optionally sets one or more variables that affect the build. Each
variable is specified in upper-case; its value follows an equal sign
with no spaces. The second line can be repeated with different
variable settings, though a "clean" must be done before the rebuild.
Type "make clean" to see options for this operation.
These are the variables that can be specified. Each takes a value of Makefile.intel_cpu
{yes} or {no}. The default value is listed, which is set in the Makefile.intel_phi
lib/kokkos/Makefile.lammps file. See "this Makefile.kokkos_omp
section"_Section_accelerate.html#acc_8 for a discussion of what is Makefile.kokkos_cuda
meant by "host" and "device" in the Kokkos context. Makefile.kokkos_phi
Makefile.omp :ul
OMP, default = {yes} For the USER-INTEL package, you have 2 choices when building. You can
CUDA, default = {no} build with CPU or Phi support. The latter uses Xeon Phi chips in
HWLOC, default = {no} "offload" mode. Each of these modes requires additional settings in
AVX, default = {no} your Makefile.machine for CCFLAGS and LINKFLAGS.
MIC, default = {no}
LIBRT, default = {no}
DEBUG, default = {no} :ul
OMP sets the parallelization method used for Kokkos code (within For CPU mode (if using an Intel compiler):
LAMMPS) that runs on the host. OMP=yes means that OpenMP will be
used. OMP=no means that pthreads will be used.
CUDA sets the parallelization method used for Kokkos code (within CCFLAGS: add -fopenmp, -DLAMMPS_MEMALIGN=64, -restrict, -xHost, -fno-alias, -ansi-alias, -override-limits
LAMMPS) that runs on the device. CUDA=yes means an NVIDIA GPU running LINKFLAGS: add -fopenmp :ul
CUDA will be used. CUDA=no means that the OMP=yes or OMP=no setting
will be used for the device as well as the host.
If CUDA=yes, then the lo-level Makefile in the src/MAKE directory must For Phi mode add the following in addition to the CPU mode flags:
use "nvcc" as its compiler, via its CC setting. For best performance
its CCFLAGS setting should use -O3 and have an -arch setting that
matches the compute capability of your NVIDIA hardware and software
installation, e.g. -arch=sm_20. Generally Fermi Generation GPUs are
sm_20, while Kepler generation GPUs are sm_30 or sm_35 and Maxwell
cards are sm_50. A complete list can be found on
"wikipedia"_http://en.wikipedia.org/wiki/CUDA#Supported_GPUs. You can
also use the deviceQuery tool that comes with the CUDA samples. Note
the minimal required compute capability is 2.0, but this will give
signicantly reduced performance compared to Kepler generation GPUs
with compute capability 3.x. For the LINK setting, "nvcc" should not
be used; instead use g++ or another compiler suitable for linking C++
applications. Often you will want to use your MPI compiler wrapper
for this setting (i.e. mpicxx). Finally, the lo-level Makefile must
also have a "Compilation rule" for creating *.o files from *.cu files.
See src/Makefile.cuda for an example of a lo-level Makefile with all
of these settings.
HWLOC binds threads to hardware cores, so they do not migrate during a CCFLAGS: add -DLMP_INTEL_OFFLOAD and
simulation. HWLOC=yes should always be used if running with OMP=no LINKFLAGS: add -offload :ul
for pthreads. It is not necessary for OMP=yes for OpenMP, because
OpenMP provides alternative methods via environment variables for
binding threads to hardware cores. More info on binding threads to
cores is given in "this section"_Section_accelerate.html#acc_8.
AVX enables Intel advanced vector extensions when compiling for an And also add this to CCFLAGS:
Intel-compatible chip. AVX=yes should only be set if your host
hardware supports AVX. If it does not support it, this will cause a
run-time crash.
MIC enables compiler switches needed when compling for an Intel Phi -offload-option,mic,compiler,"-fp-model fast=2 -mGLOB_default_function_attrs=\"gather_scatter_loop_unroll=4\"" :pre
processor.
LIBRT enables use of a more accurate timer mechanism on most Unix For the KOKKOS package, you have 3 choices when building. You can
platforms. This library is not available on all platforms. build with OMP or Cuda or Phi support. Phi support uses Xeon Phi
chips in "native" mode. This can be done by setting the following
variables in your Makefile.machine:
DEBUG is only useful when developing a Kokkos-enabled style within for OMP support, set OMP = yes
LAMMPS. DEBUG=yes enables printing of run-time debugging information for Cuda support, set OMP = yes and CUDA = yes
that can be useful. It also enables runtime bounds checking on Kokkos for Phi support, set OMP = yes and MIC = yes :ul
data structures.
These can also be set as additional arguments to the make command, e.g.
make g++ OMP=yes MIC=yes :pre
Building the KOKKOS package with CUDA support requires a Makefile
machine that uses the NVIDIA "nvcc" compiler, as well as an
appropriate "arch" setting appropriate to the GPU hardware and NVIDIA
software you have on your machine. See
src/MAKE/OPTIONS/Makefile.kokkos_cuda for an example of such a machine
Makefile.
For the USER-OMP package, your Makefile.machine needs additional
settings for CCFLAGS and LINKFLAGS.
CCFLAGS: add -fopenmp and -restrict
LINKFLAGS: add -fopenmp :ul
For the OPT package, your Makefile.machine needs an additional
settings for CCFLAGS.
CCFLAGS: add -restrict :ul
:line :line
2.4 Building LAMMPS via the Make.py script :h4,link(start_4) 2.4 Building LAMMPS via the Make.py script :h4,link(start_4)
The src directory includes a Make.py script, written The src directory includes a Make.py script, written in Python, which
in Python, which can be used to automate various steps can be used to automate various steps of the build process. It is
