compiling WRF-Chem 4.6

[cindy-d]

I've been testing WRF-Chem 4.6 on Linux x86_64 with dmpar INTEL (ifort/icc), and I've encountered a segmentation fault at the start of my simulation. To prevent this, I've had to add the following flags:

-O0 -fno-inline -no-ip

While this resolves the issue, I'm concerned that these flags could significantly increase the model's runtime. Is there a more efficient fix or workaround available? When the flags were removed, I received the error message below:

calculate MEGAN emissions at ktau, gmtp, tmidh = 1 0.0000000E+00 1.6666668E-02
photolysis_driver: called for domain 1
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
wrf.exe 00000000050AE8A1 Unknown Unknown Unknown
wrf.exe 00000000050AC9DB Unknown Unknown Unknown
wrf.exe 00000000050385C4 Unknown Unknown Unknown
wrf.exe 00000000050383D6 Unknown Unknown Unknown
wrf.exe 0000000004FB9DC9 Unknown Unknown Unknown
wrf.exe 0000000004FC1636 Unknown Unknown Unknown
libpthread-2.12.s 000000375CA0F710 Unknown Unknown Unknown
wrf.exe 000000000306A02B module_phot_fastj 2352 module_phot_fastj.f90
wrf.exe 000000000305ECDE module_phot_fastj 1817 module_phot_fastj.f90
wrf.exe 000000000305BBE3 module_phot_fastj 349 module_phot_fastj.f90
wrf.exe 00000000021FADE3 photolysis_driver 181 photolysis_driver.f90
wrf.exe 00000000021E2ADA chem_driver_ 1142 chem_driver.f90
wrf.exe 0000000002008DA9 solve_interface_ 189 solve_interface.f90
wrf.exe 00000000005C9EED module_integrate_ 338 module_integrate.f90
wrf.exe 0000000000412884 module_wrf_top_mp 326 module_wrf_top.f90
wrf.exe 000000000041283F MAIN__ 44 wrf.f90
wrf.exe 00000000004127DE Unknown Unknown Unknown
libc-2.12.so 000000375C61ED5D __libc_start_main Unknown Unknown
wrf.exe 00000000004126E9 Unknown Unknown Unknown