;; GCC machine description for SSE instructions
-;; Copyright (C) 2005
+;; Copyright (C) 2005, 2006
;; Free Software Foundation, Inc.
;;
;; This file is part of GCC.
;;
;; You should have received a copy of the GNU General Public License
;; along with GCC; see the file COPYING. If not, write to
-;; the Free Software Foundation, 59 Temple Place - Suite 330,
-;; Boston, MA 02111-1307, USA.
+;; the Free Software Foundation, 51 Franklin Street, Fifth Floor,
+;; Boston, MA 02110-1301, USA.
;; 16 byte integral modes handled by SSE, minus TImode, which gets
;; Mix-n-match
(define_mode_macro SSEMODE12 [V16QI V8HI])
(define_mode_macro SSEMODE24 [V8HI V4SI])
+(define_mode_macro SSEMODE14 [V16QI V4SI])
(define_mode_macro SSEMODE124 [V16QI V8HI V4SI])
(define_mode_macro SSEMODE248 [V8HI V4SI V2DI])
(define_insn "*mov<mode>_internal"
[(set (match_operand:SSEMODEI 0 "nonimmediate_operand" "=x,x ,m")
- (match_operand:SSEMODEI 1 "vector_move_operand" "C ,xm,x"))]
+ (match_operand:SSEMODEI 1 "nonimmediate_or_sse_const_operand" "C ,xm,x"))]
"TARGET_SSE && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (which_alternative)
{
case 0:
- if (get_attr_mode (insn) == MODE_V4SF)
- return "xorps\t%0, %0";
- else
- return "pxor\t%0, %0";
+ return standard_sse_constant_opcode (insn, operands[1]);
case 1:
case 2:
if (get_attr_mode (insn) == MODE_V4SF)
else
return "movdqa\t{%1, %0|%0, %1}";
default:
- abort ();
+ gcc_unreachable ();
}
}
[(set_attr "type" "sselog1,ssemov,ssemov")
(set (attr "mode")
- (cond [(eq (symbol_ref "TARGET_SSE2") (const_int 0))
- (const_string "V4SF")
-
- (eq_attr "alternative" "0,1")
- (if_then_else
- (ne (symbol_ref "optimize_size")
- (const_int 0))
- (const_string "V4SF")
- (const_string "TI"))
- (eq_attr "alternative" "2")
- (if_then_else
- (ior (ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
- (const_int 0))
- (ne (symbol_ref "optimize_size")
- (const_int 0)))
- (const_string "V4SF")
- (const_string "TI"))]
- (const_string "TI")))])
+ (if_then_else
+ (ior (ior (ne (symbol_ref "optimize_size") (const_int 0))
+ (eq (symbol_ref "TARGET_SSE2") (const_int 0)))
+ (and (eq_attr "alternative" "2")
+ (ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
+ (const_int 0))))
+ (const_string "V4SF")
+ (const_string "TI")))])
(define_expand "movv4sf"
[(set (match_operand:V4SF 0 "nonimmediate_operand" "")
(define_insn "*movv4sf_internal"
[(set (match_operand:V4SF 0 "nonimmediate_operand" "=x,x,m")
- (match_operand:V4SF 1 "vector_move_operand" "C,xm,x"))]
+ (match_operand:V4SF 1 "nonimmediate_or_sse_const_operand" "C,xm,x"))]
"TARGET_SSE"
- "@
- xorps\t%0, %0
- movaps\t{%1, %0|%0, %1}
- movaps\t{%1, %0|%0, %1}"
+{
+ switch (which_alternative)
+ {
+ case 0:
+ return standard_sse_constant_opcode (insn, operands[1]);
+ case 1:
+ case 2:
+ return "movaps\t{%1, %0|%0, %1}";
+ default:
+ abort();
+ }
+}
[(set_attr "type" "sselog1,ssemov,ssemov")
(set_attr "mode" "V4SF")])
(define_insn "*movv2df_internal"
[(set (match_operand:V2DF 0 "nonimmediate_operand" "=x,x,m")
- (match_operand:V2DF 1 "vector_move_operand" "C,xm,x"))]
+ (match_operand:V2DF 1 "nonimmediate_or_sse_const_operand" "C,xm,x"))]
"TARGET_SSE && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
{
switch (which_alternative)
{
case 0:
- if (get_attr_mode (insn) == MODE_V4SF)
- return "xorps\t%0, %0";
- else
- return "xorpd\t%0, %0";
+ return standard_sse_constant_opcode (insn, operands[1]);
case 1:
case 2:
if (get_attr_mode (insn) == MODE_V4SF)
else
return "movapd\t{%1, %0|%0, %1}";
default:
- abort ();
+ gcc_unreachable ();
}
}
[(set_attr "type" "sselog1,ssemov,ssemov")
(set (attr "mode")
- (cond [(eq (symbol_ref "TARGET_SSE2") (const_int 0))
- (const_string "V4SF")
- (eq_attr "alternative" "0,1")
- (if_then_else
- (ne (symbol_ref "optimize_size")
- (const_int 0))
- (const_string "V4SF")
- (const_string "V2DF"))
- (eq_attr "alternative" "2")
- (if_then_else
- (ior (ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
- (const_int 0))
- (ne (symbol_ref "optimize_size")
- (const_int 0)))
- (const_string "V4SF")
- (const_string "V2DF"))]
- (const_string "V2DF")))])
+ (if_then_else
+ (ior (ior (ne (symbol_ref "optimize_size") (const_int 0))
+ (eq (symbol_ref "TARGET_SSE2") (const_int 0)))
+ (and (eq_attr "alternative" "2")
+ (ne (symbol_ref "TARGET_SSE_TYPELESS_STORES")
+ (const_int 0))))
+ (const_string "V4SF")
+ (const_string "V2DF")))])
(define_split
[(set (match_operand:V2DF 0 "register_operand" "")
(define_insn "sse3_lddqu"
[(set (match_operand:V16QI 0 "register_operand" "=x")
(unspec:V16QI [(match_operand:V16QI 1 "memory_operand" "m")]
- UNSPEC_LDQQU))]
+ UNSPEC_LDDQU))]
"TARGET_SSE3"
"lddqu\t{%1, %0|%0, %1}"
[(set_attr "type" "ssecvt")
[(set_attr "type" "sse")
(set_attr "mode" "SF")])
+;; These versions of the min/max patterns implement exactly the operations
+;; min = (op1 < op2 ? op1 : op2)
+;; max = (!(op1 < op2) ? op1 : op2)
+;; Their operands are not commutative, and thus they may be used in the
+;; presence of -0.0 and NaN.
