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/18046] New: [ldbl-96] atanhl inaccurate


https://sourceware.org/bugzilla/show_bug.cgi?id=18046

            Bug ID: 18046
           Summary: [ldbl-96] atanhl inaccurate
           Product: glibc
           Version: 2.21
            Status: NEW
          Severity: normal
          Priority: P2
         Component: math
          Assignee: unassigned at sourceware dot org
          Reporter: jsm28 at gcc dot gnu.org

The threshold in ldbl-96 atanhl for when to return the argument, 0x1p-28, is a
bit too big (the relevant condition being x^3/3 being < 0.5ulp of x), resulting
in errors a bit above the limits of those considered acceptable in glibc, e.g.
on x86_64:

Failure: Test: atanh (0x8p-32)
Result:
 is:          1.86264514923095703125e-09   0x8.00000000000000000000p-32
 should be:   1.86264514923095703347e-09   0x8.00000000000000b00000p-32
 difference:  2.22143230910236924404e-27   0xb.00000000000000000000p-92
 ulp       :  11.0000
 max.ulp   :  1.0000

-- 
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]