[cig-commits] r19513 - in seismo/3D/SPECFEM3D/trunk: . doc doc/USER_MANUAL
dkomati1 at geodynamics.org
dkomati1 at geodynamics.org
Mon Jan 30 11:18:43 PST 2012
Author: dkomati1
Date: 2012-01-30 11:18:43 -0800 (Mon, 30 Jan 2012)
New Revision: 19513
Modified:
seismo/3D/SPECFEM3D/trunk/doc/USER_MANUAL/manual_SPECFEM3D.pdf
seismo/3D/SPECFEM3D/trunk/doc/USER_MANUAL/manual_SPECFEM3D.tex
seismo/3D/SPECFEM3D/trunk/doc/time_scheme_we_use_in_the_code_from_Hughes_1987.pdf
seismo/3D/SPECFEM3D/trunk/flags.guess
Log:
updated the manual to describe the options one may need to change in flags.guess
Modified: seismo/3D/SPECFEM3D/trunk/doc/USER_MANUAL/manual_SPECFEM3D.pdf
===================================================================
(Binary files differ)
Modified: seismo/3D/SPECFEM3D/trunk/doc/USER_MANUAL/manual_SPECFEM3D.tex
===================================================================
--- seismo/3D/SPECFEM3D/trunk/doc/USER_MANUAL/manual_SPECFEM3D.tex 2012-01-30 16:51:02 UTC (rev 19512)
+++ seismo/3D/SPECFEM3D/trunk/doc/USER_MANUAL/manual_SPECFEM3D.tex 2012-01-30 19:18:43 UTC (rev 19513)
@@ -295,7 +295,33 @@
./configure~FC=ifort~MPIFC=mpif90
\end{lyxcode}
+Before running the \texttt{configure} script, you should probably edit file \texttt{flags.guess} to make sure that it contains the best compiler options for your system. Known problems or things to check are:
+\begin{description}
+\item [{\texttt{Intel ifort compiler}}] See if you need to add \texttt{-assume byterecl} for your machine.
+\item [{\texttt{IBM or Cray compilers}}] See if you need to add \texttt{-qsave} or \texttt{-qnosave} for your machine.
+\item [{\texttt{Mac OS}}] You will probably need to install \texttt{XCODE}.
+In addition, the \texttt{clock\_gettime} routine, which is used by the \texttt{SCOTCH} library that we use, does not exist in Mac OS.
+You will need to replace it with \texttt{clock\_get\_time} if you want to use \texttt{SCOTCH}.
+\end{description}
+
+When compiling on an IBM machine with the \texttt{xlf} and \texttt{xlc} compilers, we suggest running the \texttt{configure} script
+with the following options:\\
+
+\noindent\texttt{./configure FC=xlf90\_r MPIFC=mpif90 CC=xlc\_r CFLAGS="-O3 -q64" FCFLAGS="-O3 -q64" -with-scotch-dir=...}.\\
+
+On SGI systems, \texttt{flags.guess} automatically informs \texttt{configure}
+to insert `\texttt{`TRAP\_FPE=OFF}'' into the generated \texttt{Makefile}
+in order to turn underflow trapping off.\\
+
+If you run very large meshes on a relatively small number
+of processors, the memory size needed on each processor might become
+greater than 2 gigabytes, which is the upper limit for 32-bit addressing.
+In this case, on some compilers you may need to add \texttt{``-mcmodel=medium}''
+to the compiler options otherwise the compiler will display an error
+message (for instance \texttt{``relocation truncated to fit: R\_X86\_64\_PC32 against .bss''} or something similar);
+on an IBM machine with the \texttt{xlf} and \texttt{xlc} compilers, using \texttt{-q64} is usually sufficient.
+
The SPECFEM3D software package relies on the SCOTCH library to partition meshes created with CUBIT.
Note that we use CUBIT to create meshes of hexahedra, but other packages can be used as well, for instance GiD from \url{http://gid.cimne.upc.es} or Gmsh from \url{http://geuz.org/gmsh} \citep{GeRe09}. Even mesh creation packages that generate tetrahedra, for instance TetGen from \url{http://tetgen.berlios.de}, can be used because each tetrahedron can then easily be decomposed into four hexahedra as shown in the picture of the TetGen logo at \url{http://tetgen.berlios.de/figs/Delaunay-Voronoi-3D.gif}; while this approach does not generate hexahedra of optimal quality, it can ease mesh creation in some situations and it has been shown that the spectral-element method can very accurately handle distorted mesh elements \citep{OlSe11}.