of the build process. particularly useful for working with the accelerator packages, as well
as other packages which require auxiliary libraries to be built.
You can run the script from the src directory by typing either: You can run Make.py from the src directory by typing either:
Make.py Make.py -h
python Make.py :pre python Make.py -h :pre
which will give you info about the tool. For the former to work, you which will give you help info about the tool. For the former to work,
may need to edit the 1st line of the script to point to your local you may need to edit the first line of Make.py to point to your local
Python. And you may need to insure the script is executable: Python. And you may need to insure the script is executable:
chmod +x Make.py :pre chmod +x Make.py :pre
The following options are supported as switches: Here are examples of build tasks you can perform with Make.py:
-i file1 file2 ... Install/uninstall packages: Make.py -p no-lib kokkos omp intel
-p package1 package2 ... Build specific auxiliary libs: Make.py lib-atc lib-meam
-u package1 package2 ... Build libs for all installed packages: Make.py -p cuda gpu -gpu mode=double arch=31 lib-all
-e package1 arg1 arg2 package2 ... Create a Makefile from scratch with a compiler and MPI: Make.py -m none -cc g++ -mpi mpich file
-o dir Augment Makefile.serial with settings for installed packages: Make.py -p intel -intel cpu -m serial file
-b machine Add JPG and FFTW support to Makefile.mpi: Make.py -m mpi -jpg -fft fftw file
-s suffix1 suffix2 ... Build LAMMPS with a parallel make using Makefile.mpi: Make.py -j 16 -m mpi exe
-l dir Build LAMMPS and libs it needs using Makefile.serial with accelerator settings: Make.py -p gpu intel -intel cpu lib-all file serial :tb(s=:)
-j N
-h switch1 switch2 ... :ul
Help on any switch can be listed by using -h, e.g. The bench and examples directories give Make.py commands that can be
used to build LAMMPS with the various packages and options needed to
run all the benchmark and example input scripts. See these files for
more details:
Make.py -h -i -p :pre bench/README
bench/FERMI/README
bench/KEPLER/README
bench/PHI/README
examples/README
examples/accelerate/README
examples/accelerate/make.list :ul
At a hi-level, these are the kinds of package management All of the Make.py options and syntax help can be accessed by using
and build tasks that can be performed easily, using the "-h" switch.
the Make.py tool:
install/uninstall packages and build the associated external libs (use -p and -u and -e) E.g. typing "Make.py -h" gives
install packages needed for one or more input scripts (use -i and -p)
build LAMMPS, either in the src dir or new dir (use -b)
create a new dir with only the source code needed for one or more input scripts (use -i and -o) :ul
The last bullet can be useful when you wish to build a stripped-down Syntax: Make.py switch args ... {action1} {action2} ...
version of LAMMPS to run a specific script(s). Or when you wish to actions:
move the minimal amount of files to another platform for a remote lib-all, lib-dir, clean, file, exe or machine
LAMMPS build. zero or more actions, in any order (machine must be last)
switches:
-d (dir), -j (jmake), -m (makefile), -o (output),
-p (packages), -r (redo), -s (settings), -v (verbose)
switches for libs:
-atc, -awpmd, -colvars, -cuda
-gpu, -meam, -poems, -qmmm, -reax
switches for build and makefile options:
-intel, -kokkos, -cc, -mpi, -fft, -jpg, -png :pre
Note that using Make.py is not a substitute for insuring you have a Using the "-h" switch with other switches and actions gives additional
valid src/MAKE/Makefile.foo for your system, or that external library info on all the other specified switches or actions. The "-h" can be
Makefiles in any lib/* directories you use are also valid for your anywhere in the command-line and the other switches do not need their
system. But once you have done that, you can use Make.py to quickly arguments. E.g. type "Make.py -h -d -atc -intel" will print:
include/exclude the packages and external libraries needed by your
input scripts. -d dir
dir = LAMMPS home dir
if -d not specified, working dir must be lammps/src :pre
-atc make=suffix lammps=suffix2
all args are optional and can be in any order
make = use Makefile.suffix (def = g++)
lammps = use Makefile.lammps.suffix2 (def = EXTRAMAKE in makefile) :pre
-intel mode
mode = cpu or phi (def = cpu)
build Intel package for CPU or Xeon Phi :pre
Note that Make.py never overwrites an existing Makefile.machine.
Instead, it creates src/MAKE/MINE/Makefile.auto, which you can save or
rename if desired. Likewise it creates an executable named
src/lmp_auto, which you can rename using the -o switch if desired.
The most recently executed Make.py commmand is saved in
src/Make.py.last. You can use the "-r" switch (for redo) to re-invoke
the last command, or you can save a sequence of one or more Make.py
commands to a file and invoke the file of commands using "-r". You
can also label the commands in the file and invoke one or more of them
by name.
A typical use of Make.py is to start with a valid Makefile.machine for
your system, that works for a vanilla LAMMPS build, i.e. when optional
packages are not installed. You can then use Make.py to add various
settings (FFT, JPG, PNG) to the Makefile.machine as well as change its
compiler and MPI options. You can also add additional packages to the
build, as well as build the needed supporting libraries.
You can also use Make.py to create a new Makefile.machine from
scratch, using the "-m none" switch, if you also specify what compiler
and MPI options to use, via the "-cc" and "-mpi" switches.
:line :line

