Bug 6801

Summary: fma() does not set errno for various errors
Product: glibc Reporter: Michael Kerrisk <mtk.manpages>
Component: mathAssignee: Not yet assigned to anyone <unassigned>
Status: NEW ---    
Severity: normal CC: glibc-bugs
Priority: P2 Flags: fweimer: security-
Version: unspecified   
Target Milestone: ---   
Host: Target:
Build: Last reconfirmed:
Attachments: test program

Description Michael Kerrisk 2008-07-31 10:49:50 UTC
For domain, overflow, and underflow errors, fma() correctlry raises exceptions.
 However errno is not set in any case.  errno should be set (respectively EDOM,
ERANGE, ERANGE).


Background: 
On error, many glibc math functions both set errno and raise an exception
(fetestexcept(3)).  For example, the following  function all do this: acos(),
asin(), cosh(), sinh(), acosh(), asinh(), exp(), exp2(), ldexp(), log(),
log10(), log2().  However, there is much inconsistency.  Some functions raise an
exception, but don't set errno.  Some functions set errno for some errors, but
not others.  A few set errno, but don't raise an exception.  This series of bug
reports documents deviations from what I consider the ideal: all functions
should BOTH set errno AND raise an exception for all errors.

All of these reports relate to tests on glibc 2.8 (as provided by SUSE 11.0).
Comment 1 Michael Kerrisk 2008-07-31 10:52:15 UTC
Created attachment 2865 [details]
test program

Sample runs demonstrating the problem:

$ /tmp/mt_fma inf 0 0
errno == 0
fetestexcept() says:  FE_INVALID
fma(inf,0.00000000000000000e+00,0.00000000000000000e+00)=nan
0 FE_INVALID nan

$ /tmp/mt_fma dbl_max dbl_max 0
errno == 0
fetestexcept() says:  FE_OVERFLOW FE_INEXACT
fma(1.79769313486231571e+308,1.79769313486231571e+308,0.00000000000000000e+00)=inf

0 FE_OVERFLOW +inf

$ /tmp/mt_fma dbl_min dbl_min 0
errno == 0
fetestexcept() says:  FE_UNDERFLOW FE_INEXACT
fma(2.22507385850720138e-308,2.22507385850720138e-308,0.00000000000000000e+00)=0.00000000000000000e+00

0 FE_UNDERFLOW +0
Comment 2 Joseph Myers 2012-03-01 17:23:47 UTC
Confirmed with current sources on x86 and x86_64.
Comment 3 Michael Kerrisk 2020-05-23 11:34:30 UTC
Still present in glibc 2.31 (tested on x86_64).