Use ip@5: if available instead of sp

[climbfuji]

The sp library is deprecated and replaced by ip@5 and newer (drop-in replacement). While [email protected] is still in spack-stack-1.8.0, ip-5.0.0 is also available. We expect sp to be removed in spack-stack-1.9.0.

This PR goes with NCAR/ccpp-physics#1090

[dustinswales]

All tests pass.
Looks good to me.

[climbfuji]

ping

[grantfirl]

@scrasmussen I saw #551. I'm going to run some tests on Hera to verify that this works with spack-stack 1.8.0 before I approve. If it works, we can merge it, although I don't think that we want the SCM to get ahead of where the UFS is in terms of which version spack-stack is being used by default: https://github.com/ufs-community/ufs-weather-model/blob/develop/modulefiles/ufs_hera.intel.lua

[scrasmussen]

@grantfirl If we don't want the SCM to get ahead I can just answer gabersek with instructions to checkout my updated fature/ip5 branch and use that as a workaround until the default spack-stack gets updated and we merge this? (Hopefully I'm understanding the situation correctly)

[grantfirl]

@grantfirl If we don't want the SCM to get ahead I can just answer gabersek with instructions to checkout my updated fature/ip5 branch and use that as a workaround until the default spack-stack gets updated and we merge this? (Hopefully I'm understanding the situation correctly)

I think that it should be OK to merge before UFS updates their spack-stack version because Dom changed the CMakeLists to look for IP first and then SP if not found. This should work fine with spack-stack 1.6 until it is updated to coincide with UFS. gabersek was already doing something "nonstandard" by trying the SCM with spack-stack 1.8, but it wouldn't work because CMakeLists.txt didn't even know to look for IP yet. It's weird, though, because sp supposedly still exists in 1.8, so I'm not 100% sure why the issue cropped up yet.