altivec.md (p8_vmrgow): Add support for V2DI, V2DF, V4SI, V4SF types.

gcc/ChangeLog:

2018-01-09  Carl Love  <cel@us.ibm.com>

	* config/rs6002/altivec.md (p8_vmrgow): Add support for V2DI, V2DF,
	V4SI, V4SF types.
	(p8_vmrgew): Add support for V2DI, V2DF, V4SF types.
	* config/rs6000/rs6000-builtin.def: Add definitions for FLOAT2_V2DF,
	VMRGEW_V2DI, VMRGEW_V2DF, VMRGEW_V4SF, VMRGOW_V4SI, VMRGOW_V4SF,
	VMRGOW_V2DI, VMRGOW_V2DF.  Remove definition for VMRGOW.
	* config/rs6000/rs6000-c.c (VSX_BUILTIN_VEC_FLOAT2,
	P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VEC_VMRGOW):  Add definitions.
	* config/rs6000/rs6000-protos.h: Add extern defition for
	rs6000_generate_float2_double_code.
	* config/rs6000/rs6000.c (rs6000_generate_float2_double_code): Add
	function.
	* config/rs6000/vsx.md (vsx_xvcdpsp): Add define_insn.
	(float2_v2df): Add define_expand.

gcc/testsuite/ChangeLog:

2017-01-09 Carl Love  <cel@us.ibm.com>

	* gcc.target/powerpc/builtins-1.c (main): Add tests for vec_mergee and
	vec_mergeo builtins with float, double, long long, unsigned long long,
	bool long long  arguments.
	* gcc.target/powerpc/builtins-3-runnable.c (main): Add test for
	vec_float2 with double arguments.
	* gcc.target/powerpc/builtins-mergew-mergow.c: New runable test for the
	vec_mergew and vec_mergow builtins.

From-SVN: r256395

11 files changed, 453 insertions, 14 deletions

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 60c5704..fee06da 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,20 @@
+2018-01-09  Carl Love  <cel@us.ibm.com>
+
+	* config/rs6002/altivec.md (p8_vmrgow): Add support for V2DI, V2DF,
+	V4SI, V4SF types.
+	(p8_vmrgew): Add support for V2DI, V2DF, V4SF types.
+	* config/rs6000/rs6000-builtin.def: Add definitions for FLOAT2_V2DF,
+	VMRGEW_V2DI, VMRGEW_V2DF, VMRGEW_V4SF, VMRGOW_V4SI, VMRGOW_V4SF,
+	VMRGOW_V2DI, VMRGOW_V2DF.  Remove definition for VMRGOW.
+	* config/rs6000/rs6000-c.c (VSX_BUILTIN_VEC_FLOAT2,
+	P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VEC_VMRGOW):  Add definitions.
+	* config/rs6000/rs6000-protos.h: Add extern defition for
+	rs6000_generate_float2_double_code.
+	* config/rs6000/rs6000.c (rs6000_generate_float2_double_code): Add
+	function.
+	* config/rs6000/vsx.md (vsx_xvcdpsp): Add define_insn.
+	(float2_v2df): Add define_expand.
+
 2018-01-09  Uros Bizjak  <ubizjak@gmail.com>
 
 	PR target/83628
diff --git a/gcc/config/rs6000/altivec.md b/gcc/config/rs6000/altivec.md
index 4f8eeea..733d920 100644
--- a/gcc/config/rs6000/altivec.md
+++ b/gcc/config/rs6000/altivec.md
@@ -1318,6 +1318,24 @@
 }
   [(set_attr "type" "vecperm")])
 
+;; Power8 vector merge two V2DF/V2DI even words to V2DF
+(define_expand "p8_vmrgew_<mode>"
+  [(use (match_operand:VSX_D 0 "vsx_register_operand"))
+   (use (match_operand:VSX_D 1 "vsx_register_operand"))
+   (use (match_operand:VSX_D 2 "vsx_register_operand"))]
+  "VECTOR_MEM_VSX_P (<MODE>mode)"
+{
+  rtvec v;
+  rtx x;
+
+  v = gen_rtvec (2, GEN_INT (0), GEN_INT (2));
+  x = gen_rtx_VEC_CONCAT (<VS_double>mode, operands[1], operands[2]);
+
+  x = gen_rtx_VEC_SELECT (<MODE>mode, x, gen_rtx_PARALLEL (VOIDmode, v));
+  emit_insn (gen_rtx_SET (operands[0], x));
+  DONE;
+})
+
 ;; Power8 vector merge two V4SF/V4SI even words to V4SF
 (define_insn "p8_vmrgew_<mode>"
   [(set (match_operand:VSX_W 0 "register_operand" "=v")
@@ -1336,12 +1354,12 @@
 }
   [(set_attr "type" "vecperm")])
 
