source: LMDZ6/trunk/libf/phylmd/ecrad-acc/Makefile_include.intel @ 6016

Last change on this file since 6016 was 6016, checked in by yann meurdesoif, 3 months ago

Add new ecrad version from DWD ported onto OpenACC, closed from original ecrad ECMWF starting point for LMDZ ecrad version.

Modification from ecrad-lmdz version has been included.

YM

  • Property svn:eol-style set to native
File size: 911 bytes
Line 
1# This is a Makefile for the Intel compiler: run "make
2# PROFILE=intel". Note that by default the Intel compiler puts arrays
3# on the stack, which is likely to cause a stack overflow - hence the
4# use of "-heap-arrays" which puts them on the heap like most other
5# compilers.  If you remove this argument you will likely need to set
6# the environment variable OMP_STACKSIZE to 64MB.
7
8FC = ifort
9CC = icc
10CPPFLAGS =
11
12ifndef DEBUG
13OPTFLAGS = -O3
14else
15OPTFLAGS = -O0 -check bounds -init=snan
16endif
17
18LDFLAGS = -lrt
19WARNFLAGS = -warn all
20
21# We used to have the flag "-convert big_endian" here because the RRTM
22# input files are big endian Fortran unformatted files, but now the
23# file ordering has been specified at the OPEN command so no compiler
24# flags are needed.
25BASICFLAGS = -module ../mod -fpe0 -fp-model precise -ftz -fp-speculation safe -heap-arrays
26
27DEBUGFLAGS = -g
28OMPFLAG = -qopenmp -qopenmp-lib=compat
Note: See TracBrowser for help on using the repository browser.