30
doc/accelerate_cuda.html
View File

@ -74,16 +74,30 @@ projects can be compiled without problems.
<P>This requires two steps (a,b): build the USER-CUDA library, then build <P>This requires two steps (a,b): build the USER-CUDA library, then build
LAMMPS with the USER-CUDA package. LAMMPS with the USER-CUDA package.
</P> </P>
<P>You can do both these steps in one line, using the src/Make.py script,
described in <A HREF = "Section_start.html#start_4">Section 2.4</A> of the manual.
Type "Make.py -h" for help. If run from the src directory, this
command will create src/lmp_cuda using src/MAKE/Makefile.mpi as the
starting Makefile.machine:
</P>
<PRE>Make.py -p cuda -cuda mode=single arch=20 -o cuda lib-cuda file mpi
</PRE>
<P>Or you can follow these two (a,b) steps:
</P>
<P>(a) Build the USER-CUDA library <P>(a) Build the USER-CUDA library
</P> </P>
<P>The USER-CUDA library is in lammps/lib/cuda. If your <I>CUDA</I> toolkit <P>The USER-CUDA library is in lammps/lib/cuda. If your <I>CUDA</I> toolkit
is not installed in the default system directoy <I>/usr/local/cuda</I> edit is not installed in the default system directoy <I>/usr/local/cuda</I> edit
the file <I>lib/cuda/Makefile.common</I> accordingly. the file <I>lib/cuda/Makefile.common</I> accordingly.
</P> </P>
<P>To set options for the library build, type "make OPTIONS", where <P>To build the library with the settings in lib/cuda/Makefile.default,
simply type:
</P>
<PRE>make
</PRE>
<P>To set options when the library is built, type "make OPTIONS", where
<I>OPTIONS</I> are one or more of the following. The settings will be <I>OPTIONS</I> are one or more of the following. The settings will be
written to the <I>lib/cuda/Makefile.defaults</I> and used when written to the <I>lib/cuda/Makefile.defaults</I> before the build.
the library is built.
</P> </P>
<PRE><I>precision=N</I> to set the precision level <PRE><I>precision=N</I> to set the precision level
N = 1 for single precision (default) N = 1 for single precision (default)
@ -107,11 +121,8 @@ the library is built.
0 = no CUFFT support (default) 0 = no CUFFT support (default)
in the future other CUDA-enabled FFT libraries might be supported in the future other CUDA-enabled FFT libraries might be supported
</PRE> </PRE>
<P>To build the library, simply type: <P>If the build is successful, it will produce the files liblammpscuda.a and
</P> Makefile.lammps.
<PRE>make
</PRE>
<P>If successful, it will produce the files libcuda.a and Makefile.lammps.
</P> </P>
<P>Note that if you change any of the options (like precision), you need <P>Note that if you change any of the options (like precision), you need
to re-build the entire library. Do a "make clean" first, followed by to re-build the entire library. Do a "make clean" first, followed by
@ -123,8 +134,7 @@ to re-build the entire library. Do a "make clean" first, followed by
make yes-user-cuda make yes-user-cuda
make machine make machine
</PRE> </PRE>
<P>No additional compile/link flags are needed in your Makefile.machine <P>No additional compile/link flags are needed in Makefile.machine.
in src/MAKE.
</P> </P>
<P>Note that if you change the USER-CUDA library precision (discussed <P>Note that if you change the USER-CUDA library precision (discussed
above) and rebuild the USER-CUDA library, then you also need to above) and rebuild the USER-CUDA library, then you also need to

30
doc/accelerate_cuda.txt
View File

@ -71,16 +71,30 @@ projects can be compiled without problems.
This requires two steps (a,b): build the USER-CUDA library, then build This requires two steps (a,b): build the USER-CUDA library, then build
LAMMPS with the USER-CUDA package. LAMMPS with the USER-CUDA package.
You can do both these steps in one line, using the src/Make.py script,
described in "Section 2.4"_Section_start.html#start_4 of the manual.
Type "Make.py -h" for help. If run from the src directory, this
command will create src/lmp_cuda using src/MAKE/Makefile.mpi as the
starting Makefile.machine:
Make.py -p cuda -cuda mode=single arch=20 -o cuda lib-cuda file mpi :pre
Or you can follow these two (a,b) steps:
(a) Build the USER-CUDA library (a) Build the USER-CUDA library
The USER-CUDA library is in lammps/lib/cuda. If your {CUDA} toolkit The USER-CUDA library is in lammps/lib/cuda. If your {CUDA} toolkit
is not installed in the default system directoy {/usr/local/cuda} edit is not installed in the default system directoy {/usr/local/cuda} edit
the file {lib/cuda/Makefile.common} accordingly. the file {lib/cuda/Makefile.common} accordingly.
To set options for the library build, type "make OPTIONS", where To build the library with the settings in lib/cuda/Makefile.default,
simply type:
make :pre
To set options when the library is built, type "make OPTIONS", where
{OPTIONS} are one or more of the following. The settings will be {OPTIONS} are one or more of the following. The settings will be
written to the {lib/cuda/Makefile.defaults} and used when written to the {lib/cuda/Makefile.defaults} before the build.
the library is built.
{precision=N} to set the precision level {precision=N} to set the precision level
N = 1 for single precision (default) N = 1 for single precision (default)
@ -104,11 +118,8 @@ the library is built.
0 = no CUFFT support (default) 0 = no CUFFT support (default)
in the future other CUDA-enabled FFT libraries might be supported :pre in the future other CUDA-enabled FFT libraries might be supported :pre
To build the library, simply type: If the build is successful, it will produce the files liblammpscuda.a and
Makefile.lammps.
make :pre
If successful, it will produce the files libcuda.a and Makefile.lammps.
Note that if you change any of the options (like precision), you need Note that if you change any of the options (like precision), you need
to re-build the entire library. Do a "make clean" first, followed by to re-build the entire library. Do a "make clean" first, followed by
@ -120,8 +131,7 @@ cd lammps/src
make yes-user-cuda make yes-user-cuda
make machine :pre make machine :pre
No additional compile/link flags are needed in your Makefile.machine No additional compile/link flags are needed in Makefile.machine.
in src/MAKE.
Note that if you change the USER-CUDA library precision (discussed Note that if you change the USER-CUDA library precision (discussed
above) and rebuild the USER-CUDA library, then you also need to above) and rebuild the USER-CUDA library, then you also need to

13
doc/accelerate_gpu.html
View File

@ -76,6 +76,16 @@ install the NVIDIA Cuda software on your system:
<P>This requires two steps (a,b): build the GPU library, then build <P>This requires two steps (a,b): build the GPU library, then build
LAMMPS with the GPU package. LAMMPS with the GPU package.
</P> </P>
<P>You can do both these steps in one line, using the src/Make.py script,
described in <A HREF = "Section_start.html#start_4">Section 2.4</A> of the manual.
Type "Make.py -h" for help. If run from the src directory, this
command will create src/lmp_gpu using src/MAKE/Makefile.mpi as the
starting Makefile.machine:
</P>
<PRE>Make.py -p gpu -gpu mode=single arch=31 -o gpu lib-gpu file mpi
</PRE>
<P>Or you can follow these two (a,b) steps:
</P>
<P>(a) Build the GPU library <P>(a) Build the GPU library
</P> </P>
<P>The GPU library is in lammps/lib/gpu. Select a Makefile.machine (in <P>The GPU library is in lammps/lib/gpu. Select a Makefile.machine (in
@ -120,8 +130,7 @@ Makefile.linux clean", followed by the make command above.
make yes-gpu make yes-gpu
make machine make machine
</PRE> </PRE>
<P>No additional compile/link flags are needed in your Makefile.machine <P>No additional compile/link flags are needed in Makefile.machine.
in src/MAKE.
</P> </P>
<P>Note that if you change the GPU library precision (discussed above) <P>Note that if you change the GPU library precision (discussed above)
and rebuild the GPU library, then you also need to re-install the GPU and rebuild the GPU library, then you also need to re-install the GPU

