Bug 15426 - Wrong sign of infinity from tgamma overflow
Summary: Wrong sign of infinity from tgamma overflow
Status: RESOLVED FIXED
Alias: None
Product: glibc
Classification: Unclassified
Component: math (show other bugs)
Version: 2.17
: P2 normal
Target Milestone: ---
Assignee: Not yet assigned to anyone
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-05-01 22:17 UTC by Joseph Myers
Modified: 2014-06-13 18:09 UTC (History)
0 users

See Also:
Host:
Target:
Build:
Last reconfirmed:
fweimer: security-

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Joseph Myers 2013-05-01 22:17:58 UTC 
When tgamma of a negative argument very close to 0 overflows (the result is essentially 1/x), it should produce negative infinity but wrongly produces positive infinity.  Example on x86_64: tgamma (-0x1p-1074).

(Discovered while working on a fix for the various known bugs about large ulps from tgamma.)
Comment 1 Joseph Myers 2013-05-08 12:00:35 UTC 
Fixed for 2.18 by:

commit d8cd06db62d92f86cc8cc3c0d6a489bd207bb834
Author: Joseph Myers <joseph@codesourcery.com>
Date:   Wed May 8 11:58:18 2013 +0000

    Improve tgamma accuracy (bugs 2546, 2560, 5159, 15426).