-(define_insn "p8_vmrgow"
-  [(set (match_operand:V4SI 0 "register_operand" "=v")
-	(vec_select:V4SI
-	  (vec_concat:V8SI
-	    (match_operand:V4SI 1 "register_operand" "v")
-	    (match_operand:V4SI 2 "register_operand" "v"))
+(define_insn "p8_vmrgow_<mode>"
+  [(set (match_operand:VSX_W 0 "register_operand" "=v")
+	(vec_select:VSX_W
+	  (vec_concat:<VS_double>
+	    (match_operand:VSX_W 1 "register_operand" "v")
+	    (match_operand:VSX_W 2 "register_operand" "v"))
 	  (parallel [(const_int 1) (const_int 5)
 		     (const_int 3) (const_int 7)])))]
   "TARGET_P8_VECTOR"
@@ -1353,6 +1371,23 @@
 }
   [(set_attr "type" "vecperm")])
 
+(define_expand "p8_vmrgow_<mode>"
+  [(use (match_operand:VSX_D 0 "vsx_register_operand"))
+   (use (match_operand:VSX_D 1 "vsx_register_operand"))
+   (use (match_operand:VSX_D 2 "vsx_register_operand"))]
+  "VECTOR_MEM_VSX_P (<MODE>mode)"
+{
+  rtvec v;
+  rtx x;
+
+  v = gen_rtvec (2, GEN_INT (1), GEN_INT (3));
+  x = gen_rtx_VEC_CONCAT (<VS_double>mode, operands[1], operands[2]);
+
+  x = gen_rtx_VEC_SELECT (<MODE>mode, x, gen_rtx_PARALLEL (VOIDmode, v));
+  emit_insn (gen_rtx_SET (operands[0], x));
+  DONE;
+})
+
 (define_insn "p8_vmrgew_<mode>_direct"
   [(set (match_operand:VSX_W 0 "register_operand" "=v")
 	(unspec:VSX_W [(match_operand:VSX_W 1 "register_operand" "v")
diff --git a/gcc/config/rs6000/rs6000-builtin.def b/gcc/config/rs6000/rs6000-builtin.def
index 0f7da6a..06c0c11 100644
--- a/gcc/config/rs6000/rs6000-builtin.def
+++ b/gcc/config/rs6000/rs6000-builtin.def
@@ -1650,6 +1650,7 @@ BU_VSX_2 (CMPLE_U16QI,        "cmple_u16qi",    CONST,  vector_ngtuv16qi)
 BU_VSX_2 (CMPLE_U8HI,         "cmple_u8hi",     CONST,  vector_ngtuv8hi)
 BU_VSX_2 (CMPLE_U4SI,         "cmple_u4si",     CONST,  vector_ngtuv4si)
 BU_VSX_2 (CMPLE_U2DI,         "cmple_u2di",     CONST,  vector_ngtuv2di)
+BU_VSX_2 (FLOAT2_V2DF,        "float2_v2df",    CONST,  float2_v2df)
 BU_VSX_2 (FLOAT2_V2DI,        "float2_v2di",    CONST,  float2_v2di)
 BU_VSX_2 (UNS_FLOAT2_V2DI,    "uns_float2_v2di",    CONST,  uns_float2_v2di)
 