13
doc/accelerate_gpu.txt
View File

@ -73,6 +73,16 @@ Run lammps/lib/gpu/nvc_get_devices (after building the GPU library, see below) t
This requires two steps (a,b): build the GPU library, then build This requires two steps (a,b): build the GPU library, then build
LAMMPS with the GPU package. LAMMPS with the GPU package.
You can do both these steps in one line, using the src/Make.py script,
described in "Section 2.4"_Section_start.html#start_4 of the manual.
Type "Make.py -h" for help. If run from the src directory, this
command will create src/lmp_gpu using src/MAKE/Makefile.mpi as the
starting Makefile.machine:
Make.py -p gpu -gpu mode=single arch=31 -o gpu lib-gpu file mpi :pre
Or you can follow these two (a,b) steps:
(a) Build the GPU library (a) Build the GPU library
The GPU library is in lammps/lib/gpu. Select a Makefile.machine (in The GPU library is in lammps/lib/gpu. Select a Makefile.machine (in
@ -117,8 +127,7 @@ cd lammps/src
make yes-gpu make yes-gpu
make machine :pre make machine :pre
No additional compile/link flags are needed in your Makefile.machine No additional compile/link flags are needed in Makefile.machine.
in src/MAKE.
Note that if you change the GPU library precision (discussed above) Note that if you change the GPU library precision (discussed above)
and rebuild the GPU library, then you also need to re-install the GPU and rebuild the GPU library, then you also need to re-install the GPU

49
doc/accelerate_intel.html
View File

@ -41,6 +41,10 @@ suffix to "omp" so that styles from the USER-OMP package will be used
if available, after first testing if a style from the USER-INTEL if available, after first testing if a style from the USER-INTEL
package is available. package is available.
</P> </P>
<P>When using the USER-INTEL package, you must choose at build time
whether you are building for CPU-only acceleration or for using the
Xeon Phi in offload mode.
</P>
<P>Here is a quick overview of how to use the USER-INTEL package <P>Here is a quick overview of how to use the USER-INTEL package
for CPU-only acceleration: for CPU-only acceleration:
</P> </P>
@ -50,6 +54,9 @@ for CPU-only acceleration:
<LI>specify how many OpenMP threads per MPI task to use <LI>specify how many OpenMP threads per MPI task to use
<LI>use USER-INTEL and (optionally) USER-OMP styles in your input script <LI>use USER-INTEL and (optionally) USER-OMP styles in your input script
</UL> </UL>
<P>Note that many of these settings can only be used with the Intel
compiler, as discussed below.
</P>
<P>Using the USER-INTEL package to offload work to the Intel(R) <P>Using the USER-INTEL package to offload work to the Intel(R)
Xeon Phi(TM) coprocessor is the same except for these additional Xeon Phi(TM) coprocessor is the same except for these additional
steps: steps:
@ -74,25 +81,41 @@ Phi(TM) coprocessors.
Intel(R) compiler. Use of other compilers may not result in Intel(R) compiler. Use of other compilers may not result in
vectorization or give poor performance. vectorization or give poor performance.
</P> </P>
<P>Use of an Intel C++ compiler is reccommended, but not required. The <P>Use of an Intel C++ compiler is recommended, but not required (though
compiler must support the OpenMP interface. g++ will not recognize some of the settings, so they cannot be used).
The compiler must support the OpenMP interface.
</P> </P>
<P><B>Building LAMMPS with the USER-INTEL package:</B> <P><B>Building LAMMPS with the USER-INTEL package:</B>
</P> </P>
<P>Include the package(s) and build LAMMPS: <P>You must choose at build time whether to build for CPU acceleration or
to use the Xeon Phi in offload mode.
</P>
<P>You can do either in one line, using the src/Make.py script, described
in <A HREF = "Section_start.html#start_4">Section 2.4</A> of the manual. Type
"Make.py -h" for help. If run from the src directory, these commands
will create src/lmp_intel_cpu and lmp_intel_phi using
src/MAKE/Makefile.mpi as the starting Makefile.machine:
</P>
<PRE>Make.py -p intel omp -intel cpu -o intel_cpu -cc icc file mpi
Make.py -p intel omp -intel phi -o intel_phi -cc icc file mpi
</PRE>
<P>Note that this assumes that your MPI and its mpicxx wrapper
is using the Intel compiler. If it is not, you should
leave off the "-cc icc" switch.
</P>
<P>Or you can follow these steps:
</P> </P>
<PRE>cd lammps/src <PRE>cd lammps/src
make yes-user-intel make yes-user-intel
make yes-user-omp (if desired) make yes-user-omp (if desired)
make machine make machine
</PRE> </PRE>
<P>If the USER-OMP package is also installed, you can use styles from <P>Note that if the USER-OMP package is also installed, you can use
both packages, as described below. styles from both packages, as described below.
</P> </P>
<P>The lo-level src/MAKE/Makefile.machine needs a flag for OpenMP support <P>The Makefile.machine needs a "-fopenmp" flag for OpenMP support in
in both the CCFLAGS and LINKFLAGS variables, which is <I>-openmp</I> for both the CCFLAGS and LINKFLAGS variables. You also need to add
Intel compilers. You also need to add -DLAMMPS_MEMALIGN=64 and -DLAMMPS_MEMALIGN=64 and -restrict to CCFLAGS.
-restrict to CCFLAGS.
</P> </P>
<P>If you are compiling on the same architecture that will be used for <P>If you are compiling on the same architecture that will be used for
the runs, adding the flag <I>-xHost</I> to CCFLAGS will enable the runs, adding the flag <I>-xHost</I> to CCFLAGS will enable
@ -102,10 +125,10 @@ vectorization with the Intel(R) compiler.
coprocessor, the flag <I>-offload</I> should be added to the LINKFLAGS line coprocessor, the flag <I>-offload</I> should be added to the LINKFLAGS line
and the flag -DLMP_INTEL_OFFLOAD should be added to the CCFLAGS line. and the flag -DLMP_INTEL_OFFLOAD should be added to the CCFLAGS line.
</P> </P>
<P>Note that the machine makefiles Makefile.intel and <P>Example makefiles Makefile.intel_cpu and Makefile.intel_phi are
Makefile.intel_offload are included in the src/MAKE directory with included in the src/MAKE/OPTIONS directory with settings that perform
options that perform well with the Intel(R) compiler. The latter file well with the Intel(R) compiler. The latter file has support for
has support for offload to coprocessors; the former does not. offload to coprocessors; the former does not.
</P> </P>
<P>If using an Intel compiler, it is recommended that Intel(R) Compiler <P>If using an Intel compiler, it is recommended that Intel(R) Compiler
2013 SP1 update 1 be used. Newer versions have some performance 2013 SP1 update 1 be used. Newer versions have some performance

