diff --git a/ChangeLog b/ChangeLog
index 5663110435..aec70b6932 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,21 @@
 2017-06-26  Joseph Myers  <joseph@codesourcery.com>
 
+	* sysdeps/ia64/Implies: Add ieee754/float128.
+	* sysdeps/ia64/bits/floatn.h: New file.
+	* sysdeps/ia64/float128-abi.h: Likewise.
+	* manual/math.texi (Mathematics): Document support for _Float128
+	on ia64.
+	* sysdeps/ia64/Makefile [$(subdir) = math] (CPPFLAGS): Append to
+	Makefile variable.
+	* sysdeps/ia64/fpu/e_sqrtf128.c: New file.
+	* sysdeps/ia64/fpu/k_rem_pio2.c: Remove file.
+	* sysdeps/ia64/fpu/sfp-machine.h: New file.  Based on libgcc.
+	* sysdeps/ia64/math-tests.h: New file.
+	* math/libm-test-support.h (XFAIL_FLOAT128_PAYLOAD): Also define
+	based on TEST_COND_binary128 for [__ia64__].
+	* sysdeps/unix/sysv/linux/ia64/libc.abilist: Update.
+	* sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise,
+
 	[BZ #21457]
 	* sysdeps/arm/sys/ucontext.h (struct ucontext): Rename to struct
 	ucontext_t.
diff --git a/NEWS b/NEWS
index a662be27f0..b8deb9c2b6 100644
--- a/NEWS
+++ b/NEWS
@@ -124,7 +124,7 @@ Version 2.26
   C Library is GCC 4.9.  Older GCC versions, and non-GNU compilers, can
   still be used to compile programs using the GNU C Library.
 
-* Support is added, on powerpc64le, x86_64 and x86, for interfaces
+* Support is added, on powerpc64le, x86_64, x86 and ia64, for interfaces
   supporting the _Float128 type from ISO/IEC TS 18661-3:2015.  Most of the
   interfaces are taken from TS 18661-3.  The type-generic macros in <math.h>
   support this type, but those in <tgmath.h> do not.  The GNU C Library now
diff --git a/manual/math.texi b/manual/math.texi
index 39d5b96849..d58b0e531d 100644
--- a/manual/math.texi
+++ b/manual/math.texi
@@ -67,7 +67,7 @@ these functions are described along with the @code{double},
 
 Currently, support for @code{_Float@var{N}} or @code{_Float@var{N}x}
 types is only provided for @code{_Float128} on powerpc64le (PowerPC
-64-bits little-endian), x86_64 and x86.
+64-bits little-endian), x86_64, x86 and ia64.
 
 @menu
 * Mathematical Constants::      Precise numeric values for often-used
diff --git a/math/libm-test-support.h b/math/libm-test-support.h
index 24e3cceec2..44716d0670 100644
--- a/math/libm-test-support.h
+++ b/math/libm-test-support.h
@@ -118,7 +118,8 @@ extern const char doc[];
 /* On some architectures, glibc can be built with compilers that do
    not have suitable built-in functions for setting the payload of a
    _Float128 NaN.  */
-#if (defined __x86_64__ || defined __i386__) && !__GNUC_PREREQ (7, 0)
+#if ((defined __x86_64__ || defined __i386__ || defined __ia64__)	\
+     && !__GNUC_PREREQ (7, 0))
 # define XFAIL_FLOAT128_PAYLOAD (TEST_COND_binary128 ? XFAIL_TEST : 0)
 #else
 # define XFAIL_FLOAT128_PAYLOAD 0
diff --git a/sysdeps/ia64/Implies b/sysdeps/ia64/Implies
index 8524257020..22e4836e74 100644
--- a/sysdeps/ia64/Implies
+++ b/sysdeps/ia64/Implies
@@ -1,5 +1,6 @@
 wordsize-64
 # ia64 uses IEEE 754 floating point.
+ieee754/float128
 ieee754/ldbl-96
 ieee754/dbl-64
 ieee754/flt-32
diff --git a/sysdeps/ia64/Makefile b/sysdeps/ia64/Makefile
index 5ebca89afa..a4305d524d 100644
--- a/sysdeps/ia64/Makefile
+++ b/sysdeps/ia64/Makefile
@@ -1,6 +1,11 @@
 # The ia64 `long double' is a distinct type we support.
 long-double-fcts = yes
 
+ifeq ($(subdir),math)
+# sqrtf128 requires soft-fp.