@@ -1924,8 +1925,14 @@ BU_P8V_AV_2 (VMINSD,		"vminsd",	CONST,	sminv2di3)
 BU_P8V_AV_2 (VMAXSD,		"vmaxsd",	CONST,	smaxv2di3)
 BU_P8V_AV_2 (VMINUD,		"vminud",	CONST,	uminv2di3)
 BU_P8V_AV_2 (VMAXUD,		"vmaxud",	CONST,	umaxv2di3)
+BU_P8V_AV_2 (VMRGEW_V2DI,	"vmrgew_v2di",	CONST,	p8_vmrgew_v2di)
+BU_P8V_AV_2 (VMRGEW_V2DF,	"vmrgew_v2df",	CONST,	p8_vmrgew_v2df)
 BU_P8V_AV_2 (VMRGEW_V4SI,	"vmrgew_v4si",	CONST,	p8_vmrgew_v4si)
-BU_P8V_AV_2 (VMRGOW,		"vmrgow",	CONST,	p8_vmrgow)
+BU_P8V_AV_2 (VMRGEW_V4SF,	"vmrgew_v4sf",	CONST,	p8_vmrgew_v4sf)
+BU_P8V_AV_2 (VMRGOW_V4SI,	"vmrgow_v4si",	CONST,	p8_vmrgow_v4si)
+BU_P8V_AV_2 (VMRGOW_V4SF,	"vmrgow_v4sf",	CONST,	p8_vmrgow_v4sf)
+BU_P8V_AV_2 (VMRGOW_V2DI,	"vmrgow_v2di",	CONST,	p8_vmrgow_v2di)
+BU_P8V_AV_2 (VMRGOW_V2DF,	"vmrgow_v2df",	CONST,	p8_vmrgow_v2df)
 BU_P8V_AV_2 (VBPERMQ,		"vbpermq",	CONST,	altivec_vbpermq)
 BU_P8V_AV_2 (VBPERMQ2,		"vbpermq2",	CONST,	altivec_vbpermq2)
 BU_P8V_AV_2 (VPKUDUM,		"vpkudum",	CONST,	altivec_vpkudum)
diff --git a/gcc/config/rs6000/rs6000-c.c b/gcc/config/rs6000/rs6000-c.c
index 123e46a..7a6954a 100644
--- a/gcc/config/rs6000/rs6000-c.c
+++ b/gcc/config/rs6000/rs6000-c.c
@@ -1531,6 +1531,8 @@ const struct altivec_builtin_types altivec_overloaded_builtins[] = {
     RS6000_BTI_V4SF, RS6000_BTI_V4SI, 0, 0 },
   { VSX_BUILTIN_VEC_FLOAT, VSX_BUILTIN_XVCVUXWSP_V4SF,
     RS6000_BTI_V4SF, RS6000_BTI_unsigned_V4SI, 0, 0 },
+  { VSX_BUILTIN_VEC_FLOAT2, VSX_BUILTIN_FLOAT2_V2DF,
+    RS6000_BTI_V4SF, RS6000_BTI_V2DF, RS6000_BTI_V2DF, 0 },
   { VSX_BUILTIN_VEC_FLOAT2, VSX_BUILTIN_FLOAT2_V2DI,
     RS6000_BTI_V4SF, RS6000_BTI_V2DI, RS6000_BTI_V2DI, 0 },
   { VSX_BUILTIN_VEC_FLOAT2, VSX_BUILTIN_UNS_FLOAT2_V2DI,
@@ -5520,6 +5522,17 @@ const struct altivec_builtin_types altivec_overloaded_builtins[] = {
     RS6000_BTI_unsigned_V2DI, RS6000_BTI_unsigned_V2DI,
     RS6000_BTI_unsigned_V2DI, 0 },
 
+  { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V2DI,
+    RS6000_BTI_V2DI, RS6000_BTI_V2DI, RS6000_BTI_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V2DI,
+    RS6000_BTI_unsigned_V2DI, RS6000_BTI_unsigned_V2DI,
+    RS6000_BTI_unsigned_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V2DI,
+    RS6000_BTI_bool_V2DI, RS6000_BTI_bool_V2DI, RS6000_BTI_bool_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V4SF,
+    RS6000_BTI_V4SF, RS6000_BTI_V4SF, RS6000_BTI_V4SF, 0 },
+  { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V2DF,
+    RS6000_BTI_V2DF, RS6000_BTI_V2DF, RS6000_BTI_V2DF, 0 },
   { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V4SI,
     RS6000_BTI_V4SI, RS6000_BTI_V4SI, RS6000_BTI_V4SI, 0 },
   { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V4SI,
@@ -5528,13 +5541,26 @@ const struct altivec_builtin_types altivec_overloaded_builtins[] = {
   { P8V_BUILTIN_VEC_VMRGEW, P8V_BUILTIN_VMRGEW_V4SI,
     RS6000_BTI_bool_V4SI, RS6000_BTI_bool_V4SI, RS6000_BTI_bool_V4SI, 0 },
 