49
doc/accelerate_intel.txt
View File

@ -38,6 +38,10 @@ suffix to "omp" so that styles from the USER-OMP package will be used
if available, after first testing if a style from the USER-INTEL if available, after first testing if a style from the USER-INTEL
package is available. package is available.
When using the USER-INTEL package, you must choose at build time
whether you are building for CPU-only acceleration or for using the
Xeon Phi in offload mode.
Here is a quick overview of how to use the USER-INTEL package Here is a quick overview of how to use the USER-INTEL package
for CPU-only acceleration: for CPU-only acceleration:
@ -47,6 +51,9 @@ include the USER-INTEL package and (optionally) USER-OMP package and build LAMMP
specify how many OpenMP threads per MPI task to use specify how many OpenMP threads per MPI task to use
use USER-INTEL and (optionally) USER-OMP styles in your input script :ul use USER-INTEL and (optionally) USER-OMP styles in your input script :ul
Note that many of these settings can only be used with the Intel
compiler, as discussed below.
Using the USER-INTEL package to offload work to the Intel(R) Using the USER-INTEL package to offload work to the Intel(R)
Xeon Phi(TM) coprocessor is the same except for these additional Xeon Phi(TM) coprocessor is the same except for these additional
steps: steps:
@ -71,25 +78,41 @@ Optimizations for vectorization have only been tested with the
Intel(R) compiler. Use of other compilers may not result in Intel(R) compiler. Use of other compilers may not result in
vectorization or give poor performance. vectorization or give poor performance.
Use of an Intel C++ compiler is reccommended, but not required. The Use of an Intel C++ compiler is recommended, but not required (though
compiler must support the OpenMP interface. g++ will not recognize some of the settings, so they cannot be used).
The compiler must support the OpenMP interface.
[Building LAMMPS with the USER-INTEL package:] [Building LAMMPS with the USER-INTEL package:]
Include the package(s) and build LAMMPS: You must choose at build time whether to build for CPU acceleration or
to use the Xeon Phi in offload mode.
You can do either in one line, using the src/Make.py script, described
in "Section 2.4"_Section_start.html#start_4 of the manual. Type
"Make.py -h" for help. If run from the src directory, these commands
will create src/lmp_intel_cpu and lmp_intel_phi using
src/MAKE/Makefile.mpi as the starting Makefile.machine:
Make.py -p intel omp -intel cpu -o intel_cpu -cc icc file mpi
Make.py -p intel omp -intel phi -o intel_phi -cc icc file mpi :pre
Note that this assumes that your MPI and its mpicxx wrapper
is using the Intel compiler. If it is not, you should
leave off the "-cc icc" switch.
Or you can follow these steps:
cd lammps/src cd lammps/src
make yes-user-intel make yes-user-intel
make yes-user-omp (if desired) make yes-user-omp (if desired)
make machine :pre make machine :pre
If the USER-OMP package is also installed, you can use styles from Note that if the USER-OMP package is also installed, you can use
both packages, as described below. styles from both packages, as described below.
The lo-level src/MAKE/Makefile.machine needs a flag for OpenMP support The Makefile.machine needs a "-fopenmp" flag for OpenMP support in
in both the CCFLAGS and LINKFLAGS variables, which is {-openmp} for both the CCFLAGS and LINKFLAGS variables. You also need to add
Intel compilers. You also need to add -DLAMMPS_MEMALIGN=64 and -DLAMMPS_MEMALIGN=64 and -restrict to CCFLAGS.
-restrict to CCFLAGS.
If you are compiling on the same architecture that will be used for If you are compiling on the same architecture that will be used for
the runs, adding the flag {-xHost} to CCFLAGS will enable the runs, adding the flag {-xHost} to CCFLAGS will enable
@ -99,10 +122,10 @@ In order to build with support for an Intel(R) Xeon Phi(TM)
coprocessor, the flag {-offload} should be added to the LINKFLAGS line coprocessor, the flag {-offload} should be added to the LINKFLAGS line
and the flag -DLMP_INTEL_OFFLOAD should be added to the CCFLAGS line. and the flag -DLMP_INTEL_OFFLOAD should be added to the CCFLAGS line.
Note that the machine makefiles Makefile.intel and Example makefiles Makefile.intel_cpu and Makefile.intel_phi are
Makefile.intel_offload are included in the src/MAKE directory with included in the src/MAKE/OPTIONS directory with settings that perform
options that perform well with the Intel(R) compiler. The latter file well with the Intel(R) compiler. The latter file has support for
has support for offload to coprocessors; the former does not. offload to coprocessors; the former does not.
If using an Intel compiler, it is recommended that Intel(R) Compiler If using an Intel compiler, it is recommended that Intel(R) Compiler
2013 SP1 update 1 be used. Newer versions have some performance 2013 SP1 update 1 be used. Newer versions have some performance