@@ -382,23 +408,6 @@
and in double precision on your system and compare the seismograms.
If they are identical (and in most cases they will), you can select single precision for your future runs.\\
-When compiling on an IBM machine with the \texttt{xlf} and \texttt{xlc} compilers, we suggest running the \texttt{configure} script
-with the following options:\\
-
-\noindent\texttt{./configure FC=xlf90\_r MPIFC=mpif90 CC=xlc\_r CFLAGS="-O3 -q64" FCFLAGS="-O3 -q64" -with-scotch-dir=...}.\\
-
-On SGI systems, \texttt{flags.guess} automatically informs \texttt{configure}
-to insert `\texttt{`TRAP\_FPE=OFF}'' into the generated \texttt{Makefile}
-in order to turn underflow trapping off.\\
-
-If you run very large meshes on a relatively small number
-of processors, the memory size needed on each processor might become
-greater than 2 gigabytes, which is the upper limit for 32-bit addressing.
-In this case, on some compilers you may need to add \texttt{``-mcmodel=medium}''
-to the compiler options otherwise the compiler will display an error
-message (for instance \texttt{``relocation truncated to fit: R\_X86\_64\_PC32 against .bss''} or something similar);
-on an IBM machine with the \texttt{xlf} and \texttt{xlc} compilers, using \texttt{-q64} is usually sufficient.
-
\chapter{\label{cha:Mesh-Generation}Mesh Generation}
The first step in running a spectral-element simulation consists of constructing a high-quality mesh for the region under consideration. We provide two possibilities to do so: (1) relying on the external, hexahedral mesher CUBIT, or (2) using the provided, internal mesher \texttt{xmeshfem3D}. In the following, we explain these two approaches.
Modified: seismo/3D/SPECFEM3D/trunk/doc/time_scheme_we_use_in_the_code_from_Hughes_1987.pdf
===================================================================
(Binary files differ)
Modified: seismo/3D/SPECFEM3D/trunk/flags.guess
===================================================================
--- seismo/3D/SPECFEM3D/trunk/flags.guess 2012-01-30 16:51:02 UTC (rev 19512)
+++ seismo/3D/SPECFEM3D/trunk/flags.guess 2012-01-30 19:18:43 UTC (rev 19513)
@@ -26,21 +26,21 @@
# Intel ifort Fortran90
#
if test x"$FLAGS_CHECK" = x; then
- FLAGS_CHECK="-O3 -assume byterecl -traceback -ftrapuv -ftz"
+ FLAGS_CHECK="-O3 -traceback -ftrapuv -ftz"
# useful for debugging...
# for debugging: change -check nobounds to -check all -debug -g -O0 -fp-stack-check -traceback -ftrapuv
# ifort version 11 or 10.1 with these flags shows good performance
# -xSSE4.2 -ftz -funroll-loops -unroll5
- #FLAGS_CHECK="-O3 -ftz -xS -traceback -ftrapuv -vec-report0 -std95 -implicitnone -check nobounds -assume byterecl -static-intel -warn truncated_source -warn argument_checking -warn unused -warn declarations -warn alignments -warn ignore_loc -warn usage "
- #FLAGS_CHECK="-O3 -xS -static-intel -r8 -mcmodel=large -implicitnone -warn truncated_source -warn argument_checking -warn unused -warn declarations -check nobounds -align sequence -assume byterecl -ftrapuv -ftz -traceback"
+ #FLAGS_CHECK="-O3 -ftz -xS -traceback -ftrapuv -vec-report0 -std95 -implicitnone -check nobounds -static-intel -warn truncated_source -warn argument_checking -warn unused -warn declarations -warn alignments -warn ignore_loc -warn usage "
+ #FLAGS_CHECK="-O3 -xS -static-intel -r8 -mcmodel=large -implicitnone -warn truncated_source -warn argument_checking -warn unused -warn declarations -check nobounds -align sequence -ftrapuv -ftz -traceback"
fi
if test x"$FLAGS_NO_CHECK" = x; then
# standard options (leave option -ftz, which is *critical* for performance)
# add -Winline to get information about routines that are inlined
# add -vec-report2 to get information about loops that are vectorized or not
- FLAGS_NO_CHECK="-O3 -assume byterecl -check nobounds -ftz"
+ FLAGS_NO_CHECK="-O3 -check nobounds -ftz"
fi
;;
More information about the CIG-COMMITS
mailing list