about summary refs log tree commit diff
path: root/src/math/asinf.c
diff options
context:
space:
mode:
authorSzabolcs Nagy <nsz@port70.net>2013-05-15 23:08:52 +0000
committerSzabolcs Nagy <nsz@port70.net>2013-05-15 23:08:52 +0000
commite216951f509b71da193da2fc63e25b998740d58b (patch)
tree391381eafd8b0b7eb4fdbb693608072df5c181d5 /src/math/asinf.c
parent2c184264eae3797de028403ed1e86c1f7ae5b813 (diff)
downloadmusl-e216951f509b71da193da2fc63e25b998740d58b.tar.gz
musl-e216951f509b71da193da2fc63e25b998740d58b.tar.xz
musl-e216951f509b71da193da2fc63e25b998740d58b.zip
math: use double_t for temporaries to avoid stores on i386
When FLT_EVAL_METHOD!=0 (only i386 with x87 fp) the excess
precision of an expression must be removed in an assignment.
(gcc needs -fexcess-precision=standard or -std=c99 for this)

This is done by extra load/store instructions which adds code
bloat when lot of temporaries are used and it makes the result
less precise in many cases.
Using double_t and float_t avoids these issues on i386 and
it makes no difference on other archs.

For now only a few functions are modified where the excess
precision is clearly beneficial (mostly polynomial evaluations
with temporaries).

object size differences on i386, gcc-4.8:
             old   new
__cosdf.o    123    95
__cos.o      199   169
__sindf.o    131    95
__sin.o      225   203
__tandf.o    207   151
__tan.o      605   499
erff.o      1470  1416
erf.o       1703  1649
j0f.o       1779  1745
j0.o        2308  2274
j1f.o       1602  1568
j1.o        2286  2252
tgamma.o    1431  1424
math/*.o   64164 63635
Diffstat (limited to 'src/math/asinf.c')
-rw-r--r--src/math/asinf.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/src/math/asinf.c b/src/math/asinf.c
index 462bf043..51fe6c61 100644
--- a/src/math/asinf.c
+++ b/src/math/asinf.c
@@ -26,7 +26,7 @@ qS1 = -7.0662963390e-01;
 
 static float R(float z)
 {
-	float p, q;
+	float_t p, q;
 	p = z*(pS0+z*(pS1+z*pS2));
 	q = 1.0f+z*qS1;
 	return p/q;