Fix Debug compilation (#800)

Update (and simplify) the Intel "Debug" compile flags to match those required by NCO.

Minor updates to the regional_esg_grid code so it can be compiled in
"Debug" mode on WCOSS2.

Fixes #735.

CMakeLists.txt

@@ -64,7 +64,7 @@ endif()
 if(CMAKE_Fortran_COMPILER_ID MATCHES "^(Intel)$")
   set(CMAKE_Fortran_FLAGS "${CMAKE_Fortran_FLAGS} -g -traceback")
   set(CMAKE_Fortran_FLAGS_RELEASE "-O3 -fp-model precise")
-  set(CMAKE_Fortran_FLAGS_DEBUG "-O0 -check -check noarg_temp_created -check nopointer -fp-stack-check -fstack-protector-all -fpe0 -debug -ftrapuv")
+  set(CMAKE_Fortran_FLAGS_DEBUG "-O0 -check all -ftrapuv")
   if(APPLE)
     # The linker on macOS does not include `common symbols` by default.
     # Passing the -c flag includes them and fixes an error with undefined symbols.
@@ -77,9 +77,7 @@ elseif(CMAKE_Fortran_COMPILER_ID MATCHES "^(GNU)$")
     set(CMAKE_Fortran_FLAGS "${CMAKE_Fortran_FLAGS} -fallow-argument-mismatch -fallow-invalid-boz")
   endif()
   set(CMAKE_Fortran_FLAGS_RELEASE "-O3")
-#  set(CMAKE_Fortran_FLAGS_DEBUG "-ggdb -Wall")
-  set(CMAKE_Fortran_FLAGS_DEBUG "-O1 -ggdb -Wall -fno-unsafe-math-optimizations -frounding-math -fsignaling-nans -ffpe-trap=invalid,zero,overflow -fbounds-check -fno-omit-frame-pointer -fno-optimize-sibling-calls")
-#  set(CMAKE_Fortran_FLAGS_DEBUG "-ggdb -Wall -fno-unsafe-math-optimizations -frounding-math -fsignaling-nans -ffpe-trap=invalid,zero,overflow -fbounds-check -fno-omit-frame-pointer -fno-optimize-sibling-calls")
+  set(CMAKE_Fortran_FLAGS_DEBUG "-O0 -ggdb -Wall -fcheck=all")
 endif()
 
 if(CMAKE_C_COMPILER_ID MATCHES "^(Intel)$")

sorc/grid_tools.fd/regional_esg_grid.fd/pesg.f90

@@ -406,13 +406,14 @@ subroutine xmtoxc_vak1(ak,xm,xc,xcd,xc1,xcd1,ff)!                  [xmtoxc_ak]
 logical,                  intent(out):: ff
 real(dp),dimension(3,2,2):: xcdd
 real(dp),dimension(2,2,2):: xsd1,xsdd
-real(dp),dimension(2,2)  :: xtd,xsd,xs1,xtd1,xsdk
+real(dp),dimension(2,2)  :: xtd,xsd,xs1,xtd1,xsdk,mat22
 real(dp),dimension(2)    :: xt,xt1,xs,xsk
 integer(spi)             :: i
 call xmtoxt1(ak(1),xm,xt,xtd,xt1,xtd1,ff);      if(ff)return
 call xttoxs1(ak(2),xt,xs,xsd,xsdd,xsk,xsdk,ff); if(ff)return
 xs1(:,2)=xsk; xs1(:,1)=matmul(xsd,xt1)
-xsd1(:,:,1)=matmul(xsd,xtd1)+matmul(xsdd(:,:,1)*xt1(1)+xsdd(:,:,2)*xt1(2),xtd)
+mat22=xsdd(:,:,1)*xt1(1)+xsdd(:,:,2)*xt1(2)
+xsd1(:,:,1)=matmul(xsd,xtd1)+matmul(mat22,xtd)
 xsd1(:,:,2)=matmul(xsdk,xtd)
 xsd=matmul(xsd,xtd)
 call xstoxc(xs,xc,xcd,xcdd)

sorc/grid_tools.fd/regional_esg_grid.fd/psym2.f90

@@ -411,7 +411,7 @@ subroutine logsym2d(x,z,zd)!                                         [logsym2]
 real(dp),dimension(2,2),    intent(in ):: x
 real(dp),dimension(2,2),    intent(out):: z
 real(dp),dimension(2,2,2,2),intent(out):: zd
-real(dp),dimension(2,2):: vv,oo,d11,d12,d22,pqr
+real(dp),dimension(2,2):: vv,oo,d11,d12,d22,pqr,d11pqr,d12pqr,d22pqr
 real(dp)               :: c,s,cc,cs,ss,c2h,p,q,r,lp,lq,L
 integer(spi)           :: i
 call eigensym2(x,vv,oo)
@@ -424,9 +424,12 @@ subroutine logsym2d(x,z,zd)!                                         [logsym2]
 d12(1,:)=(/-cs,c2h/); d12(2,:)=(/c2h,cs/)
 d22(1,:)=(/ ss,-cs/); d22(2,:)=(/-cs,cc/)
 pqr(1,:)=(/p,r/)    ; pqr(2,:)=(/r,q/)
-zd(:,:,1,1)=matmul(vv,matmul(d11*pqr,transpose(vv)))
-zd(:,:,1,2)=matmul(vv,matmul(d12*pqr,transpose(vv)))
-zd(:,:,2,2)=matmul(vv,matmul(d22*pqr,transpose(vv)))
+d11pqr=d11*pqr
+d12pqr=d12*pqr
+d22pqr=d22*pqr
+zd(:,:,1,1)=matmul(vv,matmul(d11pqr,transpose(vv)))
+zd(:,:,1,2)=matmul(vv,matmul(d12pqr,transpose(vv)))
+zd(:,:,2,2)=matmul(vv,matmul(d22pqr,transpose(vv)))
 zd(:,:,2,1)=zd(:,:,1,2)
 end subroutine logsym2d