Hello,
This patch extends support of masking instructions.
Bootstrapped.
Is it ok for trunk?
gcc/
* config/i386/i386.md
(define_attr "isa"): Add avx512bw,noavx512bw.
(define_attr "enabled"): Ditto.
(define_split): Add 32/64-bit mask logic.
(define_insn "*k<logic>qi"): New.
(define_insn "*k<logic>hi"): New.
(define_insn "*anddi_1"): Add mask version.
(define_insn "*andsi_1"): Ditto.
(define_insn "*<code><mode>_1"): Ditto.
(define_insn "*<code>hi_1"): Ditto.
(define_insn "kxnor<mode>"): New.
(define_insn "kunpcksi"): New.
(define_insn "kunpckdi"): New.
(define_insn "*one_cmpl<mode>2_1"): Add mask version.
(define_insn "*one_cmplhi2_1"): Ditto.
--
Thanks, K
diff --git a/gcc/config/i386/i386.md b/gcc/config/i386/i386.md
index 4867e7e..7f14ddb 100644
--- a/gcc/config/i386/i386.md
+++ b/gcc/config/i386/i386.md
@@ -740,7 +740,8 @@
;; Used to control the "enabled" attribute on a per-instruction basis.
(define_attr "isa" "base,x64,x64_sse4,x64_sse4_noavx,x64_avx,nox64,
sse2,sse2_noavx,sse3,sse4,sse4_noavx,avx,noavx,
- avx2,noavx2,bmi,bmi2,fma4,fma,avx512f,noavx512f,fma_avx512f"
+ avx2,noavx2,bmi,bmi2,fma4,fma,avx512f,noavx512f,
+ fma_avx512f,avx512bw,noavx512bw"
(const_string "base"))
(define_attr "enabled" ""
@@ -771,6 +772,8 @@
(eq_attr "isa" "noavx512f") (symbol_ref "!TARGET_AVX512F")
(eq_attr "isa" "fma_avx512f")
(symbol_ref "TARGET_FMA || TARGET_AVX512F")
+ (eq_attr "isa" "avx512bw") (symbol_ref "TARGET_AVX512BW")
+ (eq_attr "isa" "noavx512bw") (symbol_ref "!TARGET_AVX512BW")
]
(const_int 1)))
@@ -7510,21 +7513,45 @@
})
(define_split
- [(set (match_operand:SWI12 0 "mask_reg_operand")
- (any_logic:SWI12 (match_operand:SWI12 1 "mask_reg_operand")
- (match_operand:SWI12 2 "mask_reg_operand")))
+ [(set (match_operand:SWI1248x 0 "mask_reg_operand")
+ (any_logic:SWI1248x (match_operand:SWI1248x 1 "mask_reg_operand")
+ (match_operand:SWI1248x 2 "mask_reg_operand")))
(clobber (reg:CC FLAGS_REG))]
- "TARGET_AVX512F && reload_completed"
+;;TODO removed avx512f check because mask_reg implies it.
+ "reload_completed"
[(set (match_dup 0)
- (any_logic:SWI12 (match_dup 1)
- (match_dup 2)))])
+ (any_logic:SWI1248x (match_dup 1)
+ (match_dup 2)))])
-(define_insn "*k<logic><mode>"
- [(set (match_operand:SWI12 0 "mask_reg_operand" "=k")
- (any_logic:SWI12 (match_operand:SWI12 1 "mask_reg_operand" "k")
- (match_operand:SWI12 2 "mask_reg_operand" "k")))]
+(define_insn "*k<logic>qi"
+ [(set (match_operand:QI 0 "mask_reg_operand" "=k")
+ (any_logic:QI (match_operand:QI 1 "mask_reg_operand" "k")
+ (match_operand:QI 2 "mask_reg_operand" "k")))]
+ "TARGET_AVX512F"
+{
+ return TARGET_AVX512DQ ? "k<logic>b\t{%2, %1, %0|%0, %1, %2}"
+ : "k<logic>w\t{%2, %1, %0|%0, %1, %2}";
+}
+ [(set_attr "mode" "QI")
+ (set_attr "type" "msklog")
+ (set_attr "prefix" "vex")])
+
+(define_insn "*k<logic>hi"
+ [(set (match_operand:HI 0 "mask_reg_operand" "=k")
+ (any_logic:HI (match_operand:HI 1 "mask_reg_operand" "k")
+ (match_operand:HI 2 "mask_reg_operand" "k")))]
"TARGET_AVX512F"
"k<logic>w\t{%2, %1, %0|%0, %1, %2}";
+ [(set_attr "mode" "HI")
+ (set_attr "type" "msklog")
+ (set_attr "prefix" "vex")])
+