+
+(define_insn "*ieee_sminv4sf3"
+ [(set (match_operand:V4SF 0 "register_operand" "=x")
+ (unspec:V4SF [(match_operand:V4SF 1 "register_operand" "0")
+ (match_operand:V4SF 2 "nonimmediate_operand" "xm")]
+ UNSPEC_IEEE_MIN))]
+ "TARGET_SSE"
+ "minps\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sseadd")
+ (set_attr "mode" "V4SF")])
+
+(define_insn "*ieee_smaxv4sf3"
+ [(set (match_operand:V4SF 0 "register_operand" "=x")
+ (unspec:V4SF [(match_operand:V4SF 1 "register_operand" "0")
+ (match_operand:V4SF 2 "nonimmediate_operand" "xm")]
+ UNSPEC_IEEE_MAX))]
+ "TARGET_SSE"
+ "maxps\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sseadd")
+ (set_attr "mode" "V4SF")])
+
+(define_insn "*ieee_sminv2df3"
+ [(set (match_operand:V2DF 0 "register_operand" "=x")
+ (unspec:V2DF [(match_operand:V2DF 1 "register_operand" "0")
+ (match_operand:V2DF 2 "nonimmediate_operand" "xm")]
+ UNSPEC_IEEE_MIN))]
+ "TARGET_SSE2"
+ "minpd\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sseadd")
+ (set_attr "mode" "V2DF")])
+
+(define_insn "*ieee_smaxv2df3"
+ [(set (match_operand:V2DF 0 "register_operand" "=x")
+ (unspec:V2DF [(match_operand:V2DF 1 "register_operand" "0")
+ (match_operand:V2DF 2 "nonimmediate_operand" "xm")]
+ UNSPEC_IEEE_MAX))]
+ "TARGET_SSE2"
+ "maxpd\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sseadd")
+ (set_attr "mode" "V2DF")])
+
(define_insn "sse3_addsubv4sf3"
[(set (match_operand:V4SF 0 "register_operand" "=x")
(vec_merge:V4SF
[(set_attr "type" "sseadd")
(set_attr "mode" "V4SF")])
+(define_expand "reduc_splus_v4sf"
+ [(match_operand:V4SF 0 "register_operand" "")
+ (match_operand:V4SF 1 "register_operand" "")]
+ "TARGET_SSE"
+{
+ if (TARGET_SSE3)
+ {
+ rtx tmp = gen_reg_rtx (V4SFmode);
+ emit_insn (gen_sse3_haddv4sf3 (tmp, operands[1], operands[1]));
+ emit_insn (gen_sse3_haddv4sf3 (operands[0], tmp, tmp));
+ }
+ else
+ ix86_expand_reduc_v4sf (gen_addv4sf3, operands[0], operands[1]);
+ DONE;
+})
+
+(define_expand "reduc_smax_v4sf"
+ [(match_operand:V4SF 0 "register_operand" "")
+ (match_operand:V4SF 1 "register_operand" "")]
+ "TARGET_SSE"
+{
+ ix86_expand_reduc_v4sf (gen_smaxv4sf3, operands[0], operands[1]);
+ DONE;
+})
+
+(define_expand "reduc_smin_v4sf"
+ [(match_operand:V4SF 0 "register_operand" "")
+ (match_operand:V4SF 1 "register_operand" "")]
+ "TARGET_SSE"
+{
+ ix86_expand_reduc_v4sf (gen_sminv4sf3, operands[0], operands[1]);
+ DONE;
+})
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel single-precision floating point comparisons
[(set_attr "type" "ssecomi")
(set_attr "mode" "SF")])
+(define_expand "vcondv4sf"
+ [(set (match_operand:V4SF 0 "register_operand" "")
+ (if_then_else:V4SF
+ (match_operator 3 ""
+ [(match_operand:V4SF 4 "nonimmediate_operand" "")
+ (match_operand:V4SF 5 "nonimmediate_operand" "")])
+ (match_operand:V4SF 1 "general_operand" "")
+ (match_operand:V4SF 2 "general_operand" "")))]
+ "TARGET_SSE"
+{
+ if (ix86_expand_fp_vcond (operands))
+ DONE;
+ else
+ FAIL;
+})
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel single-precision floating point logical operations
[(set_attr "type" "sselog")
(set_attr "mode" "V4SF")])
+;; Also define scalar versions. These are used for abs, neg, and
+;; conditional move. Using subregs into vector modes causes register
+;; allocation lossage. These patterns do not allow memory operands
+;; because the native instructions read the full 128-bits.
