This is the mail archive of the glibc-bugs@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug math/15287] New: casinh inaccurate for imaginary part of argument = 1 and small real part


http://sourceware.org/bugzilla/show_bug.cgi?id=15287

             Bug #: 15287
           Summary: casinh inaccurate for imaginary part of argument = 1
                    and small real part
           Product: glibc
           Version: 2.17
            Status: NEW
          Severity: normal
          Priority: P2
         Component: math
        AssignedTo: unassigned@sourceware.org
        ReportedBy: jsm28@gcc.gnu.org
    Classification: Unclassified


The implementation of casinh is substantially inaccurate for arguments with
imaginary part 1 (or -1) and small real part.  For example, on x86_64, casinh
(0x1p-30 + 1.0i) should return 0x1.64564057723e9p-15 + 0x1.921cec97c2264p+0i
but returns 0x1.0000000155515p-15 + 0x1.921db54442d43p+0i, with both parts
substantially inaccurate.  For subnormal arguments (e.g. 0x1.fp-1025), the
inaccurate results underflow where the correct results would not, with
associated underflow exceptions.  Much the same issues affect casin and cacos
for appropriate inputs to those functions.

Patch: http://sourceware.org/ml/libc-alpha/2013-03/msg00465.html

-- 
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.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]