Add new stanza for ARM aarch64/openmpi

[yip111]

TYPE：enhancement

KEYWORDS: ARM

DESCRIPTION OF CHANGES:
Problem:
There are no ARM-only stanzas in arch/configure.defaults. As the suggestion from #1535 , we need a separate PR for ARM specific update.

Solution:
Edit the arch/configure.defaults file and add a stanza for aarch64 and armv7l, using the GNU compiler options.

The default optimization is -03, up from the usual -02. On two 36-h nested simulations on a raspberry pi, the difference
is a consistent 3% speed increase. This used CONUS, so the -03 seems robust.

LIST OF MODIFIED FILES:
M arch/configure.defaults

RELEASE NOTE: Added a new stanza that is only for ARM processors (right now aarch64 and armv7l) with GNU.

[davegill]

@yip111
Would you please confirm that the current version of WRF, v4.3, does NOT already provide build support for ARM.

The base branch was changed.

[davegill]

@yip111
We want to know the following.
Does the existing WRF release v4.3 work with ARM and gfortran, WITHOUT this update?

[davegill]

@yip111
There are a couple of remaining questions about the previous version of the model, and the requirements for certain flags.

[davegill]

@yip111
First of all, we very much appreciate your time. When users provide suggested modifications, it is beneficial to the entire WRF community. There are a couple of questions about this pull request (PR):

1. Without this PR, does the WRF model (branch release-v4.3.1) successfully build on your ARM machine?

2. Regarding the -march=armv8.2-a option, I see this in the gcc man page, and native looks to be a better choice for software compiled and run on the host architecture.

       -march=name
           Specify the name of the target architecture and, optionally,
           one or more feature modifiers.  This option has the form
           -march=arch{+[no]feature}*.

           The permissible values for arch are armv8-a, armv8.1-a,
           armv8.2-a, armv8.3-a, armv8.4-a, armv8.5-a or native.

           The value armv8.5-a implies armv8.4-a and enables compiler
           support for the ARMv8.5-A architecture extensions.

           The value armv8.4-a implies armv8.3-a and enables compiler
           support for the ARMv8.4-A architecture extensions.

           The value armv8.3-a implies armv8.2-a and enables compiler
           support for the ARMv8.3-A architecture extensions.

           The value armv8.2-a implies armv8.1-a and enables compiler
           support for the ARMv8.2-A architecture extensions.

           The value armv8.1-a implies armv8-a and enables compiler
           support for the ARMv8.1-A architecture extension.  In
           particular, it enables the +crc, +lse, and +rdma features.

           The value native is available on native AArch64 GNU/Linux and
           causes the compiler to pick the architecture of the host
           system.  This option has no effect if the compiler is unable
           to recognize the architecture of the host system,

3. With the m4 -G suggestion, are there GNU extensions to m4 that are causing troubles for you?

[yip111]

@davegill
Sorry I'm late.
1.Without this pr, WRF model (branch release-v4.3.1) can successfully build on ARM machine.
2.Adding these flags can bring more performance gains,and it also can provide another reference option for users.

[davegill]

Using the optimization level -03 on arm7l. This is using CONUS physics suite, d02 at 6 km. This is Saturday at 11 AM through Sunday at 5 PM.

@mgduda
Michael,
For no other reason than here is a WRF simulation on a raspberry pi, just wanted to include you.

[davegill]

@kkeene44
Kelly,
No one else is going to review a PR about ARM processors. I have tested this PR on my ARM machine. I have cleaned up the configure.defaults file to be consistent with what we have now as default options. I have modified the commit message to be in final form. Would you please, per this request, approve this additional ARM-processor stanza.

[martin-g]

Thank you!