+
+(define_insn "*andsf3"
+ [(set (match_operand:SF 0 "register_operand" "=x")
+ (and:SF (match_operand:SF 1 "register_operand" "0")
+ (match_operand:SF 2 "register_operand" "x")))]
+ "TARGET_SSE"
+ "andps\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V4SF")])
+
+(define_insn "*nandsf3"
+ [(set (match_operand:SF 0 "register_operand" "=x")
+ (and:SF (not:SF (match_operand:SF 1 "register_operand" "0"))
+ (match_operand:SF 2 "register_operand" "x")))]
+ "TARGET_SSE"
+ "andnps\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V4SF")])
+
+(define_insn "*iorsf3"
+ [(set (match_operand:SF 0 "register_operand" "=x")
+ (ior:SF (match_operand:SF 1 "register_operand" "0")
+ (match_operand:SF 2 "register_operand" "x")))]
+ "TARGET_SSE"
+ "orps\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V4SF")])
+
+(define_insn "*xorsf3"
+ [(set (match_operand:SF 0 "register_operand" "=x")
+ (xor:SF (match_operand:SF 1 "register_operand" "0")
+ (match_operand:SF 2 "register_operand" "x")))]
+ "TARGET_SSE"
+ "xorps\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V4SF")])
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel single-precision floating point conversion operations
"TARGET_SSE"
"cvtps2pi\t{%1, %0|%0, %1}"
[(set_attr "type" "ssecvt")
+ (set_attr "unit" "mmx")
(set_attr "mode" "DI")])
(define_insn "sse_cvttps2pi"
"TARGET_SSE"
"cvttps2pi\t{%1, %0|%0, %1}"
[(set_attr "type" "ssecvt")
+ (set_attr "unit" "mmx")
(set_attr "mode" "SF")])
(define_insn "sse_cvtsi2ss"
[(set (match_operand:V4SF 0 "nonimmediate_operand" "=x,x,m")
(vec_select:V4SF
(vec_concat:V8SF
- (match_operand:V4SF 1 "nonimmediate_operand" " 0,o,x")
- (match_operand:V4SF 2 "nonimmediate_operand" " x,0,0"))
- (parallel [(const_int 4)
- (const_int 5)
+ (match_operand:V4SF 1 "nonimmediate_operand" " 0,0,0")
+ (match_operand:V4SF 2 "nonimmediate_operand" " x,o,x"))
+ (parallel [(const_int 6)
+ (const_int 7)
(const_int 2)
(const_int 3)])))]
"TARGET_SSE && !(MEM_P (operands[1]) && MEM_P (operands[2]))"
"@
movhlps\t{%2, %0|%0, %2}
- movlps\t{%H1, %0|%0, %H1}
- movhps\t{%1, %0|%0, %1}"
+ movlps\t{%H2, %0|%0, %H2}
+ movhps\t{%2, %0|%0, %2}"
[(set_attr "type" "ssemov")
(set_attr "mode" "V4SF,V2SF,V2SF")])
[(set (match_operand:V2SF 0 "register_operand" "=x,x,*y,*y")
(vec_concat:V2SF
(match_operand:SF 1 "nonimmediate_operand" " 0,m, 0, m")
- (match_operand:SF 2 "vector_move_operand" " x,C,*y, C")))]
+ (match_operand:SF 2 "reg_or_0_operand" " x,C,*y, C")))]
"TARGET_SSE"
"@
unpcklps\t{%2, %0|%0, %2}
"TARGET_SSE2"
"sqrtsd\t{%1, %0|%0, %1}"
[(set_attr "type" "sse")
- (set_attr "mode" "SF")])
+ (set_attr "mode" "DF")])
;; ??? For !flag_finite_math_only, the representation with SMIN/SMAX
;; isn't really correct, as those rtl operators aren't defined when
[(set_attr "type" "sseadd")
(set_attr "mode" "V2DF")])
+(define_expand "reduc_splus_v2df"
+ [(match_operand:V2DF 0 "register_operand" "")
+ (match_operand:V2DF 1 "register_operand" "")]
+ "TARGET_SSE3"
+{
+ emit_insn (gen_sse3_haddv2df3 (operands[0], operands[1], operands[1]));
+ DONE;
+})
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel double-precision floating point comparisons
[(set_attr "type" "ssecomi")
(set_attr "mode" "DF")])
+(define_expand "vcondv2df"
+ [(set (match_operand:V2DF 0 "register_operand" "")
+ (if_then_else:V2DF
+ (match_operator 3 ""
+ [(match_operand:V2DF 4 "nonimmediate_operand" "")
+ (match_operand:V2DF 5 "nonimmediate_operand" "")])
+ (match_operand:V2DF 1 "general_operand" "")
+ (match_operand:V2DF 2 "general_operand" "")))]
+ "TARGET_SSE2"
+{
+ if (ix86_expand_fp_vcond (operands))
+ DONE;
+ else
+ FAIL;
+})
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel double-precision floating point logical operations
[(set (match_operand:V2DF 0 "register_operand" "=x")
(and:V2DF (match_operand:V2DF 1 "nonimmediate_operand" "%0")
(match_operand:V2DF 2 "nonimmediate_operand" "xm")))]
- "TARGET_SSE2 && ix86_binary_operator_ok (AND, V4SFmode, operands)"
+ "TARGET_SSE2 && ix86_binary_operator_ok (AND, V2DFmode, operands)"
"andpd\t{%2, %0|%0, %2}"
[(set_attr "type" "sselog")
(set_attr "mode" "V2DF")])
[(set_attr "type" "sselog")
(set_attr "mode" "V2DF")])
+;; Also define scalar versions. These are used for abs, neg, and
+;; conditional move. Using subregs into vector modes causes register
+;; allocation lossage. These patterns do not allow memory operands
+;; because the native instructions read the full 128-bits.
