diff --git a/lib/msun/Makefile b/lib/msun/Makefile
--- a/lib/msun/Makefile
+++ b/lib/msun/Makefile
@@ -122,7 +122,8 @@
 COMMON_SRCS+=	s_copysignl.c s_fabsl.c s_llrintl.c s_lrintl.c s_modfl.c
 .if ${LDBL_PREC} != 53
 # If long double != double use these; otherwise, we alias the double versions.
-COMMON_SRCS+=	b_tgammal.c catrigl.c \
+.PATH: ${SRCTOP}/contrib/arm-optimized-routines/math
+COMMON_SRCS+=	b_tgammal.c tgamma128.c catrigl.c \
 	e_acoshl.c e_acosl.c e_asinl.c e_atan2l.c e_atanhl.c \
 	e_coshl.c e_fmodl.c e_hypotl.c \
 	e_lgammal.c e_lgammal_r.c e_powl.c \
diff --git a/lib/msun/ld128/b_tgammal.c b/lib/msun/ld128/b_tgammal.c
--- a/lib/msun/ld128/b_tgammal.c
+++ b/lib/msun/ld128/b_tgammal.c
@@ -1,8 +1,7 @@
 /*-
  * SPDX-License-Identifier: BSD-2-Clause
  *
- * Copyright (c) 2013 David Chisnall
- * All rights reserved.
+ * Copyright (c) 2024 Mark Murray
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
@@ -26,30 +25,15 @@
  * SUCH DAMAGE.
  */
 
-#include <float.h>
-#include <math.h>
-
-/*
- * If long double is not the same size as double, then these will lose
- * precision and we should emit a warning whenever something links against
- * them.
- */
-#if (LDBL_MANT_DIG > 53)
-#define WARN_IMPRECISE(x) \
-	__warn_references(x, # x " has lower than advertised precision");
-#else
-#define WARN_IMPRECISE(x)
-#endif
 /*
- * Declare the functions as weak variants so that other libraries providing
- * real versions can override them.
+ * This is a pure C function generously donated by ARM.
+ * See src/contrib/arm-optimized-routines/math/tgamma128.[ch].
  */
-#define	DECLARE_WEAK(x)\
-	__weak_reference(imprecise_## x, x);\
-	WARN_IMPRECISE(x)
+long double tgamma128(long double x);
 
-#define DECLARE_IMPRECISE(f) \
-	long double imprecise_ ## f ## l(long double v) { return f(v); }\
-	DECLARE_WEAK(f ## l)
+long double
+tgammal(long double x)
+{
 
-DECLARE_IMPRECISE(tgamma);
+	return tgamma128(x);
+}
diff --git a/lib/msun/man/lgamma.3 b/lib/msun/man/lgamma.3
--- a/lib/msun/man/lgamma.3
+++ b/lib/msun/man/lgamma.3
@@ -25,7 +25,7 @@
 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 .\" SUCH DAMAGE.
 .\"
-.Dd December 8, 2017
+.Dd March 1, 2024
 .Dt LGAMMA 3
 .Os
 .Sh NAME
@@ -167,15 +167,6 @@
 For large non-integer negative values,
 .Fn tgamma
 will underflow.
-.Sh BUGS
-To conform with newer C/C++ standards, a stub implementation for
-.Nm tgammal
-was committed to the math library, where
-.Nm tgammal
-is mapped to
-.Nm tgamma .
-Thus, the numerical accuracy is at most that of the 53-bit double
-precision implementation.
 .Sh SEE ALSO
 .Xr math 3
 .Sh STANDARDS
@@ -212,3 +203,11 @@
 .Fn tgamma
 function appeared in
 .Fx 5.0 .
+The 128-bit
+.Ft long double
+version of
+.Fn tgammal
+replaced the 80-bit stub version in
+version in
+.Fx 16 ,
+thanks to an appropriate implementation from ARM.