Build an OpenMP version of jt9 where possible

This change  introduces the program jt9_omp  which is a testbed  for a
multi-threaded version of the jt9 decoder program. The program jt9_omp
should be  a directly substitutable for  jt9 except that JT65  and JT9
decodes are computed in parallel.

Also enable  the OpenMP directives in  decoder.f90 - note this  is not
yet a working multi-threaded decoder and the existing jt9 is still the
correct decoder to be used in WSJT-X.

Increased the available  stack size for jt9_omp.exe as this  is a hard
limit on  Windows and  the default  is not big  enough for  the OpenMP
version of jt9.

Also  Fortran  array  bounds  checking is  now  disabled  for  Release
configuration builds so as to improve performance a little.

git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@4922 ab8295b8-cf94-4d9e-aec4-7959e3be5d79

CMakeLists.txt

@@ -241,6 +241,7 @@ set (wsjt_FSRCS
   lib/dcoord.f90
   lib/decode65a.f90
   lib/decode65b.f90
+  lib/fftw3mod.f90
   lib/jt9fano.f90
   lib/decoder.f90
   lib/decjt9.f90
@@ -256,7 +257,6 @@ set (wsjt_FSRCS
   lib/fano232.f90
   lib/fchisq.f90
   lib/fchisq65.f90
-  lib/fftw3mod.f90
   lib/fil3.f90
   lib/fil4.f90
   lib/fil6521.f90
@@ -492,7 +492,7 @@ endif (APPLE)
 #
 # Fortran setup
 #
-set (General_FFLAGS "-fbounds-check -Wall -Wno-conversion -fno-second-underscore")
+set (General_FFLAGS "-Wall -Wno-conversion -fno-second-underscore")
 
 # FFLAGS depend on the compiler
 get_filename_component (Fortran_COMPILER_NAME ${CMAKE_Fortran_COMPILER} NAME)
@@ -509,7 +509,7 @@ if (Fortran_COMPILER_NAME MATCHES "gfortran.*")
   endif (CMAKE_OSX_SYSROOT)
 
   set (CMAKE_Fortran_FLAGS_RELEASE "${CMAKE_Fortran_FLAGS_RELEASE} -funroll-all-loops -fno-f2c ${General_FFLAGS}")
-  set (CMAKE_Fortran_FLAGS_DEBUG   "${CMAKE_Fortran_FLAGS_DEBUG} -fno-f2c ${General_FFLAGS}")
+  set (CMAKE_Fortran_FLAGS_DEBUG   "${CMAKE_Fortran_FLAGS_DEBUG} -fbounds-check -fno-f2c ${General_FFLAGS}")
 elseif (Fortran_COMPILER_NAME MATCHES "ifort.*")
   # ifort (untested)
   set (CMAKE_Fortran_FLAGS_RELEASE "${CMAKE_Fortran_FLAGS_RELEASE} -f77rtl ${General_FFLAGS}")
@@ -517,13 +517,13 @@ elseif (Fortran_COMPILER_NAME MATCHES "ifort.*")
 elseif (Fortran_COMPILER_NAME MATCHES "g77")
   # g77
   set (CMAKE_Fortran_FLAGS_RELEASE "${CMAKE_Fortran_FLAGS_RELEASE} -funroll-all-loops -fno-f2c -m32 ${General_FFLAGS}")
-  set (CMAKE_Fortran_FLAGS_DEBUG   "${CMAKE_Fortran_FLAGS_DEBUG} -fno-f2c -m32 ${General_FFLAGS}")
+  set (CMAKE_Fortran_FLAGS_DEBUG   "${CMAKE_Fortran_FLAGS_DEBUG} -fbounds-check -fno-f2c -m32 ${General_FFLAGS}")
 else (Fortran_COMPILER_NAME MATCHES "gfortran.*")
   message ("CMAKE_Fortran_COMPILER full path: " ${CMAKE_Fortran_COMPILER})
   message ("Fortran compiler: " ${Fortran_COMPILER_NAME})
   message ("No optimized Fortran compiler flags are known, we just try -O2...")
   set (CMAKE_Fortran_FLAGS_RELEASE "${CMAKE_Fortran_FLAGS_RELEASE} -O2 ${General_FFLAGS}")
-  set (CMAKE_Fortran_FLAGS_DEBUG   "${CMAKE_Fortran_FLAGS_DEBUG} ${General_FFLAGS}")
+  set (CMAKE_Fortran_FLAGS_DEBUG   "${CMAKE_Fortran_FLAGS_DEBUG} -fbounds-check ${General_FFLAGS}")
 endif (Fortran_COMPILER_NAME MATCHES "gfortran.*")
 
