The ldbl-128 / ldbl-128ibm implementations of Bessel function do not check if the argument can be doubled before doing so, resulting in spurious overflows for arguments above LDBL_MAX / 2.0L. This is reported for ldbl-128 at http://sourceware.org/ml/libc-alpha/2013-03/msg00434.html and the same applies for ldbl-128ibm (tested for y1l (0x1p1023L) on powerpc64).
Fixed for 2.18 by: commit 98c48fe5cc4317123a168490a8fb37540e23f104 Author: Joseph Myers <joseph@codesourcery.com> Date: Thu Mar 21 13:57:21 2013 +0000 Fix Bessel function spurious overflows for ldbl-128 / ldbl-128ibm (bug 15285).