109
doc/accelerate_kokkos.html
View File

@ -61,14 +61,17 @@ one or the other of the two modes. The first mode is called the
processor (running in native mode, not offload mode like the processor (running in native mode, not offload mode like the
USER-INTEL package) are supported. The second mode is called the USER-INTEL package) are supported. The second mode is called the
"device" and is an accelerator chip of some kind. Currently only an "device" and is an accelerator chip of some kind. Currently only an
NVIDIA GPU is supported. If your compute node does not have a GPU, NVIDIA GPU is supported via Cuda. If your compute node does not have
then there is only one mode of execution, i.e. the host and device are a GPU, then there is only one mode of execution, i.e. the host and
the same. device are the same.
</P> </P>
<P>Here is a quick overview of how to use the KOKKOS package <P>When using the KOKKOS package, you must choose at build time whether
for GPU acceleration: you are building for OpenMP, GPU, or for using the Xeon Phi in native
mode.
</P> </P>
<UL><LI>specify variables and settings in your Makefile.machine that enable GPU, Phi, or OpenMP support <P>Here is a quick overview of how to use the KOKKOS package:
</P>
<UL><LI>specify variables and settings in your Makefile.machine that enable OpenMP, GPU, or Phi support
<LI>include the KOKKOS package and build LAMMPS <LI>include the KOKKOS package and build LAMMPS
<LI>enable the KOKKOS package and its hardware options via the "-k on" command-line switch <LI>enable the KOKKOS package and its hardware options via the "-k on" command-line switch
<LI>use KOKKOS styles in your input script <LI>use KOKKOS styles in your input script
@ -105,14 +108,23 @@ and GPU packages for details of how to check and do this.
</P> </P>
<P><B>Building LAMMPS with the KOKKOS package:</B> <P><B>Building LAMMPS with the KOKKOS package:</B>
</P> </P>
<P>Unlike other acceleration packages discussed in this section, the <P>You must choose at build time whether to build for OpenMP, Cuda, or
Kokkos library in lib/kokkos does not have to be pre-built before Phi.
building LAMMPS itself. Instead, options for the Kokkos library are
specified at compile time, when LAMMPS itself is built. This can be
done in one of two ways, as discussed below.
</P> </P>
<P>Here are examples of how to build LAMMPS for the different compute-node <P>You can do any of these in one line, using the src/Make.py script,
configurations listed above. described in <A HREF = "Section_start.html#start_4">Section 2.4</A> of the manual.
Type "Make.py -h" for help. If run from the src directory, these
commands will create src/lmp_kokkos_omp, lmp_kokkos_cuda, and
lmp_kokkos_phi. The OMP and PHI options use src/MAKE/Makefile.mpi as
the starting Makefile.machine. The CUDA option uses
src/MAKE/OPTIONS/Makefile.cuda since the NVIDIA nvcc compiler is
required.
</P>
<P>Make.py -p kokkos -kokkos omp -o kokkos_omp file mpi
Make.py -p kokkos -kokkos cuda arch=31 -o kokkos_cuda file kokkos_cuda
Make.py -p kokkos -kokkos phi -o kokkos_phi file mpi
</P>
<P>Or you can follow these steps:
</P> </P>
<P>CPU-only (run all-MPI or with OpenMP threading): <P>CPU-only (run all-MPI or with OpenMP threading):
</P> </P>
@ -164,15 +176,76 @@ in <A HREF = "Section_start.html#start_3_4">Section 2.3.4</A> of the manual, as
as other settings that must be included in the machine makefile, if as other settings that must be included in the machine makefile, if
you create your own. you create your own.
</P> </P>
<P>There are other allowed options when building with the KOKKOS package.
As above, They can be set either as variables on the make command line
or in the machine makefile in the src/MAKE directory. See <A HREF = "Section_start.html#start_3_4">Section
2.3.4</A> of the manual for details.
</P>
<P>IMPORTANT NOTE: Currently, there are no precision options with the <P>IMPORTANT NOTE: Currently, there are no precision options with the
KOKKOS package. All compilation and computation is performed in KOKKOS package. All compilation and computation is performed in
double precision. double precision.
</P> </P>
<P>There are other allowed options when building with the KOKKOS package.
As above, they can be set either as variables on the make command line
or in Makefile.machine. This is the full list of options, including
those discussed above, Each takes a value of <I>yes</I> or <I>no</I>. The
default value is listed, which is set in the
lib/kokkos/Makefile.lammps file.
</P>
<UL><LI>OMP, default = <I>yes</I>
<LI>CUDA, default = <I>no</I>
<LI>HWLOC, default = <I>no</I>
<LI>AVX, default = <I>no</I>
<LI>MIC, default = <I>no</I>
<LI>LIBRT, default = <I>no</I>
<LI>DEBUG, default = <I>no</I>
</UL>
<P>OMP sets the parallelization method used for Kokkos code (within
LAMMPS) that runs on the host. OMP=yes means that OpenMP will be
used. OMP=no means that pthreads will be used.
</P>
<P>CUDA sets the parallelization method used for Kokkos code (within
LAMMPS) that runs on the device. CUDA=yes means an NVIDIA GPU running
CUDA will be used. CUDA=no means that the OMP=yes or OMP=no setting
will be used for the device as well as the host.
</P>
<P>If CUDA=yes, then the lo-level Makefile in the src/MAKE directory must
use "nvcc" as its compiler, via its CC setting. For best performance
its CCFLAGS setting should use -O3 and have an -arch setting that
matches the compute capability of your NVIDIA hardware and software
installation, e.g. -arch=sm_20. Generally Fermi Generation GPUs are
sm_20, while Kepler generation GPUs are sm_30 or sm_35 and Maxwell
cards are sm_50. A complete list can be found on
<A HREF = "http://en.wikipedia.org/wiki/CUDA#Supported_GPUs">wikipedia</A>. You can
also use the deviceQuery tool that comes with the CUDA samples. Note
the minimal required compute capability is 2.0, but this will give
signicantly reduced performance compared to Kepler generation GPUs
with compute capability 3.x. For the LINK setting, "nvcc" should not
be used; instead use g++ or another compiler suitable for linking C++
applications. Often you will want to use your MPI compiler wrapper
for this setting (i.e. mpicxx). Finally, the lo-level Makefile must
also have a "Compilation rule" for creating *.o files from *.cu files.
See src/Makefile.cuda for an example of a lo-level Makefile with all
of these settings.
</P>
<P>HWLOC binds threads to hardware cores, so they do not migrate during a
simulation. HWLOC=yes should always be used if running with OMP=no
for pthreads. It is not necessary for OMP=yes for OpenMP, because
OpenMP provides alternative methods via environment variables for
binding threads to hardware cores. More info on binding threads to
cores is given in <A HREF = "Section_accelerate.html#acc_8">this section</A>.
</P>
<P>AVX enables Intel advanced vector extensions when compiling for an
Intel-compatible chip. AVX=yes should only be set if your host
hardware supports AVX. If it does not support it, this will cause a
run-time crash.
</P>
<P>MIC enables compiler switches needed when compling for an Intel Phi
processor.
</P>
<P>LIBRT enables use of a more accurate timer mechanism on most Unix
platforms. This library is not available on all platforms.
</P>
<P>DEBUG is only useful when developing a Kokkos-enabled style within
LAMMPS. DEBUG=yes enables printing of run-time debugging information
that can be useful. It also enables runtime bounds checking on Kokkos
data structures.
</P>
<P><B>Run with the KOKKOS package from the command line:</B> <P><B>Run with the KOKKOS package from the command line:</B>
</P> </P>
<P>The mpirun or mpiexec command sets the total number of MPI tasks used <P>The mpirun or mpiexec command sets the total number of MPI tasks used

