This is the mail archive of the
glibc-bugs@sourceware.org
mailing list for the glibc project.
[Bug math/14869] New: ldbl-128ibm hypotl inaccurate and underflows for arguments around 2^-500
- From: "jsm28 at gcc dot gnu.org" <sourceware-bugzilla at sourceware dot org>
- To: glibc-bugs at sources dot redhat dot com
- Date: Thu, 22 Nov 2012 02:05:16 +0000
- Subject: [Bug math/14869] New: ldbl-128ibm hypotl inaccurate and underflows for arguments around 2^-500
- Auto-submitted: auto-generated
http://sourceware.org/bugzilla/show_bug.cgi?id=14869
Bug #: 14869
Summary: ldbl-128ibm hypotl inaccurate and underflows for
arguments around 2^-500
Product: glibc
Version: 2.16
Status: NEW
Severity: normal
Priority: P2
Component: math
AssignedTo: unassigned@sourceware.org
ReportedBy: jsm28@gcc.gnu.org
Classification: Unclassified
The ldbl-128ibm implementation of hypotl scales arguments up by 2^600 if below
2^-500. But for IBM long double, a number around 2^-1000 is subnormal (less
than 106 bits precision); that is, some numbers above 2^-500 need scaling up to
avoid large loss of precision from squaring. hypotl
(0x1.23456789abcdef0123456789ab8p-500L, 0x1.23456789abcdef0123456789ab8p-500L)
gives large errors and spurious underflow exceptions, for example.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.