-  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW,
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V4SI,
     RS6000_BTI_V4SI, RS6000_BTI_V4SI, RS6000_BTI_V4SI, 0 },
-  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW,
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V4SI,
     RS6000_BTI_unsigned_V4SI, RS6000_BTI_unsigned_V4SI,
     RS6000_BTI_unsigned_V4SI, 0 },
-  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW,
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V4SI,
     RS6000_BTI_bool_V4SI, RS6000_BTI_bool_V4SI, RS6000_BTI_bool_V4SI, 0 },
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V2DI,
+    RS6000_BTI_V2DI, RS6000_BTI_V2DI, RS6000_BTI_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V2DI,
+    RS6000_BTI_V2DI, RS6000_BTI_V2DI, RS6000_BTI_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V2DI,
+    RS6000_BTI_unsigned_V2DI, RS6000_BTI_unsigned_V2DI,
+    RS6000_BTI_unsigned_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V2DI,
+    RS6000_BTI_bool_V2DI, RS6000_BTI_bool_V2DI, RS6000_BTI_bool_V2DI, 0 },
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V2DF,
+    RS6000_BTI_V2DF, RS6000_BTI_V2DF, RS6000_BTI_V2DF, 0 },
+  { P8V_BUILTIN_VEC_VMRGOW, P8V_BUILTIN_VMRGOW_V4SF,
+    RS6000_BTI_V4SF, RS6000_BTI_V4SF, RS6000_BTI_V4SF, 0 },
 
   { P8V_BUILTIN_VEC_VPMSUM, P8V_BUILTIN_VPMSUMB,
     RS6000_BTI_unsigned_V8HI, RS6000_BTI_unsigned_V16QI,
diff --git a/gcc/config/rs6000/rs6000-protos.h b/gcc/config/rs6000/rs6000-protos.h
index 83487eb..a6c9db9 100644
--- a/gcc/config/rs6000/rs6000-protos.h
+++ b/gcc/config/rs6000/rs6000-protos.h
@@ -71,6 +71,7 @@ extern void rs6000_expand_extract_even (rtx, rtx, rtx);
 extern void rs6000_expand_interleave (rtx, rtx, rtx, bool);
 extern void rs6000_scale_v2df (rtx, rtx, int);
 extern void rs6000_generate_float2_code (bool, rtx, rtx, rtx);
+extern void rs6000_generate_float2_double_code (rtx, rtx, rtx);
 extern void rs6000_generate_vsigned2_code (bool, rtx, rtx, rtx);
 extern int expand_block_clear (rtx[]);
 extern int expand_block_move (rtx[]);
diff --git a/gcc/config/rs6000/rs6000.c b/gcc/config/rs6000/rs6000.c
index f5b7f97..a6f3e4e 100644
--- a/gcc/config/rs6000/rs6000.c
+++ b/gcc/config/rs6000/rs6000.c
@@ -39440,6 +39440,43 @@ rs6000_atomic_assign_expand_fenv (tree *hold, tree *clear, tree *update)
 }
 
 void
+rs6000_generate_float2_double_code (rtx dst, rtx src1, rtx src2)
+{
+  rtx rtx_tmp0, rtx_tmp1, rtx_tmp2, rtx_tmp3;
+
+  rtx_tmp0 = gen_reg_rtx (V2DFmode);
+  rtx_tmp1 = gen_reg_rtx (V2DFmode);
+
+  /* The destination of the vmrgew instruction layout is:
+     rtx_tmp2[0] rtx_tmp3[0] rtx_tmp2[1] rtx_tmp3[0].
+     Setup rtx_tmp0 and rtx_tmp1 to ensure the order of the elements after the
+     vmrgew instruction will be correct.  */
+  if (VECTOR_ELT_ORDER_BIG)
+    {
+       emit_insn (gen_vsx_xxpermdi_v2df_be (rtx_tmp0, src1, src2,
+					    GEN_INT (0)));
+       emit_insn (gen_vsx_xxpermdi_v2df_be (rtx_tmp1, src1, src2,
+					    GEN_INT (3)));
+    }
+  else
+    {
+       emit_insn (gen_vsx_xxpermdi_v2df (rtx_tmp0, src1, src2, GEN_INT (3)));
+       emit_insn (gen_vsx_xxpermdi_v2df (rtx_tmp1, src1, src2, GEN_INT (0)));
+    }
+
+  rtx_tmp2 = gen_reg_rtx (V4SFmode);
+  rtx_tmp3 = gen_reg_rtx (V4SFmode);
+
+  emit_insn (gen_vsx_xvcdpsp (rtx_tmp2, rtx_tmp0));
+  emit_insn (gen_vsx_xvcdpsp (rtx_tmp3, rtx_tmp1));
+
+  if (VECTOR_ELT_ORDER_BIG)
+    emit_insn (gen_p8_vmrgew_v4sf (dst, rtx_tmp2, rtx_tmp3));
+  else
+    emit_insn (gen_p8_vmrgew_v4sf (dst, rtx_tmp3, rtx_tmp2));
+}
+
+void
 rs6000_generate_float2_code (bool signed_convert, rtx dst, rtx src1, rtx src2)
 {
   rtx rtx_tmp0, rtx_tmp1, rtx_tmp2, rtx_tmp3;
diff --git a/gcc/config/rs6000/vsx.md b/gcc/config/rs6000/vsx.md
index 0323e86..effa6c1 100644
--- a/gcc/config/rs6000/vsx.md
+++ b/gcc/config/rs6000/vsx.md
@@ -372,6 +372,7 @@
    UNSPEC_VSX_XVCVSXDDP
    UNSPEC_VSX_XVCVUXDDP
    UNSPEC_VSX_XVCVDPSXDS
+   UNSPEC_VSX_XVCDPSP
    UNSPEC_VSX_XVCVDPUXDS
    UNSPEC_VSX_SIGN_EXTEND
    UNSPEC_VSX_XVCVSPSXWS
@@ -2171,6 +2172,14 @@
   "xvcvuxdsp %x0,%x1"
   [(set_attr "type" "vecdouble")])
 
