[libclc] r269807 - math: Use single precision fmax in sp path

[Jan Vesely via cfe-commits]

Author: jvesely
Date: Tue May 17 14:44:01 2016
New Revision: 269807

URL: http://llvm.org/viewvc/llvm-project?rev=269807&view=rev
Log:
math: Use single precision fmax in sp path

Fixes fdim piglit on Turks

v2: use CL fmax instead of __builtin

Signed-off-by: Jan Vesely <jan.vesely at rutgers.edu>
Reviewed-by: Tom Stellard <tom.stellard at amd.com>

Modified:
    libclc/trunk/generic/lib/math/fdim.inc

Modified: libclc/trunk/generic/lib/math/fdim.inc
URL: http://llvm.org/viewvc/llvm-project/libclc/trunk/generic/lib/math/fdim.inc?rev=269807&r1=269806&r2=269807&view=diff
==============================================================================
--- libclc/trunk/generic/lib/math/fdim.inc (original)
+++ libclc/trunk/generic/lib/math/fdim.inc Tue May 17 14:44:01 2016
@@ -25,7 +25,7 @@
 _CLC_OVERLOAD _CLC_DEF __CLC_GENTYPE fdim(__CLC_GENTYPE x, __CLC_GENTYPE y) {
     if (__builtin_isnan(x) || __builtin_isnan(y))
         return as_float(QNANBITPATT_SP32);
-    return __builtin_fmax(x - y, 0);
+    return fmax(x - y, 0.0f);
 }
 #define __CLC_FDIM_VEC(width) \
 _CLC_OVERLOAD _CLC_DEF float##width fdim(float##width x, float##width y) { \