+CPPFLAGS += -I../soft-fp
+endif
+
 ifeq ($(subdir),gmon)
 sysdep_routines += _mcount
 endif
diff --git a/sysdeps/ia64/bits/floatn.h b/sysdeps/ia64/bits/floatn.h
new file mode 100644
index 0000000000..c7b1df6d2c
--- /dev/null
+++ b/sysdeps/ia64/bits/floatn.h
@@ -0,0 +1,94 @@
+/* Macros to control TS 18661-3 glibc features on ia64.
+   Copyright (C) 2017 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, see
+   <http://www.gnu.org/licenses/>.  */
+
+#ifndef _BITS_FLOATN_H
+#define _BITS_FLOATN_H
+
+#include <features.h>
+
+/* Defined to 1 if the current compiler invocation provides a
+   floating-point type with the IEEE 754 binary128 format, and this
+   glibc includes corresponding *f128 interfaces for it.  The required
+   libgcc support was added some time after the basic compiler
+   support.  */
+#if __GNUC_PREREQ (4, 4)
+# define __HAVE_FLOAT128 1
+#else
+# define __HAVE_FLOAT128 0
+#endif
+
+/* Defined to 1 if __HAVE_FLOAT128 is 1 and the type is ABI-distinct
+   from the default float, double and long double types in this glibc.  */
+#if __HAVE_FLOAT128
+# define __HAVE_DISTINCT_FLOAT128 1
+#else
+# define __HAVE_DISTINCT_FLOAT128 0
+#endif
+
+/* Defined to concatenate the literal suffix to be used with _Float128
+   types, if __HAVE_FLOAT128 is 1. */
+#if __HAVE_FLOAT128
+# if !__GNUC_PREREQ (7, 0) || defined __cplusplus
+/* The literal suffix f128 exists only since GCC 7.0.  */
+#  define __f128(x) x##q
+# else
+#  define __f128(x) x##f128
+# endif
+#endif
+
+/* Defined to a complex binary128 type if __HAVE_FLOAT128 is 1.  */
+#if __HAVE_FLOAT128
+# if !__GNUC_PREREQ (7, 0) || defined __cplusplus
+/* Add a typedef for older GCC compilers which don't natively support
+   _Complex _Float128.  */
+typedef _Complex float __cfloat128 __attribute__ ((__mode__ (__TC__)));
+#  define __CFLOAT128 __cfloat128
+# else
+#  define __CFLOAT128 _Complex _Float128
+# endif
+#endif
+
+/* The remaining of this file provides support for older compilers.  */
+#if __HAVE_FLOAT128
+
+/* The type _Float128 exists only since GCC 7.0.  */
+# if !__GNUC_PREREQ (7, 0) || defined __cplusplus
+typedef __float128 _Float128;
+# endif
+
+/* __builtin_huge_valf128 doesn't exist before GCC 7.0.  */
+# if !__GNUC_PREREQ (7, 0)
+#  define __builtin_huge_valf128() ((_Float128) __builtin_huge_val ())
+# endif
+
+/* Older GCC has only a subset of built-in functions for _Float128 on
+   ia64, and __builtin_infq is not usable in static initializers.
