+2008-03-15 Uros Bizjak <ubizjak@gmail.com>
+
+ * config/i386/mmx.md ("sse2_umulv1siv2di3"): Rename from
+ "sse2_umulsidi3". Use V1DI mode for operand 0.
+ ("mmx_psadbw"): Use V1DI mode for operand 0.
+ * config/i386/i386-modes.def (V1SI): New vector mode.
+ * config/i386/i386.c (struct builtin_description)
+ [IX86_BUILTIN_PMULUDQ]: Use CODE_FOR_sse2_umulv1siv1di3.
+ (v1di_ftype_v8qi_v8qi): Rename from di_ftype_v8qi_v8qi.
+ (v1di_ftype_v2si_v2si): Rename from di_ftype_v2si_v2si.
+ (ix86_init_mmx_sse_builtins) [__builtin_ia32_psadbw]: Use
+ v1di_ftype_v8qi_v8qi type.
+ [__builtin_ia32_pmuludq]: Use v1di_ftype_v2si_v2si type.
+
+ * doc/extend.texi (X86 Built-in Functions) [__builtin_ia32_psadbw,
+ __builtin_ia32_pmuludq]: Fix the mode of return value.
+
2008-03-15 Richard Guenther <rguenther@suse.de>
PR middle-end/35595
{ OPTION_MASK_ISA_SSE2, CODE_FOR_umulv8hi3_highpart, "__builtin_ia32_pmulhuw128", IX86_BUILTIN_PMULHUW128, UNKNOWN, 0 },
{ OPTION_MASK_ISA_SSE2, CODE_FOR_sse2_psadbw, 0, IX86_BUILTIN_PSADBW128, UNKNOWN, 0 },
- { OPTION_MASK_ISA_SSE2, CODE_FOR_sse2_umulsidi3, 0, IX86_BUILTIN_PMULUDQ, UNKNOWN, 0 },
+ { OPTION_MASK_ISA_SSE2, CODE_FOR_sse2_umulv1siv1di3, 0, IX86_BUILTIN_PMULUDQ, UNKNOWN, 0 },
{ OPTION_MASK_ISA_SSE2, CODE_FOR_sse2_umulv2siv2di3, 0, IX86_BUILTIN_PMULUDQ128, UNKNOWN, 0 },
{ OPTION_MASK_ISA_SSE2, CODE_FOR_sse2_pmaddwd, 0, IX86_BUILTIN_PMADDWD128, UNKNOWN, 0 },
tree v4si_ftype_v8hi_v8hi
= build_function_type_list (V4SI_type_node,
V8HI_type_node, V8HI_type_node, NULL_TREE);
- tree di_ftype_v8qi_v8qi
- = build_function_type_list (long_long_unsigned_type_node,
+ tree v1di_ftype_v8qi_v8qi
+ = build_function_type_list (V1DI_type_node,
V8QI_type_node, V8QI_type_node, NULL_TREE);
- tree di_ftype_v2si_v2si
- = build_function_type_list (long_long_unsigned_type_node,
+ tree v1di_ftype_v2si_v2si
+ = build_function_type_list (V1DI_type_node,
V2SI_type_node, V2SI_type_node, NULL_TREE);
tree v2di_ftype_v16qi_v16qi
= build_function_type_list (V2DI_type_node,
def_builtin (OPTION_MASK_ISA_SSE | OPTION_MASK_ISA_3DNOW_A, "__builtin_ia32_sfence", void_ftype_void, IX86_BUILTIN_SFENCE);
- def_builtin_const (OPTION_MASK_ISA_SSE | OPTION_MASK_ISA_3DNOW_A, "__builtin_ia32_psadbw", di_ftype_v8qi_v8qi, IX86_BUILTIN_PSADBW);
+ def_builtin_const (OPTION_MASK_ISA_SSE | OPTION_MASK_ISA_3DNOW_A, "__builtin_ia32_psadbw", v1di_ftype_v8qi_v8qi, IX86_BUILTIN_PSADBW);
def_builtin_const (OPTION_MASK_ISA_SSE, "__builtin_ia32_rcpps", v4sf_ftype_v4sf, IX86_BUILTIN_RCPPS);