+(define_insn "vsx_xvcdpsp"
+  [(set (match_operand:V4SF 0 "vsx_register_operand" "=wd,?wa")
+	(unspec:V4SF [(match_operand:V2DF 1 "vsx_register_operand" "wf,wa")]
+		     UNSPEC_VSX_XVCDPSP))]
+  "VECTOR_UNIT_VSX_P (V2DFmode)"
+  "xvcvdpsp %x0,%x1"
+  [(set_attr "type" "vecdouble")])
+
 ;; Convert from 32-bit to 64-bit types
 ;; Provide both vector and scalar targets
 (define_insn "vsx_xvcvsxwdp"
@@ -2237,6 +2246,24 @@
   "xvcvuxwsp %x0,%x1"
   [(set_attr "type" "vecfloat")])
 
+;; Generate float2 double
+;; convert two double to float
+(define_expand "float2_v2df"
+  [(use (match_operand:V4SF 0 "register_operand" "=wa"))
+   (use (match_operand:V2DF 1 "register_operand" "wa"))
+   (use (match_operand:V2DF 2 "register_operand" "wa"))]
+ "VECTOR_UNIT_VSX_P (V4SFmode)"
+{
+  rtx rtx_src1, rtx_src2, rtx_dst;
+
+  rtx_dst = operands[0];
+  rtx_src1 = operands[1];
+  rtx_src2 = operands[2];
+
+  rs6000_generate_float2_double_code (rtx_dst, rtx_src1, rtx_src2);
+  DONE;
+})
+
 ;; Generate float2
 ;; convert two long long signed ints to float
 (define_expand "float2_v2di"
diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index 6357b53..4515cfa 100644
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,13 @@
+2017-01-09 Carl Love  <cel@us.ibm.com>
+
+	* gcc.target/powerpc/builtins-1.c (main): Add tests for vec_mergee and
+	vec_mergeo builtins with float, double, long long, unsigned long long,
+	bool long long  arguments.
+	* gcc.target/powerpc/builtins-3-runnable.c (main): Add test for
+	vec_float2 with double arguments.
+	* gcc.target/powerpc/builtins-mergew-mergow.c: New runable test for the
+	vec_mergew and vec_mergow builtins.
+
 2018-01-09  Steven G. Kargl  <kargl@gcc.gnu.org>
 
 	PR fortran/83742
diff --git a/gcc/testsuite/gcc.target/powerpc/builtins-1.h b/gcc/testsuite/gcc.target/powerpc/builtins-1.h
index 89efedb..b7d8c63 100644
--- a/gcc/testsuite/gcc.target/powerpc/builtins-1.h
+++ b/gcc/testsuite/gcc.target/powerpc/builtins-1.h
@@ -1,6 +1,6 @@
 /* This test is included into builtins-1-be.c and builtins-1-le.c to test on
    Big Endian and Little Endian machines.  */
-   
+
 #include <altivec.h>
 