+   Converting a narrower sNaN to _Float128 produces a quiet NaN, so
+   attempts to use _Float128 sNaNs will not work properly with older
+   compilers.  */
+# if !__GNUC_PREREQ (7, 0)
+#  define __builtin_copysignf128 __builtin_copysignq
+#  define __builtin_fabsf128 __builtin_fabsq
+#  define __builtin_inff128() ((_Float128) __builtin_inf ())
+#  define __builtin_nanf128(x) ((_Float128) __builtin_nan (x))
+#  define __builtin_nansf128(x) ((_Float128) __builtin_nans (x))
+# endif
+
+#endif
+
+#endif /* _BITS_FLOATN_H */
diff --git a/sysdeps/ia64/float128-abi.h b/sysdeps/ia64/float128-abi.h
new file mode 100644
index 0000000000..6b954cc673
--- /dev/null
+++ b/sysdeps/ia64/float128-abi.h
@@ -0,0 +1,2 @@
+/* ABI version for _Float128 ABI introduction.  */
+#define FLOAT128_VERSION GLIBC_2.26
diff --git a/sysdeps/ia64/fpu/e_sqrtf128.c b/sysdeps/ia64/fpu/e_sqrtf128.c
new file mode 100644
index 0000000000..d5bcb79aaa
--- /dev/null
+++ b/sysdeps/ia64/fpu/e_sqrtf128.c
@@ -0,0 +1 @@
+#include <sysdeps/x86/fpu/e_sqrtf128.c>
diff --git a/sysdeps/ia64/fpu/k_rem_pio2.c b/sysdeps/ia64/fpu/k_rem_pio2.c
deleted file mode 100644
index 41254ae60a..0000000000
--- a/sysdeps/ia64/fpu/k_rem_pio2.c
+++ /dev/null
@@ -1 +0,0 @@
-/* Not needed. */
diff --git a/sysdeps/ia64/fpu/sfp-machine.h b/sysdeps/ia64/fpu/sfp-machine.h
new file mode 100644
index 0000000000..45e844d3da
--- /dev/null
+++ b/sysdeps/ia64/fpu/sfp-machine.h
@@ -0,0 +1,100 @@
+#define _FP_W_TYPE_SIZE		64
+#define _FP_W_TYPE		unsigned long
+#define _FP_WS_TYPE		signed long
+#define _FP_I_TYPE		long
+
+typedef int TItype __attribute__ ((mode (TI)));
+typedef unsigned int UTItype __attribute__ ((mode (TI)));
+
+#define TI_BITS (__CHAR_BIT__ * (int)sizeof(TItype))
+
+/* The type of the result of a floating point comparison.  This must
+   match `__libgcc_cmp_return__' in GCC for the target.  */
+typedef int __gcc_CMPtype __attribute__ ((mode (__libgcc_cmp_return__)));
+#define CMPtype __gcc_CMPtype
+
+#define _FP_MUL_MEAT_Q(R,X,Y)				\
+  _FP_MUL_MEAT_2_wide(_FP_WFRACBITS_Q,R,X,Y,umul_ppmm)
+
+#define _FP_DIV_MEAT_Q(R,X,Y)   _FP_DIV_MEAT_2_udiv(Q,R,X,Y)
+
+#define _FP_NANFRAC_S		_FP_QNANBIT_S
+#define _FP_NANFRAC_D		_FP_QNANBIT_D
+#define _FP_NANFRAC_E		_FP_QNANBIT_E, 0
+#define _FP_NANFRAC_Q		_FP_QNANBIT_Q, 0
+
+#define _FP_KEEPNANFRACP	1
+#define _FP_QNANNEGATEDP 0
+
+#define _FP_NANSIGN_S		1
+#define _FP_NANSIGN_D		1
+#define _FP_NANSIGN_E		1