+(define_insn "*k<logic><mode>"
+ [(set (match_operand:SWI48x 0 "mask_reg_operand" "=k")
+ (any_logic:SWI48x (match_operand:SWI48x 1 "mask_reg_operand" "k")
+ (match_operand:SWI48x 2 "mask_reg_operand" "k")))]
+ "TARGET_AVX512BW"
+ "k<logic><mskmodesuffix>\t{%2, %1, %0|%0, %1, %2}";
[(set_attr "mode" "<MODE>")
(set_attr "type" "msklog")
(set_attr "prefix" "vex")])
@@ -7580,10 +7607,10 @@
})
(define_insn "*anddi_1"
- [(set (match_operand:DI 0 "nonimmediate_operand" "=r,rm,r,r")
+ [(set (match_operand:DI 0 "nonimmediate_operand" "=r,rm,r,r,!k")
(and:DI
- (match_operand:DI 1 "nonimmediate_operand" "%0,0,0,qm")
- (match_operand:DI 2 "x86_64_szext_general_operand" "Z,re,rm,L")))
+ (match_operand:DI 1 "nonimmediate_operand" "%0,0,0,qm,k")
+ (match_operand:DI 2 "x86_64_szext_general_operand" "Z,re,rm,L,k")))
(clobber (reg:CC FLAGS_REG))]
"TARGET_64BIT && ix86_binary_operator_ok (AND, DImode, operands)"
{
@@ -7592,6 +7619,9 @@
case TYPE_IMOVX:
return "#";
+ case TYPE_MSKLOG:
+ return "kandq\t{%2, %1, %0|%0, %1, %2}";
+
default:
gcc_assert (rtx_equal_p (operands[0], operands[1]));
if (get_attr_mode (insn) == MODE_SI)
@@ -7600,8 +7630,8 @@
return "and{q}\t{%2, %0|%0, %2}";
}
}
- [(set_attr "type" "alu,alu,alu,imovx")
- (set_attr "length_immediate" "*,*,*,0")
+ [(set_attr "type" "alu,alu,alu,imovx,msklog")
+ (set_attr "length_immediate" "*,*,*,0,0")
(set (attr "prefix_rex")
(if_then_else
(and (eq_attr "type" "imovx")
@@ -7609,12 +7639,12 @@
(match_operand 1 "ext_QIreg_operand")))
(const_string "1")
(const_string "*")))
- (set_attr "mode" "SI,DI,DI,SI")])
+ (set_attr "mode" "SI,DI,DI,SI,DI")])
(define_insn "*andsi_1"
- [(set (match_operand:SI 0 "nonimmediate_operand" "=rm,r,Ya")
- (and:SI (match_operand:SI 1 "nonimmediate_operand" "%0,0,qm")
- (match_operand:SI 2 "x86_64_general_operand" "re,rm,L")))
+ [(set (match_operand:SI 0 "nonimmediate_operand" "=rm,r,Ya,!k")
+ (and:SI (match_operand:SI 1 "nonimmediate_operand" "%0,0,qm,k")
+ (match_operand:SI 2 "x86_64_general_operand" "re,rm,L,k")))
(clobber (reg:CC FLAGS_REG))]
"ix86_binary_operator_ok (AND, SImode, operands)"
{
@@ -7623,12 +7653,15 @@
case TYPE_IMOVX:
return "#";
+ case TYPE_MSKLOG:
+ return "kandd\t{%2, %1, %0|%0, %1, %2}";
+
default:
gcc_assert (rtx_equal_p (operands[0], operands[1]));
return "and{l}\t{%2, %0|%0, %2}";
}
}
- [(set_attr "type" "alu,alu,imovx")
+ [(set_attr "type" "alu,alu,imovx,msklog")
(set (attr "prefix_rex")
(if_then_else
(and (eq_attr "type" "imovx")
@@ -7636,7 +7669,7 @@
(match_operand 1 "ext_QIreg_operand")))
(const_string "1")
(const_string "*")))
- (set_attr "length_immediate" "*,*,0")
+ (set_attr "length_immediate" "*,*,0,0")
(set_attr "mode" "SI")])
;; See comment for addsi_1_zext why we do use nonimmediate_operand
@@ -7688,11 +7721,21 @@
(match_operand:QI 2 "general_operand" "qn,qmn,rn,k")))
(clobber (reg:CC FLAGS_REG))]
"ix86_binary_operator_ok (AND, QImode, operands)"
- "@
- and{b}\t{%2, %0|%0, %2}
- and{b}\t{%2, %0|%0, %2}
- and{l}\t{%k2, %k0|%k0, %k2}
- kandw\t{%2, %1, %0|%0, %1, %2}"
+{
+ switch (which_alternative)
+ {
+ case 0:
+ case 1:
+ return "and{b}\t{%2, %0|%0, %2}";
+ case 2:
+ return "and{l}\t{%k2, %k0|%k0, %k2}";
+ case 3:
+ return TARGET_AVX512DQ ? "kandb\t{%2, %1, %0|%0, %1, %2}"
+ : "kandw\t{%2, %1, %0|%0, %1, %2}";
+ default:
+ gcc_unreachable ();
+ }
+}
[(set_attr "type" "alu,alu,alu,msklog")
(set_attr "mode" "QI,QI,SI,HI")])
@@ -7715,10 +7758,22 @@
(match_operand:SWI12 2 "register_operand" "r,r,k")))
(clobber (reg:CC FLAGS_REG))]
"TARGET_AVX512F"
- "@
- andn\t{%k2, %k1, %k0|%k0, %k1, %k2}
- #
- kandnw\t{%2, %1, %0|%0, %1, %2}"
+{
+ switch (which_alternative)
+ {
+ case 0:
+ return "andn\t{%k2, %k1, %k0|%k0, %k1, %k2}";
+ case 1:
+ return "#";
+ case 2:
+ if (TARGET_AVX512DQ && <MODE>mode == QImode)
+ return "kandnb\t{%2, %1, %0|%0, %1, %2}";
+ else
+ return "kandnw\t{%2, %1, %0|%0, %1, %2}";
+ default:
+ gcc_unreachable ();
+ }
+}
[(set_attr "isa" "bmi,*,avx512f")
(set_attr "type" "bitmanip,*,msklog")
(set_attr "prefix" "*,*,vex")
@@ -8082,14 +8137,17 @@
"ix86_expand_binary_operator (<CODE>, <MODE>mode, operands); DONE;")
(define_insn "*<code><mode>_1"
- [(set (match_operand:SWI48 0 "nonimmediate_operand" "=r,rm")
+ [(set (match_operand:SWI48 0 "nonimmediate_operand" "=r,rm,k")
(any_or:SWI48
- (match_operand:SWI48 1 "nonimmediate_operand" "%0,0")
- (match_operand:SWI48 2 "<general_operand>" "<g>,r<i>")))
+ (match_operand:SWI48 1 "nonimmediate_operand" "%0,0,k")
+ (match_operand:SWI48 2 "<general_operand>" "<g>,r<i>,k")))
(clobber (reg:CC FLAGS_REG))]
"ix86_binary_operator_ok (<CODE>, <MODE>mode, operands)"
- "<logic>{<imodesuffix>}\t{%2, %0|%0, %2}"
- [(set_attr "type" "alu")
+ "@
+ <logic>{<imodesuffix>}\t{%2, %0|%0, %2}
+ <logic>{<imodesuffix>}\t{%2, %0|%0, %2}
+ k<logic><mskmodesuffix>\t{%2, %1, %0|%0, %1, %2}"
+ [(set_attr "type" "alu,alu,msklog")
(set_attr "mode" "<MODE>")])
(define_insn "*<code>hi_1"
@@ -8177,19 +8235,36 @@
(match_operand:SWI12 2 "register_operand" "r,k"))))
(clobber (reg:CC FLAGS_REG))]
"TARGET_AVX512F"
+{
+ if (which_alternative == 1 && <MODE>mode == QImode && TARGET_AVX512DQ)
+ return "kxnorb\t{%2, %1, %0|%0, %1, %2}";
+ return "kxnorw\t{%2, %1, %0|%0, %1, %2}";
+}
+ [(set_attr "type" "*,msklog")
+ (set_attr "prefix" "*,vex")
+ (set_attr "mode" "<MODE>")])
+
+(define_insn "kxnor<mode>"
+ [(set (match_operand:SWI48x 0 "register_operand" "=r,!k")
+ (not:SWI48x
+ (xor:SWI48x
+ (match_operand:SWI48x 1 "register_operand" "0,k")
+ (match_operand:SWI48x 2 "register_operand" "r,k"))))
+ (clobber (reg:CC FLAGS_REG))]
+ "TARGET_AVX512BW"
"@
#
- kxnorw\t{%2, %1, %0|%0, %1, %2}"
+ kxnor<mskmodesuffix>\t{%2, %1, %0|%0, %1, %2}"
[(set_attr "type" "*,msklog")
(set_attr "prefix" "*,vex")
(set_attr "mode" "<MODE>")])
(define_split
- [(set (match_operand:SWI12 0 "general_reg_operand")
- (not:SWI12
- (xor:SWI12
+ [(set (match_operand:SWI1248x 0 "general_reg_operand")
+ (not:SWI1248x
+ (xor:SWI1248x
(match_dup 0)
- (match_operand:SWI12 1 "general_reg_operand"))))
+ (match_operand:SWI1248x 1 "general_reg_operand"))))
(clobber (reg:CC FLAGS_REG))]
"TARGET_AVX512F && reload_completed"
[(parallel [(set (match_dup 0)
@@ -8199,6 +8274,8 @@
(set (match_dup 0)
(not:HI (match_dup 0)))])
+;;There are kortrest[bdq] but no intrinsics for them.
