riscv: Add a soft-float implementation of fabs()

We could just use a C implementation using __builtin_fabs(), but using
this assembly version guarantees that there is no additional prolog/epilog
code. Additionally, clang generates worse code for masking off the top bit
than GCC: https://bugs.llvm.org/show_bug.cgi?id=49377.

This fixes the RISCV64 softfloat world build after cf97d2a1da. That commit
added -fno-builtin to the msun tests which resulted in the first references to
fabs (previously the compiler inlined all calls).

Reviewed By:	dim
Reported by:	mjg
Differential Revision: https://reviews.freebsd.org/D28994

lib/libc/riscv/gen/fabs.S

@@ -1,5 +1,6 @@
 /*-
  * Copyright (c) 2015-2017 Ruslan Bukin <br@bsdpad.com>
+ * Copyright (c) 2021 Alex Richardson <arichardson@FreeBSD.org>
  * All rights reserved.
  *
  * Portions of this software were developed by SRI International and the
@@ -10,6 +11,9 @@
  * Computer Laboratory as part of the CTSRD Project, with support from the
  * UK Higher Education Innovation Fund (HEIF).
  *
+ * This work was supported by Innovate UK project 105694, "Digital Security
+ * by Design (DSbD) Technology Platform Prototype".
+ *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
  * are met:
@@ -35,9 +39,12 @@
 #include <machine/asm.h>
 __FBSDID("$FreeBSD$");
 
-#ifdef __riscv_float_abi_double
 ENTRY(fabs)
+#ifdef __riscv_float_abi_double
 	fabs.d	fa0, fa0
+#else
+	slli a0,a0,1
+	srli a0,a0,1
+#endif
 	ret
 END(fabs)
-#endif