mpp_check_field fix for test_mpp_domains2

[laurenchilutti]

Description
The test_mpp_domains2 test fails at line 234 in mpp_update_domains.fh with a seg fault. Tracking through the backtrace of this error (you can see the full backtrace in issue #1290) I think I have found the root cause:

In file mpp_domains_misc.inc subroutine mpp_check_field_2d_type2 we call mpp_redistribute at line 426 with the following arguments:

FMS/mpp/include/mpp_domains_misc.inc

Line 426 in 2be8aa4

call mpp_redistribute(domain1,field_in,domain2,field1)

Because field1 and field2 are 2D arrays, mpp_redistribute will use the subroutine MPP_REDISTRIBUTE_2D_ from mpp_update_domains2D.fh

field1 in mpp_check_field_2d_type2 is field_out in MPP_REDISTRIBUTE_2D_ and the program fails when we try to set

FMS/mpp/include/mpp_update_domains2D.fh

Line 234 in 2be8aa4

field_out = 0

I think that this is happening because field1 was only allocated on pelist2 but mpp_redistribute is called from all pes . So pes from pelist1 will not have allocated field1.

FMS/mpp/include/mpp_domains_misc.inc

Lines 412 to 420 in 2be8aa4

	! Comparison is made on pelist2.
	if(any(pelist2 == pe)) then
	call mpp_get_data_domain(domain2, isd, ied, jsd, jed)
	call mpp_get_compute_domain(domain2, is, ie, js, je)
	allocate(field1(isd:ied, jsd:jed),field2(isd:ied, jsd:jed))
	if((size(field_in,1) .ne. ied-isd+1) .or. (size(field_in,2) .ne. jed-jsd+1)) &
	call mpp_error(FATAL,'mpp_check_field: input field is not on the data domain')
	field2(isd:ied, jsd:jed) = field_in(:,:)
	endif

The solution I propose in this PR is to allocate field1 outside of the if(any(pelist2 == pe)) condition so that all pes allocate.

Fixes #1290

How Has This Been Tested?
Tested with GCC 13.1.0 on GFDL dev box with openmpi make and make check

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published in downstream modules
• New check tests, if applicable, are included
• make distcheck passes

[laurenchilutti]

This also fixes Issue #1101