This is the mail archive of the
glibc-bugs@sourceware.org
mailing list for the glibc project.
[Bug math/16271] New: dbl-64 e_sqrt.c not correctly rounding in non-default rounding modes
- From: "jsm28 at gcc dot gnu.org" <sourceware-bugzilla at sourceware dot org>
- To: glibc-bugs at sourceware dot org
- Date: Wed, 27 Nov 2013 21:42:06 +0000
- Subject: [Bug math/16271] New: dbl-64 e_sqrt.c not correctly rounding in non-default rounding modes
- Auto-submitted: auto-generated
https://sourceware.org/bugzilla/show_bug.cgi?id=16271
Bug ID: 16271
Summary: dbl-64 e_sqrt.c not correctly rounding in non-default
rounding modes
Product: glibc
Version: 2.19
Status: NEW
Severity: normal
Priority: P2
Component: math
Assignee: unassigned at sourceware dot org
Reporter: jsm28 at gcc dot gnu.org
sysdeps/ieee754/dbl-64/e_sqrt.c does not produce correct results for all inputs
in rounding modes other than round-to-nearest, as illustrated by (on ARM):
Failure: Test: sqrt_towardzero (0xf.fffffp+124)
Result:
is: 1.84467435239537295360e+19 0x1.fffffeffffffc0000000p+63
should be: 1.84467435239537274880e+19 0x1.fffffeffffffb0000000p+63
difference: 2.04800000000000000000e+03 0x1.00000000000000000000p+11
ulp : 1.0000
max.ulp : 0.0000
Failure: Test: sqrt_towardzero (0xf.fffffp-4)
Result:
is: 9.99999970197677168215e-01 0x1.fffffeffffffc0000000p-1
should be: 9.99999970197677057193e-01 0x1.fffffeffffffb0000000p-1
difference: 1.11022302462515654042e-16 0x1.00000000000000000000p-53
ulp : 1.0000
max.ulp : 0.0000
Maximal error of `sqrt_towardzero'
is : 1 ulp
accepted: 0 ulp
Failure: Test: sqrt_downward (0xf.fffffp+124)
Result:
is: 1.84467435239537295360e+19 0x1.fffffeffffffc0000000p+63
should be: 1.84467435239537274880e+19 0x1.fffffeffffffb0000000p+63
difference: 2.04800000000000000000e+03 0x1.00000000000000000000p+11
ulp : 1.0000
max.ulp : 0.0000
Failure: Test: sqrt_downward (0xf.fffffp-4)
Result:
is: 9.99999970197677168215e-01 0x1.fffffeffffffc0000000p-1
should be: 9.99999970197677057193e-01 0x1.fffffeffffffb0000000p-1
difference: 1.11022302462515654042e-16 0x1.00000000000000000000p-53
ulp : 1.0000
max.ulp : 0.0000
Maximal error of `sqrt_downward'
is : 1 ulp
accepted: 0 ulp
Failure: Test: sqrt_upward (0x1.000002p+0)
Result:
is: 1.00000005960464299904e+00 0x1.000000ffffff80000000p+0
should be: 1.00000005960464322108e+00 0x1.000000ffffff90000000p+0
difference: 2.22044604925031308085e-16 0x1.00000000000000000000p-52
ulp : 1.0000
max.ulp : 0.0000
Maximal error of `sqrt_upward'
is : 1 ulp
accepted: 0 ulp
--
You are receiving this mail because:
You are on the CC list for the bug.