+;;We probably don't need to implement them.
(define_insn "kortestzhi"
[(set (reg:CCZ FLAGS_REG)
(compare:CCZ
@@ -8238,6 +8315,28 @@
(set_attr "type" "msklog")
(set_attr "prefix" "vex")])
+(define_insn "kunpcksi"
+ [(set (match_operand:SI 0 "register_operand" "=k")
+ (ior:SI
+ (ashift:SI
+ (match_operand:SI 1 "register_operand" "k")
+ (const_int 16))
+ (zero_extend:SI (subreg:HI (match_operand:SI 2 "register_operand"
"k") 0))))]
+ "TARGET_AVX512BW"
+ "kunpckwd\t{%2, %1, %0|%0, %1, %2}"
+ [(set_attr "mode" "SI")])
+
+(define_insn "kunpckdi"
+ [(set (match_operand:DI 0 "register_operand" "=k")
+ (ior:DI
+ (ashift:DI
+ (match_operand:DI 1 "register_operand" "k")
+ (const_int 32))
+ (zero_extend:DI (subreg:SI (match_operand:DI 2 "register_operand"
"k") 0))))]
+ "TARGET_AVX512BW"
+ "kunpckdq\t{%2, %1, %0|%0, %1, %2}"
+ [(set_attr "mode" "DI")])
+
;; See comment for addsi_1_zext why we do use nonimmediate_operand
;; ??? Special case for immediate operand is missing - it is tricky.
(define_insn "*<code>si_2_zext"
@@ -8837,11 +8936,15 @@
"ix86_expand_unary_operator (NOT, <MODE>mode, operands); DONE;")
(define_insn "*one_cmpl<mode>2_1"
- [(set (match_operand:SWI48 0 "nonimmediate_operand" "=rm")
- (not:SWI48 (match_operand:SWI48 1 "nonimmediate_operand" "0")))]
+ [(set (match_operand:SWI48 0 "nonimmediate_operand" "=rm,k")
+ (not:SWI48 (match_operand:SWI48 1 "nonimmediate_operand" "0,k")))]
"ix86_unary_operator_ok (NOT, <MODE>mode, operands)"
- "not{<imodesuffix>}\t%0"
- [(set_attr "type" "negnot")
+ "@
+ not{<imodesuffix>}\t%0
+ knot<mskmodesuffix>\t{%1, %0|%0, %1}"
+ [(set_attr "isa" "*,avx512bw")
+ (set_attr "type" "negnot,msklog")
+ (set_attr "prefix" "*,vex")
(set_attr "mode" "<MODE>")])
(define_insn "*one_cmplhi2_1"
@@ -8861,10 +8964,21 @@
[(set (match_operand:QI 0 "nonimmediate_operand" "=qm,r,!k")
(not:QI (match_operand:QI 1 "nonimmediate_operand" "0,0,k")))]
"ix86_unary_operator_ok (NOT, QImode, operands)"
- "@
- not{b}\t%0
- not{l}\t%k0
- knotw\t{%1, %0|%0, %1}"
+{
+ switch (which_alternative)
+ {
+ case 0:
+ return "not{b}\t%0";
+ case 1:
+ return "not{l}\t%k0";
+ case 2:
+ if (TARGET_AVX512DQ)
+ return "knotb\t{%1, %0|%0, %1}";
+ return "knotw\t{%1, %0|%0, %1}";
+ default:
+ gcc_unreachable ();
+ }
+}
[(set_attr "isa" "*,*,avx512f")
(set_attr "type" "negnot,negnot,msklog")
(set_attr "prefix" "*,*,vex")