+
+(define_insn "*anddf3"
+ [(set (match_operand:DF 0 "register_operand" "=x")
+ (and:DF (match_operand:DF 1 "register_operand" "0")
+ (match_operand:DF 2 "register_operand" "x")))]
+ "TARGET_SSE2"
+ "andpd\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V2DF")])
+
+(define_insn "*nanddf3"
+ [(set (match_operand:DF 0 "register_operand" "=x")
+ (and:DF (not:DF (match_operand:DF 1 "register_operand" "0"))
+ (match_operand:DF 2 "register_operand" "x")))]
+ "TARGET_SSE2"
+ "andnpd\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V2DF")])
+
+(define_insn "*iordf3"
+ [(set (match_operand:DF 0 "register_operand" "=x")
+ (ior:DF (match_operand:DF 1 "register_operand" "0")
+ (match_operand:DF 2 "register_operand" "x")))]
+ "TARGET_SSE2"
+ "orpd\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V2DF")])
+
+(define_insn "*xordf3"
+ [(set (match_operand:DF 0 "register_operand" "=x")
+ (xor:DF (match_operand:DF 1 "register_operand" "0")
+ (match_operand:DF 2 "register_operand" "x")))]
+ "TARGET_SSE2"
+ "xorpd\t{%2, %0|%0, %2}"
+ [(set_attr "type" "sselog")
+ (set_attr "mode" "V2DF")])
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel double-precision floating point conversion operations
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(define_insn "sse2_cvtpi2pd"
- [(set (match_operand:V2DF 0 "register_operand" "=x")
- (float:V2DF (match_operand:V2SI 1 "nonimmediate_operand" "ym")))]
+ [(set (match_operand:V2DF 0 "register_operand" "=x,x")
+ (float:V2DF (match_operand:V2SI 1 "nonimmediate_operand" "y,m")))]
"TARGET_SSE2"
"cvtpi2pd\t{%1, %0|%0, %1}"
[(set_attr "type" "ssecvt")
+ (set_attr "unit" "mmx,*")
(set_attr "mode" "V2DF")])
(define_insn "sse2_cvtpd2pi"
"TARGET_SSE2"
"cvtpd2pi\t{%1, %0|%0, %1}"
[(set_attr "type" "ssecvt")
+ (set_attr "unit" "mmx")
(set_attr "mode" "DI")])
(define_insn "sse2_cvttpd2pi"
"TARGET_SSE2"
"cvttpd2pi\t{%1, %0|%0, %1}"
[(set_attr "type" "ssecvt")
+ (set_attr "unit" "mmx")
(set_attr "mode" "TI")])
(define_insn "sse2_cvtsi2sd"
operands[0] = adjust_address (operands[0], DFmode, 8);
})
+;; Not sure these two are ever used, but it doesn't hurt to have
+;; them. -aoliva
+(define_insn "*vec_extractv2df_1_sse"
+ [(set (match_operand:DF 0 "nonimmediate_operand" "=m,x,x")
+ (vec_select:DF
+ (match_operand:V2DF 1 "nonimmediate_operand" "x,x,o")
+ (parallel [(const_int 1)])))]
+ "!TARGET_SSE2 && TARGET_SSE
+ && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+ "@
+ movhps\t{%1, %0|%0, %1}
+ movhlps\t{%1, %0|%0, %1}
+ movlps\t{%H1, %0|%0, %H1}"
+ [(set_attr "type" "ssemov")
+ (set_attr "mode" "V2SF,V4SF,V2SF")])
+
+(define_insn "*vec_extractv2df_0_sse"
+ [(set (match_operand:DF 0 "nonimmediate_operand" "=m,x,x")
+ (vec_select:DF
+ (match_operand:V2DF 1 "nonimmediate_operand" "x,x,m")
+ (parallel [(const_int 0)])))]
+ "!TARGET_SSE2 && TARGET_SSE
+ && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+ "@
+ movlps\t{%1, %0|%0, %1}
+ movaps\t{%1, %0|%0, %1}
+ movlps\t{%1, %0|%0, %1}"
+ [(set_attr "type" "ssemov")
+ (set_attr "mode" "V2SF,V4SF,V2SF")])
+
(define_insn "sse2_movsd"
[(set (match_operand:V2DF 0 "nonimmediate_operand" "=x,x,m,x,x,o")
(vec_merge:V2DF
movlpd\t{%2, %0|%0, %2}
movlpd\t{%2, %0|%0, %2}
shufpd\t{$2, %2, %0|%0, %2, 2}
- movhps\t{%H1, %0|%0, %H1
- movhps\t{%1, %H0|%H0, %1"
+ movhps\t{%H1, %0|%0, %H1}
+ movhps\t{%1, %H0|%H0, %1}"
[(set_attr "type" "ssemov,ssemov,ssemov,sselog,ssemov,ssemov")
(set_attr "mode" "DF,V1DF,V1DF,V2DF,V1DF,V1DF")])
[(set_attr "type" "sseiadd")
(set_attr "mode" "TI")])