+#define _FP_NANSIGN_Q		1
+
+/* Here is something Intel misdesigned: the specs don't define
+   the case where we have two NaNs with same mantissas, but
+   different sign. Different operations pick up different NaNs.  */
+#define _FP_CHOOSENAN(fs, wc, R, X, Y, OP)			\
+  do {								\
+    if (_FP_FRAC_GT_##wc(X, Y)					\
+	|| (_FP_FRAC_EQ_##wc(X,Y) && (OP == '+' || OP == '*')))	\
+      {								\
+	R##_s = X##_s;						\
+	_FP_FRAC_COPY_##wc(R,X);				\
+      }								\
+    else							\
+      {								\
+	R##_s = Y##_s;						\
+	_FP_FRAC_COPY_##wc(R,Y);				\
+      }								\
+    R##_c = FP_CLS_NAN;						\
+  } while (0)
+
+#define FP_EX_INVALID		0x01
+#define FP_EX_DENORM		0x02
+#define FP_EX_DIVZERO		0x04
+#define FP_EX_OVERFLOW		0x08
+#define FP_EX_UNDERFLOW		0x10
+#define FP_EX_INEXACT		0x20
+#define FP_EX_ALL \
+	(FP_EX_INVALID | FP_EX_DENORM | FP_EX_DIVZERO | FP_EX_OVERFLOW \
+	 | FP_EX_UNDERFLOW | FP_EX_INEXACT)
+
+#define _FP_TININESS_AFTER_ROUNDING 1
+
+void __sfp_handle_exceptions (int);
+
+#define FP_HANDLE_EXCEPTIONS			\
+  do {						\
+    if (__builtin_expect (_fex, 0))		\
+      __sfp_handle_exceptions (_fex);		\
+  } while (0);
+
+#define FP_TRAPPING_EXCEPTIONS	(~_fcw & FP_EX_ALL)
+
+#define FP_RND_NEAREST		0
+#define FP_RND_ZERO		0xc00L
+#define FP_RND_PINF		0x800L
+#define FP_RND_MINF		0x400L
+
+#define FP_RND_MASK		0xc00L
+
+#define _FP_DECL_EX \
+  unsigned long int _fcw __attribute__ ((unused)) = FP_RND_NEAREST
+
+#define FP_INIT_ROUNDMODE					\
+  do {								\
+    __asm__ __volatile__ ("mov.m %0 = ar.fpsr" : "=r" (_fcw));	\
+  } while (0)
+
+#define FP_ROUNDMODE		(_fcw & FP_RND_MASK)
+
+#define	__LITTLE_ENDIAN	1234
+#define	__BIG_ENDIAN	4321
+
+#define __BYTE_ORDER __LITTLE_ENDIAN
+
+/* Define ALIASNAME as a strong alias for NAME.  */
+#define strong_alias(name, aliasname) _strong_alias(name, aliasname)
+#define _strong_alias(name, aliasname) \
+  extern __typeof (name) aliasname __attribute__ ((alias (#name)));
diff --git a/sysdeps/ia64/math-tests.h b/sysdeps/ia64/math-tests.h
new file mode 100644
index 0000000000..439ac75e5d
--- /dev/null
+++ b/sysdeps/ia64/math-tests.h
@@ -0,0 +1,25 @@
+/* Configuration for math tests.  ia64 version.
+   Copyright (C) 2017 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, see
+   <http://www.gnu.org/licenses/>.  */
+
+/* Before GCC 7, there is no built-in function to provide a __float128