109
doc/accelerate_kokkos.txt
View File

@ -58,14 +58,17 @@ one or the other of the two modes. The first mode is called the
processor (running in native mode, not offload mode like the processor (running in native mode, not offload mode like the
USER-INTEL package) are supported. The second mode is called the USER-INTEL package) are supported. The second mode is called the
"device" and is an accelerator chip of some kind. Currently only an "device" and is an accelerator chip of some kind. Currently only an
NVIDIA GPU is supported. If your compute node does not have a GPU, NVIDIA GPU is supported via Cuda. If your compute node does not have
then there is only one mode of execution, i.e. the host and device are a GPU, then there is only one mode of execution, i.e. the host and
the same. device are the same.
Here is a quick overview of how to use the KOKKOS package When using the KOKKOS package, you must choose at build time whether
for GPU acceleration: you are building for OpenMP, GPU, or for using the Xeon Phi in native
mode.
specify variables and settings in your Makefile.machine that enable GPU, Phi, or OpenMP support Here is a quick overview of how to use the KOKKOS package:
specify variables and settings in your Makefile.machine that enable OpenMP, GPU, or Phi support
include the KOKKOS package and build LAMMPS include the KOKKOS package and build LAMMPS
enable the KOKKOS package and its hardware options via the "-k on" command-line switch enable the KOKKOS package and its hardware options via the "-k on" command-line switch
use KOKKOS styles in your input script :ul use KOKKOS styles in your input script :ul
@ -102,14 +105,23 @@ and GPU packages for details of how to check and do this.
[Building LAMMPS with the KOKKOS package:] [Building LAMMPS with the KOKKOS package:]
Unlike other acceleration packages discussed in this section, the You must choose at build time whether to build for OpenMP, Cuda, or
Kokkos library in lib/kokkos does not have to be pre-built before Phi.
building LAMMPS itself. Instead, options for the Kokkos library are
specified at compile time, when LAMMPS itself is built. This can be
done in one of two ways, as discussed below.
Here are examples of how to build LAMMPS for the different compute-node You can do any of these in one line, using the src/Make.py script,
configurations listed above. described in "Section 2.4"_Section_start.html#start_4 of the manual.
Type "Make.py -h" for help. If run from the src directory, these
commands will create src/lmp_kokkos_omp, lmp_kokkos_cuda, and
lmp_kokkos_phi. The OMP and PHI options use src/MAKE/Makefile.mpi as
the starting Makefile.machine. The CUDA option uses
src/MAKE/OPTIONS/Makefile.cuda since the NVIDIA nvcc compiler is
required.
Make.py -p kokkos -kokkos omp -o kokkos_omp file mpi
Make.py -p kokkos -kokkos cuda arch=31 -o kokkos_cuda file kokkos_cuda
Make.py -p kokkos -kokkos phi -o kokkos_phi file mpi
Or you can follow these steps:
CPU-only (run all-MPI or with OpenMP threading): CPU-only (run all-MPI or with OpenMP threading):
@ -161,15 +173,76 @@ in "Section 2.3.4"_Section_start.html#start_3_4 of the manual, as well
as other settings that must be included in the machine makefile, if as other settings that must be included in the machine makefile, if
you create your own. you create your own.
There are other allowed options when building with the KOKKOS package.
As above, They can be set either as variables on the make command line
or in the machine makefile in the src/MAKE directory. See "Section
2.3.4"_Section_start.html#start_3_4 of the manual for details.
IMPORTANT NOTE: Currently, there are no precision options with the IMPORTANT NOTE: Currently, there are no precision options with the
KOKKOS package. All compilation and computation is performed in KOKKOS package. All compilation and computation is performed in
double precision. double precision.
There are other allowed options when building with the KOKKOS package.
As above, they can be set either as variables on the make command line
or in Makefile.machine. This is the full list of options, including
those discussed above, Each takes a value of {yes} or {no}. The
default value is listed, which is set in the
lib/kokkos/Makefile.lammps file.
OMP, default = {yes}
CUDA, default = {no}
HWLOC, default = {no}
AVX, default = {no}
MIC, default = {no}
LIBRT, default = {no}
DEBUG, default = {no} :ul
OMP sets the parallelization method used for Kokkos code (within
LAMMPS) that runs on the host. OMP=yes means that OpenMP will be
used. OMP=no means that pthreads will be used.
CUDA sets the parallelization method used for Kokkos code (within
LAMMPS) that runs on the device. CUDA=yes means an NVIDIA GPU running
CUDA will be used. CUDA=no means that the OMP=yes or OMP=no setting
will be used for the device as well as the host.
If CUDA=yes, then the lo-level Makefile in the src/MAKE directory must
use "nvcc" as its compiler, via its CC setting. For best performance
its CCFLAGS setting should use -O3 and have an -arch setting that
matches the compute capability of your NVIDIA hardware and software
installation, e.g. -arch=sm_20. Generally Fermi Generation GPUs are
sm_20, while Kepler generation GPUs are sm_30 or sm_35 and Maxwell
cards are sm_50. A complete list can be found on
"wikipedia"_http://en.wikipedia.org/wiki/CUDA#Supported_GPUs. You can
also use the deviceQuery tool that comes with the CUDA samples. Note
the minimal required compute capability is 2.0, but this will give
signicantly reduced performance compared to Kepler generation GPUs
with compute capability 3.x. For the LINK setting, "nvcc" should not
be used; instead use g++ or another compiler suitable for linking C++
applications. Often you will want to use your MPI compiler wrapper
for this setting (i.e. mpicxx). Finally, the lo-level Makefile must
also have a "Compilation rule" for creating *.o files from *.cu files.
See src/Makefile.cuda for an example of a lo-level Makefile with all
of these settings.
HWLOC binds threads to hardware cores, so they do not migrate during a
simulation. HWLOC=yes should always be used if running with OMP=no
for pthreads. It is not necessary for OMP=yes for OpenMP, because
OpenMP provides alternative methods via environment variables for
binding threads to hardware cores. More info on binding threads to
cores is given in "this section"_Section_accelerate.html#acc_8.
AVX enables Intel advanced vector extensions when compiling for an
Intel-compatible chip. AVX=yes should only be set if your host
hardware supports AVX. If it does not support it, this will cause a
run-time crash.
MIC enables compiler switches needed when compling for an Intel Phi
processor.
LIBRT enables use of a more accurate timer mechanism on most Unix
platforms. This library is not available on all platforms.
DEBUG is only useful when developing a Kokkos-enabled style within
LAMMPS. DEBUG=yes enables printing of run-time debugging information
that can be useful. It also enables runtime bounds checking on Kokkos
data structures.
[Run with the KOKKOS package from the command line:] [Run with the KOKKOS package from the command line:]
The mpirun or mpiexec command sets the total number of MPI tasks used The mpirun or mpiexec command sets the total number of MPI tasks used