+(define_expand "mulv16qi3"
+ [(set (match_operand:V16QI 0 "register_operand" "")
+ (mult:V16QI (match_operand:V16QI 1 "register_operand" "")
+ (match_operand:V16QI 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx t[12], op0;
+ int i;
+
+ for (i = 0; i < 12; ++i)
+ t[i] = gen_reg_rtx (V16QImode);
+
+ /* Unpack data such that we've got a source byte in each low byte of
+ each word. We don't care what goes into the high byte of each word.
+ Rather than trying to get zero in there, most convenient is to let
+ it be a copy of the low byte. */
+ emit_insn (gen_sse2_punpckhbw (t[0], operands[1], operands[1]));
+ emit_insn (gen_sse2_punpckhbw (t[1], operands[2], operands[2]));
+ emit_insn (gen_sse2_punpcklbw (t[2], operands[1], operands[1]));
+ emit_insn (gen_sse2_punpcklbw (t[3], operands[2], operands[2]));
+
+ /* Multiply words. The end-of-line annotations here give a picture of what
+ the output of that instruction looks like. Dot means don't care; the
+ letters are the bytes of the result with A being the most significant. */
+ emit_insn (gen_mulv8hi3 (gen_lowpart (V8HImode, t[4]), /* .A.B.C.D.E.F.G.H */
+ gen_lowpart (V8HImode, t[0]),
+ gen_lowpart (V8HImode, t[1])));
+ emit_insn (gen_mulv8hi3 (gen_lowpart (V8HImode, t[5]), /* .I.J.K.L.M.N.O.P */
+ gen_lowpart (V8HImode, t[2]),
+ gen_lowpart (V8HImode, t[3])));
+
+ /* Extract the relevant bytes and merge them back together. */
+ emit_insn (gen_sse2_punpckhbw (t[6], t[5], t[4])); /* ..AI..BJ..CK..DL */
+ emit_insn (gen_sse2_punpcklbw (t[7], t[5], t[4])); /* ..EM..FN..GO..HP */
+ emit_insn (gen_sse2_punpckhbw (t[8], t[7], t[6])); /* ....AEIM....BFJN */
+ emit_insn (gen_sse2_punpcklbw (t[9], t[7], t[6])); /* ....CGKO....DHLP */
+ emit_insn (gen_sse2_punpckhbw (t[10], t[9], t[8])); /* ........ACEGIKMO */
+ emit_insn (gen_sse2_punpcklbw (t[11], t[9], t[8])); /* ........BDFHJLNP */
+
+ op0 = operands[0];
+ emit_insn (gen_sse2_punpcklbw (op0, t[11], t[10])); /* ABCDEFGHIJKLMNOP */
+ DONE;
+})
+
(define_expand "mulv8hi3"
[(set (match_operand:V8HI 0 "register_operand" "")
(mult:V8HI (match_operand:V8HI 1 "nonimmediate_operand" "")
[(set_attr "type" "sseiadd")
(set_attr "mode" "TI")])
+(define_expand "mulv4si3"
+ [(set (match_operand:V4SI 0 "register_operand" "")
+ (mult:V4SI (match_operand:V4SI 1 "register_operand" "")
+ (match_operand:V4SI 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx t1, t2, t3, t4, t5, t6, thirtytwo;
+ rtx op0, op1, op2;
+
+ op0 = operands[0];
+ op1 = operands[1];
+ op2 = operands[2];
+ t1 = gen_reg_rtx (V4SImode);
+ t2 = gen_reg_rtx (V4SImode);
+ t3 = gen_reg_rtx (V4SImode);
+ t4 = gen_reg_rtx (V4SImode);
+ t5 = gen_reg_rtx (V4SImode);
+ t6 = gen_reg_rtx (V4SImode);
+ thirtytwo = GEN_INT (32);
+
+ /* Multiply elements 2 and 0. */
+ emit_insn (gen_sse2_umulv2siv2di3 (gen_lowpart (V2DImode, t1), op1, op2));
+
+ /* Shift both input vectors down one element, so that elements 3 and 1
+ are now in the slots for elements 2 and 0. For K8, at least, this is
+ faster than using a shuffle. */
+ emit_insn (gen_sse2_lshrti3 (gen_lowpart (TImode, t2),
+ gen_lowpart (TImode, op1), thirtytwo));
+ emit_insn (gen_sse2_lshrti3 (gen_lowpart (TImode, t3),
+ gen_lowpart (TImode, op2), thirtytwo));
+
+ /* Multiply elements 3 and 1. */
+ emit_insn (gen_sse2_umulv2siv2di3 (gen_lowpart (V2DImode, t4), t2, t3));
+
+ /* Move the results in element 2 down to element 1; we don't care what
+ goes in elements 2 and 3. */
+ emit_insn (gen_sse2_pshufd_1 (t5, t1, const0_rtx, const2_rtx,
