132 lines
5.7 KiB
CMake
132 lines
5.7 KiB
CMake
###################################################################################################
|
|
# GNU Compiler
|
|
###################################################################################################
|
|
|
|
if (OPENMP)
|
|
set (OPENMP_FLAGS "-fopenmp")
|
|
endif ()
|
|
|
|
if (OPTIMIZATION STREQUAL "OFF")
|
|
set (OPTIMIZATION_FLAGS "-O0" )
|
|
elseif (OPTIMIZATION STREQUAL "DEFENSIVE")
|
|
set (OPTIMIZATION_FLAGS "-O2")
|
|
elseif (OPTIMIZATION STREQUAL "AGGRESSIVE")
|
|
set (OPTIMIZATION_FLAGS "-O3 -ffast-math -funroll-loops -ftree-vectorize")
|
|
endif ()
|
|
|
|
set (STANDARD_CHECK "-std=f2008ts -pedantic-errors" )
|
|
set (LINKER_FLAGS "${LINKER_FLAGS} -Wl")
|
|
# options parsed directly to the linker
|
|
set (LINKER_FLAGS "${LINKER_FLAGS},-undefined,dynamic_lookup" )
|
|
# ensure to link against dynamic libraries
|
|
|
|
#------------------------------------------------------------------------------------------------
|
|
# Fine tuning compilation options
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -xf95-cpp-input")
|
|
# preprocessor
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -ffree-line-length-132")
|
|
# restrict line length to the standard 132 characters (lattice.f90 require more characters)
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -fimplicit-none")
|
|
# assume "implicit none" even if not present in source
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wall")
|
|
# sets the following Fortran options:
|
|
# -Waliasing: warn about possible aliasing of dummy arguments. Specifically, it warns if the same actual argument is associated with a dummy argument with "INTENT(IN)" and a dummy argument with "INTENT(OUT)" in a call with an explicit interface.
|
|
# -Wampersand: checks if a character expression is continued proberly by an ampersand at the end of the line and at the beginning of the new line
|
|
# -Warray-bounds: checks if array reference is out of bounds at compile time. use -fcheck-bounds to also check during runtime
|
|
# -Wconversion: warn about implicit conversions between different type
|
|
# -Wsurprising: warn when "suspicious" code constructs are encountered. While technically legal these usually indicate that an error has been made.
|
|
# -Wc-binding-type:
|
|
# -Wintrinsics-std: only standard intrisics are available, e.g. "call flush(6)" will cause an error
|
|
# -Wno-tabs: do not allow tabs in source
|
|
# -Wintrinsic-shadow: warn if a user-defined procedure or module procedure has the same name as an intrinsic
|
|
# -Wline-truncation:
|
|
# -Wtarget-lifetime:
|
|
# -Wreal-q-constant: warn about real-literal-constants with 'q' exponent-letter
|
|
# -Wunused: a number of unused-xxx warnings
|
|
# and sets the general (non-Fortran options) options:
|
|
# -Waddress
|
|
# -Warray-bounds (only with -O2)
|
|
# -Wc++11-compat
|
|
# -Wchar-subscripts
|
|
# -Wcomment
|
|
# -Wformat
|
|
# -Wmaybe-uninitialized
|
|
# -Wnonnull
|
|
# -Wparentheses
|
|
# -Wpointer-sign
|
|
# -Wreorder
|
|
# -Wreturn-type
|
|
# -Wsequence-point
|
|
# -Wstrict-aliasing
|
|
# -Wstrict-overflow=1
|
|
# -Wswitch
|
|
# -Wtrigraphs
|
|
# -Wuninitialized
|
|
# -Wunknown-pragmas
|
|
# -Wunused-function
|
|
# -Wunused-label
|
|
# -Wunused-value
|
|
# -Wunused-variable
|
|
# -Wvolatile-register-var
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wextra")
|
|
# sets the following Fortran options:
|
|
# -Wunuses-parameter:
|
|
# -Wcompare-reals:
|
|
# and sets the general (non-Fortran options) options:
|
|
# -Wclobbered
|
|
# -Wempty-body
|
|
# -Wignored-qualifiers
|
|
# -Wmissing-field-initializers
|
|
# -Woverride-init
|
|
# -Wsign-compare
|
|
# -Wtype-limits
|
|
# -Wuninitialized
|
|
# -Wunused-but-set-parameter (only with -Wunused or -Wall)
|
|
# -Wno-globals
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wcharacter-truncation")
|
|
# warn if character expressions (strings) are truncated
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wunderflow")
|
|
# produce a warning when numerical constant expressions are encountered, which yield an UNDERFLOW during compilation
|
|
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wsuggest-attribute=pure")
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wsuggest-attribute=noreturn")
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wconversion-extra")
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wimplicit-procedure")
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -Wno-unused-parameter")
|
|
set (COMPILE_FLAGS "${COMPILE_FLAGS} -ffpe-summary=all")
|
|
# print summary of floating point exeptions (invalid,zero,overflow,underflow,inexact,denormal)
|
|
|
|
# Additional options
|
|
# -Warray-temporarieswarnings: because we have many temporary arrays (performance issue?)
|
|
# -Wimplicit-interface: no interfaces for lapack/MPI routines
|
|
# -Wunsafe-loop-optimizations: warn if the loop cannot be optimized due to nontrivial assumptions
|
|
|
|
#------------------------------------------------------------------------------------------------
|
|
# Runtime debugging
|
|
set (DEBUG_FLAGS "${DEBUG_FLAGS} -ffpe-trap=invalid,zero,overflow")
|
|
# stop execution if floating point exception is detected (NaN is silent)
|
|
|
|
set (DEBUG_FLAGS "${DEBUG_FLAGS} -g")
|
|
# Generate symbolic debugging information in the object file
|
|
|
|
set (DEBUG_FLAGS "${DEBUG_FLAGS} -fbacktrace")
|
|
set (DEBUG_FLAGS "${DEBUG_FLAGS} -fdump-core")
|
|
set (DEBUG_FLAGS "${DEBUG_FLAGS} -fcheck=all")
|
|
# checks for (array-temps,bounds,do,mem,pointer,recursion)
|
|
|
|
# Additional options
|
|
# -ffpe-trap=precision,denormal,underflow
|
|
|
|
#------------------------------------------------------------------------------------------------
|
|
# precision settings
|
|
set (PRECISION_FLAGS "${PRECISION_FLAGS} -fdefault-real-8")
|
|
# set precision to 8 bytes for standard real (=8 for pReal). Will set size of double to 16 bytes as long as -fdefault-double-8 is not set
|
|
set (PRECISION_FLAGS "${PRECISION_FLAGS} -fdefault-double-8")
|
|
# set precision to 8 bytes for double real, would be 16 bytes if -fdefault-real-8 is used
|