22
doc/accelerate_omp.html
View File

@ -42,17 +42,27 @@ MPI task running on a CPU.
</P> </P>
<P><B>Building LAMMPS with the USER-OMP package:</B> <P><B>Building LAMMPS with the USER-OMP package:</B>
</P> </P>
<P>Include the package and build LAMMPS: <P>To do this in one line, use the src/Make.py script, described in
<A HREF = "Section_start.html#start_4">Section 2.4</A> of the manual. Type "Make.py
-h" for help. If run from the src directory, this command will create
src/lmp_omp using src/MAKE/Makefile.mpi as the starting
Makefile.machine:
</P>
<PRE>Make.py -p omp -o omp file mpi
</PRE>
<P>Or you can follow these steps:
</P> </P>
<PRE>cd lammps/src <PRE>cd lammps/src
make yes-user-omp make yes-user-omp
make machine make machine
</PRE> </PRE>
<P>The CCFLAGS setting in your src/MAKE/Makefile.machine needs "-fopenmp" <P>The CCFLAGS setting in Makefile.machine needs "-fopenmp" to add OpenMP
to add OpenMP support. This works for both the GNU and Intel support. This works for both the GNU and Intel compilers. Without
compilers. Without this flag the USER-OMP styles will still be this flag the USER-OMP styles will still be compiled and work, but
compiled and work, but will not support multi-threading. For the will not support multi-threading. For the Intel compilers the CCFLAGS
Intel compilers the CCFLAGS setting also needs to include "-restrict". setting also needs to include "-restrict".
</P>
<P><B>Run with the USER-OMP package from the command line:</B>
</P> </P>
<P>The mpirun or mpiexec command sets the total number of MPI tasks used <P>The mpirun or mpiexec command sets the total number of MPI tasks used
by LAMMPS (one or multiple per compute node) and the number of MPI by LAMMPS (one or multiple per compute node) and the number of MPI

22
doc/accelerate_omp.txt
View File

@ -39,17 +39,27 @@ MPI task running on a CPU.
[Building LAMMPS with the USER-OMP package:] [Building LAMMPS with the USER-OMP package:]
Include the package and build LAMMPS: To do this in one line, use the src/Make.py script, described in
"Section 2.4"_Section_start.html#start_4 of the manual. Type "Make.py
-h" for help. If run from the src directory, this command will create
src/lmp_omp using src/MAKE/Makefile.mpi as the starting
Makefile.machine:
Make.py -p omp -o omp file mpi :pre
Or you can follow these steps:
cd lammps/src cd lammps/src
make yes-user-omp make yes-user-omp
make machine :pre make machine :pre
The CCFLAGS setting in your src/MAKE/Makefile.machine needs "-fopenmp" The CCFLAGS setting in Makefile.machine needs "-fopenmp" to add OpenMP
to add OpenMP support. This works for both the GNU and Intel support. This works for both the GNU and Intel compilers. Without
compilers. Without this flag the USER-OMP styles will still be this flag the USER-OMP styles will still be compiled and work, but
compiled and work, but will not support multi-threading. For the will not support multi-threading. For the Intel compilers the CCFLAGS
Intel compilers the CCFLAGS setting also needs to include "-restrict". setting also needs to include "-restrict".
[Run with the USER-OMP package from the command line:]
The mpirun or mpiexec command sets the total number of MPI tasks used The mpirun or mpiexec command sets the total number of MPI tasks used
by LAMMPS (one or multiple per compute node) and the number of MPI by LAMMPS (one or multiple per compute node) and the number of MPI

14
doc/accelerate_opt.html
View File

@ -38,12 +38,22 @@ input script.
</P> </P>
<P>Include the package and build LAMMPS: <P>Include the package and build LAMMPS:
</P> </P>
<P>To do this in one line, use the src/Make.py script, described in
<A HREF = "Section_start.html#start_4">Section 2.4</A> of the manual. Type "Make.py
-h" for help. If run from the src directory, this command will create
src/lmp_opt using src/MAKE/Makefile.mpi as the starting
Makefile.machine:
</P>
<PRE>Make.py -p opt -o opt file mpi
</PRE>
<P>Or you can follow these steps:
</P>
<PRE>cd lammps/src <PRE>cd lammps/src
make yes-opt make yes-opt
make machine make machine
</PRE> </PRE>
<P>If you are using Intel compilers, then the CCFLAGS setting in your <P>If you are using Intel compilers, then the CCFLAGS setting in
src/MAKE/Makefile.machine needs to include "-restrict". Makefile.machine needs to include "-restrict".
</P> </P>
<P><B>Run with the OPT package from the command line:</B> <P><B>Run with the OPT package from the command line:</B>
</P> </P>

14
doc/accelerate_opt.txt
View File

@ -35,12 +35,22 @@ None.
Include the package and build LAMMPS: Include the package and build LAMMPS:
To do this in one line, use the src/Make.py script, described in
"Section 2.4"_Section_start.html#start_4 of the manual. Type "Make.py
-h" for help. If run from the src directory, this command will create
src/lmp_opt using src/MAKE/Makefile.mpi as the starting
Makefile.machine:
Make.py -p opt -o opt file mpi :pre
Or you can follow these steps:
cd lammps/src cd lammps/src
make yes-opt make yes-opt
make machine :pre make machine :pre
If you are using Intel compilers, then the CCFLAGS setting in your If you are using Intel compilers, then the CCFLAGS setting in
src/MAKE/Makefile.machine needs to include "-restrict". Makefile.machine needs to include "-restrict".
[Run with the OPT package from the command line:] [Run with the OPT package from the command line:]