 vector double y = { 2.0, 4.0 };
@@ -14,7 +14,7 @@ int main ()
   vector float fc = vec_cpsgn (fa, fb);
   vector float fe = vec_mergeh (fa, fb);
   vector float ff = vec_mergel (fa, fb);
-	  
+
   vector double da = {1.0, 2.0};
   vector double db = {-2.0, -3.0};
   vector double dz = vec_and (da, db);
@@ -36,7 +36,7 @@ int main ()
   vector long long lg = vec_andc (la, lb);
   vector long long lh = vec_andc (la, ld);
   vector long long li = vec_andc (ld, lb);
-  
+
   vector unsigned long long ug = vec_andc (ua, ub);
   vector unsigned long long uh = vec_andc (ua, ud);
   vector unsigned long long ui = vec_andc (ud, ub);
@@ -176,7 +176,7 @@ int main ()
   vector unsigned short usc = vec_cntlz (usa);
   vector unsigned short uscd = vec_mergeh (usa, usb);
   vector unsigned short usce = vec_mergel (usa, usb);
-  
+
   vector signed char sca = {-4, 3, -9, 15, -31, 31, 0, 0,
 		            1, 117, -36, 99, 98, 97, 96, 95};
   vector signed char scb = vec_cntlz (sca);
@@ -199,9 +199,15 @@ int main ()
   vector unsigned long long u4 = vec_splat (u2, 1);
   vector bool long long l5 = vec_splat (ld, 0);
   vector bool long long l6 = vec_splat (ld, 1);
+  vector bool long long l10 = vec_mergee (ld, ld);
+  vector bool long long l11 = vec_mergeo (ld, ld);
 
   vector long long l7 = vec_div (l3, l4);
   vector unsigned long long u5 = vec_div (u3, u4);
+  vector long long l12 = vec_mergee (la, lb);
+  vector long long l13 = vec_mergeo (la, lb);
+  vector unsigned long long u8 = vec_mergee (u3, u4);
+  vector unsigned long long u9 = vec_mergeo (u3, u4);
 
   vector long long l8 = vec_mul (l3, l4);
   vector unsigned long long u6 = vec_mul (u3, u4);
@@ -213,5 +219,11 @@ int main ()
   vector unsigned long long u7 = vec_ctu (di, 2);
   vector unsigned int usz = vec_ctu (fa, 0x1F);
 
+  vector float f1 = vec_mergee (fa, fb);
+  vector float f2 = vec_mergeo (fa, fb);
+
+  vector double d1 = vec_mergee (da, db);
+  vector double d2 = vec_mergeo (da, db);
+
   return 0;
 }
diff --git a/gcc/testsuite/gcc.target/powerpc/builtins-3-runnable.c b/gcc/testsuite/gcc.target/powerpc/builtins-3-runnable.c
index 21535b3..1549817 100644
--- a/gcc/testsuite/gcc.target/powerpc/builtins-3-runnable.c
+++ b/gcc/testsuite/gcc.target/powerpc/builtins-3-runnable.c
@@ -266,6 +266,10 @@ int main()
 	vec_flt_result = vec_float2 (vec_ll_uns_int0, vec_ll_uns_int1);
 	test_result_sp(ALL, vec_flt_result, vec_flt_expected);
 
