This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Fix i386 acosh (-qNaN) spurious "invalid" exception (bug 19032) [committed]
- From: Joseph Myers <joseph at codesourcery dot com>
- To: <libc-alpha at sourceware dot org>
- Date: Wed, 30 Sep 2015 21:45:35 +0000
- Subject: Fix i386 acosh (-qNaN) spurious "invalid" exception (bug 19032) [committed]
- Authentication-results: sourceware.org; auth=none
The i386 versions of acoshf and acosh raise a spurious "invalid"
exception for an argument that is a quiet NaN with the sign bit set.
The integer arithmetic to detect arguments < 1 also detects -NaN, and
then the computation 0 / 0 in that case raises the exception. This
patch fixes this by using (x - x) / (x - x) as the computation in that
case instead, which will always raise the exception for non-NaN
arguments reaching that code, but not for quiet NaN arguments.
Tested for x86_64 and x86. Committed.
2015-09-30 Joseph Myers <joseph@codesourcery.com>
[BZ #19032]
* sysdeps/i386/fpu/e_acosh.S (__ieee754_acosh): For arguments < 1,
compute result as (x - x) / (x - x) not as 0 / 0.
* sysdeps/i386/fpu/e_acoshf.S (__ieee754_acoshf): Likewise.
* math/libm-test.inc (acosh_test_data): Add another test of acosh.
diff --git a/math/libm-test.inc b/math/libm-test.inc
index 744c66d..8615957 100644
--- a/math/libm-test.inc
+++ b/math/libm-test.inc
@@ -1790,6 +1790,7 @@ static const struct test_f_f_data acosh_test_data[] =
TEST_f_f (acosh, plus_infty, plus_infty, ERRNO_UNCHANGED),
TEST_f_f (acosh, minus_infty, qnan_value, INVALID_EXCEPTION|ERRNO_EDOM),
TEST_f_f (acosh, qnan_value, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
+ TEST_f_f (acosh, -qnan_value, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
/* x < 1: */
TEST_f_f (acosh, 0.75L, qnan_value, INVALID_EXCEPTION|ERRNO_EDOM),
diff --git a/sysdeps/i386/fpu/e_acosh.S b/sysdeps/i386/fpu/e_acosh.S
index c5cd447..263c01a 100644
--- a/sysdeps/i386/fpu/e_acosh.S
+++ b/sysdeps/i386/fpu/e_acosh.S
@@ -91,9 +91,10 @@ ENTRY(__ieee754_acosh)
fyl2x // log(2*x+1/(x+sqrt(x^2-1)))
ret
- // x < 1 => NaN
+ // x < 1 (or -NaN) => NaN
.align ALIGNARG(4)
-5: fldz
+5: fldl 4(%esp)
+ fsub %st
fdiv %st, %st(0)
ret
END(__ieee754_acosh)
diff --git a/sysdeps/i386/fpu/e_acoshf.S b/sysdeps/i386/fpu/e_acoshf.S
index 710267b..779a02c 100644
--- a/sysdeps/i386/fpu/e_acoshf.S
+++ b/sysdeps/i386/fpu/e_acoshf.S
@@ -91,9 +91,10 @@ ENTRY(__ieee754_acoshf)
fyl2x // log(2*x+1/(x+sqrt(x^2-1)))
ret
- // x < 1 => NaN
+ // x < 1 (or -NaN) => NaN
.align ALIGNARG(4)
-5: fldz
+5: flds 4(%esp)
+ fsub %st
fdiv %st, %st(0)
ret
END(__ieee754_acoshf)
--
Joseph S. Myers
joseph@codesourcery.com