[libc-commits] [libc] 135cea4 - [libc][math] Fix setting exceptional value for tanf to work with gcc.

[Tue Ly via libc-commits]

Author: Tue Ly
Date: 2023-02-01T15:53:38-05:00
New Revision: 135cea495bc1f13a3f6538ccbd8db723e35cbb9d

URL: https://github.com/llvm/llvm-project/commit/135cea495bc1f13a3f6538ccbd8db723e35cbb9d
DIFF: https://github.com/llvm/llvm-project/commit/135cea495bc1f13a3f6538ccbd8db723e35cbb9d.diff

LOG: [libc][math] Fix setting exceptional value for tanf to work with gcc.

See https://github.com/llvm/llvm-project/issues/59866

Reviewed By: sivachandra

Differential Revision: https://reviews.llvm.org/D143098

Added: 
    

Modified: 
    libc/src/math/generic/tanf.cpp

Removed: 
    


################################################################################
diff  --git a/libc/src/math/generic/tanf.cpp b/libc/src/math/generic/tanf.cpp
index 8b720c48e5de..0f3961ee5422 100644
--- a/libc/src/math/generic/tanf.cpp
+++ b/libc/src/math/generic/tanf.cpp
@@ -97,7 +97,12 @@ LLVM_LIBC_FUNCTION(float, tanf, (float x)) {
     // |x| = 0x1.143ec4p0
     float sign = x_sign ? -1.0f : 1.0f;
 
-    return fputil::multiply_add(sign, 0x1.ddf9f4p0f, sign * 0x1.1p-24f);
+    // volatile is used to prevent compiler (gcc) from optimizing the
+    // computation, making the results incorrect in 
diff erent rounding modes.
+    volatile float tmp = 0x1.ddf9f4p0f;
+    tmp = fputil::multiply_add(sign, tmp, sign * 0x1.1p-24f);
+
+    return tmp;
   }
 
   // |x| > 0x1.ada6a8p+27f