Bug 6779 – remainder() wrongly gives a domain error for (NaN, 0)

Bug 6779 - remainder() wrongly gives a domain error for (NaN, 0)


Summary:	remainder() wrongly gives a domain error for (NaN, 0)

Status:	RESOLVED FIXED

Product:	glibc
Classification:	Unclassified
Component:	math
Version:	unspecified

Importance:	P2 normal
Target Milestone:	---
Assigned To:	Andreas Jaeger

URL:
Keywords:

Depends on:
Blocks:
	Show dependency tree / graph

Reported:	2008-07-30 11:14 UTC by Michael Kerrisk
Modified:	2011-10-07 19:16 UTC (History)
CC List:	2 users (show)

See Also:
Host:
Target:
Build:
Last reconfirmed:

Attachments
test program (11.55 KB, text/plain) 2008-07-31 08:34 UTC, Michael Kerrisk	Details
View All 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 Michael Kerrisk 2008-07-30 11:14:27 UTC

The posix reaminder() spec says:

==
If x or y is NaN, a NaN shall be returned.

If x is infinite or y is 0 and the other is non-NaN, a domain error shall occur,
and either a NaN (if supported), or an implementation-defined value shall be
returned
==

On glibc 2.8, remainder(NaN, 0) gives a domain error (EDOM, FE_INVALID).  As per
the above spec, it should return a NaN without error.

Comment 1 Michael Kerrisk 2008-07-31 08:34:57 UTC

Created attachment 2844 [details]
test program

Sample run showing problem:

$ /tmp/mt_remainder nan 0
errno == EDOM
fetestexcept() says:  FE_INVALID
remainder(nan,0.00000000000000000e+00)=nan
EDOM FE_INVALID nan

Comment 2 Andreas Jaeger 2011-09-29 09:33:37 UTC

Thanks, patch send to libc-alpha mailinglist.

Comment 3 Ulrich Drepper 2011-10-07 19:16:05 UTC

I've checked in the patch.