+	vec_flt_expected = (vector float){34.0, 97.0, 214.0, -5.5};
+	vec_flt_result = vec_float2 (vec_dble0, vec_dble1);
+	test_result_sp(ALL, vec_flt_result, vec_flt_expected);
+
 	/* conversion of even words in double precision vector to single precision vector */
 	vec_flt_expected = (vector float){-12.00, 00.00, -12345678901234.00, 0.00};
 	vec_flt_result = vec_floate (vec_ll_int0);
diff --git a/gcc/testsuite/gcc.target/powerpc/builtins-mergew-mergow.c b/gcc/testsuite/gcc.target/powerpc/builtins-mergew-mergow.c
new file mode 100644
index 0000000..24df2cf
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/builtins-mergew-mergow.c
@@ -0,0 +1,263 @@
+/* { dg-do run } */
+/* { dg-require-effective-target vsx_hw } */
+/* { dg-options "-maltivec -mvsx" } */
+
+#include <altivec.h> // vector
+#include <stdlib.h>
+
+#ifdef DEBUG
+#include <stdio.h>
+#endif
+
+void abort (void);
+
+int main() {
+  vector signed int vec_si_arg1, vec_si_arg2;
+  vector signed int vec_si_expected, vec_si_result;
+  vector unsigned int vec_ui_arg1, vec_ui_arg2;
+  vector unsigned int vec_ui_expected, vec_ui_result;
+  vector signed long long vec_sll_arg1, vec_sll_arg2;
+  vector signed long long vec_sll_expected, vec_sll_result;
+  vector unsigned long long vec_ull_arg1, vec_ull_arg2;
+  vector unsigned long long vec_ull_expected, vec_ull_result;
+  vector bool long long vec_bll_arg1, vec_bll_arg2;
+  vector bool long long vec_bll_expected, vec_bll_result;
+  vector float vec_f_arg1, vec_f_arg2;
+  vector float vec_f_expected, vec_f_result;
+  vector double vec_d_arg1, vec_d_arg2;
+  vector double vec_d_expected, vec_d_result;
+  int i;
+  unsigned long long int value;
+
+  /* Tests for vec_mergeo, Merges the odd-numbered halves of two vectors.  */
+  vec_si_arg1 = (vector int){-100, -101, 102, 103};
+  vec_si_arg2 = (vector int){200, 201, 202, 203};
+
+  vec_si_result = vec_mergeo (vec_si_arg1, vec_si_arg2);
+
+  vec_si_expected = (vector int){-101, 201, 103, 203};
+
+  for (i = 0; i < 4; i++)
+    if (vec_si_result[i] != vec_si_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_si_result[%d] = %d, vec_si_expected[%d] = %d\n",
+	     i, vec_si_result[i], i, vec_si_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_ui_arg1 = (vector unsigned int){100, 101, 102, 103};
+  vec_ui_arg2 = (vector unsigned int){200, 201, 202, 203};
+
+  vec_ui_result = vec_mergeo (vec_ui_arg1, vec_ui_arg2);
+
+  vec_ui_expected = (vector unsigned int){101, 201, 103, 203};
+
+  for (i = 0; i < 4; i++)
+    if (vec_ui_result[i] != vec_ui_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_ui_result[%d] = %d, vec_ui_expected[%d] = %d\n",
+	     i, vec_ui_result[i], i, vec_ui_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_sll_arg1 = (vector long long int){-300, -301};
+  vec_sll_arg2 = (vector long long int){400, 401};
+
+  vec_sll_result = vec_mergeo (vec_sll_arg1, vec_sll_arg2);
+
+  vec_sll_expected = (vector long long int){-301, 401};
+
+  for (i = 0; i < 2; i++)
+    if (vec_sll_result[i] != vec_sll_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_sll_result[%d] = %lld, vec_sll_expected[%d] = %lld\n",
+	     i, (long long int)vec_sll_result[i],
+	     i, (long long int)vec_sll_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_ull_arg1 = (vector unsigned long long int){500, 501};
+  vec_ull_arg2 = (vector unsigned long long int){600, 601};
+
+  vec_ull_result = vec_mergeo (vec_ull_arg1, vec_ull_arg2);
+
+  vec_ull_expected = (vector unsigned long long int){501, 601};
+
+  for (i = 0; i < 2; i++)
+    if (vec_ull_result[i] != vec_ull_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_ull_result[%d] = %lld, vec_ull_expected[%d] = %lld\n",
+	     i, (unsigned long long int)vec_ull_result[i],
+	     i, (unsigned long long int)vec_ull_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_bll_arg1 = (vector bool long long){0, 0};
+  vec_bll_arg2 = (vector bool long long){1, 1};
+
+  vec_bll_result = vec_mergeo (vec_bll_arg1, vec_bll_arg2);
+
+  vec_bll_expected = (vector bool long long){0, 1};
+
+  for (i = 0; i < 2; i++)
+    if (vec_bll_result[i] != vec_bll_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_bll_result[%d] = %lld, vec_bll_expected[%d] = %lld\n",
+	     i, vec_ull_result[i],
+	     i, vec_ull_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_f_arg1 = (vector float){100.0, 101.1, 102.2, 103.3};
+  vec_f_arg2 = (vector float){200.0, 201.1, 202.2, 203.3};
+
+  vec_f_result = vec_mergeo (vec_f_arg1, vec_f_arg2);