+   sNaN, so most sNaN tests for this type cannot work.  */
+#if !__GNUC_PREREQ (7, 0)
+# define SNAN_TESTS_float128	0
+#endif
+
+#include_next <math-tests.h>
diff --git a/sysdeps/unix/sysv/linux/ia64/libc.abilist b/sysdeps/unix/sysv/linux/ia64/libc.abilist
index 39ccf2b5ee..62b67b859d 100644
--- a/sysdeps/unix/sysv/linux/ia64/libc.abilist
+++ b/sysdeps/unix/sysv/linux/ia64/libc.abilist
@@ -1885,11 +1885,18 @@ GLIBC_2.25 strfromd F
 GLIBC_2.25 strfromf F
 GLIBC_2.25 strfroml F
 GLIBC_2.26 GLIBC_2.26 A
+GLIBC_2.26 __strtof128_internal F
+GLIBC_2.26 __wcstof128_internal F
 GLIBC_2.26 preadv2 F
 GLIBC_2.26 preadv64v2 F
 GLIBC_2.26 pwritev2 F
 GLIBC_2.26 pwritev64v2 F
 GLIBC_2.26 reallocarray F
+GLIBC_2.26 strfromf128 F
+GLIBC_2.26 strtof128 F
+GLIBC_2.26 strtof128_l F
+GLIBC_2.26 wcstof128 F
+GLIBC_2.26 wcstof128_l F
 GLIBC_2.3 GLIBC_2.3 A
 GLIBC_2.3 __ctype_b_loc F
 GLIBC_2.3 __ctype_tolower_loc F
diff --git a/sysdeps/unix/sysv/linux/ia64/libm.abilist b/sysdeps/unix/sysv/linux/ia64/libm.abilist
index d7d80065fb..f165eb3957 100644
--- a/sysdeps/unix/sysv/linux/ia64/libm.abilist
+++ b/sysdeps/unix/sysv/linux/ia64/libm.abilist
@@ -402,4 +402,142 @@ GLIBC_2.25 ufromfpl F
 GLIBC_2.25 ufromfpx F
 GLIBC_2.25 ufromfpxf F
 GLIBC_2.25 ufromfpxl F
+GLIBC_2.26 GLIBC_2.26 A
+GLIBC_2.26 __acosf128_finite F
+GLIBC_2.26 __acoshf128_finite F
+GLIBC_2.26 __asinf128_finite F
+GLIBC_2.26 __atan2f128_finite F
+GLIBC_2.26 __atanhf128_finite F
+GLIBC_2.26 __coshf128_finite F
+GLIBC_2.26 __exp10f128_finite F
+GLIBC_2.26 __exp2f128_finite F
+GLIBC_2.26 __expf128_finite F
+GLIBC_2.26 __finitef128 F
+GLIBC_2.26 __fmodf128_finite F
+GLIBC_2.26 __fpclassifyf128 F
+GLIBC_2.26 __gammaf128_r_finite F
+GLIBC_2.26 __hypotf128_finite F
+GLIBC_2.26 __iseqsigf128 F
+GLIBC_2.26 __isinff128 F
+GLIBC_2.26 __isnanf128 F
+GLIBC_2.26 __issignalingf128 F
+GLIBC_2.26 __j0f128_finite F
+GLIBC_2.26 __j1f128_finite F
+GLIBC_2.26 __jnf128_finite F
+GLIBC_2.26 __lgammaf128_r_finite F
+GLIBC_2.26 __log10f128_finite F
+GLIBC_2.26 __log2f128_finite F
+GLIBC_2.26 __logf128_finite F
+GLIBC_2.26 __powf128_finite F
+GLIBC_2.26 __remainderf128_finite F
+GLIBC_2.26 __signbitf128 F
+GLIBC_2.26 __sinhf128_finite F
+GLIBC_2.26 __sqrtf128_finite F
+GLIBC_2.26 __y0f128_finite F
+GLIBC_2.26 __y1f128_finite F
+GLIBC_2.26 __ynf128_finite F
+GLIBC_2.26 acosf128 F
+GLIBC_2.26 acoshf128 F
+GLIBC_2.26 asinf128 F
+GLIBC_2.26 asinhf128 F
+GLIBC_2.26 atan2f128 F
