port-alpha/59161: lib/libc/gen/t_fpsetround test failures on Alpha

[thorpej%me.com@localhost]

>Number:         59161
>Category:       port-alpha
>Synopsis:       lib/libc/gen/t_fpsetround test failures on Alpha
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    port-alpha-maintainer
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Mar 09 15:20:00 +0000 2025
>Originator:     Jason Thorpe
>Release:        NetBSD 10.1
>Organization:
RISCy Business
>Environment:
NetBSD avanti.ktnet 10.1 NetBSD 10.1 (GENERIC) #0: Mon Dec 16 13:08:11 UTC 2024  mkrepro%mkrepro.NetBSD.org@localhost:/usr/src/sys/arch/alpha/compile/GENERIC alpha

hw.model = AlphaStation 200 4/233
>Description:
avanti:thorpej 8$ ./t_fpsetround fpsetround_basic                              
t_fpsetround: WARNING: Running test cases without atf-run(1) is unsupported
t_fpsetround: WARNING: No isolation nor timeout control is being applied; you may get unexpected failures; see atf-test-case(4)
Illegal instruction 
avanti:thorpej 9$ ./t_fpsetround fpsetround_noftz 
t_fpsetround: WARNING: Running test cases without atf-run(1) is unsupported
t_fpsetround: WARNING: No isolation nor timeout control is being applied; you may get unexpected failures; see atf-test-case(4)
*** Check failed: /usr/src/tests/lib/libc/gen/t_fpsetround.c:174: machine runs flush-to-zero by default
*** Check failed: /usr/src/tests/lib/libc/gen/t_fpsetround.c:186: machine runs flush-to-zero after one fpsetround call
*** Check failed: /usr/src/tests/lib/libc/gen/t_fpsetround.c:192: machine runs flush-to-zero after two fpsetround calls
failed: 3 checks failed; see output for more details
avanti:thorpej 10$

This is almost certainly because this code doesn't properly integrate with the kernel's FP completion code, and fiddles with the FPCR directly.
>How-To-Repeat:
See above.
>Fix:
Consult the Alpha Architecture Reference Manual for all of the gory details.