+
+  vec_f_expected = (vector float){101.1, 201.1, 103.3, 203.3};
+
+  for (i = 0; i < 4; i++)
+    if (vec_f_result[i] != vec_f_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_f_result[%d] = %f, vec_f_expected[%d] = %f\n",
+	     i, vec_f_result[i], i, vec_f_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_d_arg1 = (vector double){300.0, 301.1};
+  vec_d_arg2 = (vector double){400.0, 401.1};
+
+  vec_d_result = vec_mergeo (vec_d_arg1, vec_d_arg2);
+
+  vec_d_expected = (vector double){301.1, 401.1};
+
+  for (i = 0; i < 2; i++)
+    if (vec_d_result[i] != vec_d_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergeo(): vec_d_result[%d] = %f, vec_d_expected[%d] = %f\n",
+	     i, vec_d_result[i], i, vec_d_expected[i]);
+#else
+      abort();
+#endif
+
+  /* Tests for vec_mergee, Merges the even-numbered halves of two vectors.  */
+  vec_si_arg1 = (vector int){-100, -101, 102, 103};
+  vec_si_arg2 = (vector int){200, 201, 202, 203};
+
+  vec_si_result = vec_mergee (vec_si_arg1, vec_si_arg2);
+
+  vec_si_expected = (vector int){-100, 200, 102, 202};
+
+  for (i = 0; i < 4; i++)
+    if (vec_si_result[i] != vec_si_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergee(): vec_si_result[%d] = %d, vec_si_expected[%d] = %d\n",
+	     i, vec_si_result[i], i, vec_si_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_ui_arg1 = (vector unsigned int){100, 101, 102, 103};
+  vec_ui_arg2 = (vector unsigned int){200, 201, 202, 203};
+
+  vec_ui_result = vec_mergee (vec_ui_arg1, vec_ui_arg2);
+
+  vec_ui_expected = (vector unsigned int){100, 200, 102, 202};
+
+  for (i = 0; i < 4; i++)
+    if (vec_ui_result[i] != vec_ui_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergee(): vec_ui_result[%d] = %d, vec_ui_expected[%d] = %d\n",
+	     i, vec_ui_result[i], i, vec_ui_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_sll_arg1 = (vector signed long long int){-300, -301};
+  vec_sll_arg2 = (vector signed long long int){400, 401};
+
+  vec_sll_result = vec_mergee (vec_sll_arg1, vec_sll_arg2);
+
+  vec_sll_expected = (vector signed long long int){-300, 400};
+
+  for (i = 0; i < 2; i++)
+    if (vec_sll_result[i] != vec_sll_expected[i])
+#ifdef DEBUG
+	printf("ERROR vec_mergee(): vec_sll_result[%d] = %lld, vec_sll_expected[%d] = %lld\n",
+		i, (signed long long int)vec_sll_result[i],
+		i, (signed long long int)vec_sll_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_ull_arg1 = (vector unsigned long long int){500, 501};
+  vec_ull_arg2 = (vector unsigned long long int){600, 601};
+
+  vec_ull_result = vec_mergee (vec_ull_arg1, vec_ull_arg2);
+
+  vec_ull_expected = (vector unsigned long long int){500, 600};
+
+  for (i = 0; i < 2; i++)
+    if (vec_ull_result[i] != vec_ull_expected[i])
+#ifdef DEBUG
+	printf("ERROR vec_mergee(): vec_ull_result[%d] = %lld, vec_ull_expected[%d] = %lld\n",
+	       i, (unsigned long long int)vec_ull_result[i],
+	       i, (unsigned long long int)vec_ull_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_bll_arg1 = (vector bool long long){0, 0};
+  vec_bll_arg2 = (vector bool long long){1, 1};
+
+  vec_bll_result = vec_mergee (vec_bll_arg1, vec_bll_arg2);
+
+  vec_bll_expected = (vector bool long long){0, 1};
+
+  for (i = 0; i < 2; i++)
+    if (vec_bll_result[i] != vec_bll_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergee(): vec_bll_result[%d] = %lld, vec_bll_expected[%d] = %lld\n",
+	     i, vec_ull_result[i],
+	     i, vec_ull_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_f_arg1 = (vector float){100.0, 101.1, 102.2, 103.3};
+  vec_f_arg2 = (vector float){200.0, 201.1, 202.2, 203.3};
+
+  vec_f_result = vec_mergee (vec_f_arg1, vec_f_arg2);
+
+  vec_f_expected = (vector float){100.0, 200.0, 102.2, 202.2};
+
+  for (i = 0; i < 4; i++)
+    if (vec_f_result[i] != vec_f_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergee(): vec_f_result[%d] = %f, vec_f_expected[%d] = %f\n",
+	     i, vec_f_result[i], i, vec_f_expected[i]);
+#else
+      abort();
+#endif
+
+  vec_d_arg1 = (vector double){300.0, 301.1};
+  vec_d_arg2 = (vector double){400.0, 401.1};
+
+  vec_d_result = vec_mergee (vec_d_arg1, vec_d_arg2);
+
+  vec_d_expected = (vector double){300.0, 400.0};
+
+  for (i = 0; i < 2; i++)
+    if (vec_d_result[i] != vec_d_expected[i])
+#ifdef DEBUG
+      printf("ERROR vec_mergee(): vec_d_result[%d] = %f, vec_d_expected[%d] = %f\n",
+	     i, vec_d_result[i], i, vec_d_expected[i]);
+#else
+      abort();
+#endif
+}