def_builtin_const (OPTION_MASK_ISA_SSE, "__builtin_ia32_rcpss", v4sf_ftype_v4sf, IX86_BUILTIN_RCPSS);
def_builtin (OPTION_MASK_ISA_SSE2, "__builtin_ia32_loaddqu", v16qi_ftype_pcchar, IX86_BUILTIN_LOADDQU);
def_builtin (OPTION_MASK_ISA_SSE2, "__builtin_ia32_storedqu", void_ftype_pchar_v16qi, IX86_BUILTIN_STOREDQU);
- def_builtin_const (OPTION_MASK_ISA_SSE2, "__builtin_ia32_pmuludq", di_ftype_v2si_v2si, IX86_BUILTIN_PMULUDQ);
+ def_builtin_const (OPTION_MASK_ISA_SSE2, "__builtin_ia32_pmuludq", v1di_ftype_v2si_v2si, IX86_BUILTIN_PMULUDQ);
def_builtin_const (OPTION_MASK_ISA_SSE2, "__builtin_ia32_pmuludq128", v2di_ftype_v4si_v4si, IX86_BUILTIN_PMULUDQ128);
def_builtin_const (OPTION_MASK_ISA_SSE2, "__builtin_ia32_pslldqi128", v2di_ftype_v2di_int, IX86_BUILTIN_PSLLDQI128);
[(set_attr "type" "mmxmul")
(set_attr "mode" "DI")])
-(define_insn "sse2_umulsidi3"
- [(set (match_operand:DI 0 "register_operand" "=y")
- (mult:DI
- (zero_extend:DI
- (vec_select:SI
+(define_insn "sse2_umulv1siv1di3"
+ [(set (match_operand:V1DI 0 "register_operand" "=y")
+ (mult:V1DI
+ (zero_extend:V1DI
+ (vec_select:V1SI
(match_operand:V2SI 1 "nonimmediate_operand" "%0")
(parallel [(const_int 0)])))
- (zero_extend:DI
- (vec_select:SI
+ (zero_extend:V1DI
+ (vec_select:V1SI
(match_operand:V2SI 2 "nonimmediate_operand" "ym")
(parallel [(const_int 0)])))))]
"TARGET_SSE2 && ix86_binary_operator_ok (MULT, V2SImode, operands)"
(set_attr "mode" "DI")])
(define_insn "mmx_psadbw"
- [(set (match_operand:DI 0 "register_operand" "=y")
- (unspec:DI [(match_operand:V8QI 1 "register_operand" "0")
- (match_operand:V8QI 2 "nonimmediate_operand" "ym")]
- UNSPEC_PSADBW))]
+ [(set (match_operand:V1DI 0 "register_operand" "=y")
+ (unspec:V1DI [(match_operand:V8QI 1 "register_operand" "0")
+ (match_operand:V8QI 2 "nonimmediate_operand" "ym")]
+ UNSPEC_PSADBW))]
"TARGET_SSE || TARGET_3DNOW_A"
"psadbw\t{%2, %0|%0, %2}"
[(set_attr "type" "mmxshft")
v4hi __builtin_ia32_pmulhuw (v4hi, v4hi)
v8qi __builtin_ia32_pavgb (v8qi, v8qi)
v4hi __builtin_ia32_pavgw (v4hi, v4hi)
-v4hi __builtin_ia32_psadbw (v8qi, v8qi)
+v1di __builtin_ia32_psadbw (v8qi, v8qi)
v8qi __builtin_ia32_pmaxub (v8qi, v8qi)
v4hi __builtin_ia32_pmaxsw (v4hi, v4hi)
v8qi __builtin_ia32_pminub (v8qi, v8qi)
void __builtin_ia32_mfence (void)
v16qi __builtin_ia32_loaddqu (const char *)
void __builtin_ia32_storedqu (char *, v16qi)
-unsigned long long __builtin_ia32_pmuludq (v2si, v2si)
+v1di __builtin_ia32_pmuludq (v2si, v2si)
v2di __builtin_ia32_pmuludq128 (v4si, v4si)
v8hi __builtin_ia32_psllw128 (v8hi, v8hi)
v4si __builtin_ia32_pslld128 (v4si, v4si)