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/16338] New: ldbl-128 logl does not handle subnormals


https://sourceware.org/bugzilla/show_bug.cgi?id=16338

            Bug ID: 16338
           Summary: ldbl-128 logl does not handle subnormals
           Product: glibc
           Version: 2.19
            Status: NEW
          Severity: normal
          Priority: P2
         Component: math
          Assignee: unassigned at sourceware dot org
          Reporter: jsm28 at gcc dot gnu.org

The ldbl-128 logl implementation treats subnormals as if their exponent is
normal, resulting in large errors (e.g. logl (0x1p-16494L) is
-0x1.62dea45ee3e064dbc66918126b26p+13 when it should be
-0x1.6546282207802c89d24d65e96274p+13).  I think this is in turn the cause of
large lgammal errors in libm testing.

-- 
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]