From: Szabolcs Nagy <nsz@port70.net>
Date: Tue, 16 Oct 2018 22:20:39 +0000 (+0000)
Subject: define FP_FAST_FMA* when fma* can be inlined
X-Git-Tag: v1.1.23~66
X-Git-Url: https://git.librecmc.org/?a=commitdiff_plain;h=e980ca7a571465e8a4c887a199491c2cd8d0c0ee;p=oweals%2Fmusl.git

define FP_FAST_FMA* when fma* can be inlined

FP_FAST_FMA can be defined if "the fma function generally executes about
as fast as, or faster than, a multiply and an add of double operands",
which can only be true if the fma call is inlined as an instruction.

gcc sets __FP_FAST_FMA if __builtin_fma is inlined as an instruction,
but that does not mean an fma call will be inlined (e.g. it is defined
with -fno-builtin-fma), other compilers (clang) don't even have such
macro, but this is the closest we can get.

(even if the libc fma implementation is a single instruction, the extern
call overhead is already too big when the macro is used to decide between
x*y+z and fma(x,y,z) so it cannot be based on libc only, defining the
macro unconditionally on targets which have fma in the base isa is also
incorrect: the compiler might not inline fma anyway.)

this solution works with gcc unless fma inlining is explicitly turned off.
---

diff --git a/include/math.h b/include/math.h
index fea34686..14f28ec8 100644
--- a/include/math.h
+++ b/include/math.h
@@ -36,6 +36,18 @@ extern "C" {
 #define FP_SUBNORMAL 3
 #define FP_NORMAL    4
 
+#ifdef __FP_FAST_FMA
+#define FP_FAST_FMA 1
+#endif
+
+#ifdef __FP_FAST_FMAF
+#define FP_FAST_FMAF 1
+#endif
+
+#ifdef __FP_FAST_FMAL
+#define FP_FAST_FMAL 1
+#endif
+
 int __fpclassify(double);
 int __fpclassifyf(float);
 int __fpclassifyl(long double);