+GLIBC_2.26 atanf128 F
+GLIBC_2.26 atanhf128 F
+GLIBC_2.26 cabsf128 F
+GLIBC_2.26 cacosf128 F
+GLIBC_2.26 cacoshf128 F
+GLIBC_2.26 canonicalizef128 F
+GLIBC_2.26 cargf128 F
+GLIBC_2.26 casinf128 F
+GLIBC_2.26 casinhf128 F
+GLIBC_2.26 catanf128 F
+GLIBC_2.26 catanhf128 F
+GLIBC_2.26 cbrtf128 F
+GLIBC_2.26 ccosf128 F
+GLIBC_2.26 ccoshf128 F
+GLIBC_2.26 ceilf128 F
+GLIBC_2.26 cexpf128 F
+GLIBC_2.26 cimagf128 F
+GLIBC_2.26 clog10f128 F
+GLIBC_2.26 clogf128 F
+GLIBC_2.26 conjf128 F
+GLIBC_2.26 copysignf128 F
+GLIBC_2.26 cosf128 F
+GLIBC_2.26 coshf128 F
+GLIBC_2.26 cpowf128 F
+GLIBC_2.26 cprojf128 F
+GLIBC_2.26 crealf128 F
+GLIBC_2.26 csinf128 F
+GLIBC_2.26 csinhf128 F
+GLIBC_2.26 csqrtf128 F
+GLIBC_2.26 ctanf128 F
+GLIBC_2.26 ctanhf128 F
+GLIBC_2.26 erfcf128 F
+GLIBC_2.26 erff128 F
+GLIBC_2.26 exp10f128 F
+GLIBC_2.26 exp2f128 F
+GLIBC_2.26 expf128 F
+GLIBC_2.26 expm1f128 F
+GLIBC_2.26 fabsf128 F
+GLIBC_2.26 fdimf128 F
+GLIBC_2.26 floorf128 F
+GLIBC_2.26 fmaf128 F
+GLIBC_2.26 fmaxf128 F
+GLIBC_2.26 fmaxmagf128 F
+GLIBC_2.26 fminf128 F
+GLIBC_2.26 fminmagf128 F
+GLIBC_2.26 fmodf128 F
+GLIBC_2.26 frexpf128 F
+GLIBC_2.26 fromfpf128 F
+GLIBC_2.26 fromfpxf128 F
+GLIBC_2.26 getpayloadf128 F
+GLIBC_2.26 hypotf128 F
+GLIBC_2.26 ilogbf128 F
+GLIBC_2.26 j0f128 F
+GLIBC_2.26 j1f128 F
+GLIBC_2.26 jnf128 F
+GLIBC_2.26 ldexpf128 F
+GLIBC_2.26 lgammaf128 F
+GLIBC_2.26 lgammaf128_r F
+GLIBC_2.26 llogbf128 F
+GLIBC_2.26 llrintf128 F
+GLIBC_2.26 llroundf128 F
+GLIBC_2.26 log10f128 F
+GLIBC_2.26 log1pf128 F
+GLIBC_2.26 log2f128 F
+GLIBC_2.26 logbf128 F
+GLIBC_2.26 logf128 F
+GLIBC_2.26 lrintf128 F
+GLIBC_2.26 lroundf128 F
+GLIBC_2.26 modff128 F
+GLIBC_2.26 nanf128 F
+GLIBC_2.26 nearbyintf128 F
+GLIBC_2.26 nextafterf128 F
+GLIBC_2.26 nextdownf128 F
+GLIBC_2.26 nextupf128 F
+GLIBC_2.26 powf128 F
+GLIBC_2.26 remainderf128 F
+GLIBC_2.26 remquof128 F
+GLIBC_2.26 rintf128 F
+GLIBC_2.26 roundevenf128 F
+GLIBC_2.26 roundf128 F
+GLIBC_2.26 scalblnf128 F
+GLIBC_2.26 scalbnf128 F
+GLIBC_2.26 setpayloadf128 F
+GLIBC_2.26 setpayloadsigf128 F
+GLIBC_2.26 sincosf128 F
+GLIBC_2.26 sinf128 F
+GLIBC_2.26 sinhf128 F
+GLIBC_2.26 sqrtf128 F
+GLIBC_2.26 tanf128 F
+GLIBC_2.26 tanhf128 F
+GLIBC_2.26 tgammaf128 F
+GLIBC_2.26 totalorderf128 F
+GLIBC_2.26 totalordermagf128 F
+GLIBC_2.26 truncf128 F
+GLIBC_2.26 ufromfpf128 F
+GLIBC_2.26 ufromfpxf128 F
+GLIBC_2.26 y0f128 F
+GLIBC_2.26 y1f128 F
+GLIBC_2.26 ynf128 F
 GLIBC_2.4 GLIBC_2.4 A