 
@@ -593,6 +593,12 @@ if (NOT "${QT_LIBRARY_DIR}" STREQUAL "/lib" AND NOT "${QT_LIBRARY_DIR}" STREQUAL
 endif ()
 
 
+#
+# OpenMP
+#
+find_package (OpenMP)
+
+
 #
 # fftw3 single precsion library
 #
@@ -698,8 +704,17 @@ endif (WIN32)
 # targets
 #
 
-# build a library of package functionality
+# build a library of package functionality (without and optionally with OpenMP support)
 add_library (wsjt STATIC ${wsjt_CSRCS} ${wsjt_CXXSRCS} ${wsjt_FSRCS})
+if (${OPENMP_FOUND})
+  add_library (wsjt_omp STATIC ${wsjt_CSRCS} ${wsjt_CXXSRCS} ${wsjt_FSRCS})
+  file (MAKE_DIRECTORY ${CMAKE_BINARY_DIR}/fortran_modules_omp)
+  set_target_properties (wsjt_omp
+    PROPERTIES
+    COMPILE_FLAGS ${OpenMP_C_FLAGS}
+    Fortran_MODULE_DIRECTORY ${CMAKE_BINARY_DIR}/fortran_modules_omp
+    )
+endif (${OPENMP_FOUND})
 
 # build a library of package Qt functionality
 add_library (wsjt_qt STATIC ${wsjt_qt_CXXSRCS} ${wsjt_qt_GENUISRCS} ${GENAXSRCS})
@@ -721,6 +736,18 @@ add_executable (jt9 lib/jt9.f90 lib/jt9a.f90 lib/jt9b.f90 lib/jt9c.f90 ${jt9_CXX
 target_link_libraries (jt9 wsjt ${FFTW3_LIBRARIES})
 qt5_use_modules (jt9 Core)
 
+if (${OPENMP_FOUND})
+  add_executable (jt9_omp lib/jt9.f90 lib/jt9a.f90 lib/jt9b.f90 lib/jt9c.f90 ${jt9_CXXSRCS} wsjtx.rc)
+  set_target_properties (jt9_omp
+    PROPERTIES
+    COMPILE_FLAGS ${OpenMP_C_FLAGS}
+    Fortran_MODULE_DIRECTORY ${CMAKE_BINARY_DIR}/fortran_modules_omp
+    LINK_FLAGS "${OpenMP_C_FLAGS} -Wl,--stack,8388608"
+    )
+  target_link_libraries (jt9_omp wsjt_omp ${FFTW3_LIBRARIES})
+  qt5_use_modules (jt9_omp Core)
+endif (${OPENMP_FOUND})
+
 # build configuration dialog and transceiver interface test application
 #add_executable (ConfigTest ${ConfigTest_CXXSRCS} ${ConfigTest_GENUISRCS} wsjtx.rc)
 #target_link_libraries (ConfigTest wsjt wsjt_qt ${hamlib_LIBRARIES})
@@ -790,7 +817,7 @@ install (TARGETS wsjtx
   BUNDLE DESTINATION . COMPONENT runtime
   )
 
-install (TARGETS jt9 jt65code jt9code
+install (TARGETS jt9 jt65code jt9code jt9_omp
   RUNTIME DESTINATION ${WSJT_BIN_DESTINATION} COMPONENT runtime
   BUNDLE DESTINATION ${WSJT_BIN_DESTINATION} COMPONENT runtime
   )

lib/decoder.f90

@@ -50,17 +50,16 @@ subroutine decoder(ss,id2)
   if(nmode.eq.65) go to 800
 
 !  print*,'A'
-!!$OMP PARALLEL PRIVATE(id)
-!!$OMP SECTIONS
+!$OMP PARALLEL SECTIONS PRIVATE(id)
 
-!!$OMP SECTION
+!$OMP SECTION
 !  print*,'B'
   call timer('decjt9  ',0)
   call decjt9(ss,id2,nutc,nfqso,newdat,npts8,nfa,nfsplit,nfb,ntol,nzhsym,  &
        nagain,ndepth,nmode)
   call timer('decjt9  ',1)
 
-!!$OMP SECTION
+!$OMP SECTION
   if(nmode.ge.65 .and. (.not.done65)) then
      if(newdat.ne.0) dd(1:npts65)=id2(1:npts65)
      nf1=nfa
@@ -71,8 +70,7 @@ subroutine decoder(ss,id2)
      call timer('jt65a   ',1)
   endif
 
-!!$OMP END SECTIONS NOWAIT
-!!$OMP END PARALLEL 
+!$OMP END PARALLEL SECTIONS
 !  print*,'D'
 
 ! JT65 is not yet producing info for nsynced, ndecoded.