+ const0_rtx, const0_rtx));
+ emit_insn (gen_sse2_pshufd_1 (t6, t4, const0_rtx, const2_rtx,
+ const0_rtx, const0_rtx));
+
+ /* Merge the parts back together. */
+ emit_insn (gen_sse2_punpckldq (op0, t5, t6));
+ DONE;
+})
+
+(define_expand "mulv2di3"
+ [(set (match_operand:V2DI 0 "register_operand" "")
+ (mult:V2DI (match_operand:V2DI 1 "register_operand" "")
+ (match_operand:V2DI 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx t1, t2, t3, t4, t5, t6, thirtytwo;
+ rtx op0, op1, op2;
+
+ op0 = operands[0];
+ op1 = operands[1];
+ op2 = operands[2];
+ t1 = gen_reg_rtx (V2DImode);
+ t2 = gen_reg_rtx (V2DImode);
+ t3 = gen_reg_rtx (V2DImode);
+ t4 = gen_reg_rtx (V2DImode);
+ t5 = gen_reg_rtx (V2DImode);
+ t6 = gen_reg_rtx (V2DImode);
+ thirtytwo = GEN_INT (32);
+
+ /* Multiply low parts. */
+ emit_insn (gen_sse2_umulv2siv2di3 (t1, gen_lowpart (V4SImode, op1),
+ gen_lowpart (V4SImode, op2)));
+
+ /* Shift input vectors left 32 bits so we can multiply high parts. */
+ emit_insn (gen_lshrv2di3 (t2, op1, thirtytwo));
+ emit_insn (gen_lshrv2di3 (t3, op2, thirtytwo));
+
+ /* Multiply high parts by low parts. */
+ emit_insn (gen_sse2_umulv2siv2di3 (t4, gen_lowpart (V4SImode, op1),
+ gen_lowpart (V4SImode, t3)));
+ emit_insn (gen_sse2_umulv2siv2di3 (t5, gen_lowpart (V4SImode, op2),
+ gen_lowpart (V4SImode, t2)));
+
+ /* Shift them back. */
+ emit_insn (gen_ashlv2di3 (t4, t4, thirtytwo));
+ emit_insn (gen_ashlv2di3 (t5, t5, thirtytwo));
+
+ /* Add the three parts together. */
+ emit_insn (gen_addv2di3 (t6, t1, t4));
+ emit_insn (gen_addv2di3 (op0, t6, t5));
+ DONE;
+})
+
+(define_expand "sdot_prodv8hi"
+ [(match_operand:V4SI 0 "register_operand" "")
+ (match_operand:V8HI 1 "nonimmediate_operand" "")
+ (match_operand:V8HI 2 "nonimmediate_operand" "")
+ (match_operand:V4SI 3 "register_operand" "")]
+ "TARGET_SSE2"
+{
+ rtx t = gen_reg_rtx (V4SImode);
+ emit_insn (gen_sse2_pmaddwd (t, operands[1], operands[2]));
+ emit_insn (gen_addv4si3 (operands[0], operands[3], t));
+ DONE;
+})
+
+(define_expand "udot_prodv4si"
+ [(match_operand:V2DI 0 "register_operand" "")
+ (match_operand:V4SI 1 "register_operand" "")
+ (match_operand:V4SI 2 "register_operand" "")
+ (match_operand:V2DI 3 "register_operand" "")]
+ "TARGET_SSE2"
+{
+ rtx t1, t2, t3, t4;
+
+ t1 = gen_reg_rtx (V2DImode);
+ emit_insn (gen_sse2_umulv2siv2di3 (t1, operands[1], operands[2]));
+ emit_insn (gen_addv2di3 (t1, t1, operands[3]));
+
+ t2 = gen_reg_rtx (V4SImode);
+ t3 = gen_reg_rtx (V4SImode);
+ emit_insn (gen_sse2_lshrti3 (gen_lowpart (TImode, t2),
+ gen_lowpart (TImode, operands[1]),
+ GEN_INT (32)));
+ emit_insn (gen_sse2_lshrti3 (gen_lowpart (TImode, t3),
+ gen_lowpart (TImode, operands[2]),
+ GEN_INT (32)));
+
+ t4 = gen_reg_rtx (V2DImode);
+ emit_insn (gen_sse2_umulv2siv2di3 (t4, t2, t3));
+
+ emit_insn (gen_addv2di3 (operands[0], t1, t4));
+ DONE;
+})
+
(define_insn "ashr<mode>3"
[(set (match_operand:SSEMODE24 0 "register_operand" "=x")
(ashiftrt:SSEMODE24
[(set_attr "type" "sseishft")
(set_attr "mode" "TI")])
+(define_expand "vec_shl_<mode>"
+ [(set (match_operand:SSEMODEI 0 "register_operand" "")
+ (ashift:TI (match_operand:SSEMODEI 1 "register_operand" "")
+ (match_operand:SI 2 "general_operand" "")))]
+ "TARGET_SSE2"
+{
+ if (!const_0_to_255_mul_8_operand (operands[2], SImode))
+ FAIL;
+ operands[0] = gen_lowpart (TImode, operands[0]);
+ operands[1] = gen_lowpart (TImode, operands[1]);
+})
+
(define_insn "sse2_lshrti3"
[(set (match_operand:TI 0 "register_operand" "=x")
(lshiftrt:TI (match_operand:TI 1 "register_operand" "0")
[(set_attr "type" "sseishft")
(set_attr "mode" "TI")])
+(define_expand "vec_shr_<mode>"
+ [(set (match_operand:SSEMODEI 0 "register_operand" "")
+ (lshiftrt:TI (match_operand:SSEMODEI 1 "register_operand" "")
+ (match_operand:SI 2 "general_operand" "")))]
+ "TARGET_SSE2"
+{
+ if (!const_0_to_255_mul_8_operand (operands[2], SImode))
+ FAIL;
+ operands[0] = gen_lowpart (TImode, operands[0]);
+ operands[1] = gen_lowpart (TImode, operands[1]);
+})
+
(define_expand "umaxv16qi3"
[(set (match_operand:V16QI 0 "register_operand" "")
(umax:V16QI (match_operand:V16QI 1 "nonimmediate_operand" "")
[(set_attr "type" "sseiadd")
(set_attr "mode" "TI")])
+(define_expand "umaxv8hi3"
+ [(set (match_operand:V8HI 0 "register_operand" "=x")
+ (us_minus:V8HI (match_operand:V8HI 1 "register_operand" "0")
+ (match_operand:V8HI 2 "nonimmediate_operand" "xm")))
+ (set (match_dup 3)
+ (plus:V8HI (match_dup 0) (match_dup 2)))]
+ "TARGET_SSE2"
+{
+ operands[3] = operands[0];
+ if (rtx_equal_p (operands[0], operands[2]))
+ operands[0] = gen_reg_rtx (V8HImode);
+})
+
+(define_expand "smax<mode>3"
+ [(set (match_operand:SSEMODE14 0 "register_operand" "")
+ (smax:SSEMODE14 (match_operand:SSEMODE14 1 "register_operand" "")
+ (match_operand:SSEMODE14 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx xops[6];
+ bool ok;
+
+ xops[0] = operands[0];
+ xops[1] = operands[1];
+ xops[2] = operands[2];
+ xops[3] = gen_rtx_GT (VOIDmode, operands[1], operands[2]);
+ xops[4] = operands[1];
+ xops[5] = operands[2];
+ ok = ix86_expand_int_vcond (xops);
+ gcc_assert (ok);
+ DONE;
+})
+
+(define_expand "umaxv4si3"
+ [(set (match_operand:V4SI 0 "register_operand" "")
+ (umax:V4SI (match_operand:V4SI 1 "register_operand" "")
+ (match_operand:V4SI 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx xops[6];
+ bool ok;
+
+ xops[0] = operands[0];
+ xops[1] = operands[1];
+ xops[2] = operands[2];
+ xops[3] = gen_rtx_GTU (VOIDmode, operands[1], operands[2]);
+ xops[4] = operands[1];
+ xops[5] = operands[2];
+ ok = ix86_expand_int_vcond (xops);
+ gcc_assert (ok);
+ DONE;
+})
+
(define_expand "uminv16qi3"
[(set (match_operand:V16QI 0 "register_operand" "")
(umin:V16QI (match_operand:V16QI 1 "nonimmediate_operand" "")
[(set_attr "type" "sseiadd")
(set_attr "mode" "TI")])
+(define_expand "smin<mode>3"
+ [(set (match_operand:SSEMODE14 0 "register_operand" "")
+ (smin:SSEMODE14 (match_operand:SSEMODE14 1 "register_operand" "")
+ (match_operand:SSEMODE14 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx xops[6];
+ bool ok;
+
+ xops[0] = operands[0];
+ xops[1] = operands[2];
+ xops[2] = operands[1];
+ xops[3] = gen_rtx_GT (VOIDmode, operands[1], operands[2]);
+ xops[4] = operands[1];
+ xops[5] = operands[2];
+ ok = ix86_expand_int_vcond (xops);
+ gcc_assert (ok);
+ DONE;
+})
+
+(define_expand "umin<mode>3"
+ [(set (match_operand:SSEMODE24 0 "register_operand" "")
+ (umin:SSEMODE24 (match_operand:SSEMODE24 1 "register_operand" "")
+ (match_operand:SSEMODE24 2 "register_operand" "")))]
+ "TARGET_SSE2"
+{
+ rtx xops[6];
+ bool ok;
+
+ xops[0] = operands[0];
+ xops[1] = operands[2];
+ xops[2] = operands[1];
+ xops[3] = gen_rtx_GTU (VOIDmode, operands[1], operands[2]);
+ xops[4] = operands[1];
+ xops[5] = operands[2];
+ ok = ix86_expand_int_vcond (xops);
+ gcc_assert (ok);
+ DONE;
+})
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel integral comparisons
[(set_attr "type" "ssecmp")
(set_attr "mode" "TI")])
+(define_expand "vcond<mode>"
+ [(set (match_operand:SSEMODE124 0 "register_operand" "")
+ (if_then_else:SSEMODE124
+ (match_operator 3 ""
+ [(match_operand:SSEMODE124 4 "nonimmediate_operand" "")
+ (match_operand:SSEMODE124 5 "nonimmediate_operand" "")])
+ (match_operand:SSEMODE124 1 "general_operand" "")
+ (match_operand:SSEMODE124 2 "general_operand" "")))]
+ "TARGET_SSE2"
+{
+ if (ix86_expand_int_vcond (operands))
+ DONE;
+ else
+ FAIL;
+})
+
+(define_expand "vcondu<mode>"
+ [(set (match_operand:SSEMODE124 0 "register_operand" "")
+ (if_then_else:SSEMODE124
+ (match_operator 3 ""
+ [(match_operand:SSEMODE124 4 "nonimmediate_operand" "")
+ (match_operand:SSEMODE124 5 "nonimmediate_operand" "")])
+ (match_operand:SSEMODE124 1 "general_operand" "")
+ (match_operand:SSEMODE124 2 "general_operand" "")))]
+ "TARGET_SSE2"
+{
+ if (ix86_expand_int_vcond (operands))
+ DONE;
+ else
+ FAIL;
+})
+
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parallel integral logical operations
(vec_merge:V4SI
(vec_duplicate:V4SI
(match_operand:SI 2 "nonimmediate_operand" "mr,m,x"))
- (match_operand:V4SI 1 "vector_move_operand" " C,C,0")
+ (match_operand:V4SI 1 "reg_or_0_operand" " C,C,0")
(const_int 1)))]
"TARGET_SSE"
"@
operands[1] = gen_rtx_REG (DImode, REGNO (operands[1]));
})
+(define_insn "*vec_extractv2di_1_sse2"
+ [(set (match_operand:DI 0 "nonimmediate_operand" "=m,x,x")
+ (vec_select:DI
+ (match_operand:V2DI 1 "nonimmediate_operand" "x,0,o")
+ (parallel [(const_int 1)])))]
+ "TARGET_SSE2 && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+ "@
+ movhps\t{%1, %0|%0, %1}
+ psrldq\t{$4, %0|%0, 4}
+ movq\t{%H1, %0|%0, %H1}"
+ [(set_attr "type" "ssemov,sseishft,ssemov")
+ (set_attr "mode" "V2SF,TI,TI")])
+
+;; Not sure this is ever used, but it doesn't hurt to have it. -aoliva
+(define_insn "*vec_extractv2di_1_sse"
+ [(set (match_operand:DI 0 "nonimmediate_operand" "=m,x,x")
+ (vec_select:DI
+ (match_operand:V2DI 1 "nonimmediate_operand" "x,x,o")
+ (parallel [(const_int 1)])))]
+ "!TARGET_SSE2 && TARGET_SSE
+ && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+ "@
+ movhps\t{%1, %0|%0, %1}
+ movhlps\t{%1, %0|%0, %1}
+ movlps\t{%H1, %0|%0, %H1}"
+ [(set_attr "type" "ssemov")
+ (set_attr "mode" "V2SF,V4SF,V2SF")])
+
(define_insn "*vec_dupv4si"
[(set (match_operand:V4SI 0 "register_operand" "=Y,x")
(vec_duplicate:V4SI
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
-;; Miscelaneous
+;; Miscellaneous
;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(match_operand:SI 1 "register_operand" "c")]
UNSPECV_MWAIT)]
"TARGET_SSE3"
- "mwait\t%0, %1"
+;; 64bit version is "mwait %rax,%rcx". But only lower 32bits are used.
+;; Since 32bit register operands are implicitly zero extended to 64bit,
+;; we only need to set up 32bit registers.
+ "mwait"
[(set_attr "length" "3")])
(define_insn "sse3_monitor"
(match_operand:SI 1 "register_operand" "c")
(match_operand:SI 2 "register_operand" "d")]
UNSPECV_MONITOR)]
- "TARGET_SSE3"
+ "TARGET_SSE3 && !TARGET_64BIT"
"monitor\t%0, %1, %2"
[(set_attr "length" "3")])
+
+(define_insn "sse3_monitor64"
+ [(unspec_volatile [(match_operand:DI 0 "register_operand" "a")
+ (match_operand:SI 1 "register_operand" "c")
+ (match_operand:SI 2 "register_operand" "d")]
+ UNSPECV_MONITOR)]
+ "TARGET_SSE3 && TARGET_64BIT"
+;; 64bit version is "monitor %rax,%rcx,%rdx". But only lower 32bits in
+;; RCX and RDX are used. Since 32bit register operands are implicitly
+;; zero extended to 64bit, we only need to set up 32bit registers.
+ "monitor"
+ [(set_attr "length" "3")])